The Homeric Hymn to Hermes 110701204X, 9781107012042 [PDF]

Zitiervorschau

CAMBRIDGE CLASSICAL TEXTS AND C O M M E N TA R I E S

editors J. D I G G L E N. H O P K I N S O N S . P. OA K L E Y J. G. F. P O W E L L M . D. R E E V E D. N. S E D L E Y R . J. TA R R A N T

62 T H E H O M E R I C H Y M N TO H E R M E S

THE HOMERIC HYMN TO HERMES E D I T E D W I T H I N T RO D U C T I O N, T R A N S L AT I O N, A N D C O M M E N TA RY BY

OLIVER THOMAS University of Nottingham

University Printing House, Cambridge cb2 8bs, United Kingdom One Liberty Plaza, 20th Floor, New York, ny 10006, USA 477 Williamstown Road, Port Melbourne, vic 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107012042 doi: 10.1017/9780511997792 © Oliver homas 2020 his publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2020 Printed in the United Kingdom by TJ International Ltd, Padstow Cornwall A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data names: homas, Oliver R. H. (Oliver Robert Havelock), 1982– editor. title: he Homeric hymn to Hermes / edited by Oliver homas. other titles: Hymn to Hermes. 2020. | Cambridge classical texts and commentaries. description: New York : Cambridge University Press, 2020. | series: Cambridge classical texts and commentaries | Includes bibliographical references and index. identifiers: lccn 2019059897 | isbn 9781107012042 (hardback) | isbn 9780511997792 (epub) subjects: lcsh: Hymn to Hermes. classification: lcc pa4025.h85 h66 2020 | ddc 883/.01–dc23 LC record available at https://lccn.loc.gov/2019059897 isbn 978-1-107-01204-2 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

CONTENTS Acknowledgements Notes to this Edition

page vi viii

INTRODUCTION

1

1 Dating the Hymn to Hermes

1

2 Locating the Hymn to Hermes

23

3 Generating charis

36

4 Inluence on Later Texts

63

5 The Manuscript Tradition and This Edition

73

SIGLA

90

TEXT AND TRANSLATION

94

COMMENTARY

135

Bibliography and Abbreviations Index of passages Index of Greek Words Index of Subjects

471 521 524 525

v

ACKNOWLEDGEMENTS This whole book is but a draught – nay, but the draught of a draught. Oh, Time, Strength, Cash, and Patience! Melville, Moby Dick ch. 32

I began work on the Hymn to Hermes back in 2005. Over the years, I have fretted increasingly about time and patience, but not so much about strength and cash, thanks to numerous people and organizations whose assistance it is my pleasure to acknowledge. Without the inancial support and library resources of various institutions, this book would not exist. A studentship from the Classics Faculty at the University of Oxford allowed me to start my postgraduate studies on the Homeric Hymns, and the Arts and Humanities Research Council funded my doctorate. Balliol College provided a conducive environment for it, after which Christ Church aforded me three further years of research time by appointing me to the Christopher Tower Junior Research Fellowship in Greek Mythology. The University of Cambridge allowed me to learn the ropes of lecturing in a tremendously stimulating environment; to my delight, St John’s College housed me in the very rooms where E. E. Sikes had worked on the Homeric Hymns some 110 years before. Most recently, the University of Nottingham has given me the stability of a long-term post, and a wonderful set of colleagues. I can only pick out a small number of the many friends and colleagues whose advice has improved this book. As my doctoral supervisor, Armand D’Angour provided insightful steers while giving me considerable free rein to explore and develop my interests. My examiners, Richard Hunter and Gregory Hutchinson, have both given me characteristically wide-ranging guidance. Many others, including Adrian Kelly, Bruno Currie, Martin West, Andrew Faulkner and Nicholas vi

ACKNOWLEDGEMENTS

Richardson, ofered encouragement and advice, the last with the outstanding generosity and patience which are familiar to many scholars of the Homeric Hymns. Regular evening discussions with Tom Phillips helped me reine central points in my interpretation of the hymn. He and Lyndsay Coo both commented on draft sections of my commentary; Emily Kneebone commented on parts of the Introduction. Rebecca Lämmle, Luuk Huitink and Steven Kaye have all ielded numerous queries, while Kathryn Murphy taught me more about close reading than anyone. Most recently, I received extremely helpful comments on my draft typescript from Michael Reeve, James Diggle and Neil Hopkinson, who saved me from more errors than I would like to admit. The production team at Cambridge University Press has been exemplary, and I was fortunate to receive help with indexing from someone as sharp-eyed as Thomas Sims. Naturally, I could never have started such a project without inspirational teaching at earlier stages of my studies. I would like to register special thanks in this regard to Ann Higham, Stephen Anderson, John Falconer, Caroline Butler, David Raeburn, Robin Lane Fox and Peter Wilson, and to Jane Lightfoot, who irst pointed me to the Homeric Hymns. Maroula Perisanidi introduced me to the expression τον φάγαμε τον γάιδαρο, η ουρά μας έμεινε, and sharing the last four years with her has indeed made the tail-end of this project ininitely more pleasant. Finally, I dedicate the book to my parents, Alan and Fran, whose constant and generous inluence is at the root of every part of my life.

vii

N O T E S TO T H I S E D I T I O N This book is directed at researchers and advanced students who seek information about a point of detail in the Homeric Hymn to Hermes or an unfolding interpretation of the whole composition. Since I started my research, readers have beneited from a new student-level commentary on the hymn (Richardson 2010), two more detailed commentaries (Vergados 2013, Schenck zu Schweinsberg 2017), and two valuable book-length discussions (Jaillard 2007, Nobili 2011).1 These works are signs of the poem’s interest, but they also raise the question of what I hope to contribute, especially in comparison to Vergados’ commentary, which is on a similar scale. Naturally there are occasions where Vergados and I independently found the same parallels or thought the same scholarship worth citing. But comparison will reveal diferences of opinion or emphasis in the majority of lines, including on matters of dating, original performance context, and the themes picked out in our introductory essays. I have more to say about the last third of the poem, and have taken a diferent editorial approach by re-examining the manuscripts and being substantially more critical of transmitted readings. While we are both interested in lexicography and the formulaic system, I have replicated less information from dictionaries, and have given more space to interpretation, using diferent methodologies including (for example) politeness theory and psycholinguistics. I have tried to avoid pointing out every place where our work diverges, while still alerting readers to enough points of disagreement to stimulate future debate. At this juncture, I should give readers an essential piece of advice about navigating the commentary. I ind it more 1

Hereafter I will abbreviate the hymn as Herm., and refer to editions and commentaries by author’s name only, where this is possible without confusion; these can be found asterisked in the bibliography. I reviewed Schenck zu Schweinsberg in Thomas (2019).

viii

N O T E S TO T H I S E D I T I O N

consistent and intuitive to draw together the threads of particular passages after discussing the details, rather than before, as is the more common practice. My summative comment on (e.g.) lines 1–9 therefore comes after the comments on line 9, rather than before those on line 1 – a fact which I have indicated regularly throughout the commentary. This ordering gains particular importance in discussion the hymn’s riddles, whose efect is ruined by a pre-emptive summary; but more generally, it suits the fact that the way meaning is clariied piecemeal is integral to the experience of an audience receiving a piece of literature for the irst time.2 As usual, summative comments are not mere summaries, and readers should consult the one which covers the line they are interested in as well as comments on the line itself. Five preliminary points about notation should be made: 1. I follow the abbreviations of ancient authors listed in LSJ, OLD and the Oxford Classical Dictionary (3rd edn), with minor exceptions which are unlikely to cause confusion (such as ‘H.Dem.’ instead of ‘H.Cer.’). ‘Σ’ means ‘scholion to’. ‘Epos’ is a short-hand for Il., Od., the Hesiodic corpus, H.Hom., and the Cycle, excluding later Homerica such as the Certamen and early philosophical hexameter such as Parmenides. ‘Homeric’ is used as a short-hand for ‘in the Iliad and/or Odyssey’, without implying a stance on the authorship of those poems. For other bibliographical abbreviations see p. 495. 2. Statistical information is based on the most recent editions currently used by the Thesaurus Linguae Graecae (www.tlg.uci. edu/), unless otherwise stated. For the editions used for texts (esp. fragmentary) with variable numeration see the list at the start of the Bibliography. Citations generally relect my beliefs about ancient texts rather than adhering uniformly

2

See Most (1985) 38, in connection with commentary-writing. My ordering is of course nothing new: it is also to be found, for example, in Fraenkel (1950).

ix

N O T E S TO T H I S E D I T I O N

to particular editions, though for consistency I have used iota adscript even in late texts. 3. Except in bibliographical references, dates are bce unless stated. 4. In describing formulas, | marks the line-end, ¦ the main caesura. + indicates points of lexibility, e.g. τὸν+ δ’ ἀπαμειβόμενος acknowledges that τὴν δ’ ἀπαμειβόμενος is also found. ‘C’ and ‘V’ are sometimes used for ‘a consonant’ and ‘a vowel’ in describing word-shapes. 5. I use asterisks throughout the book to distinguish references to a passage in Herm. from references to my commentary. Thus ‘134–5*’ means ‘see lines 134–5 and my comment(s) on them’.

x

I N T RO D U C T I O N 1 DAT I N G T H E H Y M N TO H E R M E S The Hymn to Hermes was composed by an unknown poet whose virtuoso deployment of formulaic traditions shows that he was a rhapsode – one of the highly trained and emotive live performers of epic (both new and, increasingly, canonical) in ancient Greece.1 He may well have performed his composition multiple times in different forms. However, as it is transmitted (see below, §5.1), we have evidence only for minor, localized rephrasing, and not for major recasting of the story. Hence we can consider our text to derive in its essence from one act of transcription, and to analyse when and for what circumstances the transcribed poem might have been designed. I examine the date in this section, and the circumstances in the next. 1.1 Linguistic Arguments The epic Kunstsprache evolved over many centuries. As spoken varieties of Greek changed around it, many formulas were retained despite their archaism, some were updated where possible, and new formulas were added to the traditional repertoire. Newer and older linguistic features therefore rub shoulders, and one expects the proportion of newer features broadly 1

For their activities see e.g. West (2010). The poet’s understanding and manipulation of formulaic tradition is discussed throughout the commentary; see Index s.v. ‘Formulas’. Van Nortwick (1975: 15, 24–6) established the largely traditional nature of Herm.’s noun–epithet formulas, though those for Hermes himself go beyond epic norms in their breaches of economy (Sbardella 1993: 39–41). While the proportion of Herm. taken up by known formulas and their modifications is lower than for other texts (Postle­thwaite 1979, Cantilena 1982, and already Sterrett 1881), this may have as much to do with the hymn’s unusual tone as with chronology.

1

I N T RO D U C T I O N

to increase over time. However, the rate of development is difficult to model, and several further variables may affect the distribution of features in a given poem, such as the author’s stylistic preferences (including openness to a range of registers), training and locale, and the poem’s content (especially if it means that certain morphologically interesting formulas recur). In Herm., two features which are unlikely to have been consciously manipulated show the Kunstsprache at an advanced state. One has been discussed many times, namely the very low proportion of syllables which had once contained a digamma and which scan as if it were still present.2 The other feature, which has largely dropped out of discussion, is the r­ emarkable degree to which ὁ has progressed from being mainly an anaphoric demonstrative towards articular uses: by my count, there are 31 articular versus 54 demonstrative uses, plus 5 as a relative pronoun.3 Such apparently ‘late’ features are compatible with the poet having had a deliberate stylistic preference for others which are ‘early’, such as the 2nd-declension genitive singular in ‑οιο (also affected by Apollonius) and the longer forms of 1st- and 2nd-declension dative plurals.4 A range of other stylo See Janko (1982) 42–8 with statistics, and already Windisch (1867) 38–41. ‘Articular’ and ‘demonstrative’ are usable labels, even if they involve some simplification of a spectrum of usage, such as an anaphoric demonstrative with noun reprised a few words later (365*), or the various shades of definiteness conferred where the article substantivizes an adverb, particularizes where an attributive adjective is added, etc.; cf. Ambrosini (1986). Stummer (1886: 59–60) gave very similar overall figures for Herm., which suggests that it is meaningful to compare my figures with the ones he gave for other texts: these involve the proportion (articular/demonstrative) being, in ascending order, 218/c.3000 in Il., 171/2178 in Od., 3/37 in H.Aphr., 9/71 in H Dem., 62/404 in Hesiod (unfortunately Stummer counted Th., Op.  and  Sc.  to­ gether), 14/55 in H.Ap., and finally Herm. as above. Content skews the number of uses such as τὸ πρῶτον and τὸ πάροιθεν in Hesiod, but otherwise the figures corroborate other evidence about relative chronology. 4 For statistics on these see Janko (1982), including pp. 133–40 for a study of other potentially datable features of Herm.’s language (in favour of a late sixth-century date). Herm. is the text for which Janko’s morphological data 2 3

2

1 . 1 DAT I N G T H E H Y M N TO H E R M E S

metric eccentricities, in particular Herm.’s tendencies towards metrical purism, do not seem to me indicative for chronology, interesting though they are for the composer’s technique.5 The semantic evolution of ὁ can be analysed statistically. For less common words, similar semantic or even lexical innovations must be approached differently. The obvious difficulty is that our very patchy evidence for archaic Greek prevents us from estimating how long an apparently ‘new’ word or meaning existed before its first extant occurrence.6 Even if such evidence cannot, therefore, be decisive for chronology, it is suggestive to gather cases involving some of the following conditions: (a) there is no indirect evidence of the pre-existence of the form; (b) the form is used frequently once it is recorded; (c) the meaning is such that one might have expected the form to occur in earlier texts if it had existed; (d) we appear to be able to track the word’s semantic development step by step over time; (e) the form belongs to a register (such as that of interstate treat­ ies) where we have some evidence about the circumstances in which it could have developed. The following table collects such forms where the earliest attestation outside Herm. dates to the sixth century or later, together presented the least clear indications; see also Jones (2010) 94–5. Jones’ article critiques Janko’s attempts to take his data beyond claims about relative chronology; Janko (2012) is a response to some details. 5 For Herm.’s avoidance of hiatus, fondness for dactyls, respect for third-foot caesuras (with H.Aphr.) and for Meyer’s first ‘law’ (with H Dem.), and several other metrical features, see Vergados 57–62. Janko (1982: 36–41) took this as a sign of literate composition. I assume that a skilful oral poet could hone his work to suit his personal aural tastes. 6 Three cautionary tales: 273 χθές is avoided in Il. and Od. but presupposed by χθιζός, which they do use; 512 ἀκουστός is two steps logically prior to νηκουστέω (Il. 21.14); 187 ἁγνός was inherited from Indo-European, but avoided in the Iliad.

3

I N T RO D U C T I O N

with a necessarily impressionistic count of the number of attestations to around 360 bce, and linguistic notes where necessary.7 254 μηνυ332 σπουδαῖος 53 πλῆκτρον 111 ἀναδίδωμι 208 ἔδοξα 255 διαφέρομαι 92 κωφός ‘deaf ’ 460 ἀκόντιον

129 τέλεος

Stesichorus fr. 42(?), Hipponax fr. 102, then 200+ uses by c.360 bce, c.35 in verse. Indo-European heritage is not secure (Beekes 2010 s.v.). Theognis 65 then 100+ uses, c.11 verse. H.Ap. 185 then c.15 uses. Asius fr. 8 then c.30 uses, c.7 verse. IG I3 1 (c.505 bce), then very frequent; replaced ἐδόκησα (Od.; rare from 5th c.) Heraclitus 22 B10, then c.25 prose uses; S. Aj. 511 for verse. Heraclitus 22 B34, then c.20 uses; development from ‘dull, blunt’ (Il. x3) to ‘insensitive’ (first Anacr. fr. eleg. 4) to ‘deaf, mute’. Acusilaus FGrH 2 F22(?), then c.75 uses; prose except Aristophanes frr. 492–3. Earlier epic has ἄκων (x23); the diminutive suffix -ιον first appears in the sixth century (Schwyzer i.480). IG IX.12 iii.718.47 (East Locris, before 475?), then 100+ uses. Vernacular Attic–Ionic for epic τέλειος < *τελεσ-ιος. Few σ-stems form epic adjectives in -εος (Risch 1974: 133).

The table relies heavily on TLG, Hansen (1983), and the PHI epigraphic database (http://epigraphy.packhum.org/). Since it can at best give a persuasive impression, not a conclusive demonstration, an elaborate description and justification of my counting procedure (such as including Theognidea and excluding Aesopica) would not serve our purpose. For possible syntactic additions to the list of datable features see 208*, 378–9*.   Listing ‘innovative’ forms has long been popular, but often without considering the conditions which give an innovation a chance of being dat­ able. See e.g. Voss (1794) i.102–11; Fietkau (1863); Greve (1867) 26–34, 57–90; Eberhard (1874: 7–36) conflated ‘new’ and ‘un-Homeric’. Zumbach (1955), though rightly censured by Hoekstra (1969: 7–20), did draw attention to the hymnist’s love of agent nouns, nouns in -ίη (see table), and adjectives compounded from preposition + noun. 7

4

1 . 1 DAT I N G T H E H Y M N TO H E R M E S

208 σαφές

Pindar P. 2.25, then frequent. σαφής (first extant in Simonides) < σαφέως (first H.Dem.) HJK) have been taken to share a parent f against x (e.g. Càssola 598–9). This is not required, and takes us further away from the manuscripts’ relationships in other texts. The conjunctions of Dz against xp at H.Ap. 49 ἐβήσατο (for ἐβήσετο) and 166 ἐμεῖο (for ἐμοῖο) may well be stemmatically insignificant attempts to correct Θ, while their transposition of H.Ap. 36–40 to follow 41 could have arisen if the scribe of Θ omitted 36–40 by eyeskip, and added them in the margin with an ambiguous mark next to 41. At Hy. 16.3 φλεγύος Dz (φλεγύου xp), Θ might for example have had the unusual name flagged in the margin (as in KΓ) in a way that looked like a variant. 133 For D see Breuning (1929) 117–23, Pfeiffer (1953) lvi, Vian (1974) xlviii, Schade and Eleuteri (2008) 43–5. Wilson (1974) clarified the position of At; see also Bulloch (1985) 63–6. Facsimiles of D and Q are available from http://ambrosiana.comperio.it/biblioteca-digitale/. 132

76

5.1 THE MS TRADITION AND THIS EDITION

either to Θ after D’s branch, or to x) at Herm. 42~79.134 The first pair descend from a book (a) which Konstantinos Laskaris found in poor condition in Milan in 1464. He copied it that year in T (signed and dated on f. 100v), and shared both a and T with several students, who like their teacher showed particular interest in the Orphic Argonautica. One of them, the teenage Giorgio Valla, copied the Homeric Hymns in E, before Laskaris’ departure for Naples in 1465. Valla’s age explains why E is generally less accurate than T. a had generally incorporated the smaller variants of x, and has several careless omissions (H Ap. 261–89, 506–8, Herm. 535, H.Aphr. 97, Hy. 20.8) and errors.135 The other copy of x (b) was much more faithful, preserved more of the variants, and included a series of relatively sophisticated linguistic marginalia. However, both extant copies of it (L and Π) introduce numerous mistakes of their own, including words which are not morphologically viable. Π’s priority is shown by a series of small omissions in L (some marked by gaps, others not), some of which were not fixable by emendation, especially Herm. 183 om. Διὸς ~ 207 om. γουνὸν (on recto and verso of b). The scribe of Π collaborated with Demetrios Raoul Kavakes on Monacensis gr. 71, and the two books share a watermark known in Rome in 1479 (Thomas 2016: 280). A terminus ante quem for L is its use by Chalkondyles in 1488, and it may well have been used for L3 a few years before (see below); Ianos Laskaris seems to have removed the gatherings following Hy. 7.33 for his edition of Callimachus’ Hymns in 1496.136 Where variants in x are found in M, they probably go back to Ω: e.g. Herm. 212, 224, 451, H.Aphr. 214, Hy. 7.37. This could be true of all the variants, or Ψ may have picked up new ones. 135 For T see Andrés (1987) 35–8, Vian (1979) 23–6 (siglum M, including Laskaris’ comments about finding a). For E see Puntoni (1965), and Thomas (2016) 279–80 on the date. 136 A facsimile of L with bibliography is available from http://mss.bmlonline. it/. A facsimile and detailed description of Π is available from http://gallica.bnf.fr/. For Maturanzio’s acquisition of Π see Hoffmann (1983) 138–9, Zappacosta (1970) 21–2. 134

77

I N T RO D U C T I O N

The last family is headed by p, whose scribe made a series of bold conjectures in the texts of the Orphic Argonautica, Orphic Hymns and Homeric Hymns.137 The descendants are less clearly characterized. Càssola’s focus on four representatives (P, V and, in my sigla, β1, q) was an adequate means to reconstruct p’s interesting readings. However, my study of this family in order to understand the reception of the Hymns in Renaissance Florence uncovered a different model, according to which p can be reconstructed simply from P and q. The manuscript with the most telling readings and the most influence is P, owned by Giannozzo Manetti (signed on f. Iv), who died in 1459. The hand has not previously been identified but in my view is that of Ioannes Eugenikos, the scribe of M. P has few indicative errors which would sustain a prima facie case for relating later manuscripts to it, such as H Ap. 51 ἐμεῖο for ἐμοῖο, 62 λοιτοῖ for λητοῖ, Herm. 451 μέλυσι (indistinctly corrected to μέλουσι). However, over time various passages of P were annotated, and the influence of these annotations is a further indicator that a later manuscript derives from P; we can in fact trace how these annotations filtered through the family. I discuss the source of the annotations at the end of the section.138 O alone shares with P’s uncorrected phase the postponement of p’s γύων from the end of Herm. 20 to the start of 21. P’s priority is evident at H Ap. 232, which O skipped. O shows signs of some interlinear variants which were added to P by a See Vian (1979), siglum ζ; Quandt (1955), siglum π. Keydell (1942) 74 refuted Quandt’s original idea that p derives from Ψ indirectly. Good corrections in p include Herm. 356 κατέερξε, 547 ἐθέλησι, H.Ap. 306 τυφάονα for τυφλόν. An intelligent but less successful attempt can be seen at Herm. 214 φηλωτήν (cf. the connection of φηλήτης to φηλόω in Ps.-Zon. φ.1804), and cf. Commentary n. 80 on H.Ap. 515. H.Ap. 78 ἕκαστά τε φῦλα νεπούδων (for ἀκηδέα χήτεϊ λαῶν) probably shows a scribe willing to compose some verse when his exemplar failed him. 138 A microfilm of P is available from http://digi.vatlib.it/. Càssola 604–6 also analysed its variants, though I differ in several details.

137

78

5.1 THE MS TRADITION AND THIS EDITION

later hand: H.Ap. 293 βωμῶ ss. νιῶ (ss. νηῶ P2), O.Arg. 1018 ss. ἀκούσας, 1168 κόλπων ss. ο.139 A later copy of P is L2, written by Ioannes Scutariotes probably c.1460.140 Scutariotes shows knowledge of two further annotations in P, at Herm. 21, where the erroneous γύων οὐκ ἔτι was erased and replaced by οὐκέτι ---, and at 371 νέον, after which a reader added interlinear γ’ to try to smooth out the metre; O is lacking and so may already have contained the latter. L2 still shares P’s errors at 114 τ λόσε and Hy. 6.2 om. ἄσομαι. A series of new errors can be traced directly to P’s handwriting, e.g. H.Ap. 108 μεσήγας and 428 ἄιπας (misunderstanding J = -ύ as the ligature for -ας), 226 θύβη, Hy. 7.5 θάρος (P has φάρος with a τ visible under φ). Scutariotes copied L2 twice soon afterwards. In R2 (signed on f. 73) he introduced a large number of new small errors, but also found some simple corrections.141 L4, which was owned in Florence by Giorgio Antonio Vespucci, and so probably predates Scutariotes’ departure for Pisa in 1473, is marked by different errors. In four clear cases Scutariotes sought information from a source other than L2, and the only one which suits all four is the annotation in P: the restoration of γύων at the end of Herm. 20, 114 τιλόσε, the

At H.Ap. 202 ἀμφι φαείνειὴ O, ἀμφιφαείνει ss. ὴ P, it is hard to tell whether the interlinear variant in P was written by a later hand. 140 A facsimile is available at http://mss.bmlonline.it/. Eleuteri (1981) 156 based his dating on the development of Scutariotes’ handwriting. Scutariotes had been Giannozzo Manetti’s preferred copyist in the 1440s and was the Greek tutor of his son Agnolo. Giannozzo’s books went with him to Naples in 1453; Agnolo brought them back to Florence at some point in 1459–63 (Den Haan 2019). 141 Conjunctions of L2R2, besides the unusual readings of L2 already mentioned: H.Ap. 128 ἀσπέροντα, 210 ἐλαντινιονίδη, Herm. 254 λύκνω, 273 ἁπαλλοὶ, 375 οὔνεχ’, 424 ἐπαριστερὰ, 496 ὄρεξ’, 504 τέκνας, 520 ἔρδοις, H.Aphr. 56 φιλομηδεὶς, 152 προοίοι. New errors include 49 Herm. ἀμὶ, 50 ἥραρεν, 86 οἶα, 95 ὅρη, 113 φλὼξ, 168 ἅπλαστοι, 194 κῦνες, 311 τόδε, 406 ὦδε, etc. Attempted corrections include 25 τεκτηνατ’, 67 φιληταὶ, 410 ἁγνοῦ ταὶ, 465 ἡμετέρης. Hollander (1886) 11 already proposed P > L2 > R2. 139

79

I N T RO D U C T I O N

marginal restoration of Hy. 6.2 ἄσομαι, and 6.18 ἐυστεφάνου ss. ἰο- (from the Θ family).142 Càssola argued that N used both P and V, the latter being an independent child of p. However the independence of V – a composite volume (a Homeric corpus incorporating Quintus of Smyrna) made for Cardinal Bessarion, most likely c.1467 – was based on a few passages which could equally be explained as quite simple corrections made between V and P. Càssola noted (p. 606) but did not solve the awkwardness for his supposition that at H.Ap. 293 P2 and V would have the variant νηῶ independently. V and N do have conjunctive errors and corrections; but V also commits a large number of minor errors based on spelling and phonetics, while N commits a few distinctive phonetic errors despite other signs of the scribe’s carefulness.143 New minor errors in L4 include Herm. 1 ὑμνοῦσα, 2 δέοντα, 121 ἀμφοβελοῖσι, 308 ἔνεχ’, 368 ἥτοι, 383 ἐπιδέομαι. At 78 ὄπιθεν L2 vs ὄπισθεν PR2L4, Scutariotes twice overlooked the correction he had made in L2: this is explicable given the phrasing πρόσθεν ὄπισθεν, | τὰς δ’ ὄπι(σ)θεν πρόσθεν.   The relationships outlined above accord with the evidence of O.Arg.: see Vian (1979) 17–20, where QqDd correspond respectively to POL2L4. Vian also cites evidence for P’s influence over N R1 Harv (his sigla okl), though he was wary about taking P as the head of the family. Donadi (2016) argues that L2 was a copy of P for Gorgias’ Encomium. The evidence from Musaeus is compatible: Eleuteri (1981). At least the titles of Orph.H. in L2 (and Mon N) had a different source, since they were never rubricated in P. 143 For V see Mioni (1976) 300, and Martinelli Tempesta (2015) 304–9 for the date of the Iliad + Quintus part of the ms.; the main scribe (but not in H.Hom.) is Cosmas of Trebizond, and the whole predates Bessarion’s bequest to the Marciana in 1468. For N see De Meyier (1965) 137. Conjunctive errors include H.Ap. 403 ἀνασείσασκε, Herm. 143 ῥόθιος, 337 ὄρεσι, 413 κρεψίφρονος, 470 ἔπορε. Shared corrections include H.Ap. 166 ἐμεῖο ss οι, 311 θεοὶ, Herm. 20 γυίων, 114 τηλόσε. Errors particular to V (Herm. 1–100 only): 12 εἴ for εἴς, 23 ὑψηρεφέως, 25 τ’ ἐκτήσατοα ss. ’ἀ, 26 ῥάοι, 29 ἔοιπε, 44 ἐπιστροφῶσι, 46 ἕπος, 47 δόνακος, 49 πραπήδεσσϊν, 63 . for μὲν, 90 ἐπικαμπύλους, 98 δημιουργὸς, 100 μεγαμηδείδας; also the delightful 361 αἰγὰς. Errors particular to N include 181 ἔκε, 209 ὀπίδει, 250 om. δὲ, 278 βλέφαρον, 420 βοῖβος, 562 ἑδωδής. N’s caution about making corrections is 142

80

5.1 THE MS TRADITION AND THIS EDITION

Neither hypothesis of contamination (V < N + ?, N < V + P) suits: V is too careless to be systematically combining two sources, while N would have had to overlook V’s superior text at e.g. Herm. 472 μαντείας δ’, 509 σήματ’, Hy. 30.8 πάρεστι, and esp. 6.2, where V has ἄσομαι, which N omits. I infer a shared ancestor for NV (ν), derived from P (as shown by N’s gap at Hy. 6.2), after the restoration of Herm. 20 γύων (which ν corrected). At 6.2 V was able to draw on more information, though it did not do so on other occasions. The derivation of Mon from P is visible at H.Ap. 393 μινωίνου (where P has μινωί ss. ου = μινωίου, which a later reader altered to μινωί ss. νου: see below), Herm. 20 om. γύων (with PL2R2). The former makes it look as if Mon was copied after ν, the latter before. It is the latter which is less significant: Mon tends to exclude marginalia since it has very small margins of unfinished vellum, and γύων only appears in P’s inner margin. Mon also shows awareness of P2 at Herm. 114.144 So far P had indicative mistakes to share with its offspring. The evidence for deciding whether other manuscripts depend on the most correct state of P or directly on p is more slender. An instructive example is the pair BΓ, characterized by a long series of small errors along with three corrections of the inherited reading (H.Ap. 156 ὅου with a, Herm. 292 φιλητέων with R2, H Aphr. 87 ἐπιγναμπτάς). Their parent (which I call β1) was at Apostoles’ workshop in Crete c.1490, where the two scribes of B who worked on Herm. struggled with phonetic errors and less common words, while Aristoboulos Apostoles did a more

visible in e.g. 451 μέλ[[ο]]υσι, H.Aphr. 146 ἀγορεύεις with inherited ἀγοράζεις piously saved in the margin. 144 Mon is digitized at http://daten.digitale-sammlungen.de/bsb00050127/. It has a conjunctive error with P in the Pseudo-Herodotean life (Vasiloudi 2013: 88). In theory Mon could be a rebound part of, or be derived from, a book including the Pseudo-Herodotean life and H.Hom. which was owned by Pico della Mirandola (Kibre 1936 no. 110).

81

I N T RO D U C T I O N

accurate job on this part of Γ (signed f. 95).145 β1 shares several errors with P2 (H Ap. 51 ἐμεῖο, Herm. 371 νέον γ’, H.Aphr. 82 τε καί for καί; also 451 μέλσι B and μέλησι Γ ~ the apparent μέλυσι in P). Moreover, in an example of how the traditional casual collation of derivative manuscripts can lead scholars into error about even carefully collated ones, there is an un­­noticed sequence of conjunctive errors between β1 and R1, which was written by Ioannes Rhosos for a member of the Rinaldi family in the early 1470s.146 The haplography Herm. 64 -γγα φυρὴν for -γγα γλαφυρὴν is particularly telling, but just within Herm. I also noted 37 ἐπηλυσίας, 105 βοτάνεις, 234 ἠρόεν, 519 ὄμβριμον, and the unusual μ-like kappas in R1 at 254 and 358 λίκνω, where β1 had λήμνω, λίμνω. These, and doubtless other alignments in the other hymns, alongside separative errors in each of R1 and β1, suggest that they are siblings, whose parent I call β. Conjunctive readings of βP2 can be found at Herm. 371 and H Aphr. 82, while R1 also matches P2 at O Arg. 1168. Harv shares a scribe, layout, approximate date and illuminated initial with R1, but not the errors of β.147 The set of distinctive readings shared with P put it at the same stage as β or at a later one.

For B see the facsimile and Cariou’s detailed description available from http://gallica.bnf.fr/. The nearest match I have found for the size of the watermark (crossbow in circle, c.38 mm wide) is Harlfinger (1974) s.v. Arbalète 33 (1487/8). For Γ see Thomas (2016a). It is of interest for early reading practices, because of its annotations in at least six hands; even the earliest level seems to show knowledge of the first printed edition (at e.g. H.Ap. 450, H.Aphr. 20), while a later one relates to collations noted in the Marciana’s copy specifically: see Càssola 614–16. 146 The date of R1 and Harv is implied by the attribution of illuminations in them to Francesco d’Antonio del Chierico (Lazzi 1998: 62–4) and Benedetto di Silvestro, respectively; the hand suits Rhosos’ other manuscripts of roughly this date. 147 For a description see Kavrus-Hoffman (2011) 7–16. Distinctive errors include H.Ap. 523 δεῖ δ’, Herm. 139 βαθυδίδην, 237 θνήεντ’, 272 μετὰ γραύλοισε, 468 θάασεις, 492 βουσὶν ὁμοὺς, 506 μητιεύτα, 518 ἤ μοι. 145

82

5.1 THE MS TRADITION AND THIS EDITION

C and L3 were written by Demetrios Damilas for the Medici c.1485, and share similar decorations.148 These repeat the main distinctive features of P, and their treatment of the other passages where P was annotated is consistent with their late date. The balance of errors at least up to Hy. 9.3 suggests that L3 was copied from C.149 However, in that line L3 has the correction μελήτης, probably from a descendant of Θ; from at least Hy. 27.13 it is C’s text that derives from Θ with lesser input from p (mainly in titles and at 28.8, 32.6), while L3 sits within p. Damilas had access to copies derived from both p and Θ (most likely L, which Poliziano read in Florence in the 1480s); his copying procedure, and so the relationships among the four books, varied over time. Finally we come to a pair with a lengthy list of conjunctive readings, including several corrections – A (written by Niccolo Leonico Tomeo c.1488) and Q (written by Demetrios Damilas c.1490); they are also twins in the Orphic Hymns, Callimachus’ Hymns and Musaeus.150 Their parent (which I call q) did not C can be dated between Damilas’ arrival in Florence in 1484 and Wrocław, Biblioteka Uniwersytecka, Rehdigeranus 35, which copied its text of Sc. in 1488 (Pérez 1994: 190–6). It is digitized at http://gallica.bnf.fr/. L3 is digitized with bibliography at http://mss.bmlonline.it/. Its illuminations have been traced to the workshop of Francesco Rosselli (Di Domenico 2005: 112). It also has a terminus ante quem of 1488, when Chalkondyles used its text of Ps.-Hdt. Vita Homeri. 149 Distinctive shared readings: H.Ap. 393 μίνου, 518 οἷτε, Herm. 168 ἄπλιστοι, 178 εἶμε, 187 ἀγνὸν, 216 εἱλύποδας, 295 ἰωνὸν, 325 ἀθάνατον, 511 αὖ, 523 ἀτὰρ, Hy. 7.56 εἰ μὴ, 28.8 ἵθ’. They even share a peculiar way of writing ῥεῖα with circumflex over ῥ in Herm. 351, 412, 417, 485. L3 introduces further errors at e.g. Herm. 9 om. τε, 20 ἀπαθανάτων, 62 μενοινά, 163 μήτηρ, 177 τίοι, 187 ἐρισφαράγγου, 196 δὲ for δὴ, χεῖρας τε, 298 ἑρμήν, 340 ἐμᾶς, 539 κασίγνυτε. 150 For a facsimile and detailed description of A see http://gallica.bnf.fr/. For Tomeo see Vendruscolo (1996), Jackson (2009). The watermarks are close to some found in the early 1490s (~ Briquet 1907: nos. 82, 3391, Harlfinger 1974 s.vv. Aigle 22, Chapeau 21), but A was the source of O.Arg. in Rehdigeranus 35 in 1488 (Vian 1979: 19). Tomeo visited Florence in 1485–8. For 148

83

I N T RO D U C T I O N

derive from P, because otherwise it would treat P’s variants inconsistently: at H.Ap. 51 ἐμοῖο q would present a reading over-written in P by the time of L2, whereas in H Ap. 393, Herm. 20, 114 it would be later than L2. The independence of q from P makes it the most likely source of four incorrect annotations in P2 which start being replicated in the copies after L2: Herm. 20 γύων (in P2 by the time of ν) 114 τιλόσε (in P2 by the earliest of L4 β Mon Harv) H.Aphr. 82 μέγεθοςτε καί (in P2 by β Harv) More complex is H.Ap. 393: μονωίνου A, μινωίνου Q < μινωί ss. νου q, whose scribe took p μινωίου (=P) as indicating a variant. q’s extra nu appeared in P2 between ν and Mon Harv. Presumably, therefore, q is also the source of P’s correction ἄσομαι at 6.2, added between ν and the earliest of L4 β Mon V. Although Vian (1979) 19 argued that A derives from P in O.Arg., the evidence presented is weak and in fact O Arg. 195 αἰσήποιο Ψ: ἀσωποῖο A: ἄλλως ἀσώποιο P2,m is another sign of P being corrected from q. In the intellectual culture of 1460s Florence, the readings may have passed from q to P by, for example, collation, information exchanged in letters, or notes taken at a lecture. By contrast, P’s variants at H.Ap. 293 and Hy. 6.18 seem to call for knowledge coming from outside p. This should be dated before O, with the small variant at Hy. 6.18 being ignored by P’s descendants other than L4. P also preserves two marginalia (H.Ap. 71, Herm. 36) which are found in b and fit the pattern of philological comments in that source; however, at least the note at Herm. 36 was written by Eugenikos himself, so it was probably inherited through p.

Q see n. 133, Eleuteri (1981) 21–2: it presumably predates Damilas’ move to Rome in 1491, but shows characteristics of his later handwriting. For the characteristics of q see Càssola 602, 607, Olson (2012) 47–8.

84

5.1 THE MS TRADITION AND THIS EDITION

This analysis refines the evidence that P was the favoured manuscript in Florence for copying the Homeric Hymns, particularly in c.1460–75, when P was used for L2 ν β L4 Mon Harv and perhaps V, and was studied in connection with q both before and after ν. Interest in P was probably driven more by the Orphic Argonautica rather than the Homeric Hymns. However, we have a little external evidence for reading of the latter, in that the period of greatest activity is delimited by Ficino’s (lost) translation of both texts c.1462, and by Poliziano’s borrowings from Hy. 7 in epigrams of 1472/3.151 The model proposed here and represented below in the stemma may be refined by further study of manuscripts, since I have focused on the readings in the Hymn to Hermes. My model faces the challenge that some of P’s foibles (e.g. H.Ap. 283 τετραρμένον, perhaps another early annotation derived from Θ; Herm. 146 δια-) were corrected with great regularity among its descendants, while others (notably H.Ap. 62 λοιτοῖ, Herm. 451 μέλυσι) were rarely changed.152 Nevertheless, it is the most effective interpretation of the evidence available to me. Because of the importance of P changing over time in the preceding discussion, in the stemma I have used vertical lines with horizontal branches to indicate the relative dates of copy­ ing where these are known, and the usual oblique branches otherwise; as the preceding discussion makes clear, the relative chronology is not always certain. For Ficino’s early translations of O.Arg., Orph.H., Procl. H. and Hes. Th., see his letter to Martin Preninger of June 1492 (1561: 933); the date is based on Laur. 54.10 f. 81, a copy of a letter of September 1462 which contains his translation of Orph.H. 4 to Cosimo de’ Medici. That translation is certainly based on a manuscript of p’s family. It is tantalizing that Ficino and Agnolo Manetti, the owner of P, were acquainted (Ficino 1561: 632). For Hy. 7 in Poliziano Epig. 3, 10 see Pontani (2002), Schwab (2016) 306 n. 21. 152 Scholars have claimed other instances of P’s errors being widely corrected, at H.Aphr. 260 τελήεσσας, 496 δέλφιος, H.Aphr. 147 ἕκατι, 216 γεγήθε. The facsimile shows that the first three are not in P; the last is unclear. 151

85

I N T RO D U C T I O N Ω ψ

Θ

p x

D

M

a

At

P

q

b

O L2 R2

v V

β T

L4

E

Mon

β1 R1 N

Harv Π

L L3 C

1st ed.

B

Γ

A

Q

5.2 Rationale of Edition and Apparatus I think of this as a conservative edition, and I always started by trying to defend manuscript readings to my satisfaction. However, I am aware that my threshold for satisfaction is higher than that of recent commentators including Vergados, Schenck zu Schweinsberg and Richardson. I have been less conservative regarding orthography, in line with what scholars have reconstructed about rhapsodic practices, which in many cases seem to have endured so that Alexandrian scholars could make coherent claims about them. There is a clear risk in assuming too great a homogeneity in rhapsodic practices: it is theoretically possible that an Attic composer would have caused a smattering of Atticisms (for example), or that rhapsodes (or transcribers) varied the forms ἐς and εἰς 86

5.2 THE MS TRADITION AND THIS EDITION

before a consonant whimsically or for purely sonic reasons. However, in my view it is much more likely that such features are part of the very widespread tendency of Attic and koine to creep into the paradosis of epic, a process whose tail-end we can observe in our manuscripts. That tendency, in any case, leaves arguments such as ‘Ω contains several Atticisms so perhaps the composer was Attic’ on such shaky ground that my orthographic changes do not affect any cogent arguments about the poem’s origins. In the three cases where two ancient variants of a whole line are preserved (288, 366, 563) I have printed both, marked with a bracket in the margin. For other information about smaller ancient variants see the Index s.v. ‘Transmission, ancient variants’. My running translation is intended as a preliminary point of orientation towards my interpretation, and does not generally aim at elegance or at finding English equivalents of the many wordplays which are discussed in the commentary. Any critical apparatus has to navigate between utility and comprehensiveness. I have tried to err on the side of the latter.153 I cite variants at the level of MΘp, with the following exceptions, which themselves have not been applied too rigidly: • Unmetrical or morphologically impossible forms in Θ (or p) where Mp (or MΘ) share an acceptable reading. • Unmetrical or morphologically impossible forms in M which differ from an acceptable reading in Ψ only by a diacritic, a basic sonic error (such as itacism), or a shift to a more prosaic equivalent. Vergados gave a misleading sense of comprehensiveness by citing a selection of errors in manuscripts near the bottom of the stemma, but only a tiny fraction of the variations that exist. Allen and Halliday lviii–lxiv list examples of the kinds of minor variation which I have generally not mentioned.

153

87

I N T RO D U C T I O N

• Anything involving punctuation that does not affect understanding of syntax. I have also reported variants lower down the stemma where they cause doubt over what Θ and p contained, or where a later manuscript made a good correction. Where that hesitation is minimal, I try to keep the focus on deeper divisions by using parentheses. For example ‘42 αἰῶν’ Ψ (αἰῶν x)’ means that Dp had αἰῶν’, which must have been inherited from Ψ, even though x omitted the apostrophe. I have tried to note all editorial interventions which I have accepted, but only on the first occurrence if an orthographical principle is involved. I have also taken the liberty of claiming some readings which appeared in my doctoral thesis before others published them independently. Most of the information about conjectures and testimonia is located in the commentary, where they can receive discussion. Testimonia also have their own index entry. Ω, Ψ, Θ, x, a, b, p, q and β1 are used in the apparatus to mark readings in a particular lost manuscript (see the stemma above) which we can generally reconstruct with confidence. With marginalia, however, this would be misleading: a reading marked xm is found in the margin in TEΠL; one found in the text in TE and in the margin in ΠL is notated abm, even though it was doubtless in the margin of x. My edition has iota adscripts whereas the manuscripts rarely include the iota even as a subscript. Where a lemma containing iota adscript is attributed to a manuscript, you may assume that that particular letter is not actually represented. Finally I have used e.g. ὅ()δ’ to notate a sequence which could be interpreted either as ὅδ’ or as ὅ δ’ (i.e. ὃ δ’ by modern principles of accentuation). I have avoided switching into Latin for the apparatus. It is neither the case that someone without confident Latin should be discouraged from finding out about transmission, nor that everyone interested in manuscript readings in Greek texts does 88

5.2 THE MS TRADITION AND THIS EDITION

nowadays have good Latin. However, several standard abbreviations translate easily: om. X: omitted in X ss. αβγ: with superscript (interlinear) αβγ supp. X: supplemented in X + marg. αβγ: with αβγ in the margin (though Xm is used where possible) X2 denotes a reading added in X by a hand other than the original scribe (not necessarily the second scribe in particular).

89

SIGLA I have retained the traditional sigla. Folio numbers in the list are those containing the Homeric Hymns. A: Paris, Bibliothèque Nationale, grec 2763 (ff. 91–129v) At: Mt Athos, Mone Vatopedi 671 (ff. 191–218) B: Paris, Bibliothèque Nationale, grec 2765 (ff. 24v–58) C: Paris, Bibliothèque Nationale, grec 2833 (ff. 44–85) D: Milan, Biblioteca Ambrosiana, B98 sup. (120) (ff. 178v–209) E: Modena, Biblioteca Estense Universitaria, α.W.5.16 = gr. 164 (ff. 50–84v) Harv: Cambridge Ma., Houghton Library, MS Typ 18 (ff. 60–98) L: Florence, Biblioteca Medicea Laurenziana, Plut. 32.45 (ff. 155v–181v) L2: Florence, Biblioteca Medicea Laurenziana, Plut. 70.35 (ff. 68v–103v) L3: Florence, Biblioteca Medicea Laurenziana, Plut. 32.04 (ff. 450v–76r) L4: Florence, Biblioteca Medicea Laurenziana, Aedil. 220 (ff. 51–84v) M: Leiden, Universiteitsbibliotheek, BPG 33H (ff. 31–50) Mon: Munich, Bayerische Staatsbibliothek, Cod. graec. 333 (ff. 72–90v; ends at Herm. 192) N: Leiden, Universiteitsbibliotheek, BPG 74C (ff. 63–104v) O: Milan, Biblioteca Ambrosiana, C10 inf. (845) (ff. 127–43v; ends at Herm. 80) P: Vatican, Biblioteca Apostolica Vaticana, Pal. gr. 179 (ff. 86–129) Q: Milan, Biblioteca Ambrosiana, S31 sup. (734) (ff. 39–89) R1: Florence, Biblioteca Riccardiana, 53 (ff. 61–99) R2 : Florence, Biblioteca Riccardiana, 52 (ff. 31–72v) T: Madrid, Biblioteca Nacional, 4562 (ff. 56–83v) 90

SIGLA

V: Venice, Biblioteca Marciana, 456 (ff. 509–33v) Γ: Brussels, Koninklijke Bibliotheek, 11377–80 (ff. 31–67) Π: Paris, Bibliothèque Nationale, supplément grec 1095 (ff. 225–45)

91

T H E H O M E R I C H Y M N TO HERMES

T E X T A N D T R A N S L AT I O N Ἑρμῆν ὕμνει, Μοῦσα, Διὸς καὶ Μαιάδος υἱόν Κυλλήνης μεδέοντα καὶ Ἀρκαδίης πολυμήλου, ἄγγελον ἀθανάτων ἐριούνιον, ὃν τέκε Μαῖα νύμφη ἐϋπλόκαμος Διὸς ἐν φιλότητι μιγεῖσα, αἰδοίη· μακάρων δὲ θεῶν ἠλεύαθ’ ὅμιλον5 ἄντρον ἔσω ναίουσα παλίσκιον, ἔνθα Κρονίων νύμφηι ἐϋπλοκάμωι μισγέσκετο νυκτὸς ἀμολγῶι ὄφρα κατὰ γλυκὺς ὕπνος ἔχοι λευκώλενον Ἥρην, λήθων ἀθανάτους τε θεοὺς θνητούς τ’ ἀνθρώπους. ἀλλ’ ὅτε δὴ μεγάλοιο Διὸς νόος ἐξετελεῖτο,10 τῆι δ’ ἤδη δέκατος μεὶς οὐρανῶι ἐστήρικτο, ἔς τε φόως ἄγαγεν, ἀρίσημά τε ἔργα τέτυκτο· καὶ τότε γείνατο παῖδα πολύτροπον, αἱμυλομήτην, ληϊστῆρ’, ἐλατῆρα βοῶν, ἡγήτορ’ ὀνείρων, νυκτὸς ὀπωπητῆρα, πυληδόκον, ὃς τάχ’ ἔμελλεν15 ἀμφανέειν κλυτὰ ἔργα μετ’ ἀθανάτοισι θεοῖσιν· ἠῶιος γεγονὼς μέσωι ἤματι ἐγκιθάριζεν, ἑσπέριος βοῦς κλέψεν ἑκηβόλου Ἀπόλλωνος, τετράδι τῆι προτέρηι τῆι μιν τέκε πότνια Μαῖα. ὃς καὶ ἐπεὶ δὴ μητρὸς ἀπ’ ἀθανάτων θόρε γυίων20 οὐκέτι δηρὸν ἔκειτο μένων ἱερῶι ἐνὶ λίκνωι, ἀλλ’ ὅ γ’ ἀναΐξας ζήτει βόας Ἀπόλλωνος οὐδὸν ὑπερβαίνων ὑψηρεφέος ἄντροιο. ἔνθα χέλυν εὑρὼν ἐκτήσατο μυρίον ὄλβον· 25 Ἑρμῆς τοι πρώτιστα χέλυν τεκτήνατ’ ἀοιδόν,  ἥ ῥά οἱ ἀντεβόλησεν ἐπ’ αὐλείηισι θύρηισιν βοσκομένη προπάροιθε δόμων ἐριθηλέα ποίην, σαῦλα ποσὶν βαίνουσα. Διὸς δ’ ἐριούνιος υἱός ἀθρήσας ἐγέλασσε καὶ αὐτίκα μῦθον ἔειπεν· [Title] τοῦ αὐτοῦ [i.e. Ὁμήρου] ὕμνοι εἰς ἑρμῆν M: εἰς ἑρμῆν Db: ὕμνος δεύτερος εἰς ἑρμῆν a: εἰς τὸν ἑρμῆν p  1 ὕμνει Mp: ὑμνεῖ Θ  11 μεὶς xp: μὴς ss. εὶς M: μεῖς D   12 ἔς Hermann: εἴς Ω  13 τότε γείνατο Mxp: τότ’

94

T H E H O M E R I C H Y M N TO H E R M E S

Celebrate Hermes, Muse, the son of Zeus and Maia who is patron of Kyllene and Arcadia rich in flocks, the beneficent messenger of the immortals, who was born to Maia, the nymph with beautiful hair, after she mingled in Zeus’s love, (5) but she was bashful, and avoided the throng of blessed gods, dwelling inside a cave of shadows upon shadows. It was there that the son of Kronos used to mingle with the nymph with beautiful hair in the dead of night while sweet sleep gripped whitearmed Hera, unbeknownst to immortal gods and mortal men. (10) But when the intent of great Zeus came to fulfilment, and for her the tenth moon had finally taken its position in the heavens, he(?) led him into the light and notable deeds were accomplished. Then it was that she bore a versatile child of coaxing counsels, brigand, driver of cattle, the guide of dreams, (15) who is on the lookout at night, who waits at gates, and who was soon to reveal famous deeds among the immortal gods: born at dawn, he played the lyre in the middle of the day, and in the evening stole the cattle of far-shooting Apollo, on the fourth from the start of the month, when lady Maia bore him. (20) He indeed, when he had jumped from his mother’s immortal limbs, did not then stay and lie for long in his sacred winnowing-fan cradle: he rather leapt up and began to seek Apollo’s cows, stepping over the threshold of the high-roofed cave. There, he found a tortoise, and gained immeasurable prosperity: (25) yes, Hermes first engineered a tortoise to be a singer – the one who crossed his path at the courtyard gates, browsing on the flourishing vegetation in front of the house, stepping with a swagger. The beneficent son of Zeus spotted

ἐγείν- D   15 πυληδόκον Ψ: πολύδοκον M   20 καὶ Ψ: om. M   ἐπεὶ δὴ Sterrett (1881) xlviii: ἐπειδὴ Ω  22 ζήτει Ψ: ζητεῖ M   26 αὐλείηισι Turnebus: αὐλίησι Ω  θύρηισιν Hermann: θύρησι Ω  28 σαῦλα Ψ: σκύλα M  

95

T H E H O M E R I C H Y M N TO H E R M E S

“σύμβολον ἤδη μοι μέγ’ ὀνήσιμον· οὐκ ὀνοτάζω. 30 χαῖρε, φυὴν ἐρόεσσα, χοροιτύπε, δαιτὸς ἑταίρη, ἀσπασίη προφανεῖσα. πόθεν τόδε καλὸν ἄθυρμα; αἰόλος ὄστρακον ἐσσί, χέλυς ὄρεσι ζώουσα. ἀλλ’ οἴσω σ’ ἐς δῶμα λαβών. ὄφελός τί μοι ἔσσεαι, 35 οὐδ’ ἀπατιμήσω (σὺ δέ με πρώτιστον ὀνήσεις)· ‘οἴκοι βέλτερον εἶναι, ἐπεὶ βλαβερὸν τὸ θύρηφιν’. ἦ γὰρ ἐπηλυσίης πολυπήμονος ἔσσεαι ἔχμα ζώουσ’, ἢν δὲ θάνηις τότε κεν μάλα καλὸν ἀείδοις.” ὣς ἄρ’ ἔφη, καὶ χερσὶν ἅμ’ ἀμφοτέρηισιν ἀείρας 40 ἂψ εἴσω κίε δῶμα φέρων ἐρατεινὸν ἄθυρμα. ἔνθ’ ἀναμηλώσας γλυφάνωι πολιοῖο σιδήρου αἰῶν’ ἐξετόρησεν ὀρεσκώιοιο χελώνης. ὡς δ’ ὁπότ’ ὠκὺ νόημα διὰ στέρνοιο περήσει ἀνέρος ὅν τε θαμειναὶ ἐπιστρωφῶσι μέριμναι, 45 ἠ’ ὅτε δινηθῶσιν ἀπ’ ὀφθαλμῶν ἀμαρυγαί, ὣς ἅμ’ ἔπος τε καὶ ἔργον ἐμήδετο κύδιμος Ἑρμῆς. πῆξε δ’ ἄρ’ ἐν μέτροισι ταμὼν δόνακας καλάμοιο, πειρήνας διὰ νῶτα, διὰ ῥινοῖο χελώνης, ἀμφὶ δὲ δέρμα τάνυσσε βοὸς πραπίδεσσιν ἑῆισιν καὶ πήχυς ἐνέθηκ’, ἐπὶ δὲ ζυγὸν ἤραρεν ἀμφοῖν,50 ἑπτὰ δὲ συμφώνους ὀΐων ἐτανύσσατο χορδάς. αὐτὰρ ἐπεὶ δὴ τεῦξε, φέρων ἐρατεινὸν ἄθυρμα πλήκτρωι ἐπειρήτιζε κατὰ μέλος, ἣ δ’ ὑπὸ χειρός σμερδαλέον κονάβησε. θεὸς δ’ ὑπὸ καλὸν ἄειδεν ἐξ αὐτοσχεδίης πειρώμενος, ἠΰτε κοῦροι55 ἡβηταὶ θαλίηισι παραίβολα κερτομέουσιν, ἀμφὶ Δία Κρονίδην καὶ Μαιάδα καλλιπέδιλον, 30 punctuation after ὀνήσιμον Ψ: no punctuation M   32 punctuation Krause (1790) 418   33 αἰόλος ὄστρακόν ἐσσι or αἰόλον ὄστρακόν ἐστι Thomas: αἰόλον ὄστρακον ἐσσὶ Ω  34 ἔσσεαι Ilgen: ἔσση Ω  35 ἀπατιμήσω Matthiae (1800) 215: ἀπο- Ω  37 ἦ Ψ: εἰ M   ἔχμα Ruhnken (1780) 41: αἶχμα M: αἰχμὰ Ψ  38 ζώουσ’, ἢν Ψ (-σα· ἢν D): ζώουσι M  θάνηις MD: θάνοις xp  τότε κεν Hermann: τότ’ ἂν M: τότε ἂν Ψ   41 ἀναμηλώσας Ruhnken (1751) 62: ἀναπηλήσας Ω  42 αἰῶν’ Ψ (αἰῶν x): αἰὼν + marg. γρ′ ὡς δοκεῖ μοι, ἀγῶν’ ἐξετό M   ὀρεσκώιοιο

96

T H E H O M E R I C H Y M N TO H E R M E S

her, laughed, and straightaway spoke: (30) ‘A sign already – and a very beneficial one for me! I don’t disparage it. Joy to you, gorgeous-bodied girl who stamps in the chorus, companion of the feast, a welcome appearance. Where’s this pretty plaything from? You have a shifty shell-colour, mountain-dwelling tortoise! Anyhow, I shall take you and carry you into the house. You’ll be a help for me, (35) and I’ll not completely dishonour you – for you will be the first to benefit me: “Better to be at home, since the outdoors is harmful”. To be sure, you will hold off a painful seizure while you’re alive, and if you die, you may then sing very beautifully.’ So he spoke, and lifting it at the same time with both hands (40) he went back into the house carrying his lovely plaything. There, probing it open(?) with a gouge of grey iron he bored out the life-marrow of the mountain-dwelling tortoise. As when a swift thought crosses into the chest of a man in whom a throng of worries roam, (45) or when twinkling glances are whirled from eyes, so glorious Hermes contrived both speech and deed together. And he affixed reed-canes which he had cut down to their measures, making them span across the tortoise’s back, through its hide, and he stretched cowskin around it using his wits, (50) inserted arms and fitted a bridge onto both, and stretched seven gut-strings from sheep into attunement. When he had made it, he took his lovely plaything and with a plectrum tested it in a scale, and at his touch (54) it reverberated fearsomely. The god sang to it well, making an extempore trial as youths in their prime make bantering retorts at festivities, about Zeus son of Kronos and fair-shod Maia, who previously used to court χελώνης MDp: ὀρεσκώ λώνης x (-ηι supp. T: -ιο κο- supp. Π)  44 θαμειναὶ Allen (1894) 218: θαμιναὶ Ω  45 ἠ’ Thomas (2010a): ἢ M Γ2,m: αἳ Θ: ἃς p  ἀμαρυγαί MDbp: ἀμαλδύναι abm  48 διὰ νῶτα, διὰ ῥινοῖο Ω: κατὰ ν., δ. ῥ. Allen (1912): δ. ν. λιθοῤῥίνοιο Pierson (1752) 156   50 πήχυς M: πήχεις Ψ  δὲ Ψ: om. M   51 συμφώνους Ω: θηλυτέρων Antigonus  53 μέλος D’Orville (n.d.) f. 246: μέρος Ω  ἣ δ’ Hermann: ἥ()δ’ Ω  54 κονάβησε M: κονάβισσε Ψ (-ισε D)   55 ἐξ αὐτοσχεδίης MDb: ἐξαυτ- ap  ἠΰτε κοῦροι Ψ: ἥντε κοράοι M  

97

T H E H O M E R I C H Y M N TO H E R M E S

οἳ πάρος ὠρίζεσκον ἑταιρείηι φιλότητι, ἥν αὐτοῦ γενεὴν ὀνομάκλυτον ἐξονομάζων, ἀμφιπόλους τ’ ἐγέραιρε καὶ ἀγλαὰ δώματα νύμφης60 καὶ τρίποδας κατὰ οἶκον ἐπηετανούς τε λέβητας. καὶ τὰ μὲν οὖν ἤειδε, τὰ δὲ φρεσὶν ἄλλα μενοίνα, καὶ τὴν μὲν κατέθηκε φέρων ἱερῶι ἐνὶ λίκνωι φόρμιγγα γλαφυρήν, ὃ δ’ ἄρα κρειῶν ἐρατίζων ἄλτο κατὰ σκοπιὴν εὐώδεος ἐκ μεγάροιο,65 ὁρμαίνων δόλον αἰπὺν ἐνὶ φρεσίν, οἷά τε φῶτες φιλῆται διέπουσι μελαίνης νυκτὸς ἐν ὥρηι. Ἠέλιος μὲν ἔδυνε κατὰ χθονὸς Ὠκεανόνδε αὐτοῖσίν θ’ ἵπποισι καὶ ἅρμασιν, αὐτὰρ ἄρ’ Ἑρμῆς Πιερίης ἀφίκανε θέων ὄρεα σκιόεντα,70 ἔνθα θεῶν μακάρων βόες ἄμβροτοι αὖλιν ἔχεσκον βοσκόμεναι λειμῶνας ἀκηρασίους ἐρατεινούς. τῶν τότε Μαιάδος υἱὸς ἐΰσκοπος Ἀργεϊφόντης πεντήκοντ’ ἀγέλης ἀπετάμνετο βοῦς ἐριμύκους, πλανοδίας δ’ ἤλαυνε διὰ ψαμαθώδεα χῶρον75 ἴχνι’ ἀποστρέψας· δολίης δ’ οὐ λήθετο τέχνης, ἀντία ποιήσας ὁπλάς, τὰς πρόσθεν ὄπισθεν τὰς δ’ ὄπιθεν πρόσθεν, κατὰ δ’ ἔμπαλιν αὐτὸς ἔβαινεν. σάνδαλα δ’ αὐτίκα ῥιψὶν ἐπὶ ψαμάθοισ’ ἁλίηισιν ἄφραστ’ ἠδ’ ἀνόητα διέπλεκε, θαυματὰ ἔργα,80 συμμίσγων μυρίκας καὶ μυρσινοειδέας ὄζους. τῶν τότε συνδήσας νεοθηλέος ἀγκάλα ὥρης ἀβλαβέως ὑπὸ ποσσὶν ἐδήσατο σάνδαλα κοῦφα αὐτοῖσιν πετάλοισι, τὰ κύδιμος Ἀργεϊφόντης ἔσπασε, Πιερίηθεν ὁδοιπορίην ἀλεείνων,85 οἷά τ’ ἐπειγόμενος δολιχὴν ὁδὸν αὐτοτρόπησεν. 58 οἳ Clarke (in notes): ὡς Châteillon: ὃν Ω  ὠρίζεσκον ‘I. T. P.’ (1624): ὡρ- Ω  ἑταιρείηι Ψ: καὶ ἑταιρείη M   59 ἥν Wecklein (1920) 37: ἥν τ’ Ω  ὀνομάκλυτον Stadtmüller: ὀνομακλυτὸν M: ὄνομα κλυτὸν Θ: ὀνομακλυτὴν p  ἐξονομάζων M: ὀνομάζων Ψ  60 τ’ ἐγέραιρε Ilgen: τε γέρ- Ω  65 ἄλτο Wecklein (1920) 37: ἆλτο MD: ὦτο Eb: ὦρτο Tp  67 φιλῆται Ernesti: φιληταὶ M: φηληταὶ Ψ  69 αὐτοῖσίν Stephanus: -ιν Θ: -ι Mp  70 θέων D: θεῶν Mxp  72 ἀκηρασίους Mp: ἀκειρ- Θ  

98

T H E H O M E R I C H Y M N TO H E R M E S

in friendly love, naming his own genealogy of famous name; (60) and he honoured the nymph’s attendants and treasured house, and the tripods and abundant cauldrons in the home. Of these things he was singing, but in his mind he was planning the other business. His concave lyre he carried and deposited in his sacred winnowing-fan, while he, lusting for meat, (65) jumped out of the fragrant hall with a view to reconnoitring, planning sheer trickery in his mind, such as men who are brigands conduct in the hours of dark night. Helios was sinking beneath the Earth towards Okeanos, together with his horses and chariot, and Hermes (70) at a run reached the shadowy mountains of Pieria, where the blessed gods’ immortal cows had their sleeping-place and browsed on the lovely virgin meadows. Then of their number the son of Maia, the sharp-sighted slayer of Argus, cut fifty loud-lowing cows off from the herd, (75) and led them on a misleading path across sandy ground, turning their tracks backwards: for he did not forget his tricksy craft, but made the hooves their opposite – made the fore ones be behind, and the hind ones at the fore; and he himself began to step backwards downhill. But on the seaside sands he straightaway wove sandals from withies – (80) marvellous works, never contemplated or devised – mixing together tamarisks and myrtle-like twigs. After then binding up two armfuls of these, belonging to the season which was newly burgeoning, he painlessly bound under his feet, leaves and all, his light sandals, which the glorious slayer of Argus (85) had pulled up as he avoided a road-journey from Pieria, and such as he had made uniquely as he hastened his long journey. 74 ἀγέλης Ψ: ἀγέλας M   76 ἴχνι’ Hermann: ἴχνη Ω  78 ὄπιθεν MbL2: ὄπισθεν Dap  πρόσθεν Ψ: πρώτας M   79 σάνδαλα Ψ: σάλδαλα M  δ’ αὐτίκα ῥιψὶν Postgate cited in Allen (1896): δ’ αὐτίκ’ ἔριψεν MDp: ἔριψεν a: κ’ ἔριψεν b (αὐτί- supp. Π)  ψαμάθοισ’ Thomas: ψαμάθοις Ω  82 νεοθηλέος ἀγκάλα   ὥρης Thomas: νεοθηλέαν ἀγκαλωρήν M: νεοθηλέος ἀγκαλὸν ὕλης Ψ  84 κύδιμος MΘ: κύδιμα p  86 αὐτοτρόπησεν Thomas: αὐτοτροπήσας Mbmp: αὐτοπρεπὴς ὣς Db: αὐτοτροπήσας ὣς a: αὐτοτροπῆσαι Adorjáni (2011) 135  

99

T H E H O M E R I C H Y M N TO H E R M E S

τὸν δὲ γέρων ἐνόησε δέμων ἀνθέουσαν ἀλωήν, ἱέμενον πεδίονδε δι’ Ὀγχηστὸν λεχεποίην. τὸν πρότερος προσέφη Μαίης ἐρικυδέος υἱός· “ὦ γέρον, ὅς τε φυτὰ σκάπτεις ἐπικαμπύλος ὤμους,90 ἦ πολυοινήσεις εὖτ’ ἂν τάδε πάντα φέρηισιν. καί τε ἰδὼν μὴ ἰδὼν εἶναι καὶ κωφὸς ἀκούσας καὶ σιγᾶν, ὅτε μή τι καταβλάπτηι τὸ σὸν αὐτοῦ.” τόσσον φὰς συνέσευε βοῶν ἴφθιμα κάρηνα. 95 πολλὰ δ’ ὄρη σκιόεντα καὶ αὐλῶνας κελαδεινούς  καὶ πεδί’ ἀνθεμόεντα διήλασε κύδιμος Ἑρμῆς, ὀρφναίη δ’ ἐπίκουρος ἐπαύετο δαιμονίη Νύξ ἡ πλείων, τάχα δ’ ὄρθρος ἐγίνετο δημιοεργός, ἣ δὲ νέον σκοπιὴν προσεβήσετο δῖα Σελήνη Πάλλαντος θυγάτηρ Μεγαμηδείοιο ἄνακτος.100 τῆμος ἐπ’ Ἀλφειὸν ποταμὸν Διὸς ἄλκιμος υἱός Φοίβου Ἀπόλλωνος βοῦς ἤλασεν εὐρυμετώπους, ἀδμῆτες δ’ ἵκανον ἐς αὔλιον ὑψιμέλαθρον καὶ ληνοὺς προπάροιθεν ἀριπρεπέος λειμῶνος. ἔνθ’ ἐπεὶ εὖ βοτάνης ἐπεφόρβει βοῦς ἐριμύκους,105 καὶ τὰς μὲν συνέλασσεν ἐς αὔλιον ἁθρόας οὔσας λωτὸν ἐρεπτομένας ἠδ’ ἑρσήεντα κύπειρον, σὺν δ’ ἐφόρει ξύλα πολλά. πυρὸς δ’ ἐπεμαίετο τέχνην· δάφνης ἀγλαὸν ὄζον ἑλὼν ἐπέλεψε σιδήρωι

110 ἄρμενον ἐν παλάμηι, ἄμπνυτο δὲ θερμὸς ἀϋτμή. Ἑρμῆι τοι πρώτιστα πυρήϊα πῦρ ἀνέδωκεν. 87 δέμων M: δόμων Ψ  ἀνθέουσαν Thomas: ἀνθοῦσαν M: αἴθουσαν Ψ  88 ἱέμενον L: ἰέ- Ω  Ὀγχηστὸν λεχεποίην Ψ: ὀγχηστῶν λεχεποίων ss. γρ′ ὸν     ην M   90 ἐπικαμπύλος ὤμους Ψ (ἐπὶ κ. ὤ. P): ἐπικαμπύλα ξύλα M   91 πολυοινήσεις Ilgen: πολὺ οἰνήσεις M: πολὺ οἰμήσεις Θ: πολύ οἱ μήσεις p  94 φὰς συνέσευε Chalkondyles: φασὶν ἔσευε Ω (ἔσσ- bp)  98 ἐγίνετο L N (ss. γ; cf. 222 γίνεται Ω): ἐγέν- M: ἐγίγνΨ  100 Μεγαμηδείοιο a: μέγα μηδείδαο M: μεγαμηδείαο D: μέγα μηδείδιο Π: μεγαμηδ:είδοιο L: μεγαμηδείδαο p (reconstruct μεγαμηδείδαο ss. οιο

100

T H E H O M E R I C H Y M N TO H E R M E S

But an old man who was building a blossoming plot noticed him as he hurried towards the plain through Onchestos couched in grass. The son of glorious Maia addressed him first: (90) ‘Old man, you who do the digging for your plants with shoulders hunched, you will certainly have plenty of wine to drink as and when all these bear: though you have seen, be one who has not seen; be deaf though you have heard; and be silent when it does no damage to your affairs.’ He said this much, and harried together the strong and valued cows. (95) Glorious Hermes drove them through many shadowy mountains, plashing gorges and flower-filled plains. His dark accomplice, divine Night in her greater part, drew to a stop, and presently that public servant pre-dawn came on, and great Selene, (100) the daughter of lord Pallas the son of Megamedes, had newly ascended her look-out place. At that time, the mighty son of Zeus drove Phoebus Apollo’s broadfronted cattle to the river Alpheus, and the unyoked animals reached troughs and a high-beamed lodge in front of an outstanding meadow. (105) There when he was done feeding the loud-lowing cows generously on fodder, he also gathered them and drove them together into the lodge as they grazed on the clover and dewy sedge, and he brought together many logs. He reached for a craft of fire: taking a splendid laurel branch, he stripped it with an iron blade (110) well fitted in his palm, and a hot blast was exhaled. Yes, it was for Hermes that fire-sticks first produced fire. Ψ?)  103 ἀδμῆτες Ω: ἀκμῆτες Ilgen   ἵκανον Mxp: ἤλαυνον D   105 ἐπεὶ M: ἔπει Ψ  106 ἁθρόας LSJ s.v. ἀθρόος: ἀθρ- Ω  108 τέχνην Ψ: τύνη M   109 ἐπέλεψε σιδήρωι Ψ: ἐνίαλλε σιδήρω M: ἐνίαλλε σιδείωι Ludwich (1905) 9   lacuna Schneidewin (1848) 668   110 ἄρμενον L, Chalkondyles: ἅρ- Ω  ἄμπνυτο δὲ M: ἀνὰ δ’ ἄμπνυτο Ψ  θερμὸς ἀϋτμή Ψ: θυμὸς ἀυτμῆ M   111 Ἑρμῆι ... πῦρ Ludwich (1886) 809: ἑρμῆς ... πῦρ τ’ Ω  

101

T H E H O M E R I C H Y M N TO H E R M E S

πολλὰ δὲ κάγκανα κᾶλα κατουδαίωι ἐνὶ βόθρωι οὖλα λαβὼν ἐπέθηκεν ἐπηετανά, λάμπετο δὲ φλόξ τηλόσε φῦσαν ἱεῖσα πυρὸς μέγα δαιομένοιο. 115 ὄφρα δὲ πῦρ ἀνέκαιε βίη κλυτοῦ Ἡφαίστοιο, τόφρα δ’ ὑποβροχίας ἕλικας βοῦς εἷλκε θύραζε δοιὰς ἄγχι πυρός· δύναμις δέ οἱ ἔπλετο πολλή· ἀμφοτέρας δ’ ἐπὶ νῶτα χαμαὶ βάλε φυσιοώσας, ἐγκλίνων δ’ ἐκύλινδε δι’ αἰῶνας τετορήσας. 120 ἔργωι δ’ ἔργον ὄπαζε ταμὼν κρέα πίονα δημῶι, ὤπτα δ’ ἀμφ’ ὀβελοῖσι πεπαρμένα δουρατέοισιν, σάρκας ὁμοῦ καὶ νῶτα γεράσμια καὶ μέλαν αἷμα ἐργμένον ἐν χολάδεσσι. τὰ δ’ αὐτοῦ κεῖτ’ ἐπὶ χώρης, ῥινοὺς δ’ ἐξετάνυσσε καταστυφέλωι ἐνὶ πέτρηι, 125 ὡς ἔτι νῦν τὰ μέτασσα πολυχρόνιοι πεφύασιν δηρὸν δὴ μετὰ ταῦτα καὶ ἄκριτον. αὐτὰρ ἔπειτα Ἑρμῆς χαρμόφρων εἰρύσσατο πίονα ἔργα λείωι ἐπὶ πλαταμῶνι, καὶ ἔσχισε δώδεκα μοίρας κληροπαλεῖς, τέλεον δὲ γέρας προσέθηκεν ἑκάστηι. 130 ἔνθ’ ὁσίης κρεάων ἠράσσατο κύδιμος Ἑρμῆς· ὀδμὴ γάρ μιν ἔτειρε καὶ ἀθάνατόν περ ἐόντα ἡδεῖ’. ἀλλ’ οὐδ’ ὧς οἱ ἐπείθετο θυμὸς ἀγήνωρ καί τε μάλ’ ἱμείροντι περᾶν ἱερῆς κατὰ δειρῆς, ἀλλὰ τὰ μὲν κατέθηκεν ἐς αὔλιον ὑψιμέλαθρον, 135 δημὸν καὶ κρέα πολλά, μετήορα δ’ αἶψ’ ἀνάειρεν σῆμα νέης φωρῆς. τὰ δ’, ἐπὶ ξύλα κάγκαν’ ἀείρας, οὐλόποδ’ οὐλοκάρηνα πυρὸς κατεδάμνατ’ ἀϋτμῆι. αὐτὰρ ἐπεὶ δὴ πάντα κατὰ χρέος ἤνυσε δαίμων, σάνδαλα μὲν προέηκεν ἐς Ἀλφειὸν βαθυδίνην, 112 κᾶλα p: καλὰ M: κάλα Θ  κατουδαίωι Γ: κατ’ οὐδ- Ω  114 φῦσαν D’Orville (n.d.) f. 247: φύζαν Ω (φύσαν E)   ἱεῖσα p: ἰ- MΘ  116 ὑποβροχίας Thomas: ὑποβρυχίας Ω: ὑπὸ βρυχαῖς Shackle (1915)   119 ἐγκλίνων Ψ: ἐκκρίνας M   δι’ αἰῶνας Θ: διαιῶνας M: δι’ αἰῶνος p  120 πίονα M: πίονι Ψ  121 ὤπτα δ’ Θ: ὠπτὰ δ’ M: ὦπτο δ’ ἄρ’ p  122 γεράσμια Ψ: γ’ ἐρ- M   123 ἐργμένον ap: ἔργμενον M: ἑργμένον Db  124 καταστυφέλωι At x: κατὰ στυφελῆ M: κατὰ στυφέλω Dp  ἐνὶ Ω: ἐπὶ Barnes (cf. 404)   125 τὰ μέτασσα M: τάμετ’ ἄσσα Ψ  

102

T H E H O M E R I C H Y M N TO H E R M E S

He took many desiccated bits of dense timber and added an abundant supply of them in a pit dug into the ground. And the flame began to shine, sending far and wide the puff of the blazing fire. (115) While famous Hephaestus’ force was kindling the fire, by a noose he dragged a pair of spiral-horned cows outdoors to near the fire – his strength was great – and cast both of them puffing to the ground upon their backs; as he leaned on them he pierced their vital parts and sent them rolling. (120) To deed he added deed after cutting up the meat rich with fat, and he roasted it skewered on long wooden spits – the flesh together with the honorific chines and the dark blood enclosed within the intestines. Other parts lay there in their place, but he stretched out the skins within a rugged rockface, (125) as still now in later times they are planted there indistinguishably, of great duration, long indeed after these events. Next joyful Hermes pulled his rich handiwork off onto a smooth flat rock, split twelve portions to be assigned by ballot, and added a perfect honorific piece to each. (130) Then glorious Hermes felt a lust for the right to the meat, since the smell was tormenting him despite his immortality, so sweet was it. But not even so did his manly spirit listen, though he greatly desired to pass the meat down his sacred neck. Rather, he deposited some parts in the high-beamed lodge – (135) the fat and plentiful meat – and straightaway raised it up on high, a sign of his recent theft. The rest, including all of the feet and all of the heads, he totally destroyed with the blast of fire, after piling on desiccated logs. But when the deity had completed everything as required, he cast his sandals away into 127 χαρμόφρων Stephanus xxvi (cf. Hsch. χ.209): χαρμοφέρων MΘ: χάρμα φέρων p  εἰρύσσατο Θ: -ύσατο Mp  130 ἠράσσατο x: -άσατο MDp  132 ἡδεῖ’ Ruhnken (1781) 35: ᾔδει Ω  ὧς Gemoll: ὥς Ω  οἱ ἐπείθετο Ψ: ἐπεπείθετο M   133 περᾶν Barnes: περῆν M: πέρην’ Θ: πέρην p   136 Ψ: om. M   φωρῆς Hermann: φωνῆς Ψ  τὰ δ’,  ἐπὶ Thomas: ἐπὶ δὲ Ψ  κάγκαν’ ἀείρας ΘΓ: κάγκ’ ἀναείρας p  137 οὐλοκάρηνα Ψ: οὐλοκάρηβα M  138 δὴ M (-δὴ): om. Ψ  ἤνυσε Ψ: ηὔλησε M  

103

T H E H O M E R I C H Y M N TO H E R M E S

ἀνθρακιὴν δ’ ἐμάρανε, κόνιν δ’ ἀμάθυνε μέλαιναν παννύχιος, καλὸν δὲ φόως ἐπέλαμπε Σελήνης. Κυλλήνης δ’ αἶψ’ αὖτις ἀφίκετο δῖα κάρηνα ὄρθριος, οὐδέ τίς οἱ δολιχῆς ὁδοῦ ἀντεβόλησεν οὔτε θεῶν μακάρων οὔτε θνητῶν ἀνθρώπων, οὐδὲ κύνες λελάκοντο. Διὸς δ’ ἐριούνιος Ἑρμῆς  δοχμωθεὶς μεγάροιο διὰ κλήϊθρον ἔδυνεν αὔρηι ὀπωρινῆι ἐναλίγκιος, ἠΰτ’ ὀμίχλη. ἰθύσας δ’ ἄντρου ἐξίκετο πίονα νηόν ἦκα ποσὶ προβιβῶν· οὐ γὰρ κτύπεν ὥς περ ἐπ’ οὔδει· ἐσσυμένως δ’ ἄρα λίκνον ἐπώιχετο κύδιμος Ἑρμῆς·  σπάργανον ἀμφ’ ὤμοις εἰλυμένος, ἠΰτε τέκνον νήπιον, ἐν παλάμηισι περὶ γνυσὶ λαῖφος ἀθύρων κεῖτο, χέλυν ἐρατὴν ἐπ’ ἀριστερὰ χειρὸς ἐέργων. μητέρα δ’ οὐκ ἄρ’ ἔληθε θεὰν θεός, εἶπέ τε μῦθον· “τίπτε σύ, ποικιλομῆτα, πόθεν τόδε νυκτὸς ἐν ὥρηι ἔρχε’, ἀναιδείην ἐπιειμένε; νῦν σε μάλ’ οἴω ἦ τάχ’ ἀμήχανα δεσμὰ περὶ πλευρῆισιν ἔχοντα Λητοΐδου ὑπὸ χερσὶ διὲκ προθύροιο περήσειν, ἤ σε φέροντα μέταζε κατ’ ἄγκεα φιλητεύσειν. ἔρρε πάλιν· μεγάλην σε πατὴρ ἐφύτευσε μέριμναν θνητοῖσ’ ἀνθρώποισι καὶ ἀθανάτοισι θεοῖσιν.” τὴν δ’ Ἑρμῆς μύθοισιν ἀμείβετο κερδαλέοισιν· “μῆτερ ἐμή, τί με ταῦτα δεδίσκεαι, ἠΰτε τέκνον νήπιον, ὃς μάλα παῦρα μετὰ φρεσὶν αἴσυλα οἶδεν, ταρβαλέον, καὶ μητρὸς ὑπαιδείδοικεν ἐνιπάς; αὐτὰρ ἐγὼ τέχνης ἐπιβήσομαι ἥ τις ἀρίστη, βουκολέων ἐμὲ καὶ σὲ διαμπερές. οὐδὲ θεοῖσιν νῶϊ μετ’ ἀθανάτοισιν ἀδώρητοι καὶ ἄπαστοι αὐτοῦ τῆιδε μένοντες ἀνεξόμεθ’, ὡς σὺ κελεύεις.

140

145

150

155

160

165

141 παννύχιος Ψ: -ον M   ἐπέλαμπε Ψ: κατέλαμπε M   148 ἰθύσας M: ἰθύνας Ψ  ἄντρου Ψ: -ον M   ἦκα Da: ἧκα Mbp  προβιβῶν At p: -ὼν MΘ  151 εἰλυμένος xp: ἠλ- M: εἱλ- D   152 περὶ γνυσὶ Forssman (1964): περιγνύσι M: περ’ ἰγνύσι Θ: παρ’ ἰγνύσι p  155 τόδε Wolf (1784): τάδε Ω  ἔρχε’ Thomas: ἔρχη Ω  157 ἦ τάχ’ Ψ: δύσαχ’ M  πλευρῆισιν MΘ: -οῖσιν p  158 διὲκ Matthiae (1800) 264: διεκ- MΘ:

104

T H E H O M E R I C H Y M N TO H E R M E S

the deep-whirling Alpheus, (140) made the embers die down, and levelled the black ash into the sand for the rest of the night; and Selene’s fair light shone upon him. Presently he got back to the divine peaks of Kyllene, in the pre-dawn, and no blessed god or mortal man met him during his long journey, (145) nor did dogs bark. Zeus’s beneficent Hermes turned sideways and entered through the hall’s locked door, resembling an autumn breeze, like mist. Then heading straight, he reached the rich sanctum of the cave, advancing with gentle steps (he made no thud, as one does on a floor), (150) and hastily glorious Hermes approached his winnowing-fan. With swaddling wrapped around his shoulders, like a witless infant, he lay toying with the sheet around his knees in his hands, enclosing the lovely tortoise-lyre on his left-hand side. But the god certainly did not escape the notice of his goddess mother, and she spoke out: (155) ‘You of the intricate mind, whyever and from where are you arriving here during the night-time, you who are clothed in shamelessness? Now I strongly suspect that you will soon cross through the entranceway and out, but with unmanageable bindings around your sides, at the hands of Leto’s son, or that in the future you will be a brigand, pillaging in the glades. (160) Be off again: your father sowed you as a huge source of worry for human mortals and immortal gods.’ Hermes replied to her with canny words: ‘Mother dear, why do you try to frighten me so – like a witless infant whose mind knows very few crimes – (165) timid, and he cowers at his mother’s talking-to? I shall embark on whatever skill is best, tending to you and me continually. Nor will we two among the immortal gods endure without gifts or food, just staying here as you urge. δι’ ἐκ p  159 φέροντα μέταζε Càssola (κακὸν τὰ μέταζε Schmitt 1856: 151): φέροντα μεταξὺ M: λαβόντα μεταξὺ Ψ  163 δεδίσκεαι Pierson (1759) 119: τιτύσκεαι Ω  164 παῦρα μετὰ φρεσὶν   αἴσυλα Ψ: πολλὰ ἐνὶ φρεσὶν ἄρμενα M   165 ὑπαιδείδοικεν MΘ Mon: ὑπαὶ δείδ- p  167 βουκολέων Ludwich (1886) 809: βουλεύων Ω  168 ἄπαστοι MDVΓ: ἄλιστοι a: ἄπαστοι ss. λι bP (< Ψ): ἄπλιστοι q  169 ἀνεξόμεθ’ Ψ: ἀεξ- M  

105

T H E H O M E R I C H Y M N TO H E R M E S

βέλτερον ἤματα πάντα μετ’ ἀθανάτοισ’ ὀαρίζειν πλούσιον ἀφνειὸν πολυλήϊον, ἢ κατὰ δῶμα ἄντρωι ἐν ἠερόεντι θαασσέμεν. ἀμφὶ δὲ τιμῆς κἀγὼ τῆς ὁσίης ἐπιβήσομαι ἧς περ Ἀπόλλων, εἰ δέ κε μὴ δώηισι πατὴρ ἐμός, ἤτοι ἐγώ γε πειρήσω (δύναμαι) φιλητέων ὄρχαμος εἶναι. εἰ δέ μ’ ἐρευνήσει Λητοῦς ἐρικυδέος υἱός, ἄλλό τί οἱ καὶ μέζον ὀΐομαι ἀντιβολήσειν· εἶμι γὰρ ἐς Πυθῶνα μέγαν δόμον ἀντιτορήσων, ἔνθεν ἅλις τρίποδας περικαλλέας ἠδὲ λέβητας πορθήσω καὶ χρυσόν, ἅλις τ’ αἴθωνα σίδηρον καὶ πολλὴν ἐσθῆτα. σὺ δ’ ὄψεαι, αἴ κ’ ἐθέληισθα.” ὣς οἳ μέν ῥ’ ἐπέεσσι πρὸς ἀλλήλους ἀγόρευον υἱός τ’ αἰγιόχοιο Διὸς καὶ πότνια Μαῖα. Ἠὼς δ’ ἠριγένεια φόως θνητοῖσι φέρουσα ὤρνυτ’ ἀπ’ Ὠκεανοῖο βαθυρρόου. αὐτὰρ Ἀπόλλων Ὀγχηστόνδ’ ἀφίκανε κιών, πολυήρατον ἄλσος ἁγνὸν ἐρισφαράγου Γαιηόχου, ἔνθα γέροντα κνώδαλον ηὗρε νέμοντα παρὲξ ὁδοῦ, ἐκτὸς ἀλωῆς. τὸν πρότερος προσέφη Λητοῦς ἐρικυδέος υἱός· “ὦ γέρον, Ὀγχηστοῖο βατοδρόπε ποιήεντος,  βοῦς ἀπὸ Πιερίης διζήμενος ἐνθάδ’ ἱκάνω, πάσας θηλείας, πάσας κεράεσσιν ἑλικτάς, ἐξ ἀγέλης. ὁ δὲ ταῦρος ἐβόσκετο μοῦνος ἀπ’ ἄλλων κυάνεος, χαροποὶ δὲ κύνες κατόπισθεν ἕποντο τέσσερες, ἠΰτε φῶτες ὁμόφρονες. οἳ μὲν ἔλειφθεν οἵ τε κύνες ὅ τε ταῦρος, ὃ δὴ περὶ θαῦμα τέτυκται, ταὶ δ’ ἔβαν ἠελίοιο νέον καταδυομένοιο ἐκ μαλακοῦ λειμῶνος, ἀπὸ γλυκεροῖο νομοῖο. ταῦτά μοι εἰπέ, γεραιὲ παλαιγενές, εἴ που ὄπωπας ἀνέρα ταῖσδ’ ἐπὶ βουσὶ διαπρήσσοντα κέλευθον.” τὸν δ’ ὁ γέρων μύθοισιν ἀμειβόμενος προσέειπεν·

170

175

180

185

190

195

200

174 ἐγώ γε Humbert (267 ἐμοί γε etc. Allen 1896): ἔγωγε Ω  175 δύναμαι Chalkondyles, punctuation by Bothe: δύναμαι δὲ Ω  φιλητέων Wolf (1784): φιλητέον M: φιλητεύων Ψ  177 μέζον Thomas: μεῖζον Ω  

106

T H E H O M E R I C H Y M N TO H E R M E S

(170) Better to chat with the immortals for all one’s days, rich, wealthy and with lots of spoils, than to sit at home in a murky cave. In the struggle for status, I too will embark on just the same pious behaviour as Apollo, but if my father should withhold it, then I tell you I (175) shall try (I’m able!) to be the chief of brigands. And if the son of glorious Leto searches for me, I rather suspect another greater business will meet with him: I shall go to Pytho to pierce into his great house, (179) from where I shall plunder plenty of beautiful tripods and cauldrons and gold, plenty of flashing iron and many clothes. You’ll see, if you like.’ So they, lady Maia and the son of aegis-bearing Zeus, conversed with each other. But early-born Dawn, bringing light to mortals, (185) rose from deep-flowing Okeanos. And Apollo went and reached Onchestos, the delightful, venerable grove of the loud-crashing Earth-bearer, where he found an old man (?) pasturing his beast beside the road, outside(?) his plot. The son of glorious Leto addressed him first: (190) ‘Old man, bramble-clearer of grassy Onchestos, I arrive here from Pieria seeking cattle – all female, all with spiralling horns – from my herd. The black bull was grazing apart from the rest, and following behind were four amber-eyed dogs, (195) like men whose hearts are in sympathy. They were left behind – both dogs and bull – which is a truly amazing outcome, but the cows soon after the sun began to set left the soft meadow, left their sweet pasture. Tell me this, old man born long ago – whether you have perchance seen (200) a man progressing along the road in charge of these cows.’ The old man addressed him with a reply: ‘My friend, it’s an effort to list all one happens to see with one’s eyes. Many 181 αἴ κ’ ἐθέληισθα MD: αἴ()κε θελ- ap: αἶκε θέλ- b  183 Μαῖα Ψ: μήτηρ M  188 ηὗρε West: εὗρε Ω  ἐκτὸς Allen and Sikes in commentary: ἕρκος Ω  190 βατοδρόπε Ψ: βατόδροπε M   195 τέσσερες Thomas: -αρες Ω  200 ταῖσδ’ Châteillon (has Velareus 1528): ταῖς δ’ Ω  κέλευθον Ψ: -θα M  

107

T H E H O M E R I C H Y M N TO H E R M E S

“ὦ φίλος, ἀργαλέον μὲν ὅσ’ ὀφθαλμοῖσιν ἴδοιτο πάντα λέγειν. πολλοὶ γὰρ ὁδὸν πρήσσουσιν ὁδῖται, τῶν οἳ μὲν κακὰ πολλὰ μεμαότες, οἳ δὲ μάλ’ ἐσθλά φοιτῶσιν, χαλεπόν δὲ δαήμεναί ἐστιν ἕκαστον. αὐτὰρ ἐγὼ πρόπαν ἦμαρ ἐς ἠέλιον καταδύντα ἔσκαπτον περὶ γουνὸν ἀλωῆς οἰνοπέδοιο, παῖδα δ’ ἔδοξα, φέριστε, σαφὲς δ’ οὐκ οἶδα, νοῆσαι, ὅς τις ὁ παῖς ἅμα βουσὶν ἐϋκραίρηισιν ὀπήδει νήπιος, εἶχε δὲ ῥάβδον, ἐπιστροφάδην δ’ ἐβάδιζεν, ἐξοπίσω δ’ ἀνέεργε, κάρη δ’ ἔχον ἀντίον αὐτῶι.” φῆ ῥ’ ὁ γέρων, ὃ δὲ θάσσον ὁδὸν κίε μῦθον ἀκούσας. οἰωνὸν δ’ ἐνόει τανυσίπτερον, αὐτίκα δ’ ἔγνω Φιλήτην γεγαῶτα, Διὸς παῖδα Κρονίωνος. ἐσσυμένως δ’ ἤϊξεν ἄναξ Διὸς υἱὸς Ἀπόλλων ἐς Πύλον ἠγαθέην, διζήμενος εἰλίποδας βοῦς, πορφυρέηι νεφέληι κεκαλυμμένος εὐρέας ὤμους, ἴχνιά τ’ εἰσενόησεν Ἑκηβόλος, εἶπέ τε μῦθον· “ὦ πόποι, ἦ μέγα θαῦμα τόδ’ ὀφθαλμοῖσιν ὁρῶμαι. ἴχνια μὲν τάδε γ’ ἐστὶ βοῶν ὀρθοκραιράων ἀλλὰ πάλιν τέτραπται ἐς ἀσφοδελὸν λειμῶνα, βήματα δ’ οὔτ’ ἀνδρὸς τάδε γίνεται οὔτε γυναικός οὔτε λύκων πολιῶν οὔτ’ ἄρκτων οὔτε λεόντων, οὔτέ τι κενταύρου λασιαύχενος ἔλπομαι εἶναι, ὅς τις τοῖα πέλωρα βιβᾶι ποσὶ καρπαλίμοισιν. αἰνὰ μὲν ἔνθεν ὁδοῖο, τὰ δ’ αἰνότερ’ ἔνθεν ὁδοῖο.” ὣς εἰπὼν ἤϊξεν ἄναξ Διὸς υἱὸς Ἀπόλλων, Κυλλήνης δ’ ἀφίκανεν ὄρος καταειμένον ὕληι πέτρης ἐς κευθμῶνα βαθύσκιον, ἔνθά τε νύμφη ἀμβροσίη ἐλόχευσε Διὸς παῖδα Κρονίωνος. ὀδμὴ δ’ ἱμερόεσσα δι’ οὔρεος ἠγαθέοιο κίδνατο, πολλὰ δὲ μῆλα ταναύποδα βόσκετο ποίην.

205

210

215

220

225

230

202 ἴδοιτο Ψ: ἴδοιμι M   205 φοιτῶσιν Ψ: πρήσσουσιν M   207 ἀλωῆς Chalkondyles: ἁλ- Ω  208 νοῆσαι Ψ: νοήσας M   209 ἐϋκραίρηισιν Θ: -ῆσιν M: -οισιν p  ὀπήδει β1: ὁπ- Ω  211 ἔχον Hermann: ἔχεν Ω  212 φῆ ῥ’ MΘ (φή ῥ’ x): φῆ δ’ p  θάσσον Thomas (2010a): θᾶσσον Ω  μῦθον ἀκούσας Mxm: φοῖβος ἀπόλλων Ψ  214 Φιλήτην Thomas

108

T H E H O M E R I C H Y M N TO H E R M E S

wayfarers journey on the road, of whom some go to and fro full of evil intentions, others with very good ones, (205) and it is difficult to learn each man’s type. However, I was digging all through the slope of my vineyard plot for the entire day until sunset, and I thought, good sir (but I don’t know for sure), I noticed some child who was accompanying cows with fine horns – (210) an infant, but he was holding a staff, walked now on one side of them and now on the other, and rounded them up backwards, and they held their heads facing him.’ The old man spoke, and the other went on his way more quickly after hearing the speech. He observed a bird of omen with wings stretched wide, and immediately he knew that the Brigand had been born, the son of Zeus son of Kronos. (215) Hastily lord Apollo son of Zeus darted to holy Pylos, seeking his shambling cattle, his broad shoulders cloaked in a purple cloud. And the Far-shooter noticed the tracks and broke into speech: ‘Oh dear! My, this is a great wonder on which my eyes gaze. (220) These are the tracks of the cattle with upright horns, but they are turned back towards the asphodel meadow; but these are the steps of neither man nor woman, nor grey wolves, nor bears, nor lions, nor do I expect for a moment that they belong to a shaggy-necked centaur, (225) who makes such monstrous steps with his swift feet. Infuriating things on one side of the road, and more infuriating yet on the other.’ So lord Apollo the son of Zeus spoke, and darted away and arrived at Mount Kyllene, cloaked in woods, to the deep-shadowed hideaway in the rockface, where the ambrosial nymph (230) had given birth to the son of Zeus son of Kronos. A delightful smell was spreading over the holy mountain, and many sheep on tapering legs were browsing on the grass. There (φιλήτην Ilgen in errata): φιλητὴν MDb: φιλοτὴν a: φηλωτὴν p  216 εἰλίποδας M: εἱλ- Ψ  218–19 Ψ: om. M   224 ἔλπομαι εἶναι Mabm: ἔστιν ὁμοῖα Dp: ἤστιν (-ην L) ὁμοῖα b  228 καταειμένον Πp: καταείμενον MΘ  230 Κρονίωνος Ψ: κρονίωνα M   232 ταναύποδα Θ: τανύ- Mp  

109

T H E H O M E R I C H Y M N TO H E R M E S

ἔνθα τότε σπεύδων κατεβήσετο λάϊνον οὐδόν ἄντρον ἐς ἠερόεν ἑκατηβόλος αὐτὸς Ἀπόλλων. τὸν δ’ ὡς οὖν ἐνόησε Διὸς καὶ Μαιάδος υἱός235 χωόμενον περὶ βουσὶν ἑκηβόλον Ἀπόλλωνα, σπάργαν’ ἔσω κατέδυνε θυήεντ’. ἠΰτε πολλήν πρέμνων ἀνθρακιὴν ὕλης σποδὸς ἀμφικαλύπτει, ὣς Ἑρμῆς Ἑκάεργον ἰδὼν ἀνεείλε’ ἓ αὐτόν, ἐν δ’ ὀλίγωι συνέλασσε κάρη χεῖράς τε πόδας τε,240 φή ῥα νεόλλουτος προκαλεόμενος ἥδυμον ὕπνον, ἐγρήσσων ἐτεόν γε, χέλυν δ’ ὑπὸ μασχάληι εἶχεν. γνῶ δ’ οὐδ’ ἠγνοίησε Διὸς καὶ Λητοῦς υἱός νύμφην τ’ οὐρείην περικαλλέα καὶ φίλον υἱόν, 245 παῖδ’ ὀλίγον δολίηις εἰλυμένον ἐντροπίηισιν. παπτήνας δ’ ἀνὰ πάντα μυχὸν μεγάλοιο δόμοιο τρεῖς ἀδύτους ἀνέωιγε, λαβὼν κληῗδα φαεινήν, νέκταρος ἐκπλείους ἠδ’ ἀμβροσίης ἐρατεινῆς, πολλὸς δὲ χρυσός τε καὶ ἄργυρος ἔνδον ἔκειτο, 250 πολλὰ δὲ φοινικόεντα καὶ ἄργυφα εἵματα νύμφης,  οἷα θεῶν μακάρων ἱεροὶ δόμοι ἐντὸς ἔχουσιν. ἔνθ’ ἐπεὶ ἐξερέεινε μυχοὺς μεγάλοιο δόμοιο Λητοΐδης μύθοισι προσηύδα κύδιμον Ἑρμῆν· “ὦ παῖ, ὃς ἐν λίκνωι κατάκειαι, μήνυέ μοι βοῦς θάσσον, ἐπεὶ τάχα νῶϊ διοισόμεθ’ οὐ κατὰ κόσμον.  255 ῥίψω γάρ σε λαβὼν ἐς Τάρταρον ἠερόεντα, ἐς ζόφον αἰνόμορον καὶ ἀμήχανον, οὐδέ σε μήτηρ ἐς φάος οὐδὲ πατὴρ ἀναλύσεται, ἀλλ’ ὑπὸ γαίηι ἐρρήσεις, ὀλοοῖσι μετ’ ἀνδράσιν ἡγεμονεύων.” 260 τὸν δ’ Ἑρμῆς μύθοισιν ἀμείβετο κερδαλέοισιν· “Λητοΐδη, τίνα τοῦτον ἀπηνέα μῦθον ἔειπες; καὶ βοῦς ἀγραύλους διζήμενος ἐνθάδ’ ἱκάνεις; 238 ὕλης σποδὸς Ψ: ὁλοσποδὸς M   ἀμφικαλύπτει D: -οι Mxp  239 ἀνεείλε’ ἓ   αὐτόν Postgate (ἀνεείλε’) cited in Allen (1896) + Hermann (ἓ αὐτόν): ἀλέεινεν ἑαυτὸν Ω: or ἀνεείλεεν αὐτόν?  241 φή ῥα νεόλλουτος Hermann: δή ῥα νεόλλουτος Ω: ἐν ἄλλω οὕτως· θῆρα νέον λοχάων προκαλεύμενος ἡδύ xm  προκαλεόμενος West: -ούμενος M: -εύμενος Ψ  242 ἐγρήσσων ἐτεόν γε, χέλυν δ’ Hermann (ἐγρήσσων, ἐρατὴν

110

T H E H O M E R I C H Y M N TO H E R M E S

great Apollo the far-shooter then hastened and descended over the stone threshold into the murky cave. (235) Now when the son of Zeus and Maia noticed far-shooting Apollo in anger about his cattle, he sank into his scented swaddling. As wood-ash covers over a large pile of coals formed from thick logs, so Hermes, when he saw the Far-worker, wrapped himself up (240) and drew into a small compass his head and hands and feet, like one newly-bathed summoning sweet sleep. But really he was awake, and held the tortoise-lyre under his armpit. But the son of Zeus and Leto recognized and did not fail to recognize the beautiful mountain nymph and her beloved son, (245) the small child wrapped in deceptive entwinings. He glanced around every recess of the great house, took a shining key, and opened up three treasuries overflowing with nectar and lovely ambrosia; and much gold and silver lay inside, (250) and the nymph’s many red-dyed or bright white clothes – such things as the sacred houses of the blessed gods contain within them. Then when he had made a thorough enquiry in the recesses of the great house, the son of Leto addressed glorious Hermes sternly: ‘Boy, you lying in the winnowing-fan: give me information about my cows, (255) and quickly – since soon we will differ, against the order of things. I will seize you and fling you into murky Tartarus, into the gloom of painful fates where no stratagem brings escape. Nor will your mother nor your father have you released up into the light. Instead, beneath the Earth you will go to ruin, being a leader only among destructive men.’ (260) Hermes replied to him with canny words: ‘Son of Leto, what is this harsh speech you uttered? Have you really come here looking for your cows, which live in the fields? I didn’t see; I τε χέλυν Martin (1605) f. 21): ἄγρης· εἰνέτεόν (εἰνετ- M) τε χέλυν Ω  246 ἀνὰ M: ἄρα Ψ  248 ἐκπλείους Ψ: ἐμ- M   249 τε Ψ: γε M   254 λίκνωι Mabmp: κλίνη Db  κατάκειαι MΘN: -κηαι p  255 θάσσον, ἐπεὶ Ilgen (θᾶσ-): M: θᾶττον ἐπεὶ Ψ  256 λαβὼν Ilgen (arreptum Velareus 1528): βαλὼν Ω  259 ὀλοοῖσι μετ’ Schneidewin (1848) 677 (ὀλοοῖσιν ἐν Bothe in notes): ὀλίγοισι μετ’ M: ὀλίγοισιν ἐν Ψ  261 ἔειπες D: -ας Mxp  262 punctuation as independent question Stephanus  

111

T H E H O M E R I C H Y M N TO H E R M E S

{

οὐκ ἴδον, οὐ πυθόμην, οὐκ ἄλλου μῦθον ἄκουσα. οὐκ ἂν μηνύσαιμ’· οὐκ ἂν μήνυτρον ἀροίμην, 265 οὔτε βοῶν ἐλατῆρι, κραταιῶι φωτί, ἔοικα. οὐκ ἐμὸν ἔργον τοῦτο, πάρος δέ μοι ἄλλα μέμηλεν· ὕπνος ἐμοί γε μέμηλε καὶ ἡμετέρης γάλα μητρός, σπάργανά τ’ ἀμφ’ ὤμοισιν ἔχειν καὶ θερμὰ λοετρά. μή τις τοῦτο πύθοιτο, πόθεν τόδε νεῖκος ἐτύχθη· καί κεν δὴ μέγα θαῦμα μετ’ ἀθανάτοισι γένοιτο270 παῖδα νέον γεγαῶτα διὰ προθύροιο περῆσαι βουσὶ μετ’ ἀγραύλοισι. τὸ δ’ ἀπρεπέως ἀγορεύεις· χθὲς γενόμην, ἁπαλοὶ δὲ πόδες, τρηχεῖα δ’ ὑπὸ χθών. εἰ δ’ ἐθέλεις, πατρὸς κεφαλὴν μέγαν ὅρκον ὀμέομαι· μὴ μὲν ἐγὼ μήτ’ αὐτὸς ὑπίσχομαι αἴτιος εἶναι275 μήτέ τιν’ ἄλλον ὄπωπα βοῶν κλοπὸν ὑμετεράων (αἵ τινες αἱ βόες εἰσί), τὸ δὲ κλέος οἶον ἀκούω.” ὣς ἄρ’ ἔφη, καὶ πυκνὸν ἀπὸ βλεφάρων ἀμαρύσσων ὀφρύσι ῥιπτάζεσκεν ὁρώμενος ἔνθα καὶ ἔνθα, μάκρ’ ἀποσυρίζων, ἅλιον ὡς μῦθον ἀκούων.280 τὸν δ’ ἁπαλὸν γελάσας προσέφη ἑκάεργος Ἀπόλλων· “ὦ πέπον ἠπεροπευτὰ δολοφραδές, ἦ σε μάλ’ οἴω πολλάκις ἀντιτορέοντα δόμους εὖ ναιετάοντας ἔννυχον οὔ χ’ ἕνα μοῦνον ἐπ’ οὔδει φῶτα καθέσσαι, 285 σκευάζοντα κατ’ οἶκον ἄτερ ψόφου, οἷ’ ἀγορεύεις. πολλοὺς δ’ ἀγραύλους ἀκαχήσεις μηλοβοτῆρας οὔρεος ἐν βήσσηισ’, ὁπότ’ ἂν κρειῶν ἐρατίζων 288(a) ἀντήσεις ἀγέληισι βοῶν καὶ πώεσι μήλων.  ἀντᾶις βουκολίοισι καὶ εἰροπόκοισ’ ὀΐεσσιν.288(b) ἀλλ’ ἄγε, μὴ πύματόν τε καὶ ὕστατον ὕπνον ἰαύσεις, ἐκ λίκνου κατάβαινε, μελαίνης νυκτὸς ἑταῖρε.290 τοῦτο γὰρ οὖν καὶ ἔπειτα μετ’ ἀθανάτοις γέρας ἕξεις·

265 κραταιῶι Ψ: κρατερῶ M   272 ἀγραύλοισι Ψ: -ησι M   273 ὑπὸ Ma: ὑπο- Dbp  274 δ’ ἐθέλεις Ilgen (cf. Wolf 1795: ccxlvi n. 32): δὲ θέλεις Ω  277 οἶον Mx: οἷον Dp  279 ῥιπτάζεσκεν Ψ: ῥιπάζ- M  280 ἅλιον MP: ἄλιον Θq  ὡς M: τὸν Da: τὸν ὡς Π: ὡς ss. τὸν L (= Ψ): ὡς τὸν p  284 οὔ χ’ Tucker in Allen 1912: οὐδ’ M: οὐχ Ψ (οὐχ’ LQ , ου χ’ A)  

112

T H E H O M E R I C H Y M N TO H E R M E S

didn’t overhear; I didn’t hear the story from another. I couldn’t give information – I couldn’t earn an informer’s reward. (265) Nor do I resemble a cattle-rustler, a strong man. This is not my kind of deed. So far I have been concerned with other things: sleep is my concern, and my mother’s milk, and having swaddling around my shoulders, and warm bathwater. I hope nobody finds out the source of this dispute: (270) to be sure, it would be a great source of amazement among the immortals if a newborn child crossed through his doorway in search of cows which live in the fields. It is unsuitable for you to declare as much: I was born yesterday; my feet are soft, the ground underneath rough. If you want, I shall swear a great oath on my father’s head: (275) neither, I say, do I profess to be guilty myself, nor have I seen any other thief of your cows (whatever these cows are), but have only heard the report.’ So he spoke, and glinted a quick series of glances from his eyelids, and from round his eyes he cast them in a whirl as he looked this way and that, (280) whistling out loudly as if the speech he had heard was empty. Breaking into a gentle laugh, far-working Apollo addressed him: ‘My dear tricky-minded deceiver, I strongly suspect you may on many occasions bore into well-appointed homes by night, and leave not just one man sat on the ground (285) as you noiselessly collect up the possessions in his household, to judge from your words. And many are the shepherds who live in the fields that you will reduce to grief in mountain glens, when in your lust for meat you come upon herds of cattle and flocks of sheep and goats.1 But come – in case you lie down for your last and final sleep, (290) step down out of your winnowing-fan, comrade of dark night. This is a prize you will have among the

καθέσσαι Van Herwerden (1907) 186: καθίσαι MD: καθίσσαι xp  287 κρειῶν Ψ: μήλων M   288(a) Ω  ἀντήσεις DaΓC: -ης Mbp  288(b) xm  ἀντᾶις Schneidewin (1848) 679: ἄντην xm   289 ἀλλ’ ἄγε Ψ: ἀλλά γε M  μὴ πύματόν Ψ: πήματόν M   ἰαύσεις M: -ης Ψ   1

Alternative: ‘you search out herds of cows and wool-fleeced sheep.’

113

T H E H O M E R I C H Y M N TO H E R M E S

ἀρχὸς φιλητέων κεκλήσεαι ἤματα πάντα.” ὣς ἄρ’ ἔφη, καὶ παῖδα λαβὼν φέρε Φοῖβος Ἀπόλλων. σὺν δ’ ἄρα φρασσάμενος, τότε δὴ κρατὺς Ἀργεϊφόντης 295 οἰωνὸν προέηκεν ἀειρόμενος μετὰ χερσίν,  τλήμονα γαστρὸς ἔριθον, ἀτάσθαλον ἀγγελιώτην, ἐσσυμένως δὲ μετ’ αὐτὸν ἐπέπταρε. τοῖο δ’ Ἀπόλλων ἔκλυεν, ἐκ χειρῶν δὲ χαμαὶ βάλε κύδιμον Ἑρμῆν, ἕζετο δὲ προπάροιθε, καὶ ἐσσύμενός περ ὁδοῖο, Ἑρμῆν κερτομέων, καί μιν πρὸς μῦθον ἔειπεν·300 “θάρσει, σπαργανιῶτα, Διὸς καὶ Μαιάδος υἱέ· εὑρήσω καὶ ἔπειτα βοῶν ἴφθιμα κάρηνα τούτοισ’ οἰωνοῖσι, σὺ δ’ αὖθ’ ὁδὸν ἡγεμονεύσεις.” ὣς φάθ’, ὃ δ’ αὖτ’ ἀνόρουσε θοῶς Κυλλήνιος Ἑρμῆς σπουδῆι ἰών. ἄμφω δὲ παρ’ οὔατα χερσὶν ἐώθει305 σπάργανον ἀμφ’ ὤμοισιν ἐελμένος, εἶπε δὲ μῦθον· “πῆι με φέρεις, Ἑκάεργε, θεῶν ζαμενέστατε πάντων; ἦ με βοῶν ἕνεχ’ ὧδε χολούμενος ὀρσολοπεύεις; ὦ πόποι, εἴθ’ ἀπόλοιτο βοῶν γένος· οὐ γὰρ ἐγώ γε ὑμετέρας ἔκλεψα βόας οὐδ’ ἄλλον ὄπωπα310 (αἵ τινές εἰσι βόες), τὸ δὲ δὴ κλέος οἶον ἀκούω. δὸς δὲ δίκην καὶ δέξο παρὰ Ζηνὶ Κρονίωνι.” αὐτὰρ ἐπεὶ τὰ ἕκαστα διαρρήδην ἐρέεινον Ἑρμῆς τ’ οἰοπόλος καὶ Λητοῦς ἀγλαὸς υἱός ἀμφὶς θυμὸν ἔχοντες – ὃ μὲν νημερτέα φωνέων315 οὐκ ἀδίκως ἐπὶ βουσὶν ἐλάζετο κύδιμον Ἑρμῆν, αὐτὰρ ὃ τέχνηισίν τε καὶ αἱμυλίοισι λόγοισιν ἤθελεν ἐξαπατᾶν Κυλλήνιος Ἀργυρότοξον… αὐτὰρ ἐπεὶ πολύμητις ἐὼν πολυμήχανον ηὗρεν ἐσσυμένως δἤπειτα διὰ ψαμάθοιο βάδιζεν320 πρόσθεν, ἀτὰρ κατόπισθε Διὸς καὶ Λητοῦς υἱός. 292 ἀρχὸς Ψ: αὖχος M   294 φρασσάμενος Chalkondyles: φρασάΩ  303 τούτοισ’ Ψ (without elision): αὐτοῖς M   οἰωνοῖσι, σὺ Mp: οἰωνοῖσιν εὖ Θ  304 φάθ’ anon. (1567): ἔφαθ’ M: φάτ’ Ψ  306 ἐελμένος M: ἐλιγμένος Θ: ἑλιγμένος p: ἐελμένον Schneidewin (1848) 679   308 ἦ με Θ: ἧμε M: ἥ με p  ἕνεχ’ ὧδε Ψ: ἐνέχωνδε M   311 οἶον b: οἷον MDap  312

114

T H E H O M E R I C H Y M N TO H E R M E S

immortals in the future – all your days you will be addressed as “commander of robbers”.’ So he spoke, and Phoebus Apollo took the child and began to carry him. But after consideration, at that moment the strong slayer of Argus (295) as he was lifted between his hands sent forth a bird of omen – the stubborn dogsbody of the belly, an insolent message-man – and after it he gave an energetic sneeze of confirmation. Apollo heard him, and let glorious Hermes fall to the ground from his hands. Though eager for the journey he sat down in front of him, (300) teasing Hermes, and addressed this speech to him: ‘Have confidence, my swaddling-man, son of Zeus and Maia: I will find my strong and valued cows in the future, thanks to these auspices, and you for your part will lead the way.’ So he spoke, and for his part Hermes of Kyllene sprang up swiftly (305) and went in haste. With his hands he pushed past both ears the swaddling which was wrapped around his shoulders, and spoke out: ‘Where are you taking me, Far-worker, most vehement of all the gods? Are you snapping at me in anger like this because of cows? Oh dear – may the species of cows perish! For I didn’t (310) steal your cows nor have I seen another do so (whatever these cows may be), but I tell you I have only heard the report. Agree to arbitration before Zeus son of Kronos.’ When Hermes the shepherd and the treasured son of Leto had explored each point in detail (315) with diverging aims – for the one was speaking unerringly, and not unjustly trying to seize glor­ ious Hermes on a charge about the cows, whereas the Kyllenian wanted to deceive the Silver-bowed with crafty acts and coaxing words… When despite his many wiles he found him possessed of many stratagems, (320) then he began to step eagerly across the sand, out in front, with the son of Zeus and Leto behind.

δέξο παρὰ MΘ: δέξαι πὰρ p  313 ἐπεὶ τὰ Ψ: ἔπειτα M   ἐρέεινον p: -εν MΘ   315 φωνέων Wolf (1807) (-ῶν): φωνὴν Ω  316 ἐλάζετο Thomas: -οιτο M: -υτο Ψ  317 αἱμυλίοισι xp: αἰμ- MD   320 δἤπειτα Càssola: δ’ ἤπ- Ω  

115

T H E H O M E R I C H Y M N TO H E R M E S

αἶψα δὲ τέρθρον ἵκοντο θυώδεος Οὐλύμποιο ἐς πατέρα Κρονίωνα Διὸς περικαλλέα τέκνα, κεῖθι γὰρ ἀμφοτέροισι Δίκης κατέκειτο τάλαντα. εὐκολίη δ’ ἔχ’ Ὄλυμπον ἀγάννιφον, ἀθάνατοι δέ325 ἄφθιτοι ἠγερέθοντο μετὰ χρυσόθρονον ἠῶ. ἔστησαν δ’ Ἑρμῆς τε καὶ ἀργυροτόξος Ἀπόλλων πρόσθε Διὸς γούνων. ὃ δ’ ἀνείρετο φαίδιμον υἱόν Ζεὺς ὑψιβρεμέτης καί μιν πρὸς μῦθον ἔειπεν· “Φοῖβε, πόθεν ταύτην μενοεικέα ληΐδ’ ἐλαύνεις,330 παῖδα νέον γεγαῶτα φυὴν κήρυκος ἔχοντα; σπουδαῖον τόδε χρῆμα θεῶν μεθ’ ὁμήγυριν ἦλθεν.” τὸν δ’ αὖτε προσέειπεν ἄναξ ἑκάεργος Ἀπόλλων· “ὦ πάτερ, ἦ τάχα μῦθον ἀκούσεαι οὐκ ἀλαπαδνόν, κερτομέων ὡς οἶος ἐγὼ φιλολήϊός εἰμι·335 παῖδά τιν’ ηὗρον τόνδε διαπρύσιον κεραϊστήν Κυλλήνης ἐν ὄρεσσι, πολὺν διὰ χῶρον ἀνύσσας, κέρτομον οἷον ἐγώ γε θεῶν οὐκ ἄλλον ὄπωπα οὐδ’ ἀνδρῶν, ὁπόσοι λησίμβροτοί εἰσ’ ἐπὶ γαῖαν. κλέψας δ’ ἐκ λειμῶνος ἐμὰς βοῦς ὤιχετ’ ἐλαύνων340 ἑσπέριος παρὰ θῖνα πολυφλοίσβοιο θαλάσσης, εὐθὺ Πύλονδ’ ἐλάων. τὰ δ’ ἄρ’ ἴχνια δοιά, πέλωρα, οἷά τ’ ἀγάσσασθαι καὶ ἀγαυοῦ δαίμονος ἔργα· τῆισιν μὲν γὰρ βουσὶν ἐς ἀσφοδελὸν λειμῶνα ἀντία βήματ’ ἔχουσα κόνις ἀνέφαινε μέλαινα,345 αὐτὸς δ’ †οὗτος ὅδ’ ἐκτὸς† ἀμήχανος οὔτ’ ἄρα ποσσίν οὔτ’ ἄρα χερσὶν ἔβαινε διὰ ψαμαθώδεα χῶρον, ἀλλ’ ἄλλην τινὰ μῆτιν ἔχων διέτριβε κέλευθα τοῖα πέλωρ’ ὡς εἴ τις ἁραιῆισι δρυσὶ βαίνοι. 322 δὲ τέρθρον ἵκοντο MDb: δ’ ἵκοντο κάρηνα abmp  324 Δίκης Thomas: δίκης Ω, editors.   325 εὐκολίη Thomas: εὐμιλίη M: εὐμυλίη Ψ  326 ἄφθιτοι Ω: ἀθρόοι Groddeck (1786) 89   μετὰ χρυσόθρονον  ἠῶ abm: ποτὶ πτύχας Οὐλύμποιο MDbp  335 οἶος MDa: οἷος bp   337 ἀνύσσας xp: ἀνύσας MD  338 κέρτομον Ψ: τέρτομον M  339 λησίμβροτοί εἰσ’ E, Aldine (1504): -οι εἴσ’ M: -οι εἶσ’ Ψ  γαῖαν M: γαίη Ψ  342 εὐθὺ Πύλονδ’ Clarke in notes: εὐθύπυλονδ’ M: εὐθυπόρον()δ’ Ψ  ἐλάων MDa: ἑλ- bp  

116

T H E H O M E R I C H Y M N TO H E R M E S

Presently the handsome children of Zeus reached the threshold(?) of scented Olympus, into the presence of their father, Kronos’ son. For there it was that Justice’s balance lay for both. (325) Ease occupied snow-covered Olympus, and the unperish­ ing immortals had gathered after gold-throned dawn. Both Hermes and Apollo of the silver bow took their stand before Zeus’s knees. For his part, Zeus who thunders on high questioned his illustrious son and addressed this speech to him: (330) ‘Phoebus, where are you driving this heart-warming booty from – a new-born child with the natural body of a herald? This is a serious item which has come before the assembly of the gods!’ In turn, the far-working lord Apollo addressed him: ‘Father, soon, I tell you, you will hear a tale which is not robbed of force, (335) though now you tease me that I am the only fan of booty: I found a child, this wide-ranging pillager, in the peaks of Kyllene after I’d got through a great tract of ground – a teaser whose like I at least have not seen in any other of the gods or of all the men there are across the Earth who elude mortals. (340) He stole my cows from the meadow, and went driving them at evening along the strand of the roaring sea, driving them directly to Pylos. As for the tracks, they were of two kinds, monstrous, of a kind to make one marvel, and the works of some marvellous deity: in the case of the cows, (345) the dark sand containing their steps showed them reversed towards the asphodel meadow, while this unmanageable guide(?) himself walked across the sandy ground neither on his feet nor on his hands, but with some other stratagem he pressed down the whole path to look as monstrous as if someone were walking on slender-oak(?). 343 ἀγάσσασθαι Ilgen: ἀγάσασθαι M: ἀγάσσεσθαι Ψ  344 τῆισιν Ψ: τοῖσι M   346 οὗτος ὅδ’ ἐκτὸς Ω: οὗτος + ὁδηγὸς Stoll (1861) 9, ὁδαῖος Ludwich (1887) 700, ὁδουρὸς or ἄλεκτος Thomas (2010a): αὖθ’ ὁδοῦ ἐκτὸς Shackle (1915)   349 ἁραιῆισι P: ἀρ- Ω  βαίνοι Ψ: βαίνων M  

117

T H E H O M E R I C H Y M N TO H E R M E S

{

ὄφρα μὲν οὖν ἐδίωκε διὰ ψαμαθώδεα χῶρον,350 ῥεῖα μάλ’ ἴχνια πάντα διέπρεπεν ἐν κονίηισιν, αὐτὰρ ἐπεὶ ψαμάθοιο πολὺν στίβον ἐξεπέρησεν, ἄφραστος γένετ’ ὦκα βοῶν στίβος ἠδὲ καὶ αὐτοῦ χῶρον ἀνὰ κρατερόν. τὸν δ’ ἐφράσατο βροτὸς ἀνήρ ἐς Πύλον εὐθὺς ἐλῶντα βοῶν γένος εὐρυμετώπων.355 αὐτὰρ ἐπεὶ δὴ τὰς μὲν ἐν ἡσυχίηι κατέερξεν καὶ διαπυρπαλάμησεν ὁδοῦ τὸ μὲν ἔνθα τὸ δ’ ἔνθα, ἐν λίκνωι κατέκειτο μελαίνηι νυκτὶ ἐοικώς ἄντρωι ἐν ἠερόεντι κατὰ ζόφον, οὐδέ κεν αὐτόν αἰετὸς ὀξὺ λάων ἐσκέψατο. πολλὰ δὲ χερσίν360 αὐγὰς ὠμόργαζε δολοφροσύνην ἀλεγύνων, αὐτὸς δ’ αὐτίκα μῦθον ἀπηλεγέως ἀγόρευεν· ‘οὐκ ἴδον, οὐ πυθόμην, οὐκ ἄλλου μῦθον ἄκουσα, οὐδέ κε μηνύσαιμ’, οὐδ’ ἂν μήνυτρον ἀροίμην.’” ἤτοι ὅ γ’ ὣς εἰπὼν κατ’ ἄρ’ ἕζετο Φοῖβος Ἀπόλλων.365 Ἑρμῆς δ’ αὖθ’ ἑτέρωθεν ἀμειβόμενος ἔπος ηὔδα,366(a) 366(b) Ἑρμῆς δ’ ἄλλον μῦθον ἐν ἀθανάτοισιν ἔειπεν,  δείξατο δ’ ἐς Κρονίωνα θεῶν σημάντορα πάντων· “Ζεῦ πάτερ, ἤτοι ἐγώ σοι ἀληθείην ἀγορεύσω· νημερτής τε γάρ εἰμι καὶ οὐκ οἶδα ψεύδεσθαι. ἦλθεν ἐς ἡμετέρου διζήμενος εἰλίποδας βοῦς  370 σήμερον, ἠελίοιο νέον ἐπιτελλομένοιο, οὐδὲ θεῶν μακάρων ἄγε μάρτυρας οὐδὲ κατόπτας, μηνύειν δ’ ἐκέλευεν ἀναγκαίης ὕπο πολλῆς, πολλὰ δέ μ’ ἠπείλησε βαλεῖν ἐς Τάρταρον εὐρύν, οὕνεχ’ ὃ μὲν τέρεν ἄνθος ἔχει φιλοκυδέος ἥβης375 αὐτὰρ ἐγὼ χθιζὸς γενόμην· τὰ δέ τ’ οἶδε καὶ αὐτός· οὔ τι βοῶν ἐλατῆρι, κραταιῶι φωτί, ἐοικώς. πείθεο· καὶ γὰρ ἐμεῖο πατὴρ φίλος εὔχεαι εἶναι· ὡς οὐκ οἴκαδ’ ἔλασσα βόας (ὣς ὄλβιος εἴην) οὐδ’ ὑπὲρ οὐδὸν ἔβην, τόδε δ’ ἀτρεκέως ἀγορεύω,  380

352 πολὺν M: μέγαν Ψ  356 κατέερξεν p (-ε): -έρεξεν M: -έρεξε Θ  357 διαπυρπαλάμησεν Ilgen: διαπῦρ παλάμησεν M: διὰ πῦρ μάλ’ ἄμησεν Ψ  361 ὠμόργαζε Ilgen (  fricabat Châteillon): ὠμάρταζε Ω (ὡμ- Lp)  ἀλεγύνων

118

T H E H O M E R I C H Y M N TO H E R M E S

(350) ‘Well, as long as he hurried across the sandy ground, all the tracks were perfectly easy and clear to see in the powder, but when he had completely traversed his huge trail in the sand, the cows’ trail and his own over firm ground quickly became unobservable. However, a mortal man observed him (355) driving the stock of broad-fronted cows directly to Pylos. And when he had confined them where they wouldn’t be disturbed, and had completed both parts of his quick, scheming journey, he lay down in his winnowing-fan like dark night, in a murky cave facing the Western gloom. Not even (360) a sharp-eyed eagle would have spotted him. He rubbed his eyes a lot with his hands, cooking up trickery, and in person he immediately spoke without circumspection: “I didn’t see; I didn’t overhear; I didn’t hear the story from another, nor could I give information, nor earn an informer’s reward.’” (365) So Phoebus Apollo spoke, and sat down. And on the other side Hermes in turn spoke in response,2 and addressed himself to the son of Kronos, the director of all the gods: ‘Father Zeus, I shall tell you the truth, for I am unerring and do not know how to lie. (370) He came to our house today seeking his shambling cows, soon after the sun began to rise, and bringing neither witnesses nor observers from among the blessed gods. He ordered me to give information under much duress, and much did he threaten to throw me into broad Tartarus, (375) all because he has the soft flower of ambitious youth, whereas I was born yesterday (he himself knows this) – and in no way do I resemble a cattle-rustler, a strong man. Believe me – for you claim to be my loving father too – that, as surely as I wish to prosper, I did not drive the cows home (380) nor did I step over our threshold, that I am talking straight, that I greatly Θ: ἀλεγίζων M: ἀλεείνων p  365 ὅ γ’ Barnes: ἂρ Ω  366(a) MDbp  366(b) abm  368 ἀγορεύσω M: καταλέξω Ψ  373 ὕπο Wolf (1807): ὑπὸ Ω  376 τὰ δέ τ’ M: τάδε τ’ Θ: τά δέ τ’ p  379 ὣς M: ὡς Ψ   Alternative: ‘And Hermes spoke a different account among the immortals.’

2

119

T H E H O M E R I C H Y M N TO H E R M E S

Ἠέλιον δὲ μάλ’ αἰδέομαι καὶ δαίμονας ἄλλους, καὶ σὲ φιλέω, καὶ τοῦτον ὀπίζομαι. οἶσθα καὶ αὐτός ὡς οὐκ αἴτιός εἰμι· μέγαν δ’ ἐπιδέξομαι ὅρκον· οὐ μὰ τάδ’ ἀθανάτων εὐκόσμητα προθύραια. καί ποτ’ ἐγὼ τούτωι τείσω, ποτὶ νηλέα φώρην,385 καὶ κρατερῶι περ ἐόντι, σὺ δ’ ὁπλοτέροισιν ἄρηγε.” ὣς φάτ’ ἐπιλλίζων Κυλλήνιος Ἀργεϊφόντης, καὶ τὸ σπάργανον εἶχεν ἐπ’ ὠλένηι οὐδ’ ἀπέβαλλεν. Ζεὺς δὲ μέγ’ ἐξεγέλασσεν ἰδὼν κακομηδέα παῖδα εὖ καὶ ἐπισταμένως ἀρνεόμενον ἀμφὶ βόεσσιν,390 ἀμφοτέρους δ’ ἐκέλευσεν ὁμόφρονα θυμὸν ἔχοντας ζητεύειν, Ἑρμῆν δὲ διάκτορον ἡγεμονεύειν καὶ δεῖξαι τὸν χῶρον ἐπ’ ἀβλαβίηισι νόοιο, ὅππηι δαὖτ’ ἀπέκρυψε βοῶν ἴφθιμα κάρηνα. νεῦσεν δὲ Κρονίδης, ἐπεπείθετο δ’ ἀγλαὸς Ἑρμῆς.395 ῥηϊδίως γὰρ ἔπειθε Διὸς νόος αἰγιόχοιο. τὼ δ’ ἄμφω σπεύδοντε Διὸς περικαλλέα τέκνα ἐς Πύλον ἠμαθόεντα ἐπ’ Ἀλφειοῦ πόρον ἷξον, ἀγροὺς δ’ ἐξίκοντο καὶ αὔλιον ὑψιμέλαθρον ἡχοῦ δὴ τὰ χρήματ’ ἀτάλλετο νυκτὸς ἐν ὥρηι.400 ἔνθ’ Ἑρμῆς μὲν ἔπειτα κιὼν παρὰ λάϊνον ἄντρον ἐς φάος ἐξήλαυνε βοῶν ἴφθιμα κάρηνα, Λητοΐδης δ’ ἀπάτερθεν ἰδὼν ἐνόησε βοείας πέτρηι ἐν ἠλιβάτωι, τάχα δ’ εἴρετο κύδιμον Ἑρμῆν· “πῶς ἐδύνω, δολομῆτα, δύω βόε δειροτομῆσαι,405 ὧδε νεογνὸς ἐὼν καὶ νήπιος; αὐτὸς ἐγώ γε θαυμαίνω κατόπισθε τὸ σὸν κράτος, οὐδέ τί σε χρή μακρὸν ἀέξεσθαι, Κυλλήνιε Μαιάδος υἱέ.” ὣς ἄρ’ ἔφη, καὶ χερσὶ περίστρεφε καρτερὰ δεσμά ἄγνου. ταὶ δ’ ὑπὸ ποσσὶ κατὰ χθονὸς αἶψα φύοντο410 381 δὲ M: om. Ψ  383 ἐπιδέξομαι Thomas: ἐπιδεύομαι M: ἐπιδαίομαι Ψ: ἐπιδώσομαι Barnes  385 τείσω Van Herwerden (1907) 189: τίσω Ω  ποτὶ M: ποτὲ Ψ  φώρην M: φωνὴν Ψ  389 κακομηδέα Stephanus: -μήδεα Ω  395 νεῦσεν b: νεῦσε MDap    397–8 σπεύδοντε … ἠμαθόεντα Mp: σπεύδοντο ... ἠμαθόεντα δ’ Θ  398 ἷξον

120

T H E H O M E R I C H Y M N TO H E R M E S

fear the disapproval of Helios and the other gods, that I love you, and am scared of his anger. You know for yourself that I am not guilty, but I shall further take on a great oath – not by this ornamented porch-space belonging to the immortals. (385) I will pay him back one day, in proportion to his brutal ransack, despite his strength. But as for you, assist the young.’ So the Kyllenian slayer of Argus spoke with a shifty glance, and he kept the swaddling over his forearm and did not cast it off. But Zeus laughed out loud when he saw his scheming son (390) making a good skilful denial concerning the cows. He ordered both of them to search with their hearts in sympathy, and Hermes the guide to lead the way and without injurious intent to indicate where exactly the place was that he had hidden the strong and valued cows. (395) The son of Kronos bowed his head, and treasured Hermes complied, since it was easy for the plan of aegis-bearing Zeus to persuade him. Both the handsome children of Zeus hastened to sandy Pylos and reached the passage of the Alpheus, and they arrived at the fields and the high-beamed lodge (400) where the goods had been reared during the hours of night. There Hermes then went past the stone cave, and drove the strong and valued cows out into the light. But the son of Leto, looking on from a distance, noticed the cow-hides in the sheer rockface, and promptly questioned glorious Hermes: (405) ‘How were you able, trickyminded one, to slaughter two cows, when you are such a newborn infant? Even I marvel at your strength which is to come: you definitely shouldn’t grow tall, Kyllenian son of Maia.’ So he spoke, and twisted round his forearms strong bonds (410) made of vitex. But they immediately started to sprout Chalkondyles: ίξον M: ἵξον Ψ  400 ἡχοῦ Fick (1896) 272: ὅχου M: ἦχ’ οὐ Ψ (ἧχ’ οὗ D)   δὴ Ψ: δὲ M   χρήματ’ ἀτάλλετο Chalkondyles: χρήματα τιτάλλετο M: χρήματ’ ἀτιτάλλετο Ψ  401 παρὰ Ψ: ἐς M   402 φάος Hermann: φῶς Ω  403 ἀπάτερθεν Ψ: ἀπάνευθεν M   404 πέτρηι ἐν Thomas (cf. 124): πέτρηι ἐπ’ Ψ: γαίη κατ’ M   εἴρετο M: ἤρετο Ψ  406 νεογνὸς ἐὼν Ψ: νεογνοίων M   408 ἀέξεσθαι Ψ: ἀεξάσθαι M  410 ἄγνου. ταὶ Châteillon: ἅγνου· ταί M: ἁγνοῦ ταὶ Θ: ἁγνοῦται p  

121

T H E H O M E R I C H Y M N TO H E R M E S

αὐτόθεν, ἐμβολάδην ἐστραμμέναι ἀλλήληισιν, ῥεῖά τε καὶ πάσηισιν ἐπ’ ἀγραύλοισι βόεσσιν Ἑρμέω βουλῆισι κλεψίφρονος· αὐτὰρ Ἀπόλλων θαύμασεν ἀθρήσας. τότε δὴ κρατὺς Ἀργεϊφόντης χῶρον ὑποβλήδην ἐσκέψατο, πῦρ ἀμαρύσσον415 ἐγκρύψαι μεμαώς. Λητοῦς δ’ ἐρικυδέος υἱόν

417 ῥεῖα μάλ’ ἐπρήϋνεν Ἑκηβόλον ὡς ἔθελ’ αὐτός,  καὶ κρατερόν περ ἐόντα· λαβὼν δ’ ἐπ’ ἀριστερὰ χειρός πλήκτρωι ἐπειρήτιζε κατὰ μέλος, ἣ δ’ ὑπὸ χειρός σμερδαλέον κονάβησε, γέλασσε δὲ Φοῖβος Ἀπόλλων420 γηθήσας, ἐρατὴ δὲ διὰ φρένας ἤλυθ’ ἰωή θεσπεσίης ἐνοπῆς, καί μιν γλυκὺς ἵμερος ἥιρει θυμὸν ἀκουάζοντα. λύρηι δ’ ἐρατὸν κιθαρίζων στῆ ῥ’ ὅ γε θαρσήσας ἐπ’ ἀριστερὰ Μαιάδος υἱός Φοίβου Ἀπόλλωνος. τάχα δὲ λιγέως κιθαρίζων425 γηρύετ’ ἀμβολάδην, ἐρατὴ δέ οἱ ἕσπετο φωνή, κραίνων ἀθανάτους τε θεοὺς καὶ Γαῖαν ἐρεμνήν ὡς τὰ πρῶτα γένοντο καὶ ὡς λάχε μοῖραν ἕκαστος. Μνημοσύνην μὲν πρῶτα θεῶν ἐγέραιρεν ἀοιδῆι μητέρα Μουσάων· ἣ γὰρ λάχε Μαιάδος υἱόν·430 τοὺς δὲ κατὰ πρέσβιν τε καὶ ὡς γεγάασιν ἕκαστος ἀθανάτους ἐγέραιρε θεοὺς Διὸς ἀγλαὸς υἱός πάντ’ ἐνέπων κατὰ κόσμον, ὑπωλένιον κιθαρίζων. τὸν δ’ ἔρος ἐν στήθεσσιν ἀμήχανος αἴνυτο θυμόν, καί μιν φωνήσας ἔπεα πτερόεντα προσηύδα·435 “βουφόνε, μηχανιῶτα, πονεόμενε δαιτὸς ἑταῖρε, πεντήκοντα βοῶν ἀντάξια ταῦτα μέμηλεν· ἡσυχίως καὶ ἔπειτα διακρινέεσθαι ὀΐω. 411 ἐμβολάδην Ψ: ἀμβολάδην M   412 ἀγραύλοισι p: ἀγραύλησι MΘ  415 ὑποβλήδην Ω: ὑποβλέβδην Passow (1823) 949: παρβλήδην?  ἀμαρύσσον Lohsee (1872) 48: ἀμαρύσσων Ω  416 lacuna Radermacher   υἱόν Ω: υἱός?  417 ἔθελ’ Ψ: ἔθετ’ M   418 λαβὼν Ω: λύρην Stephanus xxvii   χειρός Ψ: λύρην M   419 ἐπειρήτιζε Ψ: ἐπηρ- M   422 M: om. Ψ  423 θυμὸν West (1966) 149: θυμῶ Ω  

122

T H E H O M E R I C H Y M N TO H E R M E S

underfoot in the ground, from that very spot, entwined as though grafted into each other, easily and covering all the field-dwelling cows, by the designs of beguiling Hermes. Apollo beheld and marvelled. Then the strong slayer of Argus (415) examined the place furtively(?), eager to cover over the glinting fire. But the son of glorious Leto ... he calmed the Far-shooter with great ease just as he desired, despite the latter’s strength. He took it on his left-hand side and with the plectrum tested it in a scale, and at his touch it (420) reverberated fearsomely. But Phoebus Apollo laughed with joy; lovely was the god-given clamour’s sound as it shot through his senses, and sweet desire pervaded his spirit as he listened. And he, the son of Maia, playing lovely music on the lyre, took confidence and stood on the left (425) of Phoebus Apollo. Presently, as he played clearly, he began to sing as for a prelude, and lovely was the voice which accompanied him. He declared both the immortal gods and dark Earth, how they were born originally, and how each was allotted their portion. First of the gods he honoured Mnemosyne in his song, (430) the mother of the Muses – since she had been allotted the son of Maia. Then the treasured son of Zeus honoured the other immortal gods according to both seniority and each one’s nature, uttering everything in order as he played the lyre which hung from his forearm. Unmanageable desire seized the spirit in the other’s chest, (435) and giving voice he addressed him with winged words: ‘Cow-killer, contrivance-man, working companion of the feast, this concern is well worth fifty cows. I reckon we’ll reach a relaxed settlement later, but come on, first tell me this, versatile 428 τὰ πρῶτα bB: ταπρῶτα Ω  429 ἀοιδῆι Ψ: ἀοιδὸν M   431 δὲ Ψ: δὲ καὶ M   πρέσβιν Matthiae (1800) 290: πρέσβην Ω  γεγάασιν MDa: γεγάασσιν bp  ἕκαστος Ψ: ἅπαντες M   433 ὑπωλένιον Barnes (sub cubito Châteillon): ἐπ- Ω  437 μέμηλεν Thomas: μέμηλας Ω  438 διακρινέεσθαι xp: διακρίνεσθαι MD  

123

T H E H O M E R I C H Y M N TO H E R M E S

νῦν δ’ ἄγε μοι τόδε εἰπέ, πολύτροπε Μαιάδος υἱέ· ἢ σοί γ’ ἐκ γενετῆς τάδ’ ἅμ’ ἕσπετο θαυματὰ ἔργα,440 ἦέ τις ἀθανάτων ἠὲ θνητῶν ἀνθρώπων δῶρον ἀγαυὸν ἔδωκε καὶ ἔφρασε θέσπιν ἀοιδήν; θαυμασίην γὰρ τήνδε νεήφατον ὄσσαν ἀκούω, ἣν οὔ πώ ποτέ φημι δαήμεναι οὔτέ τιν’ ἀνδρῶν οὔτέ τιν’ ἀθανάτων οἳ Ὀλύμπια δώματ’ ἔχουσιν,445 νόσφι σέθεν, φιλῆτα, Διὸς καὶ Μαιάδος υἱέ. τίς τέχνη; τίς Μοῦσα ἀμηχανέων μελεδωνῶν; τίς τρίβος; ἀτρεκέως γὰρ ἅμα τρία πάντα πάρεστιν εὐφροσύνην καὶ ἔρωτα καὶ ἥδυμον ὕπνον ἑλέσθαι. καὶ γὰρ ἐγὼ Μούσηισιν Ὀλυμπιάδεσσιν ὀπηδός,450 τῆισι χοροί τε μέλουσι καὶ ἀγλαὸς οἷμος ἀοιδῆς καὶ μολπὴ τεθαλυῖα καὶ ἱμερόεις βρόμος αὐλῶν, ἀλλ’ οὔ πώ τί μοι ὧδε μετὰ φρεσὶν ἄλλο μέλησεν, οἷα νέων θαλίηισ’ ἐνδέξια ἔργα πέλονται· θαυμάζω, Διὸς υἱέ, τάδ’ ὡς ἐρατὸν κιθαρίζεις.455 νῦν δ’ ἐπεὶ οὖν ὀλίγος περ ἐὼν κλυτὰ μήδεα οἶδας, ἵζε, πέπον, καὶ θυμὸν ἐπαίνει πρεσβυτέροισιν· νῦν γάρ τοι κλέος ἔσται ἐν ἀθανάτοισι θεοῖσιν, σοί τ’ αὐτῶι καὶ μητρί. τὸ δ’ ἀτρεκέως ἀγορεύσω· ναὶ μὰ τόδε κρανάϊνον ἀκόντιον, ἦ μὲν ἐγώ σε460 κυδρὸν ἐν ἀθανάτοισι καὶ ὄλβιον ἡγεμόν’ ἕσσω, δώσω τ’ ἀγλαὰ δῶρα καὶ ἐς τέλος οὐκ ἀπατήσω.” τὸν δ’ Ἑρμῆς μύθοισιν ἀμείβετο κερδαλέοισιν· “εἰρωτᾶις μ’, Ἑκάεργε, περιφραδές. αὐτὰρ ἐγὼ σοί 465 τέχνης ἡμετέρης ἐπιβήμεναι οὔ τι μεγαίρω· σήμερον εἰδήσεις, ἐθέλω δέ τοι ἤπιος εἶναι βουλῆι καὶ μύθοισι. σὺ δὲ φρεσὶ πάντ’ εὖ οἶδας· πρῶτος γάρ, Διὸς υἱέ, μετ’ ἀθανάτοισι θαάσσεις ἠΰς τε κρατερός τε, φιλεῖ δέ σε μητίετα Ζεύς 440 ἢ M: ἦ Ψ  γενετῆς M: γενεῆς Ψ  441 ἦέ Allen (1896): ἠέ Ω  446 φιλῆτα Aldine (1504): φιλητὰ MΘ: φηλητὰ p  447 μελεδωνῶν Agar (1929) 19: μελεδόνων M: μελεδώνων Ψ  450 ὀπηδός p: ὁπ- MΘ  451 χοροί τε Ψ: χορὸς M   οἷμος L: ὕμνος Mxm: οἶμος

124

T H E H O M E R I C H Y M N TO H E R M E S

son of Maia: (440) did these wondrous works accompany you in particular from birth, or did one of the immortals or of mortal humans give you the marvellous gift, and show you inspired song? For this newly spoken voice which I hear is wondrous, and I declare that never yet did either any man learn it (445) or any of the immortals who have Olympian homes, except you, brigand, son of Zeus and Maia. What is this craft? Who is the Muse of unmanageable cares? What is its path? Truly it is possible to lay hold of all three things together – good cheer, desire, and sweet sleep. (450) You see, I am a follower of the Olympian Muses, whose concerns are dancing, the treasured course of song, luxuriant chorusing, and the desirable blare of pipes. However, up to now nothing else has concerned my senses in this manner – a thing like the left-to-right business at young men’s feasts. (455) Son of Zeus, I marvel at how lovely this lyre-playing of yours is. ‘But now, since despite your smallness you know schemes which will be famous, sit down, dear fellow, and approve your elders’ impulse for them, since presently you will have fame among the immortal gods – both yourself and your mother. I shall declare it precisely: (460) truly, by this cornel-wood javelin, I shall seat you as a glorious and prosperous leader among the immortals, I shall give you treasured gifts, and I shall not ever deceive you.’ Hermes replied to him with canny words: ‘You ask me, Farworker, very cleverly. And I (465) in no way begrudge you embarking on my craft: you will know it today, and I am willing to be kind to you in thought and word. But your mind knows everything well, son of Zeus, since you sit in first place among the immortals, both brave and strong, and Zeus the planner loves you Ψ  453 ἄλλο M: ὧδε Ψ  μέλησεν Ψ: μέλησιν M   456 οἶδας Ψ: οἶσθα M  457–8 M: om. Ψ  457 θυμὸν M: μῦθον Ruhnken (1781) 46   460 κρανάϊνον Càssola: κρανάϊον Ω: κρανέϊνον Ilgen   461 ἡγεμόν’ ἕσσω Agar (1896) 389: ἡγεμονεύσω Ω: ἡγεμόν’ εἵσω Tyrrell (1894) 47   464 σοί Hermann: σοι Ω  466 ἤπιος b: ἥπιος MDap  468 θαάσσεις Ψ: θο- M   469 μητίετα Θ: -ιέτα Mp 

125

T H E H O M E R I C H Y M N TO H E R M E S

ἐκ πάσης ὁσίης, ἔπορεν δέ τοι ἀγλαὰ δῶρα470 καὶ τιμάς· σέ γέ φασι δαήμεναι ἐκ Διὸς ὀμφῆς μαντείας, Ἑκάεργε (Διὸς πάρα θέσφατα πάντα), καὶ νῦν αὐτὸς ἐγώ γε παναφνειὸν δεδάηκα. σοὶ δ’ αὐτάγρετόν ἐστι δαήμεναι ὅττι μενοινᾶις. ἀλλ’ ἐπεὶ οὖν τοι θυμὸς ἐπιθύει κιθαρίζειν,475 μέλπεο καὶ κιθάριζε καὶ ἀγλαΐας ἀλέγυνε δέγμενος ἐξ ἐμέθεν, σὺ δέ μοι, φίλε, κῦδος ὄπαζε. εὐμόλπει μετὰ χερσὶν ἔχων λιγύφωνον ἑταίρην καλὰ καὶ εὖ κατὰ κόσμον ἐπισταμένην ἀγορεύειν· εὔκηλος μὲν ἔπειτα φέρειν ἐς δαῖτα θάλειαν480 καὶ χορὸν ἱμερόεντα, καὶ ἐς φιλοκυδέα κῶμον, εὐφροσύνην νυκτός τε καὶ ἤματος. ὅς τις ἂν αὐτήν τέχνηι καὶ σοφίηι δεδαημένος ἐξερεείνηι, φθεγγομένη παντοῖα νόωι χαρίεντα διδάσκει, 485 ῥεῖα συνηθείηισιν ἀθυρομένη μαλακῆισιν,  ἐργασίην φεύγουσα δυήπαθον. ὃς δέ κεν αὐτήν νῆϊς ἐὼν τὸ πρῶτον ἐπιζαφελῶς ἐρεείνηι, μὰψ αὔτως κεν ἔπειτα μετήορά τε θρυλίζοι. σοὶ δ’ αὐτάγρετόν ἐστι δαήμεναι ὅττι μενοινᾶις. καί τοι ἐγὼ δώσω ταύτην, Διὸς ἀγλαὲ κοῦρε,490 ἡμεῖς δ’ αὖτ’ ὄρεός τε καὶ ἱπποβότου πεδίοιο βουσὶ νομούς, Ἑκάεργε, νομεύσομεν ἀγραύλοισιν. ἔνθεν ἅλις τέξουσι βόες ταύροισι μιγεῖσαι μίγδην θηλείας τε καὶ ἄρσενας. οὐδέ τί σε χρή 495 κερδαλέον περ ἐόντα περιζαμενῶς κεχολῶσθαι.” ὣς εἰπὼν ὤρεξ’, ὃ δ’ ἐδέξατο Φοῖβος Ἀπόλλων, Ἑρμῆι δ’ ἐγγυάλιξεν ἑκὼν μάστιγα φαεινήν βουκολίας τ’ ἐπέτελλεν, ἔδεκτο δὲ Μαιάδος υἱός 471 τιμάς· σέ γέ Aldine (1504) (with γε: γέ Allen 1896): τιμάς· σὲ δέ MV: τιμὰς σέ γε Ψ  472 μαντείας Matthiae (1805): μαντείας τ’ Ω (μ. θ’ MΠq: μ. δ’ V)   parenthesis Matthiae (1800) 295   πάρα Stephanus: παρα  M: παρὰ Ψ  473 καὶ MDbp: τῶν abm  γε παναφνειὸν Thomas: -γε παῖ()δ’ ἀφνειὸν Ω: σε (Hermann) παῖ ἀ. Allen (1897) 265–6: σε μάλ’ ἀ. Evelyn-White   474 αὐτάγρετόν Chalkondyles: αὖτ’

126

T H E H O M E R I C H Y M N TO H E R M E S

(470) for all your pious behaviour, and has given you treasured gifts and honours: you, they say, learned prophecies from Zeus’s pronouncement, Far-worker (all divine decrees come from Zeus), and I have now learnt personally of your great wealth. You can seize for yourself any learning you desire. (475) But since your heart is set upon playing the lyre, sing and play it and busy yourself with its splendours – having received them from me: and for your part, friend, bestow glory on me. Be a fine singer, holding in your hands your clear-voiced companion, who is skilled at speaking fine words in good order: (480) hereafter, free from all care, take her to the lavish banquet and gorgeous dance, and to the celebratory revel – she will bring good cheer both by day and by night. To whoever should question her knowledgeably with skill and expertise, she voices all kinds of teachings to please the mind, (485) easily toyed with in delicate intimacy, shunning agonious labour. To whoever from lack of knowledge should question her angrily at first, however, she would then drone quite emptily and ungroundedly. But you can seize for yourself any learning you desire. (490) ‘And I will give you her, treasured boy of Zeus, but for my part I will graze the pastures of both mountain and horse-rearing plain with the field-dwelling cows, Far-worker. Then the cows will mingle with bulls and bear a plentiful mix of females and males. And you definitely shouldn’t (495) be so vehemently angry, despite your eye for a profit.’ So he spoke and proffered it, and Phoebus Apollo received it and willingly handed Hermes the shining whip and prescribed

ἄγρετόν Ω (-ον M)   475 τοι Ψ: τι M   479 ἐπισταμένην Barnes: -ως Ω  481 φιλοκυδέα MΘ: φιλομειδέα p  482 ἂν Ψ: ἂν καὶ M   483 δεδαημένος Ψ: δαήμενος M   486 φεύγουσα M: φθέγγουσα Ψ  487 ἐὼν Ψ: ἰὼν M   ἐρεείνηι M: ἐρέεινε Ψ  488 αὔτως Hermann: αὕτως Ω  μετήορά  τε  θρυλίζοι Schneidewin (1848) 691: μετήορά τε θρυαλίζοι MbQ: μετήορα τε()θρυαλίζοι Dap  489 αὐτάγρετόν Aldine (1504): αὖτ᾽ ἄγρετόν Ω  μενοινᾶις Ψ: μενοιῆς M   493 τέξουσι Ψ: θ’ ἕξM  497 ἑκὼν Martin (1605) f. 37: ἔχων Ω: ἔχειν D’Orville (n.d.) f. 255  

127

T H E H O M E R I C H Y M N TO H E R M E S

γηθήσας. κίθαριν δὲ λαβὼν ἐπ’ ἀριστερὰ χειρός Λητοῦς ἀγλαὸς υἱός, ἄναξ ἑκάεργος Ἀπόλλων,500 πλήκτρωι ἐπειρήτιζε κατὰ μέλος, ἣ δ’ ὑπένερθεν ἱμερόεν κονάβησε, θεὸς δ’ ὑπὸ καλὸν ἄεισεν. καί ῥα βόας μὲν ἔπειτα ποτὶ ζάθεον λειμῶνα ἐτραπέτην, αὐτοὶ δὲ Διὸς περικαλλέα τέκνα ἄψορροι πρὸς Ὄλυμπον ἀγάννιφον ἐρρώσαντο505 τερπόμενοι φόρμιγγι. χάρη δ’ ἄρα μητίετα Ζεύς, ἄμφω δ’ ἐς φιλότητα συνήγαγε. καὶ τὸ μὲν Ἑρμῆς Λητοΐδην ἐφίλησε διαμπερές, ὡς ἔτι καὶ νῦν σήματ’, ἐπεὶ κίθαριν μὲν Ἑκηβόλωι ἐγγυάλιξεν ἱμερτήν, δεδαὼς ὃς ὑπωλένιον κιθάριζεν,  510 αὐτὸς δ’ αὖθ’ ἑτέρης σοφίης ἐκμάσσατο τέχνην· συρίγγων ἐνοπὴν ποιήσατο τηλόθ’ ἀκουστήν. καὶ τότε Λητοΐδης Ἑρμῆν πρὸς μῦθον ἔειπεν· “δείδια, Μαιάδος υἱέ, διάκτορε ποικιλομῆτα, μή μοι ἅμα κλέψηις κίθαριν καὶ καμπύλα τόξα.515 τιμὴν γὰρ πὰρ Ζηνὸς ἔχεις ἐπαμοίβιμα ἔργα θήσειν ἀνθρώποισι κατὰ χθόνα πουλυβότειραν. ἀλλ’ εἴ μοι τλαίης γε θεῶν μέγαν ὅρκον ὀμόσσαι, ἢ κεφαλῆι νεύσας ἢ ἐπὶ Στυγὸς ὄβριμον ὕδωρ, πάντ’ ἂν ἐμῶι θυμῶι κεχαρισμένα καὶ φίλα ἔρδοις.”520 καὶ τότε Μαιάδος υἱὸς ὑποσχόμενος κατένευσεν μή ποτ’ ἀποκλέψειν ὅσ’ Ἑκηβόλος ἐκτεάτισται, μηδέ ποτ’ ἐμπελάσειν πυκινῶι δόμωι. αὐτὰρ Ἀπόλλων Λητοΐδης κατένευσεν ἐπ’ ἀρθμῶι καὶ φιλότητι μή τινα φίλτερον ἄλλον ἐν ἀθανάτοισιν ἔσεσθαι.525

“... μήτε θεὸν μήτ’ ἄνδρα Διὸς γόνον. ἐκ δὲ τέλειον  526 σύμβολον ἀθανάτων ποιήσομαι ἠδ’ ἅμα πάντων, 499 Ψ: om. M   501 ὑπένερθεν Thomas (-ε Bothe): ὑπὸ νέρθεν M: ὑπὸ καλὸν Ψ  502 ἱμερόεν Ψ: σμερδαλέον M   καλὸν M: μέλος Ψ  ἄεισεν Ω: ἄειδεν Ilgen (cf. 54)   503 καί ῥα βόας M: ἔνθα βόες Ψ  ποτὶ Ψ: κατὰ M  504 ἐτραπέτην Ψ: δραπέτην M   505 πρὸς Ψ: πρὸ M   507 τὸ Ψ: τὰ M   509 σήματ’ M, Stephanus: σῆμα τ’ Ψ  κίθαριν M: κιθάρην Ψ  510 Ψ: om. M   comma placed before δεδαώς by Ernesti   ὃς Zimmermann (1920) 238: ὅ δ’ Ω  513 Ἑρμῆν Ψ: ἑρμῆ[[ς]] M  

128

T H E H O M E R I C H Y M N TO H E R M E S

cow-herding for him, and the son of Maia received it with joy. Then taking the lyre on his left-hand side (500) the treasured son of Leto, the far-working lord Apollo, tested it in a scale with the plectrum, beneath which it resounded gorgeously, and the god sang to it beautifully. Then the pair turned the cows towards the holy meadow, and the handsome children of Zeus themselves (505) sped back to snowy Olympus, taking pleasure in the lyre. And Zeus the planner rejoiced, and drew both together into affection. On his side, Hermes conceived affection for the son of Leto for ever – just as there are signs even now, since he handed the desirable lyre to the Far-shooter, (510) who began to play it knowledge­ ably as it hung from his forearm, while he in turn sought out the craft of a second expertise: he created for himself the panpipes’ clamour, audible from afar. The son of Leto at that point addressed a speech to Hermes: ‘I am afraid, son of Maia, guide and intricate planner, (515) that you will steal my lyre together with my curved bow! For from Zeus you have the prerogative that you will lay down the business of interchange for humankind across the nurturing Earth. However, if you could bear to swear for me the great oath of the gods, either nodding with your head or by the threatening water of Styx, (520) your actions would be entirely welcome and dear to my heart.’ Then the son of Maia did nod assent, promising never to steal all that the Far-shooter owned – never even to go up to his strong-built house. And Apollo son of Leto nodded assent, on conditions of harmony and affection, (525) that among the immortals nobody else would be dearer to him ‘... neither god nor man of Zeus’s line. And I will conclude a ratified covenant proper to immortals 515 ἅμα M: ἀνα- Ψ  κίθαριν M: κιθάρην Ψ  516 ἐπαμοίβιμα Wolf (1807): ἐπ’ ἀμοίβημα M: ἐπαμοίβια Ψ  517 πουλυβότειραν PQ2: πολυ- MΘq  518 μέγαν Ψ: [[κατὰ μέγαν]] M (κατὰ M2)  519 ὄβριμον Mp: ὄμβρ- Θ  520 ἔρδοις ML2: ἕρδ- Ψ  522 μή ποτ’ Ψ: μήτ’ M   524 ἀρθμῶι Ψ: ἀριθμῶ M  525 lacuna Thomas: Allen (1896) placed one after 526   527 πάντων Ω: πάντως Richardson  

129

T H E H O M E R I C H Y M N TO H E R M E S

πιστὸν ἐμῶι θυμῶι καὶ τίμιον. αὐτὰρ ἔπειτα ὄλβου καὶ πλούτου δώσω περικαλλέα ῥάβδον 530 χρυσείην τριπέτηλον, ἀκήριον ἥ σε φυλάξει, πάντας ἐπικραίνουσα θεμοὺς ἐπέων τε καὶ ἔργων τῶν ἀγαθῶν, ὅσα φημὶ δαήμεναι ἐκ Διὸς ὀμφῆς. μαντείην δέ, φέριστε διοτρεφές, ἣν ἐρεείνεις, οὔτε σὲ θέσφατόν ἐστι δαήμεναι οὔτέ τιν’ ἄλλον 535 ἀθανάτων· τὰ γὰρ οἶδε Διὸς νόος, αὐτὰρ ἐγώ γε πιστωθεὶς κατένευσα καὶ ὤμοσα καρτερὸν ὅρκον μή τινα νόσφιν ἐμεῖο θεῶν αἰειγενετάων ἄλλόν γ’ εἴσεσθαι Ζηνὸς πυκινόφρονα βουλήν. καὶ σύ, κασίγνητε χρυσόρραπι, μή με κέλευε 540 θέσφατα πιφαύσκειν ὅσα μήδεται εὐρύοπα Ζεύς. ἀνθρώπων δ’ ἄλλον δηλήσομαι, ἄλλον ὀνήσω, πολλὰ περιτροπέων ἀμεγάρτων φῦλ’ ἀνθρώπων. καὶ μὲν ἐμῆς ὀμφῆς ἀπονήσεται ὅς τις ἂν ἔλθηι φωνῆι τ’ ἠδὲ ποτῆισι τεληέντων οἰωνῶν· οὗτος ἐμῆς ὀμφῆς ἀπονήσεται οὐδ’ ἀπατήσω.545 ὃς δέ κε μαψιλόγοισι πιθήσας οἰωνοῖσιν μαντείην ἐθέληισι παρὲκ νόον ἐξερεείνειν ἡμετέρην, νοέειν δὲ θεῶν πλέον αἰὲν ἐόντων, φήμ’, ἁλίην ὅδον εἶσιν· ἐγὼ δέ κε δῶρα δεχοίμην. ἄλλο δέ τοι ἐρέω, Μαίης ἐρικυδέος υἱέ550 καὶ Διὸς αἰγιόχοιο, θεῶν ἐριούνιε, δαῖμον· σεμναὶ γάρ τινές εἰσι κασίγνηται γεγαυῖαι, παρθένοι ὠκείηισιν ἀγαλλόμεναι πτερύγεσσιν, τρεῖς, κατὰ δὲ κρατὸς πεπαλαγμέναι ἄλφιτα λευκά οἰκία ναιετάουσιν ὑπὸ πτυχὶ Παρνησοῖο,555 530 comma placed before ἀκήριον by Preller (1846) 514   531 θεμοὺς Ludwich (1908); cf. Hsch. θ.255: θεοὺς Ω  532 δαήμεναι – 534  ἐστι MΘ: om. p  533 διοτρεφές Θ: διαμπερὲς M   534 ἄλλον Ψ: -ων M   535 τὰ Matthiae (1800) 308: τὸ Ω  537 ἐμεῖο Ψ: ἐμοῖο M   540 μήδεται Mxp: βούλεται D   542 περιτροπέων Ψ: περιτραπὼν M   543 μὲν Ψ: μὴ M   ὅς τις ἂν ἔλθηι Θ: οὐδ’ ἀπατήσω M (cf. 545): ὅ.

130

T H E H O M E R I C H Y M N TO H E R M E S

and indeed to all, which my heart trusts and esteems. After that, I shall give you the beautiful staff of prosperity and wealth, (530) golden and with three leaves, which will guard you from harm as it brings to fulfilment all the suggestions for all the words and deeds of great men which I claim to have learnt from Zeus’s pronouncements. But regarding the prophecy which you question me about, good ward of Zeus, it is neither authorized for you to learn it nor for any other (535) of the immortals: the mind of Zeus knows such things, and I nodded assent as a pledge, and swore a mighty oath that apart from me no other of the gods born for all time would know Zeus’s close-constructed intent. So for you, gold-staffed brother: don’t keep bidding me (540) to reveal all the divine decrees which far-seeing Zeus plans. ‘As for humans, I will bring harm to one and benefit to another, as I round up the numerous tribes of unenviable humans. Indeed, from my pronouncements benefits will go to whoever comes at the cry and flight of decisive birds of omen: (545) this man will benefit from my pronouncements, and I will not deceive him. Whoever places his trust in vain-talking birds, however, and wishes to question my prophecy unwisely and to gain more wisdom than the eternal gods, I declare he will make a futile journey ... though I would accept his gifts! (550) ‘But I will tell you another thing, son of glorious Maia and aegis-bearing Zeus, beneficent among the gods, fateful deity. There live certain reverend sisters, maidens exulting in swift wings, three in number, their heads sprinkled with white barley-meal. (555) They dwell in homes below the fold of Parnassus, separate teachers of a prophetic art which I

τ. ἂ. ἔλθοι p  544 φωνῆι τ’ ἠδὲ  ποτῆισι M (with φωνή τ’ ἠδε), Ruhnken (1781) 50: φωνῆ καὶ πτερύγεσσι Ψ  547 ἐθέληισι p: -ήσει MΘ  παρὲκ Stephanus: παρεκ(-) Ω  549 ἁλίην Ψ: ἀλ- M   δέ κε Ψ: δὲ καὶ M   550 υἱέ Ψ: υἱός M   552 σεμναὶ M: μοῖραι Ψ  555 Παρνησοῖο Ψ: -σσοῖο M  

131

T H E H O M E R I C H Y M N TO H E R M E S

{

μαντείης ἀπάνευθε διδάσκαλοι ἣν ἐπὶ βουσίν παῖς ἔτ’ ἐὼν μελέτησα· πατὴρ δ’ ἐμὸς οὐκ ἀλέγυνεν. ἐντεῦθεν δἤπειτα ποτώμεναι ἄλλοτε ἄλληι κηρία βόσκονται καί τε κραίνουσιν ἕκαστα. αἳ δ’ ὅτε μὲν θυίωσιν ἐδηδυῖαι μέλι χλωρόν560 προφρονέως ἐθέλουσιν ἀληθείην ἀγορεύειν, ἢν δ’ ἀπονοσφισθῶσι θεῶν ἡδεῖαν ἐδωδήν 563(a) πειρῶνται δἤπειτα παρὲξ ὁδὸν ἡγεμονεύειν.  563(b) ψεύδονται δἤπειτα δι’ ἀλλήλων δονέουσαι.  τάς τοι ἔπειτα δίδωμι, σὺ δ’ ἀτρεκέως ἐρεείνων σὴν αὐτοῦ φρένα τέρπε, καὶ εἰ βροτὸν ἄνδρα δαήσαις 565 πολλάκι σῆς ὀμφῆς ἐπακούσεται, αἴ κε τύχηισιν. ταῦτ’ ἔχε, Μαιάδος υἱέ, καὶ ἀγραύλους ἕλικας βοῦς ἵππους τ’ ἀμφιπόλευε, καὶ ἡμιόνους ταλαεργούς ...”

569 καὶ χαροποῖσι λέουσι καὶ ἀργιόδουσι σύεσσιν 570 καὶ κυσὶ καὶ μήλοισιν, ὅσα τρέφει εὐρεῖα χθών, πᾶσι δ’ ἐπὶ προβάτοισιν ἀνάσσειν κύδιμον Ἑρμῆν, οἶον δ’ εἰς Ἀΐδην τετελεσμένον ἄγγελον εἶναι, ὅς τ’ ἄδοτός περ ἐὼν δώσει γέρας οὐκ ἐλάχιστον. οὕτω Μαιάδος υἱὸν ἄναξ ἐφίλησεν Ἀπόλλων 575 παντοίηι φιλότητι, χάριν δ’ ἐπέθηκε Κρονίων. πᾶσι δ’ ὅ γε θνητοῖσι καὶ ἀθανάτοισιν ὁμιλεῖ· παῦρα μὲν οὖν ὀνίνησι, τὸ δ’ ἄκριτον ἠπεροπεύει νύκτα δι’ ὀρφναίην φῦλα θνητῶν ἀνθρώπων. καὶ σὺ μὲν οὕτω χαῖρε, Διὸς καὶ Μαιάδος υἱέ, 580 αὐτὰρ ἐγὼ καὶ σεῖο καὶ ἄλλης μνήσομ’ ἀοιδῆς.

556 διδάσκαλοι ἣν Ψ: διδασκαλίαν M   557 ἀλέγυνεν MΘ: -εινεν p  558 ἄλλοτε Schneidewin (1848) 698: ἄλλοτ’ ἐπ’ Ω  560 θυίωσιν M: θυΐσωσιν Θ (-ι D): θύσωσιν p  563(a) Ω   563(b) xm  δονέουσαι Baumeister: δενέουσαι xm  565 ἄνδρα δαήσαις Thomas: ἄνδρ’ ἀδαῆ M: ἄνδρα δαείης

132

T H E H O M E R I C H Y M N TO H E R M E S

practised while still a child supervising the cattle (my father did not see to its provision). From there then, as they flit now one way now another, they browse on honeycombs and determine each thing. (560) When they rush on after eating pale honey, they are ready and willing to speak truthfully; but if they abandon the gods’ sweet food, then they try to lead one off track.3 Well then, these I give to you. Question them accurately (565) and amuse your own mind, and if you should teach a mortal man he will often hearken to the pronouncements you own, if he is successful. Have these, son of Maia, and attend to ­spiral-horned, field-dwelling cows, and horses and hard-working mules. … and amber-eyed lions and white-tusked boars (570) and dogs and all the sheep and goats that the broad earth nurtures, but glorious Hermes to rule while in charge of all flocks, and to be the sole initiated messenger to Hades who, though he receives no gifts, yet will give no small prize of honour. Thus lord Apollo cherished the son of Maia (575) with every kind of affection, and the son of Kronos added his favour. The god associates with all mortals and immortals: few indeed are the benefits he bestows, but he indiscriminately deceives the tribes of mortal humans throughout the dark night. And with that, joy to you, son of Zeus and Maia, (580) and I in turn will call to mind both you and another song.

Ψ  566 πολλάκι Ψ: -ις M   568 lacuna Wolf (1807)   572 οἶον Θ: οἷον Mp  576 ἀθανάτοισιν ὁμιλεῖ Stephanus (Diis conuersatur Velareus 1528): -σι νομίζων M: -σιν ὁμίλει Ψ Alternative: ‘then they lie as they whirl amongst each other.’

3

133

C O M M E N TA RY

C O M M E N TA RY [ Title] I doubt whether any one preferred way of referring to this composition existed at an early stage. Ancient testimonia suggest προοίμιον εἰς Ἑρμῆν and later ὕμνος εἰς Ἑρμῆν would have seemed natural.1 Testimonia sometimes ascribe the Homeric Hymns to Ὅμηρος and ὁ ποιητής (e.g. Th. 3.104.4, PHib. 2.173, Phldm. SH 78, D.S. 1.15.7, 3.66.3, 4.2.4, Σ Ar. Av. 575). Athenaeus more cautiously says Ὅμηρος ἢ τῶν Ὁμηριδῶν τις (1 22b); for Homeridae and hymns see also Pindar N. 2.1–5 and Introduction §2.3. One scholion speaks of Ὁμηρικοὶ ὕμνοι (Σ Pi. P. 3.14); or there is the periphrasis οἱ εἰς Ὅμηρον ἀναφερόμενοι ὕμνοι (Σ Nic. Al. 130; restored in POxy. 2737 fr. 1.i (2nd c. ce); cf. Vita Homeri 9.3).

1–9 1 Ἑρμῆν ὕμνει Μοῦσα: A proper name of the dedicatee appears in the first line x30/31 in H.Hom., and also in the proemic hymn of Hesiod’s Theogony.2 These hymns evince a preference for placing the name as early as is metrically convenient, and in the accusative (‘evocation’; contrast lyric hymns, which often deploy a vocative invocation).3 In such cases, either the performer introduces himself as the subject of a verb indicating performance (x18), or the Muse is asked to perform (x10, including the plural Muses in Hy. 32–3). For ὑμνεῖν in the latter context cf. Hy. 9.1 Ἄρτεμιν ὕμνει Μοῦσα, 31.1 Ἥλιον ὑμνεῖν... ἄρχεο Μοῦσα, 14.1–2 Μητέρα ... ὕμνει Μοῦσα. These passages, and the repeated use of ὑμνεῖν of the performer’s job in H.Ap. (19 = 207, 178), demonstrate a traditional proemic use, which one can translate ‘celebrate’ even though the stem ὑμν- had broader application in the archaic period

εἰς is used in all manuscripts, and in citations at e.g. Paus. 4.30.4, 10.37.5, Ath. 1 22b, St. Byz. s.v. Τευμησσός. For προοίμιον and ὕμνος see 1*, Introduction §2.3. POxy. 4667 (third century ce) may introduce Hy. 7 with something like Ληισ]τ̣αί (ἐστὶ δέ̣[ τις εἰς τὸν Διόνυ]σον ὕμν̣[ος). 2 The count excludes Hy. 1, whose beginning is lost, and H.Ares. For Hy. 33 and H.Pan, see Thomas (2011) 151. For the openings of H.Hom. see also Pavese (1991) 160–2, Calame (1995) 6–8; De Cristofaro (2006) 260–7 handily surveys all early epic proems. 3 Ἀφροδίτη gravitates to line-end, so H.Aphr. calls for a song about the ἔργα ... Ἀφροδίτης|. Uses of vocative: Ἱστίη for metrical reasons (Hy. 24, 29); Hy. 21 (Φοῖβε) was perhaps originally a citharodic proem also attributed to Terpander (see POxy. 2737 fr. 1.i). Uses of genitive: H.Ap. 1 μνήσομαι οὐδὲ λάθωμαι Ἀπόλλωνος ἑκάτοιο, Hy. 25 Μουσάων ἄρχωμαι, Hes. Th. 1 Μουσάων … ἀρχώμεθ’. 1

137

C O M M E N TA RY 2 (e.g. Od. 8.429, Hes. Op. 662; of legendary narratives at e.g. Th. 51, H.Ap. 161, Ibycus SLG S151.6, 12). Hermes’ name has various forms: epic formulas were built initially around Mycenaean Hermāhās; Ἑρμείας replaced this, perhaps as a compromise with Ionic Ἑρμέης; contracted Ἑρμῆς acquired formulas of its own, and generated Ἑρμείης.4 Herm. uses only contracted forms (x32), in stark contrast to Il. (x1/16, at 20.372) and Od. (x4/21). Other parts of Epos are more balanced (Hes. x2/3, Hes. fr. x4/7, other H.Hom. x5/9); Herm.’s exclusivity seems to betray the composer’s conscious preference. Διὸς καὶ Μαιάδος υἱόν: Genealogical attributes occur at the start of several other hymns: parentage at Hy. 7.1 ~ 26.2, 15.1–3, 16.2, 17.2 ~ 33.2; siblinghood at 9.1 ~ 27.3. They may allude to the potential for a birth-narrative (here, this hint will be taken up in 3), but they are also a preliminary move to situate the addressee within divine society. Zeus is ubiquitous whereas Maia is localized on Mt Kyllene (see 5, 244*). The lemma thus sets up a tension between local and global which will become relevant to Hermes’ desire to range beyond Kyllene and enter Olympian society. Μαῖα (secondarily Μαίη at Hes. Th. 938, back-formed from genitive Μαίης), as well as a proper name, means roughly ‘nan’, and mothering Hermes is indeed Maia’s main mythological role. The alternative declension in -άδ- of her oblique cases, as here, is typical of nymphs: Herodian already compared πέλεια vs πελειάς (ii.566; Maia was one of the P(e)leiades); cf. Chantraine (1933) 357. 2 Κυλλήνης μεδέοντα καὶ Ἀρκαδίης πολυμήλου: Kyllene, in the NE corner of ancient Arcadia, has the second-highest peak in the Peloponnese (2374 m, 7789 ft). It is evidently singled out here, immediately after mention of Hermes’ parentage, as his birthplace – a claim which was recognized across Greece, whereas that of Tanagra (Paus. 9.20.3) was localized.5 H.Pan 31 gives Hermes a τέμενος on Kyllene; Geminus Elem. 17 (drawing on earlier sources) says he received annual sacrifices there, above the clouds; Pausanias (8.17.1–2) found a ruined temple and a xoanon. Archaic and classical finds from a cavecult on Kyllene’s western peak are compatible with worship of a nymph like Maia, though other interpretations are possible.6

See Ruijgh (1967) 12. Peters (1980) 263–73 noted the poor evidence for Ἑρμέης and suggested instead that Ἑρμείας was analogical to Αἰνείας. 5 Hermes is Κυλλήνιος at Od. 24.1; cf. Hes. fr. 170, the inscription hερμες εἰμι κϙλελνιος on Berlin F1704 (Kyllenios painter, c.560), and many other ­sources. Κυλλήνιος Ἀργεϊφόντης+| appears to be a ‘late’ formula: 387, Hes. fr. 66.4, Hy 18.1, varied in Herm. 318, 408*. 6 See Erath (1999) 242–6, Kusch (1999), Tausend (1999) 361–2; also Jost (1985) 33–5, 441–9 on how Hermes’ cults on Kyllene and at Pheneos may have been related. For 4

138

C O M M E N TA RY 3 Alcaeus’ Hymn to Hermes began χαῖρε Κυλλάνας ὀ μέδεις (fr. 308), exemplifying the continuity of phrasing between epic and lyric traditions.7 Epos restricts singular active μέδων/μεδέων to gods’ benevolent influence over places, especially in prayers and hymns; the verb remained a vox propria for this relationship (e.g. Alcaeus fr. 354, Sophocles fr. 371; Keyssner 1932: 75–6). The nuance of benevolence becomes explicit with Hermes’ broader patronage of Arcadia: πολυμήλου not only describes the pastoral economy of the region, but relates its success to Hermes’ patronage, via his role in increasing flocks.8 Where line 1 positions Hermes in relation to his divine family, line 2 implies later positive interactions with humans. Vian (1987) noted that Herm. 2 seems to have been used by the poet of O.Arg., split between 137 Κυλλήνης μεδέων and 198 ἀπ’ Ἀρκαδίης πολυμήλου; this is the most convincing case for Herm.’s influence on that poem. 3 ἄγγελον ἀθανάτων ἐριούνιον: From the Odyssey on, Hermes is primarily Zeus’s messenger, but he is frequently called ‘messenger/herald of the immortals’ when the Olympians are in agreement: epic examples are Hes. Th. 939, Op. 85, H.Dem. 408–9 (with ἐριούνιος, showing that it is not a separate predicate here), H.Pan 29, Hy. 29.8; also perhaps IG I3 784a (c.490 bce). Being a messenger figures prominently near start and end of Herm. (572*), though it will not have great relevance for the intervening myth. But the epithet has relevance of a different kind: as a means of communicating instructions from gods to humans (among others), Hermes’ work complements the hymnist’s; and this function makes Hermes (indeed, only Hermes) ἐριούνιος – normally derived in antiquity from ἐρι- + ὀνίνημι (cf. 30* ὀνήσιμον).9 The epithet therefore, like 2 πολυμήλου, presupposes the healthy exchange of worship (including hymns) for divine favours on which the present hymn is itself premised. Even though Hermes’ geographical range increases in line 3, this activity based on his having to carry out instructions brings Hermes ‘down to earth’ the early finds see IG V.2 pp. 82–4; they were already connected to Herm. by Scheffler (1884) 1–5. 7 A god’s geographical sphere of influence is in fact a rare source of epithets in the openings of extant Homeric Hymns, though compare Hy. 6.2 Κύπρου κρήδεμνα λέλογχεν. Similar attributes appear in the envoi at H.Dem. 490–2, H Aphr. 292, Hy. 10.4–5. 8 Arcadian pastoralism: e.g. Il. 2.605 Ὀρχομενὸν πολύμηλον, Hes. fr. 23a.32–3 Τεγ[έης ἠδ’ Ἀρκαδίης] πολυμήλου | ἀφνειὸς ἤνασσε, H.Pan 30 Ἀρκαδίην… μητέρα μήλων, IvO 266.3 ἐν Ἀρκαδίαι πολυμέλo[ι], Roy (1999) 329–33. Jost (1985) 439–56 surveys Hermes’ worship in Arcadia. For Hermes and flocks see Introduction n. 79. 9 Cf. 577*. Phoronis fr.5 relates the epithet to κέρδεσι κλεπτοσύνηισί τ’. The etymology may be ‘very fast’ or (Willi 1999) ‘speeding on high’: see Hsch. ο.1793 οὖνον· Κύπριοι δρόμον; H.Pan 28–9 has ἐριούνιον ... θοός in ­proximity. On this and other ancient interpretations, see Reece (1997) 29–32, (1999) 87–93, Langella (2013).

139

C O M M E N TA RY 4 – 5 and invokes the amusingly human, accessible side of his character (compare in particular Aristophanes’ Wealth), in contrast with line 2, where Hermes’ power sets him on a pedestal. ὃν τέκε Μαῖα: Opening clauses in the Homeric Hymns are almost always expanded by a relative pronoun referring to the addressee (Janko 1981: 9–11), whereas this technique is used more sporadically in other types of Greek hymn. Here, unusually, the poet returns to expand on something already implied in line 1 (see 1–9* below). Alcaeus’ hymn similarly had τὸν ... Μαῖα γέννατο Κρονίδαι μίγεισα (fr. 308.2–3), but without having mentioned his parentage before the relative pronoun. 4 νύμφη ἐϋπλόκαμος: For Maia as a mountain-nymph see 244*, and for nymphs in Greek thought, see Larson (2001). πλόκαμος (~ πλέκω) refers to braids and probably also to the curled tresses widespread in Greek art (Marinatos 1967: B2, NP s.v. ‘Hairstyle’). The prominence of hair in female beauty is evident from this formulaic epithet, καλλιπλόκαμος, ἠΰκομος, and καλλίκομος (cf. Osborne 2005: 10–14). Here the formula leads neatly on to Zeus’s amorous interest. Διὸς ἐν φιλότητι μιγεῖσα: (ἐν) φιλότητι μιγῆναι is an old euphemism with a complex system of formulaic variations; whether ἐν is present or not, the other participant may appear in either genitive or dative. Our lemma recurs in Hy. 7.57; compare e.g. Hes. Th. 944 |μιχθεῖσ’ ἐν φιλότητι Διός, 374 ὑποδμηθεῖσ’ Ὑπερίονος ἐν φιλότητι|. For this family of expressions see De Hoz (1964) 284–9. 5 αἰδοίη· μακάρων δὲ θεῶν ἠλεύαθ’ ὅμιλον: αἰδώς is an inhibiting sense that others will regard your actions as dishonourable, or respect for another’s honour. αἰδοῖος thus means both ‘deserving respect’ (straightforwardly applicable to Maia) and ‘with a sense of modesty’, generally referring to female sexual inhibition, especially of brides (see 381–2* and the brilliant study of Cairns 1993, esp. 120–5). The meanings of such polysemous words are cognitively activated and processed in parallel (see e.g. Traxler 2011: 81–127 for an introduction). Processing the suitability of the latter sense of αἰδοίη here falls into two stages: initially the runover word appears surprising, since Maia is not being chaste; but the rest of line 5 can be read as evidence of her bashfulness and desire to avoid critical gossip. Compare Iliad 2.513–15 οὓς τέκεν Ἀστυόχη ... παρθένος αἰδοίη ὑπερώιον εἰσαναβᾶσα | Ἄρηϊ κρατερῶι, ὃ δέ οἱ παρελέξατο λάθρηι, and Edwards (1966) 137–43 for some related effects of enjambement. Il. 20.7–9 suggests that nymphs’ attendance at Olympus was not normally compulsory. The norms of αἰδώς governing females’ public appearances varied from place to place; for the expectation of goddesses’ αἰδώς, see e.g. Od. 8.324.

140

C O M M E N TA RY 6 – 7 For the aorist ἠλεύαθ’, see LfgrE s.v. ἀλεύομαι 2 on aspect, and GH i.384–5 on morphology. ἠλεύαθ’ ὅμιλον probably had a traditional background: at Od. 17.67 Telemachus ἀλεύατο πουλὺν ὅμιλον of the suitors. 6 ἄντρον ἔσω ναίουσα παλίσκιον: The paradosis is unusual but viable. For ‘dwelling within a cave’ one expects ἄντρου ἔσω ... παλισκίου, which Ernesti considered and Baumeister printed. But adverbial ἔσω might qualify παλίσκιον: ‘inhabiting a cave which was internally shadowy’; compare e.g. Plato R. 3 407d τὰ δ’ εἴσω διὰ παντὸς νενοσηκότα σώματα. Or conceivably it could modify the participle (‘inhabiting a cave, inwardly’), though phrases like Hdt. 8.129.2 εἶναι ἔσω ἐν τῆι Παλλήνηι offer only loose parallels. Either way, ἔσω is semantically redundant, and so intensifies Maia’s shunning of publicity.10 παλι- may be another intensifier, meaning ‘repeatedly, doubly’; cf. παλιμμήκης, Dürbeck (1977) 159. The alternative interpretation of παλίσκιος as ‘with shadows falling backwards’ seems banal. However, any intensifying force seems to have faded by the time Theophrastus used παλίσκιος of shady habitats. Caves are a favoured home of nymphs (Larson 2001: 226–58), and hence often feature as sites of erotic and/or kourotrophic activity, as here. But cave-dwelling is also typical of beasts, monsters like Polyphemus or Scylla, and humans who exist in or have been reduced to a pre-civilized state.11 Though Maia’s home will turn out to be well appointed (23*), the cave’s connotations contribute to the contrast between it and Olympus, in a way which primes us for Hermes’ aspirations for the conveniences of Olympian lifestyle; he will tendentiously criticize the cave’s primitive conditions at 172*. 7 νύμφηι ἐϋπλοκάμωι: For the repetition see below, 1–9*. μισγέσκετο: The iterative form shows that the affair was no one-night stand, as Zeus’s so often were. Cf. 8*, 58*; for the various nuances of the suffix -σκ-, see GH i.316–24.12 Amyntor’s extramarital liaison is similarly marked with frequentatives at Il. 9.450.

Similarly Clay (1989) 104 n. 30. εἴσω mainly bears the original sense ‘to within’ in Epos, but I see the sense ‘within’ at Od. 7.13 (adverb), Cypr. fr. 15.5 (preposition). D.P. 253 παραλίην ναίουσιν ἔσω Σερβωνίδα λίμνην looks like a parallel, but the scholiast read Σερβωνίδος ἅλμης with λίμνη as part of a gloss; even with the ms. reading, one might interpret ‘inhabiting the coast, (extending) into the Serbonian marsh’. 11 See e.g. Hy. 20.3–4 ἀνθρώπους ... οἳ τὸ πάρος περ | ἄντροις ναιετάασκον ἐν οὔρεσιν ἠΰτε θῆρες, PV 452, S. Ph. 31–44, D.S. 1.8.7 (< Democritus?). 12 That μίσγω already derived from μίγ-σκω was no longer felt; cf. Od. 12.355 βοσκέσκετο. 10

141

C O M M E N TA RY 8 – 9 νυκτὸς ἀμολγῶι: Here a lengthy period of quiet. The formula’s origins are disputed, as is the existence of a coherent meaning for the audiences of Epos. Its uses in the simile-pair Il. 22.27–8 ~ 317–18 refer to different phases of the night, and they presuppose human observers, whereas people are asleep in Herm. (~ Hy. 18.7), Il. 15.324–5 (probably) and Od. 4.841. Possibly the folketymology via ἀ- + μολ- (Philoxenos fr. 435) set in early, since Epos’ uses imply unexpected motion to which the stillness of night is a foil: predators attack suddenly at Il. 11.173 (μολὼν ἐν νυκτὸς ἀμολγῶι), 15.324–5; Sirius’ journey is compared to Achilles’ sudden appearance at 22.27–8, the evening star’s to his spear-cast at 317–18; the unexpected vision ‘rushed upon’ Penelope in Od. 4; Zeus’s visits to Kyllene contrast with Hera’s expectation; similarly Heracles ‘flees’ the νυκτὸς ἀμολγόν in the sun’s cup at Aeschylus fr. 69.7.13 Janko (1982) 149 observed that Herm.’s poet elsewhere chose νυκτὸς ἐν ὥρηι| (e.g. 67), and suggested that the presence of ἀμολγῶι may be a sign that these opening lines were more closely tied to traditional hymnic phrasing. 8 ὄφρα κατὰ γλυκὺς ὕπνος ἔχοι: Sleep is by turns powerful and comforting in Epos (Gitlbauer 1884: 22–5). Here both associations are combined, with irony: ‘sweet’ (to whom?) sleep ‘held Hera down’ (with connotations of overpowering and restraining), and allowed Zeus to cheat. ὄφρα ‘while’ + optative is anomalous, but not implausible. An indefinite nuance (‘while sleep held Hera, on each occasion that she slept’), inspired by the common indefinite construction after ‘when’ (KG ii.451), would reinforce the iterative implications of 7 μισγέσκετο*. λευκώλενον Ἥρην: In chiastic balance with 7 νύμφηι ἐϋπλοκάμωι: Zeus’s two bedfellows receive complementary epithets expressing beauty.14 1–9 Hermes’ parentage: To an audience familiar with Hermes’ character, the description of Zeus’s furtive nocturnal dealings suggests ‘like father, like son’ – and so it will prove in the narrative.15 Maia’s avoidance

13

There are useful discussions in Koller (1969) 110–13, Edgeworth (1988); Tsagalis (2003) emphasized not movement but the connotation of danger, relevant in the Iliad but less so in the Hymn to Hermes. 14 The Iliad-poet strongly favoured θεὰ λευκώλενος Ἥρη (x18, mostly after a main verb); only x5 is θεά absent, in which case the formula may appear in various cases. In stark contrast (p < 0.003 by Fisher’s exact test), H.Hom. only use the shorter phrase, always in oblique cases (x5). 15 Cf. ΣD Il. 24.24 ἐπιθυμίαν ἔσχεν τοῦ κλέπτειν, ὅτι καὶ Ζεὺς κλέψας Ἥραν ἐμίγη Μαίαι, derived from Myth. Hom. (POxy. 4096 fr. 11). Zeus had also been born in secret from Kronos, and conversely the adult Hermes will have sex with nymphs in caves (H.Aphr. 262–3). The birth of Dionysus ­similarly involves Zeus deceiving Hera in

142

C O M M E N TA RY 8 – 9 of publicity also primes us elegantly for Hermes’ preoccupation with passing from a secretive, inhibited, and geographically restricted child to the famous, mobile, ubiquitous Olympian we know. This tension is posed in nuce by Διὸς καὶ Μαιάδος υἱόν (1) and the contrast of Κυλλήνης μεδέοντα (2) with the wide-ranging ἄγγελον ἀθανάτων (3). Meanwhile, 2 πολυμήλου and 3 ἐριούνιον imply Hermes’ healthy exchange of favours with worshippers, and as a messenger he establishes clear and fruitful communication between gods and humans: these attributes therefore intersect with the pragmatic concerns of hymnography (see Introduction §3.4). Despite these ways in which the opening birth-myth dovetails with the narrative to follow, it had an independent existence, in that it resembles Hy. 18, quoted here except for its doublet of envois (10–11 vs 12); variations are underlined: Ἑρμῆν ἀείδω Κυλλήνιον Ἀργεϊφόντην, Κυλλήνης μεδέοντα καὶ Ἀρκαδίης πολυμήλου, ἄγγελον ἀθανάτων ἐριούνιον, ὃν τέκε Μαῖα Ἄτλαντος θυγάτηρ Διὸς ἐν φιλότητι μιγεῖσα, αἰδοίη· μακάρων δὲ θεῶν ἀλέεινεν ὅμιλον5 ἄντρωι ναιετάουσα παλισκίωι, ἔνθα Κρονίων νύμφηι ἐϋπλοκάμωι μισγέσκετο νυκτὸς ἀμολγῶι εὖτε κατὰ γλυκὺς ὕπνος ἔχοι λευκώλενον Ἥρην, λάνθανε δ’ ἀθανάτους τε θεοὺς θνητούς τ’ ἀνθρώπους. ἀλέεινεν for 5 ἠλεύαθ’*, the avoidance of 6 ἔσω*, εὖτε for 8 ὄφρα*, and the avoidance of the potentially misleading sequence ὕπνος ... λήθων (8–9) all replace linguistically unusual features of Herm.; the changed ending of line 1 and 4 Ἄτλαντος θυγάτηρ avoids the mannered way that Herm. first calls Hermes Διὸς καὶ Μαιάδος υἱόν, then repeats this concept with additions in 3–4, then in turn repeats the elements of line 4 in 6–7 Kρονίων | νύμφηι ἐϋπλοκάμωι μισγέσκετο, before adding νυκτός, whose characterizing potential is explored in 8–9. This apparent pattern in the divergences inclines me to think that Hy. 18 is derivative from Herm. rather than an independent reflex of a tradition of describing Hermes’ birth in hymns. Unfortunately the circumstances which may have motivated this engaged rewriting of Herm., and its preservation as part of the collection of hymns, are unclear.16 The allusions to 1–4 at H.Pan 27–34 have a clearer point: Thomas (2011) 165–8. an out-of-the-way place (Hy. 1.A8), whereas Apollo and Heracles are born into an immediate opposition with her. For this mythological theme see Sowa (1984) 146–57. 16 Admittedly, the change in Hy. 18.1 does produce a repetition of its own: Κυλλήνιον … Κυλλήνης. Càssola 357 supposed that the compiler of the collection found a text

143

C O M M E N TA RY 1 0 – 1 1

10–19 10 Διὸς νόος ἐξετελεῖτο: For νόος ‘plan’, the result of τὸ νοεῖν, see LfgrE s.v. νόος 1a. For Zeus’s νόος/βουλή reaching fulfilment, see Il. 1.5, Cypria fr. 1.7, Od. 11.294–7 (each Διὸς δ’ ἐτελείετο βουλή, in the last case aligned with a lapse of months), and Hesiod Th. 1002 (μεγάλου δὲ Διὸς νόος ἐξετελεῖτο); GH ii.192 discusses the imperfective aspect in these phrases (cf. LfgrE s.v. τελέω 4aγ, η). Various events are ascribed in Greek literature to Zeus’s planning, including Apollo’s birth at Aristonous Ap. 5–8 Κοίου τε κόρας ... ἄγαλμα καὶ Ζηνὸς ὑψίστου μακάρων βουλαῖς. See Richardson (1974) on H Dem. 9, Sullivan (1994) 105–8 and Allan (2008) on the Homeric scope of Zeus’s plans. The effect here of suddenly grounding his passion for Maia in a ‘plan’ may be humorous. 11 τῆι δ’ ἤδη δέκατος μεὶς οὐρανῶι ἐστήρικτο: Periods of time are often the subject of (ἐκ)τελοῦμαι (LfgrE s.v. τελέω 7), so there is a smooth transition from the topic in line 10. μείς can mean ‘month’, but here στηρίζω (‘fix in position’) – and elsewhere φθίνω and ἵσταμαι – applies properly to the moon qua marker of time. For celestial uses of στηρίζω see Il. 11.27–8 ἴρισσιν …, ἅς τε Κρονίων | ἐν νέφεϊ στήριξε τέρας, Aratus 10 Zeus τά γε σήματ’ ἐν οὐρανῶι ἐστήριξεν. The verb varies ἑστάναι in the similar passage Il. 19.117–18 ὃ δ’ ἕβδομος ἑστήκει μείς,| ἐκ δ’ ἄγαγε πρὸ φόωσδε (cf. Anacreon PMG 362, Ps.-Thphr. Sign. 27). The hymnist may have been influenced by theories that heavenly bodies were physically affixed to a celestial shell or belt, though if so he does not distinguish it from the οὐρανός and so from the stars’ orbits, as theory required.17 See also 19*. Human births occur on average c.285 days after the end of the mother’s last period, and the tenth lunar month (days c.266–95, minus the part of the first month before conception occurred) was frequently mentioned as a pregnancy-term.18 τῆι clarifies that the ‘tenth’ moon is focalized by Maia tracking her pregnancy; ἤδη adds a touch of anticipation (‘now at last’).

of Herm. with variants, which he wrote up as a separate hymn. Hy. 13 is too short to say whether its relationship to H.Dem. 1–2 is similar to that between Herm. 1–9 and Hy. 18; the variations between Hy. 33 and Hy. 17 are greater. 17 For such theories see e.g. Anaximenes 13 A14, Empedocles 31 A54 (fixing of stars), and the theories of multiple belts in e.g. Pl. Tim. 38cd, Arist. Cael. 18 The Hippocratic author of Oct. Part. often uses ‘month’ to mean ‘menstrual cycle’, but it is still worth comparing 10.1 ‘For even so-called “ten-monthers”, which I declare are born in roughly seven forty-day periods – and they are most suitable for rearing …’. For the ten-month norm see also e.g. Nat. mul. 11, Nat. puer. 30, Mul. 240. Taking μείς as the sidereal month in such contexts (Richardson) seems to produce

144

C O M M E N TA RY 1 2 – 1 3 12 ἔς τε φόως ἄγαγεν: A single continuant occasionally lengthens a princeps syllable (West 1982: 38–9) and the hymnist admitted this in ἄγαγεν despite the availability of ἔκ τ’ ἄγαγε πρὸ φόωσδ’ (cf. Il. 16.188, 19.118). In those parallels, the subject is a divinity controlling the birth (Eileithyia and Hera, respectively) and the child is the object. Here the subject could be Zeus or Maia, and the object either ‘Zeus’s plan’ (from 10) or ‘Hermes’: the latter supplement is not difficult considering the Iliadic parallels, other references to the ‘light of the living’ in descriptions of births, and the contrast with the darkness of Maia’s cave (6).19 The phrase thus builds neatly on the thematic dimensions dark/light and secret/famous introduced in 1–9*. The suggestion that Hermes is ‘in the light’ immediately upon birth refines any idea that the narrative will track a simple trajectory from dark/secret to light/famous, and indeed, some darkness will be left at the end (578*). I think it likely that rhapsodes avoided εἰς before a consonant, as in Ionic (Smyth 1894: 598–601), and that it was sporadically introduced during transmission under Attic/koine influence. Hermann was first to avoid it in editing the Homeric Hymns. Ω had instances of both ἐς and εἰς before c­ onsonants (e.g. 216 ἐς Πύλον versus here). For the derivation of both from ἐν/ἐνς (~ ἐκ/ἐξ), see Schwyzer ii.454–6. ἀρίσημά τε ἔργα τέτυκτο: The pluperfect switches focus from the event of Hermes’ birth to its lasting effects. The vagueness of ἔργα τέτυκτο puts the emphasis on ἀρίσημα, which formulaic parallels suggest is attributive (LfgrE iv.437.36–55, plus Phor. fr. 2.7) rather than predicative (‘their deeds were made manifest’). The adjective covers ‘conspicuous’ (Ps.-Theoc. 25.158) as well as ‘famous’, and so neatly continues the line’s intertwining of dark/light with secretive/famous: this enhances the rhetorical poise of the line’s halves. 13 καὶ τότε γείνατο παῖδα: For the structure ὅτε δή (10) … καὶ τότε, see Denniston 308. Despite Hes. fr. 26.4 ἥ [ο]ἱ ἐ̣[γεί]ν̣ατο παῖδα, words of the shape ⏑–⏔ are unusual in this position (Hagel 2004: 173–5, 197; Taida 2010: 253), so there is reason to reject Chrysokokkes’ articulation τότ’ ἐγείνατο (D). πολύτροπον: An adjective not applied to Hermes outside Herm. (here, 439), and in earlier epic applied prominently and uniquely to Odysseus (Od. 1.1, 10.330). The early debate about how to weigh up its implications for Odysseus’ cleverness and deviousness (preserved in Antisthenes fr. 51, Plato Hipp. Min. 364–5, 370a) may have coloured Herm.’s uses. too short a timespan. On the distribution of modern human gestation periods see e.g. Jukic et al. (2013). 19 ‘Light of living’ at births: e.g. LfgrE iv.820.14–34, Hp. Oct. Part. 11 ἐς τοὐμφανὲς ἐξίηι, Pl. Prt. 320d4. For εἰς φῶς at the revelation of secrets compare e.g. Herm. 402, S. Ph. 581, fr. 918.

145

C O M M E N TA RY 1 4 αἱμυλομήτην: A hapax. Compounds in -μήτης (often -λο-μήτης) are based on ἀγκυλομήτης, originally used of ‘crooked-sickled’ Kronos. Since he used his sickle cunningly to castrate Ouranos, the stem was linked to μῆτις (see e.g. West 1966: 158); for μῆτις in general see Detienne and Vernant (1978). The stem αἱμυλ- usually applies to speech which is persuasive (often wheedling) but detrimental to the addressee; it was extended from speech to planning, through instances such as Prometheus’ verbal αἱμύλας μηχανάς to save the Titans.20 Persuasion by wheedling and defrauding is a basis of Hermes’ μῆτις, in Herm. (Introduction §3.3.5) and elsewhere: Pandora’s αἱμύλιοι λόγοι are furnished by Hermes, who has αἱμύλοι relatives in Sisyphus and Odysseus; Hesiod also aligns his αἱμύλα-coaxing woman with a φιλήτης, which seems to be an epithet of Hermes (214*).21 Indeed, Hermes’ family is heavily implicated in the broader nexus of -μήτης/-μητις compounds: cf. 155, 514 ποικιλομῆτα, 405 δολομῆτα, POxy. 5102 fr. 1.17 ]λόμη[τι]ν ἐπισταδὸν Ἑρμ[ (Hellenistic?); ἀγκυλομήτης Kronos was his grandfather; his father is called δολομῆτα and ποικιλομῆτα by Hera (Il. 1.540, H.Ap. 322), Sisyphus is αἰολομήτης (Hes. fr. 10a), Odysseus ποικιλομήτης, κακόμητις and δολομήτης (in epic and at E. Or. 1403, Q.S. 5.292); finally, Hermes has various links to Prometheus (Introduction §1.3), who is ἀγκυλομήτης twice in Hesiod and αἰπυμῆτα at PV 18. 14 ληϊστῆρ’, ἐλατῆρα βοῶν: Elsewhere in Epos ληϊστής, -ήρ and -ωρ are used in the plural of ‘pirates’; but for the land-based ληιστής see e.g. E. Alc. 766, S. OR 122, Ar. Ach. 1077, Th. 1.5.3, X. An. 6.1.8; ληΐς recurs in Herm. at 171, 330. An ἐλατήρ can simply be a chariot’s ‘driver’ (Il. x3, H.Ap. 232; cf. ἐλαύνω in Od. 3.422). But the stem is strongly attached in Herm. to the sense of cattle-rustling (Schenck zu Schweinsberg 102; for ἐλαύνω ‘drive off’ see also e.g. Il. 1.154). When the noun itself occurs at 265 and 377 Hermes is responding to Apollo’s accusations, and here ληϊστήρ precedes. See further below on 10–19* for how the juxtaposition of epithets in these lines develops meaning. Hermes’ activities stealing cows are surprisingly rare in extant sources, given that cows were a prime target for theft. Besides the myth which follows, compare Aesop 91 (Hermes tests Teiresias’ powers by stealing his cattle) and Hermes’ theft of Io, who has been turned into a cow (juxtaposed with our myth by the Brygos painter on Princeton y1990–2); at one remove, Autolycus uses Hermes’ gift of thievery to steal cows (Hes. frr. 66–7). PV 206. αἱμυλ- of speakers: Herm. 317, Solon fr. 11 the fox-like man, S. Aj. 388, E. fr. 715, etc.; used of Hermes in Nonnus D. 8.127. The context is often erotic as in e.g. Od. 1.56–7, Hes. Op. 373–4. Of planning: Ar. Lys. 1269 foxes, Pl. Lg. 7 823e bird-hunting. 21 Pandora: Hes. Op. 78. Sisyphus: Thgn. 704. Odysseus: e.g. Pi. N. 8.33, S. Aj. 388, E. fr. 715, Ps.-E. Rh. 498, 709. φιλήτης: Hes. Op. 375. 20

146

C O M M E N TA RY 1 5 – 1 6 ἡγήτορ’ ὀνείρων: For Hermes the dream-bringer see A.R. 4.1732–3 μνήσατ’ ἔπειτ’ Εὔφημος ὀνείρατος ἐννυχίοιο, | ἁζόμενος Μαίης υἷα κλυτόν (+ Σ οὗτος γὰρ ἐπὶ τῶν ὀνείρων ὁ θεός), Cornutus ND 16.7 ὀνείρους ἐπιπέμπειν ἔδοξε, the etymology of Ἀργεϊφόντης as ‘creator of bright appearances’ (Trypho Περὶ σχημάτων fr. 2), anon. De viribus herbarum 40–3 (the plant ‘Hermes’ finger’ induces prophetic dreams, including θέσφατα πάντα|, whose closest parallel is Herm. 472), Σ Od. 7.138. This function lies at the intersection of his roles as guide of the souls of the dead (at Od. 24.12 past the δῆμος ὀνείρων), messenger, and controller of sleep.22 15 νυκτὸς ὀπωπητῆρα, πυληδόκον: Both phrases are ambivalent, encapsulating the principle whereby a god in control of X governs both occurrences and preventions of X: Hermes is both thief and protector against thieves. ὀπωπητήρ is a hapax, replacing ὀπτήρ.23 For νυκτός ‘at night’, compare e.g. Od. 13.278 ἱκάνομεν ἐνθάδε νυκτός. The phrase simultaneously suggests a ‘night-scouting thief ’ (in Donne’s words) and a watchman on the lookout for thieves. Just so πυληδόκος, also a hapax. The analogy with ὁδοιδόκος (first extant at Polybius 13.8.2) suggests a brigand who ‘awaits at the gates’, for which Radermacher compared Aristophanes Av. 496–7 κἄρτι προκύπτω | ἔξω τείχους καὶ λωποδύτης παίει ῥοπάλωι με τὸ νῶτον; for thieves around the house-door (a less common sense of πύλαι), compare ΣV Ar. Pl. 1153b Στροφαῖον: ἐπωνυμία Ἑρμοῦ, παρὰ τὸ ταῖς θύραις ἱδρῦσθαι ἐπὶ φυλακῆι τῶν ἄλλων κλεπτῶν· οὗτοι γὰρ ὀπίσω τῶν θυρῶν εἰώθασι καὶ ἀναδύεσθαι καὶ ὅλως πανουργεύεσθαι. However, Hermes’ primary association with gates is as the προπύλαιος whose presumed role is to guard against unwelcome access.24 This role was not restricted to night-time, so one should punctuate before πυληδόκον. 15–16 ὃς τάχ’ ἔμελλεν | ἀμφανέειν κλυτὰ ἔργα μετ’ ἀθανάτοισι θεοῖσιν: Hermes will effect not only a visual (φαν-) but also an aural (κλυ-) revelation for the other gods, contrasting with his secretive, shadowy birth; for the

22

For Hermes’ control of sleep see 241–2*, 448–9*, and e.g. Il. 24.343–4, 445, Od. 5.47, 24.3; Apollod. Hist. FGrH 244 F129 relates the control of sleep and dreams to the practices of fashioning the feet of the bed in Hermes’ image (Od. 23.198), and of offering him the final libation before bed (so e.g. Philostr. Her. 10.8, Hld. Aith. 3.5.1). Herms have been found in bedrooms: Herter (1976) 219 n. 94. 23 Vergados 105 compared IG XIV 2557.11, but the supplement ὀπωπ[ητῆρα is impossible (Herzog 1937: 131, Siede 2005). ὀπωπέω is itself rare: ὀπωπήσασθαι occurs in Euphorion fr. 56.3, the imperfect at O.Arg. 183, 1022. 24 For Hermes and city gates, Farnell (1909) 66 is still useful. For Hermes and other doors, see 384*.

147

C O M M E N TA RY 1 7 – 1 8 importance of phrases like ‘among the immortals’, see 170*. κλυτὰ ἔργα may formulaically suggest ‘works’, i.e. products of craft (Il. 6.324, Od. 20.72, Sc. 297, 313): the lyre will turn out to be central to this meaning. However, the use of short-term prolepsis in τάχ’ ἔμελλεν aligns Hermes’ imminent revelation of ἔργα with the narrative’s imminent revelation of them (17–19), and this promotes the sense ‘deeds’ as well. The increased emphasis on sound in κλυτά (contrast 12*) inscribes the hymn’s narrative in oral tradition, as the latest vehicle of Hermes’ κλέος; the same metapoetic effect is evident in the passages of Sc. cited. 17 ἠῶιος: For adjectives replacing an adverbial temporal expression, see KG i.273–4, and 10–19* below for further comment on the structure of 17–18. γεγονώς: Earlier epic uses γεγαώς (x20); the participle based on the singular indicative stem γεγον- is first attested in Alcaeus fr. 72.11. μέσωι ἤματι ἐγκιθάριζεν: ἐν- is probably redundant: it is rare with times of day (e.g. A.R. 4.244 ἠοῖ ἐν τριτάτηι), nor will Hermes play ‘among’ the gods, as in the verb’s other ancient use at H.Ap. 201.25 Bergk (in Schulze 1868: 16) proposed εὖ. Herm.’s account shows great temporal compression. Contrast Hermes inventing the lyre on his third day in Σ Arat. 268, apparently after his first day in Eratosthenes Hermes fr. 13, and probably as an adult in Metiochus and Parthenope; in several artistic sources he is adult when he steals the cows (LIMC s.v. Hermes 245–8; see Hägg 1989: 49–51). While μέσωι ἤματι simply marks time here, it may be from our context that Theocritus repurposed the phrase for the trope of bucolic noon-time music-making at 6.4; cf. Rudoni (2015), arguing that Hermes’ music in Herm. is taken as proto-bucolic in Theoc. 7. Herm. uses κίθαρις and φόρμιγξ as well as χέλυς and λύρη, whereas in other texts the former pair entail wooden sound-boxes. West (1992) 50–1 discusses both the nomenclature and poetic licence in applying it. Hermes appears in art with a syrinx from c.575, and with auloi or a lyre from c.520 on (LIMC s.v. nos. 310–29). Herter (1976: 226) suggested that he played the syrinx first, and that his musicality derived from his pastoral role. Even if this is so, our ignorance about whether (for example) Alcaeus’ Hymn to Hermes included the lyre prevents us from dating the development. 18 ἑσπέριος βοῦς κλέψεν ἑκηβόλου Ἀπόλλωνος: For the epithet Ἑκηβόλος see 218*. βοῦς ... Ἀπόλλωνος raises two points of uncertainty about ancient audiences’ expectations: would the genitive readily be taken to indicate

25

I have argued that Francesco Filelfo derived his use of the verb (Psych. 2.14.52) from our line: Thomas (2016) 293.

148

C O M M E N TA RY 1 9 that Apollo owns the cows; and would the audience have been surprised at Hermes inventing the lyre before the theft? Apollo’s ownership of cows does not seem to have been a stable mythological idea. Some later reflexes of our myth betray a desire to specify why Apollo has cows to steal, using the stories that he served as herdsman to Admetus and (attested earlier, but more likely derivative) Laomedon: Lactantius Placidus Narr. 2.11 says that Hermes stole Admetus’ cows, though Ovid had not specified as much (Met. 2.680); so does Σ Nic. Alex. 560a; in Ant. Lib. 23.2 Apollo’s cows are with those of Admetus, and in Pherecydes FGrH 3 F131 Apollo’s exchange with Hermes occurs during this servitude. ΣVat. D.T. p. 174 conflates the cows with a herd belonging to Helios (cf. Od. 12, Hdt. 9.93), but Herm. avoids Apollo’s syncretism with him (381–2). By contrast, Philostr. Im. 1.26.3 seems to imagine cows dedicated (by humans) to Apollo. Apollo’s epikleseis which relate to herding are either entirely general (e.g. Νόμιος: IG IV 1080, A.R. 4.1218, Call. H.Ap. 47), or specific to smaller animals (e.g. Ἐπιμηλίδιος: Tit. Cam. 135, cf. Macr. Sat. 1.17.45), and he is rarely depicted with cows in art. The alternative interpretation of the genitive (also of 22, 102, 276, 310, 340 ἐμάς, and of the general storyline) is that Apollo has been entrusted with a herd owned by the gods, like epic’s other princely herdsmen, such as Anchises (H.Aphr. 55; compare Athena’s disguise at Od. 13.222); see also 556–7 for Apollo’s days of active herding. The natural inference at 71* θεῶν μακάρων βόες supports this view. At 493–5* Hermes frames his request in terms of taking over not sole ownership but day-to-day care, and Apollo answers in kind at 567–8 when he generalizes Hermes’ role to cover management of all cattle, horses and mules. As for the second question, the ordering where Hermes steals the cows first allows him to use their hide in his lyre: so Sophocles Ichn. 372–6, Apollodorus Bibl. 3.112, Ps.-Eratosthenes Cat. 24, Nonnus D. 1.338–9. The tidiness of that ordering has been taken as evidence both for its priority and for its derivativeness, but per se it is neither (Scheffler 1884: 13–14); see also 49*, 51*. 19 τετράδι τῆι προτέρηι τῆι μιν τέκε πότνια Μαῖα: The special name τετράς shows the symbolic importance assigned to the fourth of the month, perhaps owing to the difficulty of seeing the moon in its first three days.26 τετρὰς προτέρη is an unparalleled way of describing the date, but comprehensible since the common system had to distinguish fourth (τετρὰς ἱσταμένου) from fourth-from-end (τετρὰς φθίνοντος); a decadic system known from Hesiod and Argos distinguished the fourth, fourteenth and twenty-fourth as τετρὰς πρώτη, μέση and δευτάτη (Samuel 1972: 91). Aratus 781; this detail suits the phrasing of 11 μεὶς ... ἐστήρικτο. Thphr. Sign. 8 deems the fourth important in judging weather-patterns.

26

149

C O M M E N TA RY 1 0 – 1 9 Hesiod calls the fourth sacred (Op. 770); its particular associations are with Hermes, Aphrodite and Heracles.27 At least for Hermes and Heracles it was taken as a birthday: see Plutarch Q.Conv. 9.3 738f πολλοὶ δὲ καὶ τετράδι μηνὸς ἱσταμένου γενέσθαι τὸν θεὸν [i.e. Hermes] ἱστοροῦσιν. The model of Heracles, and also in Philochorus’ view of Hermes, created proverbial wisdom that people born on the fourth were destined for a laborious existence – a nuance which could well be active here.28 10–19 Revealing the plot: Of the initial thematic material noted in 1–9*, 12* picks up on secrecy and the twin axes of light/dark and famous/ unknown. This suggestion of a narrative trajectory is joined first by a series of epithets in asyndeton (unusually long by the standards of the Homeric Hymns). Such an appositional series prompts the question as to how closely to relate the elements. In this case, four agent-noun morphemes (-τῆρ’, -τῆρα, -τορ’, -τῆρα), reinforced by further assonance of eta (from -μήτην to πυληδόκον) and a chain of semantic links (brigand/rustler, drive/lead, dreams/night) act as centripetal forces in interpretation; νυκτὸς ὀπωπητῆρα and πυληδόκον are unified in their ambivalence. However, the superficially friendly confrontation of the αἱμυλομήτης is juxtaposed with the forceful confrontation normally involved in being a ληϊστήρ, while dream-bringing is bound up with Hermes’ ability to put people to sleep, and so to defraud them without any confrontation at all. These elements suggest the many forms of Hermes’ πολυτροπία, and leave it open that in Hermes’ specific case even the deeds of a ληϊστήρ might be undertaken subtly. Lines 16–18 next confirm the narrative path (up to line 74, as it will turn out) and thereby reveal the programmatic status of at least the epithets πολύτροπον, αἱμυλομήτην, ληϊστῆρ’, ἐλατῆρα βοῶν and νυκτὸς ὀπωπητῆρα. This process of revealing the narrative is aligned with Hermes being about to reveal famous deeds (15–16), which continues in a new vein the hymnist’s adumbration of a mimetic relationship with Hermes (cf. 3 ἄγγελον*; Introduction §3.4). Some form of proemic summary immediately before a narrative opening appears in H Dem. (2–3), each ‘half ’ of H.Ap. (25–7,

Σ Hes. Op. 770 (Heracles), 800, Thphr. Char. 16.10 (Hermes and Aphrodite), Alexis fr. 260 (τετραδισταί worshipping Aphrodite). According to Σvet. Ar. Plut. 1126, ἡ τετρὰς ἐνομίζετο τοῦ Ἑρμοῦ· καὶ καθ’ ἕκαστον μῆνα ταύτηι τῆι ἡμέραι ἀπετίθεντο [sc. cakes] τῶι Ἑρμῆι. Αt Erchia, Hermes’ sacrifice on Thargelion 4 involved notable perquisites for the herald (CGRN 52 E.47–58). Other connections between Hermes and the number four are made at Paus. Att. ε.71, Lydus Mens. 2.9. 28 Philochorus FGrH 328 F85b, versus Eust. Comm. Il. iv.913–14; see already Ameipsias fr. 27 teasing the early stages of Aristophanes’ career. For monthly birthday celebrations, see Schmidt (1908) 12–13, and 88–109 on divine birthdays. 27

150

C O M M E N TA RY 2 0 – 2 1 214–15), and H.Aphr. (45–53), though the level of detail varies. Herm. presents the summary in a very striking form in 17–18, where successive events are each assigned a solar time, which is placed first in the phrase and indicates the temporal compression of this version of the story.29 The whole is then dated in 19, with ring-composition both to the mention of lunar time in 11, and to the beginning of the introduction (cf. 3 τέκε Μαῖα|). Besides solar and lunar calendars, Hermes’ birth has been set within the time-frame of Zeus’s plan (10). The multiple time-keeping systems suggest already that time will be an important motif for articulating Hermes’ precocity.

20–9 20 ὃς καί: ὅς for demonstrative ὅ: compare Hy. 15.4, Ruijgh (1971) 321. καί throws emphasis (Denniston 307–8): ‘He didn’t lie around …’ The contrast to normal babies will continue in 21–2* (οὐκέτι, γ’). ἐπεὶ δή: Felt as two distinct words in early epic: see Wackernagel (1916) 32–3. μητρὸς ἀπ’ ἀθανάτων θόρε γυίων: Offspring also ‘spring off’ at e.g. Hes. Th. 281 ἐξέθορε Χρυσάωρ … καὶ Πήγασος ἵππος, H.Ap. 119 ἐκ δ’ ἔθορε (see Introduction §1.3), GDRK 51.10 ἀνὰ δ’ ἐξέθορες, Nonnus D. 38.146 ἀποθρώισκοντα λοχείης. The verb suits Hermes’ precocious mobility (Eitrem 1906: 250), though it is unclear that ἐξέθορε does so for Chrysaor, and later Maximus Init. 226 even applies the metaphor to a stillbirth. Servius on Aen. 8.139 comments that a fast birth was an omen of the child’s prosperity, and that Mercury’s was characteristic of his later speed (Scheffler 1884: 8). The Greeks had various birthing positions which might be imagined here: see H.Ap. 117 (kneeling), LIMC s.v. Leto 6 (seated, leaning forward), Call. H Del. 209 (seated, leaning backward), Gourevitch (1988) 45–7 (standing, squatting). 21–2 οὐκέτι δηρὸν ἔκειτο μένων ἱερῶι ἐνὶ λίκνωι, | ἀλλ’ ὅ γ’ ἀναΐξας: A λίκνον is an open-wickerwork basket without a lip on one side, etymologically 29

The phrasing seems to have attracted the attention of Call. at Epig. 32.1–2 G–P ἠῶιοι Μελάνιππον ἐθάπτομεν, ἠελίου δέ | δυομένου Βασιλὼ κάτθανε. Hobbes called these lines the ‘greatest’ of the hymn’s expressions of praise (Leviathan i.10 = Malcolm 2012: ii.142–3). There may be a folktale quality. Compare the Cretan song of Porphyris in Probonas (1987), τὴν ταχινὴν ἐγέννησε κι ἀργὰ τὸν ἐβαφτίσα | καὶ τ’ ἀποξημερώματα ζητᾷ ψωμὶ νὰ φάῃ, ‘She bore him in the morning and baptised him in the afternoon, and at first light he asks for bread to eat’ (Probonas, implausibly, saw a specific connection to Herm.); or Balázs’s libretto for Bluebeard’s Castle (‘The first [wife] I found at daybreak … The second one I found at noon … The third I found at evening… The fourth I found at midnight’: Hassall 1963).

151

C O M M E N TA RY 2 2 for winnowing (λικμάω, PIE *neik-), i.e. shaking so that the lighter chaff falls out while the grain remains inside. But cradles could also be constructed from wickerwork with a higher lip behind the head than at the feet, and be rocked from side to side (Galen San. Tu. 1.8). Few sources (Galen, Hsch. λ.1016) suggest that λίκνα-cradles were common; they mostly contain Hermes or Dionysus. Hermes’ λίκνον is depicted twice by the Brygos Painter (LIMC s.v. Hermes 242a; Princeton y1990–2) and on LIMC 242b (Attic, 430s); Herm.’s six mentions may well have influenced Sophocles Ichn. 275, Aratus 268, and Callimachus’ decision to place Zeus in a λίκνον at H.Zeus 48. It may be that actual winnowing-fans were used as cradles in order to symbolize fertility, as they were in agricultural dedications.30 οὐκέτι counters an expected follow-up (Wilson 1987), viz. that the newborn will lie still. Two verbs expressing Hermes’ movements (20 θόρε, 22 ἀναΐξας) frame two expressing ‘normal’ stillness (ἔκειτο μένων): this recalls the contrast of Maia staying in her cave while Zeus paid her visits (5–9), and lays further groundwork for Hermes’ idea that he needs to break out of Maia’s sedentary quietude.31 The mention of a λίκνον contributes to this: it is a means to distinguish the mobile from the weighty and valuable – a point which neatly suggests that, from Maia’s perspective, Hermes’ mobility is undesirable. Moreover, Greek cradles were often designed so that the baby could be strapped in place: so H.Ap. 128 (Apollo), Soranus Gyn. 2.14 (a ‘harsh’ practice known in Hellenistic Thessaly). A Greek audience may well have assumed that Hermes has been swaddled, and wondered also how he escaped that constraint: the point is clarified at 151*. The lack of description makes Hermes’ escapology uncannily effortless; Apollo, by contrast, snaps his cradle-straps (H.Ap. 128–9; Introduction §1.3). 22 ζήτει βόας Ἀπόλλωνος: Line 18 has just implied that the cows will occupy Hermes’ attentions in the evening: we are forced to remodel our expectations so that the lyre-episode becomes a ‘digression’. The jerky presentation obfuscates the traces of Hermes’ plan: unlike the ἔργα mentioned in 16–18, we are made to work this out gradually, and so to sense Hermes’ ability to wrong-foot and to outstrip expectation (Introduction §3.4). Philostratus Vit.

For Hermes’ λίκνον much later see Wroth (1892) pl. 17 no. 10 (coin of Hadriani, c.200 ce). For λίκνα in ritual, see Harrison (1903, esp. 313–17), interpreting their use as cradles as a fertility-charm or symbol of purification; Bérard (1976); Frazer (1925) 5–12. Σ Call. H.Zeus 48 gives an ancient interpretation: τὸ γὰρ παλαιὸν ἐν κοσκίνωι κατεκοίμιζον τὰ βρέφη πλοῦτον καὶ καρποὺς οἰωνιζόμενοι. Soranus Gyn. 2.16, 37 discusses cradle-design, but not λίκνα. On Aratus’ knowledge of Herm. see 11*, 42*. 31 See esp. 169–72. For a possible imitation with opposition of ἀναΐξας at H.Pan 38, see Thomas (2011) 167.

30

152

C O M M E N TA RY 2 3 – 2 4 Ap. 5.15, by contrast, locates the source of Hermes’ desire for Apollo’s cows in a cow-fable told to him by his nurses. 23 οὐδὸν ὑπερβαίνων ὑψηρεφέος ἄντροιο: Epos’ thresholds are normally raised, and hence crossed with καταβαίνω and ὑπερβαίνω (e.g. Od. 4.680, Hes. Th. 749–50); the height is magnified from a baby’s perspective. Moreover thresholds – in Greece as elsewhere – became freighted with symbolism as the boundary between ‘in’ and ‘out’; for example, one had to take care to cross a threshold ‘successfully’ (Ogle 1911, Lateiner 1995: 121–3). Hence ὑπερβαίνω marks not only an obstacle scaled, but also a transgression (a sense extant already at Il. 9.501 ὑπερβήηι καὶ ἁμαρτῆι) of where babies belong. For the impressive and surprising nature of the crossing see also 271–2*. Hermes’ control of traversing boundaries led the hymnist to elevate them to a structural device (Introduction §3.2–3). The cave’s threshold and high roof align it with epic palaces; indeed, the anomaly ὑψηρεφέοςˉ ἄντροιο is caused by formular modification: cf. Il. 9.582 οὐδοῦ ἐπεμβεβαὼς ὑψηρεφέος θαλάμοιο (in Oineus’ palace). For the threshold itself see further on 233–4*. The cave also has lockable doors (26, 146), a gated αὐλή (26), a main front room (μέγαρον: 65, 146), storerooms (247) with wealthy contents (249–50; also 61), and presumably maids’ quarters (see 60 ἀμφιπόλους). It is pleasantly fragrant (65, 231). It may be described as an οἶκος (61) or δῶμα (34, 60), but as a god’s home it is also a νηός (148; see 60–1*, 231*). For the alternative negative focalization of it, see 172*, 233–4, 359.32 24 ἔνθα χέλυν εὑρὼν ἐκτήσατο μυρίον ὄλβον: χέλυν εὑρών forms a chiasmus both with 22 ζήτει βόας (hence one naturally interprets ‘found a tortoise’, with ‘invented a lyre’ resonating in the background) and with ἐκτήσατο… ὄλβον. The doubled chiasmus suggests a cognitive to-and-fro: Hermes’ search first appears to end at an inferior animal, which is then paradoxically and hyperbolically esteemed.33 The switch in focus foretells

32

Parallel ‘domestic’ features in other epic caves (Herter (1981) 191–3) tell against Vergados’ idea (2011: 6–7) that the poet is creating an incongruous image of the cave which is to be explained as Hermes’ naive focalization. Analyst discussions of Herm. often located incompatibilities in the descriptions of the cave. The fact that it is described as both a cave and a house proved easier to defend (Ludwich 1887a: 321–2) than the divergent views about whether it is appealing or dank. 33 See Eitrem (1906) 250, Croci (1977–8) 177. The ‘wrong animal’ motif: cf. 232*. The Testudines attract paradox: the tortoise outstrips the hare (Aesop 254; cf. Zeno 29 A26) or crow (SEG 39.1377b.16); it weds a ram to give birth to the lyre (AP 14.30); Croesus makes an unguessable tortoise and lamb stew (Hdt. 1.47); the eagle offers the turtle fast flight (Babrius 115), and so on.

153

C O M M E N TA RY 2 5 the eventual exchange of lyre for cattle, but there is also something childlike in its apparent whimsy. Fortuitous finds are ἕρμαια in Greek, and so appropriate to Hermes.34 ὄλβος is predominantly prosperity, of any sort, bestowed on humans by gods; where gods are ὄλβιοι, material wealth is not the focus.35 But κτήσασθαι and its cognates are strongly materialistic, and given Hermes’ patronage of profit-making there is already room to suspect that the tortoise’s value will be economic. In that case, Hermes is humanized by his materialist aspirations: so it will prove (171*); cf. his humanizing dietary aspirations (64*). Krause (1790) 418 already compared Pausanias 8.17.5 on the hill named Chelydorea beside Kyllene, and 8.54.7 on Mount Parthenion further south, which παρέχεται ... ἐς λύρας ποίησιν χελώνας ἐπιτηδειοτάτας. Currently the local species are Testudo marginata and the smaller Testudo hermanni, neither of which ranges as high as Kyllene’s upper slopes (Hailey and Willemsen 2003). Lyres made from both species have been found (Dumoulin 1992: 101–4). Herm. always has χέλῡς/-ῡν before vowels (here, 33, 153), though χέλῠς/-ῠν is the norm in other texts.36 χέλῡν εὑρών in a late, possibly neo-Greek hymn to Hermes probably derives from our line.37 25 Ἑρμῆς τοι πρώτιστα χέλυν τεκτήνατ’ ἀοιδόν: Either ‘Hermes was the very first to engineer …’ or ‘First of all, Hermes engineered …’; the similar phrasing at 111* supports the former, for which compare e.g. Bacchylides 9.10–12 κε[ῖθι] ... ἡμίθεοι πρ[ώτιστ]ον ... ἄθλησαν. The renaming of both Hermes and χέλυς in this line creates a slight narrative discontinuity, which suggests that τοι (originally dative of σύ, often retaining a hint of ‘I tell/ 34

This relevance was well noted by Schneidewin (1848) 662 (in passing) and Baumeister 185, in a period when the analyst paradigm tended to take the ‘digressiveness’ of the lyre-episode as a sign of interpolation (e.g. Windisch 1867: 37). 35 E.g. E. Hipp. 1440 (freedom from death), Ar. Thesm. 107–29 (rich in praise), Ra. 452 (association with initiates); Zeus Olbios may be ‘bestower of prosperity’ (Cook 1940: 628–53). 36 Cf. KB i.438–9. Beekes (2010) s.v. suggests the ῡ may not be inherited, despite Slavic parallels; perhaps χελνη was influential. At Timotheus Pers. 222 ῡν is doubtful. χέλ[ῡες] is restored in SEG 39.1377b (Hierapolis, c.150 ce: read χελ[ῶναι]?), as is χρυσοχέλ[ῡ] in two metrically irregular hymns to Apollo from Talmis (Bernand 1969 nos. 167, 169; first–third century ce). 37 Cougny iv.68.3, following (with new errors) Piccolos (1853) 196. Piccolos got it from Laurentianus Plut. 57.29 f. 172v, where it is subsequent to the main act of copying c.1493. Line 3 describes the lyre as ἐρατή (cf. Herm. 153), 1–2 have the sequence , like Herm. 1–3, and Hermes’ gift of |εὐφροσύνην (7) may relate to Herm. 449, 482.

154

C O M M E N TA RY 2 6 remind you’) does here mark a ‘turn’ to engage the external narratee.38 That engagement is a challenge to match Hermes’ imagination. After the paradox of v. 24, this line is set up as an asyndetic explanation for the narratee; the jingle ἐκτήσατο ... τεκτήνατ’ underscores the fact that an idea is being restated. But the ‘explanation’ replaces one paradox with two others. τεκταίνω does not normally have natural categories such as tortoises and singers as its object, and a tortoise’s squeaks and sibilants (Arist. HA 4.536a8) are not promising repertoire for a singer. For the latter paradox see Sappho fr. 118 (probably – the text is corrupt), Nicander Alex. 560 αὐδήεσσαν ἔθηκεν ἀναύδητόν περ ἐοῦσαν, Horace C. 3.11.3–5 testudo… nec loquax olim.39 Contrast Herm. 38 (whence Sophocles Ichn. 300), Pacuvius Antiope 4 (perhaps from Sophocles), Archias 21 G–P, and Symphosius 20, which focus more on the change occurring at death; Borthwick (1970) tracks that paradox across different centuries and countries. Call. H.Ap. 16 χέλυς οὐκέτ’ ἄεργος succinctly takes the idea in a new direction. For the Indo-European metaphor of ‘crafting song’, see West (2007) 34–43. Here it is particularly in place because of Hermes’ skill at invention and crafts: Introduction §3.3.6. It may be that ‘Boeo’ fr. 2.2 πρῶτος ... τεκτάνατ’ ἀοιδάν reveals formulaic background which Herm. is modifying to express the oddness of Hermes’ craft, but derivation from Herm. is also possible, and marginally supported by the fact that the context was a hexameter hymn to Apollo which rewrote the myth of the oracle’s foundation as told in the Hymn to Apollo. 26 ἥ ῥά οἱ ἀντεβόλησεν: In ἀντιβολέω the deictic implication of βολ-, that the subject is directed towards the indirect object, is often focalized in a surprising way.40 The verb of meeting, suitable for Hermes’ sudden arrivals (e.g. Il. 24.375, Od. 10.277, A.R. 4.121), is here configured as if the grazing tortoise

The conversational nature of τοι is shown by its near-confinement to direct speech in Homeric narrative (exceptions: δή τοι at Il. 10.316 ~ Od. 20.289: see Denniston 537). Several times in Hes. Th. the primary narrator uses τοι in assertions about the gods (e.g. 94, 126, 448). For noun-repetition and discourse coherence see Gordon and Chan (1995), though none of their experiments quite matches the situation here. 39 Nicander’s passage shows more sustained engagement with Herm.: the tortoise eats vegetation in the mountain (Alex. 559 οὐρείης κυτησινόμου ~ Herm. 27 + 33); Hermes separated the flesh from the shell which was αἰόλος (Alex. 562 ~ Herm. 41–2 + 33), then inserted two ‘arms’ (Alex. 562 ἀγκῶνας ~ Herm. 50 πήχυς). 40 LfgrE s.v. 1a; a more extreme example is Sc. 439 πάγος δέ οἱ [= a moving boulder] ἀντεβόλησεν; at Od. 18.272 Penelope says ‘marriage ἀντιβολήσει me’ to imply that she is not moving to greet it. 38

155

C O M M E N TA RY 2 7 – 2 9 were the mobile one – a further touch of paradox, which prepares for the tortoise developing into an omen (30*, also with -βολ-). ἐπ’ αὐλείηισι θύρηισιν: A formula (Od. 18.239, 23.49) meaning ‘at/just beyond the doors between the αὐλή and the outside’.41 The epic αὐλή is a fenced, open-air courtyard in front of a palace; its presence here, as outside Polyphemus’ home (Od. 9.184), aggrandizes the cave. For the second time in succession, Hermes’ progress is mapped against boundaries (cf. 23*; Christopoulos 2006: 311). Early Ionic generally followed a convention of adding paragogic nu before the pause at line-end; from around the time of Aristarchus, it was added only when the next line began with a vowel.42 27 βοσκομένη προπάροιθε δόμων ἐριθηλέα ποίην: βοσκομένη is associated with herding; this again emphasizes that the tortoise is an alternative to the herd of cows (see 24*; also 72*). The transitive middle occurs first elsewhere in Anacreon PMG 417.5; see DGE s.v. βόσκω B I.2. In the Hellenistic novel Metiochus and Parthenope Metiochus gave a different story: Hermes found the carcass of a tortoise and was inspired by the wind singing in its sinews (Hägg and Utas 2003). 28 σαῦλα ποσὶν βαίνουσα: The stem σαυλ- refers to a gait with a mobile torso; its main connotations are vanity and flirtatiousness. See Thomas (2015) for sources, adding Lucian Lexiph. 10, where a dandy, formerly σαυλούμενος (‘prancing’), is now immobile. With respect to a tortoise, the underlying point must be swaying the body. Herm. strikingly avoids the characterization of the tortoise’s gait as a paradigm of sluggishness (e.g. Arist. HA 2.503b9), in favour of two connotations which seem paradoxical. That of flirtatiousness will be developed brilliantly in 31* and throughout the hymn, as Sophocles appreciated (Thomas 2015). Hsch. σ.266 glosses ἥσυχα: of extant passages, a simplistic interpretation of our line is the most likely source. ἐριούνιος: 3*, 30 ὀνήσιμον*. 29 ἐγέλασσε: Hermes’ expression unites at least two types of joy: that of an infant faced with something to play with, and, as we are about to find out (30*), the joy of seeing a positive omen.43 This is the first of Herm.’s four For the manuscripts’ αὐλίηισι cf. ΣH Od. 18.239. See Bolling (1945), West (1998) xxv–xxvi. Hermann first restored such nus in the Homeric Hymns. There is sporadic manuscript support, in M at 111, 295 (also Πq), 356, 502, 521, 556, in P at 49, 162, 553, in q at 260. 43 For the first type see e.g. Aeschylus fr. 47a.786 ~ Sophocles fr. 171, Halliwell (2008) 551, Rühfel (1984) 176–84. For the second, Od. 17.542 γέλασσε and LfgrE s.v. γηθέω I 1c. 41

42

156

C O M M E N TA RY 3 0 c­arefully patterned instances of the stem γελ-: see 281*, 389*, 420*, and Bungard (2011). The stem covers both ‘laugh’ and ‘smile’. For the verb, an audible quality is rarely made explicit in Epos, where in particular with inanimate subjects visual facets of calmness and brightness, and their psychological implications, are often to the fore (e.g. Il. 19.362–3 γέλασσε δὲ πᾶσα περὶ χθών | χαλκοῦ ὑπὸ στεροπῆς). However, audibility is implied in most uses of γέλως and in Hesiod Th. 40–1 γελᾶι δέ τε δώματα ... θεᾶν ὀπὶ λειριοέσσηι. See the different interpretations in LfgrE s.v. γελάω, γέλως, Arnould (1990) 138–41 (demoting the auditory), Clarke (2005) 39–45, Halliwell (2008) 13–15, 522–7 (recuperating the auditory). For the interpenetration of brightness and joy elsewhere in Greek see Latacz (1966). 20–9 An unexpected meeting: After the summary in 17–18 and the closure of the introduction in 19, the narrative proper begins, and is rife with surprises. Lines 20–2 emphasize Hermes’ mobility, which is extraordinary for a baby; this culminates in a movement which draws our attention to boundaries (23, 26) being a favoured site for Hermes’ activities – one which will recur throughout the hymn (Introduction §3.3.6). The poet also designed this passage so that Hermes eludes our expectations. He looks for the cows in the morning, precociously in advance of the expectation set up by 18 ἑσπέριος βοῦς κλέψεν. The hymnist implicitly contrasts our limitations in processing the unexpected plot-shift with Hermes, for whom the tortoise is a surprising ἕρμαιον which he takes in his stride. Is Hermes interested in cows and tortoises from a childishly short attention-span, or with un-childish ulterior motives and flexible planning? The hint towards the latter conclusion takes the form of further surprises – the tortoise constitutes ὄλβος and has σαῦλα steps (24, 28).

30–8 30 σύμβολον ἤδη μοι μέγ’ ὀνήσιμον· οὐκ ὀνοτάζω: One should understand ἔστιν, then punctuate after ὀνήσιμον. For verbless comments on new sense-­ input, cf. Ar. Eq. 1326 καὶ γὰρ ἀνοιγνυμένων ψόφος ἤδη τῶν προπυλαίων, ‘In fact, already there’s the sound of …’. Hermes’ find is unexpected, so it is natural to take ἤδη as ‘already’ (~ sooner than expected) rather than the rarer ‘for now’; other senses do not come into play. If that is right, ἤδη cannot qualify ὀνήσιμον, and removing the punctuation after ὀνήσιμον affords no further alternatives. ἀντι-βολ-έω in 26 primes us to examine σύμ-βολ-ον carefully; three ­associations are especially relevant. (i) On Hermes and unexpected helpful meetings see 24*, 26*. (ii) A σύμβολον is an omen, normally at the beginning of an enterprise. It was often a bird, but various other animals could be

157

C O M M E N TA RY 3 0 deemed significant, as could chance remarks, sneezes, etc.44 (iii) A σύμβολον is one half of a pair of tallies; from this develops (iii.a) the sense of a token (including figurative language) requiring interpretation using a non-public code. Given our knowledge that the tortoise will become a lyre, and the paradox of 28 σαῦλα, this meaning of σύμβολον may suggest – as his subsequent words confirm – that Hermes’ insight sees some hidden complement beyond the tortoise with its filled shell opening towards the ground – namely the tortoiseshell lyre and the courtesan who uses it (Kahn 1978: 124). A σύμβολον is also (iii.b) a proof of identity and (iii.c) a means of admission, and the lyre will indeed be a means of admission to Olympus and a proof to Apollo of Hermes’ characteristic abilities at engineering, ingenuity, bargaining, and so on.45 Hermes’ ambivalent use of a word from semiotics is a rich touch of characterization. That this is (for us) Hermes’ first word enhances its characterizing power in three ways. First, there was a literary convention that speakers’ opening gambits encapsulate their interests: see e.g. the Corcyreans and Corinthians at Thucydides 1.32, 37. Secondly, a baby’s first word was a matter of discussion in antiquity (Thomas 2010: 197–8); ordinary babies, of course, do not say anything phonologically or semantically like σύμβολον. Thirdly, one location where Hermes combines linguistic and prophetic σύμβολα is where the first utterance one hears after visiting his sanctuary is prophetic (Pharai: Pausanias 7.22.2–3): Hermes’ first word, σύμβολον, foretells his control of ambivalent linguistic σύμβολα, but also – if the Pharai cult had early parallels – his control of ominous σύμβολα by which first words acquire predictive power. Omens seem rarely to be described in terms of ὄνησις. This reinforces the hint in 24* ἐκτήσατο … ὄλβον that Hermes intends to treat the tortoise not just as a sign, but as a resource. The end of the line adds οὐκ ὀνοτάζω, a variation on δέχομαι as the formal declaration that one accepts an omen (e.g. A. Ag. 1653, Ar. Pl. 63, X. An. 1.8.17, D.H. Ant. Rom. 19.5.3; Powell 1938: 81). The resulting alliteration ὀν-ήσιμον … ὀν-οτάζω (cf. Iliad 17.24–5 οὐδ’…

44

There seems to be no particular lore about tortoises in this context. Some examples of σύμβολα: Il. 10.274 (heron), 24.290–321 (eagle), Od. 15.160 (eagle), Archil. fr. 218 (probably human), Pi. O. 12.7, A. Ag. 144 (birds), fr. 36, Ps.-A. PV 487, Ar. Av. 721 (with Dunbar 1995: one’s slaves; an ass), Ra. 196, Ec. 792 (a γαλῆ augurs badly – similarly Thphr. Char. 16.3), X. Mem. 1.1.3–4 ~ Ap. 13, Luc. Pseudolog. 17 (people lame in the right leg, priests and monkeys augur badly); also Horace C. 3.27.2–5 with Nisbet and Rudd (2004). 45 For σύμβολα as tallies and ‘complementary halves’, see 527* and e.g. Democritus(?) 68 B5 in Diodorus 1.8.3 (of the signifier and signifié), Eur. Or. 1130 (‘watchword’). On the word σύμβολον in general, see Müri (1931), Falus (1981) 117–31, Struck (2004) 78–94.

158

C O M M E N TA RY 3 1 ἀπόνηθ’ ὅτε μ’ ὤνατο) probably initiates a wordplay, continued in 35, explaining the hymnist’s interpretation of ἐριούνιος from ἐρι- + ὀνίνημι (3*); Reece (1997) 37–9 attractively saw this not just in terms of the tortoise’s similarity to Hermes (see n. 55 below), but also in terms of the charis-cycle on which the hymn is built, with the lyre (now belonging to hymnists) benefiting Hermes, who in turn benefits worshippers. 31 χαῖρε, φυὴν ἐρόεσσα, χοροιτύπε, δαιτὸς ἑταίρη: Similarly at Od. 13.228–9 Odysseus says χαῖρε to the youth (Athena in disguise) since the first person one comes across (ἀντιβολεῖν) is auspicious; at Aeschylus Ag. 22 the watchman welcomes the beacon-signal with χαῖρε and treats it as a sign of the future. However, the greeting is given a more literal force than normal by the rest of the line, which signals the tortoise’s relationship to χάρις and gradually clarifies 28 σαῦλα* as an idiosyncratic focalization produced by Hermes’ racy imagination. φυὴν ἐρόεσσα¦ crosses Hesiod’s φυὴν ἐρατή¦ (Th. 259, 355) and ἐρόεσσα¦ (Th. 251, 357), boldly transferring them from nymphs to a wizened tortoise; indeed φυή, referring to natural or underlying corporeal qualities including species-characteristics, itself suggests that tortoises and nymphs belong in different categories.46 χοροιτύπε then extends the paradox by relating the tortoise’s lumbering motion not just to a hip-swaying walk as in 28 but to energetic dancing. τύπτω normally describes blunt impacts.47 Its application to dancers’ feet on the ground is expressively unusual in the earliest use of χοροιτυπ- (Il. 24.261), where Priam is opposing the impacts of choral dancing to those in the ‘dance’ of war. Cf. Il. 23.764 ἴχνια τύπτε πόδεσσιν, Od. 8.264 πέπληγον δὲ χορὸν θεῖον ποσίν, Ar. Av. 1145 ὑποτύπτω ‘splash with the feet’, Aratus 950 ‘imprint’ the ground. The reference to group dancing maintains the connotations of physical allure, but also introduces music and the tortoise’s future life as a lyre. This hint is brought forward in δαιτὸς ἑταίρη, given Od. 17.270–1 φόρμιγξ … ἣν ἄρα δαιτὶ θεοὶ ποίησαν ἑταίρην – a divine action which Herm. could be glossing cheekily.48 Here, unlike in the Odyssey, the other sense of ἑταίρη, ‘courtesan’, cannot be suppressed since the tortoise has just been sexualized. Such a

For φυή see 331* and esp. Od. 8.134–7, where it contrasts with the t­ emporary physical ravages of misfortune; for ‘zoological’ uses see e.g. S. Ichn. 307, A.R. 2.1237, 1254, Moschus Eur. 45. 47 It seems not to be used of strumming a lyre, so Matthiae’s χοροίτυπε ‘struck in the chorus’ cannot stand. 48 The lyre is also among ἀναθήματα δαιτός (Od. 1.152–3), or is δαιτὶ συνήορος (8.99). Cf. IG XII.5 ii.893 (Tenos, c.200 ce) πη]κτ̣  ί̣ δα δ’ οὐρανίων ἑτάρην θαλίης τ[ε χορῶν τε, and the literal companions in Thgn. 115 πόσιος καὶ βρώσιος … ἑταῖροι. For further metaphorical companions in Herm. see 290*, 358, 436*. 46

159

C O M M E N TA RY 3 2 courtesan might both play the lyre and dance at a symposium – the successor to the kind of indoor banquet imagined in the Odyssey.49 Panyassis fr. 16, with its reference to χοροιτυπία (15) caused by inebriation at a δαίς, probably refers to such an occasion. δαίς can also refer to public festivals with ritualized dancing, and courtesans sometimes attended these (e.g. Ps.-Dem. 59.21); however, they called for instruments louder than the tortoiseshell lyre. In short, Hermes’ words create an open-ended space encompassing tortoise, lyre, dancer, lyre-playing danseuse, and different occasions (particularly the symposium) which ground the connection between the four. The allure of music (notably expressed in the ἐρόεις mouth of minstrels at Hy. 32.19–20) means it was often figured in terms of mistresses or prostitutes, at least in humorous sources.50 The thought will continue in Herm. with Apollo’s highly eroticized response to the instrument and Hermes’ description of it (421*–34, 482–8*). The humour is enhanced by Hermes’ inventive inclusion of the tortoise in this nexus, and it sets the tone for Herm.’s playful view of lyre-music (see also 32* ἄθυρμα, 54*, Thomas 2018). 32 ἀσπασίη προφανεῖσα: The tortoise ‘came forth into view’, and the lyre ‘was shown in advance’; προφαίνομαι does not mean ‘appear in front of ’. The Odyssey makes a theme out of associating the stems ἀσπασ- and φαιν(5.394, 9.466, 23.60, 233). πόθεν τόδε καλὸν ἄθυρμα; : Krause (1790: 418) rightly placed the question-mark here. There do not seem to be parallels for taking πόθεν τόδε as a colloquial ‘Whence [sc. do you get] here?’, nor (at least in Epos) for punctuating πόθεν ... ἐσσί (or ἐστί) as a single question with αἰόλον ὄστρακον in apposition to καλὸν ἄθυρμα and the verb delayed to the end of its clause.51 The adult humour of δαιτὸς ἑταίρη is complemented by viewing the tortoise (strictly its shell, assuming that Hermes is continuing to address the tortoise) as a child’s plaything. An Apulian chous (British Museum F101; 350s bce) depicts a girl dangling a tortoise before her dog. Oppian (Hal. 5.403–7)

49

For courtesans with lyres in sympotic and comastic scenes from Athens c.515–450 see Peschel (1987) nos. 8, 40, 151, 194–6, 198, 203. For a later ­period see Ps.-Arist. Ath. Pol. 50.2 on hiring κιθαριστρίαι. χοροιτύπε ... ἑταίρη might be designed to evoke χαμαιτύπη, though that stem is only attested from the fourth century on. 50 E.g. Pherecrates fr. 155 (composers as Music’s abusive boyfriends), Ar. Ra. 1304–6 (Euripides’ tawdry muse), Call. Ia. 8 fr. 222 (Simonides’ ‘working girl’). For parallels in other societies see Solie (1993) 13 on music as a male performer’s ‘mistress’; Zecher (2007) 131–61, esp. 158, with the gender-roles reversed. 51 The long delay of ἔστι at Od. 17.159 (after several predicates, rather than an apposition) is different, because designed for heavy emphasis. Regarding a colloquial πόθεν τόδε, perhaps compare Pl. Phdr. 227a ποῖ δὴ καὶ πόθεν;

160

C O M M E N TA RY 3 3 – 3 4 depicts a boy playing with a tortoise. Apollo gets close to Dryope by disguising himself as a tortoise (Ant. Lib. 32). Girls played a game in which one was the ‘torty-tortoise’ (chelichelona: Erinna SH 401.16, Pollux 9.125).52 An audience might have made a contrast with the ἀθύρματα of other young gods. The Titans killed Dionysus by distracting him with toys (Orphic frr. 306, 578 i.28 – the latter from the third century); Persephone’s καλὸν ἄθυρμα (the narcissus at H Dem. 16) threatens her childhood. Hermes’ apparent ‘distractedness’, however, is already part of a plan.53 33 αἰόλος ὄστρακόν ἐσσι: The manuscripts’ αἰόλον ὄστρακόν ἐσσι is a blatant false equation, even if it is the shell that Hermes has his eye on particularly. I suggest either αἰόλος (also feminine in Nic. Ther. 173) with ὄστρακον as an accusative of respect, or αἰόλον ὄστρακόν ἐστι. Both explain καλόν: hence the asyndeton.54 Tortoiseshell is often lovingly depicted in vase-paintings of lyres, from the Analatos hydria on, but the phrasing is also richly suggestive of the nexus of ideas Hermes is developing. αἰόλος connects tortoise and lyre, since it can describe ‘variegation’ in melody, including lyre-music.55 The addition of ὄστρακον connects the dappled tortoise to the decorated pottery vessels of the δαίς whose ἑταίρη it is (31). Finally, after ἄθυρμα, ὄστρακον may suggest the boys’ game played by rolling a potsherd (Pollux 9.111; allusions already at Plato Phdr. 241b, R. 7 521c). χέλυς ὄρεσι ζώουσα: Hermes situates the tortoise paradoxically far from the urbane spaces of courtesan-led lyric monody; it will make the transition from mountainside to civilization, as he hopes to do himself (cf. 1–9*). 34 ἀλλ’ οἴσω σ’ ἐς δῶμα λαβών: ἀλλά switches from present musings to future plan (see Denniston 8), and from πόθεν to the apparent assertion of 52

Alcaeus fr. 359 may refer to children playing with a turtle, but the animal referred to was disputed in antiquity. 53 At A.R. 3.132 Aphrodite claims that the baby Zeus played with a marvellous ball, Διὸς περικαλλὲς ἄθυρμα. See Introduction §4.2. 54 Matthiae (1800) 214 suggested ἕσσο for ἐσσί. This had astonishing popularity, even after Agar (1921) pointed out that the pluperfect cannot mean ‘you are wearing’. Few other verbs are short enough. Gemoll accepted ἐσσί and understood 33 to flow into 34: ‘You are (only) a dappled shell … but I will …’. But its dappling makes it καλόν and so is a mark of esteem. 55 Eitrem (1906) 250. See S. Ichn. 327 αἰόλισμα τῆς λ[ύ]ρ̣ας̣ ̣ (near other i­ntertexts with Herm.: Introduction §1.3); Theoc. 16.44 (Simonides αἰόλα φωνέων in sympotic lyric; see Gow 1950 ad loc.); Limenius i.15–16 (set to a ‘chromatic’ run). Since I am hesitant about the text, I merely footnote the possibility that αἰόλος, prominently placed, could also link the tortoise to Hermes’ versatility (Kahn 1978: 123–4). For the lyre’s similarity to Hermes elsewhere, see on ὄρεσι ζώουσα, 35*, 54–5*, 63*, 290*, 293*, 436*, 478*, 488*.

161

C O M M E N TA RY 3 4 – 3 6 hospitality in ἐς δῶμα. The shift ἀλλ᾽ οἴσω – and perhaps originally the rhapsodes’ gestures – cues us to imagine Hermes picking up the tortoise; 39 ἅμ᾽ suggests that this happened during his speech. 34–5 ὄφελός τί μοι ἔσσεαι, | οὐδ’ ἀπατιμήσω: I marginally prefer Matthiae’s emendation to Ω’s ἀποτιμήσω. The latter verb normally means ‘do a valuation of, mortgage’, but is transmitted in the sense ‘dishonour’ at Plutarch Q.Conv. 5.3 677b (quoting Callimachus). ἀπατιμάω is unambiguous, and transmitted in Epos at Il. 13.113. The pair of poems in IG XIV 1389 (Rome, 166 ce) are split between ἀπητίμησεν (i.54) and ἀποτιμήσοι (ii.33). The phrasing recalls 30 σύμβολον ... μοι ὀνήσιμον· οὐκ ὀνοτάζω, but now the context places more emphasis on reciprocity. Hermes’ rationale is not simply to acquire something helpful. He will not disrespect the tortoise’s favour, but offers prompt hospitality (34) and protection (36*). σὺ δέ με πρώτιστον ὀνήσεις: Hermes expands on ὄφελός τί μοι ἔσσεαι, with emphasis on adverbial πρώτιστον. The repetition of πρώτιστ- from 25 (Hermes was first to make a tortoise sing) suggests that the tortoise will benefit Hermes by singing – but the poet still keeps the mechanism implicit. 36 ‘οἴκοι βέλτερον εἶναι, ἐπεὶ βλαβερὸν τὸ θύρηφιν’: The sentiment ‘Better to be at home’ may be proverbial (Pearson 1917 iii.102), but the precise phrasing is found only at Hes. Op. 365. The context creates various subtle differences. In Hesiod, βλαβερὸν τὸ θύρηφιν contrasts with 364 οὐδὲ τό γ’ εἰν οἴκωι ... κήδει, referring to the potential damage an asset left outdoors can do its owner. In Herm., the line explains what has preceded. (It cannot mark a new departure, since 37–8, with γάρ, explain 34–5, not 36.) Hermes’ point is therefore to offer the tortoise security against ‘the outdoors’ (article + adverb, apparently a post-archaic usage of ὁ) or ‘being outdoors’ (with εἶναι understood from the first half of the line: so Ilgen). The deviation from Hesiod underscores Hermes’ claim not to be treating the tortoise merely as a possession. Moreover, the hymnist introduces an elegant irony: Hermes is addressing a creature so agoraphobic that it carries its home with it and proverbially says οἶκος φίλος, οἶκος ἄριστος (Cercidas fr. 2, Aesop 108, App. Prov. 4.15), whereas Hermes is archetypally ‘agoraphilic’.56 Hermes’ choice of citation shows his ability to adopt another’s perspective and to use a γνώμη rhetorically to encourage the tortoise’s consent.57 See further 170 βέλτερον*. Hermann drily commented ‘ipse, immemor huius praecepti, in Pieriam proficiscitur’ (Ilgen 365); Baumeister already considered a connection to the tortoise being a stay-at-home. 57 Schenck zu Schweinsberg observed that recalling Hesiod’s comment just before – ἅρπαξ δὲ κακὴ θανατοῖο δότειρα (Op. 356) – would add to the irony of Hermes’ rheto56

162

C O M M E N TA RY 3 7 – 8 The line attracted a scholion: ση(μείωσαι) τὸν Ἡσίοδον κλέψαντα (b) / κεκλοφότα (P) τὸν στίχον. The scholion on H.Ap. 71 is also common to bP. For the assumption of Homeric authorship (and hence priority over Hesiod) in b’s philological marginalia see Thomas (2016) 282. 37–8 For the sense of ἔχμα + genitive, compare A.R. 4.201 ἔχμα βολάων. Both passages have parallel manuscript corruptions of ε to αι, which sounded the same in Byzantine Greek; compare Hsch. α.2200 αἴχματα (i.e. ἔχματα): ἐχέματα, κωλύματα. Two basic implications seem possible. The first, which has been neglected, is that the tortoise will restrain Hermes from attacking Apollo’s possessions, but only for the remainder of its life; the dead tortoise will benefit him. Alternatively, the tortoise will help Hermes already while alive by offering a form of protection, in return for Hermes’ offering it security indoors (36). Either way, though more outrageously in the latter case, Hermes continues to conceal his intention to kill the tortoise immediately. On that interpretation, ἐπηλυσίη is probably an ‘attack’ on health, whether or not a personified ‘attacker’ is implied. The most specific source, Pollux 4.187, places ἐπηλυσία amid illnesses involving inflammation of internal organs; at H Dem. 228–30, Demeter can protect Demophon against ἐπηλυσίης ¦ πολυπήμονος. The intertext could suggest that Hermes is precociously taking charge of his own nursing.58 Most medicinal uses of tortoises (e.g. Pliny Nat. 32.32–4, Dsc. Mat. 2.78.4–79.2) involve the animal’s death, which suggests that ζώουσ’ in runover position would be a paradox to complement that of the remainder of line 38 (Fernández Delgado 1990: 203). But living tortoises may well have had apotropaic values about which we are ignorant.59 For the dead singer paradox see 25*. 30–8 Hermes to the tortoise: Hermes’ volubility is displayed by manipulation of polysemous words (30 σύμβολον, 31 ἑταίρη, 33 αἰόλος and ὄστρακον), imaginative and witty transfer of language (31 φυὴν ἐρόεσσα, χοροιτύπε), wordplay (30 ὀνήσιμον ... ὀνοτάζω), paradoxical ideas (38), rhetorical use of quotation (36) and deception (35–8). At a stylistic level, the feminine caesuras

ric. Tzifopoulos (2000) further analysed Hermes’ use of adages, which as he noted (p. 149) are not a common feature of the Homeric Hymns. 58 The passage of H.Dem. is controversial: Richardson (1974) took ἐπηλυσίη as a demonic attack on a child’s health; Kledt (1996) understood ‘incantation’ based on Hsch. ε.4556, which may be corrupt; Faraone (2001) saw demonic attacks on health in the neighbouring words too; Leclerc (2006) understood an ‘attack’ not specific to health. For intertexts between Herm. and H.Dem. see Introduction §1.3. 59 An apotropaic use of a living turtle (‘marsh-tortoise’), to keep hail from a vineyard, is found in Gp. 1.14.8–10 < Sextus Iulius Africanus; structuralist analysis in Svenbro (1992) 155–8. For Roman apotropaic tortoises see Jahn (1855) 41, 98–9, 106.

163

C O M M E N TA RY 3 9 in 31–3 after -εσσα, -εῖσα and ἐσσι suggest a sing-song quality in performance, the doubled vocatives there may be a hint of childish language (see Thomas 2010: 209–11), the repeated omission of ἐστι (30, 32, 36) and the asyndeta in each of lines 30–4 seem fluent and chatty, while three further copulas (33, 34, 37) and several short clauses could suggest syntactic simplicity. However, these features are juxtaposed with the copious ways to describe a tortoise in 31–2, and a rich vein of unusual vocabulary.60 The shiftiness of the style may be part of the humour, as in the register-shifts characteristic of Aristophanes (Silk 2000: 35–6, 111–20). Hermes’ use of language obviously outstrips his age, just as his range of references combines child’s play and prostitution (32, 31); indeed, even qua toy, the tortoise is part of a strategy rather than a simple distraction. The logic betrays a similar duality: Hermes’ swift alternation between welcoming the tortoise and noting its utility can seem unfocused, but it brings out a possible strand of self-consciousness as he gradually strengthens the implication that the tortoise will also benefit from being welcomed – as if the child’s eagerness to grab the ἕρμαιον were really part of men’s code of xenia. This (tendentious) presentation of their relationship perhaps foreshadows Hermes’ hope to enter Olympus with Apollo’s support, after considerations of his usefulness have led to a friendship based on exchange of τιμαί and χάρις.

39–51 39 χερσὶν ἅμ’ ἀμφοτέρηισιν ἀείρας: The painterly detail ἀμφοτέρηισιν suggests Hermes’ care. ἅμα might reinforce it (cf. H Dem. 15–16 ὠρέξατο χερσὶν ἅμ’ ἄμφω(?) | καλὸν ἄθυρμα λαβεῖν), but is more effective if it signals that Hermes picked the tortoise up while still speaking (see 34*). Dumoulin (1994) 41 cites a statue of 450–425 bce (reconstructed from Roman copies) of an older Hermes raising the tortoise in his right hand. Another statue depicting 60

Compare Vergados 23; on p. 250 he related Hermes’ need for rhetoric to the proverb ‘a shell from a tortoise’, used of prising things from people who will not want to share them – but the proverb is not attested until the later Byzantine period. Brillante (1999, esp. 98–100, 105–7) identified specifically ‘Hermetic’ qualities of the speech in how the addressee flickers between being a tortoise and a lyre, and in the prominence of reversals (living in mountains/home; living/dead; quiet/voluble). The use of asyndeta continues in character-text (Hermes at 170, 263–4, 267, 269, 273, 370, 378, 466, 472, 478, 482; Apollo at 199, 336, 363, 406, 438, 447, 567; Maia at 160; Zeus at 332); see Radermacher 230–3. The narrator-text is more in line with other epic, though see 151*.

164

C O M M E N TA RY 4 0 – 4 1 this moment existed at Argos (Paus. 2.19.7 Ἑρμῆς ἐς λύρας ποίησιν χελώνην ἡιρηκώς).61 The action will be echoed at 295*. 40 ἂψ εἴσω κίε δῶμα: Cantilena (1982) 244 noted the sonic echoes which mark the fulfilment of Hermes’ plan in 34 ἀλλ’ οἴσω σ’ ἐς δῶμα. φέρων ἐρατεινὸν ἄθυρμα: The adjective is rarely applied to objects (though its application to a girl’s ball in anon. 41 G–P is similar) and hence draws attention to Hermes’ connection of the tortoise/lyre to desire (31*). For the structural function of φέρων in this part of Herm. see 52*, 63*. 41 ἔνθ’ ἀναμηλώσας: Ω’s ἀναπηλήσας could derive from ἀναπηλέω ~ ἀναπάλλω (cf. θηλέω ~ θάλλω), but Hermes does not ‘shake’ the tortoise. Context suggests instead a verb meaning ‘cut open’ or ‘turn belly up’, since Hermes’ next step would be to cut through the tough plastron.62 Ruhnken’s ἀναμηλώσας ‘having probed open’ (1751: 62) seems the best attempt in the former category, and a good complement to 42 ἐξετόρησεν* and the thematic association of Hermes with boring.63 Medical writers use μηλόω. The compound may be preserved in Hesychius α.4442 ἀναμιλῶσαι: †μὴ† ἀναγλύψαι: since μιλόω is not known, the corrupt μή may have originated as an interlinear correction over μι-; the juxtaposition here with γλύφανος would explain why a verb about probing was glossed with ἀναγλύψαι. (Alternatively, μιλόω did exist, and Phot. α.1563 ἀναμιλώσαιμι: ἀναγλύψαιμι points to how μή arose.) No palaeographically plausible conjectures have been offered with the meaning ‘turn belly up’, though the sense could be supported by the parallel motion when Hermes kills the cows (118–19*), by the tortoise’s ­no­torious helplessness on its back, by Sophocles’ use of ἐξ ὑπτίας in a passage describing the invention of the lyre (Ichn. 287), and by Nicander’s description of the slaughter of a turtle, τήνδ’ ἀνακυπώσας κεφαλῆς ἀπὸ θυμὸν ἀράξαι | μαύλιδι χαλκείηι.64 Palaeographically simpler conjectures produce poor

61

More obliquely, Pausanias saw a statue of a tortoise in Hermes’ temple in Megalopolis (8.30.6). 62 43–4*. Galen Comp. med. per gen. 7 p. 1022 (quoted by Allen and Halliday) gives instructions for killing a tortoise by piercing the plastron beside each foreleg. 63 Stephanus’ ἀναπειρήνας, Ludwich’s ἀναμοχλεύσας (1899: 1164) and Wecklein’s ἀνακοιλώσας (1920: 37) are palaeographically less convincing, and the first is not an attested verb. Ludwich’s ἔνθα ῥαφὴν (or δ’ ἁφήν?) λύσας (1905: 5) supposed an obscure reference to the plastron ligaments as the origin of the corruption. 64 Th. 705–6. For Nicander’s acquaintance with this part of Herm. see n. 39. Tortoise on back: e.g. Oppian as cited on 32*, Geoponica as in n. 59. Ernesti’s ἀναδινήσας

165

C O M M E N TA RY 4 1 – 2 sense: *ἀναπιλήσας (Hermann), ἀνακηλήσας (Ilgen), ἀναπηδήσας (Barnes), -εν ἀπειλήσας (Agar 1916). 41–2 γλυφάνωι πολιοῖο σιδήρου | αἰῶν’ ἐξετόρησεν: Different senses of αἰών combine: Hermes removes the tortoise’s ‘vital juices’ (thought to reside especially in the marrow and spinal column), and with it its ‘life’.65 Vergados suggested that αἰῶν’ (‘marrow’) ἐξετόρησεν would suggest the boring of finger-holes in making bone auloi: this is attractive when one adds the similar nod towards other instruments in the phrasing of 47*. Hermes’ use of a γλύφανος (a gouge, or chisel with a curved edge) adds a sculptural quality to his craftsmanship. The word’s other ancient uses both refer to carving wood (Theoc. 1.28, Call. Aet. fr. 100); cf. Alex. Eph. SH 21.3 χελυοξόου Ἑρμείαο. However, ἐκτορέω specifies that the gouge is used as the bit of a drill, perhaps spun by a one-handed bow as depicted on a red-figure hydria by the Gallatin Painter (Boston MFA 13.200, c.490 bce); the use of gouges as drill-bits in sculpture (see Adam 1966: 40–50) was briefly brought to bear on our passage by Johnston and Mulroy (2009) 4. Though Hermes may well subsequently carry on with handheld use of the gouge (Dumoulin 1992: 104 discusses three types of chisel-mark in the classical tortoiseshell lyre from Bassae), the hymnist’s synecdoche for the range of actions required is motivated by a thematic association of Hermes with boring, which is part of his characterization as a god who can control crossing at boundaries without destroying them (Introduction §3.3.6). The word-choice probably influenced Aratus 269, where the whole fashioning of the lyre is reduced to ἐτόρησε (Barnes; see Kidd 1997). The whirling drill will be recalled in 43–4*, 109–10*. The hymnist uses ἐξετόρησε here, διατετορήσας at 119 (cf. Ar. Pax 381 τετορήσω), and ἀντιτορέω (future, 178; present, 283). The derivation of these forms is disputed. Chantraine took ἐτόρησα as a byform of ἔτορον (GH i.416; ἔτορον ~ τρώω much as ἔθορον ~ θρώισκω: i.391); Döderlein (1853) 163–4 argued for the misdivision ἀν-τετόρησα > ἀντ-ετόρησα (cf. Il. 5.337 and West 1998 on 10.267; contrast also GH ii.92, deriving ἀντιτορέω straight from ἀντίτορος). The unusual combination of reduplication and sigmatic ending in τετορῆσαι was perhaps itself based on a reduplicated future τετορήσειν (Schwyzer i.783). However that may be, ἐκτορεῖν is extremely rare, so that Photius’ gloss ἐκτρυπῆσαι (ε.513) may derive from Herm.

(cf. Hsch. α.4275, unsourced) is too energetic. Vergados’ ἀνακυκλήσας would mean ‘bending into a circle’ or ‘making revolve’, not ‘turning over’. 65 Marrow and spine: Hp. Epid. 7.122, Pindar fr. 111.4–5, Plato Tim. 73b. See Degani (1961) 21–3, 35–43, Onians (1954) 200–28.

166

C O M M E N TA RY 4 3 – 4 5 43–4 ὡς δ’ ὁπότ’ ὠκὺ νόημα διὰ στέρνοιο περήσει | ἀνέρος ὅν τε θαμειναὶ ἐπιστρωφῶσι μέριμναι: The man’s idea suddenly ‘crosses into his breast’ – as we say, ‘comes to him’ – he knows not whence. This motion contrasts with what causes it – the anxieties ‘roaming’, and so cycling round to consciousness.66 The double movement perfectly suits the simile’s tenor where the drill whirls the gouge into the chest (one sense of χέλυς) or plastron of the tortoise; ὠκύ, traditionally applied to βέλος, suggests the sharpness and swiftness of the γλύφανος (Van Nortwick 1975: 7). The speed of thought was a traditional vehicle in similes, though normally with the focus on how an individual’s thinking can flit between topics: see Od. 7.36 (Phaeacian ships), Sc. 222 (Perseus flying), H.Ap. 186, 448 (Apollo flying), Thgn. 985 (youth passing), the extended similes at Il. 15.80–3 and A.R. 2.541–6, and the phrases ἅμα νοήματι and θᾶττον νοήματος. περήσει: A short-vowel subjunctive. These are attested in 5th-century Ionic inscriptions, and sometimes metrically assured in Epos; many others may have been normalized in transmission, as our instance was in B.67 In Herm., compare 288(a) ἀντήσεις Da and 289 ἰαύσεις M, where the construction makes it obvious that a subjunctive is needed. Three independent itacistic errors are possible, but seem less likely. θαμειναί: This spelling (Allen 1894: 218, later retracted; Wackernagel (1914) 119 n. 2) is supported by the fourth-century papyrus of Call. Aet. fr. 75.36 (see Pfeiffer’s apparatus). Some doubt remains: see Allen and Sikes’ parallels for Ω’s θαμῑνός. 45 ἤ ὅτε δινηθῶσιν ἀπ’ ὀφθαλμῶν ἀμαρυγαί: M’s ἤ, whether inherited or conjectural (as in Γ2, Barnes), produces a good pairing: glinting looks from whirling eyeballs are a second example of darting and circular motion combined during intense thought.68 The dominant Greek view was that the eye emits light during vision (e.g. Rakoczy 1996: 19–37). Looks which are described with the stem ἀμαρυγ- seem to connote either fierceness, sometimes with a hint of madness, or seductiveness.69 Some similar psychological

Other sudden νοήματα arise or are created by gods inside the person, e.g. Od. 2.363– 4, 14.273. 19.516–17 uses an opposite metaphor to present constant anxieties: πυκιναὶ δέ μοι ἀμφ’ ἁδινὸν κῆρ | ὀξεῖαι μελεδῶναι ὀδυρομένην ἐρέθουσιν. 67 περήσηι was also proposed by Wakefield (1796) 362. For short-vowel subjunctives see Schulze (1885) 491–2, Smyth (1894) 217–18, Bechtel iii.217, GH i.454. 68 Instead of uncorrepted ἤ before a vowel I write ἠ’: see Van Leeuwen (1918) 86–7. 69 Fierce: Hes. Th. 827 (Typhoeus), Ps.-Opp. C. 3.32 ἀπ’ ὀφθαλμῶν ἀμαρυγαί| (probably from Herm.). Seductive: e.g. Sappho fr. 16.18, A.R. 3.288, 1018–19, Ps.-Opp. C. 3.90; this is presumably the implication of Hes. frr. x5 Χαρίτων ἀμαρύγματ’ ἔχουσ–|, and cf. Theoc. 23.7 (seductive ἀμάρυγμα of a lip). Both connotations occur 66

167

C O M M E N TA RY 4 6 – 4 7 implications arise from ‘whirling’ looks: Pindar fr. 56u.13 ὀμμ]άτων ἄπο σέλας ἐδίνασεν (probably of baby Heracles reacting to the snakes sent to kill him) suggests anger or fierceness, as do Bacchylides 17.18 and A.R. 2.25; eyes roll in the throes of death or madness (e.g. Il. 16.792, E. HF 868). But eye-shifts can also be seductive, as Hermes’ will be (278–9* again with ἀμαρυγαί, 387*). The absence of evidence for eyes whirling after a brainwave is an obstacle to taking 45 closely with the situation of 43–4, or to accepting Baumeister’s conjectured line-start αἳ δέ τε, based on Θ’s αἳ ὅτε.70 46 ὣς ἅμ’ ἔπος τε καὶ ἔργον ἐμήδετο κύδιμος Ἑρμῆς: For the idea of action being coincident with speaking see Il. 19.242 ἅμα μῦθος ἔην, τετέλεστο δὲ ἔργον, Hdt. 3.135.1 εἶπε καὶ ἅμα ἔπος τε καὶ ἔργον ἐποίεε, A.R. 4.103 ἔπος ἠδὲ καὶ ἔργον ὁμοῦ πέλεν; further parallels in Allen and Halliday. Hermes was planning his actions during his speech (which itself followed directly upon seeing the tortoise: 29). Hence welcoming the tortoise so soon before killing it was no change of heart, but strategically misleading language; harm is part of the traditional referentiality of ἔργον μήδεσθαι, whereas ἔπος is not a traditional object of μήδεσθαι.71 The basic ground relating 46 to the similes in 43–5 seems to be instantaneity, but the vehicles have an open-ended connection to the tenor. Hermes’ scheme is similar to a problem-solving brainwave, involves piercing a chest with a hint of wild-eyed fierceness, and foreshadows his later darting glances. The opacity of 43–5 – what idea has arrived, what do those whirling sparkling eyes betray? – is itself significant: the hymnist, here and e­ arlier, requires the external audience to make inferences to understand Hermes’ problem and the way the tortoise can solve it (cf. 22*, Introduction §3.4). 47–8 πῆξε δ’ ἄρ’ ἐν μέτροισι ταμὼν δόνακας καλάμοιο, | πειρήνας διὰ νῶτα, διὰ ῥινοῖο χελώνης: Interpretation of these δόνακες, the sense of Sophocles fr. 36 ‘It’s as if you’re a λύρα whose κάλαμος has been removed’, and the δόναξ ὑπολύριος of Ar. Ra. 233, must start from archaeological finds of carapaces from tortoiseshell lyres, which present small man-made holes in a variety of arrangements. Carapaces from Bassae, Daphne, Argos and Locri present pairs of holes on or perpendicular to the major axis, consistent with being the ends of symmetrical struts curved under tension, i.e. made from a with ἀστραπτ-: fierceness at e.g. Il. Pers. fr. 4.8 (Ajax’s madness), PV 356 (Typhon’s crazed belligerence), X. Cyn. 6.15.6 (hunting dogs), Euphorion fr. 71.7; seduction at e.g. Pl. Phdr. 254b, Asclepiades 20 G–P, Moschus Eur. 86, Ach. Tat. 1.19. 70 Adorjáni (2013) 47–51 observed that Hölderlin used lines 43–6 (along with, one might add, 66 αἰπύν, Hermes’ lyre-playing, and his creation of fire at 109–10) to create a single image in his poem ‘Wie wenn am Feiertage’ (1800). 71 For traditional referentiality see in particular Foley (1991), Kelly (2007).

168

C O M M E N TA RY 4 7 – 4 8 strong but flexible material such as reed-stems; such struts would both form a kind of cage to keep the lyre’s arms in place, and counteract the strain on the shell created by the strings and hide. Carapaces from Daphne, Locri, Metapontum, Ephesus and the Acharnian Gate in Athens have holes along the major axis where other struts curving along the shell’s interior may have been pinned in place.72 The hymn’s description suits the former reconstruction well. After cutting the reeds to appropriate sizes (ἐν μέτροισι: compare ἔμμετρος, Eur. HF 1251 ἐν μέτρωι μοχθητέον), Hermes fitted them and thereby (coincident aorist participle) made them πείρατα across the shell.73 That difficult noun is etymologically a ‘thing which traverses’, an origin visible in its use of ropes connecting two points. It gives a suitable sense for the verb here, where the pliant struts (not unlike ropes) bridge the shell.74 The phrase διὰ ῥινοῖο is then attached to πῆξε (compare e.g. Antiphon Tetr 2 3.5 διὰ τῶν ἑαυτοῦ πλευρῶν διαπῆξαι τὸ ἀκόντιον λέγει), with ῥινός chosen as a way of describing tortoiseshell in order to attract attention to what will become a wide-ranging parallelism between the death of the tortoise and the cows.75 The word-order, and its jingle διὰ ... διά, is certainly an unusual effect, and has prompted emendations. Pierson (1752) 156 suggested λιθορρίνοιο (cf. Empedocles 31 B76.10 λιθορρίνων χελύων). Viger (1628: Notae 18) proposed κραταιρίνοιο to match Hdt. 1.47.3, with a much less plausible corruption; Vergados supported it by comparing Paul Sil. Amb. 118, but I do not see an allusion to Herm. there. Line 47 could also describe the construction of a graduated syrinx, as depicted in e.g. Haas (1985) figs. 15 [wrongly labelled ‘16’] (c.490) and 94 (c.460). This ‘misinterpretation’ will gain significance at 512*; already, alongside 41–2*, it contributes to a sense that Hermes could invent any instrument. Less obviously, a listener could fleetingly hear in δόνακας καλάμοιο the reedpens of a poet who, ‘after cutting’ a nib, ‘joins together’ words ‘in metre’. This would complement the relationship being constructed between Hermes and the poet as musicians; for Hermes and writing cf. 75–8*. 72

See Courbin (1980), Roberts (1981) 308–9, Dumoulin (1992) 100–9, 228, Psaroudakes (2006), Psaroudakes and Terzes (2013) 1–27. Earlier interpreters mostly imagined a grill of reeds supporting the ox-skin (e.g. Scaliger 1600: 420). Herm. here displays greater technical knowledge than Σvet. Ar. Ra. 233 (~ Poll. 4.62), which guessed that the δόναξ ὑπολύριος was used as the arm. 73 For coincident aorist participles see Barrett (1964) 214. 74 πειραίνω more often refers to bringing something to a limit or to completion; περαίνω also means ‘I proceed along/through’. LSJ invented the sense ‘pierce’ for our line based only on the resemblance to πείρω. 75 Introduction §3.2. The strange usage was noted by Ludwich (1908). Paus. 8.17.5 uses ἐκδέρω for Hermes’ removal of the tortoise’s shell, again with ulterior motives, namely to explain the name Χελυδόρεα: Hägg (1989) 43.

169

C O M M E N TA RY 4 9 – 5 1 49 ἀμφὶ δὲ δέρμα τάνυσσε βοός: The skin was stretched over the shell, leaving an overhanging border which was wrapped around (ἀμφί) the shell’s rim, and either stitched to it or gathered into pleats and tied fast (Roberts 1981: 310, Dumoulin 1992: 229–30). The cow-skin is better motivated in Sophocles’ Ichneutae, where Hermes probably captured the cows before making the lyre (cf. 375–6 ῥινοκόλλητον ἄλλων ἔκ̣α ̣ ρ̣ψεν βοῶν που δοράς̣ | [ἢ] ’πὸ τῶν Λοξίου). In the present context, sheep- or goat-skin would have worked more cheaply, and Maia’s access to cow-hide conflicts somewhat with Hermes’ urgency about acquiring cattle. πραπίδεσσιν ἑῆισιν: For the instrumental dative, compare Il. 1.608 Ἥφαιστος ποίησεν ἰδυίηισι πραπίδεσσιν. The πραπίδες are often the organ in which cleverness or craft is located. ἑῆισιν emphasizes that cleverness is characteristic of Hermes; πραπίδεσσι νέηισιν would be an unnecessary change. 50 πήχυς ἐνέθηκ’: πήχῡς is an accusative plural rightly accepted from M by Càssola; it was normalized in Ψ to πήχεις. Cf. West (1998) xxxiv on πολύς. Hermes appears to insert the arms after attaching the skin, using the natural recession of the shell’s edge around the tortoise’s rear legs.76 Roberts (1981: 309) found that it was more secure to nail the arms to each other and to a reed strut before adding the skin, but Psaroudakes and Terzes (2013) emphasize that several preserved carapaces have slots into which the base of the arms probably fitted. The arms and yoke of the Elgin lyre survive, and are made of sycamore (British Museum 1816,0610.501); West (1992: 57 n. 35) gives references for other materials used for these parts. Besides the familiar face-on curvature, arms could curve forward from the plane of the skin so as to keep the strings clear of the body (Dumoulin 1992: 232). The wood was probably shaped by steaming. ἐπὶ δὲ ζυγὸν ἤραρεν ἀμφοῖν: On the basis of the Elgin lyre and fifth-­ century vases, Bélis (1985) argued that normally the front of the top of each arm was sawn so as to make a deep cup; through-mortises were cut in the yoke so that it slotted over (ἐπ-ήραρεν) the arms and into these cups. Cf. Dumoulin (1992) 234. ἀμφοῖν is not found elsewhere in epic (contrast H Dem. 15 χερσὶν ἅμ’ ἄμφω), but see Alc. fr. 298.20, Pi. P. 3.57, A. Th. 814, etc. 51 ἑπτὰ δὲ συμφώνους ὀΐων ἐτανύσσατο χορδάς: The line is excerpted in related forms, (a) in the (probably) tenth-century compilation of Mirabilia in Heidelberg Palatinus gr. 398, whose title ascribes it to an Antigonus, and whose first part does probably derive from Antigonus of Carystus’ On Animals;

76

Nicander made a joke out of arms going in leg-holes: see n. 39.

170

C O M M E N TA RY 5 1 and (b) in the Palatine Paradoxographer (third century ce or later), which seems to use Antigonus only here, but cites him by name:77 Mirabilia 7

Paradoxographus Palatinus 20

ἴδιον δὲ καὶ τὸ περὶ τὰ ἔντερα τῶν προβάτων· τὰ μὲν γὰρ τῶν κριῶν ἐστιν ἄφωνα, τὰ δὲ τῶν θηλέων εὔφωνα. ὅθεν καὶ τὸν ποιητὴν ὑπολάβοι τις εἰρηκέναι, πολυπράγμονα πανταχοῦ καὶ περιττὸν ὄντα, “ἑπτὰ δὲ θηλυτέρων ὀΐων ἐτανύσσατο χορδάς”.

ἐπὶ τῶν προβάτων φησὶν Ἀντίγονος τὰ μὲν τῶν κριῶν ἄφωνα εἶναι, τὰ δὲ τῶν θηλέων ἔμφωνα· οὐ λεληθέναι δὲ τοῦτο τὸν ποιητήν. φησὶ γάρ· “ἑπτὰ δὲ θηλυτέρων οἴων ἐτανύσσατο χορδάς.”

Antigonus thus attributed Herm. to Homer and read v. 51 as unusually technical, but took that as a symptom of Homer’s inquisitiveness in various fields (compare the approach of e.g. the Pseudo-Plutarchan On Homer). The citation poses a textual problem: it has θηλυτέρων where Ω had συμφώνους. How far back does each date? If θηλυτέρων arose as an error in transmission, it was probably as an intrusive gloss in the tradition of Antigonus (as Mir. 8 has ζωήν for ψυχήν in Philitas fr. 15), clarifying how the quotation supports his point. This ascribes to Antigonus the assumption (only true most of the time) that ὀΐων are feminine; a reader added θηλυτέρων, much as the scholiast offers σύεσσι: θηλυτέραις to bring out the point at Ps.Opp. C. 1.389 κάπροι πυρόεντες ἐπαιχμάζουσι σύεσσι. The similar collocation θηλειῶν ... οἰῶν specifies ewes at Ael. NA 9.48. That is the reconstruction I find most plausible. Mechanisms to generate συμφώνους in transmission seem much harder. It is not a plausible gloss on ἐτανύσσατο. Vergados supposed a manuscript of H.Hom. annotated with the passage of Antigonus (2007) or a technical source (2013), but Antigonus’ point is not about συμφωνία, and while there is evidence for a concordance annotation from Hesiod (90* ἐπικαμπύλος ὤμους) they seem to have been rare, and either of Vergados’ proposals would be surprisingly abstruse.78 The final alternative is that both variants arose early in rhapsodes’ multiple-­ performance culture. συμφώνους supplies a fitting predicate and prepares

77

For Antigonus’ relationship to the Mirabilia see Dorandi (1999) xiv–xvii, Musso (1976). The varied sources of Paradox. Pal. tell against a very late date or use of the Mirabilia (Giannini 1964: 138). For the possibility that this Antigonus wrote the Metamorphoses which included the story of Battos, see Dorandi (1999) xvii–xxiv, cxxi. 78 For the possibility (but no more) that Sophocles already read συμφώνους see Introduction §1.3.

171

C O M M E N TA RY 5 1 for 53 κατὰ μέλος, whereas the significance of θηλυτέρων is less obvious – most likely, to fit the sheep into the series of female animals which Hermes uses to his benefit (the tortoise and later Apollo’s cows and bee-maidens).79 I infer that συμφώνους would have superseded θηλυτέρων. It may at first sight seem like a coincidence if the one verbatim citation of Herm. illuminates an ancient variant, but the testimonia to other hymns also point to variations not recorded in the manuscript tradition, and it is possible that at Antigonus’ date many more were available in local copies, possibly including ones where Herm. was transmitted on its own.80 Seven was the most common number of strings in the early fifth century, to judge by artistic conventions and literary references (Paquette 1984: 146). The poet may have known that some earlier lyres had only three or four strings (West 1981: 117–21), and in the hypothesis to Pindar’s Pythians, the fact that Hermes’ invention only had four threads for strings marks its primitivism. Our passage might therefore have been taken to emphasize how advanced Hermes’ invention was (Schenck zu Schweinsberg 125). But on the whole aetiologies often overlook such detail for the sake of anchoring the present in the past. The poet simply omits any description of how the strings are attached using a bridge, tailpiece and tuning-pegs – the last being one of the most variable and complex parts of the lyre’s construction (Dumoulin 1992: 240–3) – and of how tuning proceeded by ear, starting from the central string μέση using fourths and fifths, then filling in any remaining movable notes (Hagel 2010: 115, 121, 133). For the types of scale at this period see West (1992) 174–7. In Sophocles Ichneutae (probably) and Apollodorus Bibl. 3.113, Hermes invents the lyre after stealing the cattle, and uses their guts for the strings. Unlike with the cow-hide (49*), Herm.’s ordering has an advantage here: most 79

So Vergados (2007) 741. Antigonus’ interpretation of the point, that only female guts produce sound, is false. 80 For possible ancient variants preserved through Ω, see 224, 322, the full-line doublets at 288, 366, 563, and Introduction §5.1 on the early transmission. For Hy. 18.1–9 as a rhapsodic recomposition based on Herm., see 1–9*. Thucydides 3.104 may cite an old variant in H.Ap. 168, or misquote from memory; Sbardella (1999) is interesting on the variants at 146–50. Pausanias’ testimonia to H.Dem. are more than a corruption away from M’s text at 108–10, 154–5, 419, and in the last word of 476 (see Richardson (1974) 183–4 for the possibility of confusion between the different hymns to Demeter which Pausanias claims to know). POxy. 2379 has a possible ancient variant of H.Dem. 407. The epitome of Ath. 1 22c had χαρίεν for ἐρατόν in H.Ap. 515 (I take p’s χρυσῆν as a characteristically rash emendation using the famous phrase χρυσέα φόρμιγξ; Ψ had ατὸν). At Hy. 1.A9, ὕπατον κέρας (Σ A.R. 2.1209–15c) looks more like an ancient metrical improvement for ὕπατον ὄρος (PGen. 432, D.S. x3) than a corruption; A5, absent in both PGen. 432 and one of the two manuscript branches of D.S. 3.66.3, may have been marked as omissible in early texts.

172

C O M M E N TA RY 5 2 cow-gut is too flabby for strings – at least according to De la Lande (1786: 514–15), who gives an elaborate description of how gut-strings were prepared by hand in eighteenth-century Naples. Also – to press the logic harder than is advisable – if it is surprising that Maia has a store of sheep-gut (from the sheep mentioned in line 232?), it is odd in the other ordering that Hermes preserves the cow-guts before his decision to build a lyre (Gemoll 192). 39–51 Tortoise-engineering: As Schenck zu Schweinsberg 114 noted, more narrative attention is devoted here to establishing Hermes’ inventiveness, whereas his later inventions are fast-paced (cf. 79–85, 108–11, 511–12). The poet still negotiates between speed and technical detail (as promised in 25 τεκτήνατ’). Antigonus of Carystus already commented on the passage’s technicality, albeit misguidedly (51*): it goes beyond other passages of Epos which relate to music, and sees Hermes as sculptor (41*) and joiner (50*) as well as musician. Craftsmanship is not one of Hermes’ main specialities, but here it becomes a way to demonstrate his μῆτις.81 That is further suggested by the unnatural combination of natural materials: three animals and two types of vegetation leave arms in leg-holes, skin beside shell, and exterior guts. The poet’s description is itself artfully constructed – particularly the balance of δόνακας καλάμοιο, δέρμα βοός and χορδὰς ὀΐων (47–51). But the narration also emphasizes the speed and effortlessness of Hermes’ actions, by 39 ἅμ’ and the similes from thought and light in 43–6, and by the abruptness of 41 and the brisk pace of 47–51.82

52–61 52 αὐτὰρ ἐπεὶ δὴ τεῦξε: Vergados observed that this formula is otherwise used of particularly impressive instances of Hephaestus’ craftsmanship (Il. 18.607, Od. 8.276, Hes. Th. 585).

81

The craftiness of craft in early Greek thought is clearly seen in, for example, the story of Hephaestus’ entrapment of Ares and Aphrodite (Od. 8.297–332), or the weaving of Penelope (2.88–106) or Clytemnestra (A. fr. 375). Relevant metaphors include Il. 6.187 δόλον... ὕφαινε, 10.19 μῆτιν… τεκτήναιτο, Od. 3.119 κακὰ ῥάπτομεν … δόλοισι, and A. Ch. 220 δόλον... πλέκεις with Diggle (1981) 115. Hermes gives craftsmen a portion of ψεῦδος in Aesop 105. 82 Contrast the one-sided treatment of a hastily improvised lyre in Luc. D.Mar. 1.4, where Polyphemus did not even add tuning-pegs and so made a racket. Hermes achieves harmony without pegs being mentioned. Lucian elsewhere has Apollo describe Hermes’ actions as follows (D.Deor. 11.4): χελώνην που νεκρὰν εὑρὼν ὄργανον ἀπ’ αὐτῆς συνεπήξατο· πήχεις γὰρ ἐναρμόσας καὶ ζυγώσας, ἔπειτα κολλάβους ἐμπήξας καὶ μαγάδιον ὑποθεὶς καὶ ἐντεινάμενος ἑπτὰ χορδάς κτλ. Barnes saw that the underlined syllables suggest use of Herm.; see Strolonga (2016) 150–2.

173

C O M M E N TA RY 5 3 – 5 4 φέρων ἐρατεινὸν ἄθυρμα: The clear echo of 40 underlines the tortoise’s transformation. 53 πλήκτρωι ἐπειρήτιζε κατὰ μέλος: Hermes tests his tuning (51) with a deliberate strum in which the notes sound consecutively but distinctly. The Greek plectrum was held in the right hand and used for strumming.83 It and the differentiation of strumming and left-hand damping/plucking are not mentioned in Homer or Hesiod, but are represented in art from the Geometric period on; the plectrum does appear in H.Ap. 185. Like the chisel, cow-hide, wood and sheep-guts, a suitable object simply materializes when needed; Apollodorus (Bibl. 3.113 λύραν εὗρε καὶ πλῆκτρον) thought fit to clarify and thus to extend the list of Hermes’ inventions. The manuscripts have κατὰ μέρος here but κατὰ μέλος in 419, 501. κατὰ μέρος, ‘in turn’, is common and barely appropriate to a slow strum, whereas κατὰ μέλος is a rare and so easily corrupted technical term for ‘along the scale’ (Aristox. Harm. 1.28–9; Franklin 2003: 304–5); a further reason for preferring μέλος arises in 499–501*. D’Orville first restored κατὰ μέλος here, but glossed it ‘harmoniously’ (n.d.: ff. 246, 253). 53–4 ἣ δ’ ὑπὸ χειρός | σμερδαλέον κονάβησε: ὑπό marks the source of a noise: compare Il. 2.334, where the ships σμερδαλέον κονάβησαν ἀϋσάντων ὑπ’ Ἀχαιῶν. κοναβέω is formulaically qualified with |σμερδαλέον, which elsewhere in Epos always entails fear.84 Here the formula is misfocalized, in that Hermes is anything but repulsed by his strum. σμερδαλέον emphasizes the potential for the unfamiliar noise to shock other listeners (such as Sophocles’ satyrs: Ichn. 131, 143–4, 203–5, 260–1; cf. Cantilena 1993). It perhaps suggests that the lyre will be a kind of ‘armour’ against Apollo: see 153*, 421*; the Iliad applies σμερδαλέον κονάβησε to armour four times.85 ἥ (sc. χέλυς) follows ἄθυρμα by a simple constructio ad sensum. Accents on self-standing demonstrative uses of ὁ, ἡ, οἱ, αἱ have fairly regular manuscript support (here MDbp at least), and were restored consistently by Hermann; see West (1990) xlix. 83

Use: West (1992) 65–8. For the form see Phaklaris (1977) 230–1 (based on Polion’s bell-krater, Met. 25.78.66); some were sufficiently phallic to sustain comparison with a dildo at Herodas 6.51. LSJ s.v. give examples of metal, ivory and wood as materials. 84 LfgrE s.v. presupposes an etymological connection to Indo-European *smerd- (cf. Latin merda), with a semantic development ‘smelly’ > ‘offensive’ > ‘frightening’, which is uncertain. The formula is varied in 501*. 85 13.498, 15.648, 21.255, 593. Iwaya (1990) reached the idea of the lyre as substitute armour from a different direction.

174

C O M M E N TA RY 5 4 – 5 6 54–5 θεὸς δ’ ὑπὸ καλὸν ἄειδεν | ἐξ αὐτοσχεδίης πειρώμενος: The echo in πειρώμενος of 53 ἐπειρήτιζε underlines the experimental nature of Hermes’ actions, though the nuance has shifted from ‘testing’ against a standard to ‘trying out’ as a means of practical learning (cf. ἐμπειρία and e.g. Od. 3.23 οὐδέ τί πω μύθοισι πεπείρημαι). This characterizes Hermes’ hands-on approach (see further 440–2*), while the common application of the stem πειρα- to ‘testing’ an antagonist interacts neatly with the κερτομία in the following simile: cf. Od. 24.240 κερτομίοις ἔπεσιν διαπειρηθῆναι. Hermes’ experimentation is ἐξ αὐτοσχεδίης: as a first go, it cannot help being improvised (cf. Aristotle’s inference about the origins of poetry from αὐτοσχεδιασμός: Poet. 4 1448b23). But the derivation (‘unprepared’ < ‘immediately to hand’) well suits Hermes’ bricolage in constructing the lyre; and the phrase recalls uses of σχεδιάζω for improvisational quotation and composition of poetry (e.g. Anaxandrides fr. 16.3, Chamaeleon fr. 33, Clearchus fr. 78) where the focus is on the quick-wittedness of improvisation rather than its connection to inventions; the subsequent simile confirms this connection. Finally, καλὸν ἄειδεν aligns Hermes with the lyre itself (38; cf. n. 55). On the punctuation here see 55–6*. 55 ἠΰτε: Personal preference seems to have played a part in this word’s distribution. It is much more frequent in Herm. (x6) and Il. (x29) than in Od. (x7) or Hes. (x1). The Odyssey-poet avoids using it to introduce a clause (rather than a noun-phrase), preferring e.g. ὡς ὅτε or οἷά τε. 55–6 κοῦροι | ἡβηταὶ θαλίηισι παραίβολα κερτομέουσιν: Whereas κοῦρος can cover anything from an embryo through to a young warrior (LSJ s.v. κόρος B), and ἥβη itself can range over all post-pubescence, ἡβηταί are more narrowly connected to coming of age: see IG IX.2 234.4 (third century: the εἱβατάς at Pharsalia may own land); IX.12 718.7 (fifth century: the hεβατάς is of an age to continue the Locrian population); Diodorus 6 GP (age of 24); Tazelaar (1967), esp. 143–7. Fraenkel (1910: 120–1) compared ὑπηνή-της for the word’s form and semantics. The hymnist was varying κοῦροι | πρωθῆβαι (Od. 8.262–3). Vergados gives a good survey of the connections between ἡβάω and attending a party, including ἐνηβητήρια as the locations for the pharaoh Mycerinus’ parties in Hdt. 2.133.4; see also IvO 5 versus IED 3 for the dispute about whether ἐνηβάω is attested for ‘eating a sacrificial feast within’ the sanctuary at late archaic Olympia. This connection must be activated in our use of ἡβηταί since they are at festivities. For the variable scale of θαλίαι see LfgrE s.v. θαλίη, Schmitt Pantel (1992) 39–40. Another example specifically for youths is trag. adesp. 397 θαλίαι τε νέων. The youths are engaging in κερτομία. The semantic field of this term has been much discussed: see Gottesman (2008) and Clarke (2001) with further

175

C O M M E N TA RY 5 5 – 5 6 references. Compared to both those studies I see a looser set of family resemblances in the stem’s usage, with great variety in the speaker’s tone and tactics. The key traits of κερτομία in my view are that it is almost always verbal expression which is not straightforward and is to the hearer’s detriment, often for the speaker’s amusement.86 The sense here is of banter untrammelled by politeness. I postpone discussion of παραίβολα, and turn next to the ground of the simile. I have tentatively avoided punctuation after 54 ἄειδεν, owing to the lack of similes in Epos with the structure ‘participial phrase as tenor, adverbial pivot, finite verb in the vehicle’. This articulation suggests that the κερτομία is both improvised and sung to music. By punctuating after ἄειδεν and not πειρώμενος, one could relate 55–6 also to spoken banter, as Apollonius may have done: after 1.457–8 ἀμοιβαδὶς ἀλλήλοισιν μυθεῦνθ’, οἷά τε πολλὰ νέοι … we hear Idas speaking to rile Jason.87 However, the parallel simile at 454* makes such an approach less attractive, and I assume that unprepared lyric performances, as a forum for competitive displays of ingenuity, are already perceived to be at stake here. The combination of a party, singing, lack of preparation and κερτομία occurs at Aristophanes V. 1222–49: Bdelycleon presupposes that at an elite Attic symposium performances of snatches of lyric should follow one from another, while Philocleon uses the opportunity to be insulting. I take this to be a comic exaggeration, rather than a reversal, of sympotic banter (see further parallels on 454*). An ancient audience might have thought also of early transmission of iambus through such games of quotation, though that genre is closely associated with auloi rather than the lyre (Rotstein 2010: 233–9). It is harder to trace κερτομία in extant sympotic poetry, though some passages betray an antagonistic streak: Theognis

86

I cite some instances to justify, albeit briefly, my attempt to encapsulate the stem. Speaker lying: Od. 13.326. Dissembling: Il. 1.539 (Hera knows it was Thetis but asks anyway), 4.6 (Zeus pretends to consider stopping the war). Sarcastic: Od. 2.323. Using pungent phrasing: e.g. Il. 20.433 describing Achilles’ rhyme ἄσσον ἴθ’ ὥς κεν θάσσον ὀλέθρου πείραθ’ ἵκηαι (428). Od. 20.177 seems unusually candid for κερτομία; Il. 24.649 (Achilles’ instruction that Priam should sleep outside) is difficult on all accounts. κερτομία upsets: Od. 18.346–50, Hdt. 1.129.1. Provokes fear: Il. 4.6 (Zeus tries to make Hera worry that her efforts are in vain). Provokes fear and sense of inferiority: 20.202. Offends: Od. 20.263–6. Shames: Hdt. 8.92.2. Belittles suffering: Il. 5.419, E. Cyc. 687. Elicits false optimism: Od. 13.326. 87 For Apollonius’ extensive use of Herm. in Arg. 1 see 116–17*, Introduction §4.2. For exchange of spoken insults at parties cf. X. Cyr. 5.2.18, Plu. Lycurgus 12.4 (both probably idealizations based on Greek practice), Alexis fr. 160. The banter in adesp. eleg. 27 συμπόται ἄνδρες ὁμ[ήλικες] ... χρὴ ... σκώπτειν τοιαῦθ’ οἷα γέλωτα φέρειν could be either spoken or sung.

176

C O M M E N TA RY 5 7 581–2 probably caps and corrects 579–80; so too PMG 900–1 among the Attic skolia quoted by Athenaeus, and CA lyr. adesp. 8a–b.88 One already notices a certain mismatch between tenor and vehicle in that Hermes has no apparent audience to interact with. This is exacerbated in what follows: see 52–61* below, 454*. παραίβολα is brilliantly devised to suggest all the associations (i) ‘in rejoinder’, (ii) ‘competitive’, (iii) ‘indirect, oblique’ and (iv) ‘risky’.89 For (i) cf. LSJ s.vv. παραβάλλω A I.3, παραβολή, and Apollonius’ uses of παραβλήδην for ‘in reply’. For (ii) cf. παραβάλλομαι ‘vie’, with reference to speech at e.g. E. Andr. 289 (with Stevens 1971), and Plato Smp. 214c λόγους παραβάλλειν of verbal competition at symposia. For (iii) see LSJ s.v. παραβάλλω A IV. Greek words are often ‘cast’, and those which are not cast forwards are probably those which conceal the speaker’s thoughts; cf. Od. 4.348 παρακλιδόν ‘deceptively’. Indirection is typical of κερτομία, and provides the most plausible interpretation at Il. 4.6 κερτομίοις ἐπέεσσι παραβλήδην ἀγορεύων. This sense could pun on sympotic σκόλια, which the Greeks related to σκολιός ‘crooked’: see e.g. PMG 892, Lambin (1993) 33–4. Finally, (iv) ‘risky’ is the common sense of παράβολος.90 The fine line between κερτομία and offence can be seen from e.g. E. fr. 492, whose speaker touchily disparages χάριτες κέρτομοι as ­unbridled and juvenile. 57 ἀμφὶ Δία Κρονίδην καὶ Μαιάδα καλλιπέδιλον: For ἀμφί + acc. as a formulaic start for rhapsodic and citharodic proems and dithyrambs, see Hy. 7, 19, 22, 33, Ar. Nu. 595 with Suda α.1700 (ἀμφιανακτίζειν ~ ‘to use the start ἀμφὶ ἄνακτα’), Photius α.1303–4. The accusative identifies the hymnic dedicatee(s), so the phrasing prompts us to take Hermes’ song as a hymn to his parents; Càssola compared Hy. 27.16–21, where Artemis leads a hymn about Leto’s motherhood. Here, despite the addressee of the inset hymn not being Hermes himself, there is a clear mise-en-abyme effect, since Herm. itself began with mention of Διὸς καὶ Μαιάδος υἱόν (1).91 The juxtaposition of this hymnic colour with 56 κερτομέουσιν is

88

For further, less clear cases see Reitzenstein (1893) 3–44, Vetta (1983), Collins (2004) 84–134. Certamen 70–175 has an extensive depiction of Hesiod challenging Homer to improvise continuations, but at a public contest. 89 For the recessive accent (M; Franke), against παραιβόλα, see Probert (2003) 108. For the metrical use of epic παραί in compounds, compare παραιβάτης, παραιφασίη, παραιπεπιθεῖν, etc.; regarding its origin see García Ramón (1997). 90 The application to ‘bold’ language (e.g. Ps.-Longinus Subl. 32) would suit κερτομία but may well be later. 91 See also 58*. Vergados 4 oversimplifies the mise en abyme by claiming that Hermes is hymning himself.

177

C O M M E N TA RY 5 8 – 5 9 e­ xpressive:  its  potential  to  wrong-foot is clear from early editions, where there is no comma after κερτομέουσιν, so that the party-goers were making insults about Zeus and Maia – a risky business indeed; even with Clarke’s comma after κερτομέουσιν, the juxtaposition emphasizes the paradox of comparing hymnic song to κερτομία. The epithets Κρονίδην and καλλιπέδιλον neatly encapsulate Hermes’ themes of ancestry and Maia’s possessions. The hapax καλλιπέδιλος varies the formula Ἥρης χρυσοπεδίλου (e.g. Hes. Th. 952, fr. 25.29) to present Maia as a substitute in Zeus’s affections (cf. 58*), and perhaps to foreshadow Hermes’ own special use of sandals (e.g. 79, Od. 5.44). 58 οἳ πάρος ὠρίζεσκον ἑταιρείηι φιλότητι: The frequentative verb implies, as 7 μισγέσκετο did, that this was not one of Zeus’s one-night stands, and the stem ὠρ- < ὀαρ- points to time spent in intimate conversation.92 ἑταιρεῖος and φιλότης both apply to consummation as well as affection, but ἑταιρεῖος pulls the expression distinctly towards the latter (see LSJ s.v. I.6). For the line’s first word, I narrowly prefer Clarke’s οἵ (with a distinct palaeographical advantage) over ὡς (Γ2,m and Châteillon, with slightly stronger sense). Either suggests the transitional hymnic relative: see 3 ὃν τέκε Μαῖα*, and for transitional ὡς see H.Pan 29, Hy. 7.2, 27.19. The manuscripts’ ὅν entails a larger corruption; against construing the syntax as ἐξονομάζων (sc. τὸν ὄαρον) ὃν ὠρίζεσκον, ἥν τ’ αὐτοῦ γενεήν, one cannot name an ὄαρος, and ὅ is expected as the substitute for an inner accusative (KG i.309–10). 59 ἥν αὐτοῦ γενεὴν ὀνομάκλυτον ἐξονομάζων: For Hermes’ emphasis on naming his ancestors see 57, where Kronos appears as well as Zeus and Maia. The phrasing, and the notion of pride in the recitation of one’s ancestry, are similar to Il. 10.68–9 πατρόθεν ἐκ γενεῆς ὀνομάζων ἄνδρα ἕκαστον, | πάντας κυδαίνων.93 The manuscripts have ἥν τ’ αὐτοῦ, but the τε is awkward (Wecklein 1920: 37): it suggests that the phrase of ὀνομάζων and that of 55 πειρώμενος are logically level, and in fact 59 redescribes 57–8 rather than adding a new point.

For topoi in Greek literary depictions of ὀαριστύς see Cairns (2010). Our line may bear some relationship to Hsch. ω.345, where glosses on ὠρίζει include ὁμιλεῖ and ἀδολεσχεῖ. 93 Cf. Higbie (1995) 6–14, Dickey (1996) 52–6. The similar doubling |ἐξονομακλήδην ὀνομαζ- (Il. 22.415, Od. 4.278) also emphasizes the emotive power of names – in that case, the effect when someone is addressed by name. Faulkner (2008) 189 discusses the accent and history of ὀνομάκλυτον. The only conjecture I found in Brunck’s notes on manuscript C is ὀνομακλειτὴν here (Parisinus supp. gr. 392 f. 234v): I mention it because it is not widely known rather than because it is good. 92

178

C O M M E N TA RY 6 0 – 6 1 60–1 ἀμφιπόλους τ’ ἐγέραιρε καὶ ἀγλαὰ δώματα νύμφης | καὶ τρίποδας κατὰ οἶκον ἐπηετανούς τε λέβητας: Hermes proceeds from his parents to aspects of Maia’s οἶκος. For the use of τε to connect ἐγέραιρε all the way back to ὑπὸ ... ἄειδεν, see Denniston 497–8.94 The use of γεραίρω for conferring honour through song is common in the fifth century (e.g. Pi. O. 3.2, I. 8.62, Bacchylides x4, Eur. El. 712, Hdt. 5.67.5, Ar. Thesm. 961). However, Homeric usage relates the verb instead to the honorific portion at a feast, a connection which will become meaningful in 129*. Maia’s attendants may be compared to Circe’s four ἀμφίπολοι at Od. 10.348–51; one perhaps imagines other local nymphs, not dignified by Zeus’s attentions, with domestic tasks which may have recently included helping with Hermes’ birth and nursing.95 Nurses played larger roles in other versions of the myth: in Alcaeus’ Hymn to Hermes the god probably slipped away from the Horai (Cairns 1983: 31); in Sophocles, Kyllene has been nursing Hermes (Ichn. 272); in Eratosthenes’ Hermes, Maia’s sisters were looking after Hermes but he stole their clothes while they washed (fr. 1). Perhaps the hymnist could already allude to such stories. A particular further context for human ἀμφίπολοι was as gods’ attendants at sanctuaries (Pi. fr. 52f.117, Hdt. 2.56.2, E. IT 1114, fr. 992; cf. the ἀμφιπολεῖον at Aphaia in IG I3 1456.13). The rest of 60–1 activates this context as a further pointer to Hermes’ presentation of the cave as like a temple.96 ἀγλαά is a favourite with our hymnist: eleven uses make it over six times more frequent than in Il. (x47) or Od. (x37). Rather like the stem γαν-, the core senses seem to be ‘radiant, illustrious’, extending towards ‘beautiful’ and ‘inducing pride and joy’ (the last especially with ἀγλαΐα, ἀγλαΐζω and ἀγλαὸς υἱός: see 314*). Hermes’ proud presentation, evident in 59, continues. But homes, unlike pieces of metalwork, are rarely ἀγλαός – particularly not a ‘shadowy’ cave (6). Both here and at E. Andr. 135 (Thetis’ temple), the adjective may have been felt to apply to a sacred place by a transfer from valuable metal possessions. If so, the formulaic juxtaposition of basins and

τ’ ἐγέραιρε is marginally preferable to Ω’s τε γέραιρε, since Epos shows a general tendency to avoid feminine caesuras in the second foot (though see e.g. 49) and to place the metrical unit –⏑⏑–⏑, as ἀμφιπόλους τε would be, at line-end (Hagel 2004: 210; Taida 2010: 254–5). 95 Whereas in human society ἀμφίπολοι are distinctly elevated (see LfgrE s.v., and e.g. Pi. O. 6.32, Hdt. 2.131.2, 5.92η.3, 9.76.1, E. Su. 1115), Circe’s ἀμφίπολοι lay the table. For birthing support in Greece see French (1987), and H.Ap. 92–4 where Leto’s peers arrive. For nymphs and nursing see e.g. H.Aphr. 256–7, Hy. 26.3, Hadzisteliou Price (1978) s.v. ‘Nymph(s)’, Larson (2001) 85–6. 96 Cf. 23*, Vamvouri (2000) 100–1. Rougier-Blanc (2005) 38–9 asserted that Maia’s cave becomes temple-like only after the cattle-theft, but see also 65*. 94

179

C O M M E N TA RY 5 2 – 6 1 their tripods in 61 effectively ‘glosses’ ἀγλαός.97 Eitrem’s neat inference that Hermes knows about these cauldrons from having been bathed (1906: 253; cf. Herm. 241, 268) suits the early (pre-700?) and mythological domestic function of tripod-cauldrons for heating water. However, they mainly featured at Herm.’s date as dedications at sanctuaries, to the point where vase-painters use them to signal a sanctuary backdrop; see 64*, 179–81*, Sakowski (1997) 206–10, 354–5. As Vergados pointed out, such dedications were more typical of major sanctuaries than of a nymph-cult. Maia’s inflated metal wealth will recur at 249–51. 52–61 Hermes’ first song: The proem of Hesiod’s Theogony and several Homeric Hymns (especially Herm., H.Ap. 156–64, H.Pan, Hy. 21, 27) introduce inset songs about the gods, which prompt us to compare and contrast the surrounding hymn, and which can fruitfully be read (for example) as adumbrating the poets’ conception of their task. For Herm., see also 425–33*.98 Hermes’ first song has several hymnic elements, particularly in the way its subject is introduced (57*) and its laudatory tone – the depiction of Maia’s affair as long and tender (57–8*), his pride in his ancestry (59), 60 ἐγέραιρε, the alignment of cave and temple (60–1*). His focus on his parents’ relationship stakes a claim to the recognition he hopes to get from his absent father. It also interacts with the way Herm. itself began (1–9), though Hermes foregrounds the cave’s wealth rather than its secrecy (Nobili 2016: 49). Thus there is partial mise en abyme, though Hermes is hymning his parents (especially Maia) rather than himself, does not necessarily address an audience, breaks off before any sort of prayer, and uses a tortoiseshell lyre suited to private performance whereas the hymnist probably recited without any lyre, albeit retaining an awareness of his tradition’s lyric origins.99 Expressively, Hermes mentions himself barely if at all (in 59), and can take no further part in the praises of Maia’s οἶκος; he is not yet far enough along his trajectory from secretive love-child to acknowledged Olympian to do so. In this sense, Herm. complements its embedded song, by describing how Hermes became praiseworthy. The hymnist therefore constructs a relationship of both partial similarity and complementarity between his own work and that of Hermes, here the

Other tripods in Epos are ὠτώεντα, i.e. welded directly to the vessel and its handles. Both types are widely attested in art and archaeology. 98 For mise en abyme elsewhere in the Hymns see Vamvouri Ruffy (2004) 149–54, Calame (2011) 342–53. For H.Pan see Thomas (2011) 159–68. 99 Martin (2010). For the view that rhapsodes chanted with melodic accents but without instrumental accompaniment see also West (1971) 307–12, (1981). For hexameter hymns accompanied by kithara see Introduction §2.3 on Terpander, and West (1986). 97

180

C O M M E N TA RY 6 2 – 6 3 originator of singing hymns to the lyre.100 This emphasizes the glory of the hymnist’s poetic heritage (and heritage is one of Hermes’ hymnic themes), and is part of his broader strategy of implying fellow-feeling between himself and Hermes in order to create χάρις. The hymnist also demonstrates a taste for paradox comparable to what we have seen from Hermes already. This is shown by the simile comparing Hermes’ song to banter at parties, probably specifically in the form of competitive citation or improvisation of poetry (55–6*). This simile characterizes Hermes’ singing as a further sign of ingenuity and experimentalism, following on from his invention of the lyre; and it relates to the lyre being something to play with (32, 40) and a δαιτὸς ἑταίρη (31). But although tenor and vehicle share an improvisatory quality and a young male performer, the simile is wilfully mismatched in other respects: a solo is compared to multi-player competition at a party, hymnic praise to teasing κερτομία.101 The audience has not yet been given the tools to interpret this paradox further. For the significance of the paradox and the hymnist’s strategies for creating χάρις, see further Introduction §3.4.

62–74 62 καὶ τὰ μὲν οὖν ἤειδε, τὰ δὲ φρεσὶν ἄλλα μενοίνα: The particles introduce a heightened opposition, suitable for a summary followed by a major change of topic: see Od. 13.122, and in general Reynen (1958) 82–8. For the contrast of words versus secret thoughts, compare the Odyssey’s formula νόος δέ οἱ ἄλλα μενοινᾶι| of Penelope (2.92, 13.381, 18.283; cf. 18.344–5). Crossfertilization with the formula φρεσὶ(ν σ)ῆισι μενοινᾶι(ς)| is found also in IG I3 1204 (c.535) φρασὶν ἄλα μενοινν. Here, τὰ δὲ ... ἄλλα (‘those other things’) reminds us of Hermes’ aforementioned plan to find Apollo’s cows (22, ­reinforced by the framing in 63*). μενοινάω covers both ‘feel an impulse for’ (to the fore in 64*) and ‘plan’ (to the fore in 66*). 63 καὶ τὴν μὲν κατέθηκε φέρων ἱερῶι ἐνὶ λίκνωι: Hermes hides his toy, an action to be echoed in 134*; this is a first hint that his cot and swaddling may be a locus of deception (cf. 151–3, 237–46). The third occurrence of φέρων 100

Admittedly, the audience could imagine that other forms of lyre already exist, as in Lucian D.Deor. 11. That they do not is only clarified at 443–5, 450–2. Cf. 31 χοροιτύπε*. For another reading of Hermes’ first song as abortive see Harden and Kelly (2013) 24–9. 101 Not ‘Evidently, the closest available analogy’, as Clay (1989) 108 asserted. A somewhat similar effect, where generic mismatch flags up thematic importance, is where Pindar embeds a reference to sympotic musical παιδιά within an epinician at O. 1.16–17; dining-behaviour is a unifying theme in the ode.

181

C O M M E N TA RY 6 4 marks the end of the lyre-playing, as in 40 and 52 it marked the end of meeting the tortoise and of preparing the lyre. Meanwhile, the echo of 21 ἱερῶι ἐνὶ λίκνωι gives closure to this section and resumes the basic plan of getting Apollo’s cattle; see also 65 ἄλτο*, 150*. Hermes’ substitution of the lyre for himself during his nocturnal wanderings may suggest that the instrument was invented in Hermes’ own image (see n. 55). The hymnist does not spell out the idea that Hermes’ carers will see a lump in his cot and think he is there (Haft 1996: 33), nor a light-hearted contrast to granny Rhea’s decoy swaddling of objects (Hes. Th. 485). 64 φόρμιγγα γλαφυρήν: On the use of φόρμιγξ for a tortoiseshell lyre, see 17*. It is formulaically γλαφυρός (Od. x4, H.Ap. 183), as are caves (and ships). The epithet combines ‘smooth’ and ‘concave’, like the lyre’s soundbox on which Hermes wielded his γλύφανος. Like ΣV Od. 8.257, the hymnist may have taken γλαφυρός and γλύφω as cognate. At least by the time of Ar. Av. 1272, γλαφυρός had the further sense ‘ingenious’, which would once more connect Hermes to the lyre he produced. By the time of Lucian D Deor. 11.4, which uses γλαφυρόν of Hermes’ first song, the word had developed the critical sense ‘elegant’. κρειῶν ἐρατίζων: Hermes’ hunger is introduced abruptly, using strong lan­­­­­ guage which Apollo will repeat at 287* (see also 130*). The phrase occurs twice in the Iliad in similes comparing a hero to a lion attacking, aptly enough, cattle: Il. 11.548–52 ὡς δ’ αἴθωνα λέοντα βοῶν ἀπὸ μεσσαύλοιο | ἐσσεύαντο κύνες τε καὶ ἀνέρες ἀγροιῶται | … ὃ δὲ κρειῶν ἐρατίζων | ἰθύει (≈ 17.657–61).102 Such a bestial hunger suits the desires of a baby, which are not subordinated to reason (see e.g. Democritus 68 B70 παιδός, οὐκ ἀνδρός, τὸ ἀμέτρως ἐπιθυμεῖν). However, the content of the desire is surprising, both for a baby (who would want milk: 267) and for a god; by contrast, the baby Apollo was fed on ambrosia and nectar (H.Ap. 124; Introduction §1.3) – the normal diet of adult gods. Sources for the gods eating meat congregate in myths about the distant past and in comedy. The latter is pertinent here: the phrase suggests the humorous presentation of (especially) Hermes and Heracles as beset by human hunger among other characteristics which bring them down to earth. The comic schema is an alternative to that of Hesiod’s account of Mekone, according to which divine and human diets are now strictly separated by the institution of animal sacrifice. Kahn (1978) 56 inferred that Hermes’ hunger calls his divinity into question. But Herm. clearly promotes the comic schema: Hermes’ divinity has been stated at 54 and never brought into question, while

102

For appreciation of this simile-transfer in Ps.-Theocritus 25 see Introduction §4.3.

182

C O M M E N TA RY 6 5 his comic humanization has been introduced at 24*.103 When Apollo refers to Hermes’ hunger at 287 he has no doubt about his divinity (291). One may reinterpret 61–2 in the light of Hermes’ hunger: was his song veering towards cooking equipment because he was really thinking about beef ? Clay (1989) 111 put it the other way round: his song may remind him of his hunger. 65 ἄλτο κατὰ σκοπιήν: Alongside κρειῶν ἐρατίζων, the words reprise each of those at 22 ἀναΐξας ζήτει βόας, though with less explicit phrasing. The parallelism strengthens the sense of restart after the lyre-episode (cf. 63* ἱερῶι ἐνὶ λίκνωι; Shelmerdine 1981: 95). I take κατὰ σκοπιήν primarily as ‘with the goal of reconnoitring’ (LSJ s.vv. κατά III, σκοπιά II; cf. 73*). The strong association of σκοπιά with heights is apt for Kyllene (cf. the μηλοσκόπος Arcadian peak at H.Pan 11), but to translate ‘down along the vantage-point’ involves an unusual extended σκοπιά and a slightly awkward hysteron proteron when ἐκ μεγάροιο follows.104 For the disjunction between where Hermes reconnoitres and where he finds the cows, see 69–70*. The manuscripts suggest that in Ψ ὦρτο glossed the less common ἆλτο; Allen and Sikes noted the same variation in ΣA Il. 20.62. For the accentuation ἄλτο see West (1990) xx. εὐωδέος ἐκ μεγάροιο: The Greeks frequently make fragrance a part of ­idealized, luxurious circumstances. In Epos, a connection to gods is particularly common, perhaps encouraged by the ethereal nature of smell. Divine homes are therefore often scented, as temples were, by burning ­fragrant wood.105 The fragrance continues the idealization of Maia’s cave as ­temple-like from 60–1, now probably outside Hermes’ focalization. Perhaps

103

See 130–2, 286–8, Thomas (2017a). Gods eating meat in comedy: Ar. Pax 192–3 (Hermes), 202, 386–8, 741, Plut. 1128–37; Versnel (2011: 352–64) discussed these passages well, though elsewhere in his chapter he underestimated how specific to comic sources the evidence of Hermes’ ‘hunger’ is; cf. LIMC IV(1).798–801, 817–21 for Heracles after his apotheosis. Carnivorous gods in the distant past: Hes. Th. 538–53, Pi. O. 1.46–52 (rejecting the story, but acknowledging its fame). For the opinion that hunger calls divinity into question see also Vernant (1989) 165, Clay (1989) 111–12. 104 The occurrences of κατὰ σκοπιάς ‘each to his own look-out post’ (Od. 14.261 = 17.430) suggest that we have a traditional phrase with semantic redirection. One might entertain the emendation κατὰ σκοπιῆς, ‘down from the vantage-point’. 105 Cf. 322, and in general Lilja (1972). For scents in epiphanies see 231*. Fragrance in idealized human situations: e.g. Xenoph. fr. 1, Pi. fr. 129, E. fr. 773.13–14. Fragrant divine homes: e.g. Od. 5.59, Hy. 26.6. The gods also scent their clothes (237 θυήεντ’*) and anoint their bodies with ambrosia (e.g. H.Dem. 238, Ar. Pax 525–6, Mosch. Eur. 91–2).

183

C O M M E N TA RY 6 6 – 6 7 μέγαρον itself carries a hint of ‘temple, shrine’ (so Hdt., eleven times), though the epic sense ‘house, hall’ is appropriate enough. 66–7 ὁρμαίνων δόλον αἰπὺν ἐνὶ φρεσὶν οἷά τε φῶτες | φιλῆται διέπουσι μελαίνης νυκτὸς ἐν ὥρηι: For the phrasing of 66 cf. Od. 4.843 φόνον αἰπὺν ἐνὶ φρεσὶν ὁρμαίνοντες. It recalls 62 through ἐνὶ φρεσίν and the neatly paired verbs: like μενοινάω there, ὁρμαίνω developed from ‘feel an impulse to/ for’ to ‘plan’ to ‘imagine’. Vergados (on 62–7) observes the echo in οἷά τε φῶτες | φιλῆται of 55–6 ἠύτε κοῦροι | ἡβηταί: these similes frame Hermes’ song, and mark the movement back to the plan that was still in his mind during it. In δόλον αἰπύν (probably familiar to the hymnist from Hes. Th. 589, Op. 83: Introduction §1.3), αἰπύς seems to have passed from ‘steep’ to (roughly) ‘unmasterable’, as in its applications to πόνος, χόλος, ὄλεθρος and φόνος. Both ὁρμαίνων and αἰπύν may interact with Hermes setting off (ὁρμάομαι) across steep (as 65 σκοπιήν hinted) Kyllene as the first part of his trickery. For the shift from singular δόλον to plural οἷα, compare e.g. Od. 14.63 κτῆσιν … οἷά τε.106 A φιλήτης may burgle a house, steal animals (as here), or rob with violence (A. Ch. 1001–4).107 The stem (Herm. x6) is often connected with Hermes: he is named Philetes in Hellanicus FGrH 4 F19b (cf. 214*), is a φιλήτης at Sophocles Ichn. 340, protects one in Hipponax fr. 79, and is their king in Rhesus 217, Kaibel (1878) no. 1108 (Chios, undated); cf. 175, 292 below. The frequent connection of φιλῆται with night (here, Hipponax, Rhesus, Archil. fr. 49.7, Call. Hec. fr. 74) reflects general Greek associations of theft and the night, such as Il. 3.11 νυκτὸς δέ τε κλέπτηι ἀμείνω, Hes. Op. 605 ἡμερόκοιτος ἀνήρ, and the aggravated quality of nocturnal theft in Athens (D. 24.113). For the common apposition of φώς, ἀνήρ and γυνή to ‘ein spezieller Personenbegriff’, see Schwyzer ii.614, LfgrE iv.1081.56–60, A. Ch. 1001 φιλήτης ἀνήρ, S. fr. 933 ἀνδρὶ φιλήτηι. For ὥρη of a short period of time such as the night (later > ‘hour’), cf. Apollodorus PMG 701, A. Eu. 109, S. fr. 389, Hdt. 4.158.2. In the Iliad, ὥρη

οἷά τε rarely introduces epic similes; the one formulaic instance also relates to theft (‘like pirates’ at Od. 3.73, 9.254, H.Ap. 454), and the connection recurs at Od. 11.364–5 ἠπεροπῆά τ’ ἔμεν καὶ ἐπίκλοπον, οἷά τε πολλούς | βόσκει γαῖα. 107 The accent is paroxytone according to Philoponus Diff. Acc. s.v., Suda φ.336, EM 793.57. φιλήτης is the spelling of inscriptions and early papyri, though West’s apparatus to Hipponax fr. 79 reports ‘φι- ex inchoato φη- factum’. It is confirmed by the derivation from φιλέω made already in the Hesiodic Catalogue: see Thomas (2007), where I should have mentioned φιλατίας in FD III.1 486 (280s bce). The spelling φηλήτης was based on an assumed link to φῆλος ‘deceitful’, but the intermediary φηλέω/άω does not occur. Of Herm.’s manuscripts, M has φιλ-, Θ had φιλ- except in 67; p had φηλ- except in 175, though some members of the family write φιλ- sporadically. 106

184

C O M M E N TA RY 6 8 – 7 0 only means ‘season’; the Odyssey adds the sense ‘the natural time’ (e.g. for eating, sleeping). 68–9 Ἠέλιος μὲν ἔδυνε κατὰ χθονὸς Ὠκεανόνδε | αὐτοῖσίν θ’ ἵπποισι καὶ ἅρμασιν: There is an unusually blunt ellipsis of several hours since Hermes played the lyre in the middle of the day (17).108 The ellipsis brings out the ­relevance of 67 μελαίνης νυκτὸς ἐν ὥρηι in the preceding simile, startles us with Hermes’ mobility across the distant scene-change (see 69–70*), and avoids repetition given the attention to be lavished on his even more impressive return journey. References to the passing of time, expanding on 17–18, will be a marked feature of Hermes’ nocturnal adventure (97–100, 141–3, 182–9*). This description of sunset emphasizes Helios’ travelling, to make a contrast with Hermes (see 69–70*). Both Okeanos and a two- or four-horse chariot are fairly frequent in paintings of Helios’ rising from c.510–450 (LIMC V(1).1032– 3). Dawn has a chariot in Od. 23.244–6, Helios at Mimn. fr. 12.3.109 Helios has gates beside Okeanos at Od. 24.11–12, and may sail along it overnight back to the orient (Ath. 11 469d–70d). But first he must sink beneath the land’s horizon; for κατά + gen. of descent into the ground see e.g. LSJ s.v. κατά A II.2, Mimn. fr. 2.14, Thgn. 1278. For the comitative dative with αὐτός in Epos see GH ii.76. Janko (1982: 135) inferred a formulaic background from Iliad 23.8 ἀλλ’ αὐτοῖς ἵπποισι καὶ ἅρμασιν (not of sunset). 69–70 αὐτὰρ ἄρ’ Ἑρμῆς | Πιερίης ἀφίκανε θέων ὄρεα σκιόεντα: The audience’s expectations about the location of the cows cannot be reconstructed (cf. 18*). Philostratus’ version (Im. 1.26) – which might derive this detail, like the Horae, from Alcaeus’ Hymn to Hermes – puts the cows at Pieria, but has Hermes born near by on Olympus. Later versions place them near Delphi (hypothesis to Pindar’s Paeans), Pherae (Ant. Lib. 23) or Pylos (Ovid Met. 2.684–5). Even in Herm.’s conception, they have had various pasturages (556–7). In any case, the revelation that they are over 250 kilometres from Kyllene as the crow flies – made sudden by ellipsis (68–9*) – makes Hermes’ searching (65) preternaturally successful, so much so that one turns to symbolism to shore up the logic of the plot. The symbolism is on one level obvious: 108

For the rarity of ellipsis in the Iliad and Odyssey see Richardson (1990) 20. Od. 13.17–18 (bedtime/dawn) is comparably abrupt but there is less sense that interesting events are being skipped over. For transitions effected by a time-indication and μέν without a further connective, cf. Il. 18.241 (also Ἠέλιος μὲν ἔδυνε), 8.1, Od. 2.434 (both about dawn). 109 For solar chariots in related cultures see West (1997) 507–8; (2007) 205–11.

185

C O M M E N TA RY 7 1 since line 1, Hermes has been between Zeus and Maia, between Olympus and Kyllene; his journey to Olympus’ northern foothills is a precursor to the trip to Olympus he desires (as will be confirmed at 170, 270*, 312). Further, the most prominent denizens of Pieria are the Muses, and we have just seen Hermes’ musicality (see further 72*; Shelmerdine (1981: 103). As in 68–9, the clause consists of a traveller, verb of motion, destination and reference to mode of travel (θέων – Hermes is literally ‘brought down to earth’ by not yet having winged sandals).110 Formulaic σκιόεντα interacts specifically with the sunset: cf. the formula δύσετο δ’ ἠέλιος, σκιόωντό τε πᾶσαι ἀγυιαί (x11). In the Odyssey, μέγαρα are σκιόεντα seven times, always at shadowy times of day. (Cf. 73* for epithets hereabouts interacting with the context.) The elaborate balance casts Hermes’ movement as anti-solar, and so aligned with night (see Introduction §3.3.2). 71 ἔνθα θεῶν μακάρων βόες ἄμβροτοι αὖλιν ἔχεσκον: See 18* for the implication that Apollo cares for the cows rather than owning them.111 In Epos, αὖλις denotes a temporary sleeping-place for humans (Il. 9.232) or any sleeping-place for animals – an open enclosure (also often αὐλή), a steading, or simply the fields. ἔχεσκον suggests it is the cows’ normal habitat, though we will later hear that Apollo has previously taken them to Parnassus (555–7). The reference to sleep suits the evening setting. There is no hint of any enclosure even though it would be thematic for Hermes to break into one, nor will Apollo subsequently mention any: the natural image, then, is of the cows sleeping in the open (as in Od. 12.265 αὐλιζομενάων). This casts Pieria as a utopia innocent – until now – of theft. The word-choice also underscores the opposition of Kyllene and Olympus as mountains, homes, and the poles of Hermes’ journey, since his home on Kyllene had an αὐλή (26); cf. 72* for a similar resonance. ἄμβροτοι could probably suggest three things: (i) that the cows are immortal, like Achilles’ horses (Il. 16.867 ἄμβροτοι); (ii) that the cows will not die naturally, but might be mortal under particular circumstances, like Helios’

For μέν ... αὐτὰρ ἄρ’ cf. Il. 2.103–4. D’s θέων is also apt for a future messenger (cf. LfgrE s.v. θέω 1c; Herm. 304 θοῶς, H.Pan. 29 θοὸς ἄγγελος of Hermes). The formulaic background of ἵκανε¦ θεός+ (Il. x3, H Ap. 109) does not demand θεῶν here. For θέων ... θεῶν, cf. H.Ap. 108–9 θέειν ... θεῶν. 111 Some scholars (e.g. Shelmerdine 1981: 104, followed by Vergados and Schenck zu Schweinsberg) infer instead that numerous gods have a collective herd pastured together on Pieria, and that Hermes only steals cows belonging to Apollo. An ancient audience would have required more guidance in line 75 to reach the unusual ideas that numerous gods own cattle and that they keep them in an amalgamated herd. Further, 493–5 and 567–8 imply that Hermes is not receiving ownership just of a small number of Apollo’s cows. 110

186

C O M M E N TA RY 7 2 cattle (Od. 12.131 οὐδέ ποτε φθινύθουσιν, mentioned in this connection by e.g. Shelmerdine 1995); (iii) that they are merely ‘related to the immortals’, like Achilles’ armour (Il. 17.194) or the gods’ gifts (H.Ap. 190).112 I think an ancient audience would unavoidably have entertained the most common sense (i), and been surprised since Hermes plans to eat (64) and presently kills two of the cows (119); then one must assign more weight to (iii) or possibly (ii). This process – practically misdirection – draws attention to the change Hermes effects in the cows’ state, which will be closely tied to his power over animal reproduction (493–5). This convolution appeals to me; others may be tempted to cut through it by reading ἄμβροτον (cf. anon. PMG 926a.1 ἄμβροτοι λείμακες). 72 βοσκόμεναι λειμῶνας ἀκηρασίους ἐρατεινούς: The participle reveals that all the cows are female, in contrast to sources which include some male cows in the stolen herd.113 βοσκόμεναι recalls the previous female grazer at 27 |βοσκομένη (the tortoise; cf. 232*), building on the less obvious parallel 71* αὖλιν ~ 26 αὐλείηισι. The echoes seed the idea that the tortoise-episode and the theft-episode will be structurally paired (Introduction §3.2). The tortoise was grazing on ἐριθηλέα ποίην, and the meadows of Pieria (~ πίειρα) are fertile too. ἐρατεινός is a normal epithet for localities and for food, but not for meadows specifically, so one may sense focalization through the cows’ appreciation of the grazing (as in 198* γλυκεροῖο); ἀκηράσιος emphasizes the wealth of fodder. However, the adjectives have a richer range of associations. Uncut meadows are generally set aside for sacred purposes, like the ἀκήρατος meadow envisaged at E. Hipp. 73–6, the ἄτομοι uplands sacred to Zeus at S. Tr. 200, or ἀδρέπανος sacred land in fr. 978. This reinforces Pieria’s pristine state and hence the transgressive nature of Hermes’ act (as suggested in 71* by the lack of defences). Moreover, ἐρατεινός and ἀκηράσιος cast Pieria as ‘virgin’ territory: cf. Ibycus PMG 286.4 παρθένων κῆπος ἀκήρατος, Rhianus 3.4 G–P on Theodoros’ γυίων ἄνθος ἀκηράσιον; the ἱμερτὸν λειμῶνα of H Dem. 417 is the scene of Persephone’s abduction. And the cows will indeed need Hermes to introduce them to sex (491–5). Finally, Choerilus fr. 2 casts poets as having domesticated the parts of an originally ἀκήρατος meadow of poetry. This metaphor could make Hermes’ incursion into the uncut ­meadows of On the range of ἄμβροτος and ἀμβρόσιος see DGE s.vv., and n. 296 below. Real-world, mortal cattle formed a herd sacred to the Sun at Apollonia: Hdt. 9.93.2. 113 All female: Hor. C. 1.10.10 amotas, Philostratus Im. 1.26.3 (< Alcaeus?, but not ΣAb Il. 15.256), S. Ichn. 82, 167, etc., Apollod. Bibl. 3.112, and various scholia. Males included: LIMC s.v. Hermes nos. 242a and 245, Ant. Lib. 23.3 πόρτιας δώδεκα καὶ ἑκατὸν βοῦς ἄζυγας καὶ ταῦρον, Timotheus Anim. 19. In Ant. Lib., πόρτιας betrays a poetic source, perhaps ‘Hesiod’ (fr. 256) or Nicander (fr. 40). 112

187

C O M M E N TA RY 7 3 – 7 5 Pieria interact with his recent invention, which will transform the primordial musical landscape. 73 ἐΰσκοπος Ἀργεϊφόντης: A formula deployed so that both words suit the context: Hermes had to be sharp-eyed to find the cows (cf. 65 σκοπιήν, the lengthening shadows in 70, the distance travelled; he has justified 15 νυκτὸς ὀπωπητῆρα); Ἀργεϊφόντης was most commonly interpreted with reference to Hermes’ theft of Io after he outdid Argus’ watchfulness.114 In this sense, Ἀργεϊφόντης is proleptic; so already Pierron (1875) 542. 74 πεντήκοντ’ ἀγέλης ἀπετάμνετο βοῦς ἐριμύκους: Epos likes to use ‘fifty’ as a convenient round number, and in Od. 12.130, 14.15 this applies specifically to the size of herds. Most sources do not specify whether Hermes took all the cows, but at S. Ichn. 11–13 ἅπα]ν̣τα φρ̣[οῦδα follows references to Apollo’s cows and calves. For the 113 animals stolen in Antoninus Liberalis see n. 113. For (περι)τάμνομαι in such contexts see Il. 18.528, Od. 11.402, 24.112, A.R. 2.143; here the imperfect may suggest a protracted process, but even so the impressive action is described with extreme brevity and simplicity.115 Romani (2008) discussed the pacing of the hymn’s different descriptions of Hermes’ theft. The epithet ἐριμύκους raises tension, since the noise could betray Hermes; the cows low at Ant. Lib. 23.4, and cf. V. Aen. 8.217. This may be a ­traditional effect: at Od. 15.235 Melampous drives off the βοῦς ἐριμύκους of Iphicles. Cf. 105*.

75–8 75 πλανοδίας δ’ ἤλαυνε διὰ ψαμαθώδεα χῶρον: πλανοδίας is either a noun (inner object; plural to express convolution) or a feminine adjective (threetermination despite being a compound). Either way, the range of connotations is not greatly affected: the stem πλαν- describes misdirecting, misdirected and undirected travel. For the first see LSJ s.vv. πλάνος, πλανάω I.3: Hermes’ route The point about ἐΰσκοπος was also made by Fernández Delgado (1990) 205 n. 18. Argus’ name suggests he was originally a guard-dog, opposed to potential thieves: see West (1978) 368–9, LfgrE s.v. ὁ Ἄργος II, Zografou (1997). For ancient glosses on Ἀργεϊφόντης see esp. Et.Gud. pp. 185–6. Pelliccia (in Watkins 1995: 383–4) plausibly derived the prefix from *h2rg΄eh?- and related it to Argive ἀργᾶς ‘snake’: so S. fr. 1024 applies the word to Apollo slayer of Python. The word would have been reanalysed in a region lacking the word ἀργᾶς. Parthenius’ application of Ἀργεϊφόντης to Telephus remains mysterious: Lightfoot (1999) 197–8. 115 Of course ἀπετάμετο does not scan, but I assume that in most cases the hymnist had the facility to describe the action differently if he wanted (e.g.) an aorist. 114

188

C O M M E N TA RY 7 6 is designed to deceive Apollo, as we hear in 76–8. The second nuance suggests both that the cows go off-road on the sand (as horses πλανόωνται with poor steering: Il. 23.321) and that such deviation is reprehensible; cf. Hsch. π.2572 πληνοδίαι: παρανόμωι· τετιμημένηι [i.e. in court]· τῆι πεπλανημένηι τῆς ὀρθῆς ὁδοῦ, τουτέστιν ἀδίκωι, which also shows that πλανοδίας could be a threetermination adjective. (Here πλᾱν-, not hyper-Ionic πλην-, is the expected result of metrical lengthening: GH i.97–8.) The third, commonest sense of πλαν- pulls πλανοδίας ἤλαυνε towards an oxymoron combining directed and undirected travel; Hermes has nothing to gain from meandering, but takes a markedly unexpected route (see 75–8* below, and Introduction §1.3 for the contrast with Apollo’s walking). Forms of ἐλαύνω will come to articulate the successive phases of Hermes’ journey (cf. 96, 102, 106). For the traditional connection of proceeding from Pieria towards sand, cf. Il. 14.226 (Ἠμαθίη to the North), and H.Ap. 217 (ἠμαθόεις Lektos to the South). ψαμαθώδης terrain sets up Hermes’ track-trick in what follows. The combination with χῶρος recurs only at Q.S. 7.116 (in a different context: a simile of a river breaking its banks). The suffix -ώδης was originally separate from -οειδής and applied to smells; for its extension cf. Il. 13.53 λυσσώδης, Hes. Op. 584 καματώδης. 76 ἴχνι’ ἀποστρέψας: Epos prefers ἴχνιον to ἴχνος where metrically viable; the manuscripts’ corruption to ἴχνη is paralleled at Od. 19.436. ἀποστρέφω is often used of turning back animals in motion (e.g. charging soldiers, horses, plough-oxen) or body-parts. The hymnist, however, makes ἴχνια the object: this clarifies how Hermes’ action is designed to delude Apollo, but also leaves a gap in which the audience has freedom to conjecture what exactly Hermes did: the hymnist entices our imagination into a puzzle. δολίης δ’ οὐ λήθετο τέχνης: τέχνη combines both Hermes’ ‘craftiness’ and the elements of ‘craft’ which it will entail.116 The formulaic epithet δολίη brings out the former sense, as in the lemma’s earlier appearance at Hes. Th. 547, of Prometheus’ butchery-trick at Mekone;117 compare Hermes’ κλεπτοσύνηισι … τεχνηέσσαις in Phoronis fr. 5.3. On the other hand, Hermes’ sandals will involve craft (78–86*), and the studied vagueness of ἴχνι’ ἀποστρέψας is compatible with some veterinary wizardry.

For the craftiness of craft see 39–51*. The stem, already seen in 25 τεκταίνω, occurs more frequently in Herm. (x8) than other parts of Epos. 117 This intertext will soon gain significance (108–11*, 112–41*); cf. δολίη τέχνη also at Th. 540, 555, 560. This traditional language recurs at Od. 4.455, but references to Hesiod’s Prometheus are in the hymnist’s mind (Introduction §1.3). For butchery as a craft see Berthiaume (1982), Durand (1989). 116

189

C O M M E N TA RY 7 7 – 7 8 77 ἀντία ποιήσας ὁπλάς: Further ambiguity: the adverbial neuter ἀντία could mean that Hermes made the hooves (i) face each other (i.e. he ἀποστρέφει one pair), (ii) face the opposite direction – to their previous state or to the direction of ἤλαυνε – or (iii) have some other (non-directional) opposed quality. Compare e.g. (i) Hes. Op. 481 ἀντία δεσμεύων + Σvet. ἀλλήλοις ἐναντίως τιθέντες δεσμεύουσιν, (ii) Hp. Ulc. 26 τὰ ἀντία ὅκως ἂν ὁ ῥοῦς γίνηται τοῦ αἵματος, and (iii) Arist. Anal. 55b12 ἐὰν ἐναντίως τεθῶσιν αἱ προτάσεις. The first two interpretations raise the further puzzle, ‘How?’118 The latter two appear to be confirmed by the following words; in 345, |ἀντία βήματα will have sense (ii). Again, ancient audiences’ expectations about the story here cannot be known. Sophocles’ interpretation, however, may be glimpsed in Ichn. 121–2 εἰς τοὐπίσω τὰ πρόσθεν ἤλλακται, τὰ δ’ αὖ | ἐναντί’ ἀλλήλοισι συμπ[επλεγ]μένα, which I take to mean ‘the front steps have been shifted backwards, while the others are interlaced in opposite directions to each other’. Sophocles has Hermes outdo himself, by applying sense (i) of Herm.’s ἀντία to the rear hooves and sense (ii) to the front. See also 77–8*. 77–8 τὰς πρόσθεν ὄπισθεν | τὰς δ’ ὄπιθεν πρόσθεν: The wording represents the intricate opposition involved in the trick, but Hermes’ ploy is finally revealed as comparatively simple: he gets the cows’ front hooves to go in the rear and vice versa by driving them backwards, without any special manipulation of their legs or hooves. Some analyst critics (e.g. Schulze 1868: 21–2) complained that Hermes’ trick is unworthy of him, and Apollo easily realizes that he should follow the tracks in reverse. But it is doubtful whether Hermes hopes to get away with his theft completely, and when Apollo sees the tracks he seems not to be able to follow them and to be surprised by their form (218*, 219–26*, 342–5*), so that the method does display once more Hermes’ improvisational ingenuity. The motif of driving or dragging animals backwards may have a folkloristic background; it is found in various accounts about Cacus (L. Cassius Hemina fr. 3, D.H. Ant. Rom. 1.39, Livy 1.7.5, V. Aen. 8.209–12, etc.), and in Pliny Nat. 8.95 the marauding hippo takes care to reverse its steps. Reversing horseshoes is a more practicable and common motif.119

118

Autolycus would steal livestock and alter the owner’s brands. Sisyphus caught him by carving his initial in his animals’ hooves, which could not be so easily modified: Hes. fr. 67, Polyaen. 6.52. The story, like Herm., suggests an analogy between tracking animals and reading letters. 119 See Radermacher 192–3, Allen and Halliday 292–3, Thompson (1955–8) K534 for more or less precise parallels. Clauss (2016: 66–73) presents the argument for Vergil’s use of Herm. in constructing his version of the Cacus story, and this may well have

190

C O M M E N TA RY 7 8 For the slightly irregular ποιήσας ... ὄπισθεν, Matthiae (1805) compared Thucydides 2.83.5 τὰς πρώρας μὲν ἔξω ... ἐποίησαν. For variatio such as ὄπισθεν/ὄπιθεν see Hopkinson (1982); Richardson accepted M’s πρώτας in 78 as further variatio; but ‘first legs’ for ‘forelegs’ is hard to parallel in early texts (first Phlegon Mir. 34, Ps.-Manetho Apotel. 2.97). 78 κατὰ δ’ ἔμπαλιν αὐτὸς ἔβαινεν: Since κατέμπαλιν does not occur, κατά … ἔβαινεν is tmesis. As Hermes descends from Pieria to the shore he tries walking backwards like the cows. But soon after he encounters the sand (suggested by 75) he finds this unsatisfactory: he must constantly look over his shoulder, and leaves behind revealing baby-sized prints. Those considerations prompt his next stratagem. 75–8 Reading the cows’ traces: Hermes leaves clear but misleading traces inscribed in the sand, which Apollo must read to infer the cows’ path.120 The hymnist in several ways replicates this problem for the audience, and both the metaphors of ‘tracking down’ sense and of poetic ‘feet’ were live ones (see e.g. Plato Parm. 128c ἰχνεύεις τὰ λεχθέντα, and the sustained metapoetic wordplay on ἴχνος in Simias fr. 26). The description begins with 75 πλανοδίας, which is syntactically ambiguous, apparently oxymoronic with ἤλαυνε, and has various connotations of which the most common (πλανas meandering) is the least apt. The vagueness of 76 ἴχνι’ ἀποστρέψας and ambiguity of 77 ἀντία then force the audience to make an inference about how Hermes is creating the tracks. To this extent, the hymnist exercises his poetic τέχνη in a δολίη way for the audience, while describing Hermes’ δολίη τέχνη against Apollo. In 77–8 τὰς πρόσθεν ὄπισθεν, τὰς δ’ ὄπιθεν πρόσθεν the hymnist constructs a phrase which matches Hermes’ inversion of the forward sign and the rear. For this approach to hymnic charis by demonstrating likemindedness with Hermes, see Introduction §3.3–4.121

been an intertext open to Vergil’s readers. But Clauss did not discuss how many shared elements of the story (reversal of prints, subtlety, cave, cow-sacrifice and aetiology) seem already to have been in Cassius Hemina. Crudden (1994: 26) emphasized instead the greater similarity between Herm. and Dionysius’ account. 120 Contrast Ant. Lib. 23.3, where he ties branches to the cows’ tails in order to hide the traces. Keith (1992) 108–12 considered whether Ov. Met. 2.686 arte sua siluis occultat refers to this rather than to hiding the cows in the woods. If so, the detail probably derives from Nicander, but the plural siluae very rarely means ‘branches’ (e.g. Cons. Liv. 255). 121 Bergren (1982) 98–9 went further and discussed the hoofprints, among other features of Herm., as an analogue of a deconstructionist account of writing, namely a system of signs which appear to, but never quite, point to the signified.

191

C O M M E N TA RY 7 9 The decoding of imprints, an important skill in hunting (X. Cyn. 8, 10.5), receives various interesting treatments in Greek texts: see e.g. A. Ch. 205–10 (with the implication, as in Od. 4.149, that Hermes’ footprints here might help Apollo identify the culprit as a family-member) and E. El. 532–7, Ar. Nu. 973–6. An amusing opposite to Hermes’ deception is Clement’s report that courtesans might have messages like ‘Follow me’ hammered into their soles (Paed. 2.11). The fate of the prints (here, and where Apollo sees them) is left open; marks in a road near Agyrium were identified with the petrified tracks of Heracles and Geryon’s cattle (D.S. 4.24.2–3), but no aetiological role is expressed here.122

79–86 79 σάνδαλα δ’ αὐτίκα ῥιψίν: I shall discuss the general significance of the sandals below, 79–86*. The new topic displaces αὐτίκα from first position, expressing the urgency of Hermes’ need to change strategy (78*). αὐτίκα ῥιψίν is Postgate’s emendation for Ω’s αὐτίκ’ ἔριψεν. ῥῖπες (or ῥίπεα or ῥῖποι) are pliable shoots, often woven together in two perpendicular layers. Material, texture and usage vary, from rush matting as a cheap mattress (e.g. Aen. Tact. 29.6) through to coarser wicker-work (e.g. Od. 5.256 ῥίπεσσι … οἰσυΐνηισιν as a shield against waves; Ps.-Arist. Prob. 15.6 911b5 presupposes that ῥῖπες let through light). Vergados revived Matthiae’s αὐτίκ’ ἔραψεν with colon at lineend. The idea of ‘stitching sandals’ would at first seem literal and ordinary, before 80 ἄφραστα ... διέπλεκε reveals that it is metaphorical (stitching up as designing a trick) and extraordinary, and that actual stitching was not involved. However, this reconstruction entails an asyndeton much harsher than any other in Herm.’s narrator-text (see 151*), and it removes a plausible motivation for corruption, namely a scribe struggling with the relatively unfamiliar ῥιψίν. ἐπὶ ψαμάθοισ’ ἁλίηισιν: Epos and early Ionic inscriptions far preferred -ηισι and -οισι to -ηις and -οις (Smyth 1894: 364–70, 379–82). Herm. demonstrates this preference most strongly, with eighty certain long forms to one certain short one (291); I have added an elision in six further cases where -οις or -ηις precedes a vowel, but not where matters are complicated by a digamma (151, 245).123 122

Cf. the prints preserved as monuments in Hdt. 4.82, and the later practices discussed in Dunbabin (1990). 123 Janko (1982) 54–7 observed that the ratios (short forms/all forms) and, less clearly, (certain short forms/all short forms) increase in Epos over time. (Details of the data are criticized in Jones 2010: 300–5, but the general trend seems to stand.) Janko therefore interpreted Herm.’s figures as an archaizing preference (cf. Introduction §1.1). In elegy, there is regional variation too: mainland poets – whose vernacular used short forms – admit more of them than their Ionian contemporaries (West 1974: 93–4).

192

C O M M E N TA RY 8 0 – 8 2 80 ἄφραστ’ ἠδ’ ἀνόητα: The common senses give ‘impossible to perceive or apprehend’. This is not apt of the sandals themselves, but hints at their effect on Hermes’ footprints (cf. 353 ἄφραστος*). Context instead requires ἀνόητος to be ‘undevised’; the formal similarity of the adjectives suggests a complementary meaning for ἄφραστος – ‘unconceived’ rather than ‘indescribable’ (A. Pe. 165, S. Tr. 694). At Sophocles El. 1262–3 ἀφράστως ἀέλπτως τ’ ἐσεῖδον, the formal parallelism again argues for ‘in a way not conceived’.124 θαυματὰ ἔργα: The prints will indeed inspire θαῦμα at 219. For Hermes’ θαύματα (the noun, not to be confused with the adjective θαυματά), cf. Introduction n. 101. 81 συμμίσγων μυρίκας καὶ μυρσινοειδέας ὄζους: μυρίκη denotes one of the tamarisk species, hardy bushes often found near rivers or the shore (so e.g. Il. 21.350, Thphr. HP 1.4.3). Their slender, pliant shoots are ideal for Hermes’ work. D.S. 12.12.2 mentions a classical tamarisk garland, Dsc. Eupor. 2.63.4 a bread-basket woven from it, and Plu. Gen. Soc. 578e a palliass (ῥίψ) made from it and vitex. Strictly, Hermes does not mix ‘tamarisks’: the plural of μυρίκη is being used of countable bits of tamarisk (cf. λύγοι, ‘vitex-shoots’).125 μυρσινοειδέας may be an adjectival substitute for unmetrical μυρσίνης (cf. GH ii.14 for such adjectives). The pliancy of myrtle (Myrtus communis) was crucial for its major use in garlands, and it grows in sandy soil. However, the -ειδής suffix retains its semantic value elsewhere in Epos, so the second species may be ‘myrtle-like’; specifically, Dsc. Mat. 4.7.1 identifies μυρσινοειδές with κληματίς, perhaps the equally pliant periwinkle (Vinca herbacea).126 Overall, our line resembles Il. 10.467 συμμάρψας δόνακας μυρίκης τ’ ἐριθηλέας [cf. 82 νεοθηλέος] ὄζους, where the ‘Hermetic’ character Odysseus is creating a sign for where he is leaving the spoils of a night-raid (see Introduction §1.3). 82 τῶν τότε συνδήσας νεοθηλέος ἀγκάλα ὥρης: The line redescribes the action of 79–81, now in terms of ‘binding together’ (see 83*). The end poses a textual problem: νεοθηλέαν ἀγκαλωρήν M vs νεοθηλέος ἀγκαλὸν ὕλης Ψ.

Dionysus rises with ἀφράστωι (‘inaudible’) πεδίλωι at Nonn. D. 16.342, but by then ἄφραστος is a more common word, and a reference to Herm. is dubious. 125 Alternatively one could consider reading μυρίκης; the treatment of a genitive and an adjective as parallel qualifiers of ὄζους would resemble e.g. Pl. Tht. 163a ὅ τε σὸς καὶ ὁ Θεοδώρου λόγος, particularly if μυρσινοειδέας ~ μυρσίνης (see main text). 126 -ειδής in Epos: ἀλλοειδής, εὐειδής, ἠεροειδής, θεοειδής, ἰοειδής, μυλοειδής. However, -ώδης had developed from ‘smelling of ’ to ‘consisting of ’ (73*). For later weakening of -ειδής see e.g. Nicander’s repurposing of ἰοειδής to mean ‘poisonous’ (Th. 281). For κληματίς cf. Hsch. κ.2959 κλῆμα: ὑπόδημα, where Latte’s assumption of a misinterpretation of LXX Jeremiah 31.32 κλήματα θάλασσαν διῆλθεν is unconvincing. 124

193

C O M M E N TA RY 8 3 M’s ἀγκᾱλωρή is not a plausible compound, and νεοθηλέαν is a false ending. Ψ’s ἀγκαλός/-όν meaning ‘armful’ would have to be an unattested byform of the normal words ἀγκαλίς and ἀγκάλη (which dates back at least to the thitd century: see Hsch. α.3417). A word beginning ἀγκαλ- is eminently suitable; PLond. 131.396 (78 ce) δεσ[μεύ]οντος ἀγκάλας and POxy. 3354.9 (257 ce) δέσι̣ς̣ ἀγκαλῶν, both referring to armfuls of pruned vine-shoots, offer precise though distant parallels. Ψ’s genitive is the easiest case for νεοθηλής metrically. My conjecture supposes that Ψ’s ὕλης extruded some form of ὥρη, visible in M. Since I take ὥρη to mean ‘season (of growth)’ but not ‘seasonal growth’, the object of συνδήσας was not ὥρην but whatever began ἀγκαλ-.127 One could admit ἀγκαλόν in spite of the lack of parallels, or ἀγκαλίδ’ or ἀγκάλω. However, I have hesitantly suggested the dual ἀγκάλα, which best explains the transmitted readings.128 Scribes could have misunderstood it as a neuter plural ἀγκαλᾰ; one patched the metre with ἀγκαλόν. M’s compound perhaps arose via ἀγκάλ’, or the ligature ἀγκαλ``, with the accusative ending reintroduced at line-end. One could alternatively assume that ὕλης is original and e.g. ἐν ὥρᾳ was a marginal gloss on νεοθηλέος, but the path from there to M is harder. Many editors have been happy to accept Ψ and to leave M’s reading as an unexplained aberration. 83 ἀβλαβέως ὑπὸ ποσσὶν ἐδήσατο σάνδαλα κοῦφα: The variation of formulaic ὑπὸ ποσσὶν ἐδήσατο καλὰ πέδιλα marks Hermes’ sandals as abnormal.129 But despite their unsophisticated construction, he ties them on without hurting himself. ἀβλαβέως also suggests that they will subsequently be no hindrance to his walking: see DGE s.v. βλάπτω I, and e.g. GVI 999.3–4 (Myrina, second century) μακρὴν γήραος οἷμον ἐνὶ θνητοῖσι πορευθείς | ἀβλαβέως, Arr. Parth. fr. 85 οἱ δὲ κύκλους ἐκ λύγων τοῖς ποσὶ περιαρμόσαντες … ἀβλαβῶς ἐπήρχοντο κατὰ τῆς χιόνος. The adjective κοῦφος supports this implication: it is already used of nimble walking in Il. 13.158, Sc. 323; later the Graces

At X. HG 2.1.1 ἀπὸ τῆς ὥρας ἐτρέφοντο, ‘they lived off the season’ is comprehensible without positing the extra sense ‘seasonal growth’, pace Radermacher (1933) 156. For the genitive in my proposal cf. Pi. fr. 122.8 μαλθακᾶς ὥρας ἀπὸ καρπὸν δρέπεσθαι. 128 The first-declension feminine dual ends -ω in Mycenaean and Hes. Op. 198, but -α is borrowed from masculine first-declension nouns at e.g. Hp. Aff. 9 δεούσαιν, IPArk 14 κράναιυν (Orchomenos, 360s), IG IX.12 iii.706 κόραιν (W. Lokris, c.280), as well as in Attic. 129 This effect would be enhanced if σάνδαλα were not object but a predicate: ‘bound the armfuls underfoot as sandals’; however, this strains 84 τά. The formula: Il. x5, Od. x7; seven instances start with the extension ποσσὶ δ’ ὑπὸ λιπαροῖσιν. 127

194

C O M M E N TA RY 8 4 – 8 6 gambol in σάμβαλα κοῦφα (Diotimus 1.6 G–P). The repetition of σάνδαλα from 79 frames the description of Hermes’ engineering much as the repetition of φέρων ἐράτεινον ἄθυρμα did at 52*. 84 αὐτοῖσιν πετάλοισι: Hermes has no time to remove the leaves, which will in any case help to confuse the traces.130 For |αὐτοῖσιν + dative compare 69, and see 78–86* for the framing effect. 84–5 τὰ κύδιμος Ἀργεϊφόντης | ἔσπασε: The πέταλα have explicitly not been plucked: τά refers to σάνδαλα. But σπάω applies naturally not to σάνδαλα but to the shoots (cf. Od. 10.166 σπασάμην ῥῶπάς τε λύγους τε): this reinforces the oddity of Hermes’ materials. 85 Πιερίηθεν ὁδοιπορίην ἀλεείνων: Πιερίηθεν is much better taken with the following description of motion than with ἔσπασε. ὁδοιπορίη is ‘travelling by road’ as opposed to cross-country; Radermacher compared Hdt. 8.118.1, where it opposes sea-travel. Our phrase recalls one sense of 75 πλανοδίας* (‘going off-road’), and motivates Hermes’ coastal route: travelling by road would have left less clear traces (352–4; cf. E. El. 534–5), but would expose him to witnesses (cf. 87). The sense is similar to Il. 6.202 πάτον ἀνθρώπων ἀλεείνων|, even if the motivation is different. 86 οἷά τ’ ἐπειγόμενος δολιχὴν ὁδὸν αὐτοτρόπησεν: The three participles of the paradosis make a crowd, and I have proposed αὐτοτρόπησεν: this gives a pair of relative clauses with chiastically positioned participial phrases, as follows: ‘sandals, which (a) he plucked (b) while avoiding walking on roads, and of a kind which, (b) as he hastened his journey, (a) he made characteristically’. The corruption could rest on the similar ligatures for -ας and -εν (cf. 437). αὐτοτροπέω probably conveyed something like ‘having acted after his own manner’, here implying prompt improvisation.131 We have seen that Hermes’ behaviour was indeed characteristic and idiosyncratic; perhaps the poet coined the hapax verb to garnish the idea of uniqueness. The sequence ὁδοιπορίην ἀλεείνων ... ὁδόν even phrases his movement as a ὁδός without a ὁδός (cf. 75*). The formula δολιχὴν ὁδόν (Od. x3) gains a neat sonic

Adorjáni (2011) 132–3 suggested that the πέταλα foreshadow the wings on the adult Hermes’ sandals. 131 Cf. ἰδιότροπος, ὁμότροπος, τοιουτότροπος and ἀλλότροπος, ἑτερότροπος. Hippocratic writers use ἀλλοιοτροπέω and κακοτροπέω intransitively. For a traveller’s typical hasty improvisation see also Posidipp. 94. 130

195

C O M M E N TA RY 7 9 – 8 6 connection to δόλιος. For transitive ἐπειγόμενος see e.g. Od. 2.97 ἐπειγόμενοι τὸν ἐμὸν γάμον. I take the paradosis to mean ‘[Hermes plucked sandals while he avoided roads,] inasmuch as he acted after his own manner while he hastened his long journey’.132 οἷά τε + participle (see Ruijgh 1971: 532–3) goes with αὐτοτροπήσας since it cannot go with ἐπειγόμενος, either in the sense ‘like someone hastening’ (the substantivization of ἐπειγόμενος is harsh, and it is untypical of travellers to make their own sandals or to avoid roads), or ‘inasmuch as he was hastening’ (Hermes hurried, but was motivated by secrecy). Adorjáni (2011) 135 suggested αὐτοτροπῆσαι (transitive), with which ἐπειγομένου or ἐπειγομένωι gives ‘such [sandals] as to make his long path idiosyncratic as he hurried’. One reason to admit a transitive sense is the unattributed gloss in Hesychius α.3182 ἀλλοτροπῆσαι: μεταθεῖναι, which one could even adopt to give ‘such sandals as to change the character of his long path …’. The truth may lie farther afield. 79–86 Improvising sandals: For ancient sandals see Lau (1967), who observed that shoe-making, at least in Athens, was a prototypical craft (182–3). The sole of a sandal was normally cut from two sheets of leather, which were then stitched around a sturdier insert; weaving and tying were involved in the strapwork. Although stitching has appropriate connotations of trickery (Lau 1967: 43, 196), Hermes has no time for needles or fancy strapwork, but weaving and tying apply to the soles (διέπλεκε, συνδήσας) to create thematic consistency with his craftiness elsewhere in the Hymn.133 An audience may have seen this weaving as an ingenious borrowing of technology from a different area: Arrian (quoted on 83*) describes a snowshoe as a ‘circle made of vitex’.134 Hermes’ second piece of improvised construction bears comparison with that of the lyre: broadly, the techniques are complementary (binding and weaving versus carving, boring and fixing), but a shared characteristic is the harmonious combination of disparate materials (συμμίσγων; cf. 39–51*). As Clay (1989) 113 noted, an important difference

In x, αὐτοτροπήσας was displaced into the margin by αὐτοπρεπὴς ὥς. I take this to be a minuscule variant noted through collation, rather than a clue for emendation. αὐτοπρεπής is not otherwise attested, though Hippocratic authors use ξενοπρεπής ‘unfamiliar’. 133 Introduction §3.3.6. 134 Radermacher 247 compared Alpine snowshoes woven from small brushwood. The purpose of Hermes’ sandals is nowhere stated explicitly, but their main effect will be to confuse his track (218–26; cf. 78*); they might, like a snowshoe, make his journey over sand easier, but if one dwells on this point, the poor cows are left with a lot of wading to do.

132

196

C O M M E N TA RY 8 7 is that the lyre is a long-term contribution to culture whereas the sandals are disposable. Here the materials – myrtle (probably) and tamarisk – show harmonious alliteration, but they had contrasting social values: at least, D.S. 12.12.2 notes that in fifth-century Thurii a tamarisk garland was the false ­prosecutor’s ‘prize’, whereas myrtle garlands mark proper festival participation. Hermes demonstrates creativity by using myrtle for an ‘antiwreath’ (not a circlet but a disc, not for the head but for the feet). Finally, the humorous description of the sandals as κοῦφα (normally ‘light-weight, wieldy’), sandwiched between the ‘armfuls’ of shoots and the revelation that they still had their leaves, emphasizes the actual opposition between these sandals, whose job is not to fit like a slipper, and the laborious workmanship of the thin leather soles of good sandals.135 κοῦφα also points up the contrast with the golden shoes which allowed the adult Hermes to travel so freely, and of which his invention here is proleptic (Il. 24.340–2 = Od. 5.44–6; for art see Yalouris 1953). Lines 84–6 resume a series of details from 69–75, and so bring Hermes’ visit to Pieria to a close: cf. 84 |αὐτοῖσιν πετάλοισι* with 69 |αὐτοῖσιν, 85 motion Πιερίηθεν ~ 70 reaching Πιερίην, 86 ἐπειγόμενος ~ 70 θέων, 83 Ἀργειφόντης ~ 73, 85 ὁδοιπορίην ἀλεείνων ~ 75 πλανοδίας. If Eratosthenes Hermes fr. 9 ‘He was stitching on the sole of a light φαικάσιον’ referred to Hermes making shoes for his cattle-theft (which is uncertain), the effect was markedly different: stitching appears; to judge from Plu. Ant. 33.7, the φαικάσιον was actually ‘light’ (whereas κοῦφα is a joke), and may have connoted Hermes’ patronage of the Hellenistic gymnasium.

87–94 87 τὸν δὲ γέρων ἐνόησε δέμων ἀνθέουσαν ἀλωήν: Human witnesses are a common feature of this myth. For the view that this one has been relocated from an earlier story about Battos, see Introduction §3.3.4, and for his character in this episode see below, 87–94*. ἀλωή can mean ‘threshing-floor’, but here ἀνθέουσαν points to its other sense, a cultivated plot of land, with vegetable beds, fruit trees (esp. vines: Cf. 349*. Hermippus fr. 17 uses αὐτοσχεδές of a simply-worked woman’s shoe: Hermes’ sandals are an extreme instance of this ‘improvisation’. For ancient valorization of well-fitting shoes see e.g. Diggle (2004) on Thphr. Char. 4.2, Gal. U.Part. 3.10 pp. 235–6. Cursaru (2011) presented a very different reading of Hermes’ sandals, as contributing to the initiatory and catabatic qualities which she saw in his journey with the cows.

135

197

C O M M E N TA RY 8 8 cf. 91) and flowers, but not cereals.136 δέμω is ‘build’ in a narrow, solid sense; when it is applied to a plot of land, one might imagine the initial construction of a paved area (cf. Od. 7.123, for drying grapes) and protective wall or fence; perhaps also of trellising (cf. δέδμανθ’ of bowers at Theoc. 15.119–20).137 The implication from δέμω is that the plot is new. ἀνθέουσαν could refer to existing plants (cf. 96), or be proleptic; it replaces βρίθουσαν from Il. 18.561, since these vines are not yet bearing fruit (Vergados 50). δέμων might seem to jar with 207*, where the man will recall his occupation here as ‘digging’. But since both passages, and 90* σκάπτεις, can be taken as very brief summaries of the man’s diverse activities, there is no strong reason to emend. 88 ἱέμενον: For the assimilation of (ϝ)εμαι ‘strive’ and the middle of ἵημι (‘direct oneself ’), see LfgrE s.v. ἵημι II A 2. πεδίονδε δι’ Ὀγχηστὸν λεχεποίην: The reference to Onchestos is not found in other versions, and therefore requires interpretation (Allen and Sikes 133, for example, took it as the main evidence for a Boeotian origin of Herm.). I see four senses in which Onchestos is a good place to mention at this point. (i) The topography explains why Hermes’ attempt to evade open roads has failed, since Onchestos was something of a bottleneck for Boeotian travel. It lay on the south side of the main East–West road through Boeotia, itself part of a major route from Malia towards Attica and the Peloponnese; there were few other passable options. H.Ap. 231–6 also clearly associates Onchestos with this road.138 (ii) The topography enhances our image of the old man’s activities. Onchestos lies on a low saddle, beyond which the road descends to the South-East towards the Teneric plain (πεδίονδε), at the primarily agricultural shore of Lake Copais. The area is repeatedly described in terms of fertility; here Onchestos is λεχεποίης, and later its lovely trees are The two senses of ἀλωή may be connected by synecdoche: at Od. 7.122–6 the first things described in Alcinous’ ἀλωή (‘plot’) are open floors for drying and trampling grapes. Vines in ἀλωαί: cf. formulaic γουνὸν ἀλωῆς οἰνοπέδοιο (207), Il. 18.561–8, ΣD Il. 5.90. Fruit trees: Il. 9.540, 21.36–8. Vegetables: Od. 24.247. Flowers: 6.293; cf. Hsch. α.3251 ἄλουα (Cypriot neuter plural): κῆποι. Laertes’ ἀλωή is described at length in Od. 24.220–344. Regarding the orthography of ε-contract verbs such as ἀνθέουσαν vs ἀνθοῦσαν or 315 φωνέων vs φωνῶν, see GH i.60–3. 137 For the ἕρκος of a plot of land see 188*, Il. 5.90, 18.564, Od. 7.113, 24.224 (explicitly made of stone). For vine-training, sometimes on trellises, see e.g. Ar. V. 326, 1291, Billiard (1913) 356–67. ἀλωαί in both senses can also be ἐϋκτίμεναι (x4). But unlike δεμ-, the stem κτι- had shifted towards ‘cultivate land’ by Mycenaean times. 138 See Larson (1968) 109–10, Buck (1979) iv. The bottleneck also explains why Apollo later seeks information there. 136

198

C O M M E N TA RY 8 9 mentioned (186–7). Its gentle slopes of all kinds will be ideal for the man’s vineyard.140 (iii) Onchestos’ famous sanctuary of Poseidon meant that Herm.’s audience had a good chance of recognizing the name, and in some cases the topographical points mentioned. The road made Onchestos ideal for interstate display, and by the late sixth century its sanctuary of Poseidon hosted a horse-racing festival, and contained a poros temple and a large (c.18m x 7.75m) ‘Bouleuterion’ with a meeting-space in between.141 (iv) The hymnist alludes to the break in the narrative of Apollo’s journey at Onchestos in H.Ap.: see Introduction §1.3. 139

89 τὸν πρότερος προσέφη Μαίης ἐρικυδέος υἱός: In epic scenes where a new arrival meets someone already present, there seems to be no etiquette (whether social or formulaic) about which party starts the conversation. Formulas expressing ‘spoke first’ include |τὸν πρότερος προσέειπε, |τόν … ¦πρότερος πρὸς μῦθον ἔειπεν|, and |τὸν προτέρη ⏕ – ¦προσεφώνεε.142 The substitution of προσέφη in the first of these occurs only here and in 189, while προσέφη and μετέφη are almost entirely fixed after the caesura in Epos (x243/247; other exceptions at Il. 16.842, Od. 11.565). As Van Nortwick (1975) 31–5 observed, the formulaic modification draws attention to the parallel with 189, as do the decision to describe Hermes as Μαίης ἐρικυδέος υἱός rather than Κυλλήνιος Ἀργειφόντης, and the avoidance of the natural formula combination προσέειπε Διὸς καὶ Λητοῦς υἱός in 189. With Μαίης ἐρικυδέος υἱός compare the Massaliote text of Od. 1.38 (ΣHMa) πέμπψαντες Μαίης ἐρικυδέος ἀγλαὸν υἱόν, IG XII.2 476 (Mytilene) Ζηνὸς καὶ Μαίας ἐρικυδέος ἀγλαὸν Ἑρμῆ,143 Maximus Init. 138, and Od. 11.576 Γαίης ἐρικυδέος υἱόν. Reference to a person using the name of their mother was much less common than that using the name of the father or both parents (for epigraphic material Haliartos is ποιήεις at Il. 2.503, H.Ap. 243; cf. Apollod. Hist. FGrH 244 F157; Thebes is ποιήεις in Σ E. Ph. 638; Teumessos is λεχεποίης at H.Ap. 224. 140 See more technically Columella 3.1.3–2.6. Onchestos’ slopes are also mentioned at H.Ap. 231, Pi. I. 1.33, fr. 94b.41–9 τί-|μαθεν ... ἀμφικτιόνεσσιν ἵππων τ’ ὠκυπόδων πο[λυ]γνώτοις ̣ ἐπὶ νίκαις, αἷς ἐν ἀϊόνεσσιν Ὀγχη[στοῦ κλυ]τ̣ᾶς … ἐκόσμηθεν. 141 See Spyropoulos (1973) 380, Roesch (1982) 211, 269, Schachter (1986) 207–21. Other early mentions are Il. 2.506, Hes. fr. 219, Alc. fr. 325, Pi. I. (3/4).37. Thebes was probably in control until the Persian War, after which Haliartos took over: SEG 25.554, Étienne and Knoepfler (1976) 222 n. 773. After 338 bce, Onchestos became the centre of the Boeotian confederacy. 142 This system of formulas is alliterative: for brief thoughts on the role of alliteration in generating formulas, see Stanford (1969) 15, Lord (1960) 55–7. ὠκέες ἵπποι < *h1ōk' éues h1ék' uōs shows that it had long been a factor. 143 The date is disputed: third century in SEG 16.466, 1st c. ce in Hodot (1990) MYT 181, 2nd c. ce in Kaibel (1878) 330. 139

199

C O M M E N TA RY 9 0 see Ogden 1996: 94–6). However, sons of Zeus are often distinguished by their mothers (West 1966: 431); this is especially frequent in Herm., with close juxtapositions at e.g. 243–4, 403–8. 90 ὦ γέρον, ὅς τε φυτὰ σκάπτεις ἐπικαμπύλος ὤμους: Hermes, as usual in Epos, includes an address within the first line, to attract attention.144 I take ὅς τε to contain ‘epic’ τε, so that Hermes cites digging around the stems of plants (φυτὰ σκάπτεις), regardless of what the man is doing at the moment, as an emblematic activity of farmers, singled out because it sets up his focus on the vines in 91 (πολυοινήσεις + deictic τάδε).145 Readers who prefer to take 207 ἔσκαπτον as evidence that the man is indeed digging can take the τε in ὅς τε as colourless (as in a few cases in Homer and several in Pindar, Aeschylus, Sophocles and Apollonius), and either emend 87* δέμων or take it in an unusually broad sense encapsulating the various tasks of one setting up a new plot, including digging for e.g. weed-clearing, installing irrigation trenches, and planting.146 The man’s hunch could be a sign of age (as in e.g. Il. 24.359 Priam’s γναμπτοῖσι μέλεσσιν, Od. 2.16 γήραϊ κυφὸς ἔην, Luc. Saturnalia 9 ἐπικεκυφότες ὥσπερ οἱ πάνυ γεγηρακότες) and/or of the effect of gardening (as in Luc. Timon 7 σκάπτει ... ἐπικεκυφώς, Ach. Tat. 1.1.6 ὀχετηγὸς ... κεκυφώς). Hermes probably chose φυτὰ σκάπτεις to bring out the toil of gardening; at least, φυτοσκάφος occurs in contexts emphasizing the relentless nature of the work at e.g. A.R. 1.1172 and Ps.-Theoc. 25.27. Hermes therefore makes nature and culture conspire to emphasize the man’s pitiable lack of physical ease, though it is unclear whether his tone is more pitiful or contemptuous (ὦ γέρον is compatible with either, or with reverence – contrast e.g. Il. 9.115 and Od. 2.201).147

144

Bassett (1934), discussing the exceptions to this principle, overestimated the possibility that they are condescending. 145 Emblematic activity: cf. the noun φυτοσκάφος, discussed below. For the role of relative clauses in addresses see 254*. For the value of περίσκαψις or γύρωσις on plants of different ages see e.g. Gp. 3.11, 5.19–20, 5.26 (for manuring young plants), Aesop 235, Billiard (1913) 327–8; Vergados’ note is helpful. φυτά contrasts with sown crops, and can refer particularly to vines (e.g. Hes. Op. 571–2); it summarizes the contents of an ἀλωή also at Il. 18.57 = 438, Od. 24.227. 146 On colourless ὅς τε see e.g. Il. 11.383 and 17.174, and Ruijgh (1971) 941–2, 986–7, 1001–2 on later authors. Planting-holes: e.g. Thphr. CP 3.12, Billiard (1913) 254– 5. Weeding and clearing: X. Oec. 20.20, Gp. 3.10; see also (without digging) 190 βατοδρόπε*. 147 For the unusually limited terminology for addressing the elderly in Greek see Dickey (1996) 207–12, 248. Whereas the presence or absence of ὦ in classical Attic is a matter of deference, it is only used in c.12 per cent of addresses in Epos, the main

200

C O M M E N TA RY 9 1 – 9 3 This ambivalence turns out to be a rhetorical strategy to soften him up for a threat of further loss. M’s ἐπικαμπύλα ξύλα probably derives from a strikingly erudite concordance annotation: the adjective’s only other use is Hes. Op. 427 ἐπικαμπύλα κᾶλα, where the scholia gloss κᾶλα with ξύλα. We need not, however, import Hesiod’s phrase into Herm. as Allen and Sikes attempted; doing so leaves Ψ’s reading unexplained. 91 ἦ πολυοινήσεις: For the formation of the verb cf. πολυγαλακτέω, πολυκαρπέω. πολύοινος means both ‘having lots of wine’ (e.g. X. Vect. 5.3, Gp. 5.43) and ‘drinking a lot’ (e.g. Pl. Lg. 2 666b, Gal. QAM p. 809); one should not suppress the humorous ambiguity. εὖτ’ ἂν τάδε πάντα φέρηισιν: For the sense ‘bear fruit’ see LSJ s.v. φέρω A V. There is some doubt about whether such 3rd-person singular subjunctives should be spelled -ησι: contrast πίησι on Nestor’s Cup with SEG 26.1139.1 μέλληισι (Hipponion, c.400), Schwyzer i.661 with Peters (1998) 594–6. 92–3 καί τε ἰδὼν μὴ ἰδὼν εἶναι καὶ κωφὸς ἀκούσας | καὶ σιγᾶν, ὅτε μή τι καταβλάπτηι τὸ σὸν αὐτοῦ: εἶναι and σιγᾶν have directive force: see GH ii.316–18, Schwyzer ii.380–2, Allan (2010). καί marks ἰδών as concessive, and τε adds very little to καί, as in 133, Hes. Op. 360, Empedocles x4, A.R. 1.1105; see Ruijgh (1971) 933–5. In 93 I take καταβλάπτηι as 3rd-person active subjunctive (after ὅτε in a generalization: see GH ii.256) with ‘being silent’ as the subject. αὐτοῦ underlines a possessive as at e.g. Il. 6.446, Od. 2.45, Pi. fr. 97. If one takes καταβλάπτηι as 2nd-person passive, both τι and τὸ σὸν αὐτοῦ would be accusatives of respect; for ὅτε μή + indicative expressing a condition see KG ii.184. The lack of real connective at the start of 92 shows curtly that 91 was not a pleasantry, but a setup for the menace which follows: the old man still has plenty to lose.148 As in his previous unanswered speech, to the tortoise, Hermes deploys a proverb in a way which belies his lack of experience of the world (cf. 36). For this proverbial expression for wilfully ignoring a misdemeanour, see especially Ps.-Demosthenes 25.89 τὸ τῆς παροιμίας, ὁρῶντες μὴ ὁρᾶν καὶ ἀκούοντες μὴ ἀκούειν, Thphr. Char. 1.5; elsewhere hearing is

factor here being to front the address within Hermes’ overall utterance: see Kieckers (1908–9) 358–62, and Dickey (1996) 200–6 for Attic. ὦ is expressive at 254*. 148 Vergados objected that since Hermes does not punish the Onchestan’s indiscretion later, we do not have sufficient evidence to call the tone threatening. But the asyndeton implies that the man’s wealth is precarious, regardless of whether or not the narrative loses sight of him before any punishment. The Onchestan also shows every sign of being nervous (201–11*).

201

C O M M E N TA RY 9 4 omitted or replaced by e.g. ‘knowing’; cf. Ps.-Plu. Lib. Ed. 13e, Plaut. MG 572–3. Ps.-Demosthenes and Ps.-Plutarch give the context of parents indulging their children, which suits the age-dynamic here.149 Moreover, it is typical of Hermes to manipulate the evidence of the senses, and to ask the old man to turn his natural characteristics into their opposites (Introduction §3.3.6) – compare his promise to make the tortoise a dead singer (38). For κωφός in the sense ‘deaf ’ see Introduction §1.1. 94 τόσσον φὰς συνέσευε: τόσσον φάς varies ὣς εἰπών to emphasize Hermes’ succinctness (Vergados), and he does not wait for a reply. Chalkondyles’ ­restoration accounts better than Cobet’s φὰς ἔσσευε (1862: 309) for Ω’s φασὶν ἔσευε. Though ἔσσευε (< *e-kyeu-) is the normal epic spelling (restored in bp), the augment could be added to a single sigma as in Il. 5.293 ἐξεσύθη. For σεύω of animals, see B. 18.8–10 ληισταὶ ... σεύοντ’ ἀγέλας βίαι, Hsch. μ.1197 μηλοσόη ὁδός: δι’ ἧς τὰ πρόβατα ἐλαύνεται· Ῥόδιοι. βοῶν ἴφθιμα κάρηνα: Il. 23.260 (containing τ’), Herm. x4. I translate ‘strong and valued cows’, since their ‘staunch heads’ seem to connote value: in Il. 23.260 the formula occurs in a list of prizes; the periphrasis ‘head of X’ in Greek often emphasizes value (see LSJ s.v. κεφαλή I.2, Il. 9.407 ἵππων ξανθὰ κάρηνα in a list of gifts). The best guide to the sense of ἴφθιμος is the formulaic system whereby ‘even’ ἴφθιμοι are staved off, blinded, routed, powerless. This suggests something like ‘physically strong’ or ‘staunch’, and applications to women favour (from a Greek perspective) the latter.150 87–94 The witness at Onchestos: I argue in Introduction §3.3.4 that the human witness was a pre-existing feature of the myth, in the form of Battos, and that he has been replaced here principally to allude to the appearance of Onchestos during Apollo’s journey at H.Ap. 229–38 and to separate out 149

Unlike Ps.-D., Hermes puts the proverb into a chiastic form: see Introduction n. 63. Similar phrasing could be turned to different ends: at A. Th. 246 ‘though you see, do not see too openly’, the point is to apply self-control in the face of the approaching army; often ‘seeing but not seeing’ applies to incompetence at processing sense-data (A. Ag. 1623, Ps.-A. PV 447–8, S. fr. 923, Plb. 12.24.6, etc.). 150 ‘Even’ ἴφθιμοι: e.g. Il. 4.534, 8.144, 12.410, 16.659, 17.749, Od. 16.89, Hes. Th. 698, Op. 704; also Theognidea 1389. Compare the formula βόας καὶ ἴφια μῆλα, where the similar and similarly obscure adjective probably has traditional referentiality of quality and value (LfgrE s.v.). For applications to women, see |ἰφθίμην ἄλοχον+ (Il. 5.415, 19.116, variants at Od. 12.452, 23.92) and |θυγατέρ’ ἰφθίμηι (10.106, 15.364); also 11.287, 16.332. Ancient lexicographers predominantly linked the word to ἶφι. Athanassakis (1971) suggested an etymology from *ϝῑφίτῑμος, under metrical constraints. His synchronic claim, that the semantic emphasis remained on value rather than strength, is unconvincing.

202

C O M M E N TA RY 8 7 – 9 4 Battos’ aetiological function, which the next episode will relocate beside the Alpheus (124–6, 134). Those conclusions do not, however, determine the Onchestan’s character. This was perhaps inspired by Laertes, especially at Odyssey 24.205– 344. The emphasis in both cases is on the sympathy generated by their advanced age and the toughness of their work (our man is working late and will rise early for more: 90*, 184–7, 206); both are bent over as they dig around their plants; Laertes wears leather protection against thorns (24.228–30), as the Onchestan clears brambles (Herm. 190 βατοδρόπε); the two estates offer the only instances of γουνὸν ἀλωῆς οἰνοπέδοιο| (207; Od. 1.193, 11.193); Laertes’ men have gone to fetch stones for the ἀλωῆς ... ἕρκος (24.224), which may correspond to 188* depending on how one emends that line.151 These parallels, tantalizing as they are, do not seem particularly fruitful for interpreting the Onchestan’s role. I prefer to see the hymn’s one instance of an internal human audience as a negative foil for how the external audience can engage with Hermes, now that he receives cult. The hapless man gets no explanation to help him understand the god, and next time we meet him he will seem both nervous and baffled (208–9). The structural balance with Hermes’ ­unanswered speech to the tortoise (Introduction §3.2) points to the man’s reduced agency, and to Hermes’ lack of effusiveness in addressing him. The god is in the awkward position of having been spotted, but takes this in his stride.152 Rather than sending the Onchestan to sleep, hiding himself, or offering a bribe, he hints that the vineyard is liable to damage. This may be the more pointed since Hermes was later invoked precisely to protect a vineyard, in IG XII.2 476 (Eitrem 1906: 256). His threat to the Onchestan not only involves the thematic confusion of visual and aural signs, but specifically asks the man not to divulge oral traditions about Hermes (as Aphrodite leaves Anchises with an instruction not to retell their story: H.Aphr. 290). By contrast, the hymnist is doing just that in order to grant us narrative access to appreciate Hermes’ θαύματα.153

151

See Shelmerdine (1986), Introduction §1.3. Vergados 376 expresses scepticism. From the old man’s perspective, one can compare the suddenness of Hermes’ appearance and disappearance particularly with the way he helps Odysseus at Od. 10.277–308. 153 Like Battos, other humans in H.Hom. are on the receiving end of divine irritation (H.Dem. 251–83, Hy. 7, etc.), which contributes to the terrifying nature of a traditional epiphany. Sowa (1984) 241, 247 observed that Hermes has a different modus operandi in appearing to the Onchestan, omitting the normal traits of an epiphany such as clarity, fragrance, the human’s outright fear, and an explicit revelation of identity after being in disguise; cf. Vergados (2011a) 82–6. 152

203

C O M M E N TA RY 9 5 – 9 7

95–100 95 πολλὰ δ’ ὄρη σκιόεντα: The poet modifies formulaic ὄρεα σκιόεντα| (70*, Od. 5.279, 7.268, H.Ap. 34), but rather than |οὔρεα δὲ σκιόεντα (cf. Il. 1.157) he admits a contraction to accommodate the emphatic πολλά. ὄρη otherwise occurs in Epos at Hy. 27.4, also in this position with σκιόεντα; for occasional synizesis and contraction of ε(h)α in Epos see GH i.56. The echo of 70 is the first sign that we are entering a transitional passage in the narrative. καὶ αὐλῶνας κελαδεινούς: The stem κελαδ- is applied principally to ­natural phenomena including winds (Il. 23.208) but more often streams (Il. 18.576, 21.16, also river Keladon at Il. 7.133; cf. Sappho fr. 2.5 ὔδωρ... κελάδει). A narrow valley might roar with either. 96 καὶ πεδί’ ἀνθεμόεντα: The collocation is not otherwise extant in Epos, but compare Pi. fr. 107a.4 ἀνθεμόεν πεδίον. The adjective perhaps locates the story in spring (cf. 82). This completes a trio of noun-then-adjective phrases in 95–6, which between them encapsulate the different terrains of Hermes’ cross-country adventure. διήλασε: cf. 75* ἤλαυνε. 97–8 ὀρφναίη δ’ ἐπίκουρος ἐπαύετο δαιμονίη Νύξ | ἡ πλείων: Every word other than ‘night’ is striking. ὀρφναίη represents a modification of its only epic formula, νύκτα δι᾽ ὀρφναίην, while retaining the traditional referentiality not just to difficulty of vision but to its failure (see 578, LfgrE s.v. ὀρφναῖος). δαιμονίη substitutes for the traditional ἀμβροσίη, and so avoids hiatus. The adjective is distant from its only use elsewhere in Epos (the address δαιμόνιε+, ‘What’s got into you?’), but δαίμονες have a semantic connection to the epithets ἀμβροσίη and ἱερή (Stes. fr. 8.4, A. fr. 69.7, E. Ion 85, fr. 114, etc.), and actual nocturnal δαίμονες will presently materialize in 100. The personification ἐπίκουρος lends a new military colour to Hermes’ scouting against Apollo (cf. 65* κατὰ σκοπιήν, 290* νυκτὸς ἑταῖρε). Expressions comparable to νὺξ παύεται are remarkably rare (e.g. Galen Di. Dec. 1.8 p. 808 νυκτὸς ἤδη παυομένης, Longus 2.25 παυομένης ἡμέρας); perhaps the verb’s presence is justified by ἐπίκουρος, since παύομαι often refers to leaving off battle (LfgrE iii.1075.36). Then ἡ πλείων pulls back from the personification of ἐπίκουρος ἐπαύετο. The only close parallel is itself difficult – Il. 10.251–3 μάλα γὰρ νὺξ ἄνεται, ἐγγύθι δ᾽ ἠώς, | ἄστρα δὲ δὴ προβέβηκε, παροίχωκεν δὲ πλέων νύξ | τῶν δύο μοιράων, τριτάτη δ᾽ ἔτι μοῖρα λέλειπται, where articular τῶν and genitive δύο are both anomalous.154 Arithmetic favours the translation 154

Zenodotus neglected and Aristophanes of Byzantium athetized 253 (ΣA), but then παροίχωκεν δὲ πλέων νύξ repeats μάλα γὰρ νὺξ ἄνεται.

204

C O M M E N TA RY 9 8 – 9 9 ‘the majority of the night, consisting of two portions’ over ‘more of the night than two portions’ (which would be less of a parallel to Herm.). Allusion is plausible: Herm. has recently (81*) suggested its interest in the Doloneia as a nocturnal adventure, and the military colour of ἐπίκουρος supports this. 98 τάχα δ’ ὄρθρος ἐγίνετο δημιοεργός: τάχα ‘presently’ marks a temporal ellipse of vague length. ὄρθρος normally indicates a period before the yellows and pinks of dawn (cf. 184*); in X. HG 4.5.18 Agesilaus leaves Orchomenos in ὄρθρος and passes Mantineia, 10 miles away, ἔτι σκοταῖος. ὄρθρος is often cited as the early limit of when people get up, and thus for a thief it is when the risk of being spotted increases; herders such as Apollo were notorious early risers.155 This pre-dawn ‘does a public service’. This anthropomorphism balances that of ἐπίκουρος (97–8*); the public service could be understood as providing first light, contrasting with Hermes’ ὀρφναίη ally. Alternatively, the pre-dawn heralds the day, and heralds are δημιοεργοί at Od. 19.135. Casting the pre-dawn as one of the other artisans at Od. 17.383–5, as a senior magistrate, or as a demiurge, seems harder.156 Nor can δημιοεργός mean τὸν δῆμον ἐργάζεσθαι ποιῶν, despite the similar thought at e.g. Hes. Op. 580, and the claim at Σrec. PV 5. ἐγίνετο: Such spellings were standard in fifth-century Ionic, and Homeric manuscripts give little evidence that rhapsodic traditions preferred γιγν-. See Thumb and Scherer (1959) 263, GH i.12–13. 99 ἣ δὲ νέον σκοπιὴν προσεβήσετο δῖα Σελήνη: The implication, allowing for the vagueness of νέον, is that the moon rose or culminated in the early hours. This places us much later in the month than the fourth (19*), on which the moon sets only a few hours after the sun (see also 141*). The inconsistency, noted by Schneidewin (1848: 667), was taken by Lohsee (1872: 31–2) as poetic licence in balancing Hermes’ traditional birth-date with a desire to mention the moon at this point. That desire seems motivated by the possibility of tracking critical stages of Hermes’ journey with reference to the movements of the sun and moon (see 68–9*; σκοπιή also recalls 65 from those initial stages of

155

Cf. Hes. Op. 577 of a diligent farmer at harvest-time, Aesop 97 of shepherds, Pherecrates fr. 10 of women grinding the day’s flour, Ar. V. 216 of keen jurors, Av. 489 of select tradesmen (still νύκτωρ, as at Eccl. 741). For the expectation that few people are awake see Thgn. 863 (of a prostitute slinking home), Th. 3.112.3, Pl. Prt. 310a–b; it is frequently mentioned as a time when anxiety causes insomnia. For a connection to the dawn chorus see e.g. Ibycus PMG 303, Thgn. 863. By contrast A. Ag. 254 σύνορθρον αὐγαῖς loosely equates ὄρθρος and sunrise. 156 Ruhnken (1781) 32 cited Hsch. δ.848 δημιουργός: ὁ ἥλιος here. But for the sun as demiurge see rather Thphr. Ign. 5, Corp. Herm. 16.5, Euseb. Pr. 3.4.2.

205

C O M M E N TA RY 1 0 0 Hermes’ journey), and of redeploying the traditional role of Battos’ σκοπιά in the myth (see below, 95–100*), using the equally traditional metaphor of a high celestial position as lookout.157 The military colour of ἐπίκουρος returns, leaving the moon’s affiliations in the balance: as a form of light, the moon is liable to reveal Hermes’ theft, but as a power associated with his ‘ally’ Night, she may be keeping a lookout for his benefit. Radermacher 207 suggested that the moon has returned to the sky out of favouritism for thieves, but one might have wished for more than προσεβήσετο to describe this.158 For -βήσετο see Leumann (1953). Aristarchus found and preferred it in the Iliad (ΣT Il. 3.262, etc.), so that it is attested widely in Homeric manuscripts, whereas in parts of Epos which received less philological scrutiny, the paradosis generally offers -βήσατο (as here, 233, H.Ap. 141, Sc. 33, 338; but ἐβήσετο at H.Ap. 49). For δῖα Σελήνη| cf. Hy. 32.8, in a more detailed description of the (full) moon ascending. Though δῖα may etymologically mean ‘heavenly’, in synchronic terms the epithet adds traditional epic colour more than semantic detail; LfgrE suggests ‘impressive’ as the most applicable of the ancient glosses. 100 Πάλλαντος θυγάτηρ Μεγαμηδείοιο ἄνακτος: The Titan Pallas is otherwise the cousin of Selene, Helios and Eos, and brother of Astraeus (Hes. Th. 371–7), or the father of Aurora in Ovid (e.g. Met. 9.421). Our line offers a further variation on his genealogical position: we can no longer assess its familiarity or effect on ancient audiences. Pallas’ father is normally called Kreios, as in Hesiod. Megamedes is not known elsewhere, and editors have long taken that as a problem. A more positive approach is to interpret it as a speaking name, ‘Of great schemes’: Medea and Perimede are exemplary sorcerers by moonlight (Theoc. 2.16) – an occupation which perhaps explains the connection of Medea and Agamede (cf. Il. 11.740–1 for her ‘cleverness’ with drugs) to the family of the celestial bodies. Moreover, taking Megamedes to embody the connection of nocturnal agency through cunning to sorcery could be particularly relevant just before Hermes demonstrates cleverness with a nocturnal deformation of sacrificial rituals (albeit not a very witch-like one). Μεγαμηδείοιο: I tentatively prefer a’s reading, against previous editors. One can infer -δείδοιο b and so -δείδαο ss. οιο x, somehow reflecting inherited Compare Od. 8.302 Ἠέλιος γάρ οἱ σκοπιὴν ἔχεν, H.Dem. 62 Helios as the θεῶν σκοπὸν ἠδὲ καὶ ἀνδρῶν. These parallels tell against Shelmerdine’s idea (1981: 132) that σκοπιή here is a mountain peak and that προσεβήσετο describes the moon descending to touch it. Moreover the verb is used of ascents (LSJ s.v. 3), not descents. 158 A more ambitious approach would be that the narrator, like Medea at A.R. 3.533, has altered the moon’s natural course to help his hero with an unorthodox nocturnal ‘sacrifice’; see on 100* for a possible hint in this direction. 157

206

C O M M E N TA RY 1 0 1 – 1 0 2 variants -δεΐδαο and -δείοιο. The former is normal in Ionic, and so the more likely gloss; for patronymic adjectives in Aeolic see e.g. Thumb and Scherer (1959) 4. However, it is possible that Ω had Μεγαμηδεΐδαο alone, and that e.g. a blotch in Θ caused the other readings. 95–100 Rewriting the night: Lines 97–100 wilfully rewrite traditional diction at every turn – more so than is normal even by Herm.’s standards: see on ὀρφναίη, ἐπαύετο, δαιμονίη and δημιοεργός, and probably Selene’s unique genealogy (100*) and the allusion to a crux based on unique phrasing in Iliad 10 (97–8*). The poet truly creates a δαιμονίη night, which continues the innovation of 87–94, where Battos and his lookout were rewritten as the old man of Onchestus; in fact, as Hermes reaches the Alpheus it is the moon, not Battos, who has adopted a lookout-post, and even done so ‘newly’ (99). The phrasing thereby symbolizes, in two ways, the hymnist’s claim to novelty in this part of the myth.

101–7 101 ἐπ’ Ἀλφειὸν ποταμόν: Mention of the Alpheus, which may have been expected from some previous versions of the myth, aligns Hermes with Heracles and in my view brings the action into the world of the original performance-context (Introduction §2.1). Versions of the myth mention ­various destinations for the cows. In the later extant texts containing Battos, his lookout is between Mt Lykaon and Koryphasion (Ant. Lib. 23), or in Neleus’ territory (Ovid Met. 2.684–9); both entail the destination ‘Pylos’ (for which see 216*), and routes to Triphylian or Messenian Pylos from northern Greece can naturally include crossing the Alpheus. By contrast, media such as drama and painting favour unity of space: the Brygos Painter has Hermes hide the cows in Maia’s cave (references on 21–2*), a Caeretan hydria (LIMC s.v. Hermes 241) in a separate cave next to Maia’s, and Sophocles’ Ichneutae there or near by; Philostratus Im. 1.26 moves the whole action near Olympus. 101–2 Διὸς ἄλκιμος υἱός | Φοίβου Ἀπόλλωνος βοῦς ἤλασεν εὐρυμετώπους: The hymnist avoided several epic formulas for Hermes in favour of a Hesiodic one for Heracles.159 And although βοῦς εὐρυμετώπους appears in various combinations, βοῦς ἤλασεν εὐρυμετώπους is only extant elsewhere

159

Hes. frr. 35.5, 43(a).61, Sc. 320; borrowed by Pi. O. 10.44–5, Ps.-Theoc. 25.42. Contrast Διὸς ἀγλαὸς υἱός (432), Κυλλήνιος Ἑρμῆς, κρατὺς Ἀργεϊφόντης (Van Nortwick 1975: 34).

207

C O M M E N TA RY 1 0 3 – 1 0 4 at Hes. Th. 291, of Heracles driving Geryon’s cattle (Shelmerdine 1981: 13–14). The alignment draws attention to Hermes’ unusual (Heraclean) physical achievement and pugnacity in moving the cattle. As often, juxtaposed formulas for him and Apollo square up to each other (see e.g. 214–15, 235–6, 318, 365–6). But the phrasing also points light-heartedly to the distance between Hermes’ actions and Heracles’ modus operandi in rustling the cattle of Geryon or Augeas; the latter story is the more important, since Heracles brought Augeas’ cows to the Alpheus, to the site of Olympia (Pi. O. 10.34–45). 103 ἀδμῆτες: The paradosis seems to modify Il. 10.292–3 ~ Od. 3.382–3 βοῦν ἤνιν εὐρυμέτωπον | ἀδμήτην, where the adjectives emphasize that the cows are ideal for sacrifice (so too ἀδμής in Babrius 1.37); in Antoninus, Hermes steals a different range of animals but the cows are similarly ἄζυγες (Ant. Lib. 23.3). Here, then, ἀδμῆτες invites one to apply a sacrificial schema to the following episode. Moreover, the cows’ lack of subjugation by a bull or a yoke sets up the initiation into sex and money-making which Hermes will bring to the forty-eight who do not experience an initiation into mortality (493–5*).160 103–4 ἵκανον ἐς αὔλιον ὑψιμέλαθρον | καὶ ληνούς: The echo of 70–1 ἀφίκανε ... αὖλιν frames the cows’ journey and marks the transition to a new ­episode (cf. on the λειμών of 104*, and ἐριμύκους in 105*). αὔλιον, unlike αὖλις (71*), distinctly implies a cave or man-made structure rather than a mere ­sleeping-place. After it, ὑψιμέλαθρον (literally ‘with a high roof-beam’) ­suggests grandeur and artificial construction, like Augeas’ huge αὔλια by the Alpheus in Ps.-Theocritus 25. I disagree with the normal view that the αὔλιον here does turn out to be a cave at 401*. Rather, having Hermes hide the cows in a man-made steading is a significant departure from other versions of the story, and marks Hermes’ removal of them from a pristine natural habitat to one shaped by human pastoralism (see 491–4*). The presence of ­watering-troughs and 399* ἀγρούς reinforces this shift, which comes at the expense of raising some potentially distracting questions about who the human herders are and why they are absent.161 ὑψιμέλαθρον can still make this destination a counterpart of Maia’s ὑψηρεφής cave (23), where Hermes took the cows in some versions (101–2*). Ilgen’s neat ἀκμῆτες would produce a different point, that the gods’ cows have extraordinary stamina. 161 LSJ cite LXX Genesis 30.38–41 for ληνός in the sense ‘watering-trough’, identified here by Spanheim (1697) 246. Lines 106–7 tell against the interpretation ‘feeding-troughs’. 160

208

C O M M E N TA RY 1 0 4 – 1 0 7 104 προπάροιθεν ἀριπρεπέος λειμῶνος: The original visual focus of ἀριπρεπής would be distracting in a nocturnal scene (pace LfgrE; compare e.g. Il. 9.441): the meadow is more generally ‘outstanding’, and balances the lovely Pierian meadows of 70–2. Conceivably, then, the lack of an epithet to balance 72 ἀκηρασίους contributes to this meadow’s traces of human activity (103–4*). 72* in turn recalled the tortoise grazing outside Maia’s cave in 27, and the chain of pasturing locations is reinforced by the repetition of προπάροιθε from there. 105–8 Preliminary comment: The temporal clause could end either at 106 καί or at 108 πυρὸς δ’. I prefer the former alternative, since the latter requires three subordinate verbs with mixed tenses. For apodotic καί and δέ see Denniston 308, 179. 105 ἔνθ’ ἐπεὶ εὖ βοτάνης ἐπεφόρβει βοῦς ἐριμύκους: ἐριμύκους is another echo of 69–74; as in 74* it toys with our anxiety that Hermes may be exposed by the cattle’s lowing (Fernández Delgado 1998: 8). However, εὖ underlines Hermes’ beneficent abilities as a herder. For the prominence brought to this by the patterning of consonants in 104–5 see Introduction §3.2. The derivation of ἐπεφόρβει is not clear, though the sense is. It could be pluperfect from φέρβω, whose other instances are limited to the present stem; or one could imagine an imperfect of ἐπιφορβέω ‘supervise X feeding on Y’, though neither this nor ἐπιφορβός, nor any prepositional compound of φορβός, is attested elsewhere. 106 συνέλασσεν ἐς αὔλιον ἁθρόας οὔσας: The line extends Hermes’ herding credentials slightly, from driving the animals and feeding them, to collecting them together. The ending -ᾰς of ἁθρόας and the Attic participle οὔσας arise here from modification of a formula, probably ἁθρόοι ἦσαν| (Od. 1.27). See Janko (1982) 144–5. Morpurgo Davies (1964) 152–65 discusses the origin of forms in -ᾰς. For the breathing see LfgrE s.v. ἁθρόος E. 107 λωτὸν ἐρεπτομένας ἠδ’ ἑρσήεντα κύπειρον: λωτός in this context is an untechnical name for a range of plants found in pastureland (cf. Thphr. HP 7.15.3); for |λωτὸν ἐρεπτόμενοι see Il. 2.776 (of horses), Od. 9.97 (of Lotophagoi, secondarily). κύπειρος/-ον probably covers at least Cyperus longus (galingale) and C. rotundus (nut-sedge).162 For the combination see Il. 21.351 καίετο δὲ λωτός τε ἰδὲ θρύον ἠδὲ κύπειρον, Od. 4.603 ἔνι μὲν λωτὸς πολύς, ἐν

162

Cyperus capitatus is also currently found in Elis (Koumpli-Sovantzi 1991), but prefers sandy terrain.

209

C O M M E N TA RY 1 0 8 – 1 0 9 δὲ κύπειρον (of Sparta), Pherecrates fr. 114 πατοῦντες ἐν λειμῶνι λωτοφόρωι κύπειρόν τε δροσώδη ... (in the Underworld?). Greeks often emphasized the moist habitats of κύπειρος/-ον, as in ἑρσήεντα and Pherecrates’ δροσώδης. This suits a meadow (104) near the Alpheus (101) at night. The tense of ἐρεπτομένας implies that the cows continue to feed as Hermes attempts to herd them into their lodgings – a picturesque touch to enhance his achievement; Franke commented ‘haec boum dum in stabulum aguntur lotum et cyperum carpentium descriptio videtur ex ipsa rerum natura deprompta’ (p. 66).

108–11 108 σὺν δ’ ἐφόρει ξύλα πολλά. πυρὸς δ’ ἐπεμαίετο τέχνην: The first half of the line balances 106 τὰς μὲν συνέλασσε, but this second act of gathering is then flagged as having extra significance as a further sign of Hermes’ invention and technical skill. ἐπεμαίετο is carefully designed: Hermes ‘reaches for’ a craft by actually ‘grasping’ a fire-stick (109–10).163 Compare 511* σοφίης ἐκμάσσατο τέχνην. Bion fr. 1.2, where Apollo attempts to heal Hyakinthos’ wounds, ends σοφὰν δ᾽ ἐπεμαίετο [Crispinus: ἐπεβώσετο or ἐπεβένετο mss.] τέχναν. If Crispinus’ emendation is right, Bion may have been inspired by Herm. 108 + 511. 109–10 δάφνης ἀγλαὸν ὄζον ἑλὼν ἐπέλεψε σιδήρωι < > ἄρμενον ἐν παλάμηι: This phrase describes Hermes inventing a fire-technology (108) involving a laurel stick. Laurel is recommended as a hard wood rubbed to create sparks at Thphr. HP 5.9.6–7. It may be ἀγλαός (see 60–1*) because of its glossy leaves or because it is valued by Apollo: compare, respectively, E. Ion 79–80 ὡς πρὸ ναοῦ λαμπρὰ θῆι πυλώματα | δάφνης κλάδοισιν and Pi. fr. 94b.7–8 ὅρπακ’ ἀγλαὸν δάφνης, Theopomp. Com. fr. 48 ἀγαλματίοις ἀγαλοῦμεν ἀεὶ καὶ δάφνηι (in a prayer to Apollo), Macedonicus 5–6 ἱκτῆρ[α] κλάδον ἐν παλάμηι θέτε καλὸν ἐλαΐνεον κ[αὶ δάφνης?] | ἀ[γλ]αὸν ἔρνος (see Peek 1980: 45, pl. 4). The latter sense enhances the antagonism: Hermes uses Apollo’s sacred plant against him. Finally, the semantic development of παλάμη from ‘palm’ to ‘plot’ implies Hermes’ ingenuity, and we will see fire and Hermes’ παλάμαι recurring in a different combination below at 357.

Elsewhere in Epos, ἐπιμαίομαι takes an accusative when physical contact is implied, and a genitive for conceptual seeking-out. The rarer accusative could therefore draw attention to the wordplay. However, cf. A.R. 3.816 νόωι ἐπεμαίεθ’ ἕκαστα, Arat. 89 ἐπιμαίεο Χηλάς, and for the simplex Od. 13.367 μαιομένη κευθμῶνας ἀνὰ σπέος.

163

210

C O M M E N TA RY 1 0 9 – 1 1 0 At the end of 109, Ω had σιδήρω; before that, M and Ψ offer disconcertingly different verbs, both hapaxes – ἐνίαλλε and ἐπέλεψε. In my view, ἐπέλεψε (with a concomitant lacuna) allows one to explain how ἐνίαλλε arose, but not vice versa. (i) M’s ἐνίαλλε would mean ‘sent forwards quickly within’. But σιδήρωι must then be emended to a softer wood than laurel: Ludwich (1905) 9 found σιδείωι sc. ὄζωι, ‘a bough of pomegranate’, a word only extant in lexicographers (first Hdn. Gr. ii.442, κνίδειος καρπὸς καὶ σίδειος κλάδος) and so liable to corruption. This reading and ἄρμενον ἐν παλάμηι are suitable for a fire-plough, where a stick is rubbed to and fro in a groove to create a spark; but fire-ploughs are not explicitly attested for ancient Greece, and we have not explained Ψ’s ἐπέλεψε, which is too rare for a scribal emendation, and more than a corruption away from M’s reading.164 (ii) Conversely, ἐπέλεψε (‘stripped’, not ‘sharpened’; cf. ἐπιτέμνω ‘cut on the surface’) stands well with σιδήρωι, but the passage then fails to mention the second piece of wood required. We must therefore posit a lacuna: I cannot agree with Zanetto (1996) in accepting a narrative ellipsis of the second piece of wood and the action by which fire is generated in an aetiology for fire-sticks which extends the theme of Hermes’ characteristic imaginative juxtaposition of apparently inert objects. Conceivably the missing line contained a verb like ἐνίαλλε, was omitted, was re-added in the margin, and survived only as a single-word variant.165 Straying further from both ἐπέλεψε and ἐνίαλλε replaces one difficulty with two, and nothing seems to be gained by other alterations such as ἀρμένωι for ἄρμενον. What fire-technology does solution (ii) involve? Because of the uncertainty of what was in the lacuna, the better-attested fire-drill could enter the picture. The hand-operated fire-drill, where the laurel was placed loosely between both palms, while the other block of wood rested on the ground, would require ἐν παλάμηι and barely suits ἄρμενον.166 A bow-operated drill is more likely: such a device was already suggested at 41–2*, and adds to the numerous parallels between this part of the hymn and 20–62 (Introduction

164

For Greek fire-making see Kühn (1859) 36–7, Morgan (1890) 19–31; Cholmeley (1922) raised the possibility of understanding a fire-plough here. Of the main ancient sources, Thphr. HP 5.9 and Σ A.R. 1.1184 specify rotation (i.e. a fire-drill), whereas Thphr. Ign. and Sen. QN 2.22 speak more vaguely of friction; Seneca attributes knowledge of the technique to herdsmen in particular. 165 See Shackle (1915). If the lacuna ended κλάδωι δ᾽ ἐνίαλλε [or similar] σιδείωι, one could explain the lexicographers’ knowledge of σίδειος and the omission (by homoeoteleuton), but the truth need not be so convenient. 166 Il. 18.600, where a potter tests his wheel ἄρμενον ἐν παλάμηισιν, suggests a firmer grip than can be used in a hand-operated fire-drill, as does Od. 5.234, of an axe.

211

C O M M E N TA RY 1 1 0 – 1 1 1 §3.2). ἄρμενον ἐν παλάμηι then means that the laurel-stick was a suitable diameter for Hermes to press on its upper end with his left hand. 110 ἄμπνυτο δὲ θερμὸς ἀϋτμή: Fire-sticks proverbially required much physical effort (e.g. X. Cyr. 2.2.15), whereas the impression from the narrative pacing is that Hermes manages easily (cf. 113* for a similar effect). Both ἀναπνέω/-οή and ἀϋτμή (~ ἄημι) regularly concern the emission of hot air accompanying fire; similarly 114 φῦσαν. It may be anachronistic to try to explain the wording using Theophrastus’ physics, but he similarly notes that ivy makes the best base for a fire-drill because τάχιστα καὶ πλεῖστον ἀναπνεῖ (HP 5.9.6); he seems to have thought that in fire-sticks the friction both concentrates heat and creates wood-dust, that the heat vaporizes moisture in the wood-dust, and that the hot vapour (as in ἀϋτμή) exits from pores, which are more numerous in wood of an open texture (Ign. 17, 28–9, 69 with Coutant 1971: xii–xx versus Fortenbaugh and Gutas 1992: 262). For aorist ἄμπνῡτο see GH i.382. θερμὸς ἀϋτμή| also occurs at Hes. Th. 696 and is cited as an apparent breach of concord at e.g. ΣA Il. 18.222b, which probably goes back beyond Aristonicus to Aristarchus; Lesbonax Fig. 2 explicitly places θερμὸς ἀϋτμή in a series of Homeric examples before and (apparently) separate from a final parallel from Hesiod (ὡς τὸ παρ’ Ὁμήρωι “κλυτὸς Ἱπποδάμεια” καὶ “θερμὸς ἀϋτμή” καὶ “ἁλὸς πολιοῖο” καὶ παρ’ Ἡσιόδωι “δαϊζομένοιο πόληος”). This might be a rare sign of an ancient scholar using Herm., but I do not place much trust in Lesbonax’ precision.167 111 Ἑρμῆι τοι πρώτιστα πυρήϊα πῦρ ἀνέδωκεν: Ω’s Ἑρμῆς ... πυρήϊα πῦρ τ᾽ ἀνέδωκε would mean that Hermes ‘distributed’ fire and fire-sticks. With Ludwich’s emendation, πῦρ is the object of ἀναδίδωμι ‘give off’: compare e.g. Th. 3.88 τὴν νύκτα φαίνεται [Stromboli] πῦρ ἀναδιδοῦσα πολύ, Posidonius fr. 46 πέτρα δ᾽ ἐστὶ πῦρ ἀναδιδοῦσα, Greg. Nyss. Eunom. 2.1.118 πῶς τριβόμενα πρὸς ἑαυτὰ τὰ ξύλα πῦρ ἀναδίδωσι;. The echo of 25 contributes to the parallelism between the tortoise-episode and the cattle-rustling: see Introduction §3.2. 108–11 Inventing fire-sticks: The fire-drill (or -plough?) is a second instance alongside the sandals of Hermes combining types of wood to create a familiar part of culture which babies do not normally use. The invention is also related to that of the lyre (111*), where again inert (even dead) materials See Blank (1988) 178, Pontani (2010) on Σ Od. 4.442c. Scaliger’s ἀϋτμήν (1573: 389– 90) does not occur in the nominative, and is unnecessary given the Hesiodic parallel and Od. 12.369 ἡδὺς ἀϋτμή. Quintus varied ἄμπνυτο... ἀϋτμή at 10.62 ἄμπνυεν ἀϋτμήν, and borrowed Herm.’s other use of ἀϋτμή (137) at 8.90.

167

212

C O M M E N TA RY 1 1 2 produced something lively. As will become clearer through the sacrifice-like qualities of the following events, Hermes’ action contrasts significantly with the myth of Prometheus offering humans fire-craft by theft rather than by invention (112–41*; D.S. 5.67.2 specifies Prometheus’ gift as πυρεῖα), and with ways in which new fire was obtained for sacrifices in Greek cults (Furley 1981: 39–40). A further contrast was suggested by Detienne (1997) 26 n. 118: Hermes invents a typical human technique for starting fire, whereas Apollo at H.Ap. 444 starts the Delphic flame through sparks flying off him at an epiphany; this would contribute to Herm.’s broader engagement with the Hymn to Apollo (Introduction §1.3).

112–41 112 πολλὰ δὲ κάγκανα κᾶλα: κάγκανος is restricted to wood for burning in Epos (x4). κᾶλον (‘log’) refers to firewood also at Ion TrGF 19 F 29, Call. Hec. fr. 31; Dion. Epop. Bassarica(?) fr. 41 κάγκανα κῆλα recalls our phrase, modified probably because of intervening speculation about the connection of κᾶλον and κῆλον.168 These poets perhaps felt a connection to καίω, fostered here by the four instances of κα-, although other uses are incompatible with this (Hes. Op. 427; the Doric sense ‘ships’ timber’ at Ar. Lys. 1253, X. HG 1.1.23). Compared to 108 ξύλα, the lemma suggests Hermes may be using wood which has been specifically stored and dried for firewood, which would be another sign that the site sees occasional human visitors (103–4*). κατουδαίωι ἐνὶ βόθρωι: It is left unclear whether Hermes finds the pit or digs it. From 116, Hermes’ actions will start evoking sacrifice at every turn, as discussed in Introduction §2.1. We have already had some cues in this direction: Hermes is intending to eat beef (64), an activity almost always preceded directly by sacrifice in Greece; 103 ἀδμῆτες* has suggested a sacrificial schema. This raises the question whether a fire in a βόθρος had particular ritual connotations, and if so how pertinent they are here. βόθρος is used sporadically for temporary pits in esoteric chthonian rituals (most famously in Od. 11), and from Pausanias on for pits in some heroa such as that of Pelops at Olympia (Paus. 5.13.2). But these pits are predominantly used for libations, and not as far as we can tell for building a fire (Ekroth 2002: 60–74). Extant pits in sanctuaries (to a wide range of divinities) do not tell us about terminology, and only a few are known to have contained

See Livrea (1973) 59–60. Benaissa (2018: 178–9) suggests that Dionysius’ κακοδηνέϊ (B. fr. 33r.41) could relate to Herm. 389 κακομηδέα.

168

213

C O M M E N TA RY 1 1 3 – 1 1 6 sacrificial fires. I conclude that Hermes’ fire-pit had few if any ritual connotations. By contrast, the pit has practical utility: Hermes is small, and needs to manage a large fire and then hide its ashes; fire-pits protect against the wind. For other classical non-sacrificial fire-pits lit with such considerations in mind, see A.(?) fr. 89, X. Anab. 4.5.6, Hp. Nat. Mul. 109, Mul. 133. 169

113 οὖλα λαβὼν ἐπέθηκεν: The core of οὖλος is ‘dense and curly’, and the word is frequently applied to wood with irregular grain in Theophrastus.170 This sense seems acceptable here even if Theophrastus himself (Ign. 32, 36, 72, 74 – the most explicit Greek viewpoint we have on the matter) declares that dense wood which is difficult to split, i.e. οὖλος wood, gives an acrid smoke, which is not ideal for Hermes’ purposes. Again the narrative pacing suggests Hermes’ haste and divine nonchalance: real fires need to be built slowly and carefully with plenty of ventilation before ‘many logs’, let alone dense ones, can be packed on. ἐπηετανά: For the synizesis cf. Hes. Op. 607, Pi. N. 6.10. 114 φῦσαν: See 110* for the conception of fire exhaling. Ω’s φύζαν was defended by Janko (1982: 146), but there is no evidence for a meaning other than ‘flight’. 115 βίη κλυτοῦ Ἡφαίστοιο: Since πῦρ is the object, ‘Hephaestus’ force’ is the power of combustion, rather than a metonym for fire itself as normal (LfgrE ii.950.33–47). The phrasing suggests that despite Hermes’ current isolation and attack on Apollo’s powers, his technology (fire-sticks) can also collaborate with the power of another god; this contrasts with Prometheus’ actions when he stole fire (Introduction §1.3). 116–17 τόφρα δ’ ὑποβροχίας ἕλικας βοῦς εἷλκε θύραζε | δοιὰς ἄγχι πυρός· δύναμις δέ οἱ ἔπλετο πολλή: Compare Il. 20.403–5 ὡς ὅτε ταῦρος | ἤρυγεν ἑλκόμενος Ἑλικώνιον ἀμφὶ ἄνακτα | κούρων ἑλκόντων, γάνυται δέ τε 169

See Patera (2012) 207–23 on pits in some Thesmophoria of Magna Graecia, Ekroth (2002) 80 on the Imperial enagisterion for Palaimon at the Isthmus; Riethmüller (1999) 135–8 used ‘bothros’ as if it was a technical term, and thought they were for ‘chthonian or heroic sacrifices’ (137) despite citing the Pythion and sanctuary of Athena Poliouchos at Gortyn. 170 οὐλότης is correlated with density and hardness (HP 3.11.1–4, 4.1.4, 5.3.3, etc., with the notes in Amigues 1988–2006), and occurs where nutriment becomes εἰλουμένη (CP 6.11.8); at HP 5.2.3 it contrasts with burrs; at 5.5.1 unworkable kinds of wood have οὖλαι συστροφαί. Vergados and Schenck zu Schweinsberg took οὖλα here = ὅλα. This is banal and unsuited to the usage of κῆλον.

214

C O M M E N TA RY 1 1 6 – 1 1 7 τοῖς ἐνοσίχθων, and in particular A.R. 1.405–8, which is probably modelled on Herm. with careful variation of each detail (see Introduction §4.2): φιτροὺς ἀζαλέης στόρεσαν καθύπερθεν ἐλαίης. τείως δ’ αὖτ’ ἀγέληθεν ἐπιπροέηκαν ἄγοντες βουκόλοι Αἰσονίδαο δύω βόε· τοὺς δ’ ἐρύσαντο κουρότεροι ἑτάρων βωμοῦ σχεδόν … Cf. κάγκανα κᾶλα ... ἐπέθηκε... τόφρα δ’... βοῦς εἷλκε... δοίας ἄγχι πυρός, and the idea of animals being selected from a larger herd. Both parallels contribute to the point that Hermes is behaving like a κοῦρος at the acme of his strength; he even takes on two cows simultaneously, which goes beyond the demands of hunger (even Heracles is sated with one in Hes. fr. 265, Pi. fr. 168) or the amount of meat needed for what follows. The parallels also show the relationship of Hermes’ action to human preparations for sacrifice. For the possibility that the two cows refer to a specific ritual, see Introduction §2.1. For cultic manhandling of male cows see also Thphr. Char. 27.5 with Diggle (2004), Burkert (1984) 837 on Paus. 1.19.1, 8.19.2. ὑποβροχίας: This is the natural way to drag cows. Besides abundant artistic evidence, there are tie-rings in front of the altar at Claros (Delattre 1992: 22), Peleus compares Andromache’s binding to that of a cow (E. Andr. 720 βοῦν ... ἤλπιζες ἐντείνειν βρόχοις;), and cf. Il. 13.571–2 ἤσπαιρ’ ὡς ὅτε βοῦς τόν τ’ οὔρεσι βουκόλοι ἄνδρες | ἰλλάσιν οὐκ ἐθέλοντα βίηι δήσαντες ἄγουσιν. Naiden (2007) has shown that it was not normally a bad omen if a victim had to be pulled to its slaughter. Although ὑποβρόχιος is not found elsewhere, ὑπόβροχος is (a ‘bound’ letter: BGU ii 531 i.4; Arsinoite, c.80 ce). The rope appears rather suddenly, but so do the γλύφανος (41), cow-hide (49), βόθρος (112), whatever tool Hermes uses in 119, and so on. Ω had ὑποβρυχίας, but ὑποβρχιος and ὑπόβρῠχα mean ‘underwater’ (the former in this metrical position at Hy. 33.12), and even the later extension to ‘under the surface’ is hardly suitable for describing the cows being indoors. Meanwhile ὑποβρχιος is not otherwise found, the synizesis -χίας has no adequate parallel in Epos (see GH i.170), and the sense of ὑπο- (‘quietly’) is unconvincing.171 The best way to preserve βρῡχ- here is ὑπὸ βρυχαῖς, ‘to the sound of bellowing’ (Shackle 1915). ἕλικας βοῦς: The poet seems to have taken this formula to describe the cows’ horns (cf. 192 κεράεσσιν ἑλικτάς), as most scholars do. LfgrE ii.550 lists

The last point tells against Ludwich’s ὑποβρύχους (1890). ὑποβρῡχάομαι later means ‘wheeze’ (Greg. Nyss. ix.114, Adamantius Phys. 2.41). At PV 1082 βρῠχία δ’ ἠχὼ παραμυκᾶται, there may be a wordplay on βρῡχ-, but ‘from the depths’ gives better sense, and an allusion to water suits the preceding metaphor χθὼν σεσάλευται.

171

215

C O M M E N TA RY 1 1 8 – 1 1 9 other ancient interpretations. Before selective breeding, all cattle had horns, and Greek art from the sixth century on depicts four types: short or helical, and pointing upwards or forwards.172 In reality, if not in art, short-horned cows were probably more common (Richter 1968: 52–3). The adjective here, in assonance with εἷλκε and to an extent with ὑπο-, may have been performed so as to suggest Hermes’ puffing (Fernández Delgado 1998: 110). 118 ἀμφοτέρους δ’ ἐπὶ νῶτα χαμαὶ βάλε φυσιοώσας: Hermes effectively wrestles the cows into helplessness. For βάλλω as a non-technical wrestling metaphor compare e.g. Hdt. 6.27 ἐς γόνυ τὴν πόλιν ἔβαλε. Here φυσιοώσας suits the idea of wrestling, and brings to the fore the cows’ resistance: compare Il. 13.571–2 (with ἤσπαιρε), quoted on 116–17*. Hermes’ physical strength is a common quality of gods in hymns (Keyssner 1932: 48–54), but in Herm. it is a new point of focus. Wrestling by the Alpheus evokes Olympia, and Hermes’ palaestral role there and elsewhere (see Introduction §2.2). That he is wrestling with cows perhaps builds on the allusions in 101–2* to evoke Heracles’ wrestling with the Cretan bull and with Achelous in bull form. Yet the poet’s careful vagueness in describing the action allows these hints of mighty Olympic wrestlers and Heracles to coexist humorously with a quite different mental image: Hermes is simply too small to kill a cow by the normal method of stunning its neck from above, so getting the cows to the ground is a practical necessity for the tiny trickster. In Hdt. 4.60, the Scythians’ method of tripping their sacrificial cows (καταβάλλει) by a rope has nothing to do with their height, but it does effect a similarly marked opposition to Greek sacrificial procedures. 119 ἐγκλίνων: This reading of Ψ can probably stand as a rare intransitive use, ‘leaning on them’; compare LSJ s.v. κλίνω IV. Transitive interpretations are less satisfactory semantically: ‘making them lie down’ ignores ἐν-, while ‘making them lean aside’ is pleonastic with ἐκύλινδε. M’s ἐκκρίνας gives the wrong sense. Schneidewin conjectured ἀγκλίνας, ‘after making them lie on their back’ (1848: 669). ἐκύλινδε δι᾽ αἰῶνας τετορήσας: For the form τετορήσας see 41–2*. Hermes first bores through the cows’ αἰών, then sends them rolling like the corpses at Il. 11.147, H.Ap. 359.173 The first action clearly recalls 42–3*, where he had the tortoise on its back and chiselled through the plastron to bore out the αἰών. As in 42, αἰών is not located specifically. This time, the theme of Hermes’ drilling 172

See e.g. Richter (1930) plates 9, 28–34; Pfuhl (1923) index 12 s.v. ‘Kuh’, ‘Rind’, ‘Stier’; Morin-Jean (1911) s.v. ‘taureau’; LIMC s.vv. Europe 1, Herakles IV H, Io 1, Theseus VI. 173 The poet gave us cause to doubt that Hermes could kill the cows: 71*.

216

C O M M E N TA RY 1 2 0 – 1 2 1 contrasts with the normative slitting of cows’ throats in sacrifice, which Apollo will assume has taken place (405*). Compare 118* on wrestling versus stunning. 120 ἔργωι δ’ ἔργον ὄπαζε ταμὼν κρέα πίονα δημῶι: ἔργωι δ᾽ ἔργον ὄπαζε refers to the series of actions elaborated in the following lines, beginning either with or immediately after the initial butchery of the animals (ταμὼν κρέα). The phrase has three important connotations. (i) The uses of ἔργον centre these actions within the narrative scheme of Hermes’ κλυτὰ ἔργα (16; so 80* θαυματὰ ἔργα of the sandals). (ii) The stem ἐργ- is often associated with ­sacrifice (Casabona 1966: 301–4; cf. Richardson 1974: 303–4 on δρᾶν), and continues the prompts for us to measure Hermes’ actions against that schema. (iii) The phrase probably alludes to pre-existing proverbial phrases for cumulative toil, ἔργον ἐπ’ ἔργωι and πόνος πόνωι.174 The poetic context intertwines these three connotations, in that the narrative significance of Hermes’ meat-­preparation, implied by (i), is justified by (ii) and (iii). The emphasis on labour in (iii) characterizes Hermes’ future status as a working god; in particular, an Aristophanic conception sees him keeping the Olympian kitchen (Ar. Pax 201–2). Meanwhile (ii) enhances the connection of Hermes’ actions to the sacrificial butchery which was among the traditional early roles of κήρυκες (e.g. Il. 18.558–9, Od. 1.109–12; cf. Berthiaume 1982: 6–7, Jaillard 2007: 160 on Hermes depicted roasting σπλάγχνα), who are under Hermes’ protection; Ariston 1 G–P has a cook dedicating his utensils to Hermes. The details of butchering the victims (references in n. 117) are left implicit, and this preserves a sense of Hermes’ haste. The following lines neatly imply that he has collected the blood, skinned the carcasses, extracted and cleaned the entrails, filled them with the blood, and fashioned wooden spits. See, by contrast, 123*. 121 ὤπτα δ’ ἀμφ’ ὀβελοῖσι πεπαρμένα δουρατέοισιν: Epos shows a formulaic system where forms of μιστύλλω are followed by either καὶ ἀμφ᾽ ὀβελοῖσιν ἔπειρεν+| or πεῖράν τ᾽ ὀβελοῖσιν|. Our line, like Od. 3.463 ὤπτων δ᾽ ἀκροπόρους ὀβελοὺς ἐν χερσὶν ἔχοντες, varies this, and makes room for δουρατέοισιν. The adjective, whose meaning is ‘made of long pieces of wood’ until Phaedimus SH 669, keeps Hermes’ actions at arm’s length from ordinary human sacrificial procedure using metal spits, and also points to his skill at improvising See e.g. Hes. Op. 382 (the line introducing the agricultural calendar) ἔργον ἐπ’ ἔργωι ἐργάζεσθαι, S. Aj. 866 πόνος πόνωι with Finglass (2011), Hp. Epid. 6.3.18, Theoc. 15.20 ἔργον ἐπ’ ἔργωι of spinning, Aristid. 27.7 ἔργον ἔργωι συνάπτειν. D’Alessio (2005) 231 n. 57 suggested that Pi. I. 6.66–7 μελέταν ἔργοις | ὀπάζων may be related to Herm.’s phrase; it is true that in early usage ὀπάζω is normally ‘make attend ’.

174

217

C O M M E N TA RY 1 2 2 – 1 2 3 with whatever materials are to hand, using another act of piercing (as in 119). The hymnist omits any boiling of meat, in conformity with epic norms; this keeps the focus on Hermes’ culinary skills (see Theophrastus’ comment ἐργωδέστερόν φασι τὸ καλῶς ὀπτῆσαι καὶ μᾶλλον τέχνης ἢ τὸ ἑψῆσαι, Ign. 74). Apollodorus substitutes boiling for roasting at Bibl. 3.112. 122 νῶτα γεράσμια: Including, in modern terms, the ribeye, sirloin, tenderloin, and silverside steaks. They were often added to a regular portion as an honorific γέρας: Il. 7.321, Od. 4.65, CGRN 86 A.50–3 (Cos, c.350) γέρη τοῦ βοὸς ... [κάρυξι ν]ώ̣τ̣ου δίκρεας, ὑπώμαια, αἱματίου ὀβελὸς τρικώλιος, Νεστορίδαι̣[ς δὲ] ν̣ώ̣ τ̣ου δίκρεας. So of pork at Od. 8.475, 14.437, CGRN 156.13–14 (Myconos, c.200) [ἱ]ε̣ρέα̣ι ἄρχοντες διδόντων ὀσφύν. The species is not specified in 201.34–5 (Miletus, predating 475 at this point), 42.1–2 (Iasos, early fourth century) λαμβανέτω δὲ τῶν θυομένων σκέλος ἕν, ὁποῖον ἂν θέληι, σὺν τ[ῆι] ὀσφύϊ, 100 (Miletus, c.300). 122–3 καὶ μέλαν αἷμα | ἐργμένον ἐν χολάδεσσι: The emotive connotations of μέλας ‘dark’ enhance its applicability to shed blood (see the useful discussion in LfgrE s.v.). Blood-sausages are mostly mentioned in comedy and sanctuary documents – two types of intertext which contribute here to the impressions that Hermes’ hunger is down to earth and humorous, and that his actions are significantly like sacrifice. See e.g. Ar. Eq. 208, CGRN 39.10–12 (Miletus, c.400) ἢν δὲ [β]ν ἔρ[δη]ι, δύο κρέα καὶ χόλικα̣ [κα]ὶ αἱμάτιον, Sophilus fr. 6 χορδὴ ἡματῖτις, CGRN 86 A.52 (cited on 122*, and specifying that they were grilled on a three-pronged spit). For the beef sausages called χόλικες see Ar. Eq. 1179, fr. 83, Pherecrates fr. 113, Eubulus fr. 63. On Greek sausages in general, and the number of steps Hermes breezes through here, see Frost (1999), Ekroth (2002) 247–51. Matching the parallelism in 119* between how Hermes kills the cows and the tortoise, the detail here echoes the importance of sheep-gut strings (χορδαί) in building the lyre; see also 124*. 123 τὰ δ’ αὐτοῦ κεῖτ’ ἐπὶ χώρης: For ἐπὶ χώρης ‘in place’ (not ‘on the ground’) see e.g. Pi. P. 4.273, Epicurus fr. 96, Chrysippus fr. 523. The ‘other parts’ presumably include the bones and organs, so that the phrase quietly underlines that Hermes has no use for thigh-bones and σπλάγχνα, which were central to human sacrificial communication with the divine and featured prominently in epic descriptions of sacrifice.175 Vergados took τὰ δ’

Kirk (1981) 64. The focus on tasting roast σπλάγχνα is related to their communicative importance in hieroscopy by e.g. Durand (1989) 98–9.

175

218

C O M M E N TA RY 1 2 4 as the parts prepared for cooking in 121–3: while it makes good sense of ἐπὶ χώρης, this interpretation would be better obtained with τάδ’ in asyndeton. 124 ῥινοὺς δ’ ἐξετάνυσσε καταστυφέλωι ἐνὶ πέτρηι: While the meat is cooking, Hermes attends to the cows’ skins. The line adds the stretching of cow-hide to the parallels between the death of the cows and that of the tortoise (119*, 122–3*; see Introduction §3.2). For πέτρη ‘rock-face’, containing a cave, compare e.g. 229, E. Cyc. 195. Barnes conjectured ἐπί, based on Ψ’s πέτρηι ἐπ’ at 404, when Apollo later spots the skins; then πέτρη would mean ‘boulder’. However, Hermes has reason not to leave the hides outdoors, and 401* and 404* will independently suggest that the hides are in a cave and not on a rock. Moreover, in later texts when πέτραι are στυφελαί they are generally much bigger than boulders, and several times associated with caves.176 See also 103–4*. The cave suggests that Hermes’ natal experience with caves will after all help him in his cattle-rustling. Caves are scarce near the Western (‘Pylian’: 216) part of the Alpheus. However, at Olympia there was the artificial ‘Idaean cave’: see Introduction n. 45. Hermes’ separate treatment of the skins creates lasting debris for Apollo to be annoyed at later, and for the aetiology which follows immediately; once again it evokes sacrificial practices, in that numerous ritual regulations list a special destination for victims’ hides as a valuable by-product of sacrifice, including dedication and, more often, sale to support the sanctuary’s finances.177 In the case of dedicated skins, there is some later evidence to support the commonsense expectation that they were initially hung for drying and display: Leonidas 46 G–P (> Zonas 3 GP), D. Chrys. 1.53, Longus 2.30–1.

See the fragment of hexameter fable in Suda σ.1265 (the mountain on which the eagle drops the tortoise); Q.S. 1.295 (Niobe-formation on Mt Sipylos), 11.368, 12.409 (rock-faces), 12.449, 14.475 (rock-faces with caves), and cf. 3.579–80; 3.236, 14.624 (cliffs); 6.478 seems to refer to the ground in a giant cave. Ps.-Opp. C. 4.278 is compatible with either a boulder or a rock-face; a cave is mentioned in 267. καταστύφελος itself occurs in Hes. Th. 806 and Parthenius fr. 33.4 (coloured by Hesiod’s Stygian context). 177 Dedication: e.g. IG VII 235.30 (Oropos, c.380), SEG 28.750 (Lissos, c.300), ΣT Il. 22.159, IG II2 1366 (Attica, second century ce). Subsequent sale: CGRN 44.32–3 (c.413/2), 169.4–6 (Callatis, second century: sale in sanctuary), 212.13–14, 23–4 (Pergamon, c.130); IG II3 445.42 and II2 1496 A.iv mention a δερματικόν in sacred accounts of the 330s; LSCG 85 (Korope, Hellenistic: an annual sale at the assembly); LSAM 77.44–5 (Halicarnassus, 3rd c., funding dedications), IDélos s.v. δέρμα (Delos, second century, funding dedications). See further ThesCRA i.119–22. 176

219

C O M M E N TA RY 1 2 5 – 1 2 6 125–6 ὡς ἔτι νῦν τὰ μέτασσα πολυχρόνιοι πεφύασιν | δηρὸν δὴ μετὰ ταῦτα καὶ ἄκριτον: I shall discuss details first, then the broader meaning of the aetiology. I interpret ‘as still now in after times, long-lived, they are in a state of having grown for a very long time after these events, and indistinguishably’, i.e. the skins have fused completely with rock that is still visible. This involves (i) accepting τὰ μέτασσα, (ii) πεφύασιν not being a copula, (iii) ἄκριτον not meaning ‘continuously’. (i) For τὰ μέτασσα (M) ‘afterwards’ compare τὰ πρῶτα ‘at first’; the rare word was scrambled in Ψ. μέτασσος occurs elsewhere only in connection with Od. 9.221, of animals who are neither πρόγονοι nor ἕρσαι. Though some later Greeks derived that from μεταξύ, it relates to μετά: compare ἔπισσος ‘follower’ in Hecat. FGrH 1 F363, Call.(?) fr. 735, and Gusmani (1961) on the -tyo- suffix in ὀπίσσω, πρόσω and perhaps εἴσω. (ii) In πολυχρόνιοι πεφύασιν | δηρόν, the adjective means ‘long-lasting’ as elsewhere (pace LSJ, Janko 1982: 136); for δηρόν + perfect compare Il. 18.125, where δηρὸν πέπαυμαι could be glossed ‘I am in a state of having stopped for a long time’, i.e. ‘I have long stopped’. Taking πεφύασιν as a copula would produce ‘they have long become long-lasting’, which is feeble. It would also require West’s ἄκριτοι, since things can hardly become long-lasting ‘continuously’ (cf. Od. 18.174 πενθήμεναι ἄκριτον αἰεί), ‘indistinguishably’ or in any other sense of ἄκριτον. (iii) If πεφύασιν … ἄκριτον is right, the sense ‘continuously’ is somewhat redundant. Much stronger is ‘indistinguishably’, which clarifies that the skins have lost their original form and fused to the rock (cf. Nicaenetus 1.1 G–P ἄκριτον ὄρος of contiguous hills), and brings out the nuance of natural development in φύω, especially since the skins’ transformation has involved becoming a new part of nature.178 All this suits the aetiology admirably, and to that we now turn. The other main version of the story also involved aetiological petrifaction (Battos becoming the ‘Look-out point of Battos’), precisely in this section where Herm. diverges (Introduction §3.3.4). Our passage rewrites that story in order to connect the hymn’s audience through a present-day topographical feature near the Alpheus back to illud tempus of Hermes’ actions, and it is flagged elaborately with the temporal markers typical of aetiologies, which emphasize our awe-inspiring distance from that period (μέτασσα, πολυχρόνιοι, δηρὸν δὴ μετὰ ταῦτα). The functional value of such aetiologies very often relies on an audience’s ability to relate them to a familiar object,

178

Vergados’ denial that these lines are aetiological for a rock-formation involves accepting the weaker sense of πεφύασι (against his own parallels) and forcing δηρὸν καὶ ἄκριτον into ‘immeasurably long’.

220

C O M M E N TA RY 1 2 7 – 1 2 8 rather than having to take the object’s existence on trust.179 The emphasis placed on the aetiology tells against explaining away this cave as invented for the plot, or casually transplanted from Messenian Pylos, where, as Müller (1833) noted, in Ant. Lib. 23 Hermes hides the cows in the so-called ‘Cave of Nestor’, whose stalactites are ripe for aetiology (though not in fact aetiologized by Antoninus). However, as mentioned in Introduction §2.1, there are few caves by the Alpheus except the artificial one(s?) at Olympia, which was known to Pseudo-Pindar and Demetrius of Scepsis, but does not seem to have been famous; hence the lines would have been most effective if the original audience was on site at Olympia. 127 χαρμόφρων: A hapax cited at Hsch. χ.209 Χαρμόφρων: ὁ Ἑρμῆς. That entry seems to take the adjective as an epiklesis rather than a passing description.180 However, the adjective has a particular aptness here: Hermes’ delight at his actions, expressed as χάρμα, bodes well for human attempts to induce χάρις-exchange through animal sacrifice. εἰρύσσατο πίονα ἔργα: As in 121*, the phrasing varies but evokes the sacrificial formulas including ἐρύω, which are summarized in LfgrE ii.724.45–8 (add the variant at Od. 19.423). πίονα ἔργα not only recalls 120 ἔργωι δ’ ἔργον + πίονα δημῶι as Hermes returns to his cooked meats, but is a further and wittier instance of formulaic modification: the hymnist seems to have noted the coexistence of πίονα δῆμον+| and πίονα δημόν+|, as well as πίονα ἔργα| in the sense ‘rich agricultural fields’ (Il. 12.283, Od. 4.318, H Dem. 93), which is close to πίονα δῆμον, and completed the square by assigning to πίονα ἔργα the new meaning ‘rich handiwork’, where the richness is measured in δημός. On these phrases see Nagler (1974) 5–9, 12, 37–44. 128 λείωι ἐπὶ πλαταμῶνι: Nature’s table. The background sacrificial tables are for the carving of cooked hunks of meat, and for the presentation of divine portions as a display of honour (Gill 1991: 7–27). The first use is

179

For Eliade’s conception of illud tempus see Introduction at n. 39. Pelliccia (1989) discussed temporal markers like ὡς ἔτι νῦν in aetiologies; Hunzinger (2012: 50 n. 48) gives the other passages where H.Hom. refer to the performer’s ‘now’. For the register and/or date of μετὰ ταῦτα see Introduction §1.1. 180 Compare similarly curt entries about Hermes, none of which are demonstrably just passing descriptions: α.7022 Ἀργειφόντης, 8328 Αὐξίδημος, ε.2491 ἐμπολαῖος, 2597 ἐναγώνιος, 3207 Ἐννήιος (in Chios), 3232 Ἐνόδιος (Paros), 4120 Ἐπάκτιος (Sicyon), 4786 Ἐπιθαλαμίτης (Euboea), 5086 Ἐπιπολιαῖος (Rhodes), 5319 ἐπιτέρμιος, 6677 εὐάγγελος, 6932 Eὔκολος (Metapontum), 7139 Εὐρυμέδων, π.73 Παιδοκόρης (Metapontum), σ.2036 στροφαῖος (< Ar. Pl. 1153), φ.656 Φλυήσιος (separate from Hipponax fr. 47?), χ.799 χρυσόρραπις.

221

C O M M E N TA RY 1 2 8 – 1 2 9 immediately relevant, while the second becomes so over the remainder of 128–9. For the recurrence of this phrase at A.R. 1.365 (involving a new context, but secured by a sequence of parallels) see Introduction §4.2. It is unclear whether the Georgian Hymn to Dionysus’s πλα]ταμὼν λ[εῖ]ος (Furley 2007, line 4), and Hesychius’ comment (π.2472) that ‘others’ use πλαταμών in the sense of λεωπετρία, are also related. 128–9 καὶ ἔσχισε δώδεκα μοίρας | κληροπαλεῖς, τέλεον δὲ γέρας προσέθηκεν ἑκάστηι: Hermes is acting secretively (see 66 ὁρμαίνων δόλον, his confusion of footprints, and his threats to the Onchestan) in the middle of the night, so he does not want any guests; nor does it make sense to create twelve portions just for himself. Hence eleven of the portions have a purely symbolic value. The symbolism lies partly in their emphatic equality: their distribution by lot motivates Hermes to carve them equally; furthermore each has a τέλεον (‘full-size’) γέρας, overturning the fundamental use of γέρα to distinguish a few honorands.181 The number twelve symbolizes Hermes’ wish to carve out an equal portion of divine status within a Dodecatheon, where gods seem to have been treated equally.182 The Dodecatheon could in principle be interpreted with reference to a specific cult, or taken as a synecdoche for the whole pantheon – a question we will return to on 425–33*. There are three objections to the latter interpretation: it was not until the fourth century that the synecdoche became widespread in texts and art (Long 1987: 331–4); the audience knew that Hermes remained a ‘lesser’ god in almost all localities; and his prime focus has been on taking over Apollo’s prerogatives specifically (see esp. 173).183 I therefore think an audience would have interpreted the scene through an allusion to a particular Dodecatheon, and the only known one near the Alpheus was at Olympia, where in fact Hermes and Apollo attained the perfect symbol of fraternal equality in that they shared one of the six twin altars (Introduction §2.2).

For κληροπαλεῖς ‘for ballot-shaking’ cf. Stes. fr. 97.223 κλαροπαληδ̣ όν. Berthiaume (1982) 50–1 discusses the principle of equality of portions without a γέρας at Greek sacrifices; see esp. Plu. Q. Conv. 3.1 642f–3a which includes the principle of fair distribution by ballot. τέλεος, even in a different application here, supports the passage’s indirect connection to the language of ritual documents, where it very frequently describes a fully grown victim. For the significant contrast of Hermes’ drive for equality with the carving of Hesiod’s Prometheus, see below on 112–41*. 182 The symbolism was already outlined in Gemoll 184. On Dodecathea see Long (1987) and Georgoudi (1996), discussing Herm. on pp. 67–70. 183 Jaillard (2007) 126–7 noted that the symbolism here is less confrontational than what has so far seemed antagonism just towards Apollo, but he interpreted this as an early intimation of the more conciliatory latter stages of the hymn. 181

222

C O M M E N TA RY 1 3 0 As Kahn (1978: 65) noted, Hermes expresses a desire, using his own ballot, to tweak the pre-existing distribution of divine τιμαί, which was sometimes itself conceived as the result of a ballot (see 428–9*). Indeed, the audience probably considered that Hermes had acquired some τιμή in precisely the matter of ballots: see Ar. Pax 365; Suda κ.1785 cites Euripides’ Aeolus for the practice of drawing an olive leaf first from the ballot and calling it ‘Hermes’. 130 ὁσίης κρεάων ἠράσσατο: For this use of ὁσίη the closest parallel seems to be CGRN 99 A.21–5 (Cyrene; fourth century, with earlier material): Ἀ]καμαντίων ὁσία παντὶ καὶ ἁγνῶι καὶ βαβάλω̣[ι]. πλὰν ἀπ᾽ ἀνθρώπω Βάττω τῶ ἀρχαγέτα καὶ Τριτοπατέρων καὶ ἀπὸ Ὀνυμάστω τῶ Δελφῶ, ἀπ᾽ ἄλλω ὅπη ἄνθρωπος ἔκαμε, οὐκ ὁσία ἁγνῶ. τῶν δὲ ἱαρῶν ὁσία παντί. Everyone has ‘ὁσία of ’ the Akamantia and the hiera, but there is not ‘ὁσία for the pure from’ several tomb-cults and other places of death: this may mean a right of entry, though Parker (1983) 336–9 included the suggestion that ‘ὁσία of ’ and ‘ὁσία from’ refer to the right to eat cult-offerings.184 This strange target for ἠράσσατο makes an expressive contrast with 64 κρειῶν ἐρατίζων*. There, Hermes’ feelings about meat-consumption were notably humanized, in a broadly ‘comic’ conception, but with traditional referentiality aligning him with a lion. Here the punctual (aorist) thought of ὁσίη marks the sudden interposition of Hermes’ awareness of norms about divine diet, as opposed to both the human and the bestial (Jaillard 2007: 147): his hunger is an aberration within the world of gods constructed by Herm. (as 131 will spell out), an infant’s neglect of social propriety in favour of bodily desires. Both points were already suspected at 64*. The humorous humanization of Hermes’ impulses is not, however, abandoned with this new consideration. ὁσίη is a term pointedly drawn from restrictions on human eating at sacrifices (cf. Peels 2016: 249), here flipped to describe what Hermes should give up.185 It contributes to the sense that Hermes’ actions are a precursor to human rituals (Introduction at n. 41). κρεάων (rather than κρειῶν) is transmitted here and at Nic. Alex. 258, and is metrically secure at Alcaeus of Messene 2.2 G–P. See GH i.209–10.

184

Cf. Ebeling (1885) 89, Versnel (2011) 323 n. 45. Jeanmaire (1945), by contrast, suggested that in six passages ὁσία describes something ‘desacralized’. However, 173*, H.Dem. 211 and H.Ap. 237 all refer as normal to r­eligiously acceptable behaviour; our passage and the Cyrenaean law can be taken as straightforward specifications of that; Jeanmaire’s final passage is corrupt (Diggle 1970: 118–20). Jaillard (2007: 108) took ὁσίη to mean ‘honorific part’, based on a misunderstanding of the relationship of ὅσιος and the concept of τιμή: see 173, 470*. 185 Cf. Pi. P. 9.36, where Cheiron asks Apollo if it is ὁσία to rape Cyrene, while teasing him about his apparent lapse from omniscience.

223

C O M M E N TA RY 1 3 1 – 1 3 2 131–2 ὀδμὴ γάρ μιν ἔτειρε καὶ ἀθάνατόν περ ἐόντα | ἡδεῖ’: With ἔτειρε compare the similar application to hunger at Od. 4.369. The separation of ὀδμή and ἡδεῖα, both at line-start, conveys how Hermes’ hunger is exacerbated by the smell. That further touch of anthropomorphism runs up against ‘though he is immortal’. The other gods of Herm. are subject to the normal and different conception of how the smell of sacrifices is appreciated: the savour from burnt fat (κνίση), rather than the smell of roast meat, is an index of what they really appreciate – honour (cf. Versnel 2011: 310). The reference to Prometheus’ separation of what men and gods get from a sacrifice caps a series of contrasts between him and Hermes in this episode (Introduction §1.3).186 132–3 ἀλλ’ οὐδ’ ὧς οἱ ἐπείθετο θυμὸς ἀγήνωρ | καί τε μάλ’ ἱμείροντι περᾶν ἱερῆς κατὰ δειρῆς: Hermes’ hunger for meat was stated in 64, implied again in 108, and brought to a head in 131. Why when it comes to the crunch does he not eat? The problem perhaps troubled Apollodorus, whose Hermes tastes his cooking (Bibl. 3.112). Hermes’ motivation against eating was hinted at in 130* ὁσίης: part of him becomes aware that gods are not supposed to eat meat. This line of thought suits 133 ἱερῆς too: the modification of formulaic ἁπαλῆς ἀπὸ δειρῆς+| (Il. 3.371, 13.202, 18.177) underlines the significance of the adjective. The hymnist’s description of Hermes’ internal conflict forms part of the complicated development of Greek representations of psychology. It locates Hermes’ new motive in his θυμός (previously unmentioned), which resists his ἵμερος to eat and the anguish of sniffing the cooked meat – and resists successfully despite some expectation that it might accede to strong desires. ἐπείθετο describes responses to instructions, advice or requests: hence ‘listen’ rather than either ‘obey’ or ‘was … persuaded’; as in e.g. Od. 2.103, the dative pronoun (οἱ, there ἡμῖν) marks the owner of the θυμός. The phrase leaves it open whether the θυμός had previously been neutral, against eating, or irrelevant. The contrast of impulses – an incipient sense of social propriety in the θυμός (which considers ὁσίη), versus the desires of the senses and belly – superficially resembles the conflicts between the θυμοειδές (concerned for social propriety and positive face) and ἐπιθυμίαι identified at Pl. R. 4 439e–41c. However, Hermes’ θυμός has some executive capacity, as in the epic background, where θυμός has very wide-ranging functions including decision-making.187

186

Clay (1989: 124–6) inferred that Prometheus’ sacrifice predates Hermes’ actions. But at Hes. Op. 77–8 Hermes – apparently adult – helps to fashion Pandora. The mythical chronology is not coherent, but what is important is that the parallels to Prometheus are available for the audience to measure Hermes’ actions against. 187 For a survey of the epic θυμός see LfgrE, Sullivan (1995) 54–69; Cairns (2014) discusses the relationship to Plato, and the way characters project unworthy thoughts onto their θυμός during deliberative soliloquies.

224

C O M M E N TA RY 1 3 2 – 1 3 3 The θυμός is so protean that in other passages it is also the locus of hunger: in particular, at Il. 12.300 a lion’s θυμὸς ἀγήνωρ bids it seek out meat. We saw that at 64 Hermes’ hunger was described in the terms of an Iliadic lion-simile; perhaps here a second such simile is alluded to in order to mark how Hermes’ θυμός has now taken him away from simpler bestial impulses. However, the θυμός is still ἀγήνωρ, ‘manly’. After the preceding lines’ e­ xploration of the extent of Hermes’ humanization, one should hardly sweep this epithet aside as a loose application of the Odyssean formula -πείθετο θυμὸς ἀγήνωρ|.188 Though no longer humanized by hunger, Hermes’ restraint is still framed through a humanizing consideration about ὁσίη (130*). The balance may come out also in ἱερῆς κατὰ δειρῆς, where Vergados observed that δειρή is not a normal word to use in the context of eating. If it had its connotation of vulnerability here, the combination with ἱερῆς would again express the tension of Hermes’ decision to take the part of a normal god. Contrast Clay’s interpretation (1989: 117–26, largely followed by Vergados) that Hermes initially expects the gods to come to eat but is unsure of his own divinity, then ‘finds himself unable to swallow the meat’, an inability which is ‘unmistakable evidence of his divinity’, and goes on to hide his handiwork out of embarrassment. This interpretation of his initial expectation (shared by Leduc 2005: 158–62) goes against his secrecy throughout the cattle-­rustling episode. Moreover the hymn contains no sign of his alleged uncertainty about his status, realization now that he is a god, or subsequent embarrassment; a character who expects gods to come to dinner can hardly infer from his inability to eat that he too is a god. A better gloss on his change of mind is that he finds himself ‘able not to swallow the meat’. καί τε: For this combination used outside a permanent or gnomic situation, see 92–3*. περᾶν: Franke (περᾷν Barnes): Ω seems to have had περῆν’. Epos does not elide infinitives in -ναι (GH i.86), and the senses of περαίνω are wrong. Barnes’s suggestion remains doubtful since περάω is not otherwise causative and πέρνημι (fut. infinitive περάαν) is about trade. But διαπεράω can mean ‘make to cross water’ (LSJ s.v. III), and the thematic nature of crossing in the hymn perhaps encouraged an extended use here (Introduction §3.3.6). Other approaches to emendation involve a complex change, e.g. taking περ as an intrusive gloss on the concessive use of καί, and restoring something like ἱμείροντι διεῖναι ἧς: cf. Allen (1897) 258, comparing Il. 24.642 λαυκανίης καθέηκα, Agar (1916). But this does not account for the origin of ἱερῆς. 188

ἀγήνωρ describes non-humans also at Hes. Th. 237 (Phorkys), 641 (Hundred-handers).

225

C O M M E N TA RY 1 3 4 – 1 3 6 134–5 ἀλλὰ τὰ μὲν κατέθηκεν ἐς αὔλιον ὑψιμέλαθρον, | δημὸν καὶ κρέα πολλά: The structure closely echoes 63–4: particle, demonstrative, μὲν κατέθηκε, location | noun glossing the demonstrative. The meat must be hidden as the lyre was, as potential evidence of Hermes’ schemes. Once more this demonstrates the dynamic symmetry of 20–67 and 62–145 (Introduction §3.2). 135–6 μετήορα δ᾽ αἶψ᾽ ἀνάειρεν, | σῆμα νέης φωρῆς: After κατέθηκεν, Hermes promptly performs a contrasting action: he may suspend the meat from above (but how?), place it on a raised ledge, or pile it up from the ground (cf. μετεωρίζειν of piling up a rampart in Th. 4.90.2). The hidden meat then becomes, paradoxically, a σῆμα of the theft it also conceals.189 It is not, like Hermes’ obscured footprints, a concealing sign designed to be seen by Apollo but to confuse him. Rather, its signal is to be transmitted to future witnesses, like a victory-trophy (Bothe), if they are given the suitable background myth. The hymnist therefore suggests, in the vicinity of 124–6, that this is a further aetiology for an object beside the Alpheus. The apparent man-made nature of the αὔλιον (103*) and the vagueness of μετήορα ... ἀνάειρεν might have prompted an audience to connect this object to a structure inside a building, for example, though we can no longer be sure.190 Moreover, this aetiology is framed quite differently from the one at 124–6, since the hymnist implicates himself (as the revealer of the background myth) as an integral intermediary in Hermes’ signalling to his future worshippers. The whole phrase was perhaps inspired by Odysseus’ dedication of Dolon’s helmet, Il. 10.465–6 ἀπὸ ἕθεν ὑψόσ’ ἀείρας | θῆκεν ἀνὰ μυρίκηι, δέελον δ’ ἐπὶ σῆμά τ’ ἔθηκεν. Recall the similarity of 81* to Il. 10.467, and the obvious utility of the Doloneia for a poet with a night raid to describe. 136–7 τὰ δ᾽, ἐπὶ ξύλα κάγκαν’ ἀείρας, | οὐλόποδ’ οὐλοκάρηνα: Ω had ἐπὶ δὲ ξύλα …, where ἐπί could be taken as adverbial (cf. Hdt. 7.69, 219 for ἐπὶ

Hermann’s emendation of Ψ’s φωνῆς (M omitted 136 through homoeoteleuton) is secured by 385 φώρην M: φωνήν Ψ. ν/ρ is a simple minuscule error. For φωρά ‘theft’ add Crinagoras 30.3 GP to the parallels in LSJ. Vergados took the apposition to mean ‘he lifted them up since they were a sign’, but one would expect ὡς or ἐόντα to clarify the causal nuance. In Vergados’ interpretation, Hermes changes his mind in the middle of 135 and takes the flesh and fat back out to the fireplace for burning, but the poet expresses neither a change of psychology nor of location. Emendation, such as ἂψ ἀνάειρεν σήμαθ’ ἑῆς (cf. Schopen 1846), would be called for. 190 Eitrem (1906: 259 n. 16) suggested that Hermes’ action represents the invention of meat-curing, but in the absence of references to salting or smoking this seems farfetched. 189

226

C O M M E N TA RY 1 3 7 – 1 3 8 δέ, ‘and next’) or in tmesis. Either approach raises several problems which are solved by the inclusion of a pronoun as in my emendation. Adverbial ἐπί is not supported by epic usage; it leaves the body-parts other than the flesh, fat, heads and feet unaccounted for, and potentially lying around for Apollo to see; it requires οὐλόποδ’ οὐλοκάρηνα to modify ξύλα in the sense ‘along with all the feet and heads’ rather than ‘complete with the feet and heads’.191 If ἐπὶ ... ἀείρας were in tmesis the second and third problems would still arise. The adjectives could now modify the flesh mentioned in 135, but that only causes new problems: Hermes’ ‘sign’ gets destroyed immediately, the fireplace suddenly moves, and no antithesis is left for 134 τὰ μέν.192 The prefix οὐλο- evokes ritual language (compare e.g. ὁλόκαρπος, ὁλόκαυτος, ὁλόκληρος), and is doubled and further reinforced by the completeness suggested by the opposition feet/head. That pairing also evokes rituals: heads and feet were disposed of in one pile separate from other debris in the sanctuary of Apollo Hylates at archaic Kourion (ThesCRA i no. 530b); and they are often mentioned – normally consecutively – in lists of γέρα.193 The first of the adjectives may demonstrate the transformation enacted on the cows: their entry into the world of ritual has converted εἰλίπους into οὐλόπους; might οὐλοκάρηνος mark a similar change from οὐλόκερως? The phrasing recalls 112–13 and so initiates ring-composition around the Alpheus episode (see 112–41* below); a second load of wood is needed to raise the fire’s setting back from barbecue to burn, as Allen and Halliday observed in different terms. 137 ἀϋτμῆι: See 110*. 138 πάντα κατὰ χρέος: One can limit πάντα to the clearing up, and take κατὰ χρέος as ‘according to (his) need’ (so e.g. A.R. 3.189), i.e. to conceal the traces from Apollo. But after extensive allusions to ritual language and procedures, one should not suppress the sense ‘according to required procedure’

Compare Pherecrates fr. 113.13 σχελίδες ὁλόκνημοι, ‘ribs complete with the whole shanks’ (in one mass), not ‘ribs and all the shanks’ (separately). 192 See n. 189 for Schopen’s approach to the problem. Schneidewin (1848: 671) proposed a lacuna; Greve (1867: 41–2) saw the need for a pronoun but made no suggestion. 193 E.g. CGRN 99.98–9 (Cyrene, reinscription of archaic material), 45 fr. 3.43, 56 (Athens, c.400), 30 B.2 (Delphi, c.400), 85.58–9 (Cos, c.350), 42.3 (­Iasos, fourth century), 104.44 (Halicarnassus, c.300), 193 B.20–1 (Hyllarima, c.197), Demo FGrH 327 F1, ICos ED 216 B.5–6 (Cos, c.200), IG II2 1366.10 (Laurion, c.200 ce), and probably Hsch. ε.2822. In CGRN 169.5 (Callatis, second century) they are among parts sold off to fund the sanctuary, while in LSAM 54 (Didyma, third century) both are sold off separately. 191

227

C O M M E N TA RY 1 3 9 – 1 4 1 (e.g. A.R. 4.889; LSJ s.v. χρέος II.2), with πάντα referring to the whole set of preceding actions. δαίμων: The singular is rarely applied to known gods in Epos, but cf. 343*. Here the choice of word is conditioned by the context of butchery, since it activates the etymology ‘apportioner’ (originally of fate: cf. μοῖρα ~ μέρος, Il. 6.115, 8.166, 19.188, Od. 2.134). Compare Alcman PMG 65 δαίμονάς τ᾽ ἐδάσσατο. 139 σάνδαλα μὲν προέηκεν ἐς Ἀλφειὸν βαθυδίνην: Hermes must get rid of the sandals which would explain the strange traces left on the sand. The epithet βαθυδίνην, though traditional of large rivers, here neatly points to how the Alpheus will destroy the evidence. 140 ἀνθρακιὴν δ’ ἐμάρανε, κόνιν δ’ ἀμάθυνε μέλαιναν: Hermes’ cover-up continues with his hearth. ἀνθρακιὴν δ’ ἐμάρανε could probably have been understood as describing a shift from flaming to smouldering logs, even though ἄνθρακες are normally already at the smouldering stage. For the aorist used to describe such a shift see Il. 9.212–13, φλὸξ ἐμαράνθη (to create an ἀνθρακιή). Hermes’ still smouldering embers will (in my view) reappear at 415–16*. For the Greeks’ preservation of smouldering embers see 237–8*. For them, it was not natural for Hermes to quench the embers completely before covering them over, so there is little force in the objection (Cobet 1862: 311, proposing ἐμάραινε) that he fails to complete the job properly. A separate question, discussed already by Hermann, is whether to spell the aorist ἐμάρηνε or to invoke the parallel of Il. 21.347 ἀγξηράνηι. Asclepiades has μάραν- (3.3 G–P) but used features from various dialects (Sens 2011: lxv–lxxii). The neat phonetic doubling ἀμάρᾱνε … ἀμάθῡνε leads on seamlessly to Hermes making the ashes indistinguishable from the surrounding sand. For the development ‘turn to dust’ > ‘level with dust’ compare Q.S. 14.643–5 χείμαρροι ... πρὶν τείχεα πάντ’ ἀμαθῦναι; Demeter ἀμάθυνεν the nymph Minthe not by pulverizing her but by trampling her completely into the soil, whence she presently resurfaces as mint (Opp. H. 3.492). Hermes presumably achieves his aim by putting sand over his ash-pile, but that is not to say that ἀμαθύνω here, and only here, means ‘cover with sand’ (DGE). 141 παννύχιος: Hermes’ actions at the Alpheus are set within ὄρθρος (98*–101, 143) so that παννύχιος bears the meaning ‘to the end of the night’ (as at Il. 10.2); it and πανημέριος normally imply a more substantial duration, but this is less clear at Il. 23.105, where Achilles presents the brief dream of Patroclus as παννυχίη. Conceivably the word was once again

228

C O M M E N TA RY 1 4 2 – 1 4 5 chosen for its cultic resonance, i.e. all-night festivals (see H Dem. 292, E. Hcld. 782, Ba. 862), and to multiply the references to time which ring this episode at 97–100 and 141–3. Wecklein’s simpler ἐννύχιος (1920: 38) deserves consideration. ἐπέλαμπε: There is little to choose between this and M’s κατέλαμπε. Vergados’ claim that ἐπιλάμπω entails an unsuitable contrast with a preceding state of darkness (‘supervened and shone’) is refuted by e.g. Hp. AWP 8 οὐ γὰρ ἔτι ὁ ἥλιος ἐπιλάμπει. 112–41 Aetiological cookery: For the functions of these lines in characterizing Hermes and relating his actions to features of the audience’s world, see Introduction §2.1–2. The cohesion of the passage with the immediately preceding lines is expressed by ring-composition: 136 ἐπὶ ξύλα κάγκαν’ ἀείρας echoes 112–13 κάγκανα κᾶλα … ἐπέθηκεν; 137 ~ 110 ἀϋτμή; 138 ἐπεὶ ... κατὰ χρέος ἤνυσε may recall the well-completed action of 105 ἐπεὶ εὖ ἐπεφόρβει. Finally 139 recalls the mention of the Alpheus in 101, and 141 that of moonlight at 99.

142–5 142 δῖα: An epithet not normally used of mountains despite its notoriously broad application (99*). Perhaps Kyllene would have been understood as ‘divine’ (because it is Hermes’ birthplace) or ‘bright’ with snow (cf. Il. 2.735 Τιτάνοιό τε λευκὰ κάρηνα). The hymnist modified boldly the formulaic ἀφίκετο+ δῖα θεάων/ γυναικῶν (Cantilena 1982: 249). 143 ὄρθριος: See 112–41* for this word’s role in ring-composition framing the Alpheus episode. 145 λελάκοντο: An aorist built from a reduplicated stem: for the linguistic details, see Nussbaum (1987). 142–5 λελάκοντο. Returning home: These lines show the hymnist’s deftness in suggesting the speed and serenity of Hermes’ return home. As often αἶψα (142) qualifies an arrival, and marks the journey as something swift and negligible for narrative purposes (‘presently’ rather than ‘straightaway’). In this passage the effect is enhanced: normally in such cases a summary description of the journey precedes (see 322, Il. 2.667, 6.370, etc.), but the description here is minimal and follows. The negatives in 143–5 tell us that Hermes had an undisturbed walk, but register the risks avoided. In particular, 143 δολιχῆς ὁδοῦ recalls 86 δολιχὴν ὁδόν, which was followed immediately by

229

C O M M E N TA RY 1 4 5 – 1 4 6 a human witnessing Hermes.194 This time, ‘neither human nor god’ sees him (144). This pairing occurs sporadically throughout Herm.;195 here it resonates particularly with Zeus’s nocturnal journeys to Kyllene, unnoticed by gods and humans (9) – like father like son, at least partly (see 145–54*). Hermes also eludes dogs, which are more likely to be on the prowl at night (Franke), and which are the traditional enemy of thieves (e.g. Hes. Op. 604) and of Hermes in particular: cf. his dealings with Argus, and e.g. Hipponax fr. 3a Ἑρμῆ κυνάγχα, μηιονιστὶ Κανδαῦλα, | φωρῶν ἑταῖρε. Antoninus Liberalis 23.2 has dogs guarding Apollo’s herd (cf. 193–5*) and Hermes πρῶτα μὲν ἐμβάλλει ταῖς κυσίν, αἳ ἐφύλαττον αὐτάς, λήθαργον καὶ κυνάγχην.196 The verb ἀντεβόλησεν (143) recalls 26*, where it hinted at Hermes’ control over positive chance encounters, to add the complementary point that he has some control over avoiding negative chance encounters too. And so he traverses the 70 km back to Kyllene in peace, with Zeus-like confidence.

145–54 145 Διὸς δ᾽ ἐριούνιος Ἑρμῆς: The bare genitive patronym is only paralleled in Epos by the formula Ὀϊλῆος ταχὺς Αἴας (x7), but for later parallels see e.g. Hipponax fr. 35 (cited on 408), Finglass (2011) on S. Aj. 172–5 (five examples in that play, four with Zeus as father). There is insufficient reason to take Ἑρμῆς as an intrusive gloss on υἱός (cf. 28). 146–7 δοχμωθεὶς μεγάροιο διὰ κλήϊθρον ἔδυνεν | αὔρηι ὀπωρινῆι ἐναλίγκιος, ἠΰτ’ ὀμίχλη: Hermes avoids having to open the front door, in another sign of his preternatural skill at crossing boundaries while leaving them intact.197 Athena can similarly get past Nausicaa’s bedroom door ἀνέμου ὡς πνοιή at Od. 6.20, but whereas that passage leaves it open that her body may have transformed briefly into air, ours suggests a more mysterious freedom from physical laws. δοχμωθείς (contrasting with 148 ἰθύσας) suggests the embodied movement of squeezing through a narrow gap side on, or (LfgrE s.v.) of deviating towards the side of the door; it is hard to imagine Hermes turning into an ‘autumnal’ breeze when we have no reason to think the story is set in autumn. But κλήϊθρον (‘barred door’: Barrett 1964: 269) gives no clue 194

The parallelism extends backwards: at 86 Hermes had just taken the cows and invented his sandals; here he has just discarded both. 195 Cf. 161, 338–9, 441, 444–5, 526, 565. It is also common in H.Dem., especially towards the start: 11, 22, 45, 55, 62, 73… 196 Cf. Triphiodorus 503–4 for lack of barking as a notable part of ἡσυχία. 197 See Introduction §3.3.6. The courtyard door (26) is forgotten.

230

C O M M E N TA RY 1 4 8 to a physical gap. The juxtaposition of similes with colourless and opaque vehicles makes Hermes’ movement even harder to visualize. See 380* for some confirmation about how he enters the cave. For paired similes see e.g. 43–5, Il. 21.251–2, Od. 19.233–4; Nodell (1787) compared A.R. 4.877 πνοιῆι ἰκέλη δέμας, ἠΰτ’ ὄνειρος, where δέμας suggests that Thetis actually transforms into a breeze.199 In our case, the first vehicle is welcome, as αὖραι normally are – a gentle draught in the time of Sirius (cf. Il. 5.5); the second more sinisterly recalls mist being ‘not at all welcome for herdsmen, and better than night for a thief ’ (Il. 3.10–11). For the theme of Hermes’ ambivalence see esp. 577–8*. For the idea that the chiasmus in 147 may support the concepts of crossing and transformation see Introduction n. 63. 198

148 ἄντρου: Either a genitive of the space within which the movement ἰθύσας occurs (cf. GH ii.58–9), or the whole to which νηόν belongs. It makes little difference to the sense. νηόν: Despite there being a cave-sanctuary on Mount Kyllene (2*), the terminology of temples does not transfer easily. Here νηός denotes an inner part (separate from the ἄδυτοι of 247*) of the overall locked cave; in architecture, the overall νηός contains an inner locked cella. The most fruitful approach is to assume etymological play on ναίω.200 Though he actually has the run of Greece, and it is this range which brought him his few later νηοί, Hermes’ proper ‘dwelling-place’ at this point is only a portion of the cave, in the immediate vicinity of his cot. The jarring word thus prepares us to contrast Hermes’ mobility in 146–7 with Maia’s view in what follows about where he ought to be. See 145–54* below, and 359* for Apollo’s version of Hermes’ entrance.

Contrast Od. 4.802 where the dream of Penelope’s sister enters her chamber παρὰ κληῗδος ἱμάντα, i.e. through the strap-hole; see NP s.v. ‘lock’ and Diels (1897) 123–51 for different locking-mechanisms available at Herm.’s date. Nor is there a reference here to a keyhole (ὀπή, κλειθρία). 199 I doubt whether Apollonius, though he has just imitated H.Dem. 231ff. extensively, was consciously inspired by Herm. here. See also Q.S. 4.111 αὔρηι ὑπηώηι ἐναλίγκιος, also of Thetis, which may have been inspired by our line to complement her going to the Greek camp ἠΰτ’ ὀμίχλη at Il. 1.359. 200 Cf. H.Ap. 298 ἀμφὶ δὲ νηὸν ἔνασσαν, Cleinias in Σ A.R. 2.1085a. Temples and their dedications were frequently conceived of as the deity’s house and possessions: see ἱερὸς δόμος at Il. 6.89, Hy. 24.2; gods go to stay in their temples at e.g. Alcm. PMG 14b, Sappho fr. 2, PMG 871; see Burkert (1988) 29–36, and Vink (1995) for the houselike arrangement of early temples. 198

231

C O M M E N TA RY 1 4 9 – 1 5 2 149 ἦκα ποσὶ προβιβῶν· οὐ γὰρ κτύπεν ὥς περ ἐπ’ οὔδει: More cunning walking (cf. 75–8*), this time concealing sonic rather than visual evidence. For thematic byforms of βίβημι compare 225 βιβᾶι, and Bechtel iii.178–80 for Ionic evidence. For ὥς (περ) with the conditions in which one would expect a verb to obtain, see e.g. Il. 18.518 καλὼ καὶ μεγάλω σὺν τεύχεσιν, ὥς τε θεώ περ, Od. 9.422–3 μῆτιν ὕφαινον, ὥς τε περὶ ψυχῆς, Th. 2.65.11 πολλά, ὡς ἐν μεγάληι πόλει καὶ ἀρχὴν ἐχούσηι, ἡμαρτήθη. 151 σπάργανον ἀμφ’ ὤμοις εἰλυμένος: A hard stop after 150 leaves us with the harshest asyndeton in Herm.’s narrator-text.201 Postponing the stop until after εἰλυμένος (on the ground that ἠΰτε often stands in asyndeton) would imply that Hermes has had his upper arms swaddled even while butchering two cows, and raises a distracting question about why the narrator only mentions this marvel now. Ilgen’s conjecture σπάργανα δ᾽ oversimplifies, and the echo at 306 σπάργανον ἀμφ᾽ ὤμοισιν ἐελμένος supports the singular here. Schneidewin (1848) 671 ventured a lacuna. I prefer to take the lemma with what follows: the narrator provided the merest hint that Hermes unswaddled himself in 22*, as his actions thereafter suggest; all of 151–3 describes the result of Hermes reswaddling himself to look innocent, an action discussed below in 145–54*. For a later discussion of how Greeks swaddled babies see Soranus Gyn. 2.14–15, with Burguière, Gourevitch and Malinas (1990) 90. If it relates to an extant passage, Hsch. α.4164 ἀμφώμοις: ἀναβλήμασιν, ἀναθήμασιν ‘with things thrown back over or placed on the shoulders’ (cf. LSJ s.vv. ἀνατίθημι B 1, ἀναβάλλω B III) is most likely to have arisen by rearticulating either our phrase or 306 ἀμφ’ ὤμοισιν. 152 περὶ γνυσί: γνυσί is an old zero-grade alternative to γούνεσσι or γούνασι, identified here by Forssman (1964); cf. Katz (2007) with further references. Ω had περ᾽ ἰγνύσι, and the likely imitation at Ps.-Theoc. 25.242 μακρὴν δὲ περ᾽ ἰγνύηισιν ἕλιξε | κέρκον suggests that Herm. was articulated in this way already in the late third century. But prepositional περί is not elided in Epos.202 The unusual γνυσί led to misdivision. λαῖφος ἀθύρων: λαῖφος is used in Epos of various broad pieces of material – sails, beggar’s rags, Pan’s lynx-skin cloak. Hence λαῖφος probably denotes a broad outer wrap in Hermes’ swaddling. This is one of the methods suggested by ancient art, though one must reckon with the likelihood of stylization 201

For Herm.’s marked use of asyndeton in direct speeches, see n. 60. For elision of the preverb, see Hes. Th. 678 περ-ίαχε. For knowledge of Herm. in Ps.-Theoc. 25, overlooked by Reece (2009) 241, see also Introduction §4.3.

202

232

C O M M E N TA RY 1 4 5 – 1 5 4 and simplification. The other uses of λαῖφος might suggest rather rough material. However, σπεῖρον includes sailcloth, beggar’s clothing, and a wrap, but also a shroud of high-quality weaving; and ῥάκος is used ­pejoratively of beggar’s rags and non-pejoratively of strips of swaddling. In any case one can contrast the more overt pretension of Apollo being swaddled in a ‘fine new white φᾶρος’ (H.Ap. 121–2). Here ἀθύρω takes an outer object – a construction with hardly any parallels, though cf. the corresponding passive in 485, Nonnus D. 45.244 κόμπον ἀθύρων, and παίζω + outer object (LSJ s.v.). Normally the verb’s object is internal, i.e. it refers to the game played or something produced by play. See further 145–54* below. 203

145 Διὸς–154 Baby and adult: Maia’s threshold is a boundary both between the space of Hermes’ aspirations and that of his babyhood, and between major sections of the poem. The motifs of swift movement between threshold and cot connect these lines (esp. 150) back to the section-junctures at 20–6 and 63–5, and forwards to when Hermes will next leave the cave (false exit at 290* λίκνου; 320* ἐσσυμένως). As he recrosses that threshold and passes from ranging across Greece to being ‘where he belongs’ in his cot, he retains the precocious skills which got him out of the cave before – boundarycrossing which is emphatically marvellous (145–7*) and deceptive walking (149*). And the straining of 148 νηόν* to describe the area of his cot causes a double vision of Hermes as both baby and adult god. Hermes’ un-babylike skills at crossing and walking give way, when he returns to the nursery which is not yet a temple, to a paradoxical show of babyhood – the ability to self-swaddle, and to use swaddling with ulterior motives. Like his crossing of the threshold, this act of manipulation is characteristic, and forms part of the hymn’s focus on how binding demonstrates μῆτις (Introduction §3.3.6). By contrast, the baby Apollo simply burst through the bonds which were restraining him (H.Ap. 128–9; Introduction §1.3). The restrictive or binding nature of swaddling is frequently mentioned: in Pl. Lg. 7 789e, swaddling models the infant like wax; Lycophron 1202 uses the epithet γυιόκολλος; Soranus Gyn. 2.14 uses διαπλάσσειν, and in 2.2 speaks of ἐπίδεσμοι (cf. 157*); swaddling-bands contribute to the baby Heracles’ vulnerability to Hera’s snakes (Pi. fr. 52u.12), and so on. Some authors further attributed an ethical quality to that restrictiveness, in producing conformity (Holman 1997). Hermes makes a resource out of something emblematic of infancy (e.g. A. Ag. 1606, E. Tro. 759; Ael. VH 8.8 ἐν σπαργάνοις) and in particular of its limitations. 203

See, for example, LIMC s.v. Kronos nos. 21–3, Schefold (1981) 31–41, Neils and Oakley (2003) 185, 187, 223.

233

C O M M E N TA RY 1 4 5 – 1 5 4 The particular strip rearranged in 151 goes around the shoulders (cf. 268, 306) and so will make it look as if Hermes’ hands have been tied – though he can only rearrange it because his arms are not tied as one would expect. Next he toys with an outer wrap (152 λαῖφος ἀθύρων*) around his knees, using παλάμαι, which hold a hint of cunning (109–10*). This action seems more compatible with the physics of being swaddled, though still liberated by the standards of the passages cited above. Here Hermes’ precocity is less physical than mental. 152 ἀθύρων picks up the repeated earlier descriptions of the tortoise/lyre as an ἄθυρμα (32, 52; 40 with ἐρατός as in 153), and immediately the lyre hidden inside the swaddling under Hermes’ left arm is mentioned: he is distracting from his revealing plaything by playing at playing with his swaddling (Eitrem 1906: 261). The lyre’s hiding-place (153 ἐπ’ ἀριστερὰ χειρός) is itself rich in resonances.204 Free Greeks could afford to arrange their cloaks so that the left arm, and sometimes the right too, remained idle and wholly covered.205 The armpit (especially the left) thus became a place for concealing – notably weapons, stolen goods, and delicate animals which need sheltering.206 The lyre was briefly a pet, is delicate, and will be a useful defensive ‘weapon’ (cf. 237*, 421–2*) in a case of theft. The lyre’s position may constitute a reversal of its proper place above the left forearm (418*), and is also – finally, and to return us to the issue of baby versus adult – an innuendo, since Hermes is snuggling up in bed with the ἐρατή and previously sexualized lyre (31; cf. 478, 485). In the midst of this, Hermes is compared to, of all things, a τέκνον | νήπιον (151–2). A simile implies that the tenor is not an instance of the ­vehicle, but Hermes is definitely a child: this draws our attention to νήπιος, whose traditional enjambement – often used to inflate then burst suspense – is here given a new point. The adjective covers various aspects of childishness, prominently including lack of mental development in areas such as general knowledge, memory and foresight, emotional attachment to

Literally ‘at the hand [see LSJ s.v. χείρ II.1] on the left’. For the phrase cf. Il. 12.201 ἐπ᾽ ἀριστερὰ λαὸν ἐέργων|, Od. 5.277 ἐπ᾽ ἀριστερὰ χειρὸς ἔχοντα|, Hes. Op. 480 περὶ χειρὸς ἐέργων|, Alcman PMG 84. 205 The himation could cover both arms or the left only; the chlamys, which is associated with Hermes (LIMC V(1).383), often covered the left arm. See Losfeld (1991): 141, 178. 206 Weapons: e.g. X. HG 2.3.23, Pl. Grg. 469d. Stolen goods: e.g. Lys. fr. 197; Alciphr. 3.10.3. Animals: e.g. Pl. Lg. 7 789c, Ar. Byz. Epit.HA 2.439. Klein (1980) 138 commented on the ‘surfeit of symbols of hollowness’: the concave shell concealed in the armpit in the cradle in the cave. 204

234

C O M M E N TA RY 1 5 5 – 1 5 6 the present, and limited perspective.207 Hermes, therefore, looks but is not νήπιος in how he lies in his cradle – still a paradox, since lying in a cradle was brought up in 21* as something typical of all other babies: the intratexts κεῖτο and λίκνον prompt us to recall that line here.208 In some respects, however, Hermes is νήπιος enough: his stratagem is promptly and wrily exposed in 154. μητέρα is placed emphatically, and supported by οὐκ ἄρα ‘certainly did not …’ (Grimm 1962: 20): of course a mother, especially a divine one, would notice the return of her missing infant.209 This is consistent with other deficiencies in Hermes’ infantile trickery (e.g. 243*, 260–77*). Hermes cannot, after all, outdo his father’s secretive nocturnal entries into Maia’s cave (8–9, brought to mind by 154 ἔληθε ~ 9 λήθων, and by 142–5*), by escaping everyone’s notice.

155–61 155 τόδε: ‘to here’: see La Roche (1861) 89–90 for parallels, including Od. 1.409 τόδ’ ἱκάνει. The poet uses πόθεν τόδε in 32 and 269 in the same position. Ω’s τάδε ... ἔρχεαι would mean, if anything, ‘do these acts of coming’, but it is better to take τάδε as a minuscule error. 156 ἔρχε᾽: The orthography, with elided -εαι rather than correpted -ηι, appears sporadically in papyri and manuscripts of Homer. This suggests that the older pronunciation continued at least sometimes in rhapsodic performances of later centuries: Van Leeuwen (1886). For ἀναιδείην ἐπιειμένε see below, 155–61*. The hymnist uses νήπιος also at 164, 210, 406: each time the reference is to Hermes’ (merely) apparent childishness. In Epos as a whole the word’s most prominent use refers to lack of foresight and perspective, suggestively put in Heraclitus 22 B79 ἀνὴρ νήπιος ἤκουσε πρὸς δαίμονος ὅκωσπερ παῖς πρὸς ἀνδρός. General knowledge: Od. 6.301, Hes. Op. 40. Memory: Il. 21.441. Attachment to present: Od. 21.85; cf. shortlived enthusiasm at Il. 20.411. Failure of planning: 16.262, 18.295. νήπιος can also simply describe youth (e.g. Il. 11.561, Od. 1.297) or lack of physical development (e.g. Il. 2.338, 4.238). 208 The paradox will be enhanced by 163–4*, where Hermes asks not to be treated like a τέκνον νήπιον. See Huebner (1986) 162–71, who noted the further connections with 55–6 ἠΰτε κοῦροι | ἡβηταί and 265 οὔτε ... κραταιῶι φωτὶ ἔοικα. 209 For the application of θεά to a nymph see e.g. Larson (2001) 27–33, and for θεὰν θεός cf. Od. 5.97, Hes. Th. 380, 405. The focus on μητέρα makes the change of grammatical subject at εἶπέ τε easily comprehensible; Ilgen needlessly demanded εἶπε δέ. For the possible influence of 154 on Marcus Argentarius AP 5.127 see Vergados 124. 207

235

C O M M E N TA RY 1 5 7 – 6 1 157 ἦ τάχ’: Ψ’s reading is confirmed by S. Ichn. 198 ἦ τάχ’, where the ­chorus, as Maia is doing, imagine the criminal being taken away in chains. For Sophocles’ use of Herm. see Introduction §1.3. M’s unmetrical, unattested δύσαχ᾽ conceivably rests on a misreading of τ as ligatured υσ. 158 Λητοΐδου: For Greek metronyms see 89* (phrases like Μαίης ... υἱός), 261* (vocative metronyms). Λητοΐδης occurs in Herm., Hesiodic fragments, lyric (first Alcman PMG 48), iambus and elegy (including e.g. IG I3 1469: Ptoion, 6th c.), and on Cypselus’ ‘chest’ (Paus. 9.5.18). Ap.S. p. 163 notes the lack of Homeric metronyms in -ίδης, including specifically Λητοΐδης; Allen and Sikes liv inferred that he excluded Herm. from Homeric authorship. It is unclear whether the Attic genitive ending is original. Hermann printed Λητοΐδεω, but cf. Hes. fr. 70.39 Κομή[το]υ̣ (papyrus of c.100 ce), Hy. 16.3 Φλεγύου, Smyth (1894) 342–51. 159 The paradosis, excluding μεταξύ, makes good sense if construed as ‘or that you will be a brigand in the glens, pillaging (φέροντα M)/taking sc. things (λαβόντα Ψ)’.210 ἤ without a preceding comparative naturally means ‘or’, not the much rarer ‘rather than’ (see KG ii.303 n. 2). Either participle should be taken as modifying σε as the subject rather than object of φιλητεύσειν. The alternative construal has three main problems: a bare participle makes an awkward object; verbs in -εύω (in this case, a hapax) are more often ­intransitive; the idea would seem to be that Hermes will rob Apollo while Apollo carries/takes Hermes through glens, but Apollo has no reason to go there, whereas Apollo will also envisage Hermes’ brigandage in glens at 287* ἐν βήσσηισ’. On my preferred interpretation the second σε is redundant, and may convey a similar wonderment to the repeated σέ γ’ in Od. 8.488. The only emendation required is therefore μεταξύ to μέταζε (as at Hes. Op. 394), since Hermes’ brigandage is not envisaged before Apollo’s arrival.211 Of the participles, φέροντα is better in both sense and tense; the scribe of Ψ perhaps misunderstood and wrote λαβόντα as a gloss on the idea of Apollo taking Hermes away. 155–61 Maia’s rebuke: Maia’s only speech characterizes her as forthright and capable of punchy and intelligently playful phrasing. She begins with a spluttering double-question (155 τίπτε σύ, ποικιλομῆτα, πόθεν ...), then adopts a threatening tone as she imagines Hermes’ future punishment (see

LSJ s.v. φέρω VI. λαβόντα would be a ‘coincident’ aorist participle with future governing verb: see Barrett (1964) 214. 211 So first Schmitt (1856) 152, though he also changed φέροντα to κακὸν τά. 210

236

C O M M E N TA RY 1 5 5 – 1 6 1 οἴω with LfgrE s.v. 1aα, the emphasis on νῦν and τάχα, perhaps the assonance ἦ τάχ᾽ ἀμήχ-, and below on διὲκ προθύροιο). She expresses with increasing precision her maternal intuition about what Hermes has been up to. Both halves of 155 ποικιλο-μῆτα suggest cunning.212 νυκτὸς ἐν ὥρηι vehemently simplifies Hermes’ connection to darkness, considering 143 ὄρθριος, and suggests for the external audience the φιλῆται who act νυκτὸς ἐν ὥρηι to whom Hermes was compared in 67; in 158 Λητοΐδου she identifies the wronged party, and in such a way as to contrast her motherhood with the more fortunate state of Leto. διὲκ προθύροιο περήσειν raises threatening (though unfulfilled) expectations of the type-scene where a character is dragged outdoors, but specifically opposes Hermes’ own thematic crossing through the space in front of the cave door.213 ἀναιδείην ἐπιειμένε (‘clothed in shamelessness’, 156) is similarly pointed, since Hermes’ latest act of impudence involved arranging his clothes; the metaphor, found at Il. 1.149 and 9.372 (Achilles describing Agamemnon), may have been formulaic, but in any case one could read the intertext as endowing Maia with the vehemence of that prototypical Greek literary dispute (Introduction §1.3). The ἀμήχανα δεσμά which Apollo will apply to Hermes’ sides (157) cast his self-swaddling by contrast as a self-binding μηχανή (see further 245*, 409*). Whether or not Maia’s intuition was due to versions of the story where she could see that Hermes had brought the cows back to her cave (Eitrem 1906: 264), it is effective here at undermining and challenging his fledgling trickery (cf. 145–54*). The story that Hermes ended up giving the lyra to Apollo as the price of being released (ly-tron) may have been available to Herm.’s audience, as to the audience of Euripides’ Antiope (fr. 190); see Introduction n. 94. Maia ends with a shift towards despair and rejection. 159 κατ’ ἄγκεα φιλητεύσειν presents as unappealing the life of an outlaw in the wilderness, in bitterly punning terms which contrast it with a baby’s proper life of being kissed (φῐλη-) while cradled in its mother’s arms (ἀγκάς, ἐν ἀγκάλαις).214 Maia then dismisses Hermes with ἔρρε πάλιν (which 259* ἐρρήσεις will echo) and shifts the blame for him to Zeus’s bad influence: πατὴρ ἐφύτευσε conveni­ ently sidelines her own role in Hermes’ birth, while also intersecting with the

See respectively Detienne and Vernant (1978) 18–20, and 13* ἀγκυλομήτην. ποικιλομῆτα occurs again at 514, and in this metrical position at Od. 13.293 (Athena to Odysseus), H.Ap. 322 (Hera to Zeus). 213 Compare 271* and Introduction §3.3.6 for the theme of Hermes and ­crossing. See LfgrE s.v. πρόθυρον, esp. B4, for the lack of architectural ­detail about the word’s uses in Epos (‘space in front of door’, not ‘porch’), and the type-scene of being dragged outdoors. 214 Compare the instructive misinterpretation of early translators including Velareus (1528) ut te … in ulnis osculer, Dartona (1537), Certon (1615), Chapman (1624). 212

237

C O M M E N TA RY 1 6 2 Homeric metaphor of ‘sowing’ κακόν/death for someone, i.e. the idea that malicious trickery grows organically from an invisible start.215 Keyssner (1932: 120–4, esp. 122) usefully set μέριμναν | θνητοῖς in the context of the widespread use of the trope ‘ for humans’ in Greek hymnography; Maia assigns to Hermes the sort of abstract noun normally applied to the dedicatee’s opponents (as the Python is a δήλημα βροτοῖσι for Apollo to defeat in H.Ap. 364), so that the hymnist humorously incorporates an ‘antihymnic’ voice.216

162–81 162 τὴν δ’ Ἑρμῆς μύθοισιν ἀμείβετο κερδαλέοισιν: An implementation of a formulaic system for single-line reply-introductions, which can be characterized as based on the following choices: A. |τὸν+ δέ or B. | M. μύθοισιν or N. μιν ἔπεσσιν / μ’ ἐπέεσσιν / σφ’ ἐπέεσσιν X. ¦ἀμειβόμενος+ προσέειπεν+ or Y. ¦ἀμείβετο .217 A and N are mutually exclusive ways of mentioning the addressee. B and Y are nowhere used together to characterize a reply twice over. The remaining possibilities are distributed as follows: AMX Il. x2, Herm. 201*; AMY Il. 3.171, Herm. x3; BMX Od. x4; BNX Od. x8. The hymnist’s focus on AM_ among all possible ways of introducing replies is noticeable (see below), as is the Odyssey-poet’s liking for B_X despite showing familiarity with AMY at 2.83 Τηλέμαχον μύθοισιν ἀμείψασθαι χαλεποῖσιν. AMY, like several other traditional patterns, offers the opportunity to characterize a speech in brief.218 The hymnist uses the pattern three times to describe a progression in Hermes’ κερδαλέοι words. The stem κερδ- relates to gains acquired at another’s expense, primarily through war and commerce; hence the development of κερδαλέος towards ‘cunning’ (Cozzo 1998, esp. 13–15). The shift in Hermes’ κερδαλέοι words is from acquisitiveness See LfgrE s.v. φυτεύω. For fathers ‘planting’, i.e. inseminating, see e.g. Hes. Op. 812, Sc. 29, Ar. V. 1133. 216 Pfeiffer also compared Molorcus’ expostulation to the mice at Call. Aet. fr. 54c.12–14 (his fr. 177), including ξ]εί̣νο̣ι̣ς̣ κωκυμ̣ούς. See Seiler (1997) 93–5. 217 A further variant is H.Dem. 118 ὣς ἔφαθ’, ἣ δ’ ἐπέεσσιν ἀμείβετο πότνα θεάων. For reply-introductions see Edwards (1969), and for the traditional background of speech-introductions in general, Edwards (1970), De Jong (1987) 196–208, Friedrich (2007) 42–3, 60–3, 69–75. 218 Most commonly in Epos such characterizations describe words as joyful, upset or harsh; for κερδαλέος see Od. 6.418. De Jong (1987) 204 surveys the Iliadic material. 215

238

C O M M E N TA RY 1 6 3 – 1 6 5 via cunning to profit (260*, 463*). Here, the adjective lays the ground for Hermes’ first clear statement of ambition for wealth through competition with Apollo. 163–5 μῆτερ ἐμή, τί με ταῦτα δεδίσκεαι, ἠΰτε τέκνον | νήπιον, ὃς μάλα παῦρα μετὰ φρεσὶν αἴσυλα οἶδεν, | ταρβαλέον, καὶ μητρὸς ὑπαιδείδοικεν ἐνιπάς; : Hermes’ words gain meaning through reference to two other passages. ἠΰτε τέκνον νήπιον clearly recalls the primary narrator’s simile in 151– 2*. In 151–2 Hermes looked superficially like an ordinary baby, but was not babyish inasmuch as he was the subtle architect of that similarity; here he claims directly not to be a normal baby, in a way which has two interesting implications for his characterization of himself, Maia and Apollo. First, he reduces Maia’s speech to a mother’s attempt to frighten her children with reference to a grown-up coming to haul them off for naughty behaviour. Greek child-carers used bogeyman figures, so that Apollo is reduced to this demeaning role. See e.g. X. HG 4.4.17 (transposed to a military context), Erinna SH 401.25, Call. H.Art. 62–71 (where Hermes himself plays the Cyclops to scare naughty daughters of goddesses), Strabo 1.2.8 τοῖς τε γὰρ παισὶ προσφέρομεν ... εἰς ἀποτροπὴν τοὺς φοβεροὺς (sc. μύθους)· ἥ τε γὰρ Λάμια μῦθός ἐστι καὶ ἡ Γοργὼ καὶ ὁ Ἐφιάλτης [still the Greek for ‘nightmare’] καὶ ἡ Μορμολύκη, Luc. Philops. 2 ἀλλόκοτα καὶ τεράστια μυθίδια παίδων ψυχὰς κηλεῖν δυνάμενα, ἔτι τὴν Μορμὼ καὶ τὴν Λάμιαν δεδιότων. Secondly, Hermes’ words recall the responses of Aeneas and Hector to Achilles’ threats in Il. 20.200–2 = 431–3 Πηλεΐδη, μὴ δή μ᾽ ἐπέεσσί γε νηπύτιον ὥς | ἔλπεο δειδίξεσθαι, ἐπεὶ σάφα οἶδα καὶ αὐτός | ἠμὲν κερτομίας ἠδ᾽ αἴσυλα μυθήσασθαι (note also 430 οὐ ταρβήσας). Hermes more politely deflects Maia’s ἐνιπή (both ‘threat’ and ‘reproof ’: LfgrE s.v. ἐνίπτω) with a why-question rather than a prohibition, and tackles her unmotherly severity by introducing a neat wordplay on the two senses of δε(ι)δίσκομαι, ‘frighten’ and ‘salute’.219 The logic in Iliad 20 is that only a child lacks the pragmatic competence to (i) construct and hence (ii) see past aggressive bluster. Hermes makes point (ii) in 165. Point (i) is incorporated in 164, but αἴσυλα οἶδεν is vaguer than αἴσυλα μυθήσασθαι: Hermes encompasses the idea that babies, not being inured to criminality, will be put off simply by disapproval such as

In 163, Ω’s τιτύσκεαι cannot govern με, so Bothe changed με to νυ. However, the superiority of Pierson’s δεδίσκεαι is secured by the Iliadic parallel and the pun. δεδίσσεαι, though from the more common verb δειδίσσομαι (as in Il. 20), would not allow the wordplay and is further from the paradosis. See also Stes. fr. 15.6 ]τα δεδίσκ[εο in a reply rejecting a previous threat, Ar. Lys. 564 ἐδεδίσκετο (confirmed by Σvet.: ἐδεδίττετο Maltby).

219

239

C O M M E N TA RY 1 6 6 – 1 6 7 Maia’s. The parallel, whether alluding to Iliad 20 specifically or to formulas traditional in contexts of pre-duel threats (see Fenik 1968 s.v. ‘Threats and taunts’), casts Hermes and Maia as duellists, and so takes the young god’s macho pugnacity to a ridiculous degree (Kassel 1991: 15). For Apollo and Hermes as duellists see e.g. 237 κατέδυνε*. For the variation in gender in τέκνον ... ὅς see GH ii.21. ταρβαλέον occurs only in 165 in Epos; the sense ‘scared’ (rather than ‘scary’) is normal.221 ὑπαιδείδοικεν is based on formulaic ὑπαὶ δείους¦ (Il. 10.376, 15.4, cf. Sc. 71 ὑπαὶ δεινοῖο¦) < ὑπὸ δϝέεος. The metre-preserving ending of ὑπαί was borrowed from παραί (for which see n. 89); so e.g. Nic. Th. 178 ὑπαιφοινίσσεται, D.P. 151 ὑπαινοτίη. 220

166 τέχνης ἐπιβήσομαι: For τέχνη and its combination of trickery and practical skills (we have seen lyre-making, herding, sandal-making, firecraft and butchery) see 76*. Here the preceding contrast τέκνον νήπιον ... αὐτὰρ ἐγώ shows that in some sense Hermes’ τέχνη will be αἴσυλα, i.e. offensive to another god (for this nuance see Il. 5.403, 876 if ἀήσυλα is the same word, 21.214; Od. 2.232 ~ 5.10 of chieftains abusing their god-given privileges). But Hermes’ proud declaration reappropriates the terms of Maia’s (supposedly negative) characterization of him at 155 ποικιλομῆτα. For the metaphor in ἐπιβαίνω see 173, 465*, Hes. Op. 659 ἐπέβησαν ἀοιδῆς, Od. 22.424 ἀναιδείης ἐπέβησαν. 167–8 βουκολέων ἐμὲ καὶ σὲ διαμπερές· οὐδὲ θεοῖσιν | νῶϊ μετ’ ἀθανάτοισιν: Ludwich’s excellent conjecture βουκολέων continues both nuances of τέχνη. Hermes will ‘tend’ his and Maia’s interests (for this general use see e.g. Ar. V. 10), through the ‘herding of cows’, to which he has just staked a claim, but also by ‘deceiving’. For βουκολία and deception see Men. Sam. 530 μή με βουκολεῖς ὅρα, Aesop 28 οἱ δὲ θεοὶ βουλόμενοι αὐτὸν ἐν μέρει ἀντιβουκολῆσαι, Hld. Aith. 8.7.3, and compare LSJ s.v. ποιμαίνω II.3, as well as 542*. The deceptive association is activated as pertinent to Hermes’ exploits, even though on a grammatical level ἐμὲ καὶ σέ are unsuitable objects for it. The paradosis would have to be articulated τέχνης ἐπιβήσομαι (ἥ τις ἀρίστη | βουλεύων) ἐμὲ καὶ σέ ..., but the parenthesis and the implication that Maia will learn a τέχνη are very

Cf. 166*. In 164 M offers the variant πολλὰ ἐνὶ φρεσὶν ἄρμενα. Even if one extracts from ἄρμενα ‘fitting a child’ not ‘fitting’, this is banal and hard to believe in as an ancient variant. ἐνί (creating hiatus) could derive from a gloss on μετά + dat. (cf. 259), but the extent of the corruption suggests that one of M’s ancestors had suffered a good deal of fading in this phrase. 221 In Hsch. τ.185 ταρβαλέον: δεινόν, φοβερόν, read δειλόν? 220

240

C O M M E N TA RY 1 6 8 – 1 6 9 awkward. For the corruption Allen and Sikes compared PLond. lit. 22 at Il. 14.445 (βουλέοντι ss. κο). The emphatic pronouns ἐμὲ καὶ σέ and νῶϊ reject Maia’s attempt in 159– 61 to distance herself from Hermes’ activities: he has both their interests at heart. Hermes contrasts the pair with the group of immortal gods more broadly, a strategy discussed further on 170*. 168 ἀδώρητοι καὶ ἄπαστοι: Ψ had ἄπαστοι ss. λι, i.e. the interlinear variant ἄλιστοι. ἀδώρητοι expresses isolation from human offerings, and perhaps also the bestowal of prerogatives by senior gods (Jaillard 2007: 90); ἄπαστοι implies isolation from sacrifices. Against ἄλιστοι, the classical word is spelled ἄλλιστος and is an epithet of Hades meaning ‘inexorable’ and hence not suited even to Hermes’ strategy of casting their current state as Hades-like (172 ἠερόεντι*).222 ἄπαστοι is exaggerated (of course Maia has food: cf. 248), but the better for it. Hermes uses another absurd hyperbole in 309*. Here it suits his highly tendentious description of their situation: life in the cave is no better than that of humans in mourning (cf. Il. 19.346 for Achilles ἄκμηνος καὶ ἄπαστος|). It extends the theme of Hermes’ humorously human hunger (64, 130–1), and suggests a justification for the cattle-theft in terms of his current hunger (see Haft 1996: 30–1, with a modern Cretan parallel). The hyperbole itself could well have prompted a scribe to conjecture ἄλιστοι in the Byzantine period, when that spelling is found. Herm.’s phrase may be one of the sources for Lyc. Al. 140 ἄσιτα κἀδώρητα φορμίζων μέλη (of Paris; Hermes φορμίζει too). 169 αὐτοῦ τῆιδε μένοντες: An idiom: as well as Cert. 269 |αὐτοῦ τῆιδε μένουσα, there are close parallels in Hdt. 5.19.1, 7.141.2, 9.11.1 (of which the last has the same derogatory tone as here), and later Hld. Aith. 6.6.2. None seems to derive from our passage. ἀνεξόμεθ᾽, ὡς σὺ κελεύεις: Elsewhere the formulaic system comprising ὥς σε κελεύω|, ὥς με κελεύεις|, and ὡς/εἰ σὺ κελεύεις| is used positively: the advisee will act as advised.223 Here, the emphasis on σύ is contemptuous, and κελεύεις is (as Ilgen observed) hyperbole: Maia at most implied in 155–6 that Hermes should not be out and about at night. Hermes’ ‘disobedience’ towards traditional language may enhance our sense of his dislike of the status quo. 222

Euphorion fr. 103.4, Crinagoras 19.3 G–P, IGUR iii.290 (first century ce). Cf. 573* ἄδοτος. For the double lambda normal in derivatives of λιτή see GH i.176; πολύλιστον is transmitted for Simonides (fr. 264b) in Plutarch, but the manuscripts there are badly corrupt. 223 The poet may have recalled in particular Il. 8.35 (Athena to Zeus ~ 446 Hera to Zeus) πολέμου μὲν ἀφεξόμεθ᾽, ὡς σὺ κελεύεις.

241

C O M M E N TA RY 1 7 0 – 1 7 2 170 βέλτερον: Significantly engaging with its last use in 36*: there, Hermes amusingly advocated life in the household; here, he reveals his true colours (Fernández Delgado 1990: 208). ἤματα πάντα μετ᾽ ἀθανάτοισ’: Hermes reinforces from 167–8 both the desire for permanent status (cf. διαμπερές and 292*) and his obsession with broader divine society; the close repetition of μετ᾽ ἀθανάτοισ(ι) more particularly establishes such phrases as a motif which tracks Hermes’ success in rising to the Olympian stage. Cf., with comments, 16, 270, 291, 366, 458, 461, 525, 576, similarly 332 θεῶν μεθ’ ὁμήγυριν, and contrastingly 259 ὀλοοῖσι μετ᾽ ἀνδράσιν, 468 of Apollo. ὀαρίζειν: For the dilution from ‘woo’ to ‘chat to, be on close terms with’ see e.g. Od. 19.179 Μίνως... Διὸς μεγάλου ὀαριστής, DÉLG 771. The fact that Hermes previously praised Maia for her ὀαριστύς with Zeus in the cave (58) points up the shallowness of his rhetoric: see further on 172, 179–81. 171 πλούσιον, ἀφνειόν, πολυλήϊον: Overlapping terms are piled up ­rhetorically, as in Hesiod Op. 637 οὐκ ἄφενος φεύγων οὐδὲ πλοῦτόν τε καὶ ὄλβον. For Hermes’ materialism see 24*. Though πολυλήϊος often relates to λήϊον ‘corn(field)’ (e.g. Hes. fr. 240.1, Arat. 1058), given his cattle-rustling Hermes is evidently thinking more of acquiring wealth from ληΐς ‘booty’; cf. 335*. Ancient scholarship varies between the two derivations, and both may be relevant at A.R. 1.51 πολυλήϊοι Ἑρμείαο | υἱέες εὖ δεδαῶτε δόλους. Hesychius’ curt gloss π.2880 πολυλήϊον: πλούσιον is more probably for our passage than for the other extant use of that form (Nonnus D. 11.395), which relates specifically to corn. 172 ἄντρωι ἐν ἠερόεντι θαασσέμεν: Hermes focalizes the cave as an opposite of Olympus. ἠερόεις, which plays off the weaker but apter ἠεροειδής, is strongly linked to the Underworld in Epos, and qualifies Tartarus in 256.224 It is also used of Maia’s cave in Apollo’s negative focalizations at 234, 359, though elsewhere the cave is shown to be well appointed, and was described as such by Hermes himself at 60–1.225 For the regressive connotations of

The Underworld association of ἠερόεις is also clear in Orph.H. 69.4 ἄντρωι ἐν ἠερόεντι παρὰ Στυγὸς ἱερὸν ὕδωρ, which may derive from our line. However, Telestes PMG 805c (with Bergk’s ἀερόεν) and Hellenistic literature do not retain the nuance. The basic sense of ἀήρ is ‘cloud’ rather than ‘air’, whence the shift towards murkiness. For ἠεροειδής of a pleasant cave see Od. 13.103. 225 See also n. 32. The fact that the cave can be at one time dank and at a­ nother rich confused scholars until narratology brought the tools for thinking it through clearly. Bothe on v.148 already understood Hermes’ aversion to the cave here (‘cui sordet domus materna, honorum divinorum cupido’), but assumed he should display this 224

242

C O M M E N TA RY 1 7 3 cave-dwelling which Hermes mobilizes, see 6*. θαάσσω and its cognate θῶκος (< θοϝακ-) simultaneously connote sitting ceremonially, for example in a meeting on Olympus (468, Il. 15.124), and sitting idly (e.g. Hes. Op. 493, cf. LfgrE s.v. ἧμαι I.2, II.4b). Hermes thus sharpens the negative view of the cave as enforcing inactivity, by deploying as a simultaneous foil the ‘suppressed’ sense of the θῶκοι, where the gods sit in council or feast with full honours. ἀμφὶ δὲ τιμῆς: For ἀμφί + gen. ‘in a struggle for’ see e.g. Il. 16.824–5, μάχεσθον | πίδακος ἄμφ᾽, Sc. 402, Xnph. fr. 1.20, Pi. P. 9.105 ‘they went ἀμφὶ γυναικός’ (for a suitor-contest), 4.276 ‘devote all your efforts ἀμφὶ Κυράνας’, A. Ag. 62 ‘Zeus sends an Erinys [~ the Greek expedition to Troy] ἀμφὶ γυναικός’. This usage suits Hermes’ competitive attitude towards Apollo. By contrast, ‘Concerning τιμή, …’ would be blander and more loosely integrated into the syntax.226 τιμή here refers inseparably to status in the eyes of Olympians, to enduring spheres of control (τιμαί in the plural), and to the honorific worship which humans will pay Hermes as a consequence of them. The importance of the concept to Hermes is emphasized by the way he fronts it and separates it from the rest of the sentence by enjambement. 173 κἀγὼ τῆς ὁσίης ἐπιβήσομαι ἧς περ Ἀπόλλων: Despite the likely influence of Zeus’s original promise about divine τιμαί at Hes. Th. 396 τιμῆς καὶ γεράων ἐπιβησέμεν, here ὁσίη must be a noun rather than an adjective with τιμή understood. Otherwise, the phrase ὁσίη τιμή would lack precise parallels, and the word-order would be contorted. The idea that Apollo’s ὁσίη is logically prior to the acquisition of τιμή will be repeated at 468–71*. In both passages Hermes seems to use ὁσίη in a very human way, to mean behaviour which will not offend any gods. He portrays this as a precondition for the grant of prerogatives and cult honours; here he contrasts this course of ‘proper behaviour’ with the crimes he will perpetrate if excluded from Olympus (175–81).227 Even if the use of ὁσίη suggests, as in 130*, that aversion consistently. Radermacher explained Hermes’ inconsistency in terms of changing mood and rhetoric, but did not address Apollo’s perspective(s). Vergados (2011) overshot by arguing that all the information we get about the cave is equally partial. Rather, the rich reality is entailed in the narrative at e.g. 248–50, and contrasts with the negative focalizations by Hermes and Apollo which are occasional and tendentious in a readily interpretable way. 226 For ἀμφί + gen. ‘concerning’ see e.g. Od. 8.267, Emped. 31 B23.2, Pi. O. 12.8, S. Ph. 554–5, E. Hec. 75, Or. 756. In each case ἀμφί is governed directly by a verb, which would not be true here of ἀμφὶ τιμῆς ... ἐπιβήσομαι. Radermacher compared ἀμφί + dative introducing a new topic at Il. 7.408. 227 For the overall metaphor compare Od. 23.13 σαοφροσύνης ἐπέβησαν. Clay’s interpretation of ὁσίη here (close to ‘ritual’) causes trouble in both the passages it purports to explain: she had to suppose that Hermes was temporarily mistaken in thinking ὁσίη

243

C O M M E N TA RY 1 7 4 – 1 7 5 Hermes is not yet fully inhabiting the mentality of a god, the marked echo of 166 (ἐγώ, –ης ἐπιβήσομαι, ἥ τις / ἧς περ ⏑– –|) shows that he is restating his ­aspirations in a more serious manner than 166–72: the ambivalent words τέχνης and βουκολέων have been replaced by ὁσίη, the amusingly mundane focus on material wealth with the hunt for τιμή. The emphasis in κἀγώ and τῆς … ἧς περ Ἀπόλλων tackles Maia’s subtext in 158* that Apollo is a better son. Crasis of καί is rare in Homer (Il. 2.238 χἠμεῖς, maybe 6.260 καὐτός); cf. κἀμέ at Archil. fr. 24.16, κἀγώ in SEG 49.1049.2 (c.530 bce Black Sea Ionic), Theognidea 1349. 174 εἰ δέ κε μὴ δώηισι πατὴρ ἐμός: For the gods’ τιμαί as Zeus’s ‘gift’, see e.g. 468, Hesiod Th. 399, 412. Hermes’ reference to Zeus as ‘my father’ calls to mind that for him Zeus is not only the king who dispenses status among the gods, but also the father who has not yet publicly recognized his illegitimate son (see 274, 331, 334–64, 378 for this issue). The structure εἰ δέ κε μὴ δώηισι ... ἐγώ occurs also in Il. 1.324–5 where, as Schenck zu Schweinsberg observed, it did not end well for Agamemnon. For other undertones of Iliadic or at least macho confrontation in Maia’s conversation with Hermes see Introduction §1.3. ἤτοι ἐγώ γε: For the orthography (not ἦ τοι or ἔγωγε) see West (1998) xviii. 175 (δύναμαι): This articulation, found by Bothe, gives an unusually elliptical parenthesis, which characterizes Hermes’ exuberance. Single verbs in parentheses tend to be a means of avoiding indirect discourse (e.g. οἶμαι, ὁρᾶις). Explanatory parentheses tend to be slightly longer, including formulaic δύναμαι+ γάρ| at e.g. Od. 4.612, 827, 5.25, 16.208, A. Ch. 374.228 Hence Ω’s δύναμαι δέ arose, but then φῑλήτης does not scan. φιλητέων ὄρχαμος εἶναι: A variation on the formula (x29) ὄρχαμος+ ἀνδρῶν| / ὄρχαμε λαῶν|. Despite a few uses in the Odyssey as a general honorific, that formula’s dominant military tone of ‘commanding’ (LfgrE s.v. ὄρχαμος, Bonnafé 1984) makes for an unusual feature in a divine epithet, markedly repeated in 292 ἀρχὸς φιλητέων*.229 Hermes’ phrasing thus

logically prior to τιμή, and to emend 470 (1989: 129–31). Jaillard (2007) 88 made ὁσίη sc. τιμή (roughly ‘proper usage of one’s divine prerogatives’) posterior to τιμή, as did Peels (2016) 243–4, for whom ὁσίη refers to the ritual obligations which humans will have to honour in the future. These readings give insufficient weight to 172 ἀμφί and 470 ἐκ. Jay-Robert (1999) 8–15 interpreted ὁσίη here as a right, conferred by Zeus on other gods, to receive honours. 228 See KG ii.353–4, Schwyzer (1939), Schwyzer ii.706. 229 For more general metaphors for Hermes’ patronage of φιλῆται see 67*, and for his patronage as ‘leading’ see 259.

244

C O M M E N TA RY 1 7 6 – 1 7 8 c­onjures  a stark vision of militant organized crime in response to Maia’s warning that being a φιλήτης will not be pleasant (159). But although his cattle-theft has proved that Hermes is worth keeping on side, from the audience’s perspective the baby’s ambitions and self-confidence are faintly ridiculous in the face of Zeus’s regime; in this sense, Eitrem’s interpretation of the formulaic modification as ‘parodic’ (1906: 261) is valuable. 176 ἐρευνήσει: The stem embraces various kinds of searching: particularly pertinent here are the tracking of a hunter (e.g. Il. 18.321, Od. 19.436, Emped. 31 B101.17, E. Hec. 1172) and the search – especially a house-search – of a law-enforcer: X. Cyr. 1.2.12, IK Mylasa i.3.9 (355/4), and specifically of searching for property at Hdt. 5.92δ.1, Antiphon 5.29, X. Lac. Pol. 7.6, Pl. Leg. 12.954b, Theopomp. FGrH 115 F330, IG IV2.1 123.19 (Epidaurus, fourth century). Λητοῦς ἐρικυδέος υἱός: Cf. 89* Μαίης ἐρικυδέος υἱός. A performer may have enlivened the formulaic ἐρικυδέος with some sarcasm, to express Hermes’ contempt for the idea that Leto and Apollo are superior to him and Maia (see 155–61* on Maia’s use of Λητοΐδου in 158). 177 ἄλλό τί οἱ: For the accentuation see West (1990) xxxii.230 καὶ μέζον: In favour of the Ionic spellings μέζων, κρέσσων, etc. see West (1998) xx. Though they have only a slender presence in Homeric papyri and ancient scholarship (e.g. PHib. 23 at Od. 20.41), it is more likely that Attic/koine μείζων, κρείσσων, etc. (where ει is based on analogy with e.g. χείρων) replaced Ionic forms during transmission than that rhapsodes used them originally; see also GH i.256. ὀΐομαι ἀντιβολήσειν: Hermes uses ὀΐομαι to throw Maia’s understated threatening tone at 156 ὀΐω back at her (155–61*). For the link between ἀντιβολήσειν and Hermes’ future control of chance encounters see 26*; here the word’s more antagonistic uses of battle-encounters (e.g. Il. 7.113– 14, 12.465) are also relevant, and the rhyme with 178 ἀντιτορήσων| can be construed as a feature of aggressive flyting (compare e.g. Il. 6.143 ἄσσον … θάσσον). 178 εἶμι γὰρ ἐς Πυθῶνα μέγαν δόμον ἀντιτορήσων: Of the buildings at Delphi, Apollo’s μέγας δόμος must refer primarily to the main temple, but the phrasing allows a serious act of ἱεροσυλία to be simultaneously coloured with the techniques of private burglary.231 One category of the latter was So here at least in MVA. Ω had οὔτέ in 224, 444, 445, 534, μήτέ in 276; cf. 229 ἔνθά τε Θβ. 231 For temples as divine houses see 148*. 230

245

C O M M E N TA RY 1 7 9 – 1 8 1 called τοιχωρυχία, boring through a mud-brick wall (see 283–4*, Il. 10.267 Autolycus πυκινὸν δόμον ἀντετορήσας; Cohen 1983: 72–3, 78–9). Hermes can extend it to the stone walls guarding the temple treasures with his preternatural skill at crossing boundaries (Introduction §3.3.6).232 Here and at 523, the Delphi of the legendary past is imagined as having one main building and (in what follows) negligible outdoor dedications; H.Ap. similarly speaks only of ‘the stone temple’ (294–9, 523), against what an ancient audience would have known about the sanctuary in their times – namely multiple buildings and dedications, and between 548/7 and c.511 an incomplete temple.233 Πυθῶνα: Visser (1997: 382) provides parallels for this form as opposed to Πυθώ. 179–81 ἔνθεν ἅλις τρίποδας περικαλλέας ἠδὲ λέβητας | πορθήσω καὶ χρυσόν, ἅλις τ’ αἴθωνα σίδηρον | καὶ πολλὴν ἐσθῆτα. σὺ δ’ ὄψεαι, αἴ κ’ ἐθέληισθα: The items mentioned are typically combined in epic descriptions of κειμήλια and gifts or prizes, and ἅλις (here ‘plentiful’, not ‘sufficient’) is frequent in such situations too; the phrase τρίποδας περικαλλέας ἠδὲ λέβητας in particular occurs at Od. 13.217, alongside gold and clothes among Odysseus’ gifts from the Phaeacians. Hermes praised Maia’s cave for containing tripods and cauldrons in a similar connection at 60–1*, and we will see her storerooms at 248–51. The intratext sets up a comparison between Hermes ransacking Delphi and Apollo ransacking the cave (see 246–7*), but it also marks how manipulable Hermes’ focalization of the cave is – earlier praised in terms of rich sanctuaries, here denigrated in contrast to one. These Delphic items allude more particularly to the types of human ­dedication which make sense as divine versions of κειμήλια, while excluding certain other types (e.g. marble or ceramic statuettes) which do not. Monumental tripods were especially favoured at Delphi in the first half of the fifth century, the adyton contained ‘the’ tripod, where the Pythia sat, and at least by the fourth century, historians appreciated that tripods had been a popular early dedication. Heracles tried to rob Apollo of the famous tripod, and Hermes’ threat here therefore enhances his similarities to Heracles

232

Compare Hdt. 2.121 for another story about the cunning required to get into a stone treasury. I assume that an audience would envisage a stone temple first; there were also stories about prehistoric Delphic temples of laurel, wax and feather, and bronze (Paus. 10.5.9–12). For the derivation of ἀντιτορέω, see 41–2*; ἀντι- can be taken as ‘in opposition’. 233 See Scott (2010) figs. 3.1, 3.4, 4.1 for the spread of buildings and monuments. For the destruction of the earliest known temple see Paus. 10.5.13, and for the ‘Alcmaeonid’ rebuilding see Hdt. 2.180, 5.62, Ps.-Arist Ath. Pol. 19.3–4, Forrest (1969) 284–6.

246

C O M M E N TA RY 1 8 1 (Introduction §2.2). Delphi was notorious for gold dedications such as Croesus’ spectacular gifts, while bronze was typical of statuary. Among iron valuables we know of Glaukos’ tripod (Hdt. 1.25.2), Teisagoras’ iron statue of Heracles and the hydra (date uncertain; Paus. 10.18.6), some lances, and spits such as those of Rhodopis. The dedication of textiles at Delphi is assumed at E. Ion 327, 1143–5, and attested more directly at other sanctuaries.234 In 180, for πορθήσω + accusative of the objects removed, rather than the place they are removed from, see the corresponding passive usage at Aeschylus Su. 443 χρήμασιν ... πορθουμένοις, Eupolis fr. 162 τἀργύρια πορθεῖται, and LSJ s.v. πέρθω II. The application of αἴθωνα to iron, while traditional, stands apart from the other early applications of αἴθων to things of fiery colour (bronze, sources of fire, various animals) and/or temperament (hunger, humans); LfgrE s.v. αἴθοπ- plausibly suggests that αἴθωνι σιδήρωι+| arose as a modification of αἴθοπι χαλκῶι+|.235 181 σὺ δ᾽ ὄψεαι, αἴ κ᾽ ἐθέληισθα: As often, the formula barely conceals a threat under the guise of a polite offer; cf. Il. 8.471, 9.359, Kelly (2007) 343–4. Parenthetic εἰ + ἐθέλειν with the infinitive left implicit is a common idiom in Epos (cf. 274; LfgrE s.v. ἐθέλω 1eβdd, 1eγ). The subjunctive ἐθέληισθα suggests that the speaker does not presume to know the addressee’s will – a strategy of negative politeness which may be genuine, purely conventional, or ironic as here, where Maia has made her wishes plain.236 162–81 Rhetoric and motivation: Hermes’ response to Maia is as rich in characterization as his address to the tortoise, and affords us a first glimpse of his debating style. His prime tactics are humour and pugnacity. The 234

On tripods see Ath. 6 231e–2d, RE v.1678–89, Jacquemin (1999) 175–8. Gold and bronze at Delphi: Jacquemin (1999) 164–8. Early dedications there: Amandry (1991), Rolley and Chamoux (1991), Maass (1993) 128–44. Lances and spits: Hdt. 2.135.4 (Rhodopis), Strøm (1992). For dedication of clothing see ThesCRA i.296–7. Hollinshead (1999: 190 n. 6) gives useful orientation in the scholarship on temple inventories. 235 Beekes (1996) 16 argued instead that αἴθων fundamentally meant ‘shining’. This hardly suits an eagle (Il. 15.690) or hunger (Hes. Op. 373 αἴθοψ, fr. 43a αἴθων; cf. Sc. 265 αὐσταλέη λιμῶι; for the connections between fire and hunger in Hesiod’s account of Prometheus see Vernant 1980: 178–81), or later developments such as the applications to a fox (Pi. O. 11.19), a boar (B. 5.71) or human temperaments (A. Th. 448, counterpart to the thunderbolt of 444–5; S. Aj. 223). 236 Politeness strategy: Brown and Levinson (1987) 144, 162–3. Genuine uncertainty: Il. 9.427–9 Φοῖνιξ δ’ αὖθι… κατακοιμηθήτω, | ὄφρά μοι … ἕπηται | αὔριον, ἢν ἐθέληισιν· ἀνάγκηι δ’ οὔ τί μιν ἄξω, 18.457. Conventional politeness: e.g. Il. 10.55, Od. 4.391, 21.233, 24.511.

247

C O M M E N TA RY 1 8 2 – 1 8 3 wordplay on θαασσέμεν (172) underlines his aggressively overdrawn contrast of life in the cave and life on Olympus. The wordplay on βουκολέων (167) reappropriates Maia’s negative portrayal of Hermes’ trickery as something positive, and positive – whatever she may say – for both of them. ἐρικυδέος (176) may add a touch of sarcasm to Hermes’ rejection of Maia’s doubts that they can be as fortunate as Apollo and Leto. The wordplay on δεδίσκεαι (163) criticizes Maia’s scare-tactics as an inappropriate way for a mother to greet her baby, and wittily reduces Apollo to the status of a bogeyman. But that does not stop Hermes using scare-tactics himself, parroting Maia’s use of ὀΐω for threatening understatement (177), and ironically paying lip-service to her wishes in the closing clause. Hermes’ pugnacity is so overblown as to be risible. His characterization of the cave contradicts his own praise of it the last time he spoke (60–1). Intertexts to Il. 20 and perhaps Il. 1 in lines 163–5 and 174 challenge us to compare Maia to Achilles; 175 ὄρχαμος and 177 ἀντιβολήσειν continue the militaristic language between Hermes’ syntactically exuberant but improbable claim that he can create chaos against Zeus’s wishes (175), and his Heraclean threat to ransack the temple at Delphi (178–81). As at 152, the fact that Hermes is not a normal νήπιος infant does not prevent his behaviour from retaining a juvenile spirit. Behind this rhetoric, the speech is crucial in revealing Hermes’ deeper motives. At first his search for Apollo’s cows seemed unmotivated and whimsical, then a matter simply of eating, then a means of symbolizing his ­aspiration to divine status. Here he first presents his motivation to join Olympus in light-heartedly mundane terms of getting more wealth and food, but then more seriously as a matter of equalling Apollo in τιμή. It is perhaps only now that we catch up with his motivation in seeking the cows so abruptly at 20, or understand what kind of ὄνησις he foresaw from the tortoise as a bargaining chip (30, 34–5). For the hymnist’s strategy of revealing Hermes’ mind gradually, see Introduction §3.4.

182–9 182 ὣς οἳ μέν ῥ’ ἐπέεσσι πρὸς ἀλλήλους ἀγόρευον: For the pleonasm ‘address with words’ compare Od. 10.34 οἱ δ᾽ ἕταροι ἐπέεσσι πρὸς ἀλλήλους ἀγόρευον, and with different phrasing e.g. Il. 5.30, Od. 12.36. The normal formula (Il. x8, Od. x16) replaces ῥ᾽ ἐπέεσσι in our line with τοιαῦτα. 183 αἰγιόχοιο: The epithet was understood to relate to Zeus’s aegis (see Il. 5.739–42 αἰγίδα … αἰγιόχοιο), a shield-like object formed from a goat-skin, which when shaken caused terror. Greeks mostly related -οχος to ἔχω (see Sc. 443–4 κούρη Διὸς αἰγιόχοιο … αἰγίδ’ ἔχουσα) or to shaking, on the basis of either Poseidon Γαιήοχος (see 187*), or a stem *ueg΄h- (Meillet 1924: 253).

248

C O M M E N TA RY 1 8 4 – 1 8 6 Etymologically, however, the latter basically means ‘transport’, and Zeus, like other Indo-European thunder-gods, travels by goat in Orph. fr. 236(ii); see West (1983) 133, (1978) 366–8. Μαῖα: Ψ: μήτηρ M. Either reading could generate the other (Ψ>M: subconscious slip towards a common formula; M>Ψ: gloss). The line with μήτηρ, however, risks jarring if an audience-member momentarily understood ‘The son and mother of Zeus’. 184–5 See discussion in 182–9* below. 186 Ὀγχηστόνδ’ ἀφίκανε κιὼν πολυήρατον ἄλσος: Word-order does not permit us to construe Ὀγχηστόνδε κιών and ἄλσος ἀφίκανε separately, so ἄλσος is in apposition to Ὀγχηστόν, denoting the whole place by synecdoche (compare e.g. Pi. O. 8.9, N. 2.5). After -δε one expects e.g. ἐς ἄλσος, but the anomaly is easily explicable since the hymnist was modifying Ὀγχηστὸν … Ποσιδήϊον ἀγλαὸν ἄλσος (Il. 2.506, H.Ap. 230). ἀφίκανε κιών will prove structural (Introduction §3.2). The phrase is an instance of an epic idiom where ‘coming’ or ‘going’ modifies a verb indicating the end-point of that motion.237 Onchestos’ grove and sanctuary of Poseidon were not mentioned in 87–94, and their presence now might suggest that Apollo’s perception of space is framed by human cult in a way in which Hermes’ is not, as yet. Indeed, Apollo ‘owned’ far more sacred groves than any other god. Such groves had practical value for a sanctuary (as a source of firewood, building material, or timber to sell), but were also perceived to offer pleasant shade and water: here πολυήρατον was chosen over Ποσιδήϊον, whose content was then elaborated in 187.238 The retrojection of Onchestos into the mythical period enhances its venerability.239 Groves are easier to retroject than the sanctuary architecture,

LfgrE s.vv. εἶμι I.5, ἐλθεῖν I.3, κίε I.1c. Il. 17.129 ἂψ … ἰὼν ἀνεχάζεθ’ is ­unusually pleonastic. Dietrich (1973) 214–15 argued that in later examples the participles became grammaticalized and underscored the completion of an action rather than any movement (cf. ‘I went and overslept’). 238 For groves see Birge (1982), and for their aesthetics Jacob (1993) 40–4, Dillon (1997) 114. Call. H.Dem. 29 θεὰ δ’ ἐπεμαίνετο χώρωι is an extreme example. Many ἄλση in Pausanias (e.g. 2.13.3, 29.1) and Strabo are tourist attractions. For πολυήρατος of places see Od. 11.275, Solon fr. 4.21, Bacchylides 19.9, Ar. Nu. 301, Hdt. 4.159.3. For regulations on using wood from sacred groves see e.g. LSCG 111.2–4 (Paros, fifth century), 148 (Gortyn, third century), 84 (Korope, c.100), IG XII.4 283–4 (Cos, c.400), CGRN 99 A.8–10 (Cyrene, fourth century), SEG 22.508A.16 (Chios, fourth century), Paus. 5.13.3 (Olympia); Lupu (2005) 169, Jordan and Perlin (1984) 157. 239 Several other sources describe a very early foundation of the sanctuary: Hes. fr. 219, H.Ap. 225–30, Hdn. Gr. i.338, Σ A.R. 3.1242. 237

249

C O M M E N TA RY 1 8 7 since ‘The shade-giving tree epitomizes both beauty and continuity across the generations’ (Burkert 1985: 85; cf. Birge 1982: 192–5). 187 ἁγνόν: More ‘venerable’ (cf. ἅζομαι) than ‘pure’. For the adjective’s frequent postposition in early poetry see e.g. the formula Ἄρτεμις ἁγνή|, Od. 21.258–9 ἑορτὴ τοῖο θεοῖο | ἁγνή, Sappho fr. 2.1–2 ναῦον | ἀγνόν, A. Pe. 628 χθόνιοι δαίμονες ἁγνοί. Here it results in an unusual word-order: adjective, noun, adjective, possessive genitive. ἐρισφαράγου Γαιηόχου: Γαιήοχος is almost always applied to Poseidon (or, later, water), in both literature and cultic contexts, to the extent that (as with e.g. ἑκηβόλος or ἐννοσίγαιος) a proper name is not needed.240 The sense ‘earth-shaking’ seems to be implied at Hy. 22.2–6 (where γαιήοχε follows γαίης κινητῆρα and ἐννοσίγαιε) and Ps.-Plutarch De Homero 107, and here it suits ἐρισφαράγου (cf. Hes. Th. 706, where winds σὺν … ἐσφαράγιζον earthquakes along with dust, thunder and lightning); Poseidon’s control of earthquakes was certainly revered at Onchestos.241 The epithet’s exceptional applications to Zeus (A. Su. 816) and Artemis (S. OR 160) are based on γηοχέω ‘possess land’, but the digamma in IG V.1 213 γαιᾱϝοχος shows that the epithet was not originally related to ἔχω. Ancient scholars also offer ‘holding the Earth steady’, ‘holding in the Earth’, ‘riding [sexually?] on the Earth’ and ‘rejoicing [γαίων] in chariots’.242 187–8 ἔνθα γέροντα | κνώδαλον ηὗρε νέμοντα παρὲξ ὁδοῦ, ἐκτὸς ἀλωῆς: The juxtaposition might fleetingly suggest that the man is a tenant of sanctuary land, which was frequently rented out.243 In 188, the transmitted κνώδαλον εὗρε νέμοντα παρὲξ ὁδοῦ ἕρκος ἀλωῆς is a notorious crux. Suppose κνώδαλον is the object of νέμοντα: then since 240

For the cultic use see X. HG 6.5.30, IG II2 3538 (first century ce), Paus. 3.21.8. Epithets shared by Epos and cult are notably rare: Dionysus κισσοκόμης, Zeus ἑρκεῖος; less remarkably Zeus μέγιστος, ὕπατος, Dionysus βακχεῖος, Demeter ἁγνή, Kore ἁγνή. See Preller and Robert (1894) index III. 241 σφαραγέω = ‘cause to emit a noise’ (generally by violence, though contrast Pindar fr. 140a.60 λ̣[ι]γ̣υσφα̣ρ̣ά̣γων̣ lyres) or less often ‘make to swell’ (Od. 9.440, Nic. Ther. 553). At Od. 9.390, σφαραγεῦντο combines a reference to distension with the noise-based simile which follows it; attraction to σμαραγέω is visible at Hes. Th. 679, 693, 706. For Poseidon and earthquakes at Onchestos see SEG 30.442 β̣ιασσθενῆο | Ποτιδάονος, Pi. I. 1.52, (3/4).37 ὁ κινητὴρ δὲ γᾶς Ὀγχηστὸν οἰκέων, Spyropoulos (1973a) 270. 242 For ‘holding the Earth steady’ see Apollod. Hist. FGrH 244 F96 (> e.g. Corn. ND pp. 42–3, Plu. Thes. 36.6; garbled in ΣGe Il. 21.446–9). For the other senses see Ap.S. p. 54, ΣD Il. 23.584. 243 Birge (1982) 180–2, Walbank (1983), Carroll-Spillecke (1989) table B.

250

C O M M E N TA RY 1 8 7 – 1 8 8 νέμω + double accusative (‘pasturing his beast on the hedge …’) is not found, ἕρκος ἀλωῆς is metaphorical (‘the safeguard of his plot’) and apposed to γέροντα or κνώδαλον. Either way it is confusing, since ἀλωαί have literal ἕρκη (e.g. Il. 5.90, Od. 24.224, Moschus 4.100, Opp. H. 4.601). Moreover the metaphor’s focus on defence is irrelevant to the old man, and if apposed to κνώδαλον suggests a guard-dog, which is difficult in itself and makes νέμω the wrong verb. Alternatively, κνώδαλον would be in apposition to γέροντα. But then νέμοντα ἕρκος ἀλωῆς is unconvincing as ‘inhabiting the circuit of his plot’; and applying κνώδαλον to the personality of humans is the stuff of abrasive comedy (Cratinus fr. 251, Ar. V. 4, Lys. 476), which seems quite different from the gentle humour applied to the Onchestan’s confusion in what follows. Where to emend? I hesitantly print the smallest viable change – Allen and Sikes’ ἐκτός for ἕρκος. This retains κνώδαλον, the hardest word for a scribe to introduce.244 The oxymoron κνώδαλον νέμοντα (one word implying wildness, the other domestication) may set up a thematic joke where Apollo’s searching uncovers completely the wrong animal (as at 232*, 437); κνώδαλον in this context would probably suggest a donkey (as at Pi. P. 10.36); Allen and Halliday compared Luc. Asin. 43, where the donkey waits for his master to finish digging his garden.245 For παρέξ ... ἐκτός compare 48 διά … διά, Il. 18.576 πάρ … παρά. Other approaches have tended to alter κνώδαλον and νέμοντα: hence solutions such as καμπύλον or κηδάλωι followed by ηὗρε δέμοντα or ηὗρεν ἐλῶντα.246 This route supposes a ἕρκος of stone (so Od. 24.224: see 87–94* for this intertext), wood or hedging (Aesop 19, 233, Varro R. 1.14; see 190

244

Matthiae (1800: 252) suggested that it intruded from a gloss ‘against wild beasts’ placed against ἕρκος. 245 PV 462 uses κνώδαλα for cows, but is not a relevant parallel since they are wild and being yoked for the first time: Ludwich (1887). 246 δέμοντα Barnes (cf. 87); -ν ἐλῶντα Ilgen (cf. Il. 18.564 ἕρκος ἔλασσε; Od. 7.113, Hes. Th. 726); Tyrrell’s -ν ἀμῶντα (1894: 45) could relate to 191 βατοδρόπε. Schmitt’s καμπύλον (1856: 153, cf. line 90) is rather far from the paradosis, as are τρόχμαλον (‘round-stone wall’: Hermann), σκώλους (‘thorns’: McDaniel 1900: 82), ῥωπάδας (Harder 1912: 64). Ilgen’s κωτίλον is untrue. His κάνδαλον (‘trench’), Hermann’s νωχαλόν (‘slow’), νώδαλον (‘toothless’: Ridgeway 1888) and κώκαλον (‘ancient’: Stahl 1912) are not secure as Greek words. κηδάλωι (Courtney 2008) has a marginally more secure footing, but the most likely meanings are inappropriate, namely ‘with a pick/hoe’ (if Hesychius’ gloss σκάλαθρον, itself not attested elsewhere, is related to σκάλλω and σκαλαυθρίς), or ‘with a stirrer’ (cf. σκαλαθύρω, σκάλευθρον, σπάλαθρον, Myc. qa-ra-to-ro). Ludwich (1905) 11 made the poor man κνωδάλου ... γέμοντα, ‘infested with fleas’. Evelyn-White’s παρ’ ἔξοδον ἕρκεος αὐλῆς (1914a: 221) unfortunately tampers with ἀλωή.

251

C O M M E N TA RY 1 8 9 βατοδρόπε*), which would gain relevance here because of its protective function against thieves. For the spelling ηὗρε, preserved in some literary papyri, see West (1998) xxvii. Medieval manuscripts of Homer have εὗ-, the Attic form from the fourth century onwards (Threatte 1996: 482–3). παρέξ (MDb) rather than πάρεξ (ap) was Herodian’s preference for Homer: see West (1998) xviii–xix, though he preferred πάρεκ/ξ anyway. 189 τὸν πρότερος προσέφη Λητοῦς ἐρικυδέος υἱός: As at 89*, the poet does not pause to describe any emotion, e.g. relief at finding a potential assistant; Apollo plunges actively into speech. 182–9 A dawn arrival: Hermes’ nocturnal adventure has been measured against the progress of the night at 142–3, 97–100 and 68–9. In particular, 184–8 parallel 68–72 in order to frame the narrative of the night’s events, and to cue us to contrast the new journey (Apollo’s) with the old. Each passage has five lines, consisting of: |a celestial divinity appearing;247 reference to Ὠκεανός; αὐτάρ + name of god|; ἀφίκανε¦ + a religious place; ἔνθα; reference to inhabitants (in both cases connected to what follows by an |article). ἔδυνε … Ὠκεανόνδε and ὤρνυτ’ ἀπ’ Ὠκεανοῖο are opposed, and map the antagonism of Hermes and Apollo onto the paradigmatic antithesis night/ day.248 The context therefore enhances the structural function typical of epic dawns. Although Dawn (here personified as usual: see -γένεια, φέρουσα, ὤρνυτο) is the gradual process of the sky turning pink and yellow in the morning, epic tradition favoured it over the moment of sunrise as a structuring device for sharp scene-shifts and new narrative starts. The hymnist worked into his lines nearly all the traditional building-blocks of dawn-descriptions.249

Cantilena (1982) 249–51 discusses how |Ἠὼς δ’ ἠριγένεια is an unusual sort of formulaic modification, of a nominative name-epithet formula from line-end to linestart, without enjambement. The desire for parallelism with 68 |Ἠέλιος explains it. 248 See Introduction §3.2 on intratexts and §3.3.2 on day/night. For early signs of Apollo’s luminosity, eventually leading to identification with Helios (excluded at Herm. 381), see H.Ap. 441–4, Theagenes 8 A2, A. Th. 859, TrGF iii.138, Telesilla PMG 718, B. 13.105–6 (φοιβός ‘radiant’: cf. 330*), Oenopides A7, E. Phaethon (fr. 781) 225. 249 Whoever divided Il. and Od. into rhapsodies liked to align dawns with their junctures: Il. 8, 11, 19; Od. 2, 3, 5, 8, 16.2 (cf. 15.495), 17, though not 12.8. For Homeric scene-shifts in general, see Hellwig (1964) 88–107. For the idea that elaborate descriptions of dawn in the Iliad can carry an ‘implicature’ of a structural function by ‘flouting’ the length required for mere description, see Macleod (1982) 47, and e.g. Leech and Thomas (1990) 179–84, 194–204 for the Gricean terminology. 247

252

C O M M E N TA RY 1 8 2 – 1 8 9 These are: a temporal adverb or conjunction; Ἠώς and/or ἠριγένεια in the nominative (required); one or more epithets of Eos; a verb (generally ἐφάνη or a verb of motion); the origin or destination of Eos’ motion (often ‘from Okeanos’); recipients of her light in the dative; if the main verb was of motion, a participial phrase of bringing or announcing light (quite rare).250 By having Dawn bring light to mortals specifically, rather than to immortals alone (e.g. Il. 2.49) or to immortals and mortals (Od. 5.2), the hymnist avoids the suggestion that Apollo needed light to discover the theft, and prompts us to imagine the scene as human participants rather than as privileged external observers, just before reintroducing the hymn’s only human participant. The most surprising feature of the parallelism 184–8 ~ 68–72 is that Apollo arrives not at Pieria, the obvious start of his search, but at Onchestos. The poet could have situated the discovery of the theft after dawn, or made his presentation jump back to Apollo’s discovery (see Scodel 2008). Instead, he makes the search overlap with Hermes’ return and omits the overlap (the discovery of the theft, Apollo’s reaction, the places he searches before Onchestos) – a ­paralipsis made particularly striking by the descriptive potential of what is omitted, which Apollo’s recollections at 195–8 and 340–54 hardly flesh out.251 This decision has various effects. We have seen Hermes pass through Onchestos (87–94) and so know that Apollo is on the right path, though we will find out in 218* that he has not made progress by following the footprints by moonlight.252 We are immediately invited to measure Apollo’s progress against Hermes’ journey (see Introduction §3.2 for this parallelism). The abrupt plunge into Apollo’s search also confers a sense of his urgency (confirmed at e.g. 212); the poet does not even pause to spell out that Apollo is already searching. This urgency is also suggested by the handling in lines 185–9 of the type-scene of arrival leading to conversation (Arend 1933: 28–33). The most common elements are: (i) A goes; (ii) A arrives (iii) at location X; (iv) A finds B; (v) amplifying description (often introduced by ἔνθα); (vi) A adopts a position, normally standing; (vii) A speaks to B. For an audience attuned to this pattern, the absence here of (vi), i.e. a reference to Apollo halting, suggests his haste, even if eventually the 23-line conversation which ensues does entail that he paused.

250

See, with a different approach, Vivante (1980). Herodotus’ use of formulas for dawn and sunrise (ἠώς τε (δὴ) διέφαινε x3, ἅμα τῶι ἡλίωι ἀνιόντι x3) suggests that he recognized dawn-descriptions as a prominent example of epic formularity. One of his few other dawns, 8.23.1 ἅμα ἡλίωι σκιδναμένωι, is also modelled on Homer, as Eustathius observed (Comm. Il. ii.498.22–3). 251 That 185 αὐτάρ does not skip over any significant time-period will be confirmed at 371. Philostratus Im. 1.26.3 and Apollod. Bibl. 3.113 also skip through Apollo’s discovery that the cows are missing. 252 See 88* for why Onchestos is a strategic place to look.

253

C O M M E N TA RY 1 9 0

190–200 190 ὦ γέρον, Ὀγχηστοῖο βατοδρόπε ποιήεντος: Hermes, in his parallel encounter with the Onchestan, also began ὦ γέρον (90*), though the tone of that address is flexible and Apollo seems more respectful in what follows (below, 190–200*). He extends his address to a length characteristic of epic and tragedy (and comic parodies; Wendel 1929: 49–51), in ways which independently suggest Apollo is striving for a high register (contrast the tone of e.g. 436*). The first expansion is βατοδρόπε, a hapax from a family of rare -δροπος compounds most at home in lyric, and a word noted by Et.Gen. β.63.253 It is a noun, rather than an adjective which would entail an unconvincing interleaved word-order.254 I take it to mean ‘bramble-cutter’ rather than ‘blackberry picker’ (Clay 1989: 115 n. 69), since βάτοι had a reputation for worthlessness and thorns more than for their fruit (see e.g. Aesop 233, 263); they occur in this connection in the context of gardening in particular, as when Laertes is hindered by brambles in his plot (Od. 24.230, possibly in the hymnist’s mind: see 87–94*), or more distantly when brambles oppose vines at Luke 6:44. A vine-grower would naturally have a δρεπάνη (see e.g. Sc. 292, Pl. R. 1 333d), but use it also for cutting away brambles. Address-usage makes it likely that Apollo is expressing a typical viticultural task, regardless of what the Onchestan is doing in 188: nouns expressing activities appear in addresses occasionally, especially as divine epithets in verse and military or civic occupations in prose, but very rarely do they indicate a temporary activity, for which a relative clause is normally used as in 254.255 βατο-δρόπε probably plays on ‘Battos’, the informant in other versions of the myth (Fick 1890: 28). See Introduction §3.3.4 for Battos predating Herm., and compare Plato’s self-referential use of πλάττω of a young wordsmith at Ap. 17c, and the same pun in Timon SH 793. Apollo also includes in his address the man’s location, using a genitive toponym with its own epithet. Such genitives are extremely rare in prose addresses; in verse, they are largely confined to tragedy where the genitive depends on ‘ruler’ or ‘inhabitant’; an adjective such as Ὀγχήστιε would be more standard (Wendel 1929: 71–3, Dickey 1996: 293). For ποιήεντος see 87–8 ἀνθέουσαν … λεχεποίην.

Whence EM 191.57. The entry ends φη in manuscript A of Et.Gen., and probably derives from Philoxenus’ Rhematikon (fr. 392); Gaisford (1848), as often, thought instead of Herennius Philo’s Rhematika. 254 Hence not βατόδροπε ‘pricked by brambles’, despite the parallel at Bion 1.21–2 αἱ δὲ βάτοι νιν | … αἷμα δρέπονται. 255 On apposed nouns in addresses see Wendel (1929) 38–9, Dickey (1996) 177–84, 287– 92, 298–305. 253

254

C O M M E N TA RY 1 9 1 – 1 9 4 191 βοῦς ἀπὸ Πιερίης διζήμενος: Word-order suggests ἀπὸ Πιερίης is better taken with βοῦς (compare H.Ap. 393 Κρῆτες ἀπὸ Κνωσοῦ) than with ἱκάνω. ἐνθάδ’ ἱκάνω: A traditional slight pleonasm, which normally carries some emphasis. In our case the nuance is probably ‘I’ve come all the way here’. Often the arrival ‘here’ is surprising (262*, Od. 1.173, 7.239, 11.160, 16.31); elsewhere the emphasis is assertive (Il. 13.449) or emotive (Od. 6.206, 7.24). 192 πάσας θηλείας, πάσας κεράεσσιν ἑλικτάς: Apollo gives sparing detail about the missing cows, not even including their number. The animals stolen in other sources are discussed in n. 113. κεράεσσιν ἑλικτάς is a unique development of the formula ἕλικας βοῦς (116*); for the dative see GH ii.75. For the shape of the hexameter (two phrases with anaphora of πᾶς, underlining the cows’ uniformity) compare Call. H.Art. 14 = 43, Theoc. 15.6, CA lyr. adesp. 9.2. 193–4 ὃ δὲ ταῦρος ἐβόσκετο μοῦνος ἀπ’ ἄλλων | κυάνεος: κυάνεος is often used for dark hair, and was one of the main categories of colour for cows in antiquity, along with reddish brown (the most common: Richter 1968: 47), white and mixed. For μοῦνος ἀπό ‘separate from’ see S. Ph. 183, E. IA 669, and οἶον ἀπ’ ἄλλων| at Od. 9.192, H.Aphr. 76. ἄλλων (not ἀλλάων) leaves the gender-composition of the rest of the herd vague, but regardless of any castrated males, Apollo implies one bull and a much larger number of females – a typical ancient strategy to control reproduction.256 On his veracity, see 190–200* below. 194–6 χαροποὶ δὲ κύνες κατόπισθεν ἕποντο | τέσσερες, ἠΰτε φῶτες ὁμόφρονες. οἳ μὲν ἔλειφθεν | οἵ τε κύνες ὅ τε ταῦρος, ὃ δὴ περὶ θαῦμα τέτυκται: Dogs are frequently present in sources to help herdsmen, whether the animals are sheltered at night, on the move (Il. 18.578–74), or grazing (17.65; similarly Ps.-Hdt. Vit. Hom. 21–2). Quartets are a favoured small group in Epos (x20), including a quartet of alert dogs at Od. 14.21–2. Here they ‘were accompanying in the rear’ (there is no pleonasm: compare e.g. Il. 24.182, where Hermes ἕψεται Priam as his guide). I take them to be in the rear of the herd, not of the bull, since the latter would be dereliction of duty, incompatible with the neat encapsulation in ἠΰτε φῶτες ὁμόφρονες of how their intelligent teamwork makes them a useful alternative to having more herdsmen.257 Their

256

The proportion of bulls is low at e.g. Il. 2.480, 18.577–86, Theoc. 4.20, 25.126, Varro R. 2.5.18 and even Ps.-Archim. Bov. An alternative strategy is gender segregation, e.g. Il. 15.630–6 (all cows), Pi. P. 4.205, Aesop 49, Theoc. 27 (all bulls). See Georgoudi (1990) 233–5. 257 For ὁμοφροσύνη of small groups see e.g. Hes. Th. 60, Thgn. 81; it more commonly applies to couples or large groups. The word ὁμόφρονες will be developed at 391*. For anthropomorphism of dogs, cf. Hermog. Id. 2.4 on X. Cyn. 3.7.

255

C O M M E N TA RY 1 9 7 – 1 9 8 protective role is assumed in 195–6 (Apollo is astonished that they were left behind) and probably elaborated in χαροποί, which elsewhere in Epos qualifies lions and has a strong connotation of fierceness, suitable for guard-dogs, who may (in the world of epic similes) have to fend off precisely lions.258 Apollo’s implication seems to be that he can only imagine that the cows must have been stolen from the ‘front’ of the herd. The Ionic spelling τέσσερες is preferred in papyri of Homer; Ω’s τέσσαρες arose under the influence of Attic τέτταρες (Wackernagel 1916: 13, West 1998: xxxv). For the prolongation κύνεςˉ compare Il. 3.40 ὄφελες ἄγονος, Od. 3.322 αὐτοέτες οἰχνεῦσιν, West (1982) 38. For the accentuation of περί ‘exceedingly’ see e.g. ΣA Il. 2.831 and West (1998) xix, against LSJ s.v. περί E.II. 197 ἠελίοιο νέον καταδυομένοιο: The tense implies, if pressed, that Apollo discovered the theft almost immediately, before sunset. Otherwise, he would probably conjecture that the cattle were stolen at night (see e.g. 67, 284). For the metrical lengthening καταδῡομένοιο compare e.g. A.R. 1.924–5 νέον γε μὲν ἠελίοιο | δῡομένου. 198 ἐκ μαλακοῦ λειμῶνος, ἀπὸ γλυκεροῖο νομοῖο: μαλακός is commonly applied to meadows, e.g. Il. 14.349, Od. 5.72, 9.133, H Dem. 7, H.Ap. 118; also of things in meadows at e.g. Pherecr. fr. 114, Longus 1.4.3. By contrast, the stem γλυκ- is rarely used of places except homelands (by extension of its application to νόστος), and is frequently used of food and drinking-water.259 Apollo therefore seems to be focalizing Pieria through the cows, as a home with good food to which they must long to return. This draws sympathy for them, and increases the θαῦμα (196) that the cows left. For further description of the Pierian meadow see 72, 221.

The Greeks related χαροπός to the stem of ὄπωπα, though other derivatives have -οψ or -ωπ-: Risch (1974) 171–2. For χαροπός beside βλοσυρός describing fierce looks see Sc. 175–7 of lions, IG IX2.4 880 (Corcyra, c.600) of Ares. Fierceness is probably at stake at Ar. Pax 1065–6, where the Spartans are χαροποί apes; for the link of χαρ- to χάρμη and avidity, see ­Latacz (1966) 38–43. On the adjective as a post-­epic ­colour-term see Maxwell-Stuart (1981); the development means that Xenophon (Cyn. 3.3) makes χαροποί eyes a bad feature of a hunting dog. 259 In Pi. P. 5.23–4, γλυκύς applies to a garden because it belongs to Aphrodite. Drinking water: e.g. Od. 12.306, Xenoph. fr. 1.8, Hdt. 4.52.2, 181.2. Among human foods, the stem applies mostly to honey and fruit. A telling instance of γλυκύς applying to plants is Periander’s gloss on Hes. Op. 41 ἀσφοδελὸν μέγ’ ὄνειαρ as γλυκὺς δ’ ὁ ἀνθέρικος (Plu. Conv. Sept. Sap. 157f): this is paradoxical (for asphodel as a cheap food see 221*), to capture Hesiod’s contrast of cheap honesty with luxurious food procured unjustly. 258

256

C O M M E N TA RY 1 9 9 – 2 0 0 199 ταῦτά μοι εἰπέ: οὗτος is rarely cataphoric, but compare καί μοι τοῦτ’ ἀγόρευσον ἐτήτυμον, ὄφρ’ ἔϋ εἴδω (Od. x7, H.Ap. 467). The general usage of οὗτος suggests that both here and in that formula it is addressee-oriented (roughly ‘this topic for your input’ rather than ‘this topic which I am raising’), as a form of politeness.260 For the plural with a single yes–no question following, compare e.g. Il. 6.150, Od. 8.350. γεραιὲ παλαιγενές: Speeches in Epos do not often contain a second vocative, but where they do it can help to articulate the speaker’s structure; here it renews the appeal for the addressee’s attention at the switch from background information to climactic request.261 εἴ που ὄπωπας: The phrase is part of a formulaic system related to searching, typically following ‘tell me’ or ‘look to see’: εἴ που ὄπωπας /ἄκουσας / ἐφεύροι|, and |εἴ που ἐσαθρήσειε and variants. που could mean ‘somewhere’, but Apollo has no reason to think that the old man is a roamer, so the modal sense ‘perhaps’ is to the fore. Expressing doubt whether one’s addressee will be able to help makes it easier for them not to, and so it is a widespread form of politeness (Brown and Levinson 1987: 145–72). Apollo’s choice of εἴ που ὄπωπας over εἴ που ἄκουσας, or a combined reference to seeing or hearing, contributes to characterizing his deafness to possible aural evidence at this stage in the hymn.262 200 ἀνέρα: An incorrect assumption about the perpetrator, much like A. Ag. 1251 ἀνδρός, S. Ant. 248 ἀνδρῶν. For Apollo to function in the narrative, his time-flattening prophetic skill is made fallible, and the narrator and his audience can temporarily enjoy superior knowledge. For this feature of Apollo in narrative, see Pi. P. 9.42–9, Felson (2009). Regarding the plausibility of Apollo’s assumption, see further 265–73*. 260

Some deictics situate their referent in space and time, others within a speech as upcoming (‘cataphoric’) or previously mentioned (‘anaphoric’); for the terminology see Felson (2004). For cataphoric οὗτος see also 269, 291, Il. 15.217, 23.415, Hes. Op. 293, 388, fr. 211.6. Anaphoric οὗτος, by contrast, is common in direct speech, e.g. Il. 1.126. Bakker (1999) discusses the addressee-orientation of οὗτος in contrast to ὅδε; cf. 266*. For Apollo’s politeness see further below, 190–200*. 261 Some other examples: 208* and Il. 10.558 from tenuous comments to a proper answer; 24.618 resumes the idea from 601; Od. 14.131 switches from general to specific, 14.185 from Ithaca to the stranger. Where the tone is more emotive, repeated addresses can contribute to it, e.g. Phoenix’ appeal to Achilles at Il. 9.485–96, or Eurymachus’ attack on Halitherses at Od. 2.178, 192. 262 ‘Seen or heard’: Od. 3.93–4 = 4.323–4, 17.510–11. Only ‘heard’: Od. 17.106, where Penelope’s choice demonstrates – with irony – her assumption that Telemachus cannot have seen Odysseus. Only ‘seen’: H.Dem. 69–71, because seeing is Helios’ strength. On the theme of visual and aural signs see Introduction n. 104.

257

C O M M E N TA RY 1 9 0 – 2 0 0 ταῖσδ’ ἐπὶ βουσί: For ἐπί ‘in charge of ’, see LSJ s.v. B III.6. Anaphoric ὅδε is not particularly common (Ebeling 1885 s.v. III.B); perhaps its markedly proximal nuance encourages the addressee to visualize the cows (see Felson 2004: 260). διαπρήσσοντα κέλευθον: Formulaic, and part of a system with -πρησσον πεδίοιο, πρήσσωμεν+ ὁδοῖο. The phrase’s other occurrences suggest the interpretation ‘proceeding on their journey’ rather than ‘proceeding along the road’ (though it makes little difference). See also 203*. 190–200 A polite request: Apollo’s manner with the Onchestan contrasts with that of Hermes at 90–3 – a contrast set up by the similarity of 89 to 189 and the shared opening ὦ γέρον. But whereas Hermes’ speech was curt and hinted coercively at his divinity (93), Apollo’s approach centres on positive politeness – choosing expressions which appeal to the Onchestan’s desire to feel that Apollo is interacting with him on terms of familiarity.263 Apollo begins with an address, extended in a way which avoids bluntness and smacks of a formal register (190*), though the mundane content (βατοδρόπε) keeps it from deference. After the essentials about the stolen cows (191–2) he spins out an extended description about the rest of his herd (193–6). I submit that for the external audience these details hold little water. Apollo implies that he knows the relative positions of the animals with some precision, though he did not seem to be anywhere near Pieria at 71–4, where the narrator also failed to mention an encounter between Hermes and dogs, a traditional enemy of the thief (see 142–5*).264 Those lines (αὖλιν, βοσκόμεναι) suggested that the herd was basically static rather than on the move as Apollo implies in κατόπισθεν ἕποντο (194). Finally, the existence of a bull may have caused surprise given the model of Helios’ cows, who need no reproduction (Od. 12.130–1).265 That surprise is at any rate justified later, when Hermes contrasts the cows’ prior circumstances with a future in which they will reproduce in a mixed herd (493–5*). If Apollo is inventing these animals – a tactic which could be communicated in performance – his obvious reasons for doing so are his desire to cover his identity, to shelter the Onchestan from a sense that he is embroiled in a theft of divine property, to engage his interest

263

For the term ‘positive politeness’ see Introduction n. 60. For politeness including the delay of requests in Homer, see also Bedke (2016). 264 If tales with this version predated Herm., one could interpret its relegation to a fiction by Apollo as the poet’s polemic against his rivals, as in the neoanalyst approach to Odysseus’ Cretan tales (e.g. Grossardt 1998). 265 Admittedly the audience’s expectations here may have been coloured by pre-­ existing versions of the myth. For the presence of a bull in Ant. Lib. see n. 113.

258

C O M M E N TA RY 2 0 1 fully rather than bluntly ask for information, and to interact more naturally with him.266 Apollo’s rhetorical elaboration then comes across as an attempt to add conviction to his dissimulation. Lines 193–6 are based on a chiasmus enhanced with echoes (ὁ δὲ ταῦρος ... -οὶ δὲ κύνες /οἵ τε κύνες ὅ τε ταῦρος), alliterative panache (κυάνεος, χαροποί τε κύνες κατόπισθεν), and balancing details (κυάνεος and χαροποί juxtaposed by enjambement; ἐβόσκετο μοῦνος/ κατόπισθεν ἕποντο). The two surrounding sentences about the cows themselves balance: 198 ἐκ ... ἀπό follows 191–3 ἀπὸ Πιερίης ... ἐξ ἀγέλης, while 192 and 198 both comprise two qualifying phrases in asyndeton. A series of suggestive touches appeal to the imagination (194–6* on χαροποί and ἠΰτε φῶτες ὁμόφρονες, 198* γλυκεροῖο). The positive politeness continues in Apollo’s eventual request. Having normalized his herd, he delays the language of crime in ἔλειφθεν and ἔβαν, and by presenting his prevailing reaction as surprise (196); only line 200 implies that he suspects theft.267 The second address appeals again for the Onchestan’s attention at the crucial point (199*). The succinct phrasing of ταῦτά μοι εἰπέ need not sound curt, and in fact by eschewing the circumlocutions typical of negative politeness it is more probably friendly in tone.268 που in 199 is the kind of hedge typical of negative politeness, while ταῦτα more positively suggests the Onchestan’s ability to participate usefully in the exchange and ταῖσδε may prompt him to engage cognitively with the problem. Throughout his speech, therefore, Apollo’s urgency manifests itself not in bluntness, but in the insistence with which he attempts to ensure a helpful response.

201–11 201 τὸν δ’ ὁ γέρων μύθοισιν ἀμειβόμενος προσέειπεν: For reply-introductions belonging to this system see 162*, which this line echoes (Introduction §3.2), though it does not attempt to encapsulate the man’s complex reaction or the tone of what follows as that line did with κερδαλέοισιν. The echo

266

For engaging narrative as positive politeness see Brown and Levinson (1987) 106–7. Aphrodite’s less convincing story at H.Aphr. 81–3 is a useful contrast. Haft (1996) 47 n. 34 suggested that Apollo is signalling his Olympian status to the old man at ἀπὸ Πιερίης, but Pieria had human inhabitants. 267 Cf. Eitrem (1906) 262. θαῦμα τέτυκται could imply agency (see e.g. Od. 8.276 τεῦχε δόλον; Il. 18.549 τὸ δὴ περὶ θαῦμα τέτυκτο of Hephaestus’ craftsmanship) but is ambivalent. Contrast Apollo’s very different strategy at 254–5, 340. For θαῦμα in Herm. see Introduction n. 101. 268 Briefly phrased requests in Epos are rarely curt, though see e.g. 254, Od. 11.144. Conversely, long requests need not be leisurely (e.g. Il. 11.819).

259

C O M M E N TA RY 2 0 2 introduces μύθοισιν. As at e.g. 218 and 329, this instance shows the word’s loss of specificity since the Iliad.269 202 ὦ φίλος: Neither word is appropriate for addressing a god. For this feature of ὦ in Epos see Scott (1903) 192. The singular of φίλος is characteristic in Epos for addresses to family-members or old friends (sarcastically at Il. 21.106, Achilles to Lycaon), and to express friendliness towards new acquaintances. It is used four other times in addresses to complete strangers: Odysseus thinking he has been abandoned at Od. 13.228, as beggar at 16.91, 17.415; Theoclymenos the exiled suppliant at 15.260. Although the number of passages is not large, the turn of phrase may betray the Onchestan’s sense of his precarious position. φίλος appears in the nominative instead of the masculine singular vocative strikingly often compared with other words. In Epos the usage is mostly restricted to φίλος as a substantive; the most common metrical format is ⏑ φίλος C- (Il. x7, Od. x5), or there is a system for starting a speech with ὦ φίλ’ (Od. x10, avoided in Il.), which admits ὦ φίλος V- (here, Od. x2, Sc. 95) and δός, φίλος (Od. 17.415) among its variants.270 202–3 ἀργαλέον μὲν ὅσ’ ὀφθαλμοῖσιν ἴδοιτο | πάντα λέγειν: Here as often μέν marks a speaker’s opening gambit, without a subsequent particle balancing it (neither 205 δέ nor 206 αὐτάρ creates a suitable antithesis): Denniston 382–4. The Onchestan’s opening is particularly similar to Od. 7.241, 13.312, 19.221, which also begin ‘vocative, it is ἀργαλέον to X, because Y, but nevertheless Z’. In each case the speaker is (or wants to seem) embarrassed, and the Onchestan seems all the more so because of his further long-winded generalizations in element Y (lines 203–5).271 Both idioms ‘to X is ἀργαλέον’ and ‘Someone is ἀργαλέος to X’ (see LfgrE s.v. ἀργαλέος 2) normally do without ἐστίν and front ἀργαλέος. For ἀργαλέος in connection with recounting

For the Iliad’s categories of μῦθοι see De Jong (1992) 392–3: commands, boasts, insults, and extended recitations of remembered events. The emphasis on μῦθοι arising from deliberation in Martin (1989) 15 is a useful addition to these at Od. 24.350 at least. 270 See also adjectival φίλος ὦ Μενέλαε at Il. 4.189. For the general avoidance of bare φίλος and φίλε in biceps positions see 477*. On the nominative for vocative see Wackernagel (1912) 8–9, Schulze (1923), Svennung (1958), Dickey (1996) 134–8. 271 At Od. 7.241 Odysseus is trying to wriggle out of Arete’s questions. At 13.312 he is embarrassed that Athena deceived him. 19.221 is the most subtle use: Odysseus (in disguise) feigns hesitation to increase the effect on Penelope of his ability to describe Odysseus’ appearance; she takes his point to be that remembering is ἀργαλέον in the sense ‘effortful’, but the external audience appreciates that it is also emotionally ‘painful’ for Odysseus. 269

260

C O M M E N TA RY 2 0 3 – 2 0 5 numerous details see e.g. Il. 2.490, 12.176 (narrator-text), Od. 7.241, Hes. Th. 369. For other instances of its passage from ‘painful’ (< ἀλγ-αλέος) to ‘mentally effortful’ and ‘difficult’ see e.g. Od. 13.312, Hes. Op. 484, Egoscozábal (2003). In fifth-century Athens the word’s register seems to be colloquial. It does not occur in tragedy; for prose and comedy see e.g. Ar. V. 1279, Nu. 450 with Dover (1968), X. Hier. 6.5, and Aeschines 1.61 as part of amusing invective. An association with humour would suit Herm.’s characterization of the Onchestan’s platitudes. In ὅσ’ ὀφθαλμοῖσιν ἴδοιτο, an indefinite ‘someone’ remains unexpressed: see LSJ s.v. τις A II.15, KG i.35–6. M’s ἴδοιμι came from a scribe unfamiliar with this construction. The optative in a primary-sequence indefinite clause contributes to the tone of embarrassment since it emphasizes the range of possibilities the man faces: see GH ii.248, KG ii.428–9 (in particular the parallels at Thgn. 369–70, X. Cyr. 1.6.19). 203 πολλοὶ γὰρ ὁδὸν πρήσσουσιν ὁδῖται: A populated human world is glimpsed, but kept firmly in the background. The alliteration (on π-) and polyptoton (of ὁδ-) are typical of Greek sententiousness: such wordplay creates a memorable dictum, and may be taken as a linguistic ‘guarantee’ that the concepts in the proverb really are related: Silk (1974) 176–89, 225–7. 204 μεμαότες: Since Ionic μεμᾰότ- is unmetrical, Epos usually replaces it with μεμαῶτ- (x72), probably underlain by Aeolic μεμάοντ-. By contrast μεμᾱότ- (only x3) rests on the common metrical lengthening ⏑⏑⏑⏑  ⏑–⏑⏑. Il. 16.754 μεμᾱώς is tertiary. See GH i.429–31, 101. 205 φοιτῶσιν: The common position of φοιτάω at line-start is neatly deployed in this run-over: its sense of multiplicitous travel (repeated, in various directions, etc.) increases the man’s difficulties in cataloguing men’s journeys, and the enjambement puts it next to χαλεπόν. I take M’s variant πρήσσουσιν to have been imported from 203. χαλεπὸν δὲ δαήμεναί ἐστιν ἕκαστον: The clichés continue. The construction ‘It is χαλεπόν’ + infinitive plays a large part in early Greek gnomic and didactic discourse, for example as the opening of seven poems in the Theognidean corpus; so does the opposition κακός/ἐσθλός in 204. The difficulty of distinguishing good and bad men is mentioned several times. Often it is traced to the gods (e.g. Thgn. 133–42, 401–6, S. Ant. 622–4, E. HF 669–70) – a conceptual background which enhances the irony here that the man is expressing the difficulty of judging travellers to a traveller whom he has not identified as a god. δαῆναι normally implies learning by practice, but is used of finding out about people at Od. 16.316. For the word’s importance towards the end of Herm. see 444–5*.

261

C O M M E N TA RY 2 0 6 – 2 0 9 206 αὐτὰρ ἐγὼ πρόπαν ἦμαρ ἐς ἠέλιον καταδύντα: αὐτὰρ ἐγώ contrasts the static, trustworthy speaker with the ὁδῖται who have just been described.272 His time-reference (ἐς ἠέλιον καταδύντα) answers to 197 ἠελίοιο νέον καταδυομένοιο, and specifies a later moment which allows for Hermes’ mythologically fast travel from Pieria. Since πρόπαν ἦμαρ and ἐς ἠέλιον καταδύντα both generally occur in the context of dining, traditional referentiality may suggest that the man worked through dinner (compare Il. 19.162–3 οὐ γὰρ ἀνὴρ πρόπαν ἦμαρ ἐς ἠέλιον καταδύντα | ἄκμηνος σίτοιο δυνήσεται ἄντα μάχεσθαι, Hes. Th. 596–9). For the construction of ἐς ἠέλιον καταδύντα see Schwyzer ii.391. 207 ἔσκαπτον περὶ γουνὸν ἀλωῆς οἰνοπέδοιο: See 90* for digging as an emblematic task of the vine-grower, and for the difficulty of deciding whether ἔσκαπτον here and/or 87 δέμων specify what the man was doing when he saw Hermes, or give a general impression of his standard activities. περί + accusative means ‘all around, throughout’ (LSJ s.v. C I.4), as normal in Epos. For the distribution of γουνὸν ἀλωῆς οἰνοπέδοιο see 87–94*. 208 παῖδα δ’ ἔδοξα, φέριστε, σαφὲς δ’ οὐκ οἶδα, νοῆσαι: The man moves from background platitudes to a response. As in 199* a second vocative marks the switch of focus. φέριστε expresses general positive affect, while the halting rhythm and the hedging show the Onchestan’s continued embarrassment as he has to correct 200 |ἀνέρα by the implausible |παῖδα. νοῆσαι recalls the action of 87 ἐνόησε. Both σαφές and ἔδοξα are relatively late forms: see Introduction §1.1. Parenthetic idioms such as σάφ’ οἶδα and σάφ’ ἴσθι are themselves only common from Aeschylus on. 209 ὅς τις ὁ παῖς: ‘whoever that child was who …’. Compare Il. 3.166–7 τόνδ’ ἄνδρα … | ὅς τις ὅδ’ ἐστὶν Ἀχαιὸς ἀνήρ. ἅμα βουσὶν ἐϋκραίρηισιν ὀπήδει: A unique modification of the formula βοῶν (or νεῶν) ὀρθοκραιράων. ἐϋκραίρηισιν matches 192 κεράεσσιν ἑλικτάς partly but not fully.273 The Onchestan is cagey about whether the cows he saw can be identified with those Apollo has lost, but if they can, he flatters Apollo that they were fine animals: Fernández Delgado (1998) 11.

The transition αὐτὰρ ἐγώ occurs in Od. x74, in Il. only x16, a significant difference considering that Od. has only about 16 per cent more direct speech. 273 A two-termination compound εὔκραιρος does occur (e.g. A. Su. 300), and is reflected in p’s ἐϋκραίροισιν here. But MΘ’s ending is both difficilior and typical in compounds of -κραιρα, including ὀρθοκραιράων (Nussbaum 1986: 224–9). 272

262

C O M M E N TA RY 2 1 0 When ὀπηδέω occurs with ἅμα in Epos (LfgrE s.v. ὀπηδέω 2, plus here, H.Ap. 530, Pany. fr. 18), the subject is more powerful than the object accompanied. This contrasts with the usual sense of the peripheral which arises with both ἅμα and ἑπ-/ὀπ- individually. 210 νήπιος, εἶχε δὲ ῥάβδον: The formulaic pattern is |νήπιος in adding enjambement, apodotic δέ, clause exposing naivety. The old man, uniquely, uses νήπιος just of Hermes’ age, then adversative δέ and a clause describing far from juvenile behaviour.274 Compare the complex uses in 152 (145–54*) and 163–5* (the latter structurally related to our passage: Introduction §3.2), which grapple with Hermes’ status as an abnormal baby (Introduction §3.3.3). This ῥάβδος appears only here in the hymn; we were not told where Hermes obtained or disposed of it. The natural interpretation is that he is carrying an ad hoc cattle-prod, called a ῥάβδος in Moschus 1.1 G–P, but elsewhere κέντρον (e.g. Call. fr. 24.7, Philip 19.5 GP, Longus 1.27.2), ὅρπηξ (Hes. Op. 468), μύωψ (Cercidas fr. 8.2, Antiphilus 15.1 GP), and so on.275 More central to the usage of ῥάβδος are two kinds of staff that the adult Hermes will hold, which the ῥάβδος here foreshadows. The first is Hermes’ golden magic wand, which he uses in various ways including at Od. 24.5 while herding souls; the second is his κηρύκειον. Both staffs will be discussed further in the notes to 529–32*. For those of Herm.’s early rhapsodes who held a ῥάβδος rather than a lyre, this line furthers the connection between Hermes and the hymnist (Introduction §3.4) and provides an interesting opportunity for the performer to act out the story with a prop.276 ἐπιστροφάδην δ’ ἐβάδιζεν: ἐβάδιζεν is a colloquialism which characterizes the man’s tone (see Ar. Pax 114–17 with Olson 1998). In Epos ἐπιστροφάδην means ‘turning to and fro’.277 Hermes’ aim would be to keep the cattle in line, not to drive them to and fro (see 75*). In principle the adverb might 274

See Edwards (1966) 142 n. 67. The formulaic pattern occurs often in Il.; also e.g. Od. 3.146, H Aphr. 223, Hes. Op. 40. 275 For images, see e.g. Amouretti (1986) 92–3, 294–5, fig. 10, pl. 8a, 9c, LIMC s.v. Hermes no. 246. Lycurgus’ βουπλήξ (Il. 6.135) was interpreted as an axe (so in art), a whip or a prod (Ap.S., Paus. Att. s.v.); for cattle-whips see 497*. A less common herding instrument was a throwing-stick, used to keep cows in line at Il. 23.845, where ΣD glosses καλαύροπα: τὴν βουκολικὴν ῥάβδον. 276 On the ancient connection of ῥάβδος and ‘rhapsode’ see e.g. West (1966) on Hesiod Th. 30. 277 Four other instances, of taking on a large number of fighters; cf. LSJ s.vv. ἐπιστρεφής II.1, ἐπιστρέφω II.2, ἐπιστροφή II.4, and ἐπιστρωφάω for related words. Contrast Ps.-Opp. C. 4.138, where ἐπιστροφάδην describes lions who simply turn tail when faced with an enemy. Leduc (1998) 45–8 interpreted ἐπιστροφάδην with reference to Hermes’ epithet στροφαῖος.

263

C O M M E N TA RY 2 1 1 have a different ad hoc meaning such as ‘turning (and looking) backwards’, corresponding to more common uses of ἐπιστροφή and ἐπιστρέφω. Perhaps Hermes would do this to see if Apollo is following (cf. Hdt. 3.156.1 ἐπιστρεφόμενος ὡς ... αὐτόμολος), but the description gave no hint of this, so I find it more natural to keep his concentration on the cows. Herm. has six adverbs in -δην (here, 313, 411, 415, 426, 494), a much higher frequency than Il. (c.20) or Od. (c.10). The distributions of adverbs in -δην, in -δόν and in -δα seem to vary independently according to authorial preferences: Herodotus uses -δόν more than other authors but hardly ever -δην, whereas Aeschylus and Euripides have -δην more than -δόν, and so on. I have discounted the few words of this class which gained currency (e.g. σχεδόν, ἅδην, later βάδην). Haas (1956) 138–41 surveys the Homeric uses. 211 ἐξοπίσω δ’ ἀνέεργε: ‘restrained them (and pushed them) backwards’. ἐξοπίσω is the opposite of what one expects from a herdsman and is placed emphatically. Contrast the idea at Il. 17.752, where the Ajaxes ἀνέεργον ὀπίσσω the Trojan attack. κάρη δ’ ἔχον ἀντίον αὐτῶι: ‘holding one’s head so-and-so’ was formulaic: LfgrE ii.840.40–3. Adjectival ἀντίος occurs in Epos also at 345, Od. 17.334; for the sense ‘facing’ + dative, see e.g. Hdt. 2.34.1, X. Eq. 6.2. For singular κάρη after a plural subject, ‘they (each) held their head …’, see e.g. X. An. 4.7.16 εἶχον … μαχαίριον, KG i.14–15. Ω had ἔχεν, but ‘he held his head ἀντίον αὐτῶι’ is not a plausible shorthand for ‘he held his head in the opposite direction from his direction of travel’ (pace Ernesti).278 ἔχεν may have arisen through a scribe levelling the subject of the verbs. 201–11 An awkward reply: The Onchestan wishes to negotiate between averting the obviously superhuman infant’s threat (92–3) and not lying to his interlocutor; also, he is still bewildered by what he saw, and embarrassed to provide such an implausible account. The result is an amusing showpiece of characterization through direct speech. The man begins reticently: 202 ὦ φίλος* may already sound precarious. Then one trite disclaimer about things being difficult is fleshed out with an unusual number of further generalizations (202–3*), including another cliché about things being difficult (205*); details such as the optative ἴδοιτο (202) and the selection of φοιτάω among verbs of motion (205) contribute to expressing the complexity of the task, and hence lack of confidence in his answer. That answer, when it eventually comes (208–9), is marked by further hedging (ἔδοξα, σαφὲς δ’ οὐκ οἶδα and ὅς Vergados and Schenck zu Schweinsberg keep ἔχεν and take κάρη to be plural, ‘he kept their heads …’. I see no example of plural κάρη until Q.S. 6.250. H.Dem. 12 uses nominative plural κάρα.

278

264

C O M M E N TA RY 2 1 2 τις ὁ παῖς) and a halting rhythm. παῖδα, νήπιον and ἐξοπίσω are all placed at line-start to suggest his emphatic delivery of the most paradoxical information. But the man shies away from giving a helpful physical description of the child. The man’s lack of confidence in his information is expressed in these ways, and he mitigates his fear of disappointing Apollo by establishing a cooperative attitude. He suggests his own simple trustworthiness (206*), and picks up several things from Apollo’s question: 202–3 ὅσα ἴδοιτο … λέγειν ~ 199–200 εἰπέ … εἴ που ὄπωπας; 203 ὁδὸν πρήσσουσι ~ 200 διαπρήσσοντα κέλευθον; 206 ἐς ἠέλιον καταδύντα ~ 197 ἠελίοιο νέον καταδυομένοιο; 208 ἔδοξα... νοῆσαι reverting again to ὄπωπας; 209 ἐυκραίρηισιν ~ 192 κεράεσσιν ἑλικτάς. The opening platitudes are not merely reticent but seek to establish common ground while demoting the man’s individual responsibility for the information he has to deliver (cf. Brown and Levinson 1987: 112, 274). The colloquialism ἐβάδιζεν (210*) could be a sign of the man’s humble social position but also a semi-conscious reinforcement of the friendly tone and avoidance of formality. The Onchestan in this scenario fits the role of a traveller’s helper, met by chance when the traveller requires directions, often at wells (Thompson 1955–8: N715.1). In Epos these helpers are typically young (Reece 1993: 12–13). The Onchestan is atypical: he is old, does not know where Hermes is, and has been instructed by Hermes not to help Apollo; such ‘unhelpfulness’ is something found more frequently in quest-narratives outside Epos (Davies 2006: 200). It is no coincidence that Hermes is the Greek traveller’s helper par excellence (already Il. 24.153 ff., Od. 10.277). On this occasion, his advice to the Onchestan allows him to use his sphere of influence against Apollo.

212–18 212 φῆ ῥ’ ὁ γέρων, ὃ δὲ θάσσον ὁδὸν κίε μῦθον ἀκούσας: Apollo departs without thanks or further questions: the man has implied that he has no further answers to offer. So at Il. 6.390 ἦ ῥα γυνὴ ταμίη, ὃ δ’ ἀπέσσυτο δώματος Ἕκτωρ, Hector hurries on peremptorily without thanking the maid for her directions. Apollo’s haste will be underlined throughout this section (215 ἐσσυμένως, 227 ἤϊξεν, 233 σπεύδων). μῦθον ἀκούσας (Mxm) is less explicable as a corruption than Ψ’s Φοῖβος Ἀπόλλων, which may well have arisen from a gloss specifying ὃ δέ. The line then merges three traditional patterns for capping a speech: ‘|ἦ/φῆ ῥα ... but ’ (Il. x3, Od. x2 with ἦ; here, Hes. Th. 550, H Dem. 145 with the clearer φῆ); ‘So he spoke, μῦθον ἀκούσας|’ (Il. x3, Hes. frr. x2); ‘βῆ δ’, , ἐπεὶ τὸν μῦθον ἄκουσεν’ (Il. 2.16, Od. x3). Here the verb of motion is κίε, which the hymnist is using with a structural function in this section (Introduction §3.2).

265

C O M M E N TA RY 2 1 3 – 2 1 4 θάσσον: The accent is controversial. The alpha could be long or short etymologically (Lamberterie 1990: 584–90 vs Sihler 1995: 363). Herodian i.523 supported θᾶσσον, but θάσσον has reasonable manuscript support in the Iliad and appears to rhyme with ἄσσον, which has good support, at 6.143. West (1998) xx therefore took θᾶσσον as an Atticism arising in transmission; cf. 177* μέζον. 213 οἰωνὸν δ’ ἐνόει τανυσίπτερον, αὐτίκα δ’ ἔγνω: Apollo gets help from augury – one of the major parts of Greek divination and hence an area of his expertise.279 Augury comprised noticing a bird, particularly a large one (τανυσίπτερον), recognizing it (particularly its call or flight) as significant, then interpreting the sign. The first two actions are as normal rolled together in ἐνόει (an aorist is more common: read ἐνόησε τανύπτερον?); ἔγνω introduces the authoritative interpretation (as in S. OR 395–8, E. HF 596–7, X. Mem. 1.1.4). Such signs were normally ascribed not to the bird’s volition but to a god (see e.g. X. Mem. 1.1.3–4, diviners ὑπολαμβάνουσιν οὐ τοὺς ὄρνιθας … εἰδέναι τὰ συμφέροντα τοῖς μαντευομένοις, ἀλλὰ τοὺς θεοὺς διὰ  τούτων αὐτὰ σημαίνειν). Here, however, the bird appears of its own accord with its signal for a god; so too at e.g. Hes. fr. 60, H Dem. 46. The fact that Apollo receives the omen as he sets out on a new part of his journey marks it as a σύμβολον, which makes it part of a sequence within Herm. between 30* and 295*; Apollo will recognize as much at 303*. This omen is the main item in Apollo’s hunt which has no correlate in the episode of Hermes’ cattlerustling (Introduction §3.2): prophecy is exclusively Apollo’s for now, though it will turn out not to lead him directly to a solution. Later in the hymn, prophecy will become one of the main areas of Hermes’ rivalry with his brother (467–77), and auspices will recur at 544–6. 214 Φιλήτην γεγαῶτα, Διὸς παῖδα Κρονίωνος: I interpret ‘that the Brigand had been born, the son of Zeus son of Kronos’, with Φιλήτης as a name for Hermes (as in Hellanicus FGrH 4 F19b; for the word, see 67*). The line could also, if performed without a comma after γεγαῶτα, mean ‘that the son of Zeus had been born a brigand’. In Epos, perfect forms of γίγνομαι almost always relate to birth, which is clearly relevant here: hence ‘that the son of Zeus had come to be the brigand’ and ‘that the son of Zeus had been the brigand’ are less plausible interpretations. Διὸς παῖδα Κρονίωνος is a substitute for Διὸς καὶ Μαιάδος υἱόν, unique to Herm. and placed for structure (Introduction §3.2). It develops 208 παῖδα, as 279

See 544* and e.g. Od. 15.526, Plu. Mor. 405d with Amandry (1950) 57–9. For sources on Greek augury more generally see Bouché-Leclercq (1879) 127–45, Pollard (1977) 116–29, Dillon (1996), Posidippus 21–35.

266

C O M M E N TA RY 2 1 5 – 2 1 6 Apollo identifies that mysterious child’s father; at 228 he will prove to know which son is at stake. In contrast to the treatment of most bird-omens in Greek literature, the hymnist leaves Apollo’s interpretation beyond human comprehension, and only gradually clarifies Apollo’s insight through his subsequent actions; this is the more expressive given the tracks and the Onchestan’s words, which we understand but which do not seem to help Apollo.280 215 ἐσσυμένως δ’ ἤϊξεν: Used of flying divinities at H Dem. 449 and Hes.(?) POxy. 2509.1. Apollo might fly too; at any rate, his journeys here and at 227 (also with ἤϊξεν) are presented as practically instantaneous. For the structural role of ἀΐσσω in this section see Introduction §3.2. Διὸς υἱὸς Ἀπόλλων: Metrically equivalent to (ϝ)ἑκάεργος Ἀπόλλων. The choice produces a parallelism with 214 Διὸς παῖδα, as at Il. 7.23–4, Od. 8.334– 5. Friedrich (2007: 80–1) argued that ἑκάεργος is sometimes avoided in Il. and Od., but did not consider that introducing Διὸς υἱός could be desirable. 216 ἐς Πύλον ἠγαθέην: The hymn’s first mention of Pylos; Hermes has gone in that direction (218); 398 finally makes explicit that it is where he hid the cows. The idea that part of the Alpheus belonged to Pylos was traditional (Il. 2.591–2, 5.545, 11.711–12). Strabo explained it using a Pylos near modern Kakovatos (see RE xxiii.2129), but the Iliad does not distinguish this from Nestor’s home in Messenia, i.e. the Mycenaean palace-centre at Ano Englianos and its harbour (Od. 3, Il. 9.153). We can consider why the hymnist wanted to introduce the placename, without tackling the knotty questions of whether Strabo was right and whether different strands of tradition relocated Nestor deliberately or confusedly.281 Pylos may have featured in earlier versions of the myth; Hermes hid the cows in the famous ‘cave of Nestor’ at Messenian Pylos at least in Nicander, to judge from Ant. Lib. 23.5 and Ov. Met. 2.684. The association, whether first in Herm. or before, may be conditioned by the numerous myths involving cattle-rustling and Pylos (Nobili 2011: 23–70), which help our hymnist to bring Hermes and Heracles together (Introduction §2.2). Apollo’s decision to go to Pylos is presumably another inference from the bird-omen of 213 to which we are not given access. The omen still will not give him the cows’ exact location. Such inexact directions might more

280

For a similarly abrupt auspice and interpretation see Il. 13.821–3. For Herm.’s representation of Apollo’s prophecy in response to H.Ap. see Introduction §1.3. 281 Frame (2006) elaborated the suggestion that Nestor’s Pylos was ‘moved’ from Messenia, within epic tradition, through Spartan ideology. There was also an Elean Pylos: Coleman (1986). At H.Ap. 424 Pylos is mentioned just after the Alpheus, but is coastal. The list of places there is not sequential (nor are those at 30–44, 216–44), and Messenian Pylos is intended.

267

C O M M E N TA RY 2 1 7 – 2 1 8 comprehensibly come from human witnesses near Pylos, as in Apollodorus’ version (Bibl. 3.113). ἠγάθεος is used of several places where gods have been – Pytho, Nysa, Lemnos, Kyllene (231). Its relevance to Pylos is normally less obvious, and perhaps conditioned by the assonance with Pylos’ other main epithet ἠμαθόεις, but here the adjective gains some support from Hermes’ sacrifice-like actions in 112–41. εἰλίποδας βοῦς: For the anatomical cause of this winding gait, see Körner (1930) 40–1. The epithet is contextually relevant since it affects the depth of the footprints Apollo finds in 218 (cf. Fernández Delgado 1998: 11). 217 πορφυρέηι νεφέληι κεκαλυμμένος εὐρέας ὤμους: Divinely created, non-meteorological clouds occur over twenty times in Epos; they can be at ground-level (e.g. Il. 5.345), but the audience may be imagining Apollo in the air (215*). In the Iliad Apollo causes several of these clouds, both for others (5.345, 20.444, 23.188) and himself (15.308, 16.790, 21.547). Steigerwald (1986) discusses the range of hues of πορφύρα, whose connotation of nobility may be relevant for a god. However, the application of πορφύρεος to waves suggests possible interference from πορφρω ‘surge’ (LfgrE s.v. πορφύρεος, esp. 2a–b), and if this was active in the application to clouds the adjective may also connote Apollo’s anger. Compare Il. 17.547–52, where Athena comes to stir war πορφυρέηι νεφέληι πυκάσασα ἓ αὐτήν and is ἠΰτε πορφυρέην ἶριν: the purple colours forebode bloodshed, but a relationship to a surging stormcloud would also contribute to the terrifying sight. The hymnist never tells us about Apollo’s cloud dissipating: that is the convention for such clouds, with the dramatic exception of Od. 7.143. 218 ἴχνιά τ’ εἰσενόησεν: νοέω once again describes Apollo reacting to Hermes’ actions (cf. 208, 213, 235*). The footsteps’ importance as evidence was prepared at 75–9. There they were found on sand in northern Greece, here on a second patch of soft ground (not necessarily sand, but see 140 and 398* for the cognitive association of ‘Pylos’ with sand even when we are some kilometres inland). However, at 344–5* Apollo will suggest that he saw the prints before reaching Onchestos, and lines 219–26* would suit that sequence of events very well too. The hymnist decided to narrate Apollo’s search only from the moment he reached Onchestos (182–9*), and has perhaps relocated some poetic material which originally belonged further north. Ἑκηβόλος: This, ἑκατηβόλος, ἑκατηβελέτης and ἑκάεργος are favoured in contexts of shooting and, more generally, aggression.282 Herm.’s cluster of 282

The highest concentration is in Il. 1 (eleven occurrences), where Apollo shoots plague-­arrows; at 23.872 Apollo Ἑκηβόλος is invoked for an archery-contest; he is

268

C O M M E N TA RY 2 1 9 – 2 2 1 four such epithets between here and 239 indicates Apollo’s scariness on the warpath. εἶπέ τε μῦθον: For the expectation of authority raised by the formula, see below on 219–26*. The phrase follows ‘noticing’ also at 154, H.Ap. 256, Theb. fr. 3.1.

219–26 219 ὦ πόποι, ἦ μέγα θαῦμα τόδ’ ὀφθαλμοῖσιν ὁρῶμαι: Most soliloquies in Epos start with an exclamation. ὦ πόποι appears only in speeches, and normally first (but see 309*); it is often continued with ἦ, which similarly belongs in direct speech.283 It expresses displeasure, or ironic Schadenfreude, at a new situation. The sense of surprise here is enhanced by the pleonasm ὀφθαλμοῖσιν ὁρῶμαι, which Lechner (1882: 14–21) showed is often expressive – in this case emphasizing the unpleasant reality of the vision. For θαῦμα in Herm. see Introduction n. 101. 220 ἴχνια μὲν τάδε γ’ ἐστί: Similarly Il. 20.344–6 ὦ πόποι, ἦ μέγα θαῦμα τόδ’ ... ἔγχος μὲν τόδε ... In both passages, the repetition of ὅδε (also here in 222) is effective after ὦ πόποι in suggesting the speaker’s surprised fixation on a new stimulus. βοῶν ὀρθοκραιράων: ὀρθο- primarily means ‘upright’ rather than ‘straight’ in this compound, which also applies to ships, which in the Geometric period curved up at either end (Morrison and Williams 1968: 37, plates 1–8); compare e.g. ὀρθοπάλη ‘upright wrestling’, Dion. Epop. B. fr. 33v.35 ὀρθόκερων ἔλαφον. There is no conflict with 192* κεράεσσιν ἑλικτάς, nor between Od. 12.348 ὀρθοκραιράων and 355 ἕλικες: the horns corkscrew upwards (Georgoudi 1990: 241–2). 221 ἐς ἀσφοδελὸν λειμῶνα: ἀσφόδελος is generally taken as a true asphodel (DGE suggests Asphodelus aestivus or fistulosus), but Theophrastus’ observations shooting things at Il. 9.564, H.Ap. 357 (the culmination towards which H.Ap.’s uses of these epithets point); his bow is mentioned near by at H.Ap. 177, H.Aphr. 151. The epithets occur when he terrifies human warriors at e.g. Il. 5.439–44, 15.231–53, 16.94, 711; Od. 20.278 may hint at the μνηστηροφονία. 283 ἦ occurs in narrator-text at Od. 22.31, marking the suitors’ focalization; cf. Denniston 279–80. The casual tone of ὦ πόποι comes across in the fact that it is not used in conversations between gods and humans, with the rule-proving exception of Od. 13.383, where it shows Athena’s unusual intimacy with Odysseus. Regarding accentuation, Hdn. Gr. i.502–3 advocated ὦ πόποι, and implies that Aristarchus favoured ὦ ποποῖ (with the perispomenon characteristic of many interjections); Eust. Comm. Il. i.154 mentions authorities in favour of ὤ. See West (1990) l.

269

C O M M E N TA RY 2 2 2 that it sometimes grew in wet conditions (HP 1.4.3) suggest that the word covered other species, perhaps more suited to λειμῶνες with their strong connotation of moistness.284 To judge by the Pierian meadow’s other favourable epithets, its ‘asphodel’ was acceptable as fodder (cf. Gp. 19.7.3, Grieve 1931: 72), even if ἀσφόδελος often appears as a cheap human food (e.g. Hes. Op. 41). By Herm.’s time, the ‘asphodel meadow’ probably carried a strong traditional referentiality to the Underworld, as in Od. 11.539, 573, and 24.13, as well as ritual practices such as planting asphodel on tombs (SEG 41.855 ~ Arist. fr. 644 Rose) and adorning Hecate and Persephone with it on Rhodes (Suda α.4299).285 If so, its presence here may hint at Hermes’ future acquaintance with the Underworld meadow through his psychopompic duties (as in Od. 24.13); the same function was alluded to in 210* ῥάβδον. But whereas Hermes herds souls to the asphodel meadow, here he herds the cows away from one, so that the terms of the hint are opaque. 222 βήματα δ’ οὔτ’ ἀνδρὸς τάδε γίνεται οὔτε γυναικός: A careful antithesis with 220, as we move on to the set of prints left by Hermes’ makeshift sandals: |ἴχνια/|βήματα, μέν/δ’, τάδε γ’/τάδε , ἐστί/γίνεται, genitives of the track-maker. ‘Neither a man’s nor a woman’s’ is probably a polar expression approximating to ‘definitely not a human’s’ (parallels in LfgrE i.832.25–70), since Apollo’s surprise is caused by the form of the impressions rather than their size. Nevertheless, the ordering is effective: Apollo’s series of comparanda for Hermes’ tracks starts, like his initial guess in 200, with ‘man’, which is closest to the truth; a woman is less plausible for a cattle-rustler, and begins a sequence of increasingly outlandish guesses which indicate Apollo’s dumbfoundedness. For βῆμα in the sense ‘footprint’ rather than ‘stride, act of stepping’ see Herodotus 4.82 ἴχνος Ἡρακλέος … τὸ ἔοικε μὲν βήματι ἀνδρός …, S. Ichn. 102, 118, Ps.-E. Rh. 205. For the spelling γίνεται see 98*. For the accents on οὔτέ see n. 230.

284

For this connotation see LfgrE ii.1654.44–5. Contrast the argument in Amigues (2002), that ἀσφοδελὸς λειμών must originally have been *σποδελὸς (‘dusty’) λειμών since Asphodelaceae prefer dry habitats. Reece (2007) argued differently for the same early reanalysis. 285 The only other upper-world ‘asphodel’ meadow in Greek seems to be a worldedge battlefield at Or. Sib. 14.144 (5th c. ce?). Suggestions about the origin of the underworld association have included the plant’s odour and paleness, and the importance of its roots (for medicine and cooking). Further Indo-European otherworld meadows: Puhvel (1969), West (2007) 393. Cursaru (2014: 46–62) elaborated on the idea that the vegetation in Hermes’ night-raid is associated with the Underworld.

270

C O M M E N TA RY 2 2 3 – 2 2 4 223 οὔτε λύκων πολιῶν οὔτ’ ἄρκτων οὔτε λεόντων: Apollo continues with a standard trio of animals who might prey on cattle, if not lead them off.286 The ordering progresses away from truth and towards the exotic. Wolves were present in Greece and shared Hermes’ characteristic cunning; bears were probably rarer, though attested for the Arcadian hills, where Hermes was born (e.g. Paus. 4.11.3; also on Taygetos at 3.20.4); in the Peloponnese, lions seem to have been known primarily through oral traditions (Alden 2005). 224–5 οὔτέ τι κενταύρου λασιαύχενος ἔλπομαι εἶναι, | ὅς τις τοῖα πέλωρα βιβᾶι ποσὶ καρπαλίμοισιν: At the end of 224 Ψ had ἔστιν ὁμοῖα while MΨm had ἔλπομαι εἶναι. Θ>x>b retained the variants, while a as usual incorporated the marginal reading. Both seem viable. With ἔλπομαι εἶναι, τι goes with ἔλπομαι and 225 adds a further point about centaurs: ‘nor do I expect for a moment that they belong to a centaur, who does make such monstrous steps.’ The steps of a centaur have the right weirdness in general terms, but even they are not the solution to Apollo’s puzzle; the litotes οὐ … ἔλπομαι as often marks a wry approach to an exasperating situation. ἐστὶν ὁμοῖα requires a small ellipse (‘the steps … are not similar [to those] of a centaur’), parallelled at Hp. AWP 14 τούτων γὰρ οὐκ ἔστιν ἄλλο ἔθνος ὁμοίας τὰς κεφαλὰς ἔχον. Both interpretations suppose that ὅς τις has its secondary function of introducing separate information about an antecedent which is already adequately defined, as in e.g. Thgn. 675–6 κυβερνήτην μὲν ἔπαυσαν | ἐσθλόν, ὅτις φυλακὴν εἶχεν ἐπισταμένως, A. Pe. 744–5.287 After thinking of humans in 222 and animals in 223, Apollo completes his list of non-solutions with a hybrid of the two. For this mode of expressing completeness cf. A. Th. 197 ‘neither male nor female nor in between’, E. Hel. 1137 ‘god or non-god or something in between’. Although in art from c.500 centaurs’ legs are normally all equine, Herm.’s first audiences may well have known the earlier conception where they have human forelegs, and

286

For this trio see e.g. H Aphr. 70–1 (+ panthers), Arist. HA 6.18 571b27, Ps.-Theoc. 25.183–5, SH 900.10, Plu. Mor. 995a, etc. The three possible pairings are also all found – particularly lions and bears without the smaller wolves (Od. 11.611, H.Aphr. 159, Hy. 7.46–7, etc.). 287 Ruijgh (1971: 323–9) and Probert (2015: 280–2, 426) have described the Homeric use of ὅς τις to restrict the set to which its antecedent belongs. Emendations have been proposed with each variant to refer ὅς τις to the creator of the tracks, in line with this restrictive function. Schneidewin’s κένταυρον λασιαύχενα ἔλπομαι εἶναι gives ‘nor do I expect for a moment that whoever is making such monstrous steps is a centaur’ (1848: 676). Kuiper’s κενταύρου λασιαύχενός ἐστιν ὁμοῖος (‘nor is whoever makes such monstrous steps at all similar to a centaur’; 1910: 29 n. 2) introduced a non-epic use of ὁμοῖος + genitive (see LSJ s.v. B2).

271

C O M M E N TA RY 2 2 6 hence particularly odd tracks. Centaurs notionally lived near Pelion, and on Pholoe not far north-east of Apollo’s location. Like the animals of 223 they belong – partly like Hermes – in mountain wildernesses (esp. caves) and could be a threat to cattle, but as predators rather than cattle-rustlers; unlike the animals, they belong only to legend.288 The wildness of Apollo’s climactic candidate is enhanced by the rarity of λασιαύχενος (x4 in archaic and classical Greek, always of wild animals). The adjective suggests an equine mane (S. Ant. 350), though this does not correspond to artistic portrayals. For ποσὶ καρπαλίμοισιν compare Isoc. Helen 26 Κενταύρους … οἳ … τάχει … διενεγκόντες.289 They also suit πέλωρα, whose core associations were the awe-inspiring quality of the preternatural – especially of large size, which could be relevant to the traces of Hermes’ ad hoc shoes; the stem’s frequent application to portentous creatures fits well in the context of struggling to interpret an unusual sight, and is amusingly unsuited to Apollo’s own brother, with whom he will find so much in common. The variant ἐστὶν ὁμοῖα seems to have influenced Batr. 170b καὶ μύες Κενταύρων μεγαλαύχων ἦσαν ὁμοῖοι. The line is interpolated (170a–b are a doublet for 171, incorporated variously before or after it, and μύες is metrically unusual; see Glei 1984; 39–45) and can therefore hardly be an independent reflex of an otherwise unknown epic formula, or the source of Herm.’s variant. 226 αἰνὰ μὲν ἔνθεν ὁδοῖο, τὰ δ’ αἰνότερ’ ἔνθεν ὁδοῖο: The root αἰν- generally applies to the gut emotions of fear, anger and pain, and their external causes; derivatively (esp. in the form αἰνῶς) it can approach ‘excessive’, but rarely just ‘strange’.290 The cows’ footprints are infuriating because Apollo can identify them but not understand their direction; the others are the more infuriating because he cannot understand their form at all. Furthermore, Hermes’ tracks are πέλωρα (225); αἰνός and πελωρ- are applied to similar animals, and occur in proximity also at e.g. Il. 8.423, Od. 10.219 (> Call. H.Art. 51), Cypr. fr. 32. Apollo’s antithesis is rather laboured, to express the impasse he has reached: though both ἔνθα … ἔνθα and ἔνθεν … ἔνθεν can govern a partitive genitive of location (e.g. 357, Il. 5.222–3, H.Pan 22; see Allen and Halliday), such a genitive is not normally repeated. For the disparity between single

288

Crudden (1994) 24. For centaurs in art see Schiffler (1976) 163–8. Centaurs preying on cattle: see e.g. ‘Palaephatus’ Incr. 1, interpreting κένταυρος as ‘bull-spearer’. Cheiron and Pholos lived in caves: Pi. P. 3.63, 4.102–3, 9.30, I. 8.41, Theoc. 7.149, etc. 289 For βιβᾶι see 149* προβιβῶν. In Epos, forms of βίβημι and βιβάσκω are o­ ften immediately preceded by a neuter plural adjective (e.g. μακρά, κραιπνά, καλά). The poet uses πέλωρα to vary this while rephrasing κραιπνά. 290 αἰνῶς marks a likeness as ‘preternaturally’ close at Il. 10.547, Od. 1.208. δεινός passes more readily from ‘fearful’ to ‘strange’.

272

C O M M E N TA RY 2 1 9 – 2 6 ἔνθεν meaning ‘from there’ and correlated ἔνθεν meaning ‘on one side’ see Lejeune (1939) 313, 402–3. For ὁδός of a non-constructed path or route (here sandy ground) see Il. 16.374, LfgrE s.v. 1b. 219–26 Apollo’s reaction to the tracks: Analyst critics found a number of problems with this episode, which they took to have been spliced into the story from an originally separate version: (i) Apollo should have seen the tracks in the sand earlier, as implied at 342–54; (ii) he should not be surprised, since the Onchestan has explained that the cows were walking backwards with company, and since his augury seems to have pointed him towards Pylos (217*); (iii) he should be able to follow the tracks back to the cows’ hiding place.291 One may suspect that such a speech occurred further north in other tellings of the story, but its appearance here is not botched. (i) Meeting Apollo suddenly at Onchestos wastes no narrative time on aimlessness, but allows him to be slow to the scent (182–9*); a sense that Apollo ‘should’ have seen the tracks nearer to Pieria is a productive sign of Apollo’s failings as a searcher.292 The different sequence at 342–54 suits Apollo’s tendentiousness there (341–2*), and cannot be taken as authoritative. (ii) It is still psychologically plausible for Apollo to voice his surprise (with more gusto than narrator-text can muster) when he actually sees the cows’ footprints pointing the wrong way, beside Hermes’ mysterious traces: in fact, 220 γ’ throws emphasis upon τάδε and so sets up a contrast, implying that Apollo notices both tracks together, and they make a contrasting pair of signs – one interpretable but misleading, the other uninterpretable.293 The conjunction makes it convincing, despite the distance from Pieria, that these are indeed the tracks of his cows. (iii) The ground is probably not soft everywhere, so he cannot follow the tracks backwards to the cows (compare 352–4). Epos’ five other divine soliloquies are all decisive (Il. 17.200, 442, Od. 5.285, 376, H.Ap. 287). All are described as μῦθοι in the speech-introduction, and the formula εἶπέ τε μῦθον (218; x11 in Epos) elsewhere introduces story-telling or authoritative speech. Even human soliloquies which externalize confusion generally involve some sort of resolution, such as that of Achilles to fight people other than Aeneas (Il. 20.343–52) or to fight Lycaon (21.53–63); exceptions, such as Odysseus’ curtailed deliberations at Od. 5.298–313 and 407–25, suggest extreme vulnerability. Though Herm. uses μῦθος less specifically than earlier epic texts (201*), the formula εἶπέ τε μῦθον here may draw attention to how Apollo’s 291

See e.g. Matthiae (1800) 256. Ludwich (1887a: 327) gave an early unitarian riposte. See 218 ἴχνιά τ’ εἰσενόησεν*. Romani (2008) 90 also read the delay in Apollo’s seeing the footprints as a narrative strategy to show his difficulty in tracking Hermes. 293 This contrast will recur implicitly when Apollo’s oracular answers are juxtaposed with augury and with the oracular flight of bee-maidens: 541–9*, 560–3. For the footprints see also 75–8*, 79–86*. 292

273

C O M M E N TA RY 2 2 7 – 2 2 8 crescendo of exasperated bemusement so signally fails to match the resoluteness of that formula or of a traditional divine soliloquy. Hermes has, as it were, placed Apollo outside a god’s familiar experiences, as covered by traditional diction.

227–34 227 ὣς εἰπὼν ἤϊξεν ἄναξ Διὸς υἱὸς Ἀπόλλων: Apollo recovers from his bewilderment. The repetition of ἤϊξεν … Ἀπόλλων from 215 gives the sense of returning to purpose, i.e. to the thread of information gained from the old man (210–11) and the subsequent bird-omen (213–14). Thereafter Apollo went ἐσσυμένως. Here prompt action is part of the traditional usage of |ὣς εἰπών and the rarer lengthened forms |ἤτοι ὅ γ’ ὣς εἰπών and |ἣ μὲν ἄρ’ ὣς εἰποῦσα, which are all generally followed within the line by a main verb expressing the speaker’s next action.294 Nevertheless, Apollo’s fruitless detour to Pylos before going back to Kyllene, and the opacity of what the omen told him, give a sense that Hermes has reduced him to a flustered and fruitless zig-zag. 228–9 Κυλλήνης δ’ ἀφίκανεν ὄρος καταειμένον ὕληι | πέτρης ἐς κευθμῶνα βαθύσκιον: The move to Kyllene shows that Apollo understood from the bird-omen which son of Zeus was relevant to the theft (see 214*); he knows Maia’s address. With βαθύσκιον compare 6 ἄντρον ... παλίσκιον: in both, the shadows underline secrecy; the end of 229 also echoes ἔνθα and νύμφη from 6–7. Apollo is clearly in the right place. Line 228 echoes H.Ap. 225 Θήβης δ’ εἰσαφίκανες ἕδος καταειμένον ὕληι (Introduction §1.3). There the formula καταειμένον ὕληι (x5 in Epos) contrasted with Thebes’ later roads (226–8); so at Od. 19.431 it helps emphasize the denseness of foliage around the boar’s den (cf. 19.439–44).295 Here, Apollo has left the ὁδός (188, 212, 226) for the pathless mountain, which is inconvenient for travelling and searching. His focalization of the woods as a concealing cloak – like Hermes’ swaddling (compare 153 + 156 ἐπιειμένε, 237–8) – contrasts with the parallel moment in Hermes’ journey where, despite the time, Kyllene had δῖα κάρηνα (142; Introduction §3.2). Hermes’ birthplace next appears as a κευθμών – a cavern suited to the birth of Ouranos’ rejected children (Hes. Th. 158), a snake (fr. 204.130), Eurytion (Stes. fr. 9.7 ἐν κευθμῶνι πέτρας), or wild boars (Od. 10.283), and therefore, in

One noteworthy exception is the formula ὣς εἰπὼν ὤτρυνε μένος καὶ θυμὸν ἑκάστου, where the main verb is not a separate action. The family of three formulas (with their minor variants) occurs significantly less in Od. (x49) than in Il. (x91), given the number of speeches in each epic. 295 καταειμένον ὕληι modifies an ὄρος also at Od. 13.351, H.Aphr. 285. Contrast the metrical alternative μέγα τε ζάθεόν τε (H.Aphr. 258). 294

274

C O M M E N TA RY 2 2 9 – 2 3 1 Apollo’s eyes, a wild and inappropriate venue in which to hunt down someone with pretensions to becoming Olympian. The potential threat of entering a dark and claustrophobic cave is particularly evident in Odyssey 9. Apollo will describe his perception of Kyllene at 358–60*. For the double expression of the place reached, see LfgrE ii.1174.62–4, omitting Il. 13.20, 14.230, Od. 10.81 (appositions), and adding Il. 9.414, Od. 3.488 = 15.186, H.Ap. 282. The second expression is generally more specific, as here. 229 ἔνθά τε: For ἔνθά τε introducing past mythological events see Il. 2.594, Od. 19.178, Nic. Alex. 104 (ἔνθά τε νύμφη|), SEG 35.1233.17 (Lydia, 148/9 ce); it is a use of τε which Ruijgh struggled to explain (1971: 479, 481, 908–9). 229–30 νύμφη | ἀμβροσίη ἐλόχευσε Διὸς παῖδα Κρονίωνος: Divinities are not often ‘ambrosial’, but see Ar. Av. 1320 (Graces), Moiro 2.2 G–P (naiads), and Dionysus’ nurse Ambrosia (LIMC s.v. Lykourgos I). The relevant associations are cleanliness and sweet smell (Od. 4.445–6, Hy. 7.37, Cypr. fr. 4.5–6, Thgn. 9, etc.; see 231 ὀδμή*) as aspects of female attractiveness.296 The suspicious focalization of 228–9 is giving way to the more alluring sensations which greeted Zeus on his nocturnal visits. Van Nortwick (1980) documented the connection between the language of 229–31 and that of seduction scenes. Seduction, however, can also be deceptive, and ἐλόχευσε was perhaps chosen to hint at ἐλόχησε, given the context of lying in secret.297 Διὸς παῖδα Κρονίωνος echoes 214* and suggests that Apollo is homing in on the target identified there. 231 ὀδμὴ δ’ ἱμερόεσσα: For the fragrance of Maia’s cave see 65 εὐώδεος ἐκ μεγάροιο*. The redolence of places is often, as here, described when someone first arrives there (LfgrE ii.1092.12–13) – a species of focalization – but is also typical of divine epiphanies (e.g. H Dem. 277–8, Hy. 7.37, Thgn. 9, E. Hipp. 1391; Richardson 1974: 252). The ‘desirable’ smell contributes to the transition away from Apollo’s initial dark-tinted perception of Kyllene in 228–9. Crudden (1994: 12) helpfully contrasted the smoke which is the first signal of arrival at a human habitation (airborne, perceivable at a distance; dirty, acrid; I agree with DGE that ἀμβρόσιος was not, synchronically, just a substitute for ἄμβροτος. The most convincing case of their equality is νὺξ ἀμβροσίη (e.g. Il. 2.57), despite the observation in Roscher (1884–90) 280–2 that ambrosia ~ honey ~ honeydew ~ dew. Hermes’ golden sandals (24.341) and the Muses’ voices (Hes. Th. 69), for example, may be explained by the connection of ambrosia to oil and honey respectively: see n. 314, West (2007) 90. 297 Van Nortwick argued less cogently that this language is continued in 247–50, even though by that stage it is clear that Apollo is in no mood for seduction.

296

275

C O M M E N TA RY 2 3 2 see Od. 1.58, 9.167, 10.99, 149) and the smell which is the first sign of arrival at divine homes (airborne, perceivable at a distance; colourless, sweet). δι’ οὔρεος ἠγαθέοιο: The adjective reaffirms divine presence. Mountains including Kyllene were common cult-sites, and may be ζάθεος.298 However, ἠγάθεος itself describes a mountain only here in Epos, and was presumably chosen over epithets for mountains which occur in formulas in other cases (e.g. ὑψηλοῖο, εἰνοσιφύλλου, οἰοπόλοιο).299 This markedness draws attention to the repetition from 216, which perhaps contrasts Apollo’s fruitless trip to Pylos with his current more promising destination. 232 κίδνατο: According to a standard view, Epos used σκίδναμαι unless κίδναμαι was metrically desirable. Hence Lobeck (1853: 125) proposed σκίδνατο; cf. Shipp (1972) 43; for optional σ-, see Schwyzer i.334. However, there may have been some semantic distinction, since Epos avoids using κίδναμαι of discrete scattering (first A.R. 2.978) but is happy to use σκίδναμαι for it. κίδναμαι is sometimes used outside Epos without a metrical consideration, e.g. Pi. fr. 129.10 ὀδμὰ δ’ ἐρατὸν κατὰ χῶρον κίδναται. Θ’s κιδνᾶτο may reflect the view that (σ)κίδνημι was a form of (σ)κιδνάω (e.g. Hdn. Gr. i.451); for alternations such as βιβάω/βίβημι see 149*. πολλὰ δὲ μῆλα ταναύποδα βόσκετο ποίην: The wild mountain on closer inspection resembles an upland pasture with signs of domestication. Presumably the sheep belong to Maia and her ἀμφίπολοι; one can interpret them as the source of Hermes’ gut-strings at 51.300 Such activity in Maia’s household is appropriate to Hermes’ later herding duties as Ἐπιμήλιος (571*; cf. 314* οἰοπόλος). From Apollo’s point of view, however, his search has ended with the wrong flock – so near and yet so far. Whereas the cows were εἰλί-ποδες (216), the sheep are ταναύ-ποδα ‘with tapering lower legs’, perhaps recalling H.Ap. 304 and in any case chosen against H.Ap. 413 βαθύτριχα. βόσκομαι was applied to the cows at 72 (cf. 193); it is now transferred to the sheep. For such echoes see below, 227–34*. When Hermes encountered the wrong animal grazing in the same location (27*), he was able to turn it to much better account. The formation of ταναύποδα is explained either from ταναϝο- by vocalization of digamma (GH i.33), or from a hybrid of unmetrical τανύποδα and

298

Hes. Th. 2, H.Ap. 233, H.Aphr. 258. See e.g. Buxton (1992) 5, Langdon (1976) 100–12; ‘Arkadian mountain peaks seem to have been exceptionally well endowed with sanctuaries’ (Jost 1994: 218). 299 SH 906 has an ἠγάθεος mountain: see below, 227–34*; in Orph.H. 48 Tmolus is ἠγάθεος in the context of Dionysus’ arrival. 300 For herding nymphs see e.g. Od. 14.435, Semonides fr. 20, Orph.H. 51.12–13. Nymphs called Epimelides are related to flocks from the 2nd c. ce on (e.g. Paus. 8.4.2; probably already in the background of Apollo Epimelidios in Hellenistic Camiros: Tit. Cam. 35), though their name was also derived from ‘apple tree’ (RE vi.172.49–63).

276

C O M M E N TA RY 2 3 3 – 2 3 4 ταναός (Szemerényi 1964: 159). The metrician Heliodorus advocated the spelling τανποδα (Apion fr. 135), inherited in Mp here and at H.Ap. 304 (as in some manuscripts at Od. 9.464). This suggests the scholarly engagement of one of the Homeric Hymns’ early scribes, and a second erudite intervention by the scribe of Θ in this single line. 233–4 σπεύδων κατεβήσετο λάϊνον οὐδόν | ἄντρον ἐς ἠερόεν: The cave naturally has a stone threshold (see 23* for its importance), but the formula λάϊνον οὐδόν has a traditional application to the thresholds of the temple at Delphi.301 The phrase therefore contributes to the connections between Maia’s cave and temples (60–1*, 148* νηόν, 248–51*), and suggests a more specific comparison between Apollo entering this home and Hermes threatening to enter Delphi at 178–81. This raises the question whether Apollo is losing the moral high ground by barging into Maia’s home uninvited (see 246–7*, 370– 4). As in 228–30 versus 231–2, this presentation of Maia’s cave is juxtaposed to Apollo’s less positive focalization: for ἠερόεις and the Underworld see 172*; here it adds a new hint of danger to the wildness implied in 222–5 and 229. Waiting to be received is a standard element in the type-scene of arrival, following the description of the place and people reached, and so expected in this position (Reece 1993: 5–21). Apollo’s refusal to wait (σπεύδων) is a statement of intent. On the ending of κατεβήσετο see 99*. 234 ἑκατηβόλος αὐτὸς Ἀπόλλων: αὐτός may mean ‘the great’: compare e.g. H Dem. 317, H.Ap. 181, Ar. Nu. 219 with Dover (1968), Thphr. Char. 2.5. This verges on the more common use whereby a god is said to act ‘himself ’ to convey particular forcefulness and authority (LfgrE s.v. αὐτός I.1a). Apollo’s aggression is also implied in ἑκατηβόλος (218*). The phrase acts like a nominative ‘declension’ of the formula ἑκ(ατ)ηβόλον Ἀπόλλωνα+|, but as at H.Aphr. 151 ἑκηβόλος αὐτὸς [‘even’] Ἀπόλλων|, the pronoun’s semantic force is relevant. The alternative idea that αὐτός means ‘in person’ and implies ‘without his cloud disguise from 217’ (e.g. Schulze 1868: 29) is far-fetched. 227–34 Arriving at Kyllene: This arrival encapsulates many of the important events which have preceded. The lines share several features with 184–9, so that the pair frame Apollo’s journey. The two passages describe 301

Il. 9.404, Od. 8.80, H.Ap. 296; derivative uses in Parm. 28 B1.35, Bacchylides fr. 4.21. See Ramersdorfer (1981) 193–4. Normal thresholds were wooden throughout the archaic period; sometimes the formula applies to palaces, which had stone thresholds in the Mycenaean period (Wace 1962: 496–7). Like other features, the threshold of Eumaeus’ hut (Od. 16.41) is borrowed from traditional descriptions of posher abodes.

277

C O M M E N TA RY 2 3 5 – 2 3 7 arrivals (ἀφίκανεν¦ + location, ἔνθα, description of place and inhabitants); both times, Apollo finds animals other than his cows (188 κνώδαλον, if the text is right; 232 μῆλα). Apollo’s arrival simultaneously recalls Hermes’ two exits from the cave (22–8, 64–74): all three involve grazing animals (27, 72, 232 βοσκ-; 27, 232 ποίην, 72 λειμῶνας) whose legs we hear about (28, 216, 232); pairwise, 231 and 65 share the cave’s fragrance, while 233 and 23 share its threshold. Hermes, unlike Apollo, was delighted with the animals he found. Finally, 229–30* replays 6–7 as a description of Hermes’ birth. These lines seem to have influenced the poet of the hexameters SH 906 (for which see Perale (2014); he does not make the connection). The ends of lines 2–4 survive as ἱερὴ δ’ ἀποκίδναται ὀδμή (cf. 231–2), πολυπτύχου ἠγαθέοιο (cf. the rare application of ἠγάθεος to mountains in 231), ἑκατηβόλου Ἀπόλλωνος (cf. 234). The context appears to be unrelated. For a possible relationship to Od. 5 in this section see Introduction §1.3.

235–45 235–6 τὸν δ’ ὡς οὖν ἐνόησε Διὸς καὶ Μαιάδος υἱός | χωόμενον περὶ βουσὶν ἑκηβόλον Ἀπόλλωνα: νοέω has recently described Apollo and the Onchestan reacting to Hermes’ actions (208, 213, 218). The switch of subject signals a new dynamic, where Hermes has to take on the reactive role. The lines make full use of the energy of the formulaic pattern ‘|τὸν+ δ’ ὡς οὖν ἐνόησε | ’ (Il. x10, Od. x2, here; x9/13 a noun is added, here Ἀπόλλωνα). The hyperbaton between |τόν .. and the participle at line-start convey the dramatic ‘revelation’ of the thing noticed and its nature – a species of implicit focalization. Focalization is also implicated by ἑκηβόλον Ἀπόλλωνα following 234 ἑκατηβόλος αὐτὸς Ἀπόλλων: the repeated identification is otiose for us, but not for Hermes.302 For περί + dative with an expression of anger, see Sc. 12 |χωσάμενος περὶ βουσί (a late formula?), E. Hel. 1342–3, A.R. 1.1340–3. The development seems to have been ‘stand over (protectively)’, then ‘fight over’, ‘be anxious about’, then ‘be angry about’. Similar developments can be seen in Il. 23.88 ἀμφ’ ἀστραγάλοισι χολωθείς, ἀμφί + genitive (172*), and περί + genitive. See GH ii.127–9. 237 σπάργαν’ ἔσω κατέδυνε θυήεντ’: In the type-scene of arrival and hospitality the guest waits to be noticed; often, the host’s (youngest) son notices, and rushes to welcome the guest (see Reece 1993). This is pointedly negated:

For Grice’s maxim of relevance and implicature see n. 249. ἑκηβόλος may suggest Apollo’s aggression: 218*.

302

278

C O M M E N TA RY 2 3 7 Apollo has not waited (233–4*), and upon noticing him Hermes shrinks. The poet leaves it open for now whether this is rational fear (compare e.g. H.Aphr. 181–3) or a ploy; either way, we are far from Astyanax’s instinctive shrinking from the gleam of Hector’s helmet (Il. 6.467–70). κατέδυνε is neatly chosen: δύ(ν)ω very often results in invisibility (which Hermes seems to want just now), often connotes subtlety of motion (e.g. Il. 10.221, Od. 22.334), and often applies to donning armour. This gives a sense of Hermes improvising with the materials available in his conflict with Apollo.303 Once again he exercises unusual dexterity in manipulating his swaddling rather than being constrained by it (145–54*). Gods and humans used aromatic plants both in ritual and domestic settings, in the latter case to perfume clothes and rooms.304 The Greeks comment on how swaddling can all too easily come to smell unpleasant, e.g. A. Ch. 756–60, Diphilus fr. 73; Sor. Gyn. 2.31 advocates pressing a baby’s tummy during a bath, to make it urinate while unswaddled. The god’s swaddling is marked as aesthetically superior. In Il., Od. and Hesiod, θύω, θύος and θυήεις (< *θυεσ-ϝεντ-) refer to scents caused by heating in a ritual context (x13), whereas θύον, θυόεις, θυώδης, and τεθυωμένος have non-ritual contexts (x6). However, Mycenaean tu-we-a seem not to be specifically ritual, and the distinction blurs again from H.Aphr. 58–9 on. It is probably coincidence that Hermes’ xoanon on Kyllene was made of θύον (‘citron wood’, Paus. 8.17.2). 237–8 ἠΰτε πολλήν | πρέμνων ἀνθρακιὴν ὕλης σποδὸς ἀμφικαλύπτει: One can interpret ‘as wood-ash covers over a large heap of charred logs’, i.e. to keep them smouldering and to avoid having to relight a fire, or ‘as fine earth from woodland covers over a large charcoal clamp of logs’. The former is preferable for four reasons. (i) It uses the more common sense of σποδός (the sense ‘soil’ is extant first at Hdt. 2.140.1), and one which interacts with 237 θυήεντα, where fragrance also derives from heated plant-matter. (ii) It has an easier genitive ὕλης (of material: see Allen and Halliday). (iii) It recalls line 140, where Hermes tried to extinguish the ἀνθρακιή in his cooking-pit, and so fits a pattern of him redeploying actions from his theft more defensively, to deflect Apollo’s anger (see 240* συνέλασσε). (iv) Crucially, it expresses a contrast of sparky interior and dormant exterior, whereas charcoal clamps

For the further relevance of formulas for ‘entering Hades’, which often involve ἔσω and a verb prefixed by κατα- or δύ(ν)ω, see 256*. 304 At Il. 6.483, clothes are κηώδεα (~ καίω), scented with heated logs. The process is depicted, probably in a ritual context, on an oenochoe by the Meidias Painter of c.415 (New York, Metropolitan Museum 75.2.11). Clothes could also be imbued with perfume (Cypr. fr. 4; probably νεκτάρεος at Il. 3.385, 18.25). For subsequent methods, see Lilja (1972) 49. 303

279

C O M M E N TA RY 2 3 9 – 2 4 0 (ἀνθρακιά at Thphr. HP 5.9.4, 9.3.1–3) tend to smoke and require constant attention (Kelley 1986: 5–7, Olson 1991). The same vehicle is used at Od. 5.488–93 to contrast Odysseus’ sleeping exterior and the spark of life within him (Introduction §1.3), and intriguingly at Sotades Com. fr. 1.28–9 of fish being cooked after being ‘swaddled’ in leaves (Vergados 112). πολλήν suggests Hermes’ bulk compared to his swaddling; so might πρέμνων, since it normally denotes the lower part of the trunk (Schmidt 1878: 459–62). The subject of ἀμφικαλύπτω can be clothing (Il. 2.262, Hp. Mul. 203), which suits the tenor. But the subject is most often a literal cloud or a metaphorical one of death or sleep; the last contributes to the implication that Hermes is pretending to be snuggled up, rather than shrinking in fear; this will become clearer in the following lines. 239 Ἑκάεργον: The second element was often understood correctly from ἔργον, sometimes from εἴργω (e.g. Heraclitus Q.Hom. 7.8 vs Ap.S. p. 65). The first interpretation suits the cluster of ἑκα- epithets, which seems to express Apollo’s militant power (218*). ἀνεείλε’ ἓ αὐτόν: Ω’s ἀλέεινεν means ‘shrank away from, avoided’ (never ‘hid’). It therefore cannot govern ἓ αὐτόν, since the pronouns are too emphatic to refer to Apollo (see Od. 8.396 with Garvie 1994; contrast Gigliotti 1991). Lohsee’s recourse to ἀνέειλε (1872: 27), refined by Postgate, gives excellent sense. This family of verbs (εἴλω, εἰλέω, ἴλλω, etc.) seems to have fused three main senses – ‘press together’, ‘make curved’, ‘encircle’ – all relevant to swaddling as tight encirclement. εἰλέω is used of swaddling at 306 ἐελμένος* and e.g. Sor. Gyn. 2.14.1 τὸ βρέφος … κατειλημένον ἄχρι τῆς ὀσφύος ῥάκεσι. The reflexive here may emphasize the surprise that Hermes can encircle himself (see 145–54* on his self-swaddling) – an incomparable point which the vehicle of the simile fails to capture. ἀνα- may suggest ‘in retreat’ (so Th. 7.81.4 ἀνειληθέντες ‘backed into a small space’), or ‘upwards from the feet/back’. The rarity of ἀνειλέω and allure of Ἑκάεργον … ἀλέεινε explain the corruption. Alternatively, ἀνεείλεεν αὐτόν would avoid the elision before ἕ, where the digamma has a strong tendency to be observed (Wackernagel 1916; 107–9), though the digamma of οἷ/οἱ is neglected at e.g. 143, Od. 11.442; for reflexive αὐτός, see LfgrE i.1666.16–35. The tenor continues through to 241–2*, which restate the contrast of dormant exterior and active interior. Hence a comma is needed at the line-end, not a hard stop. 240 ἐν δ’ ὀλίγωι συνέλασσε κάρη χεῖράς τε πόδας τε: Soranus (Gyn. 2.15) underlines the importance of swaddling a baby with its arms stretched by its sides and its legs together. συνέλασσε κάρη may suggest that Hermes scrunches his shoulders towards his neck. The action corresponds in the

280

C O M M E N TA RY 2 4 1 simile to the embers being raked together, but that would happen before they are covered – a further inconcinnity to mark Hermes’ transcendent use of swaddling. Monbrun (2007: 118) observed that the line could describe a tortoise adopting a defensive posture, so that it may contribute to the connections between Hermes and his tortoise (n. 55); admittedly a tortoise retracts its limbs and head in a different way from Hermes. The application of ἐλαύνω to body-parts is very unusual, and evokes more normal uses, in particular 106 συνέλασσεν of Hermes herding the cattle. Swaddling, like herding, was thought to impose organization on sub-human creatures (Gourevitch 1995, esp. 250). Hermes ‘herds’ himself and resembles a covered fire to deflect the anger caused by his literal herding and fire-­ covering – all so that he can ‘herd’ himself and Maia (167). 241–2 φή ῥα νεόλλουτος προκαλεόμενος ἥδυμον ὕπνον, | ἐγρήσσων ἐτεόν γε: First, the text. Ω began 241 with δή ῥα, words which wrongly imply that Hermes actually is newly washed and summoning sleep, and which only modify conditionals and temporal adverbs in Epos. Hermann’s φή leaves a second simile nested inside the tenor of the first, a rare feature paralleled at e.g. Il. 5.554–60. ῥα (cf. 212 |φῆ ῥα) invites the audience to visualize the image.305 Substantivized νεόλλουτος in such a context is justified by the formula κακὸς ὥς. I prefer not to compromise with xm’s nonsensical variant θῆρα νέον λοχάων. θῆρα could have arisen as a sonic deformation of δή ῥα, or a visual one from φή ῥα or φῆρα. The corruption ἄγρης· εἰνέτεόν τε (without following δέ) immediately below may have encouraged this. The origin of λοχάων from -λουτος is less comprehensible – perhaps an intrusive gloss ἐκ λοχίων? As for the corruption in 242, even if εἰνέτεος did mean ‘nine years old’, its relevance to the tortoise would be unclear. Martin (1605: f. 21) restored ἐγρήσσων (see Radermacher 1981 for parallel corruptions) and Hermann ἐτεόν γε … δ’. Eitrem (1906: 267) noted that Hermes’ first bath was the subject of local legend at Trikrena (Paus. 8.16.1), though the origin of this is undatable. More generally, a first wash immediately after birth was a prominent association in Greek comments about babies (Ginouvès 1962: 235–7); bathing and being swaddled were states of babyhood often mentioned together.306 As bathing

305

Bakker (1993) 15–25, charting a middle course between the overly restrictive emphasis on liveliness of tone in Denniston 33–5 and Grimm’s focus on appealing to the audience’s field of experience (1962; see LfgrE s.v. ἄρα I 2a). ἄρα is not normal after the hinge of a simile. 306 268, H.Ap. 120–1, Call. H.Zeus 32–3, H.Del. 6, Str. 3.4.17, Plu. Cat. Mai. 20.5, Mor. 265a, Sor. Gyn. 2.13–14, 2.30–5; E. Ion 1493 presupposes the pairing. For other com-

281

C O M M E N TA RY 2 4 2 – 2 4 3 was taken as a way to calm a baby before it was reswaddled and ready for sleep (see Sor. Gyn. 2.30), so Hermes’ feigned sleep is ἥδυμος.307 The logic, however, betrays a mismatch: new-washed babies do not act like Hermes; they look like him after his action. Indeed, ‘summoning’ sleep is more suited to a caregiver singing a lullaby than to the baby (e.g. S. Ph. 827–9, A.R. 4.146). This inconcinnity parallels that between 238 ἀμφικαλύπτει (state) and 239 ἀνεείλε’ (action), and complements the revelation in 242 that Hermes is really awake, his eyes glinting like embers just beneath the dormant exterior. Our sense of Hermes’ agency could be enhanced further by the fact that προκαλέομαι often means ‘challenge’ in legal and other contexts: his pretence of sleep is confirmation that he will not capitulate to Apollo in this phase of their dispute.308 As a baby he is once again incomparable (cf. 145–54*), his appearance the result of calculated acting. Hermes chooses as the domain of his acting one of his adult functions, the control of waking and sleeping (14*). Allen (1897: 260) already compared 242 to Hipponax(?) fr. 177 Ἑρμῆ ... κάτυπνον οἶδας ἐγρήσσειν. Podbielski (1963: 328) emphasized the comic nature of this foretaste of Hermes’ speciality. 242 χέλυν δ’ ὑπὸ μασχάληι εἶχεν: For the lyre under Hermes’ left armpit see 153*. Its mention at this moment suggests that, even if Hermes will not fool Apollo by his sleepy appearance, he will maintain at least one trick up his sleeve. 243 γνῶ δ’ οὐδ’ ἠγνοίησε Διὸς καὶ Λητοῦς υἱός: The conceptual and verbal echo of 235 ἐνόησε Διὸς καὶ Μαιάδος υἱός| perhaps contrasts Apollo’s reaction to his recognition with that of Hermes. γνῶ δ’ οὐδ’ ἠγνοίησε probably alludes to Hes. Th. 551 γνῶ ῥ’ οὐδ’ ἠγνοίησε, in order to cast Hermes as Prometheus and Apollo as Zeus (Introduction §1.3; cf. 256*). The phrasing (‘X and not not-X’) follows a traditional pattern which flouts Grice’s maxim of

ments on babies’ baths see e.g. Hdt. 6.52.6, Hp. Reg. Sal. 6, Lys. 1.9, Ar. Lys. 19, and Horsfall (1971) 1110 on myths of special first baths. 307 Swaddling probably limits sensations which might engender a startle reflex: Gerard, Harris and Thach (2002–3). ἥδυμος, first extant in Alcman PMG 135, probably derived by a combination of the idea of ‘sweet’ sleep (ἡδύς, also γλυκύς as in 8) with νήδυμος (extant in Il., Od., H.Aphr.), an epithet of sleep whose original meaning was unclear. See Brugmann (1900) 278–9; Reece (2009: 41–5), among others, preferred a reanalysis ἔχε νήδυμος < ἔχεν ἥδυμος < ἔχε ϝήδυμος. ΣA Il. 2.2b notes that ἥδυμος is post-Homeric, citing Simonides and Antimachus but as usual not the Hymn to Hermes. 308 I see no reason to emend. Vergados preferred Van Herwerden’s προβαλεύμενος (1882: 1).

282

C O M M E N TA RY 2 4 4 – 2 4 5 quantity and generally adds a perceptible emphasis, here principally overturning Hermes’ hopes.309 Compare Il. 4.22 underlining the strength of Athena’s anger; at 9.70 Nestor is praising Agamemnon before talking him round; 5.816, 8.246, Od. 5.143, H.Ap. 1 and Hes. Th. 102–3 all underline the subject’s willingness; Th. 551 and Hp. Epid. 7.3 overturn the audience’s expectation (Hesiod diverges from a version where Zeus was genuinely tricked; a patient with shortterm memory loss αὐτὸς ἑωυτῶι ξυνήιδει τὸ πάθος, οὐδ’ ἠγνόει).310 The iota in ἠγνοίησε arose, probably under the influence of -νοια, because ἠγνόησε cannot fit in a hexameter: Wyatt (1969) 168. 244 νύμφην τ’ οὐρείην: Maia’s activity is almost entirely located on Mt Kyllene (compare e.g. Simonides fr. 20 οὐρείας; also 2* above); compare her sister Taygete ~ Mt Taygetos, and their father Atlas (a mountain at least by Hdt. 4.184.3); the Pleiades are ὀρειᾶν at Pi. N. 2.11. Land-dwelling nymphs were especially associated with water sources, trees and mountains, but the categorization into naiads, dryads and oreads is hard to trace earlier than A.R. 4.1149–51 and the conjectured contents of Callimachus’ On Nymphs (fr. 413; RE xvii.1532–3): see e.g. H.Aphr. 256–72 for mountain-dwelling hamadryads and E. Hel. 195–6 for mountain-dwelling naiads. καὶ φίλον υἱόν: The formula gains bite from the situation: Hermes is φίλος to Maia, but not to Apollo, whose focalization (as the subject of a verb of perception) is coming in 245. 245 παῖδ’ ὀλίγον: As well as size, the adjective suggests Apollo’s initial undervaluation of Hermes’ significance; he will have to backtrack at 456*. For the connection of ὀλίγος with apparent insignificance see e.g. Il. 1.167, 2.529, Od. 9.515, 20.259, Hes. fr. 43.61, ὀλιγοδρανέων, ὀλιγηπελέων. δολίηις εἰλυμένον ἐντροπίηισιν: Elsewhere in Epos, one εἰλύεται with either tangible wrappings or night. ἐντροπία most likely has the ad hoc sense ‘thing turned around’, corresponding to the literal turning found in ἐντρέπω (LSJ s.v. 1) and ἐντροπαλίζω; for the formation of such a noun in -ία compare ταινία, κειρία. The fact that the lemma could then describe a victim of a snare, but in this instance describes the perpetrator, draws attention to Hermes’ paradoxical recourse to self-binding as a means of controlling a hunting situation (compare 241–2*). Meanwhile the more common sense of ἐντρέπομαι – directing oneself in αἰδώς towards (Il. 15.554, Od. 1.60, often in 309

For Grice’s maxims see n. 249. For Indo-European parallels, see Humbach (1959). Ps.-Hermog. Meth. 37 analyses the pragmatics of choosing ‘not not-X’ over ‘X’; cf. Weyman (1887) 453–91. Many α-privative words, including ἀγνοεῖν, are used mainly after negatives: Wackernagel (1924) 297–9.

310

283

C O M M E N TA RY 2 4 6 tragedy; ἐντροπίη at Hp. Decent. 2) – adds sarcasm focalized through Apollo, since Hermes’ windings are a sign of the opposite of αἰδώς. This is supported by the echo of 151 εἰλυμένος + 156 ἀναιδείην ἐπιειμένε, the other instance where Hermes’ deceptive self-swaddling was redescribed by an adult in terms of brazenness.311 235–45 Confrontation: The two protagonists finally recognize (235, 243) and confront each other. Words signifying them are placed carefully. Straightaway their formulas stand opposed at consecutive line-ends (235–6). In 239 Ἑρμῆς and Ἑκάεργον are juxtaposed even more closely. Apollo’s recognition in 243 echoes Hermes’ in 235. As the two protagonists face off, Maia’s presence as a potential participant in the scene (see 227–34*; LIMC s.v. Hermes 241, Philostr. Im. 1.26.3) is sidelined. Indeed, the detailed echoes of 151–6 (Introduction §3.2) make Apollo take over from Maia: Hermes manipulates his swaddling in a pretence; his lyre nestles under his arm; the adult perceives his ‘wrapping’ in metaphorical terms (a cloak of brazenness or trickery). This raises expectations that Hermes will argue back against Apollo as he did against Maia. His initial position is finely poised on a scale of pro- and re-activity: the words ἐνόησε (235), ἀνθρακιή (238) and συνέλασσε (240) all recall moments from the cattle-theft and from Apollo’s search, but are reapplied to Hermes’ defensive actions; Apollo bursts in, contravening the traditional deference of a visitor to the host’s son upon arrival. On the other hand, both κατέδυνε (237) and προκαλεόμενος (241) suggest that Hermes is gearing up for a fight.

246–53 246–7 παπτήνας δ’ ἀνὰ πάντα μυχὸν μεγάλοιο δόμοιο | τρεῖς ἀδύτους ἀνέωιγε: In Epos, παπταίνω expresses looking around with urgency and/or alertness; Lonsdale (1989) proposed a connection with hunting, which could be relevant here. μυχοί are protected and (especially in a cave) shadowy, and hence secretive – a place to hide oneself (Il. 21.23, Od. 9.236, etc.) or one’s possessions (Od. 13.363). The μυχός is conceptually opposed to the threshold (compare Od. 7.87, 96), which is where we left Apollo at 233–4, so that the curt narrative suggests his directness and single-mindedness in traversing the

Unsurprisingly, many approaches to ἐντροπίηισιν have not admitted such wordplay. LSJ invented the sense ‘dodges’. Van Bennekom’s tentative ἐν στροφίηισι (in LfgrE s.v. ἐντροπίη) introduces the rare combination εἰλύω ἐν and was unlikely to be so corrupted; Gemoll’s εὐτροπίηισι is a noun limited to philosophy.

311

284

C O M M E N TA RY 2 4 7 – 2 4 8 cave. This impression may have been fostered by Apollo’s failure to request entry or to strip off – at least if the laws governing house-searches in classical Athens give a rough guide to the appropriate background of expectations here. There, a citizen could demand to search a property for stolen goods, and the owner probably had to accede in most cases. But the searcher had to remove his cloak, to stop him planting property in an enemy’s house and/or taking in a dagger.312 Moreover, the choice of ἀδύτους – etymologically ‘a place not to be entered’, as the wordplay at H.Ap. 443 ἐς δ’ ἄδυτον κατέδυσε acknowledges – casts Maia’s cave as a temple which Apollo transgresses, as well as a home. ἄδυτον almost always refers to a secluded part of a sanctuary, either an inner storeroom, found in many archaic and classical temples, or a place of oracular pronouncement; they are sometimes caves.313 Apollo’s search of Maia’s cave therefore parallels Hermes’ threatened response (176–81), the ransack of Delphi with its famous oracular ἄδυτον. μεγάλοιο δόμοιο (as opposed to, say, the formula σπείους γλαφυροῖο) echoes 178 Πυθῶνα μέγαν δόμον: Herter (1981) 192. The analogy with temple architecture does not explain τρεῖς: temples have, if at all, a single ἄδυτον; nor do other houses in Epos have three storerooms. Radermacher 124–5 took this as a fairytale motif, citing a Swiss parallel. But the detail is apt for the complicated topology of caves formed in karst landscapes like Arcadia. For the cave-cult on Kyllene see the references on 2*. ἀνέωιγε: West (1998) xxxiii discusses the form. 247 λαβὼν κληῖδα φαεινήν: Metal (‘shiny’) keys are particularly associated in Greek art with priests (Mantis 1990: 28–65): the connotations of temple-­ architecture may continue. For Greek keys and locks see 146* with references. 248 νέκταρος ἐκπλείους ἠδ’ ἀμβροσίης ἐρατεινῆς: Ψ’s ἐκπλείους (‘completely full’) is the lectio difficilior, supported by slightly later parallels such as E.

312

Cf. Isaeus 6.42 for a case of home-owners not acceding to a request to search. Isaeus (tendentiously?) presents this refusal as improper in the words οὐκ ἠθέλησαν τῶν δικαίων οὐδὲν ποιῆσαι. Presumably the law was qualified so as to prevent indiscriminate demands (unlike at Pl. Lg. 12 954a–b). See Ar. Nu. 499 + Σ for the fear of the searcher planting goods, and in general Glotz (1904) 201–8. 313 Hollinshead (1999) cites only this passage and E. Andr. 1034 for ἄδυτα outside temples; in the latter, the text is uncertain and may refer to an oracular room. For ἄδυτα and caves see Pi. fr. 95 with Lehnus (1979) 116 n. 52, E. Ion 938 for sanctuaries of Pan, Hollinshead (1999) 194 n. 22. For other terms specific to temple architecture which are applied to the cave see 148 νηόν*, 233–4*. For ἀδύτους instead of ἄδυτα, cf. Moeris ν.6 νώτους, Egli (1954) 81–6, 110–11.

285

C O M M E N TA RY 2 4 9 – 2 5 0 Cyc. 388, X. HG 3.2.11. However, M’s ἐμπλείους (‘full inside’) has epic parallels at Od. 18.119 and 22.3, and may be original. The image here is of storage jars containing divine counterparts to human staples such as wine, oil and grain. But nectar and ambrosia were not clearly defined for the original poets and audiences. In the Iliad, Odyssey and Hymns, nectar normally has the role of wine (see e.g. Il. 4.3 νέκταρ ἐοινοχόει). Elsewhere it resembles honey, perhaps by extrapolation from wine being improved to Greek tastes when sweetened with honey. Epic ambrosia, as an unguent or detergent, is the divine counterpart to olive oil; as a food, it can be the counterpart to bread, or a liquid, perhaps most often honey-like.314 In any case, Apollo does not find the cows for which Hermes had been hungry (64), and does find the resources for hospitality which he has pointedly not waited to be offered.315 249 χρυσός τε καὶ ἄργυρος: As at 60–1*, Maia’s metal goods are poised between the possessions stored in wealthy households (like the ingots of Od. 2.338 νητὸς χρυσὸς καὶ χαλκὸς ἔκειτο, or worked objects) and the dedications of a sanctuary strongroom, such as the offerings at Delphi outlined in 179–81*. Gold may be the metal of choice for gods’ possessions, but silver is quite normal too (e.g. ἀργυρότοξος Ἀπόλλων; see LfgrE s.v. ἀργύρεος 1bα). 250 πολλὰ δὲ φοινικόεντα καὶ ἄργυφα εἵματα νύμφης: The line clearly echoes and balances 249: |πολλ-, δέ, ¦καὶ ἄργυ-. Around the caesura are a pair of terms suggesting prototypical and contrasting colours. φοινικόεντα suggests the red end of the spectrum of prestigious dyes extracted from 314

See LfgrE s.vv. For nectar and honey see 562* and e.g. E. Ba. 143; so probably Alcman PMG 42 (nectar object of ἔδμεναι). For bread-like ambrosia see Od. 5.93, Il. 5.777. It is often an object of ἐδ- and βρω-, so non-liquid. However, ambrosia is like honey in Ibycus PMG 325, and in Sappho fr. 141 Hermes ἐοινοχόησε ambrosia which has been mixed in a κρατήρ, but does so using an oil-flask. Nectar and ambrosia together are liquid at Il. 19.347–53, winelike at Od. 9.359, Pi. O. 1.62, not liquid at H.Ap. 124–7. At Il. 19.38 nectar joins ambrosia as an embalming agent; wine and honey might have been used as such (see Hdt. 2.68.3 with Lloyd 1976), or the formulaic combination is loosely given the characteristics of ambrosia. Despite all the variations, Anaxandrides fr. 58 τὸ νέκταρ ἐσθίω πάνυ | μάττων [i.e. bread-like] διαπίνω τ’ ἀμβροσίαν may be comically muddled. For the idea that nectar and ambrosia replaced an earlier representation where gods consumed wine and meat see Pulleyn (2006) 66, 69. For ambrosia and Skt. amṛta, a drink of the gods, see e.g. West (1966) 342. 315 For such relocation of traditional elements in the hospitality type-scene, compare Il. 16.158–63: Achilles and the Myrmidons are refusing the normal pre-battle communal meal, which is relocated to a simile (Nimis 1987: 23–33).

286

C O M M E N TA RY 2 5 1 – 2 5 3 molluscs (Steigerwald 1986); see e.g. Xenophanes fr. 32, Arist. Mete. 3.2 372a4 for φοιν- in rainbows.316 ἄργυφα also describes both colour and prestige, the latter connotation being reinforced by the echo of ἄργυρος above. The application to clothing could activate an analysis ἄργ-ὑφ- ‘white-woven’.317 These clothes continue the poising of the ἄδυτοι between human storeroom and temple strongroom, since clothes were a form of dedication: again cf. 179– 81*, where Delphi had metals and clothes for the taking. 251 θεῶν μακάρων ἱεροὶ δόμοι: ἱερὸς δόμος is used of a temple at Il. 6.89, Hy. 24.2. For the inclusion of nymphs among θεοί see n. 209. 246–53 Apollo’s house-search: The audience probably knew that in some versions Hermes hid the cows in his cave (101–2*); if so, the poet exploits it to communicate Apollo’s frustration. There is desperation too – of a psychologically plausible nature – when Apollo searches the cave even after finding evidence of the cows at Pylos. The brevity and inconsequentiality of his search is expressed through ring-composition on a miniature scale. The difference between ignoring Hermes and questioning him occupies the outer ring, and the figurative use of 252 ἐξερέεινε (‘enquire’ > ‘explore’, as at e.g. Od. 12.259 πόρους ἁλὸς ἐξερεείνων) sets up a contrast between the failed house-search and the slightly more productive verbal enquiry which follows: 245 246 249 250–1 252 253

παῖδ’ πάντα μυχὸν μεγάλοιο δόμοιο πολλὸς δέ... καὶ ἄργυ-… ἔνδον ἔκειτο πολλὰ δέ... καὶ ἄργυ-… ἐντὸς ἔχουσιν μυχοὺς μεγάλοιο δόμοιο Ἑρμῆν

Hermes (ignored) Area of search (start) Wrong objects found Wrong objects found Area of search (end) Hermes (addressed)

Hermes’ partial success in confusing Apollo means that 253 κύδιμον ‘glorious’ is no longer merely proleptic. Moreover, the cave turns out to be rich and conspicuously like a temple (246–7; contrast Apollo’s initial impression in 234). The fact that its stores recall those of Delphi in Hermes’ threat of ἱεροσυλία (178–81) casts doubt on the propriety of Apollo having barged in (cf. 233–4, 370–2). Rather than finding his cows, he loses some of our sympathy.

I agree with Palmer (1962: 105) that the word was probably pronounced φοινῐκόεντα, a form of metrical shortening, rather than φοινῑκόεντα. 317 Cf. Le Feuvre (2004). Chantraine (1933: 262–4) argued instead for a weakly attested series of colour-terms in -φος. 316

287

C O M M E N TA RY 2 5 4

254–9 254 ὦ παῖ: ὦ grabs attention; see with caution Scott (1903) for expressive instances of its use in Epos. παῖ was not especially harsh per se, though less intimate than the personal name: Dickey (1996) 209. But the monosyllables in this context suggest a thunderous delivery. ὃς ἐν λίκνωι κατάκειαι: Relative clauses immediately after an address are rare. Rather than clarifying who is addressed, they normally single out a pertinent attribute of the addressee, especially a general and valued activity of a god.318 Here, however, Apollo picks what he believes to be an atypical and untrustworthy state for Hermes; the phrase may therefore have been performed sarcastically. κατάκειαι preserves a form intermediate between the loss of intervocalic sigma in inherited κεῖσαι and its reinsertion by analogy with verbs whose stem ends in a consonant, where -σαι had remained (Schwyzer i.668). See SEG 54.562 κεῖαι (Larisa c.457) and 37.351 -κεῖοι (third-century Arcadia; Hoffmann 1900: 203–4). It is unlikely that a scribe hit on -κειαι by accident. Schulze (1892: 443–4) even suggested that κεῖσαι should be emended out of Epos, but the reinsertion of sigma probably happened at different rates in different areas and for different verbs (Buck 1955: 55–6). 254–5 μήνυέ μοι βοῦς | θάσσον: The alliteration of μ-, the monosyllables creating a pronounced sixth-foot caesura, the curt syntax (contrast e.g. περὶ βοῶν, ὅπου βοῦς ἔκρυψας) and the explosive enjambement rebuild the ­rhetorical vigour of ὦ παῖ*. Contrast Apollo’s gambit for enquiring about cattle at 190–200*. Besides simply meaning ‘tell’, μηνύω is the technical term for informing as the first step in legal proceedings; in the fifth century it is particularly aligned with serious or abhorrent cases, and so Apollo may suggest an interpretation of the cattle-theft as ἱεροσυλία. In Attica, ὁ βουλόμενος laid information before the assembly or council, or sometimes magistrates could force someone to inform, as the supporters of Theramenes ‘forced’ Agoratus (Lysias 13.18); a reward was occasionally offered, in the form of money (see 264* μήνυτρον) or amnesty.319 Apollo arrogates to himself the power to get Hermes to own up, using a threat rather than an incentive. For the idea of informing on oneself compare Sophocles OR 1384, Ant. 84; for the hymn’s legal language, see Introduction §3.3.5.

318

See e.g. Il. 1.37, 16.514, 17.248, Od. 24.36, S. Tr. 200, E. HF 60. Procedure: MacDowell (1978) 181–3. In particularly abhorrent mythical crimes: S. OR 1384, E. Hipp. 1077, Hec. 192.

319

288

C O M M E N TA RY 2 5 5 – 2 5 6 Ψ’s θᾶττον (M has blank space) is an Atticism arising in transmission (cf. 212*). For the variable quantity of upsilon in μηνύω see LSJ s.v., GH i.372–3, and contrast Herm. 373 μηνύειν. 255 ἐπεὶ τάχα νῶϊ διοισόμεθ’ οὐ κατὰ κόσμον: For ἐπεί ‘since sc. otherwise’ see LSJ s.v. B 1; Schenck zu Schweinsberg may be right to insist that the omission of a word like ‘otherwise’ makes Apollo’s expression of the risks more forceful. The nominative pronoun νῶϊ personalizes the affront – the opposite of the depersonalizing politeness strategy discussed on 201–11*. οὐ κατὰ κόσμον implies the speaker’s moral superiority and, as often, is ‘a sinister understatement’ which sets up a rebuke or threat of punishment.320 Apollo will retract his threat with a similar understatement at 438*. Here κόσμος gains a possible cosmic resonance from 256*. The middle of διαφέρω seems to be a prosaic usage, often found in later inscriptions about dispute-resolution (Introduction §1.1). 256 ῥίψω γάρ σε λαβὼν ἐς Τάρταρον ἠερόεντα: λαβών is Ilgen’s correction of Ω’s βαλών. ῥίπτω is formulaically accompanied by a verb of seizing (x11 besides here: LfgrE iv.42.29–33), but never by a pleonastic verb of throwing. |ῥίψω is placed for threatening delivery, like 254 ὦ and 255 θάσσον. Epos contains several contexts of throwing (especially using ῥίπτω) from a height, with some shared formulas: immortals are thrown from Olympus to Earth, the sea or Tartarus; Astyanax is thrown from the walls of Troy. Closest to Herm.’s line is Zeus’s threat to send any disobedient god to Tartarus at Il. 8.12–13 (πληγεὶς οὐ κατὰ κόσμον ἐλεύσεται Οὔλυμπόνδε, | ἤ μιν ἑλὼν ῥίψω ἐς Τάρταρον ἠερόεντα), but it evokes more generally the tradition of such acts of throwing.321 Tartarus is a dank and emphatically inescapable Underworld cavern, the conceptual opposite of Olympus. People are generally despatched there at the hands of Zeus or a close accomplice, and in response to a threat to Zeus’s dominion.322 Apollo’s threat therefore arrogates the power

320

Kirk (1990) on Il. 8.12; so e.g. Od. 8.179, 20.181. The tone of moral superiority means that the formula can hardly occur in narrator-text: the exception which proves the rule is Il. 2.214, of Thersites. 321 See also Il. 1.591–3 (Hephaestus), 15.22–4, 19.130–1, 24.734–6 (Astyanax; cf. 6.466– 70), Il. Parv. fr. 21.4, H.Ap. 317–19 (similarly Il. 18.392–9); Tartarus features also at Hes. Th. 868 (Titans), fr. 30.22, fr. 54. S. Tr. 779–80 reworks the language of these scenes. 322 Besides throwing, Zeus sends people there with his thunderbolt, e.g. Hes. Th. 853–68, fr. 30.22, Pi. fr. 52d. Zeus’s accomplices: e.g. Athena at E. HF 908, an Erinys at Or. 265, the Keres at Sc. 255. For Tartarus being inescapable see 257*. Anti-Olympus: geometrically opposed at Il. 8.16, Hes. Th. 720; also in some architectural features

289

C O M M E N TA RY 2 5 7 of Zeus to effect such punishments: Brown (1947) 84, Vox (1981) 111, Harrell (1991). It is also disproportionate in that Hermes’ cattle-theft, though part of a plan to usurp some of Apollo’s prerogatives, is trivial compared with (say) the Titans’ attack on the gods.323 Both the pretensions to Zeus-like power and the over-inflation of the crime were already hinted at in 254–5* μήνυε. Given Apollo’s threat, one might reinterpret 255 as coloured by the idea ‘we will be borne our separate ways, but not [sc. for you] through the upper world’. The various meanings of κόσμος often interact, and the sense ‘(layer of) the upper world’ was established by the fourth century, with roots as far back as Anaximander (Kerschensteiner 1962: 16–17, 21–2, 45–54). Herm.’s poet may have seen a similar ambiguity in Il. 8.12–13 (quoted above) and Parmenides 28 B4.3 ‘when being scatters everywhere κατὰ κόσμον’. 257 ἐς ζόφον αἰνόμορον καὶ ἀμήχανον: ζόφος, which is often ἠερόεις as Tartarus was in 256, is both ‘where the sun sets’ and ‘Underworld’; the otherworld may once have been located in the far West (Stokes 1962: 16–21). For αἰνός, see 226*, Sc. 227 ζόφον αἰνόν. Tartarus is ἀμήχανος because its inmates were regularly shackled (cf. 258 ἀναλύσεται), as in an upper-world δεσμωτήριον: e.g. Hy. 1.C5, Hes. Th. 618–20, 726–33 with West (1966), Pindar fr. 207, PV 154–7; Anacreon also emphasizes its inescapability in PMG 395.11– 12. The audience can recall Maia’s prediction in 157 that Apollo would apply ἀμήχανα δεσμά to Hermes’ flanks (Introduction §3.2), and contrast both adult viewpoints to Hermes’ recent μηχανή where he deployed swaddling-bands (called δεσμά at H.Ap. 129, ἐπίδεσμοι at Sor. Gyn. 2.2) to suggest his innocence. For binding in Herm., see Introduction §3.3.6. 257–8 οὐδέ σε μήτηρ | ἐς φάος οὐδὲ πατὴρ ἀναλύσεται: Apollo’s phrasing is reminiscent of the bluster with which he gloats over Python at H.Ap. 367–8.324 His bluff here continues with the unsupported idea that Zeus would not release Hermes. We know that his assumption that Maia would support Hermes is too simple. In fact, she has renounced Hermes in similar terms to Apollo (see 257* ἀμήχανον, 259 ἐρρήσεις*). Vox (1981: 112–13) noted the

(Ballabriga 1986: 259). Many other features vary in different sources, e.g. Tartarus is separate from Hades (Il. 8.16), contains it (Hes. Th. 767–74) or is the worst part of it (Sc. 255); cf. n. 326. 323 Later sources do offer some parallels for less treasonable crimes which land one in Tartarus, e.g. E. Hipp. 1290, Pl. Grg. 524b. 324 See Richardson 132–3, with further parallels from epic; elsewhere there are instances with more pathos, such as IG IX.12 ii.313 (Tyrrheion, second century). Compare also Q.S. 12.209–13.

290

C O M M E N TA RY 2 5 9 further irony that Apollo himself, after killing the Cyclopes, was saved from tartarization only by Leto interceding with Zeus (Hes. fr. 54). ἐς φάος opposes 257 |ἐς ζόφον. This conceptual opposition was traditional: compare e.g. Hes. Th. 652–3, H Dem. 337–8. 259 ἐρρήσεις: A further echo of Maia’s rebuke (160 ἔρρε). The derogatory tone is particularly clear in the future tense, at least in the colloquial Attic question at Cratinus fr. 129 οὐκ ἀπερρήσεις, Ar. Pax 500 οὐκ ἐς κόρακας ἐρρήσετε, Lys. 1240 οὐκ ἐρρήσετ’. The form ἐρρήσεις is glossed, possibly from Herm., in Phot. ε.1958 with μετὰ βίας εἰσελεύσηι. ὀλοοῖσι μετ’ ἀνδράσιν ἡγεμονεύων: M’s μετ’ is preferable to Ψ’s -ν ἐν. It expresses ‘within a community of …’ more precisely (GH ii.117; with ‘lead’ at e.g. Il. 1.252, 7.227, Od. 10.204), and is the lectio difficilior since μετά + dative is largely restricted to Epos. Apollo’s threat rests on the common representation that characters in the Underworld, apart from sinners with eternal punishments (e.g. Tantalus), continued their worldly activities. He already sees in Hermes both facets of the early usage of ἡγεμονεύειν (‘lead the way for, command’).325 He has led away the cattle, and will become a guide with epikleseis such as ἡγεμόνιος (Ar. Pl. 1159; IG II2 1496.85 from 331/0 bce), ἁγεμών (ILindos 184; second century) and ἀγήτωρ (Paus. 8.31.7, at Megalopolis); his guiding will in fact take him to the Underworld, though only temporarily; Apollo and Zeus will see Hermes as a guide at 301–3*, 392, 461. Meanwhile Hermes has threatened to become a commander of thieves (175), and Apollo will explicitly predict this title at 291–2*. The similar activity lends significance to the changed circumstances: the company in Tartarus is reduced to merely μετ’ ἀνδράσι (contrast the thematic phrase ἐν/μετ’ ἀθανάτοις, which charts Hermes’ relationship to Olympian society: 170*), and an inglorious company of men at that. Bothe’s ὀλοοῖσι replaces Ω’s ὀλίγοισι(ν). The latter presumably means ‘few.326 It cannot mean ‘insignificant’ without clarification such as a contrast with μέγας (Moorhouse 1947: 37), and those in Tartarus are not ‘physically small’. While hardly any humans made it to Tartarus, numbers seem a banal 325

Vergados (on line 258) observed that Apollo’s idea here shifts from that of ἀναλύσεται, where Hermes is shackled. The idea of ‘commanding the dead’ may very well have recalled Achilles’ grim view of the Underworld at Od. 11.488–91, ending καταφθιμένοισιν ἀνάσσειν (Diggle, p.c.). 326 Hes. Th. 447, Simon.(?) fr. 256.6, A. Pe. 330, frequently in classical prose. Gemoll’s idea that ‘few’ is a litotes for ‘none’ presupposes the conception of Tartarus as specifically for immortals (Hes. Th., Pherecyd. Syr. 7 B5; contrast Hes. fr. 30.22, E. Hipp. 1290, S. fr. 442, where humans are there), and implies that Hermes would have no interest in leading Tartarus’ immortal inhabitants.

291

C O M M E N TA RY 2 6 0 – 2 6 1 factor for Apollo to pick out in his climactic point, when Tartarus’ inmates offer so many sordid opportunities. Perhaps ἀλιτοῦσι ‘who have sinned’ deserves consideration. Hermann’s δολίοισιν risks making Tartarus sound too homely for Hermes: Van Herwerden (1907) 185. 254–9 Apollo’s threats: Apollo continues his show of force, with an a­ rsenal of rhetorical strategies (rhythm, alliteration, syntactic curtness, understatement and perhaps sarcasm) thrown into relief by the contrast with his earlier approach to the Onchestan. As his bursting into the house was of dubious propriety (233–4*, 246–53*), so our sympathy for this speech is undermined: Apollo overinflates the crime, claims powers he does not have (254–5* on μήνυε, 256*), makes ungrounded assertions, mistakes Maia’s disposition, and is hypocritical (257–8*). His bluster can be taken as a sign of an intemperate bully (compare 370–7) with juvenile pretensions to his father’s power, or as a rhetorical power-play which he naively hopes will work on a figure whom the audience knew would grow into the god of rhetoric. In fact, the threat does touch a nerve, though probably not as Apollo expected: a permanent trip to dishonour in Tartarus would run diametrically against Hermes’ quest for a permanent (170 ἤματα πάντα) life of τιμαί on Olympus, away from Maia’s cave, which he focalized as underworldly at 172* ἠερόεν; and permanent ‘leadership’ there would replace the fleeting descents to the underworld for which the audience worshipped Hermes as their psychopompic guide (see 572–3).

260–77 260 τὸν δ’ Ἑρμῆς μύθοισιν ἀμείβετο κερδαλέοισιν: After the echoes of Maia’s speech (155–61) in Apollo’s, Hermes’ response is introduced as in 162*. There, κερδαλέοισιν set up his focus on his ambition for wealth; here, the focus is more on the cunning involved in acquiring κέρδος. Cf. 463*. 261 Λητοΐδη, τίνα τοῦτον ἀπηνέα μῦθον ἔειπες; : The tone and shape of the line provide a further similarity to Hermes’ exchange with Maia (cf. 163), and the unusual metronymic address Λητοΐδη recalls her suggestion that Leto is the happier mother.327 However, Hermes’ phrasing is closer to formulaic

327

158 |Λητοΐδου with 155–61*. Other metronymic addresses: 567, Alcm. PMG 48, Stes. fr. 2a, Anacr. PMG 388.11(?), Simon. fr. 263, A. Ag. 914. They appear more frequently in Euripides and Lucian: E. Hel. 616, 1098, IA 686, 819, 836, 1106, fr. 177, PMG 934.14; Luc. Cont. 1 = Prom. 5 (Μαίας παῖ); Fug. 29, Prom. 5, Dial. Mort. 28.1. Metronymic joined with patronymic: E. IT 238, Hel. 1680, IA 896, Pl. Alc.I 105d, Ps.-Pl. Epist.2 313a. For reference using the mother’s name, see 89*.

292

C O M M E N TA RY 2 6 2 – 2 6 3 ways of beginning an indignant response, namely vocative + either ποῖον τὸν μῦθον ἔειπες;| (Il. x7) or οἷον/ποῖον ἔειπες;| (Il. x5, Od. x4).328 Such criticisms do tend to be reasonable and respected, so that Hermes is pretending to lay claim to level-headed indignation.329 ἀπηνέα developed towards ‘harsh’ (e.g. Pl. Lg. 12 950b) from a more specific core of refusing to trust or cooperate; Hermes may be objecting to Apollo’s refusal to believe his innocence despite the failure of the house-search.330 The Homeric tradition prefers ἔειπες to ἔειπας (West 1998: xxx). -ας, though occasionally found in classical Ionic (Bechtel iii.202–3), was normal in Attic and so likely to creep into the paradosis. 262 καὶ βοῦς ἀγραύλους διζήμενος ἐνθάδ’ ἱκάνεις; : Stephanus first identified that καί here is used to start a surprised question (cf. Denniston 311–12), separate from line 261. This seems to be a conversational idiom not otherwise admitted in Epos. The tautology ἐνθάδ’ ἱκάνεις adds emphasis (191*), which after ἀγραύλους creates the argument that cows sleep in the fields, not in mountain caves. This argument is tendentious for two reasons. It is the first time Apollo’s cows have been called ἄγραυλος, and in fact the unspoilt Pierian meadow was less of an ἀγρός (i.e. managed pasture or agricultural land) than the meadow near the Alpheus to which Hermes led them (see 71–2*, 103–4*, 399 ἀγρούς).331 Secondly, in other versions of the myth the Greeks were happy to imagine Hermes keeping the cows temporarily on Kyllene (101–2*). Hermes deploys for advantage both the normality of Apollo’s cows in terms of habitat (which he has just effected) and the negative representation of the cave’s outof-the-way location. After the recent evocations of Delphi (233*, 246–7*), Fernández Delgado (1990: 215–16) may have been right to suggest that διζήμενος would be coloured by its frequent use in recorded Delphic oracles; see Parke and Wormell (1956: s.v.). Hermes would then be casting Apollo like a visitor to his own oracle, and will give him a blunt response of ignorance. 263–4 οὐκ ἴδον, οὐ πυθόμην, οὐκ ἄλλου μῦθον ἄκουσα. | οὐκ ἂν μηνύσαιμ’· οὐκ ἂν μήνυτρον ἀροίμην: πυθόμην in this context means ‘overhear’, not The more elaborate ποῖόν σε ἔπος φύγεν ἕρκος ὀδόντων| (Il. x2, Od. x6) would not fit Hermes’ sham simplicity. For such phrases see Kelly (2007) 185–6, 225–6. 329 The exception is that Zeus tends to ignore the question ποῖον τὸν μῦθον ἔειπες;. 330 For ἀπηνής of refusing to believe something, compare Il. 23.484, Od. 23.97, 230. Blanc (1985), (1995) discusses the word’s origins, comparing ἀν-αίν-ομαι and the development of προσηνής. 331 For the cows’ epithets, and the fact that ἄγραυλος becomes important only from now on, see Fernández Delgado (1998) 6. 328

293

C O M M E N TA RY 2 6 3 – 2 6 4 ‘enquire’ or ‘ascertain by asking’ – since that would raise the question why Hermes was making enquiries. οὐκ ἴδον, οὐ πυθόμην thus matches what Hermes enjoined the Onchestan to say at 92–3: Tzifopoulos (2000) 155. Hermes’ three assertions list without fluster different types of information – first-hand visual, first-hand aural, and second-hand.332 Chance aural evidence has a natural place in the story, and might include conspiratorial muttering (as at Od. 19.486), the cows lowing as they were dragged off (74*), dogs barking (145*), or a lyre made from the cows’ body-parts (as in S. Ichn.). In 264, superficially Hermes uses the same stem as Apollo (cf. 254 μήνυε) to indicate a desire to cooperate, but his introduction of a possible reward for information is subversively infuriating since it was obviously excluded from Apollo’s demand to admit or be tartarized. Other versions of the story do allow for a reward, as in Battos’ bribe or Apollo’s offer of a μήνυ[τρον] at S. Ichn. 87. A further irony in οὐκ ἂν μηνύσαιμι is that Hermes will become the announcer par excellence, as herald and messenger (for μηνύω of messages, see e.g. E. Ph. 1217–18). The phrasing of these lines extends the traditional asyndeton οὐκ ἴδον, οὐ πυθόμην to five clauses, and is unparalleled in Epos and quite different from Hermes’ style elsewhere in the Hymn. One could take this as a rhetorical technique for emphatically rejecting an accusation: compare the four clauses in asyndeton, each beginning with a negator, at Demosthenes 18.107, or the longer but more varied stack of negatives at Antiphon 6.17.333 However, comparative evidence suggests a further opportunity in these lines, namely that they are a partial representation of childish syntax, which Hermes is feigning. This characterization would cover the very short clauses, asyndeton, repetitious nature of 264, and persistent use of a simple sentence-functor first in each clause.334 A performer could have adopted a wheedling or petulant tone of voice, and the lines are remarkable enough for Apollo to choose to mimic them at a climactic moment at 363–4*. This would show, by ancient Greek standards, a receptivity to unusual communication phenomena which is thematic in the hymn (Introduction §3.4). As in other Greek texts, the representation is far from thoroughgoing (cf. the advanced word μήνυτρον and optatives). Hermes’ tailoring of register would neatly complement his

For the opposition of πεύθομαι and ἴδον see e.g. Od. 1.242, 10.147, 17.510–11. For ἴδον vs ἄλλων μῦθον ἀκούων see Hes. frr. 199.3, 204.62–3. The point of these traditional contrasts was variable: Od. 23.40–1 οὐκ ἴδον, οὐ πυθόμην, ἀλλὰ στόνον οἶον ἄκουσα | κτεινομένων makes ἄκουσα first-hand and πυθόμην second-hand. 333 Negated short phrases in asyndeton are much more common: see e.g. Od. 24.246–7 with KG ii.290. 334 For this and Theon’s similarly petulant negatives in POxy. 119, see Thomas (2010). 332

294

C O M M E N TA RY 2 6 5 – 2 6 7 paralinguistic pretence of being a typical baby in 235–42; both tricks rely on abilities which are anything but babyish. 265 οὔτε βοῶν ἐλατῆρι, κραταιῶι φωτί, ἔοικα: Hermes moves from denying knowledge to denying guilt. ἔοικα implies εἰκός, a central concept of classical but not Homeric rhetoric: see 260–77* below. Hermes aligns the accusation (‘cattle-rustler’) with ‘strong man’, and his dissimilarity to the latter implies his dissimilarity to the former. Understandably he passes over his own nature as more than human, though we and Apollo may well notice that it destroys the argument. For Hermes’ surprising strength, see 117, 407, 414 κρατύς. His presupposition here that appearances can ground inferences about reality contrasts with his track-trick, as Apollo knows all too well (219– 26), and his attempted trick of swaddling himself tightly at 235–42. The external audience has the extra pleasure of feeling Apollo’s frustration as his own supposition that cattle-rustlers are adult men (200 ἀνέρα*) is deployed against him, and of recalling that the poet gave Hermes’ identity as a cattle-rustler a prominent place in his initial list of epithets (14 ἐλατῆρα βοῶν, echoed here).335 The sequence ‘οὐ clause, οὔτε clause’ has parallels at e.g. Il. 22.265, S. OC 450–1, E. IA 1319–23; see KG ii.288–9, Denniston 509–10. Hermann’s οὔ τι matches 377, but the two lines need not be standardized. 266 οὐκ ἐμὸν ἔργον τοῦτο, πάρος δέ μοι ἄλλα μέμηλεν: ἔργον superficially is meant as ‘characteristic activity’ (LfgrE s.v. 1b, often the subject of μέλει), but in the background are the senses ‘criminal deed’ (LfgrE 1d), as Apollo sees it, and ‘great deed’ (cf. 16*, where the cattle-theft was one of Hermes’ κλυτὰ ἔργα). τοῦτο here is a good example of the pronoun’s common addressee-­ orientation (‘this that you’ve mentioned’: Hermes dissociates himself from it). Contrast Od. 16.207 τόδ’ (‘this which affects me’) ἔργον Ἀθηναίης. After πάρος, the perfect μέμηλεν indicates that a past state is ongoing: KG i.134, Schwyzer ii.273–4. Although very occasionally πάρος ‘sooner’ verges on the sense ‘rather’ (e.g. E. Or. 345), Ilgen’s idea that here it expresses preference and not priority lacks support. 267 ὕπνος ἐμοί γε μέμηλε: Sleep is placed first, as particularly relevant to Hermes’ nocturnal affairs, and to cohere with his acting at 235–42. For Apollo’s response, see 289*. The businesses, chores and active pursuits which are the subject of μέλει for adults are regularly opposed to sleeping (e.g. Il. The phrase βοῶν ἐλατῆρι here and in 377 may have been a source for Colluthus, who reapplies it to a cattle-prod (Rapt. 43), or for Nonnus (Introduction §4.5).

335

295

C O M M E N TA RY 2 6 8 2.24–5, 10.91–2, 24.683–4, Od. 4.414–15, 11.330–2): Hermes suggests that his pursuits diametrically oppose adult norms, including those of a cattle-rustler. The repetition of μοι μέμηλε may add to the point by building on the childish repetitiousness of 263–4*.336 But as usual, Hermes’ pretences at childishness are undercut, since the words point the audience’s minds again towards the adult Hermes’ role of controlling their sleep-patterns (14*, 241). καὶ ἡμετέρης γάλα μητρός: Whether true or not, this claim reaffirms Hermes’ pretence that he is an ordinary rather than a superhuman baby. It counters Apollo’s allegation first by denying any desire for cow’s milk or beef (though we know he was κρειῶν ἐρατίζων: 64); it also opposes Apollo’s own tastes from H.Ap. 123–5 οὐδ’ ἄρ’ Ἀπόλλωνα χρυσάορα θήσατο μήτηρ | ἀλλὰ Θέμις νέκταρ τε καὶ ἀμβροσίην ἐρατεινήν | … ἐπήρξατο (Introduction §1.3). Hermes’ claim also entails maternal supervision overnight, and covers up his spirit of independence (playing on Apollo’s assumption in 257–8 that Maia would support Hermes). Breastfeeding was a central part of the concept of motherhood in archaic and classical Greece, regardless of how much wet-nursing went on in practice.337 In Herm., the use of plural pronouns and verb-forms for singular ones is a characteristic of Hermes’ speech (see also 276, 310, 465, 491): the tone can be read alongside the evidence of tragedy, where the speakers are disproportionately often women, children talking to parents, or people of low social status. For 2nd-person plural-for-singular see 276*.338 268 σπάργανά τ’ ἀμφ’ ὤμοισιν ἔχειν καὶ θερμὰ λοετρά: Swaddling and bathing complement the activities in 267 in a rhetorically effective way: swaddling, like sleep, entails inactivity; bathing, like breastfeeding, entails However, μέλω is doubled six other times in Epos, including 451–3 (LfgrE iii.117.43– 4). The examples do not seem to be formulaic. 337 Symptomatic of the centrality of breastfeeding to motherhood is the intense supplicatory gesture of baring a breast: Lee (1997) 194. Wickert-Micknat (1982: 69–70) discussed the lack of wet-nurses in Epos. Their prevalence in the fifth century is hard to quantify; later, they seem to have been common in well-to-do households; bottle-feeding seems to have been limited to the weaning process. Étienne (1973: 35–9) and Rühfel (1988) are particularly instructive. 338 Herm. has 1st-person plural-for-singular markedly more frequently than the rest of Epos, but less than Euripides, Theognis or Sophocles. In Epos most examples involve ἡμέτερος. The origin may be cases like ‘our house’ (x9 in Epos, including 370), where the relevance of the speaker’s kin varies with context; or cases where a speaker politely includes others in his own concerns, e.g. Il. 22.393 ἐπέφνομεν Ἕκτορα δῖον. ἡμέτερος ‘my’ also qualifies family-members at Od. 6.311, 22.464. The ‘royal we’ is hardly traceable before the Hellenistic period. See Schwyzer ii.203, 243, and with caution Zilliacus (1953) 12–34. 336

296

C O M M E N TA RY 2 6 9 – 2 7 1 supervision. As a set, the four constitute a soothing bedtime routine in reverse, extending (by breastfeeding) and explicating the terms of Hermes’ pretence in 235–42*.339 The emphasis on basic physical needs recurs in the typical comforts offered to guests in the Odyssey, i.e. bath, anointment, clothing, eating and bed (Arend 1933: 124–6, Laser 1983: 140–1; Laertes’ lack of κομιδή at 24.249–55 gives the converse). The overcomfortable Phaeacian lifestyle gives itself over to such pursuits and also to music (Od. 8.248–9 αἰεὶ δ’ ἡμῖν δαίς τε φίλη, κίθαρίς τε χοροί τε | εἵματά τ’ ἐξημοιβὰ λοετρά τε θερμὰ καὶ εὐναί); Hermes cannot yet mention the κίθαρις, but Apollo will find out that it interests him (μέλει again) at 437*. For θερμά (‘warm’ here, rather than ‘hot’) cf. Sor. Gyn. 2.31 ὕδωρ θερμὸν καὶ εὔκρατον ὡς πρὸς τὴν τοῦ βρέφους εὐαρέστησιν· τὸ γὰρ ὡς πρὸς ἡμᾶς εὔκρατον ζεστότερον. 269 μή τις τοῦτο πύθοιτο, πόθεν τόδε νεῖκος ἐτύχθη: νεῖκος means ‘quarrel’ but also, in the context of 264 μήνυτρον, a more legalistic wrangle (Il. 18.497, Hes. Th. 87, Op. 29–35, etc.): the language is building towards the following episode when the brothers take their dispute to Zeus. Hermes’ words turn the adults’ attacks on him back against Apollo: his feigned wonder at the accusation (¦πόθεν τόδε νεῖκος) echoes Maia’s exasperated wonder at him (155 ¦πόθεν τόδε νυκτός); and his idea that news of the accusation could be embarrassing to the accuser (expressed with πύθοιτο) overturns the presupposition of 263 (expressed through πυθόμην) that the criminal should want to escape gossip. 270 καί κεν δὴ μέγα θαῦμα μετ’ ἀθανάτοισι γένοιτο: Hermes glosses 269 τις by μετ’ ἀθανάτοισι (for the importance of which see 170*), showing that he is already thinking of his reception by Olympians. This constitutes a change of tack, from simulating normality to presupposing that he belongs in superhuman society. μέγα θαῦμα echoes Apollo’s recent reactions (196, 219), but with this broader Olympian audience in mind. While Hermes superficially says ‘your accusation will produce scepticism’, he simultaneously and complacently hints ‘the crime will produce astonishment’. He will be proved correct (330–2*). Cf. 277*. 271 παῖδα νέον γεγαῶτα: Hermes’ argument is again about what is plausible for an infant. But ironically the hymnist gives him a phrase which echoes 214 |Φιλήτην γεγαῶτα, from when Apollo identified Hermes as a suspect.

339

Compare Theoc. 24.3–5 with the sequence bath, breastfeeding, bed; Sor. Gyn. 2.31–7 suggests bath, anointing and massage, swaddling, breastfeeding, bed.

297

C O M M E N TA RY 2 7 1 – 2 7 3 We perceive Apollo’s awareness that παῖδα νέον and φιλήτην are all too compatible, conceptually as well as metrically. Compare 331*. 271–2 διὰ προθύροιο περῆσαι | βουσὶ μετ’ ἀγραύλοισι: ἀγραύλοισι here ousts εἰλίπους (contrast Il. 6.424 |βουσὶν ἐπ’ εἰλιπόδεσσι; similarly contrast 412 ἐπ’ ἀγραύλοισι βόεσσιν with Il. 16.488 ἐν εἰλιπόδεσσι βόεσσιν, and see 567*), bringing us back towards the point of 262*: babies and cows reside in very different places. That line emphasized the unlikelihood of finding cows in a cave. I take this one to complement it with the improbability of a baby going out into the ἀγρός. The asyndeton between 272 and 273, which points out the improbability of Hermes being outdoors, suggests that 271–2 should express disdain for the idea that a newborn would travel at all. Accordingly I take μετά + dative as ‘in pursuit of ’ (compare the preverbial use at Od. 6.132–3 βουσὶ μετέρχεται … ἠὲ μετ’ … ἐλάφους) rather than the normal ‘in the company of ’, so that Hermes’ departure from the cave is more at issue than his return.340 As in 158 διὲκ προθύροιο περήσειν, echoed here, the focus is on Hermes’ crossing of the threshold: see Introduction §3.3.6.341 272 τὸ δ’ ἀπρεπέως ἀγορεύεις: A modification of formulaic ἀτρεκέως/ ἀσφαλέως ἀγορεύ- in this metrical position. The broad range of πρεπ- means that Hermes simultaneously casts Apollo’s comment as unworthy of him (see 269), not apt for the situation, and unworthy of Hermes (compare Th. 2.36.4, Isoc. Panath. 227 παραβολὴν ἀπρεπῆ πρὸς τὴν ἐκείνων δόξαν). The stem πρεπdeveloped from its core sense ‘outstanding (esp. visibly)’ (as in 351) through ‘fostering repute’ towards more general propriety (barely in Epos, but already in e.g. Alcman PMG 98.2 πρέπει παιᾶνα κατάρχην). Public speaking offered a place to shine by the time of the Iliad (9.441 ἀγορέων, ἵνα τ’ ἄνδρες ἀριπρεπέες τελέθουσιν), and τὸ πρέπον later became a critical term applied particularly to the construction of speeches which accord with the speaker’s character and circumstances: Sacks (1986). 273 χθὲς γενόμην, ἁπαλοὶ δὲ πόδες, τρηχεῖα δ’ ὑπὸ χθών: Hermes presents his argument in short, clear, irrefutable steps. The two verbless phrases, with their lapidary antithesis of ἁπαλοί and τρηχεῖα and their predicate–­ subject word-order, are typical of speakers appealing to shared knowledge rather than trying to convey information (Guiraud 1962, Lanérès 1994). From all the reasons why newborns do not go out walking, Hermes observes that Schneidewin’s βουσὶν ἐπ’ (1848: 678) sidesteps the unusual use of μετά which I propose. Stadtmüller restored the formula with διέκ here (also Il. x1, Od. x3; cf. 155–61*).

340 341

298

C O M M E N TA RY 2 7 4 the ground out in the ἀγρός is rough – a cheeky point, since it was the softness of the ground that posed a risk against which he made his ad hoc shoes (78–86). ἁπαλός describes areas of skin not hardened by toil, in an affective way which suggests vulnerability as a valued physical quality in children.342 For children’s soft feet in particular see e.g. H Dem. 287, Plato Smp. 195c–e. At Il. 19.92–4 Atē has ἁπαλοὶ πόδες· οὐ γὰρ ἐπ’ οὔδει | πίλναται, ἀλλ’ ἄρα ἥ γε κατ’ ἀνδρῶν κράατα βαίνει | βλάπτουσ’ ἀνθρώπους: she walks on men’s (normally) hairy heads, not on the ground, and so has soft feet, but her soft tread (like that of Hermes) makes her difficult to perceive. 274 εἰ δ’ ἐθέλεις, πατρὸς κεφαλὴν μέγαν ὅρκον ὀμέομαι: Ilgen’s δ’ ἐθέλεις is the normal epic form, though some Ionic inscriptions and H.Ap. 48 use θέλω. For the corruption to the koine form cf. H Dem. 160. For ὀμέομαι with both inner and outer object compare e.g. Hes. Th. 793 τὴν [Styx] ἐπίορκον … ἐπομόσσηι, Thgn. 1195 θεοὺς ἐπίορκον ἐπόμνυθι, Ar. Nu. 1232 ταῦτ’ … ἐπομόσαι μοι τοὺς θεούς. For the forms of the future in Epos see GH i.62, 451. Hermes offers an oath, and projects upon Apollo a desire to challenge him to one. Naturally offers of oaths, both promissory and assertory, were often effective in settling non-judicial disputes: e.g. Il. 9.132, 15.40, 21.373, Od. 10.345, Ps.-D. 50.31, Lys. 32.13. For oath-offers in legal proceedings, see 384*. Hermes can hardly believe that Apollo will be satisfied by this, and speakers who drop in the parenthesis ‘εἰ + ἐθέλει+ without infinitive’ are often manipulating their outlook (181*). Radermacher aptly questioned how s­eriously Hermes’ oath could be taken in view of his age; see Fletcher (2008: 36) on the lack of recorded oaths by individual children. The powerful being by whom one swore was to witness that the oathtaker was telling the truth (if referring to present or past) or to check that s/he fulfilled their promise (if referring to future), and to punish deviations. Hermes mentions Zeus’s head specifically as the body-part with which he can witness, nod (395*) and utter assent. The oath by Zeus’s head appears to be particularly awesome and limited to gods: Hera includes it at Il. 15.35–46; at H.Aphr. 26–8 (including κεφαλῆς πατρός and μέγαν ὅρκον) and Sappho fr. 44a.5, Hestia and Artemis, respectively, take hold of their father’s head while

342

ἁπαλός is used in Il. of warriors’ necks, to emphasize vulnerability during a deathwound, except for 3.371, where it emphasizes Paris’ exceptional effeminacy, and of the hands and feet of fortunate women and children elsewhere. For children’s soft skin and its positive valuation see e.g. Thgn. 1341, Pi. O. 6.55–6, E. Med. 1075, 1403, Pl. Tim. 81b–c, Golden (1990) 5, 8–9.

299

C O M M E N TA RY 2 7 5 swearing an oath to remain a virgin; Heracles overbearingly asks Hyllus to swear by Zeus’s head at S. Tr. 1185, but he swears simply by Zeus.343 275 μὴ μὲν ἐγὼ μήτ’ αὐτὸς ὑπίσχομαι αἴτιος εἶναι: The future verb in 274 left it unclear whether Hermes is offering to swear immediately (so at e.g. Il. 1.233, Od. 20.229) or later; Il. 21.373 presents the same ambiguity. A delay would allow Hermes his desired trip to Olympus, where he can actually touch Zeus’s head as he swears, and Apollo may well prefer the larger number of witnesses and the solemnity of the place. Hermes will in fact offer an oath on Olympus (383–5), but he also launches immediately into μὴ μὲν ἐγώ …, a headline (‘Certainly not I …’) which marks oaths elsewhere. He follows this μή with the indicative (276 ὄπωπα is unambiguous; see below for the construction of ὑπίσχομαι), a construction which is characteristic of oaths and cannot state the content of a future one; see KG ii.183–4, citing e.g. Il. 19.260–1 ἴστω νῦν Ζεύς... μὴ μὲν ἐγὼ κούρηι Βρισηΐδι χεῖρ’ ἐπένεικα.344 Hermes’ first oath is therefore μὴ αὐτὸς ὑπίσχομαι αἴτιος εἶναι. For ὑπίσχομαι/ ὑπισχνέομαι meaning ‘profess’ rather than the future-oriented ‘promise’, see e.g. Hdt. 2.28.1, 7.104.3. I take the whole to mean ‘I certainly do not profess to be guilty myself ’, and to be deceptive because so easily confused with ‘I certainly profess not to be guilty myself ’ – especially since it has been sprung on Apollo abruptly, and since the latter makes a more natural counterpart to what follows and corresponds to a recorded oath-challenge from Gortyn, μήτ’ αὐτὸν αἴτιον ἔμην, μήτε σὺν ἄλλι, μήτ’ ἐπ’ ἄλλι ϝισάμην.345 Hermes is demonstrating the admirable skill in formulation attributed to his son Autolycus, ὃς ἀνθρώπους ἐκέκαστο | κλεπτοσύνηι θ’ ὅρκωι τε (Od. 19.395–6); 343

For a database of Greek oaths, see https://www.nottingham.ac.uk/~brzoaths/database/ [last accessed 18 April 2019]. 344 Il. 19.261; cf. Pl. Ap. 22a ἦ μὴν ἐγώ, and also Hera’s μὴ δι’ ἐμήν ... (Il. 15.41). For uncorrelated μέν in oaths and asseverations, see Denniston 362, 389–90. For the assertory use of nominative ἐγώ see e.g. Il. 1.29 τὴν ἐγὼ οὐ λύσω, Gonda (1956) 78. Callaway (1993: 22) argued that Hermes only o­ ffers a future oath here (so Papakonstantinou 2007, Fletcher 2008), without discussing how the syntax would work. Callaway showed that Hermes’ two oaths diverge from a type-scene of oath-taking. This suggests that they are not straightforward instances of oaths (compare the divergences from formulas of sacrifice in 104–41), but does not prevent them from being oaths. 345 IC IV 47.16–21 (5th c.): A has given B a slave as a pledge, and the slave disappears; the magistrate must challenge B to swear ‘neither to be guilty himself, nor as an accomplice, nor to know that the slave is with a third party’. The more common interpretation of our line (‘I profess not to be…’) presupposes a ‘misplaced’ negative, as in οὔ φημι ‘I say that… not’. However, this is confined to very few verbs (KG ii.180–1) and there are no parallels for ὑπισχνέομαι. Crudden (2001) and Vergados share my interpretation.

300

C O M M E N TA RY 2 7 6 – 2 7 7 compare Themison’s amusing avoidance of perjury at Hdt. 4.154, or Leto avoiding an oath that Delos will host an oracle at H.Ap. 87–8. Perjury, rather than deceptive swearing, carried no such prestige – just a risk of divine punishment, magnified by explicit imprecations (see e.g. Il. 23.595, D. 54.40–1).346 It is intriguing that the fifth of the month (cf. 19) is particularly dangerous for perjurors according to Hes. Op. 803–4 (see West 1978); the likelihood of Herm.’s audience making this connection is less clear. 276 μήτέ τιν’ ἄλλον ὄπωπα βοῶν κλοπὸν: Hermes’ second asseveration, that he saw no other thief, is true but tricksy. ἄλλον ὄπωπα recalls 263 ἴδον ... ἄλλον (book-ending the speech), but also 199 ὄπωπας: Hermes is not a witness, and the poet makes him ‘answer’ negatively Apollo’s question to the real witness at 199–200 (see 277* for another echo of that episode). ὑμετεράων: Here and line 310 are probably Epos’ only instances of ‘singular’ ὑμέτερος, which perhaps occurs at Solon fr. 19.1–2 σὺ … καὶ γένος ὑμέτερον, then from Isocrates on. This interpretation is preferable to taking ὑμέτερος as a true plural, since Hermes’ speech is characterized by the use of plural for singular pronouns (267* ἡμετέρης), and since he should not project any preternatural insight that the cows belong to all the gods. 277 αἵ τινες αἱ βόες εἰσί: ‘whatever those cows are’. Hermes has shown knowledge of cows’ habitat and of cattle-rustlers, and is less likely to intend the blatant inconsistency of ‘whatever cows are’; the article marking a general category is anyway rare in Epos. The phrasing recalls 208–9 παῖδα … ὅς τις ὁ παῖς, where the Onchestan proved a slightly more helpful witness. For the punctuation, see next lemma. τὸ δὲ κλέος οἶον ἀκούω: The present of ἀκούω can be ‘timeless’ (LfgrE s.v. G). I take Hermes’ point to be that he has heard Apollo’s report about the thief but has not seen him (276). He returns to the contrast of seeing and hearing from 263, using a formulation resembling Il. 2.485–6 ὑμεῖς γὰρ θεαί ἐστε, πάρεστέ τε ἴστέ τε πάντα, | ἡμεῖς δὲ κλέος οἶον ἀκούομεν οὐδέ τι ἴδμεν, Hes. fr. 204.61–2 ὄφρ[̣ α ἴδοιτο] | … Ἑλένην μηδ’ ἄλλ̣ων ο̣ἶο̣ν̣ ἀ̣κ[ούοι. Hence αἵ τινες αἱ βόες εἰσί should be punctuated and performed as an aside, and this second half of the line finishes off the misleading oath. Normally having ‘only heard’ implies having no direct knowledge, but here it contrasts only with seeing what is outside oneself. Finally, as Hermes argued that Apollo’s accusation would be a θαῦμα (270), apparently because it is absurd, but really because the crime was a marvel, so he ends by claiming that he has only κλέος, apparently 346

The frequency of perjury cannot be assessed from rhetorical sources, which magnify the crime as if it were rare, but have constant recourse to it as a sign that an opponent is corrupt (384*).

301

C O M M E N TA RY 2 6 0 – 2 7 7 because he has no knowledge of the crime, but really because it will bring him fame.347 The alternative (e.g. Càssola) is to take the ‘report’ which Hermes has heard to be about the cows, with the clause attached loosely to αἵ τινες … εἰσί. Hermes’ speech then trails off in logical tightness, the implications of κλέος are less detailed, and he lapses into an outright lie. 260–77 Too clever by half: Hermes employs a number of more or less serious rhetorical strategies to fend off Apollo’s accusations. These aptly characterize the future patron of rhetoric, and their technicality contrasts with Apollo’s brand of rhetorical forcefulness at 334–64*. Hermes projects indignant surprise that Apollo would be looking for cows on a wild mountaintop (261–2), protests with anaphoric negatives that he has no information (263– 4), contrasts his physical and behavioural characteristics with those of cattle thieves (265–8, 271–3), predicts third-party scepticism at the claim which will embarrass Apollo (269–71), and offers a cunningly worded oath of apparent denial (274–7). The third and fourth elements play on his claim to be an ordinary baby – the opposite of his self-presentation with Maia at 163–5.348 This is supported by a linguistic veneer of childishness consisting of short clauses (most pronounced at 263–4) and repeated words (264 μηνύσαιμι/μήνυτρον, doubled μέμηλε in 266–7), use of plural-for-singular pronouns (267, 278), and a high proportion of end-stopped lines.349 All this complements Hermes’ physical pretence to be a swaddled and sleepy baby at 237–42; indeed he picks out particular characteristics in 267–8 which recall that. Hermes’ pretences, both verbal and physical, are self-refuting. Here the cunning wording of the oath (275–7), the self-satisfied wordplays on θαῦμα and κλέος (270, 277), the wilful misunderstanding in 264 of Apollo’s instruction μήνυε, and the brazenness of picking out hard terrain as a reason that he could not have stolen the cows when he left Apollo footprints in the sand (273), all point to the fact that this baby is anything but ordinary. In fact, Hermes betrays in 270 and 274 that he thinks of himself as a member of the Olympian family. The inferential path from Hermes’ babyish appearance to his behaviour is therefore, as Apollo already knows, not straightforward. This undermines his claims in 265–73, which are based on the normal lack of overlap between the categories ‘baby’ and ‘human cattle-rustler’. The nature

For interplay between κλέος as ‘report’ and ‘fame’ in the Odyssey see Goldhill (1991) 93–108. For ἀκούω in contexts of ‘hearing’ the content of one’s own reputation see DGE s.v. II.3. 348 Knudsen (2012) situated Hermes’ ability to modulate his self-presentation within developments in fifth-century rhetoric. 349 The end-stopping was pointed out by Van Nortwick (1975) 93–5. He r­ ather overstated the extent to which it recurs in Hermes’ defence-speech at 368–86. 347

302

C O M M E N TA RY 2 7 8 of the argument here may well have seemed a further example of rhetorical over-cleverness. It has attracted attention as being among the first arguments from εἰκός (i.e. ‘X is unreasonable/implausible/unnatural, so do not believe/ do X’), and is signalled as such by ἔοικα at its start (265); see Introduction §1.2. Hermes’ overall argument appears in summary near the end of Ezra Pound’s Canto 24, framed by allusions to the Otreus story of H.Aphr. at the end of 23 and in 25.350

278–81 278–9 ὣς ἄρ’ ἔφη, καὶ πυκνὸν ἀπὸ βλεφάρων ἀμαρύσσων | ὀφρύσι ῥιπτάζεσκεν ὁρώμενος ἔνθα καὶ ἔνθα: Hermes does three things with his eyes: he produces ἀμαρύγματα; these are dispatched rapidly using his eyelids, i.e. by their being fluttered; and he glances in different directions. I will discuss the meanings attributable to these gestures, then the details of how to interpret ὀφρύσι and ῥιπτάζεσκεν. We saw at 45* that the principal connotations of ἀμαρύγματα are seductiveness, fierceness and madness (the latter two are not relevant here), though the juxtaposition with 43–4 raised the possibility of a less attested relationship to intense thought. Here that possibility is stronger since Hermes’ internal glints fit with the contrast at 237–42 between fiery inner activity and passive exterior. In contrast to 45, here ἀπὸ βλεφάρων substitutes for ἀπ’ ὀφθαλμῶν. Though βλέφαρον often means ‘eye’, the focus on fast movements of the eyelids is eminently suitable.351 Quick blinking or winking could be taken as a gesture of guilt: Aristotle says that men are blinkers, starers, or middling, and that the blinker is fickle (HA 1.10 492a10–12); Paphlagon dares the Sausage-seller to deny a crime without blinking at Ar. Eq. 292, where Σ glosses σκαρδαμύττειν with πυκνῶς βλεφαρίζειν – similar to Herm.’s phrase.352 Gaze-shifts too are cross-culturally among the gestures best correlated with

350

Davenport (1983: 224) suggested that Hermes’ appearance characterizes the trickiness involved in Niccolò d’Este’s diplomatic and financial success. Pound also cites Hy. 6 from Dartona (1537) in Canto 1; near the end of Canto 79 ‘Κόρη καὶ Δήλια [sic] καὶ Μαῖα | trine as praeludio | Κύπρις Ἀφροδίτη’ points to the sequence of the long proemic Homeric Hymns, and connects the first three to earth, moon and stars: Pearlman (1969) 273. 351 For adverbs and accusatives like πυκνόν qualifying a look, see Rakoczy (1996) 42–3; compare also Il. 6.484 δακρυόεν γελάσασα, Pi. P. 9.38 χλοερὸν γελάσσαις ὀφρύϊ, etc. For the sense, compare πυκινός of fast-oscillating feet at H.Pan 20, 23. 352 Pelliccia (1995: 74) suggested that Hermes blinks to pretend he is still adapting to the light after waking (cf. Herm. 361). This gesture would be undercut by the shifting glances which follow.

303

C O M M E N TA RY 2 7 8 – 2 7 9 fear and so guilt. On Greek eye-contact see Cairns (2005); at Aristainetos 1.22.33 Doris nods μόλις, ἐφ’ ἑκάτερα παρακινοῦσα τὸ βλέμμα to fake embarrassment. One interpretation uniting all three eye-gestures is therefore that Hermes’ glints betray scheming while his blinks and eye-shifts betray guilt.353 But there is an alternative: Clement Paed. 3.11.70 equates shifting looks and τὸ βλεφαρίζειν with διὰ τῶν ὀφθαλμῶν μοιχεύειν. Thus all three gestures could be seductive; Hermes’ flirtation would not be erotic, of course, but cute and charming; like much flirtation, however (see Il. 14, Hes. Op. 373–4, etc.), it would be deceptive. Compare 387 ἐπιλλίζων*. The problems of interpreting ὀφρύσι and ῥιπτάζεσκεν are separable from the ambiguity discussed so far. One sense of ῥιπτάζω is ‘toss and turn in bed’, which is difficult to suppress completely when Hermes is in bed, his sleep disrupted.354 However, if the verb has that meaning here, Hermes is again betraying himself by failing to observe the restraints of swaddling (compare 151–2, 237–40), thanks to which only his head should turn. (I doubt whether ὀφρύσι belongs with this sense, i.e. that ‘tossed and turned with his eyebrows’ meant ‘wiggled (only) his eyebrows’.) The primary sense of ῥιπτάζω here is ‘throw’, with an object (ἀμαρύγματα) supplied from the preceding participle.355 For the common metaphor ‘casting glances’ see e.g. Od. 4.150, A. Su. 1004–5, Ag. 240–1, 742; the frequentative verb-form would suit πυκνόν. The dative of ὀφρῦς, as the location of a gesture, can mean the muscles around the eyes (e.g. in smiles at H Dem. 358, Pi. P. 9.38), and these are more pertinent here than the brows specifically, which are generally raised in grief and pride or knitted in anger (LSJ s.v. ὀφρῦς). ὀφρύσι could then be taken with both ῥιπτάζεσκεν and ἀμαρύσσων. Cf. the datives at X. Cyn. 6.15 ἀστράπτουσαι τοῖς ὄμμασιν, Ps.-Opp. C. 3.90 βλεφάροισιν ἀπ’ ὀφθαλμῶν ἀμαρυγαί | ἱμερόεν στράπτουσι, Ach. Tat. 1.4 καταστράπτει … τῶι προσώπωι.356 On this interpretation, the nuance ‘tossed and turned’

353

Children as young as three have been found to suppress non-verbal signs that they are lying, though they often give themselves away with an overconfident smile. The right verbal signals take longer to learn. See Talwar and Lee (2002), Keltner (2005) on signs of embarrassment in children. 354 There is an absolute use in the Hippocratic Diaet. Acut. 8.46 ῥιπτάζειν. More common expressions are ῥιπτάζε ν ἑαυτόν or ῥιπτάζεσθαι. They normally indicate sleeplessness caused by internal annoyance (e.g. Ar. Lys. 27–8, Hp. Morb. 2.17). 355 For the Greek flexibility in mentally supplying a noun extracted from a preceding verb see e.g. Th. 1.22.1 μέλλοντες πολεμήσειν ἢ ἐν αὐτῶι ἤδη ὄντες, KG i.34. 356 These parallels show how Hsch. α.3472 ἀμαρύσσων: λάμπων, ἀστράπτων could relate to our passage (or to G.Naz. PG xxxvii 1553.8, also of sight). The gloss λάμπων was in fact added here in T by a later hand.

304

C O M M E N TA RY 2 8 0 remains, but only as a hint that Hermes is pretending to have been woken when he would prefer to be asleep (cf. 267). Eitrem (1906: 270) suggested an allusion to Hesiod fr. 294.2, which uses ὁρώμενον ἔνθα καὶ ἔνθα of Hermes’ enemy Argus. But the point of the allusion is unclear, and the phrase, also in Orph. fr. 132 (see Bernabé’s corrigenda), could have been reached independently. 280 μάκρ’ ἀποσυρίζων, ἅλιον ὡς μῦθον ἀκούων: μακρός frequently qualifies noises, to mean ‘carrying a long way’ (LfgrE s.v. II.2), not ‘for a long time’. The compound ἀποσυρίζω may imply ‘whistle in denial’ (compare e.g. ἀπεῖπον, ἀπόμνυμι, ἀπομυθέομαι; LfgrE s.v. ἀπό I.3bβ), but means simply ‘whistle out’ at Lucian VH 2.5. Apollo’s μῦθος (also 253, 261) contained an accusation, an order and a threat. Orders and threats are ἅλιος if they will not be fulfilled: so of oaths at Il. 4.158, 5.715, S. Tr. 258, a prediction at Il. 18.324, declarations of intent at 24.92 and 224, and Zeus’s intent at Od. 5.104 = 138. Accusations could be ἅλιος because misdirected, rather as the formulaic ἅλιον βέλος misses its target.357 Both Hermes and the narrator know that at least the accusation in Apollo’s μῦθος was not ἅλιος. This rules out the variant τόν (Ψm) for ὡς, which is easily explicable as a way of avoiding the metrical anomaly ἅλιονˉ ὡς, not necessarily under the influence of Il. 5.715 ἅλιον τὸν μῦθον ὑπέστημεν. For the timeless present ἀκούων, compare 277*. Hermes’ simulation of carefree innocence is not a use of whistling attested elsewhere in Greek sources, whence the participial phrase of explanation.358 We do hear of singing and humming used as ways to fill a tense silence (A. Ag. 16–17 with Thomas 2013: 491–2; Ar. Nu. 720–1 with ΣTh), which suggests that the simulation is visibly forced. This is another instance of Hermes demonstrating skills which undermine his pretence to be as innocent as a normal baby, and like the previous two lines, it shows Herm.’s interest in exploring a wide range of means of communication, including the difficulty of reading gestures as signs of guilt or innocence (Introduction §3.4). Finally, the far-sounding whistling will prove proleptic when Hermes invents the far-sounding panpipes at 512*.

Snell (1964: 172–3) argued that the application of ἅλιος to words is derived from its use of missiles. Compare 315 νημερτέα* for Greek metaphors of ‘misdirected’ words. 358 For whistling in pastoralism and disgust see Van Stekelenburg (2000). It seems to have generally been a low-status gesture, and this may affect Hermes’ self-­ presentation. 357

305

C O M M E N TA RY 2 8 1 – 2 8 2 281 τὸν δ’ ἁπαλὸν γελάσας: ἁπαλὸν γελάσας may have been formulaic, though it is only extant in Epos here and at Od. 14.465. There, wine produces such laughter in a sensible man; Athenaeus 5 180a gives the plausible gloss ἄνανδρον. But the context here foregrounds the connection of ἁπαλός to children, as in 273*. So with the ἁπαλός laugh of the young Achilles at Philostr. Im. 2.2.2. The repetition of ἁπαλός from 273 may suggest that Apollo’s laughter is a suitable, less brutal response to Hermes’ mischief. His shift of attitude is not obviously predictable after his earlier bluster: Greeks did not always enjoy being lied to (e.g. Prometheus enrages Zeus). Nor was it Hermes’ aim to make Apollo laugh (260 κερδαλέοισιν, 318 ἤθελεν ἐξαπατᾶν). However, there was a tradition of appreciating a third party’s clever lies, which must help humour conquer outrage here. Hermes’ gestures are ridiculous; they and his speech surprise Apollo, who combines an indulgent laugh to a naive inferior ­(similarly at Od. 13.287 Athena smiles as she sees through Odysseus’ lies) and a laugh at a ‘silly’ argument (e.g. Hecat. FGrH 1 F1a); for these reasons for laughing, see Arnould (1990) 25–30, 48–9, 86–90. Herm.’s quartet of laughs make a pattern: see on 29, 389, 420.

282–92 282 ὦ πέπον ἠπεροπευτὰ δολοφραδές: ὦ πέπον is an epicism. The two principal uses are: (i) as positive politeness in the context of giving orders, criticism or advice in a risky situation – somewhat like ‘my dear fellow/ lady’ in old-fashioned English (so in the Iliad, Od. 9.447, 22.233, Sc. 350, and more menacingly Sc. 357); (ii) when one immortal advises a less powerful one, generally without a threatening backdrop. Our passage and Hes. Th. 560 combine the two. Apollo will give Hermes an order (290) to avoid danger, and the degree of menace is difficult to gauge, poised as we are between the soft laugh of 281 and the critical ἠπεροπευτὰ δολοφραδές (see also 289*). In Hesiod, Zeus addresses Prometheus with open hostility, sarcastically self-quoting Th. 544. For Herm.’s use of the Prometheus episode of the Theogony see Introduction §1.3. 282–4 ἦ σε μάλ’ οἴω | πολλάκις ἀντιτορέοντα δόμους εὖ ναιετάοντας | ἔννυχον: Apollo’s use of οἴω in predictions of something to be deprecated echoes Maia’s words of 156 (see 155–61*, Introduction §3.2). The emphatic position of πολλάκις suits Apollo’s distasteful tone. But unwittingly Apollo echoes Hermes’ threat to burgle the temple at Delphi (177–8 ὀΐομαι … δόμον ἀντιτορήσων); since Delphi is more secure even than εὖ ναιετάοντας houses, he underestimates Hermes’ ambitions, which are

306

C O M M E N TA RY 2 8 4 actually directed against him.359 For Hermes’ piercing and nocturnal trickery, see Introduction §3.3. Thefts in Athens were treated as aggravated if they occurred at night. According to Demosthenes, ‘Solonian’ law stated εἰ δέ τις νύκτωρ ὁτιοῦν κλέπτοι, τοῦτον ἐξεῖναι καὶ ἀποκτεῖναι καὶ τρῶσαι διώκοντα, καὶ ἀπαγαγεῖν τοῖς ἕνδεκα εἰ βούλοιτο· τῶι δ’ ἁλόντι ὧν αἱ ἀπαγωγαί εἰσιν … θάνατον τὴν ζημίαν (24.113); in the imagined state of Plato’s Athenian, νύκτωρ φῶρα εἰς οἰκίαν εἰσιόντα ἐπὶ κλοπῆι χρημάτων ἐὰν ἑλὼν κτείνηι, καθαρὸς ἔστω (Lg. 9 874b). Such self-help was inadmissible against many other categories of theft. Cohen (1983: 72–6) explained that the victim was particularly vulnerable at night, whereas such a licence against diurnal thefts could easily be abused. This dramatic nature of nocturnal burglary neatly suits the traditional syntax of ἔννυχον: to concentrate an adverbial phrase into such a predicative adjective is marked (Schwyzer ii.79), and in Epos ἐννύχιος and ἔννυχος always come at line-start, x4/7 markedly postponed. There is comparable drama in Il. 11.682–3 ἠλασάμεσθα … |ἐννύχιοι (a theft), 715–16 ἄγγελος ἦλθε… |ἔννυχος (an ambush), Hes. fr. 195.30–2 ὦρτο … |ἱμείρων … |ἐννύχιος (Zeus’s seduction of Alcmene). 284 οὔ χ’ ἕνα μοῦνον ἐπ’ οὔδει φῶτα καθέσσαι: Tucker, cited in Allen (1912), gave the right analysis of the first three letters, καθέσσαι was noted in passing by Van Herwerden (1907: 186) and Wackernagel (1916: 63–5) discussed its superiority to Ω’s καθίσ(σ)αι, which is a secondary Attic spelling. Apollo’s thought, if stated directly, would involve κε + aorist subjunctive referring to a possible future event, as in Il. 11.431–3 ‘either you will boast … or κεν … ἀπὸ ... ὀλέσσηις your heart, wounded by my spear’. In both passages, and at 1.184 ἐγὼ δέ κ’ ἄγω Βρισηΐδα, the speaker presents a threat with grim understatement as a mere possibility. Sitting on the ground was a sign of distress and supplication; related to the latter, it is the position of the beggar.360 These sitters are either distressed by their losses, or have no furniture left. For the latter, compare the Greek proverb urging restraint in case the cicadas sing on the ground, i.e. when your trees have been destroyed (Arist. Rhet. 1395a2, citing Stesichorus; cf. AP 7.723); at LXX Sir. 11.1–5, in the ephemerality of external glories εὖ ναιετάοντες+ brought flexibility to the formulaic system constituted by ἐϋκτίμενος/ εὖ ναιόμενος. See Shipp (1961) 43–7. The transferred sense of the participle of ναιετάω (‘provide habitation’ rather than ‘dwell at’) has numerous parallels in Greek and elsewhere: Debrunner (1944), Fraenkel (1951) 374. 360 Distressed sitting in Epos: Il. 1.349, 18.25, 24.163–4, Od. 4.539, 718, 10.567, 21.55, H.Dem. 98. Supplication: Il. 21.115, Od. 7.153. Beggars: e.g. Od. 17.339, 446, 18.17, 110. 359

307

C O M M E N TA RY 2 8 5 – 2 8 6 πολλοὶ τύραννοι ἐκάθισαν ἐπὶ ἐδάφους – an image not found in the extant Hebrew.361 285 σκευάζοντα κατ’ οἶκον ἄτερ ψόφου: σκευάζω here is ‘gather possessions’, as in the middle at Lysias fr. 197 ἐπειδὴ πάντες κατέδαρθον, ἐσκευασμένος τῶν χαλκωμάτων ὅσα οἷός τ’ ἦν …, Dinarchus fr. 35 σκευασάμενοι τὰ ἐκ τῆς οἰκίας. The stem brings an appropriate connotation of artifice, as in ‘dressing up’ food or actors, and σκευάζω means ‘trick’ colloquially at Men. Sam. 599 (see Gomme and Sandbach 1973). The pun recalls that in 167* βουκολέων. ψόφος often refers specifically to footsteps rather than other clatter (e.g. Naupactia fr. 7.2 ποδοψοφίη, S. Tr. 967, fr. 269c.22, E. Tro. 887, Ba. 687, Cratinus fr. 360), so that the line recalls Hermes’ silent steps at 149. The register of this line may be pointedly ‘unepic’, perhaps to mark Hermes’ outlandish and inappropriate behaviour. ψόφος and σκευάζω are both extant in early lyric (Sappho fr. 44.25, Archil. fr. 140) but very rare in Epos (ψόφος presupposed in Naupactia fr. 7; σκευ- nowhere else). οἷ’ ἀγορεύεις: Echoing 272 ἀγορεύεις|, and so matching Hermes’ tactic of drawing a negative implication from the opponent’s words; see further 282–92* below. 286 πολλοὺς δ’ ἀγραύλους ἀκαχήσεις μηλοβοτῆρας: After lines describing Hermes’ burglaries of κειμήλια from wealthy houses, |πολλὰς δ’ ἀγραύλους (matching 283 |πολλάκις) switches focus to a complementary rural setting and the theft of πρόβασις. The traditional polarity (Watkins 1994: 657–62) suggests that Hermes will master thefts of all kinds. 288(a–b) shows that μηλοβοτῆρας is a shorthand for shepherds and cowherds, as in Il. 18.528–9.362 Herdsmen are regularly related to the ἀγρός (e.g. Il. 18.162, Hes. Th. 26; cf. ἀγροϊώτης, ἄγρ-οικος), but ἀγραύλους here is more directly a quotation from Hermes’ speech (262*), countering the claim that ἄγραυλοι cows belong to a different sphere; Apollo in fact saw μῆλα on Kyllene (232). See further 287*. ἀκαχήσεις: Related to ἄχος, a more acute form of upset than ἄλγος (see LfgrE s.vv. on the verbs accompanying them). The form of the verb allows us to hear Hermes’ future epithet ἀκάκητα converted into a symbol of his harm.363 The tone of this is unclear, since the meaning of ἀκάκητα is unclear. 361

I am grateful to Hugh Williamson for advice on this point. μῆλα themselves very rarely include cows, but see Simonides fr. 253 (μῆλον of a bull, which surprised Ar. Byz.), A. fr. 158.3–4 μυχηθμοῖσι καὶ βρυχήμασιν… μήλων, Ps.-Theoc. 25.119, Ap.S. p. 112.17 μῆλα· κοινῶς μὲν τὰ τετράποδα ... 363 For the development ἀκαχήσω < ἀκάχησα, a sigmatic substitute for ἤκαχον (: *ἄχνυμι :: ἤγαγον : ἄγ-ω), see Risch (1974) 248, GH i.397–8. For ἀκάκητα see Il. 16.185, Od. 24.10, Hes. fr. 137. 362

308

C O M M E N TA RY 2 8 7 – 2 8 8 Many glosses offered ‘removing κακά’, which would make the wordplay seem ironic (Baumeister). However, EM 44.55 gives the gloss πανοῦργος, which may already have been Hesiod’s idea when he transferred the epithet to Prometheus (Th. 614). 287 οὔρεος ἐν βήσσηισ’, ὁπότ’ ἂν κρειῶν ἐρατίζων: Greeks who wanted to keep even small herds, and could afford supervisors, required extra summer pasturage in uplands for when crop-fields were in use. Though the journey to these pastures could in reality be short, many were situated in hilly borderregions, and both mountains and herdsmen were conceptually opposed to the civilization of sedentary farmers – a conception which sharpened the sense that they were at risk from brigands such as Hermes.364 The contexts of the formula οὔρεος ἐν βήσσηισ(ι) are more limited than those of βῆσσα, and most emphasize wildness – the location for beasts and hunting, wild winds, and wild-fires. The first of these connections contributes to Apollo’s image where Hermes, hyperbolically κρειῶν ἐρατίζων, attacks livestock like a lion (see 64*). The image pointedly rejects Hermes’ self-presentation as a harmless baby, but brings dramatic irony: Apollo does not yet realize that Hermes was κρειῶν ἐρατίζων against his own herd. The reference to mountain glens also echoes Maia’s prediction in 159 κατ’ ἄγκεα φιλητεύσειν. 288 ἀντήσεις ἀγέληισι βοῶν καὶ πώεσι μήλων/ἀντᾶις βουκολίοισι καὶ εἰροπόκοισ’ ὀΐεσσιν: The Homeric Hymns contain a set of doublets, where early variants were transmitted together (Introduction §5.1). Assuming this pair is part of the set rather than a concordance interpolation, the ἄντην transmitted at the start of 288(b) needs emendation. Schneidewin advocated ἀντᾶις (1848: 679; cf. Hes. Op. 408 τητᾶι) or hyper-Ionic ἀντῆις. With the present stem, ἀντάω means ‘search out’, as opposed to ‘come upon’ in future and aorist.365 In 288(a), ἀντήσεις is transmitted in Da (and, by

364

Regarding the practices of herding in the poet’s Greece (the natural reference for the future tense in an aetiological hymn), and the difficulties in our sources, I found Georgoudi (1974), Cherry (1988) and Hodkinson (1988) particularly useful; Shaw (1982: 8–24) analysed conceptions of herdsmen. Cows as well as flocks inhabit the uplands of epic: see e.g. Il. 13.571–2, 20.188–9. For use of border hills see e.g. Th. 5.42, Hell. Ox. col. 18.3, and mythically S. OR 1133–9. Aristotle connects pastoralism to brigandage (Pol. 1.8 1256b5), while Str. 3.3.5 says of Portuguese tribes ἦρχον δὲ τῆς ἀνομίας ταύτης οἱ ὀρεινοί, καθάπερ εἰκός· λυπρὰν γὰρ νεμόμενοι καὶ μικρὰ κεκτημένοι τῶν ἀλλοτρίων ἐπεθύμουν. On the general opposition of mountains to settlements see Buxton (1992). 365 For the morphology of -άω verbs in Ionic see Smyth (1894) 526, 566–72. The epic subjunctive from ἄντομαι would be ἄντηαι not ἄντηι.

309

C O M M E N TA RY 2 8 9 – 2 9 0 error, in ΓC) and may be an original short-vowel subjunctive which other manuscripts regularized (see 43–4*). Both lines then give good sense and fit comfortably into a formulaic context. ἀγέληισι βοῶν καὶ πώεσι μήλων is a declension of Il. 11.696 ἀγέλην τε βοῶν καὶ πῶϋ μέγ’ οἰῶν, itself part of a flexible system of ways to combine ἀγέλη and πῶϋ. For εἰροπόκοις ὀΐεσσιν see Il. 5.137. 289 ἀλλ’ ἄγε, μὴ πύματόν τε καὶ ὕστατον ὕπνον ἰαύσεις: The normal order is ‘ἄγε, μή ’, but Apollo shifts the threatening words forwards. The collocation πύματόν τε καὶ ὕστατον (as adverbial neuters) also occurs at Il. 22.203, Od. 4.685, 20.13, 116. In those four cases the pleonasm implies ‘last before death’, though that is not central to the sense of either adjective individually; here Hermes is facing a correspondingly final stint in Tartarus.366 One may infer that death was part of the formula’s traditional referentiality. Apollo phrases the threat in response to Hermes’ professed love of sleep (267). ἰαύσεις (M) is the lectio difficilior compared with Ψ’s ἰαύσης, and gains credibility from the variant 288(a) ἀντήσεις (not in M); but M’s ending could also be a sonic error. 290 ἐκ λίκνου κατάβαινε: As Apollo realizes, Hermes is in fact mobile underneath his swaddling. (Hermes will not admit this until 304.) Apollo’s initial purpose in moving Hermes is unclear: perhaps he plans to frighten him into revealing the cows’ hiding-place, but a type of ἀπαγωγή where one person arrests another and frog-marches him to a magistrate is also possible. In Athens this was a procedure for dealing with thieves caught red-handed (Cohen 1983: 40–4, 58–61; Harris 1994), though Apollo’s epistemological powers mean that the ‘red-handedness’ prescription cannot simply be transferred to our context. The reference to Hermes’ cot contrasts with 150 λίκνον ἐπώιχετο, which suggests itself as a possible framing device signalling the end of this episode in Maia’s cave (cf. 285* ~ 149). Hermes’ next move will confound that expectation: 299*. μελαίνης νυκτὸς ἑταῖρε: Apollo continues his focus in 284 ἔννυχον. Unwittingly, he also relates Hermes to his lyre (31* δαιτὸς ἑταίρη) and echoes 67, where Hermes’ theft was like a plot μελαίνης νυκτὸς ἐν ὥρηι|.367 The collocation of μέλαινα with νύξ is formulaic (Il. x10, Od. x3, Hes. Th. x4), but the See 256*. Vergados noted the possible borrowing at Q.S. 13.27 πανύστατον ὕπνον ἴαυον, of the Trojans’ last night. 367 See also 436*. For Hermes and his lyre see n. 55. Hecate is νυκτὸς … ἑταίρη in a hymn at Hippolytos Ref. 4.35.5, as is silence at Triphiodoros Il. Pers. 503. 366

310

C O M M E N TA RY 2 9 1 – 2 9 2 three uses in Herm. form a series: 358* echoes this line, which echoes 67 to suggest the aptness of Apollo’s description. 291–2 τοῦτο γὰρ οὖν καὶ ἔπειτα μετ’ ἀθανάτοις γέρας ἕξεις· | ἀρχὸς φιλητέων κεκλήσεαι ἤματα πάντα: γάρ shows that Apollo is explaining the address μελαίνης νυκτὸς ἑταῖρε and/or why Hermes should cooperate: both are in fact pertinent. οὖν throws emphasis on τοῦτο (Denniston 445–6), while καί means ‘actually’ before the temporal adverb, as in 302, 385, 438.368 This emphasis marks a change of opinion, based on Hermes’ words – not that Apollo is taken in by those words: he still sees Hermes’ thievishness. But he predicts an outcome for it which alters the threat of tartarization (256–9), with several corresponding features: ἤματα πάντα recalls the finality of his earlier threat (cf. 289*), ἀρχὸς φιλητέων ~ 259 ὀλοοῖσι(?) … ἡγεμονεύων; however, κεκλήσεαι ... μετ’ ἀθανάτοις significantly opposes ἐρρήσεις … μετ’ ἀνδράσιν in Tartarus. In this respect Apollo unwittingly echoes, and even goes beyond, Hermes’ ambitions to ἤματα πάντα μετ’ ἀθανάτοις ὀάριζειν (170) or else to φιλητέων ὄρχαμος εἶναι (175*), since now the gods will acknowledge his quasi-military organization of crime. Hermes’ earlier ambition for wealth and τιμή comes out in both γέρας – which includes the power of leading robbers (cf. Hecate’s γέρα at Hes. Th. 429–52) and the title of leader (see e.g. H.Aphr. 30–2, Call. H.Art. 7–25 for titles as γέρα) – and the implication of honorific language in κεκλήσεαι. Apollo may say γέρας with some sarcasm (cf. Posidipp. 135 for a similar light-hearted claim). But previous occurrences of the stem suggest that Hermes is quite satisfied. He set up his butchery to symbolize that his future γέρας would be distinct from but equal in scale to those of other gods (122, 128–9*); now Apollo foretells a unique γέρας. He will no longer have to offer γέρας to others (as at 60), and will become the target of invocations in his own right, as at the end of Herm. itself, or – with more pertinent wording – as in Hipponax fr. 3a Ἑρμῆ … φωρῶν ἑταῖρε, δεῦρό μοι. 282–92 Apollo’s second speech to Hermes: Apollo’s speech responds verbally to Hermes’ main claim. Hermes’ ‘ἀπρεπέως ἀγορεύεις| that I could have crossed the threshold βουσὶ μετ’ ἀγραύλοισι’ (270–2) produces Apollo’s ‘given what ἀγορεύεις|, you will often pierce houses and pain many ἀγραύλους herders’. πολλάκις, πολλούς, ἀρχός, and ἤματα πάντα are emphatically placed rejections of Hermes’ claim of innocence, and Apollo adds house-burglary to the discussion of rustling livestock (283–5 + 286–8). Compare Il. 16.498–9 καὶ ἔπειτα … ἤματα πάντα, 1.213 (καί ποτε), Od. 8.510, 21.24, Hes. fr. 33a.18; Denniston 319. Hermes’ title lies solely in the future, so that καὶ ἔπειτα cannot mean ‘also/still in the future’; nor is ἔπειτα ‘in that case’, since ἤματα πάντα steer us to take it temporally.

368

311

C O M M E N TA RY 2 9 3 – 2 9 4 But Apollo’s words also echo, with irony, Hermes’ statement of intent in 170–8 (292*, 283–4*), and his appetite for cattle from 64 (287*). The last intratext, along with those to 67 (290* νυκτὸς ἑταῖρε), 149 (285*) and 150 (290*), suggests that we are reaching a juncture-point and another crossing of Maia’s threshold – but Hermes will find a way to wriggle out of that for a little longer. Apollo updates some features of this speech at 456–62: see 436–62*.

293–8 293 ὣς ἄρ’ ἔφη, καὶ παῖδα λαβὼν φέρε Φοῖβος Ἀπόλλων: Apollo picks Hermes up, apparently not waiting to see whether he will move when threatened (289–90); but unlike the threat of 256 λαβών, he carries Hermes rather than hurling him. ὣς ἄρ’ ἔφη, καί … echoes 278, drawing attention to the difference between Hermes’ chicanery and Apollo’s direct action which responds to it. However, Apollo’s action also recalls 39–40 (ὣς ἄρ’ ἔφη, καί … φερ-), where Hermes picked up the tortoise after a prediction of its future esteem (35). This alignment of Hermes and the lyre (n. 55) will be deepened by 294–6*. 294–6 σὺν δ’ ἄρα φρασσάμενος, τότε δὴ κρατὺς Ἀργεϊφόντης | οἰωνὸν προέηκεν ἀειρόμενος μετὰ χερσίν, | τλήμονα γαστρὸς ἔριθον, ἀτάσθαλον ἀγγελιώτην: In 294–5, Hermes appears to despatch a literal οἰωνός, to match that of 213 (see further 297*, 303*, Introduction §3.2). Like most οἰωνοί this one is sent by a specific god, and its occurrence between a prediction (292) and a departure suggests two principal lines of interpretation. If it is a σύμβολον foretelling a successful journey, what success and for whom? Alternatively, it may confirm 286–8 and 292; see 299*. However, οἰωνός can be extended to non-avian omens (e.g. X. An. 3.2.9 a sneeze; cf. ὄρνις at Ar. Av. 717–22; much later Palladas AP 9.484 relates ὄρνεον to farts and winds), and 296 forces some reinterpretation. Several factors suggest that Hermes farts rather than burps. προέηκεν in Epos is particularly used for despatching a message, omen (e.g. Od. 2.147 a bird-omen), wind or offensive missile. All four uses are relevant to a fart here, and whereas the Greeks regularly called farts ‘wind’ (e.g. Ar. Pax 175, Ra. 1096–7; Hippocratic φῦσα; Katz 1999: 319), this is less natural for burps.369 Moreover, Hermes’ emission is meaningful (οἰωνός, ἀγγελιώτης) and to my knowledge burps were

369

It may be relevant that (προ)ἵημι is used by the Hippocratics for defecation (López Férez 1999), and that farting has an amusing interaction with 292 ἀρχός (also ‘fundament’).

312

C O M M E N TA RY 2 9 4 – 2 9 6 not considered so whereas numerous Greek sources attribute meaning to farts – contempt, fear, or in comedy prophetic significance.370 Aside from the prophetic readings outlined above, possible readings of Hermes’ deliberate fart include a signal of contempt, an attempt to disgust Apollo, a simulation of fear or of a baby’s physiology (when babies are picked up they often release air swallowed during feeding, or produced by their immature digestion). The last point again shows Hermes’ ability to manipulate the parts of an infant’s condition which were considered restrictive (cf. 151, 235–42, 263–4).371 For the interpretation which Apollo chooses see 299–303*. The elaborate phrasing gestures towards the embarrassment of talking about farts directly, while really relishing the humour of the topic.372 There is bathos in reducing omens to gods’ farts, and this is not the defence expected of κρατὺς Ἀργεϊφόντης, of all the ways to describe Hermes at this moment (he is rather on the receiving end of Apollo’s κράτος). Comedy revelled in bathetic, exuberantly indirect language: compare Eubulus’ periphrasis for πρωκτός (fr. 106), and the excerpts which Athenaeus 10 448b–59c drew from Clearchus. Meanwhile 294 σὺν … φρασσάμενος puns on συμφρασσόμενος, the kind of constriction which forces air from its natural circulation (see esp. Hp. Flat. 7–8), again emphasizing that Hermes uses intelligence where other babies are victims of their physiology. The neologism ἀγγελιώτην ­probably mobilizes ridicule of the suffix -ιώτης in pretentiously inflated job-titles, as recorded by Pollux 10.17 τὸν μέντοι σκευοφόρον ἐν Ταξιάρχοις Εὔπολις ‘σκευοφοριώτην’ (fr. 285) παίζων ἐκάλεσεν.373 ἔριθος may pun on ἐριθεύς ‘robin’ – a bird, though paltry for an οἰωνός. In any case, the range of ἔριθος (female maid or ‘help’; male ‘hireling’ with derogatory connotations, later ‘unprincipled

370

Lilja (1972: 139–42) discussed farts from comedy which indicate fear or contempt; see also POxy. 413 col. i, and Hdt. 2.162.3 for a contemptuous messenger-fart. For prophetic farts see e.g. Ar. Eq. 639, Daniel (1985: 129–30) on SEG 34.1051 (sixth century ce). They are a comic fantasy; Aristotle says ὁ πταρμὸς … σημεῖον οἰωνιστ κὸν καὶ ἱερὸν μόνον τῶν πνευμάτων (HA 1.11 492b7–8). 371 For deliberate farting compare Ar. Ra. 9–10, Hp. Prog. 11 ἢν μὴ ἑκὼν οὕτω [noisily] ποιῆται ὁ ἄνθρωπος τὴν ἄφεσιν τῆς φύσης, Augustine CD 14.24 nonnulli ab imo sine paedore ullo ita numerosos pro arbitrio sonitus edunt, ut ex illa etiam parte cantare uideantur. 372 For the topic as embarrassing, compare euphemistic φῦσα in Hippocratic authors; Phot. α.2732 ἀποψοφεῖν: τὸ πέρδεσθαι, εὐσχήμως λέγων· εὐσχημονέστερον δὲ ‘διαπνεῖν’ καὶ ‘ἀποπνεῖν’; Machon Chreiai fr. 11.158–9; Suda β.261 βὴξ ἀντὶ πορδῆς: … οἱ πέρδοντες λανθάνειν πειρώμενοι προσποιοῦνται βήττειν. 373 Compare 301 σπαργανιῶτα, 436 μηχανιῶτα. The suffix -ιώτης most commonly denotes ‘dweller in’, and attaches preferentially to nouns in -ια: Redard (1949) 9. ἀγγελιώτης has the latter characteristic, σπαργανιῶτα the former. For imitation of ἀγγελιώτης by Callimachus and Nonnus see Introduction §4.

313

C O M M E N TA RY 2 9 7 politician’) brings out the full range of τλήμων, from ‘long-suffering’ (as if Hermes’ flatulence were a messenger sent on various errands; cf. 568 ταλαεργός) to ‘brazen’ – a frequent quality of the epic stomach (e.g. Od. 7.216–17 οὐ γάρ τι στυγερῆι ἐπὶ γαστέρι κύντερον ἄλλο | ἔπλετο).375 The circumlocution has a less frivolous side. Both at ἔριθος and more obviously at the climactic ἀγγελιώτης, the fart resembles Hermes, who is sometimes the gods’ dogsbody, and also a flying (and currently airborne, as Baumeister noted) messenger; his action is also brazen and ἀτάσθαλος.376 Compare Hermes’ resemblances to another of his products, the lyre (293*): indeed, ἀειρόμενος μετὰ χερσίν recalls 39 χερσὶν ἅμ’ ἀμφοτέρηισιν ἀείρας; both fart and tortoise are potential σύμβολα (omens for the start of a journey; cf. 30*); tortoise and Hermes make unexpected noises, and later the possibility of a lyre doing so is added (488*). Finally, the hymnist’s phrasing is quadruply apt for Hermes’ childish prank of making a fart into an omen. Riddles were characteristic of both prophecy and childish games;377 Hermes demonstrates his creativity and sets Apollo a puzzle of interpretation – both of which the hymnist matches in his circumlocution (see Introduction §3.4). 374

297 ἐσσυμένως δέ: Echoing 215 |ἐσσυμένως δ’, which immediately followed the other οἰωνός. This confirms the value of contrasting Apollo’s reactions to the two ‘signs’ (see also 303*). ἐσσυμένως is, as normal, ‘energetically’; for the occasional extension towards ‘promptly’ see LfgrE s.v. 1 on e.g. Il. 3.84–5 ἄνεώι τ’ ἐγένοντο | ἐσσυμένως.

374

Female hired worker: Hes. Op. 602–3, S. fr. 286 (spider as spinner), Theoc. 28.1. Female helper: Od. 6.92, Ar. Pax 786, Simias fr. 16. Male hired worker: Il. 18.550, 560, Amphis Erithoi. For the political dodger (Arist. Pol. 5.3 1303a16), see also ἐριθεύω, ἐρίθεια, ἐριθευτός. Hsch. ε.5708 glossed ἔριθος as ἐρεθιστής. If this relates to an extant passage, ours is the best match. Elsewhere ‘slaves [vel sim.] to the belly’ means ‘gluttons’ (e.g. E. fr. 282.5, X. Mem. 1.6.8, Lucian AP 11.410). 375 The sense ‘brazen’ is first extant elsewhere at A. Ch. 384, though possible in Anacr. PMG 347.7. Wilson (1971) tracked the stem τλα- shifting from ‘prepared to go through with [sc. unpleasantness or danger]’ to ‘overly bold’ and ‘suffering’. Katz (1999) suggested that τλήμων γαστήρ could parody the formulaic collocation of (πολυ)τλήμων and θυμός, since the θυμός and γαστήρ are opposed (see e.g. 132), and both associated with wind. 376 Hermes’ ‘servitude’: e.g. PV 941–3 τρόχ ν … διάκονον … ἀγγελῶν, 954, 983 ὑπηρέτης, 966–9 λατρείας, S. fr. 269c.21 ἄγγελον… τρόχιν, 35 λάτρις, 269d.22 λάτρι[, E. Ion 10 λάτρις, and esp. Luc. D.Deor. 4. 377 For Greek riddles, see 550–66*; Waern (1951); Schultz (1909), esp. 22–81. Plato R. 5 479bc and Ath. 10 451b are explicitly riddles for children, while children pose riddles in Hclt. 22 B56 ~ Ps.-Hdt. Vit.Hom. 35.

314

C O M M E N TA RY 2 9 7 μετ’ αὐτὸν ἐπέπταρε: Belief in the significance of a healthy person’s sneezes, which are generally uncontrollable and so numinous, is attested from Od. 17.541–50 and throughout pre-Christian Greek culture, and elsewhere.378 Significant sneezes occur just after a wish or prediction, or at the start of a new undertaking (a kind of σύμβολον), where they can be taken as omens for good or bad. ἐπιπταίρω is a rare poetic compound, but the ἐπι- seems to mean ‘in confirmation’.379 As with the fart, there are many possible interpretations of a sneeze at this juncture: physiologically, babies often sneeze in their first few days; it may foretell success (but for whom?) on the journey they are starting, or confirm Apollo’s prediction at 292. Alternatively, as a deliberate sneeze, it might have no validity as a sign; compare Macedonius Epig. 39 Madden, where someone vainly tries to harm his wife by deliberately sneezing beside a tomb.380 Rather than there being rules for interpreting a sneeze ‘correctly’, the witness was generally a ‘layman’ who had to provide a proactive interpretation that a sneeze was significant, what it referred to, and whether it was positive or negative. Loud and/or repeated sneezing was probably more numinous (see ἐσσυμένως, Od. 17.541–2 μέγ’… σμερδαλέον, Pease 1911: 441–2); identifying a significant sneeze is also easier in the context of other unusual signs (as here, Hdt. 6.107.3, Plu. Them. 13). Some sources mention the position and identity of the sneezer and the time of day as significant factors in interpretation, but often they are neglected.381 297–8 τοῖο δ’ Ἀπόλλων | ἔκλυεν: For the modification of ¦τοῦ δ’ ἔκλυε | compare Il. 9.571–2 τῆς δ’ ἠεροφοῖτις Ἐρινύς | ἔκλυεν. Such phrases occur outside prayer contexts also at Od. 10.311, H.Aphr. 215, H Dem. 39.

Pease (1911), Lateiner (2005). Some sources disdain the belief as δεισιδαιμονία, but presuppose its popularity: e.g. Philemon fr. 101, Polyaen. 3.10.2, Plu. Gen. Soc. 581–2, D.L. 6.48. 379 Sneeze confirms prediction: X. An. 3.2.8–9, N. D. 7.107 (ἐπι-). Confirms wish: Od. 17.541–50 (ἐπι-); probably Theoc. 7.96 (ἐπι-). Promotes new undertaking: Theoc. 18.16 (ἐπι-), N. D. 13.82 (ἐπι-). Obstructs new undertaking, perhaps because of the connection to illness (cf. the Greek for ‘bless you’, Ζεῦ σῶσον): Hdt. 6.107.3, ‘Arist.’ Prob. 33.11 962b21–2 ὁ μὲν πταρμὸς μᾶλλον δοκεῖ ἐπισχεῖν τοὺς ἀρχομένους, Hsch. ε.5104 ἐπισχετικὸν γὰρ ὁ πταρμὸς πολλάκις, Men. fr. 844.9, Polyaen. 3.10.2. 380 Greek methods for inducing a sneeze are discussed at ‘Arist.’ Prob. 33.4–5 961b36– 2a16. 381 Position of sneezer (compare augury): Plu. Them. 13, Gen. Soc. 581b, D.L. 6.48. Identity of sneezer: Philemon fr. 101, Theoc. 18.16, Plu. Gen. Soc. 581b. Time of sneeze: ‘Arist.’ Prob. 33.11 962b19–21 διὰ τί οἱ μὲν ἀπὸ μέσων νυκτῶν ἄχρι μέσης ἡμέρας οὐκ ἀγαθοὶ πταρμοί, οἱ δὲ ἀπὸ μέσης ἡμέρας ἄχρι μέσων νυκτῶν;. 378

315

C O M M E N TA RY 2 9 8 – 3 0 2 Apollo saw the previous οἰωνός (213 ἐνόει): for the switch towards the aural, see Introduction at n. 104. 298 ἐκ χειρῶν δὲ χαμαὶ βάλε κύδιμον Ἑρμῆν: Apparently in disgust, as in Thebais fr. 3 ἰσχίον ὡς ἐνόησε [sc. Oedipus, who expected shoulder], χαμαὶ βάλεν. If so, κύδιμον seems as parodic as 294 κρατύς. This fall is not treated as a moment of danger for Hermes, as it would be for other babies.

299–303 299 ἕζετο δὲ προπάροιθε: Apollo moves towards Hermes’ level to communicate: a degree of haughtiness has been removed. One might envisage him sitting on a chair or on the floor; the latter is appealing since it would be a partial confirmation of Apollo’s own prediction in 283–4, in line with the possible confirmatory nature of the fart and sneeze.382 καὶ ἐσσύμενός περ ὁδοῖο: 297 ἐσσυμένως resonated with 215 through its form and position, but the adjective here resonates with the content – Apollo’s eagerness to travel after an οἰωνός. Whereas the former omen made him hurry, here he is momentarily held back. This change marks the success of Hermes’ unexpected tactic for postponing the closure of the scene, which seemed imminent in 282–92*. 300 κερτομέων: See the summative note on 299–303* below. 301 θάρσει: A formulaic opening gambit in Epos (Il. x7, Od. x7, H.Hom. x3), also in Aristophanes, Plato and others. Normally the hearer is in particular need of moral support in the face of fear or upset. Hence the word regularly, as here, ousts a vocative into second place (Il. 8.39 = 22.183, 24.171, Od. 2.372, X. Cyr. 5.1.6, etc.), and a direct reason not to worry follows, often in asyndeton. For a relationship to caregiving register see 299–303* below. 302 εὑρήσω καὶ ἔπειτα: ‘I will actually find in the future’; see 291–2* for this use of καί. Apollo is building on 291, where he predicted Hermes’ future prerogative over theft; the intervening fart and sneeze, far from just confirming that prediction, allow Apollo to predict something positive for himself. βοῶν ἴφθιμα κάρηνα: 94*.

382

Eitrem (1906) 271. Tzifopoulos (2000: 159–60) interpreted Hermes to be turning the prediction of 283–4 back on Apollo, by making Apollo put him on the ground.

316

C O M M E N TA RY 3 0 3 303 τούτοισ’ οἰωνοῖσι: The demonstrative suggests a contrast with the οἰωνός of 213, which did not help Apollo find his animals. σὺ δ’ αὖθ’ ὁδὸν ἡγεμονεύσεις: The future tense hovers between a prediction and an order. ἡγεμονεύσεις| recalls 259* ἡγεμονεύων|, in the threat of tartarization. There, Hermes’ ‘leadership’ implied both guidance and command. When Apollo envisaged an alternative future for Hermes in 292, only the idea of command was incorporated. Here, his more prominent role as a guide is also converted into something honourable (cf. 392*). The echo suggests a further reason why Hermes can feel confident: the threat of Tartarus is being superseded. 299–303 Apollo interprets the omens: Our reading of Apollo’s speech is guided by 300 κερτομέων, which suggests dissimulation (55–6*). After Hermes’ ambiguous fart and sneeze, Apollo maintains ambiguity in his response. On one level, θάρσει pretends to assume that the fart was caused by fear of the threat in 289 followed by being hauled out of his cot (Bothe). Apollo allays that ‘fear’ by foreseeing that Hermes will find some form of leadership-role other than in Tartarus or among brigands, suggesting a level of freedom which is not his intention (as Hermes realizes: 307–8*). Simultaneously, Apollo pretends to treat Hermes as an eager collaborator worried about the cows, and wilfully interprets the fart and sneeze as good omens for the new venture to find them. Thirdly, he casts himself as a keen and caring child-minder: the satyrs use θάρσει similarly at A. Dict. (fr. 47a) 804 θάρσει δή· τί κινύρηι, where they are fussing over the baby Perseus’ body-language. With σπαργανιῶτα, compare the lexical creativity of the annoying father at Thphr. Char. 20.5 who calls his son a ποπανουργίαν (another hapax) τοῦ πάππου.383 Apollo’s humour and optimism mark a recovery from the instinctive disgust at 298 to the gentle laugh of 281.384 However, the focus on Hermes’ clothing is very unusual as a feature of an address, and insightfully raises the question whether the swaddling is a proof of Hermes’ infancy or a forum for his trickery. Aesop 91 is intriguingly similar to this passage. Hermes tests Tiresias’ prophetic power by stealing his cows. Tiresias then asks Hermes to describe auspices to him, and the second bird indicates that Tiresias will retrieve his cows ‘if Hermes wishes’. Apollo’s second οἰωνός was the fart which Hermes 383

See Diggle (2004) 397 for a convincing analysis of Theophrastus’ text and its humour. I owe the idea here to a conversation with Barbara Graziosi. On the form of σπαργανιῶτα see n. 373. 384 Contrast Vergados’ interpretation that Apollo’s speech is both deceptive and aggressive.

317

C O M M E N TA RY 3 0 4 – 3 0 5 delivered to Apollo at 294, and Apollo’s interpretation is that Hermes himself will lead Apollo to his cows.

304–12 304–5 ὣς φάθ’, ὃ δ’ αὖτ’ ἀνόρουσε θοῶς Κυλλήνιος Ἑρμῆς | σπουδῆι ἰών: Hermes moves quickly (θοῶς); σπουδῆι probably adds eagerness rather than being a pleonasm. The action refutes his pretence from 273 that he is unlikely to go walking (Bothe). The two uses of δ’ αὖτε in 303 and 304 prompt us to compare Hermes’ next move with what Apollo predicted for it, i.e. cooperative leading: is Hermes going to accede? 305–6 ἄμφω δὲ παρ’ οὔατα χερσὶν ἐώθει | σπάργανον ἀμφ’ ὤμοισιν ἐελμένος: M’s ἐελμένος can hardly be a scribal invention, and is defensible. Θ’s ἐλιγμένος and p’s ἑλιγμένος are both false forms, explicable as attempted metrical corrections of ειλιγμένος, which would be a gloss on the unfamiliar ἐελμένος.385 The echo of 151 σπάργανον ἀμφ’ ὤμοις εἰλυμένος suggests that ἐελμένος would be a transitive middle, ‘who had wrapped … around his own …’; this seems viable given Pl. Prt. 342c ἱμάντας περιειλίττονται ‘they coil straps around themselves’. Hence Schneidewin’s ἐελμένον (1848: 679) is probably not needed. We must give ἐώθει the object σπάργανον and take παρ’ οὔατα as ‘past his ears’ (cf. GH ii.121–3), since παρεώθει οὔατα is nonsensical, and παρ’ οὔατα … ἐώθει ‘he pressed [intransitively] against his ears’ is inappropriate before speaking (Radermacher). The image created is that a swaddling-band (possibly the outer wrap, 152* λαῖφος) passes round the back of Hermes’ neck, and he spends some time (imperfective aspect) pushing the ends back over each shoulder – perhaps to ensure he does not expose the lyre (Gemoll), or in preparation for travel.386 As at 151 and 235–42, Hermes swaddles himself from the inside, displaying his supernatural control of his babyhood – and this time while keeping a lyre tucked under one arm.

For εἴλω and its cognates see 239* ἀνέειλε. The appropriate breathing of ελίσσω is unclear: see the paradosis of Il. 23.309, and the mix of κατελίσσω and καθελίσσω in the Hippocratic corpus. For our textual problem see also Andrisano (1988–9). 386 At Musaeus Her. 163 πολλάκις ἀμφ’ ὤμοισ ν ἑὸν συνέεργε χιτῶνα, Hero is fidgeting in modest embarrassment. Hsch. α.4167 ἄμφωτος: χιτῶνος εἶδος (called ἀμφώτας in EM 93.15) is another garment passing the ear. For Hsch. α.4164 see 151*. 385

318

C O M M E N TA RY 3 0 7 – 3 0 9 307 πῆι με φέρεις: Baumeister noted the mild but indignant hyperbole, by which Hermes puts Apollo’s recently curtailed act of carrying him (293 φέρε, 298 ἐκ χειρῶν ... βάλε) in the present tense. On the other hand, Hermes shows his recognition that in 303 ἡγεμονεύσεις Apollo was not seriously predicting his freedom to guide. The pair of indignant questions in 307–8 recalls the start of Hermes’ previous speech to Apollo (261–2). ἑκάεργε, θεῶν ζαμενέστατε πάντων: ζαμενέστατος is a hapax, with unusual double intensification (ζα-, -τατος); the word will be echoed when Apollo’s anger ends for good in 495 (περι-ζα-μενῶς κεχολῶσθαι). μένος covers strong impulses from ‘martial eagerness’ to ‘anger’ as well as physical power; 308 χολούμενος and Hesiod’s use of ζαμένησε for ‘was angry’ (Th. 928) identify Hermes’ focus as anger, a sense otherwise clustered in Il. 1, 9, Od. 11, and the Hymn to Demeter. Despite Apollo’s aggressive epithets like ἑκάεργε (218*), Hermes’ outburst seems hyperbolic, like those of Achilles in Il. 22.15 ἔβλαψάς μ’ ἑκάεργε, θεῶν ὀλοώτατε πάντων, Menelaus at 3.365, or Diomedes at 10.164–7 – but perhaps here with a particularly childish quality (cf. 309*). 308 ὀρσολοπεύεις: ὀρσολόπος was analysed as ‘arse-skinner’ by Schwyzer (1923) 21–2. Pernée (1985) suggested that the metaphor originally concerned hunting-dogs, in which case it was transferred to the military rout (so Anacreon PMG 393, Ares ὀρσολόπος), and later to verbal insults (Maximus Init. 107, Phot. ο.528; see Perpillou 1996: 118–21).387 Hermes could appropriately describe Apollo in terms of his herd’s guard-dog (hence Hermes’ enemy) and/or a hunting-dog. Or this may be evidence of the verb’s extension to ‘wounding’ words, with reference both to Apollo’s κερτομία (derived from κέαρ τέμνων by e.g. ΣD Il. 1.539) and to his accusation, which Hermes reintroduces as an affront in the following lines. Either line of thought saves us from Vergados’ inference that Apollo has actually started spanking Hermes without the narrator telling us about either the start or the end of this interesting procedure. However, given the other exaggerations in 307 and 309, it could be that Hermes is describing Apollo’s treatment of him in terms of spanking as an exaggeration based on how children were punished. 309 ὦ πόποι: 219*. The exclamation rarely occurs mid-speech; here it starts a response to the imagined answer to 308.

Harder is A. Pe. 10–13 ὀρσολοπεῖται θυμὸς ἔσωθεν· πᾶσα γὰρ ἰσχὺς Ἀσιατογενὴς οἴχωκε, νέον δ’ ἄνδρα βαΰζει. If θυμός is the subject of βαΰζει, its confused jumpiness may be being described first in terms of being bitten by a dog, then as itself like a dog.

387

319

C O M M E N TA RY 3 0 9 – 3 1 2 εἴθ’ ἀπόλοιτο βοῶν γένος: Hermes’ (faked) complaint against cows uses the stereotyping idiom εἴθ’ ἀπόλοιτο, and generalises unreasonably from Apollo’s cows to the whole species. This continues the suggestions of a childish outburst (307*). Hippolytus’ similarly overextended wish against women (E. Hipp. 664–5) sounds similarly immature. The empty rhetoric will come across more forcefully when Hermes eventually takes over cattle and their generation (491–5). Barnes observed the Callimachean echo at Aet. fr. 110.48 Χαλύβων ὡς ἀπόλοιτο γένος. 309–11 οὐ γὰρ ἐγώ γε | ὑμετέρας ἔκλεψα βόας οὐδ’ ἄλλον ὄπωπα | (αἵ τινές εἰσι βόες), τὸ δὲ δὴ κλέος οἶον ἀκούω: Hermes more or less repeats 275–7* from his previous speech, though he omits the contorted oath, and makes 277 more assertive by introducing δή. The change from αἵ τινες αἱ βόες εἰσί to αἵ τινές εἰσι βόες does not mean that he is now claiming ignorance of cows generally: compare Il. 2.487 οἵ τινες ἡγεμόνες … ἦσαν, ‘who the leaders were’. Possibly, the near-verbatim repetition contributes to his self-portrayal as childlike and helpless. 312 δὸς δὲ δίκην καὶ δέξο παρὰ Ζηνὶ Κρονίωνι: Having first asked πῆι, then ‘why?’, then claimed that there is no good reason why, Hermes circles back to suggest a destination. He does so using technical legal language for submitting to arbitration, which was used at Athens in preference to litigation for a wide range of allegations including theft (Dem. 22.25–6; Roebuck 2001); for legalism in Herm., see Introduction §3.3.5. In litigation, the complainant demands and takes (λαμβάνειν, rarely δέχεσθαι) δίκη, and the defendant gives it; originally δίκη in such phrases was probably ‘adjudicated compensation’. In arbitration, the more ‘Hermetic’ option where the aim was negotiating a compromise rather than applying prescribed laws, each party is engaged in give and take. For δίκην διδόναι καὶ δέχεσθαι of interstate arbitrations see IG I3 6 A.41–2, 127.18, Th. 1.37.5, 140.2, 5.59, and Introduction §1.1; compare δίκην διδόναι καὶ λαμβάνειν at Hdt. 5.83.1, Ps.-X. A. P. 1.18 (intrastate arbitrations with foreign jurors), D. 53.20 (entirely set within Athens).388 For the form δέξο (which p conjectured away), see GH i.296, and for παρά + dative of an adjudicator see LSJ s.v. B II 3. A defendant also persuades the claimant to go to arbitration at D. 29.58. The poet does not try to differentiate private arbitration, public arbitration, and litigation in the divine world. The present situation resembles a public arbitration before a king, but also the archetypal private arbitration – Anaximander 12 B1 already has διδόναι … δίκην καὶ τίσιν ἀλλήλοις. Sometimes (e.g. Hes. Op. 225, Thgn. 45) the adjudicator is also said to ‘give’ δίκη, i.e. to impose it, and one of the parties to ‘take’ it.

388

320

C O M M E N TA RY 3 1 3 a family squabble referred to the κύριος. Zeus is an ideal arbiter on the one hand as king, father of Δίκη, and upholder of justice, and on the other as impartial through his shared fatherhood of the disputants.389 This was emphasized at the start of this scene by the parallel formulas describing both Hermes and Apollo as ‘sons of Zeus’ (227, 230; see also 323, 378*); his paternity of Hermes, however, is yet to be made public. Though the move to Olympus corresponds to the hymnic set-piece of a god’s introduction there, no independent treatment of the myth is known to have included an arbitration scene. Indeed, the narrator presents Hermes’ suggestion as a novelty, by contrast with 309–11, which recalled Hermes’ previous speech, and with 304–5, where Hermes seemed momentarily to signal obedience (the narrator’s teasing piece of misdirection). Hermes’ tone may be childishly blunt, as Artemis’ requests at Call. H.Art. 6–25 are characterized by |δός x5 (Bornmann 1968); this would complement the immature hyperboles of 307* φέρεις, ζαμενέστατε, 309* and perhaps 308* ὀρσολοπεύεις, and the repetitious nature of 309–11. However, a single |δός can also be taken as a friendly gambit and hence positive politeness.390 Hermes’ strategy is to grasp at an alternative procedure to the self-help which Apollo favoured in his initial anger; he has the ulterior motive of debuting in Olympian society. Apollo foresees that any arbiter will adjudicate in his favour, and is more interested in recovering the cows than in punishing Hermes. He may have been intending to take Hermes to Olympus anyway (290*).

313–21 313–21 Preliminary comment: Lines 315–18 refer to and summarize the discussion in 254–312, rather than describing a state which begins afterwards.391 They therefore constitute a parenthesis within the temporal clause which starts with 313 αὐτὰρ ἐπεί, glossing ἀμφὶς θυμὸν ἔχοντες in particular. In the event, the temporal clause has no main clause, and there is an anacolouthon of a fairly standard type: 318–19 revert to 313–14 (again juxtaposing the two protagonists after αὐτὰρ ἐπεί), but the parenthetical material on Hermes’ attempted deception leads away from the original structure ‘When they had 389

For Zeus resolving disputes among other gods see also the complaint of the baby Apollo against Gaia’s invention of prophetic dreams (E. IT 1270–83, with a similarly amused reaction to the young god’s ambition), the trial of Ares for killing Halirrhothius (e.g. IT 945–6), the arbitration of Athena and Poseidon (Apollod. Bibl. 3.179), and the dispute of Aphrodite and ­Persephone over Adonis (3.184–5). 390 Compare Il. 14.198, Od. 17.345, 415, 9.355 (with irony), Hes. Op. 453. 391 Richardson (1990: 31–5) discussed such summaries which round off one episode before a transition.

321

C O M M E N TA RY 3 1 3 – 3 1 4 conversed, they set off’ to ‘When Hermes failed to deceive, they set off’.392 We become more aware of the narrator’s presence, both in syntactical and evaluative terms (see 315–16* on οὐκ ἀδίκως), just before a scene on Olympus which may well be an innovation (312*). 313 αὐτὰρ ἐπεὶ τὰ ἕκαστα διαρρήδην ἐρέεινον: Since neither god has asked a genuine fact-finding question, ἐρέεινον is ‘explore’ (as, more literally, in 252).393 The echo of 252 ἐπεὶ ἐξερέεινε serves to frame the conversation between Apollo and Hermes. διαρρήδην (first here) is prosaic and mostly legalistic; the more common sense ‘explicitly’ is not apt for what we have seen of Hermes, but the adverb also means ‘with a detailed discussion’ (e.g. Pl. Lg. 12 932e διείρηται … διαρρήδην).394 The transition from conversation using just αὐτὰρ ἐπεί is unusual, though found in two traditional contexts: αὐτὰρ ἐπεί ῥ’ ὄμοσέν+ τε τελεύτησέν+ τε τὸν ὅρκον (Il. x1, Od. x4, H.Ap. 89), and αὐτὰρ ἐπεὶ τό γ’ ἄκουσε where the final turn of conversation requires action (Il. x3, Od. x5, H.Dem. 334). Compared with these structures, Hermes’ proposal in 312 is not a natural close to a conversation. This strains διαρρήδην: the suggestion is that Hermes has had enough and is off before Apollo can respond properly. 314 Ἑρμῆς τ’ οἰοπόλος: οἰο- should indicate ‘alone’, and οἰ- ‘sheep’. However, Greeks occasionally related οἰοπόλος and οἰονόμος to sheep, assisted by e.g. αἰ-πόλος: cf. Leonidas 19.1 G–P οἰοπολεῖτε, Anyte 3.2 G–P οἰονόμος, Apollodorus in Ap.S. p. 119.25, Hsch. ο.380 οἰοσφάγωι. One can work in both meanings, since Hermes’ solitary travels during the night will lead him into the world of pastoralism (571). καὶ Λητοῦς ἀγλαὸς υἱός: As noted on 60*, the word ἀγλαός passes from ‘radiant’ through ‘beautiful’ towards ‘inducing pride and joy’. The last nuance is to the fore as parental focalization in the formulaic descriptions ἀγλαὸς υἱός and ἀγλαὰ τέκνα (compare also E. Hel. 11 Εἰδώ, τὸ μητρὸς ἀγλάϊσμ’); hence my translation ‘treasured’.

392

For Epos’ flexible uses of appositions and parentheses see GH ii.15–16, 356, Schwyzer (1939) 14–15. They can be viewed as imitating the greater flexibility of sentence-structure in conversation: Kakridis (1976); Bakker (1997), esp. 39–44; cf. KG ii.590–1 on Plato’s anacoloutha. 393 p restored ἐρέεινον, whereas MΘ have ἐρέεινεν: scribes expected the thought to be ‘After Apollo had cross-examined him, …’. 394 Jane Lightfoot pointed out to me a possible purely sonic influence of Il. 1.6 διαστήτην ἐρίσαντε. Compare the resemblance of 339* to Od. 10.191. Van Nortwick (1975: 29) independently compared 314 to Il. 1.7 since lines where two noun–epithet combinations confront each other are unusual.

322

C O M M E N TA RY 3 1 5 – 3 1 6 315 ἀμφὶς θυμὸν ἔχοντες: Contrast 391*. For the formulaic system of ‘having one/different hearts’, see LfgrE ii.1083.34–43; for the spatial metaphor in ἀμφίς compare e.g. Il. 2.13–14 ἀμφὶς … ἀθάνατοι φράζονται, 13.345 τὼ δ’ ἀμφὶς φρονέοντε, S. Ant. 510 τῶνδε χωρὶς εἰ φρονεῖς, and ἀμφισ-βητέω, ἀμφιλ-λέγω. 315–16 ὃ μὲν νημερτέα φωνέων | οὐκ ἀδίκως ἐπὶ βουσὶν ἐλάζετο κύδιμον Ἑρμῆν: For the status of this clause as an apposed summary of 313 see above, 313–21 Preliminary comment*. Ω had νημερτέα φωνὴν … ἐλάζυτο … Ἑρμῆν. To start with the verb, I have tentatively preferred ἐλάζετο since this is the form found elsewhere (x11) in Epos, and much more often in subsequent hexameter (Theoc., A.R., Moschus, Nic., AP, Orphica, etc.); Euripides’ love of λάζυμαι may have encouraged its spread during transmission.395 Apollo ‘took’ Hermes literally, as in 293 λαβών. The legalistic ἐπὶ (LSJ s.v. B III.1 ‘on a charge relating to’) βουσίν suggests this is an arrest, and there is no evidence for Ludwich’s idea that λάζομαι could mean ‘accuse’ or ‘chastise’ (1908: 118). νημερτέα φωνήν is now stranded syntactically, and the adjective is traditionally the object of a verb of speaking, so that the simplest correction is Wolf ’s φωνῶν (or, as I prefer, φωνέων: see n. 136). For φωνέω + neuter adjective in the fifth century see LSJ s.v. Plausibly a scribe wrote φωνήν to agree with νημερτέα. νημερτέα (< *n-̣ privative + ἁμαρτ-) is one of a series of archaic Greek adjectives for speech, or occasionally thought, which does not fail in some way: compare ἀ-ληθής, ἀ-τρεκής, ἀ-ψευδής versus ἐτεός ‘true’, and see Cole (1983). The ‘target’ which νημερτέα words ‘do not miss’ varies between truth and truthfulness.396 νημερτέα φωνέων identifies ὃ μέν as Apollo, since Hermes has been lying unclearly; cf. 369*. The adjective frequently applies to the truth of prophetic predictions, such as those made by Apollo for Hermes’ future in 292 and 303; indeed, H.Ap. associates Apollo with announcing Zeus’s νημερτὴς βουλή at 132, 252, 292. However, the parallelism of οὐκ ἀδίκως with *n-̣ ἁμαρτ- shows that it is Apollo’s accusations which are primarily at stake. This narratorial comment οὐκ ἀδίκως is uncharacteristically judgemental for primary narrators in the

Ptolemy Pindarion preferred λάζυμαι in Homer (ΣAT Il. 8.389), but his views were largely ignored. λάζυμαι is a variant in Theoc. 18.46 known to Hesychius λ.88; cf. Ps.-Opp. C. 2.11. Bechtel i.281 suggested it was analogical to αἴνυμαι. Hippocratic Ionic shows both λάζυμαι and λάζομαι. 396 When applied to prophecies, ‘unerring’ emphasizes truth (cf. ἁμαρτοεπής rejecting a prediction at Il. 13.824). At e.g. Od. 3.327–8 the focus seems to be more on Menelaus’ lack of deception. See Luther (1935) 33–43, and contrast Snell (1975) 13–15. As at 280 ἅλιος*, the underlying metaphor is of ‘casting’ words (e.g. LSJ s.v. ἵημι I.2, A. Ag. 785–7). 395

323

C O M M E N TA RY 3 1 7 – 3 2 0 Homeric tradition, especially when the exceptional nature of the ‘case’ means that it is not clear what right Apollo has to seize Hermes.397 317 αὐτὰρ ὃ τέχνηισίν τε καὶ αἱμυλίοισι λόγοισιν: The τέχναι contrasted with λόγοι here are the physical tricks of 76*, 237–42, 278–80, 294–7. Proteus (Od. 4.455) and octopuses (Opp. H. 2.232–40) also make their bodies a source of artifice. Hermes’ αἱμύλιοι words oppose νημερτέα φωνέων (cf. A.R. 1.792 αἱμυλίοισιν vs 797 νημερτές), and live up to 13 αἱμυλομήτην* in the introductory list of his epithets. The stem suits his words both in their attempt to deceive Apollo (cf. Pi. N. 8.33 πάρφασις, αἱμύλων μύθων ὁμόφοιτος, δολοφραδής, bearing in mind Herm. 282 δολοφραδές), and in their probably wheedling tone of innocence (see 263–4*). 319 For the anacolouthon see above, 313–21 Preliminary comment*. πολύμητις πολυμήχανον takes the opposition in 318 Κυλλήνιος ἀργυρότοξον and brings out the balance and similarities between the brothers, which have led to impasse. Hermes’ μῆτις centres more on subtlety and on planning than Apollo’s μηχαναί, which centre on being ready with a practical response to them; 299– 303* was particularly quick-witted, and he has not been close to the ἀμηχανία with which he threatened Hermes (257).398 ηὗρεν also suggests the balanced situation: Apollo found Hermes (though not his cows); here we find that Hermes has made his own explorations of (313) and discoveries about Apollo. 320 ἐσσυμένως: Reiterating Hermes’ eagerness to travel from 305 and Apollo’s from 299. The stem is carefully placed in Herm. (cf. 299*~215): our instance restores the intratextual framing suggested in 290*~150 but temporarily broken by Hermes’ antics. Now we can move on from the framing device to a scene-change, and the gods have a clearer sense of direction. διὰ ψαμάθοιο βάδιζεν: Hermes chooses to retrace his route (75) – potentially to sustain his pretence that walking on newborn feet is difficult (273), and certainly to give Apollo some normal prints to contrast with the mystifying tracks running parallel to them. The hymn seems to represent Hermes as not yet able to fly (70*); certainly, doing so would undermine his claims of innocence.

397

See De Jong (1987) 136–46, 197–208; Richardson (1990) 37–41, 161–4. ­Direct evaluations are found particularly in introductions to a character or immediately before or after direct speech. 398 For the relationship of the stem μηχ- to situations displaying μῆτις, see 346–9, Martin (1983) 14–26. Extant epic mainly uses πολυμήχανος as a substitute for πολύμητις in πολυμήχαν’ Ὀδυσσεῦ|. This does not imply that the adjectives had identical meanings.

324

C O M M E N TA RY 3 2 1 – 3 2 2 321 πρόσθεν, ἀτὰρ κατόπισθε Διὸς καὶ Λητοῦς υἱός: Hermes leads – as predicted in 303, eagerly, and in preparation for his role as guide. But the position also enables Apollo to ‘shepherd’ him along (cf. Thphr. Char. 18.8): he drives north a different beast from what he expected (see 330*).

322–9 322 αἶψα δὲ τέρθρον ἵκοντο: Compare 142 Κυλλήνης δ’ αἶψ’ … ἀφίκετο: the contrast is between Hermes’ two possible homes, his father’s realm and his mother’s (see 1–9*, 167–72). The reading αἶψα δ’ ἵκοντο κάρηνα (xmp) was probably designed to clarify the rare and obscure word τέρθρον. Since κάρηνα is largely confined to hexameter, this could be a rhapsodic variant, or a gloss inspired by a particular source such as line 142. However, the choice of τέρθρον (against e.g. H Dem. 184 αἶψα δὲ δώμαθ’ ἵκοντο) is significant. Herm.’s usage is most closely related to four early instances where τέρθρον appears in the context of doors and narrow passageways. These associations, even if they do not allow us to reconstruct a single best translation (not LSJ’s ‘summit’, at any rate; nor ‘roof ’, which Hsch. τ.526 may have extracted from our line), have plausible thematic resonance in Herm. Etymology suggests τέρθρον should be a place or instrument for occurrences of *terh1/3- ‘rub, bore’, *terh2- ‘cross’, or *dher- ‘hold upright’.399 In Apollodorus PMG 701 someone arrives ἐπὶ τέρθρον θυράων at night. This suggests that when in Euripides fr. 371 (if we skirt the likely corruption of other details) Heracles bemoans having to traverse the τέρθρον of Hades, he is thinking of the Underworld’s doorway, whose appearances in Near Eastern and Greek sources are discussed at West (1997) 141, 156–9. Senses such as ‘passageway’, ‘fastening-point’ or perhaps ‘threshold’ (where doorposts are fixed) would suit. Empedocles 31 B100 uses the words πυκιναῖς τέτρηνται ἄλοξιν | ῥινῶν ἔσχατα τέρθρα διαμπερές (3–4) apparently to describe the inner end of the nostrils, which are pierced with pores which blood cannot pass out of, but which air passes through during breathing (see O’Brien 1970). The word may again describe a constricted passageway, or the point where something (the air-channel within the nostril) is anchored in a surface (the interface with blood-vessels). At Hp. Mul. 125 τέρθρον is the epicentre of a case of wandering womb, which πνίγουσιν: this suggests ‘constriction-point’.400 The

399

For -θρον see Chantraine (1933) 372–4. The suffix remained productive (e.g. in medical language), and τέρθρον could have originated after the ­laryngeal in *terh- was lost. 400 Erot. Voc. Hp. p. 35 describes the Hippocratic usage as obscure.

325

C O M M E N TA RY 3 2 3 – 3 2 4 word’s associations seem to be with doorways, narrow passages and fastenings. Applied to the context in Herm., these turn this arrival into another act of making a difficult crossing through a closed doorway (146), or passing over a threshold (23, 233). For the palace gates on Olympus compare e.g. Il. 8.411. The other main fields for the stem τερθρ- lie in verbal fussiness (τερθρεία, τερθρεύομαι; from Isoc. Helen 4 on) and ships’ rigging (τέρθρον, τέρθριος). I see no cogent way for these to help our understanding of the stem’s development or usage here, though a hint of the former could be relevant in setting up the rhetorical cleverness of the upcoming debate.401 θυώδεος Οὐλύμποιο: Apart from this phrase (also at H Dem. 331, Q.S. 7.557), Olympus seems rarely to be described for its smell. I take the scent not to derive from plants like thyme which grow on the mountain, nor to be transferred from sacrifices conducted elsewhere, but to be that typical of divine residences including Kyllene (231*; 237* θυήεντ’). 323 ἐς πατέρα Κρονίωνα Διὸς περικαλλέα τέκνα: For Epos’ use of εἰς/ ἐς ‘to’ with a person, see GH ii.103–4. For Διὸς περικαλλέα τέκνα| see 397*. Following the juxtaposition of words for the gods in 318–19, this line unites them through their shared descent from Zeus and Cronos, an idea implicit in 312* as a reason both are happy to turn their dispute over to Zeus. 324 κεῖθι γὰρ ἀμφοτέροισι Δίκης κατέκειτο τάλαντα: τάλαντα fundamentally means a ‘balance’, used commercially to compare two weights (already

401

Lexicographers defined the terms from rigging in incompatible ways: they may not have been experts, and usage may have varied over time. (i) In Ar. Eq. 440–1 τέρθριοι are ropes slackened when a strong wind moderates; S. fr. 333 probably reapplied a τερθρία wind to farting in fear (Lämmle 2013: 73–4), possibly playing on τέρθρον as a narrow passageway (above) or ring (below). These ropes are probably clewlines on a sail’s lateral edges, by which a first corner of sail can be let out (cf. Σ Ar. Eq. 440a τέρθριοι = ἔκφοροι, Phot. η.204 ἔκφοροι = clewlines). Similarly Gal. Gloss. p. 145, Hsch. τ.525, EM 752.48. Erot. Voc. Hp. p. 126 makes τέρθριοι ‘brails at the end of the mast’, but brails belong on the sail, so read ἐπὶ τέλει τοῦ ἱστου. (ii) However, Σ Ar. Eq. 440a also says that τέρθριοι were fastened in the bow; Hsch. τ.527 positions a τερθρωτήρ on watch there, and Σ A.R. 1.564–7 lists sheets, πρόποδες, τέρθριοι and μέσοι as a series. Hsch. τ.526 τέρθρον = ‘foresail’ may be related. In images in Morrison and Williams (1968), only forestays and braces are attached in the bow, not clewlines or sheets. (iii) Phot. τ.174 defines τέρθρα as rings on the mast for halyards. Richardson (1945) suggested that τερθρεία ‘verbal fussiness’ derived from a naval role such as that of the τερθρωτήρ. Or might a sense ‘threshold’ relate to the raised foredeck of a ship and a platform from which sophists performed? A naval application to a pair of ropes tied in the same ring could explain why Poll. 2.134–5 seems to use τέρθρα for the sternocleidomastoid muscles.

326

C O M M E N TA RY 3 2 5 Il. 12.433–5); since a fair measurement required precise suspension, the balance became a symbol of even-handed justice.402 In the Iliad, Zeus uses one to demonstrate an inequality between the fates of duelling warriors; later, balances are used by various divinities, often (from c.500 bce on) including personified Justice.403 I therefore read τάλαντα Δίκης and construe it as a metonym for ‘a just weighing-up’ of their cases. ἀμφοτέροισι emphasizes fairness, and (rather like 323*) contrasts with 315 ἀμφ-ίς; cf. 391, 397, 507 on expressive uses of ἀμφ-. 325 εὐκολίη δ’ ἔχ’: M has εὐμιλίη and Ψ had εὐμυλίη. Neither is extant or readily interpretable, and the text immediately below may also be in error. When ἔχω governs a place, inanimate subjects give the local conditions – often weather or sound, but also including light, smell and pleasure.404 My suggestion rests on the minuscule confusion of μ/κ, and occurs in verse in SEG 34.27 (Attica, 370s), IG II2 12405 (fourth century?), Adaeus 6.4 GP. Other proposals which are palaeographically, linguistically and semantically reasonable include συλλαλιή (Càssola; συλλαλέω is attested from Plb., LXX), εὐθαλίη (Zimmermann 1920: 236; εὐθᾰλέω first in Nic.), and εὐμαρίη (Bothe; see LSJ s.v. εὐμάρεια for this rare spelling).405

402

Bühler (1999) on Znb. Prov. 2.64, Diotimus 1 G–P. Ridgeway (1888: 111–12) connected τάλαντα here to money staked in legal cases (Il. 18.507, Isoc. Paneg. 46, etc.), but Herm. evidently precludes any financial depositions. 403 The scales of Zeus were probably a near-Eastern import: see Dietrich (1964) 111–25, West (1997) 393–4. For Justice’s balance see B. 17.25–6, 4.11, A. Ag. 250–1, Ch. 61, Σ Arat. 88. In art, however, Δίκη is never shown with a balance, and Δικαιοσύνη is only identified by it from the Roman period on (LIMC s.vv.). 404 Weather: Od. 12.75–6 οὐδέ ποτ’ αἴθρη | κείνου ἔχει κορυφήν; 13.245 αἰεὶ δ’ ὄμβρος ἔχει [sc. Ithaca]; LSJ s.v. κατέχω II.4. Sound: Hy. 26.10 βρόμος δ’ ἔχεν… ὕλην, LSJ s.vv. κατέχω II.2, ἐπέχω VI. Light: H.Ap. 445 Κρίσην κάτεχεν σέλας. Smell: Hermippus fr. 77.9 ὀσμὴ … κατὰ πᾶν ἔχει … δῶ. Pleasure: Od. 9.6 ἐϋφροσύνη μὲν ἔχηι κατὰ δῆμον ἅπαντα, Sc. 284–5 πᾶσαν δὲ πόλιν θαλίαι τε χοροί τε | ἀγλαΐαι τ’ εἶχον. 405 Linguistically dubious conjectures include D’Orville’s εὐμελίη (n.d.: f. 251), Franke’s εὐελίη, οὐμιλίη (Allen 1912, tentatively), εὐκηλίη (Allen and Sikes) and εὐωχίη (West 2003a: 150); for the rarity of iota as glide in Epos outside proper names see 116*. Semantically dubious conjectures include στωμυλίη and εὐνομίη (both D’Orville n.d.), αἱμυλίη (Heyne in Groddeck 1786: 90); Hermann’s ἐμμελίη (perhaps at Suda ε.972) is unconvincing since Apollo has been absent. Palaeographically unconvincing: εὐδίη (Baumeister), εὐφροσύνη (Führer 2004), ἀδμωλή (Bergk), αἰθρίη (Schmitt 1856: 154). ­Allen, ever keen to accept the manuscripts, also defended both εὐμυλίη (1897; refuted by Radermacher) and εὐμιλίη (Allen and Sikes, based on an implausible etymology of ὅμιλος).

327

C O M M E N TA RY 3 2 5 – 3 2 6 Ὄλυμπον ἀγάννιφον: The formulaic epithet (here, 505, Il. 1.420, 18.186, Hes. fr. 229) is still apt for most of the year. The doubled nu reflects *sneigwh-, as 185 βαθυ-ρρόου < *sreu-. 325–6 ἀθάνατοι δέ | ἄφθιτοι ἠγερέθοντο: For the aspect of ἠγερέθοντο (‘were gathered’ not ‘were gathering’) see GH i.327–8. Ω’s reading presents a so-called ‘violent’ enjambement between subject and attributive adjective, which is hard to parallel.406 However, the possibility of enjambing a patronymic (523–4 Ἀπόλλων | Λητοΐδης) and the frequent reduplication of words expressing immortality (A. fr. 28 ἀείζων ἄφθιτον, E. Andr. 1256 ἀθάνατον ἄφθιτόν τε, ‘Ion’ FGE 1 ἄφθιτον ... ἀενάοις, Isyll. 11, Orph. fr. 691.14 ἄφθιτον ἀθάνατον, ‘Pythagoras’ Carm. Aur. 71 ἀθάνατος θεὸς ἄμβροτος) call for caution about emending. ἄφθιτος developed from ‘undiminishing’ to cover nature-deities (Hes. Th. 389), then immortalization (Mimn. fr. 4), then the Olympians (first parallel in Pi. P. 4.33; Treu (1965) 8–15). Given the corruption above, ἁθρόοι (following Groddeck 1786: 89 ἀθρόοι; see 106* for the breathing) deserves consideration. It precedes two of the three occurrences of ἠγερέθοντο before caesura (Od. 2.392, 24.468), and cf. ἀολλέες ἠγερέθοντο| x3. 326 μετὰ χρυσόθρονον ἠῶ: So abm. Ω’s ποτὶ πτύχας Οὐλύμποιο repeats ‘Olympus’ clumsily, and the gods should gather from the πτύχες. The variant may have arisen from a marginal annotation such as Il. 11.75–7 (the gods’ homes κατὰ πτύχας Οὐλύμποιο) or 20.5 (Zeus summons an assembly from πολυπτύχου Οὐλύμποιο). χρυσόθρονον is anthropomorphic, μετά not: for such slippage, compare Il. 23.227 κροκόπεπλος ... κίδναται Ἠώς. Dawn’s goldenness is readily understandable (cf. Latin aurora ~ aurum). The second element of the adjective was probably interpreted as θρόνος (as later in scholia) rather than θρόνα (‘embroidered flowers’, as in Il. 22.441 ἐν ... θρόνα ποικίλ’ ἔπασσεν). The former suits Dawn’s epithet ἐΰθρονος, ὑψίθρονος, and the application of ἀγλαόθρονος to water-nymphs in Pindar and Bacchylides; for χρυσόθρονος itself compare the golden divine thrones of e.g. Il. 8.442, 14.238. However, Dawn herself was not depicted enthroned (LIMC s.v. Eos, ΣHMN Od. 6.48). Possibly -θρονος epithets arose from θρόνα, and tenaciously retained their connection to females even after that derivation had been forgotten.407 See Higbie (1990) 28–65 on Il., Edwards (1966) 122–48. δέ occurs at line-end Il. x10, Od. x10, Thgn. x2, always after a bucolic diaeresis; the other examples do not have a comparable enjambement. 407 West (2007: 221 n. 90) noted a Latvian parallel for Dawn’s golden dress. On -θρονος epithets, see Gerber (1982) 4 n. 2, Olson (2012) on H.Aphr. 218–19. Theocritus, Lyco406

328

C O M M E N TA RY 3 2 8 – 3 3 0 The lemma continues the pattern of marking new sections with a timeindication, and reminds us how quickly the action has moved since 184. For the gathering of gods shortly after daybreak, compare Il. 1.533 (a δαίς in Zeus’s house, soon after 497 ἠερίη), 8.1–2, Od. 5.1–4; Epos’ human meetings also frequently start in the early morning (Il. 2.48–51, 19.1 + 45, Od. 2.1–7, 8.1–5, etc.). 328 πρόσθε Διὸς γούνων: Why the ‘knees’ specifically? Perhaps simply because they are the foremost part of a seated person. However, we ­anticipate that Zeus will recognize Hermes publicly as his son in this scene, and the lap is a site for welcoming a newborn into the family (so of grandparents at Il. 9.455, Od. 19.401; of Zeus as father at Call. H.Art. 4). Moreover, the brothers are referring an uncertain decision to a higher authority, i.e. θεῶν ἐν γούνασι κεῖται.408 ὃ δ’ ἀνείρετο φαίδιμον υἱόν: Zeus naturally questions Apollo first, as the older and perhaps visibly angrier of the two; it turns out that he also wants to make a joke at Apollo’s expense. φαίδιμος is not used of Hermes or Apollo elsewhere in Epos, but is more likely to suggest the latter: the ancient (false) etymological link between φαιδ- and φάος or φαίνω (see Pindar fr. 109.3 φαιδρὸν φάος, Hdn. Gr. ii.142) makes φαίδιμος more relevant to one who was probably already verging towards the sun-god (n. 248). 329 Ζεὺς ὑψιβρεμέτης: This, and its alternative ἐριβρεμέτης, favour contexts of Zeus’s destructive or punishing power, sometimes explicitly as a god of thunderstorms (e.g. Od. 23.331, Hes. Th. 568). But our passage is not the only exception (Il. 1.354, Od. 5.4).

330–2 330 Φοῖβε: The name is common in literature, remarkably rare in cult. φοιβός seems to mean ‘bright’, and later Greeks linked Φοῖβος to φάος (e.g. Ps.-Zon. φ.1816, Et.Gud. p. 555), so that the choice of title could relate to

phron and Nicander use θρόνα of ‘plant-drugs’; cf. Cleitarchus’ glosses in Σ Theoc. 2.59–62: Θεσσαλοὶ μὲν τὰ πεποικιλμένα ζῶια, Κύπριοι δὲ τὰ ἀνθινὰ ἱμάτια, Αἰτωλοὶ δὲ τὰ φάρμακα. Eust. Comm. Il. iv.648–9 compares the similar connection of drugs, plant-extracts and dyes in φάρμακον. 408 Less secure is the evidence for a synchronic connection of γόνυ and γένος, and hence for further symbolism where the knees are the source of generation (Gladigow 1968). Onians (1954: 303–9) suggested that ‘in the lap of the gods’ arose from a metaphor of gods ‘spinning’ destiny.

329

C O M M E N TA RY 3 3 1 328* φαίδιμον. Other cognates focus on Apollo’s prophetic and purificatory aspects, but these are rarely prominent in epic Φοῖβος.409 πόθεν ταύτην μενοεικέα ληΐδ’ ἐλαύνεις: Zeus starts with precisely the question-word which Hermes had warned Apollo about at 269. He is, of course, one of the few who know where Hermes hails from. He sees straight through what has happened and pre-emptively inverts the issue with the metaphor of Apollo leading Hermes as ‘plunder’. In fact, it is Hermes who is the ληϊστῆρ’, ἐλατῆρα βοῶν (14), and who is currently leading Apollo (321; Eitrem 1906: 272). ληΐς is often animate and edible, and μενοεικής seems to have spread from describing the restorative power of food over the body to its restorative power over the spirits, then to other pleasing acquisitions (including booty at Od. 14.232).410 The adjective’s application to a person is exceptional, and points up the irony of describing Hermes as Apollo’s victim in metaphorical terms which involve Hermes’ own hungry crime (64). However, Zeus’s metaphor also shows more serious insight into both gods: Hermes will prove impishly ‘heartwarming’ (389), while the force implied in taking ληΐς (14*) raises a genuine issue with Apollo’s behaviour (246–53*, 373–4). 331 παῖδα νέον γεγαῶτα: παῖδα is both ‘child’ and ‘son’. Zeus describes Hermes’ birth-state as Hermes had done (270–1: newborn παῖδα, so not a thief); this reinforces the echo in 330* πόθεν of 269. Apollo’s own initial insight about Hermes, by contrast, involved Φιλήτην γεγαῶτα (214). Zeus overturns this, while light-heartedly accusing Apollo of being the φιλήτης. φυὴν κήρυκος ἔχοντα: Zeus immediately recognizes one of Hermes’ central future roles: he is the κήρυξ of the gods (3*), was born at Mount Kerykion according to the Tanagrans (Paus. 9.30.2), and was the father of the Eleusinian Kerykes, and patron of human heralds.411 Zeus recognizes this through Hermes’ φυή, i.e. certain innate, mainly physical characteristics.412 Ruipérez (1953) argued for the oxytone accent of φοιβός at B. 13.105, and for a derivation from *bheigw- ‘illuminate’ > ‘cleanse’. 410 Mugler (1963) 143–4. ΣD Il. 23.29 usefully compares θυμ-αρής. For μένος deriving from food and drink see LfgrE iii.132.43. 411 E.g. IG I3 776, A. Ag. 514–15; Paus. 1.38.3. On heralds see LfgrE s.v. κήρυξ, Karavites (1987), with caution; Adcock and Mosley (1975) s.v. ‘herald’. ka-ru-ke had an unclear role at Mycenaean sanctuaries. The similarity to Sanskrit kārú- ‘itinerant singer (esp. of hymns)’ is tantalizing for readers of Herm., but Mondi (1978) and Schmitt (1967: 301–6) disagreed on how the words were cognate, while Beekes (2003) rejected the connection ­altogether. 412 See also 31*. φυή does not focus on the face like εἶδος, nor particularly on beauty. Gods generally adopt a human’s δέμας as disguise, not their φυή (Clay 1974: 129–30; Il. 2.58 is an exception). Outside Epos, φυή can embrace non-physical characteristics, e.g. sexuality (Archilochus fr. 25) and acumen (Pi. O. 2.86). 409

330

C O M M E N TA RY 3 3 2 One means to understand this φυή is through the typical iconography of heralds’ bodies as slim and nimble, mature, and quick to gesture (GoblotCahen 2007: 263–72). But the normal advanced age of heralds (implied in Il. 7.279 παῖδε, 24.79, Od. 19.244) immediately contrasts with Hermes’ youth and shows Zeus’s continuing irony. The ground of similarity lies partly in things which are visible but not φυή, such as Hermes’ apparent function as Apollo’s escort on a lengthy journey, and partly in Zeus’s knowledge of the non-visible, such as Hermes’ aptitude for sacrifice-like butchery (104–41) and carrying of a staff (210).413 Mainly, however, the phrase acts as a propitious acknowledgement that in the future Hermes will become a member of Zeus’s royal retinue – somewhat menial, but still valued for a skill at public speaking that will see him survive the current divine assembly in the role of defendant, and then be promoted to the summoner of future assemblies and announcer of Zeus’s messages.414 Indeed, Zeus’s comment is also propitious because it resembles Hermes’ own comment to the tortoise (see below, 330–2*). 332 σπουδαῖον τόδε χρῆμα: Hermes is already walking σπουδῆι (305). The surface sense is ‘this matter is urgent’: cf. Candaules sharing with Gyges τὰ σπουδαιέστερα τῶν πρηγμάτων (Hdt. 1.8.1), Persians deliberating on τὰ σπουδαιέστατα τῶν πρηγμάτων when drunk (1.133), Isoc. Soph. 50. σπουδαῖος first appears in Theognis, often with χρῆμα or πρῆγμα (e.g. 65–70, 115, 642–4). However, Zeus’s wit lifts the expression from cliché. The presence of a herald in 331 light-heartedly implies that matters should be treated seriously (Lohsee 1872: 29). Zeus juxtaposes the stems |σπουδand |παιδ-, as in the familiar oxymoron which asserts that, as in Herm. as a whole, humour can have an earnest purpose (compare e.g. Ar. Ra. 522–3, Isoc. Euag. 11, Anacharsis at Arist. EN 10.6 1176b33 παίζειν ὅπως σπουδάζηι). χρῆμα suggests both the meaning ‘piece of property’, reverting to the idea that Apollo is the brigand (cf. 400 where the cows are χρήματα, Semonides fr. 6 χρῆμα ... ληΐζεται), and perhaps the colloquial tone of the word’s application to people in comedy and in later prose (Bergson 1967: 85–7, 93–4). θεῶν μεθ’ ὁμήγυριν: ~ 270 μετ’ ἀθανάτοισι, continuing the signs that Hermes’ warning at 269–72 is being fulfilled. For Ianos Laskaris’ treatment of the phrase see Thomas (2016) 297–8. 413

For heralds’ staffs see 530*. For heralds and butchery see Introduction n. 38. The status of the epic herald as subservient to a king was later viewed with some disdain (e.g. Ath. 10 425d, Hsch. κ.2560 κήρυκες: … οἱ τὰς ὑπηρετικὰς ἐπιτελοῦντες πράξεις); cf. the sources for menial Hermes in n. 376. For their rhetorical skills see Idaeus’ diplomacy at Il. 7.354–98, and the names and epithets gathered at LfgrE ii.1411.31–4.

414

331

C O M M E N TA RY 3 3 4 330–2 Zeus’s speech: Zeus’s opening gambit is auspicious for Hermes. It is κερτομία against Apollo – verbal behaviour using pointed phrasing and jokes, for amusement and to defuse Apollo’s outrage publicly (55– 6*; cf. 335). Zeus also confirms Hermes’ warning at 269–72 that Apollo would receive a bemused audience on Olympus, and recognizes Hermes as in essence a herald. Furthermore, various aspects of Zeus’s address to Apollo about Hermes mirror Hermes’ address to the tortoise about its shell (31–3). Both speakers ask where their addressees have got the ‘third party’ from (32, 330 πόθεν), and express their positive feelings towards it, which derives from their similar visionary ability to see its potential (the shell’s for a lyre, Hermes’ for heraldry) in paradoxical terms related to φυή (31, 331). Hermes expanded on that paradox by presenting the tortoise as a courtesan; Zeus gives his surprising metaphorical description of Hermes first (μενοεικέα ληΐδ’). By implication, Zeus’s acceptance of Hermes will be as complete and productive as Hermes’ acceptance of the tortoise. These parallels bring together two strands of the hymn’s characterization of Hermes – that he resembles his father, and that his inventions resemble him (Introduction n. 108).

334–64 334–5 ὦ πάτερ, ἦ τάχα μῦθον ἀκούσεαι οὐκ ἀλαπαδνόν, | κερτομέων ὡς οἶος ἐγὼ φιλολήϊός εἰμι: τάχα in Epos can always have its original sense ‘soon’; A. Pe. 713 ἀκούσηι μῦθον ἐν βραχεῖ χρόνωι is therefore comparable.415 Apollo responds to Zeus’s ironic σπουδαῖον (332) by insisting that he is not wasting anyone’s time. He also deploys in ἀλαπαδνός the rhetorical trope of asserting immediately that one’s case is important (Lausberg 1998: 125–6; so e.g. Lysias 1, 15, 32). The root sense of ἀλαπαδνός seems to be ‘emptied, empty’ (cf. ἀλαπάζω ‘sack’), then ‘weak’. It is not otherwise applied to words, though κενός describes false, insignificant or ineffectual ones (LSJ s.v. I.2), while ‘weak words’ oppose ‘strong’ ones, which are widely believed (as in ὁ λόγος κρατεῖ), argument-winning (e.g. Pi. P. 2.81 ἔπος … κραταιόν), or binding (e.g. formulaic καρτερὸς ὅρκος).416 By using ἀλαπαδνόν, a stretched meta­ phor from the realm of pillaging, Apollo fights fire with fire: though Zeus’s mere κερτομία (335) labelled Apollo and not Hermes as the pillager (οἶος ἐγώ is 415

I would locate the first unambiguous instances of the development ‘perhaps’ < ‘readily’ < ‘soon’ with the indicative in the fourth century (e.g. Pl. Lg. 4 711a τάχα οὐδὲ τεθέασθε τυραννουμένην πόλιν). 416 The adjective ἀλαπαδνός is not α-privative, but is aurally and semantically similar to them. One can therefore compare formulaic οὐκ ἀλαπαδνός+| with double negatives like 243* οὐδ’ ἠγνοίησε, 316 οὐκ ἀδίκως.

332

C O M M E N TA RY 3 3 6 indignantly emphatic), Apollo will treat things seriously and will have no connection with ‘emptying’ even in speech. The hapax φιλολήϊος comes across as an ad hoc invention suiting the unspeakable extravagance of 330. Indeed, the external audience has heard Hermes’ eagerness to become πολυλήϊος (171), which proves Apollo’s point. 336 παῖδά τιν’ ηὗρον τόνδε: Apollo dives into the content of his μῦθος οὐκ ἀλαπαδνός with asyndeton. For τις + ὅδε, compare e.g. Od. 6.206 ὅδε τις δύστηνος … ἱκάνει: the speaker can point to the referent but registers a lack of comprehension. διαπρύσιον κεραϊστήν: Having echoed Zeus’s |παῖδα (cf. 331), Apollo substitutes very different predicates. διαπρύσιος is mostly ‘far-reaching’, of noises. This use is relevant to whether Hermes is more brigand or herald, and in fact διαπρύσιος is applied to a herald’s voice at e.g. D.S. 11.38.6.417 Apollo, in other words, plays off a possible διαπρύσιον κήρ-υκα, to sharpen the contrast between Zeus’s description and his own.418 When κεραϊστής arrives, we must reinterpret διαπρύσιος. At Il. 17.748 πεδίοιο διαπρύσιον τετυχηκώς and (derivatively?) Pindar N. 4.51 the adjective means topographically ‘far-reaching’ or possibly ‘continuous (in space)’; Σvet. Ar. Pax 482c and D.L. 2.143.8 apply it to enmities, probably ‘continual (across time)’. The most applicable meanings here are ‘wide-ranging’ or ‘continual’. Apollo’s imaginative flyting continues in κεραϊστής. The rare word (only otherwise at Or. Sib. 2.261, plus various senses at Hsch. κ.2239) entails particularly violent (normally lethal) looting. The stem therefore minimizes Hermes’ μῆτις, and continues the strategy in ἀλαπαδνόν of overturning Zeus’s use of φιλολήϊος (see esp. Il. 24.245 ἀλαπαζομένην τε πόλιν κεραϊζομένην τε). More specifically, the ancient derivation of κεραΐζω from κέρας aligns his ‘booty’ (in Zeus’s terms) with Hermes’ booty, which is actually horned.419 bm glosses this difficult phrase with ἤγουν φανερὸν κλέπτην. κλέπτης is only a very partial interpretation of κεραϊστής, though the stems occur together at Hdt. 2.121β.1. φανερός ‘patent’ also glosses a διαπρύσιος noise at Σ Ps.-Opp. C. 4.178, Eust. Comm. Il. ii.568.14. But as mentioned, that is not the most plausible sense here.

The κήρυξ in Epos is similarly characterized as λιγύφθογγος. Stentor is taken as Nestor’s herald by ΣbT Il. 2.96–7, and is said to have died by rivalling Hermes in ΣAbT Il. 5.785. 418 Audiences may have heard a further echo of διαπρύσιον κιθαριστήν (cf. H.Aphr. 80 διαπρύσιον κιθαρίζων), whose aptness is not something Apollo appreciates. 419 At Hdt. 8.86, 91 κεραΐζω is used of ships ramming each other at Salamis with their prow-horns. Compare Hsch. κ.2239 κεραϊστής: … ὁ κέρατα ἔχων. 417

333

C O M M E N TA RY 3 3 7 – 3 4 0 337 Κυλλήνης ἐν ὄρεσσι: ὄρεα is a collective, ‘mountainous region’, applied to a single named mountain as at e.g. H.Ap. 34, H.Aphr. 54. The phrase is used in Hes. fr. 170 to describe Hermes’ birth. Since ἐν ὄρεσσι/ὄρει is rare in this position (x2 elsewhere in Epos), the formula may have been tied to Hermes. πολὺν διὰ χῶρον ἀνύσσας: The distance magnifies the shock-value (and conversely impressiveness) of Hermes’ action, and the ground crossed and the tracks left will be important to Apollo’s story (see 342–54). Hes. Op. 635 πολὺν διὰ πόντον ἀνύσσας suggests a formulaic background. 338 κέρτομον οἷον ἐγώ γε θεῶν οὐκ ἄλλον ὄπωπα: Hermes is κέρτομος (see 301–3*) in having lied and provoked Apollo into the present phase of the quarrel. The marked echo of 335 κερτομέων ὡς οἶος ἐγώ expresses the partial similarity of father and son (330–2*), only to overturn it with ‘whose like I have never seen’: Hermes’ κερτομία is genuinely, uniquely irritating, unlike Zeus’s temporary play. 339 οὐδ’ ἀνδρῶν, ὁπόσοι λησίμβροτοί εἰσ’ ἐπὶ γαῖαν: Apollo diminishes Hermes by comparing him to the κερτομία of human thieves. The hapax λησίμβροτοι is semantically and morphologically similar to θελξίμβροτος, and the half-line was probably sonically influenced by Od. 10.191 ἠέλιος φαεσίμβροτος εἶσ’ ὑπὸ γαῖαν (Cantilena 1982: 255), which may have been formulaic.420 For λανθάνειν near κλέπτειν see e.g. Hes. Op. 52–3, Thgn. 20, 1311, Pherecydes FGrH 3 F120 ἐκ δὲ τοῦ Ἑρμοῦ Αὐτόλυκος, ὃς … εἶχε… ταύτην τὴν τέχνην παρὰ τοῦ πατρός, ὥστε τοὺς ἀνθρώπους ὅτε κλέπτοι τι λανθάνειν. Parallels favour M’s γαῖαν over Ψ’s γαίη (i.e. γαίῃ): see Il. 7.446 τίς ἐστι βροτῶν ἐπ’ ἀπείρονα γαῖαν, Hes. Op. 11–12 ἐπὶ γαῖαν | εἰσὶ δύω, LfgrE s.v. γαῖα I 1c, g, GH ii.110–11. By contrast, ἐπὶ χθονί is more common than ἐπὶ χθόνα in such expressions. 340–1 κλέψας δ’ ἐκ λειμῶνος ἐμὰς βοῦς ὤιχετ’ ἐλαύνων | ἑσπέριος: Apollo explains his general character-denunciation of Hermes as a κέρτομος κεραϊστής by encapsulating his charge as the first statement in the narration of events, and with κλέψας placed first within that. Such an early petitio Early translations until Barnes translated ληισίμβροτοι (conjectured by Fick 1896: 271). But Epos always preserves a diaeresis in ληΐς. The majority of the Greek words and names compounded from βροτός (< *mrtó-) retain the original ‘m’, and their modifier ends -σι- (see Knecht 1946). Many were probably built from each other ad hoc. Compare also the Persian name Piršamarda ‘Punisher [< ‘interrogate’ < *prek‘question’] of mortals’ (Hallock 1969: 229).

420

334

C O M M E N TA RY 3 4 1 – 3 4 3 principii can be paralleled in oratory (e.g. Antiphon T1 1.2, Isaeus 1.2) and functions like a prothesis while being more syntactically integrated with what follows: Lausberg (1998) 136. Both ἐκ λειμῶνος and ἑσπέριος recall 197–8. The echo points up Apollo’s very different rhetorical approach here: to the potential witness, he sauntered towards his main point with six lines of background information, and avoided the word κλέπτω altogether; in court, ‘theft’ comes first and the background is dropped. The simplification from ‘fifty of the cows I look after’ to ‘my cows’ is a further sign of Apollo’s rhetorical vehemence: Lohsee (1872) 36. Both ἐλαύνων and 342 ἐλάων overturn Zeus’s ἐλαύνεις in 330*. 341–2 παρὰ θῖνα πολυφλοίσβοιο θαλάσσης, | εὐθὺ Πύλονδ’ ἐλάων: παρὰ θῖνα is true (75–8), and sets up the focus on footprints in the following sentence. The earlier narrative implied that Apollo did not see any footprints until near Pylos (218*). As discussed in 219–26*, one interpretation is that the narrator relocated some poetic material and left an inconsistency. But a more generous one is that Apollo here and in 344–5 is reconstructing Hermes’ route, and relocating his experience from Pylos in order to present a more convincing sequential account of Hermes’ actions. εὐθύ(ς) here and, more clearly, in 355 ἐς Πύλον εὐθὺς ἐλῶντα means ‘directly’ rather than ‘immediately’, a piece of tendentiousness in line with Apollo’s rhetorical approach. The word is only transmitted in Epos in our two lines and in Hes. fr. 43a.63 (‘immediately’), though it is in elegy at e.g. Tyrt. fr. 4.6 and Thgn. 330; the non-Ionic spelling was still avoided by Apollonius and rare in Hellenistic hexameter generally. ἰθύς is so widespread in Epos that scribes were unlikely to introduce εὐθ- in our lines as corrections; hence Agar’s emendations to ἰθ- (1927: 48) are not cogent. The hymnist seems to treat the final sigma as a metrical convenience, though elsewhere semantic differences are observable (Herodotus allows ἰθύς but not ἰθύ to mean ‘immediately’, etc.; see LSJ s.v. εὐθύς). 342 τὰ δ’ ἄρ’ ἴχνια δοιά, πέλωρα: Verbless phrases rarely refer to the past; Guiraud (1962: 318–23) concluded that they confer variety and rapidity to a narrative. This one, however, could contribute an exclamatory quality, replaying Apollo’s original surprise at 218–26 which contained a verbless exclamation in 226, ἴχνια in 220, and πέλωρα in 225 (now extended to both sets of tracks). For the liveliness of ἄρα (also twice in 346–7), see n. 305. 343 οἷά τ’ ἀγάσσασθαι καὶ ἀγαυοῦ δαίμονος ἔργα: ἔργα (like 345* ἀνέφαινε) is coloured by 16* ἀμφανέειν κλυτὰ ἔργα. The sandals (80 ἔργα) and now the track-trick were among these. Both ἀγαυός and ἄγαμαι are built

335

C O M M E N TA RY 3 4 4 – 3 4 5 from ἀγα-, the zero-grade of μεγα-.421 ἄγαμαι is therefore both ‘wonder at’ and ‘consider an enormity’; 342 πέλωρα embraces similar nuances of awe and the preternatural (225*). ἀγαυός is normally positive, but the similar wordplay at Od. 23.63–4 κτεῖνε μνηστῆρας ἀγαυούς, | ὕβριν ἀγασσάμενος suggests that the Odyssey’s formulaic application of ἀγαυός to the suitors (cf. Hes. Th. 632, of the Titans) could be interpreted as ‘excessive’. The ambivalence continues in δαίμονος. Apollo publicly recognizes Hermes as a divinity, and in this sense corrects the various species he mentioned at 219–26 (Cursaru 2012: 40). But he mobilizes the strong connotations in δαίμων of unfathomability and danger; the implications in Epos of the address δαιμόνιε are irritated perplexity at a person’s unnatural actions, i.e. close to ἄγαμαι, though these are not normally transferred to a δαίμων himself.422 For the derivation of δαίμων from ‘distributor’ and its relevance to Hermes’ butchery and heralds see 138*; Apollo’s lack of awareness of that butchery lends to his disparagement an irony which resurfaces in 345, 353, 357, 361. 344 τῆισιν μὲν γὰρ βουσίν: For the dative approaching a genitive see GH ii.71–2. 344–5 ἐς ἀσφοδελὸν λειμῶνα | ἀντία βήματ’ ἔχουσα: ἐς λειμῶνα and ἀντία are separate predicates, the latter perhaps emphasized by its position. I have not found parallels for taking ἀντία ἐς together to mean ‘facing’; at X. An. 4.3.26 στρέψας πρὸς τοὺς Καρδούχους ἀντία τὰ ὅπλα ἔθετο, for example, the πρός probably goes with στρέψας. For the use of ἔχω compare Aeschin. Ctes. 212 τὰ τῶν κονδύλων ... ἴχνη ἔχει ἔτι φανερά, D.S. 3.27.2 ὁ ... τόπος ἴχνη τε ἔχει καὶ σημεῖα πολλά sc. of an elephant’s presence. 345 κόνις ἀνέφαινε μέλαινα: ἀνέφαινε reinforces the echo in 343 ἔργα of the programme from 15–16 ὃς τάχ’ ἔμελλεν | ἀμφανέειν κλυτὰ ἔργα μετ’ ἀθανάτοισι θεοῖσιν. The audience of gods is in place, and the appearance of Hermes’ ἔργα is being related.

421

For assonances in this speech see below, 334–64*. A category-distinction between δαίμων and θεός is mentioned by Hes. Op. 122. In Il. and especially Od., singular δαίμων generally occurs in direct speech to attribute an event to an unknown minister of fate; his work is normally short-term and often unfavourable, though for favourable interventions see e.g. Il. 3.182, Od. 9.381, 18.146, Tsagarakis (1977) 98–105. The Iliad and H.Hom. allow for the identity of a δαίμων being known: in the singular see 551, Il. 3.420, H.Dem. 300, H.Pan 22, Hy. 30.16; for the plural see e.g. 381, Il. 1.222, 6.115, H.Dem. 338, H.Ap. 11. I have found LfgrE s.v., Burkert (1985) 179–81 and Wilamowitz (1931) 362–70 helpful.

422

336

C O M M E N TA RY 3 4 6 κόνις here refers to ‘sand’, as at e.g. Il. 5.586–8, 21.271; see 341 θῖνα, 347 ψαμαθώδεα. μέλαινα hardly points to a volcanic beach. The adjective was traditionally applied to the earth (e.g. Od. 19.111, Hes. Th. 69, Alcm. PMG 89, Sappho frr. 1.10, 16.2), including at Od. 11.364–5 ἠπεροπῆά τ’ ἔμεν καὶ ἐπίκλοπον, οἷά τε πολλούς | βόσκει γαῖα μέλαινα, where it contributes to the connotation of furtive activities in darkness. The whole phrase is designed to rework 140 κόνιν δ’ ἀμάθυνε μέλαιναν|, where κόνις μέλαινα meant ‘ash’.423 Ironically, Hermes was there concealing traces from Apollo rather than ‘revealing’ them. 346 †οὗτος ὅδ’ ἐκτός†: The combination οὗτος ὅδε is unsatisfactory in itself, despite a parallel in IG XII.5 300 (Paros, first century bce), and leaves ἐκτός disconnected. If the corruption is fairly small the most plausible options are οὗτος + noun or, whether retaining οὗτος or not, an adjective parallel to ἀμήχανος (perhaps most likely οὐ ⏑⏑–⏑, or οὗτος ἀ–⏑). On the former route I favour ὁδηγός (Stoll 1861: 9; SEG 61.1432, from Gadara, appears to be a copy of an imperial-era dedication which called Hermes the [καθ]οδηγός of humankind). Ludwich’s ὁδαῖος (1887: 700) is used of Hermes at Phot. ο.28, where there is no cogent reason to emend; ὁδουρός is more distant but has an attractive ambiguity between ‘guide’ and ‘highwayman’. On the latter route οὗτος ἄλεκτος (‘unspeakable’ as in Pherecr. fr. 168) seems best. The corruption may be broader.424 ἀμήχανος: A carefully freighted word for ‘intractable’. (i) ἀμήχανος traditionally connotes the frustration one feels at stubborn people who ignore reasonable prompting. (ii) It has previously referred (157; 257 in Apollo’s mouth) to bonds which Hermes would face: both uses were undercut by coming after Hermes’ self-swaddling, itself a μηχανή (151, 245). By transferring the word to Hermes here Apollo draws out the implication that Hermes can evade others’ μηχαναί or even turn them to his own benefit. (iii) Apollo’s bafflement

So probably in Anacreon PMG 347.5–6. Aristocles deemed κόνῑς the Attic pronunciation, which tragedy supports (see Hdn. Gr. i.526); more often the iotas of κόνις, κονίη are short. The reworking of 140 means that our line is not straightforward evidence for an Atticism. In any case, the ῑ could be back-formed from metrical lengthening in κονίσαλος: Wyatt (1969) 164 n. 34. 424 Hermann’s ὅδ’ ἄϊκτος ‘unapproachable’, Ludwich’s ὁ λεπτός (1890, apparatus), Humbert’s ἄθικτος and Bothe’s ὁ δεκτός ‘the welcome’ (not receptor as Allen 1912 suggested) all give poor sense. Tucker’s ἄδερκτος ‘unseen’ (in Allen 1912) and Allen and Halliday’s ἄδεκτος ‘unacceptable’ suppose unattested senses (ἄδεκτος normally means ‘not admitting of ’, in philosophical contexts). Radermacher invented ὅδακτος. Shackle’s αὖθ’ ὁδοῦ ἐκτός (1915) is far from the paradosis, but preferable to West’s οὔθ’ ὁδοῦ ἐκτός (2003a: 150), which leaves ἀμήχανος stranded. 423

337

C O M M E N TA RY 3 4 6 – 3 4 9 in the face of Hermes’ track-trick (cf. 348 μῆτιν ἔχων) illustrates how another aspect of being ἀμήχανος is limiting others’ μηχαναί pre-emptively.425 But there is some exaggeration here, since the narrator described Hermes finding Apollo to be πολυμήχανος in the face of his own μῆτις (317–19). (iv) Apollo’s word-choice emphatically quashes the other sense of ἀμήχανος (‘having no μηχανή’), which Zeus’s teasing identification of Hermes as the victim at 330–2 might have suggested. 346–7 οὔτ’ ἄρα ποσσίν | οὔτ’ ἄρα χερσὶν ἔβαινε: The latter method may suggest both crawling (as in A. Eu. 37 τρέχω δὲ χερσίν), as is normal for babies, and walking solely on one’s hands. Greek acrobats did the latter, but not long-distance, so the second half of the polar expression would be absurd, as often (Wilamowitz 1895: 231–2). As in 222–4, Apollo’s list of options becomes increasingly bizarre. 348 ἀλλ’ ἄλλην τινὰ μῆτιν ἔχων: For Apollonius’ use of Herm., including this phrase, see Introduction §4.2. 348–9 διέτριβε κέλευθα | τοῖα πέλωρ’: For τρίβω with ‘path’ as outer object see LSJ s.v. II: the parallels tell against interpreting ‘he pressed out a journey that was so monstrous’, i.e. ‘he journeyed along, making such monstrous impressions’. Both the use of διατρίβω, with δια- adding a touch of ‘all along’ to the movement, and the application of πέλωρα to κέλευθα, are bold, in a way which matches Hermes’ action.426 The echo of 342 ἴχνια δοιά, πέλωρα frames the description of the traces. 349 ὡς εἴ τις ἁραιῆισι δρυσὶ βαίνοι: The optative after ὡς εἰ admits different implications: either, as at e.g. Il. 11.467, Apollo presents the action as incomprehensible but a possibility; or he is expressing an absurd fantasy which, ironically, we know to be nearly true. Either way, this climactically odd option completes a tricolon with ποσσὶν/χερσὶν ἔβαινε, after a slight delay, which increases its impact. At 222–4 the climax was hybrid centaurs. The hint of an animal–plant hybrid here would go beyond πελώριος: they are the kind of outré fiction that Lucian had to reach for in VH. For the breathing of ἁραιός see West (1998) xvii; previous editors have printed ἀραιῆισι. It seems an oxymoronic attribute of δρῦς, which is generally

Compare Il. 15.14, where Hera’s trickery makes her ἀμήχανος to Zeus; the net woven with μῆτις renders its victim helpless (e.g. Od. 8.296–9). 426 Agar (1927: 49) suggested transposing 349 after 351, when πέλωρα would describe ἴχνια as at 342. κέλευθα itself cannot mean ‘footsteps’. 425

338

C O M M E N TA RY 3 5 0 – 3 5 3 used of large whole trees (e.g. in Aesop 71, opposed to κάλαμος). The juxtaposition effectively conveys Apollo’s confusion. Rather than implying young saplings, however, ἁραιός might signal a separate species. Theophrastus simi­ larly describes the seaweed Sargassum vulgare C. Agardh as ποντία δρῦς (HP 4.6.9; see DGE s.v. δρῦς). This very species could fit Apollo’s reconstruction of Hermes’ foliage-and-all seaside improvisation, but clearly the identification remains speculative. 350 ἐδίωκε: Either intransitive ‘travelled fast’ (LfgrE s.v. διώκω 1, DGE s.v. A) or transitive ‘impelled’ with object understood. Compare Il. 23.423–4 ἔχε … ἵππους ... ὀλίγον δὲ παρακλίνας ἐδίωκεν. 351 ῥεῖα μάλ’ ἴχνια πάντα διέπρεπεν: Apollo emphasizes the clarity of the signifiers, glossing over the obscurity of what they signified. 352 ψαμάθοιο πολὺν στίβον ἐξεπέρησεν: στίβος retains a connection with στείβω when used of paths: they are either well beaten or (as here) a trail formed by particular footprints.427 Apollo continues 348 διέτριβε κέλευθα, both in the focus on pressing and in the grammatical forcing of the verbs (Hermes is said to ‘traverse’ the στίβος as he creates it). For the thematic importance of ‘traversing’ see Introduction §3.3.6. M’s πολύν reprises the double-edged tone of 337 πολὺν διὰ χῶρον and 343 ἀγα-: the magnitude of the deed is monstrous but also impressive. Ψ’s μέγαν is normally used of wide areas, not narrow lengths. However, Sc. 294 μεγάλων ἀπὸ ὄρχων offers a parallel, and μέγαν might be preferred by the principle utrum in alterum abiturum erat. 353 ἄφραστος γένετ’ ὦκα βοῶν στίβος ἠδὲ καὶ αὐτοῦ: στίβος has the same sense as in 352; the singular never need mean ‘footprints’, and is distinguished from them at e.g. S. Ichn. 115 ἴχνη τε χὠ στίβος (perhaps inspired by our passage: Introduction §1.3), X. An. 1.6.1. The close repetition of a stem is characteristic of Apollo’s speech (see 334–64* below); compare the narrator at e.g. 4~7, 398–9 ἷξον ... ἐξίκοντο, and intensively from around 409 (Introduction §3.2). Apollo unwittingly echoes 79–80, where the narrator described the sudden genesis of the track through ἄφραστα sandals. Unlike in 346 ἀμήχανος*, Apollo’s control of ambiguity falls short: ἄφραστος in 80 meant ‘unconceived’,

στίβος may take an objective genitive at E. Ion 743 περιφερὴς στίβος χθονός, ‘the path caused by the treading of the ground is circuitous’; cf. Diggle (1981) 104–5 on the text.

427

339

C O M M E N TA RY 3 5 4 – 3 5 7 and that is the cause of Apollo’s confusion. Compare the echo of 140 in 345* κόνις ... μέλαινα. 354 χῶρον ἀνὰ κρατερόν: For κρατ- ‘hard’, compare κραταίλεως, κραταίπεδος, esp. E. El. 534–5 πῶς δ’ ἂν γένοιτ’ ἂν ἐν κραταιλέωι πέδωι | γαίας ποδῶν ἔκμακτρον; For Hermes’ route at this point, see 88*. ἐφράσατο: Picking up 353 ἄφραστος to add cohesion: see 334–64* below. 355 ἐς Πύλον εὐθὺς ἐλῶντα βοῶν γένος εὐρυμετώπων: For the rhetoric of restating 340–2 and tweaking the meeting with the Onchestan, see 334–64* below. γένος: Apparently in the unusual sense ‘stock’, as perhaps at Od. 20.211– 12 οὐδέ κεν ἄλλως | ἀνδρί γ’ ὑποσταχύοιτο βοῶν γένος εὐρυμετώπων. ‘Stock’ would also make sense at Herm. 309, but so does the more common ‘species’. For γένος ‘breed’ – the abstract counterpart of ‘stock’ – see e.g. X. Cyn. 3.1. 356 αὐτὰρ ἐπεὶ δὴ τὰς μέν: See 357* for the echo of 105–8. ἐν ἡσυχίηι: Containing the connotation of secrecy in e.g. Th. 8.69.2 τοῖς δ’ ἐν τῆι ξυνωμοσίαι εἴρητο ἡσυχῆι …, and that of security in e.g. Pi. P. 9.22–3 ἡσύχιον | βουσὶν εἰρήναν παρέχοισα. κατέερξεν: One of the successes of p’s scribe. Ω’s κατέρεξεν was an understandable slip after the slaughter in 119–37. 357 καὶ διαπυρπαλάμησεν ὁδοῦ τὸ μὲν ἔνθα τὸ δ’ ἔνθα: For μέν ... καί see Denniston 374. Ilgen’s διαπυρπαλάμησεν is not otherwise found, but in Pi. O. 10.80 the thunderbolt is πυρπάλαμον, probably ‘dealing destruction with fire’ (see Verdenius 1988: 80), and ancient lexicography records πυρπαλάμης and πυρπαλῶμαι. Our epitome of Suet. Blasph. 6 has πυρπαλάμης: ὁ πανοῦργος· ἐξ οὗ καὶ πυρπαλαμᾶσθαι, τὸ κακοτεχνεῖν, with the etymology οἱονεὶ διὰ πυρὸς ἰέναι πανουργίας ἕνεκα. Eustathius, using a fuller epitome, proposed that πυρ- could imply bravery (Comm. Il. ii.5; Taillardat 1967: 11–22). Paus. Att. π.46 πυρπαλάμης: ὁ ταχέως τι ἐπινοῶν καὶ παλαμώμενος ἴσα πυρί seems to take πυρ- as a symbol of speed. Stolz (1903: 251–3) compared Il. 20.371 πυρὶ χεῖρας ἔοικεν, and reconstructed the development ‘with a hand (παλάμη) like fire’ > ‘dealing destruction like fire’ and ‘as adroit as fire’ (as παλάμη ‘hand’ > ‘violence’, ‘plot’) > ‘scheming quickly’ > ‘criminal’. The context here foregrounds the subtle sides of fire and παλάμη over the violent. Ironically, Apollo chooses this remarkable word for the very episode when Hermes was secretively ‘grasping’ at fire using his palm (108 πυρὸς δ’ ἐπεμαίετο τέχνην + 110 παλάμηι). The similarity of 356 to 105–8 (ἐπεὶ … τὰς μέν + the action of

340

C O M M E N TA RY 3 5 8 – 3 5 9 shutting up the cows) supports the intratext.428 What Apollo hoped to draw attention to with his hapax was quite different – the strangeness of Hermes’ action, as with the similarly unusual phrase διέτριβε κέλευθα in 348*. Regarding the grammatical connection of διαπυρπαλάμησεν to the rest of the line, I suspect the verb was taken as analogous to διαπρήσσω ὁδοῦ (genitive denoting space within which the action occurs). Adverbial τὸ μὲν ἔνθα, τὸ δ’ ἔνθα then mean ‘there and back’ and refer to the whole round trip to Kyllene; cf. Od. 2.213 οἵ κέ μοι ἔνθα καὶ ἔνθα διαπρήσσωσι κέλευθον.429 If on the other hand the verb was perceived to be transitive, Hermes made a scheme out of both sides of the road, in reference to the two mystifying sets of footprints, which Apollo described using similar phrasing at 226 μὲν ἔνθεν ὁδοῖο … δ’ ... ἔνθεν ὁδοῖο. The second set of tracks did not accompany Hermes beyond Pylos, but such precision is not Apollo’s concern. 358–9 κατέκειτο μελαίνηι νυκτὶ ἐοικώς | ἄντρωι ἐν ἠερόεντι κατὰ ζόφον: The formulaic comparison νυκτὶ ἐοικώς (Il. 1.47, Od. 11.606; similarly Il. 12.463) otherwise describes a fast, threatening motion; at Il. 1.47, ΣD takes the simile to characterize Apollo as φοβερὸς καὶ ἀόρατος, where the latter adjective rests on the dominant early Greek view that night is a substance which makes things hard to see and which is, by a slight slippage, itself hard to see: Dyer (1974). Gemoll saw a parody of that passage here, but although Apollo is overdramatizing his own fearfulness, both aspects of the scholiast’s analysis suit how he wishes to portray Hermes’ movements as a threat. He recently characterized Hermes as μελαίνης νυκτὸς ἑταῖρε (290: the first halves of 290 and 358 are similar too); he goes on to focalize the cave using terms he applied in 256*–7 to Tartarus (ἠερόεντι, ζόφον). Hermes is a friend of night, at night, moving like night, in a dark place resembling the prison of the worst criminals. For Hermes and night see further 334–64* below, Introduction §3.3.2. 359–60 οὐδέ κεν αὐτόν | αἰετὸς ὀξὺ λάων ἐσκέψατο: The etymology and development of λάω have been disputed, but the only plausible sense of ὀξὺ λάων here is ‘looking sharply’.430 For eagles’ proverbially sharp eyes see e.g. Il. For Hermes, fire and μῆτις, see Introduction §3.3.6. For irony undermining Apollo’s traditional omniscience see below, 334–64*. 429 See KG i.595 for this use of the article. 430 Prier (1980) argued that λάω etymologically meant ‘see’, and that there is no need to posit a coalescence of homonyms to explain the word’s uses. The equation λάω = βλέπω was frequently and imaginatively used by Greek etymologists: see Hsch. λ.80, and e.g. Manetho 609 FGrH F13 on λέων; Crates in ΣV Od. 19.229 on ἀλαός; Orion 428

341

C O M M E N TA RY 3 6 0 – 3 6 1 17.674–5 αἰετός, ὅν ῥά τέ φασιν | ὀξύτατον δέρκεσθαι ὑπουρανίων πετεηνῶν, Luc. Icarom. 14, Ael. NA 1.42. Apollo underlines the implication in 358 νυκτὶ ἐοικώς that Hermes’ movements were impossible to see, and does so in a way calculated to appeal to Zeus. McCail (1970) proposed ὀὺ λάοντα in Agathias AP 5.237.5, and took it as an allusion to Herm. 360–1 πολλὰ δὲ χερσὶν | αὐγὰς ὠμόργαζε: Apollo’s narrative skips over his own arrival and maintains a focus on Hermes’ activities; see 334–64* below. ὠμόργαζε is Ilgen’s correction of Ω’s minuscule error ὠμάρταζε, which is not a word, but close enough to ὁμαρτέω to have been copied (Lp in fact have ὡμάρταζε). ὀμοργάζω, not attested elsewhere, would derive from ὀμόργνυμι like μιγάζομαι from μ(ε)ίγνυμι. Such formations in -άζω may be intensive (Debrunner 1917: 125–7), and this quality is enhanced by the assonance –γας … -γαζ-. For the gesture, compare Od. 18.199–200 τὴν δὲ γλυκὺς ὕπνος ἀνῆκεν, | καί ῥ’ ἀπομόρξατο χερσὶ παρειάς (ΣH: ὃ συνήθως οἱ διυπνισθέντες ποιοῦσιν); Longus 1.26.2 τοὺς δὲ ὀφθαλμοὺς ἀπέματτεν ἔτι καθεύδειν θέλοντας, Philostr. Vit. Ap. 4.20.32–3 τὸ δὲ μειράκιον, ὥσπερ ἀφυπνίσαν, τούς τε ὀφθαλμοὺς ἔτριψε ... Apollo presents Hermes as having pretended to wake up and rub his eyes sleepily.431 This complements the narrator’s image of Hermes pretending to be asleep (241–2). However, the narrator did not mention anything of this sort at 260 or 278–80 and, considering the rest of Apollo’s rhetoric, the natural interpretation is that he is inventing. 361 δολοφροσύνην ἀλεγύνων: The verb’s formulaic usage with meals as the object (Od. x5, Empedocles 31 B137.8) is relevant here, since meal-preparation loomed large in the part of Hermes’ theft where Apollo has not seen through his δόλος.432 For the abstract object compare 476 ἀγλαΐας, A.R. 3.1105 συνημοσύνας. See also 362 ἀπηλεγέως*.

pp. 91, 105 on λαγώς, μέλαν; Choerob. Epim. p. 164 on λήμη; Eust. Comm. Il. ii.339 on ἱλάω. At Od. 19.228–30 κύων ἔχε ποικίλον ἔλλον | ἀσπαίροντα λάων· … ὃ μὲν λάε νέβρον ἀπάγχων, the verb may again refer to predatory looks, even if ancient scholars claimed other meanings such as ‘devour’ (Aristarchus in ΣV, Hsch. λ.472) or ‘bark’ (Hsch. λ.78); Rutherford (1992) demanded ‘grip’. Leumann (1950: 233–6) suggested the phrase ὀξὺ λάων had originally meant ‘crying shrilly’ (cf. λάσκω), but accepted that in Herm. λάων means ‘gazing’. 431 Also perhaps E. Or. 219–20. αὐγή ‘eye’ occurs only here in Epos, where the sense is normally ‘light’ of sun or fire. But αὐγάζομαι ‘I spot’ occurs already at Il. 23.458, Hes. Op. 478. 432 -ύνω normally creates a verb from a noun or adjective. Chantraine (DÉLG s.v.) suggested that its irregular function in ἀλεγύνω arose through analogy with ἐντύνω,

342

C O M M E N TA RY 3 6 2 – 3 6 3 ἀλεγύνων is preserved only in Θ. M’s ἀλεγίζων is the more common derivative of ἀλέγω, but ‘taking notice of ’ is inappropriate, and ἀλεγίζω almost always follows οὐκ. p’s ἀλεείνων is an error of Byzantine pronunciation: [aleˈʝi-] ~ [aleˈi-]. ἀλεγύνω caused scribes problems again at 557. 362 αὐτός: ‘in person’ – a usage at the intersection of LfgrE s.v. I 1b, 2a, 3aγ.433 Apollo presupposes and overturns any assumption that the infant (yet to speak on Olympus) might need a spokesman, and so emphasizes his full responsibility for the brazen lie that follows. For sources in which Hermes stayed quiet for longer, see LIMC s.v. Hermes 241, S. Ichn. (where his role is delayed by the appearance of Kyllene), Philostr. Im. 1.26. ἀπηλεγέως: ἀπο- ‘not’ (LSJ s.v. D.6) + ἀλέγω. The other two uses in Epos (Il. 9.309, Od. 1.373) describe speech which ‘does not attend to’ the addressee by being uncompromisingly direct – in the language of politeness theory, by neglecting to mitigate a threat to negative face.434 This usage continues in Apollonius (e.g. 1.439, 2.845, 3.439) and is apt here, though Apollonius also uses the word more generally of other kinds of casual misdemeanour (e.g. 2.17 εἰ δ’ αὖ ἀπηλεγέοντες ἐμὰς πατέοιτε θέμιστας). There is rhetorical wordplay in the contrast with 361 ἀλεγύνων (annominatio: Lausberg 1998: 292): Hermes attends to trickery but, in doing so, not to propriety. 363–4 ‘οὐκ ἴδον, οὐ πυθόμην, οὐκ ἄλλου μῦθον ἄκουσα, | οὐδέ κε μηνύσαιμ’, οὐδ’ ἂν μήνυτρον ἀροίμην.’: Apollo cites 263–4*, though he does not completely replicate the childishness of the repetitive syntax there, since he replaces οὐκ ἂν … οὐκ ἄν with οὐδέ κε … οὐδ’ ἄν. For the tone see the summative note below.

ἀρτύνω, bringing out the nuance ‘preparing’, which is not strong in ἀλέγω itself. I assume that DGE rejected this nuance when the object is a banquet since the subject sometimes appears to be a guest. How­ever, at Od. 8.38 and 13.23 the Phaeacian guests do help in meat-preparation; at 11.185–6 Τηλέμαχος … δαῖτας ἐΐσας | δαίνυται, ἃς ἐπέοικε δικασπόλον ἄνδρ’ ἀλεγύνειν, one can take the δικασπόλον ἄνδρα to be an Ithacan elder inviting Telemachus, not Telemachus himself. All passages down to A.R. 4.1203 are compatible with the basic sense ‘pay heed to the preparation of ’. 433 αὐτὸς δ’ could in principle contrast Hermes ‘as a whole’ with his αὐγάς, but by such language of ‘whole’ individuals the Greeks seem to focus on the body: see e.g. Od. 9.257 φθόγγον τε βαρὺν αὐτόν τε πέλωρον, LfgrE s.v. αὐτός III 1aγ, 3. 434 For the formation compare ἀνηλεγέως, the tendency for οὐκ to precede ἀλέγω and ἀλεγίζω, Latin nec-lego, and Albanian p-log-ët ‘lazy’ (where p- ~ ἀπό). Secondarily, the word spreads to ‘directness’ which is not related to politeness: e.g. A.R. 4.902, Nic. Th. 495. Our line makes that point with αὐτίκα.

343

C O M M E N TA RY 3 3 4 – 3 6 4 334–64 Apollo’s forensic ambition: A detailed exposition of Apollo’s rhetoric will help us understand both his characterization and Herm.’s relationship to rhetoric (which intertwines with problems such as its date: Introduction §1.2). Vergados and Schenck zu Schweinsberg adopt what has become a standard emphasis on the artlessness of this speech. But at Herm.’s date a handbook treatment was not the primary way to generate rhetorical force, and anyway it is not particularly hard to map the structure on major categories from the handbooks: after a proem (334–9), Apollo presents a narrative (διήγησις) interspersed with elements of argument (πίστεις), as Demosthenes often did. The proem’s two sentences are tied together as claim and explanation, and also by the echoes linking 338 to 335. Apollo skilfully negotiates the transition from Zeus’s speech, which might well have thrown him. He engages with Zeus’s words both in rejecting his ironic σπουδαῖον (332) with the proemic trope ‘This is a serious matter’, and in returning his humour (wordplay in ἀλαπαδνόν, διαπρύσιον; imaginative and significant use of the unusual words φιλολήϊος, κεραϊστής). Apollo’s initial focus is firmly on Hermes’ ἦθος, as was typical of proems (Lausberg 1998: 129–31). While cementing his relationship to Zeus by starting ὦ πάτερ, he refuses throughout to acknowledge Hermes’ paternity, here referring merely to παῖδά τιν’ (336). Apollo then gives a striking explanation of his characterization with the opening of the narrative (340), which encapsulates and tendentiously simplifies the kernel of his allegation. The first sentence recalls 191–8, particularly 197–8, and so draws attention to Apollo’s choice of a completely different approach, from meandering around the idea of theft to placing it front and centre. The subsequent description of the tracks (342–9) recalls 218–26 in multiple ways.435 Again the contrast sets in perspective the rhetorical conviction with which Apollo ascribes to Hermes’ μῆτις prints which had previously left him unable to reconcile what he saw with even the species of the culprit. Apollo’s excursus describing the tracks is framed with a verbal echo (342 δοιά, πέλωρα ~ 349 τοῖα πέλωρ’), before he passes from unignorable signs (351 ῥεῖα μάλ’ ἴχνια πάντα διέπρεπεν) to testimony that is allegedly even clearer. Verbal patterning again marks off the transitional contrast (esp. 350 διὰ ψαμαθώδεα χῶρον ~ 354 χῶρον ἀνὰ κρατερόν), but also unites it to what has immediately preceded (the epiphora of διὰ ψαμαθώδεα χῶρον in 347 and 350; 351 Besides topic, compare the wording (ἴχνια, and the rarer usage βήματα; 342 δοιά, πέλωρα/349 τοῖα πέλωρα ~ 225 τοῖα πέλωρα, 344–5 ἐς ἀσφοδελὸν λειμῶνα| ἀντία ~ 221 πάλιν τέτραπται ἐς ἀσφοδελὸν λειμῶνα|); sentence-structure (μέν of cow-prints … δέ of Hermes’; 345–9 and 222–4 both crescendo in oddity); and the shared concept of marvel (343 ἀγάσσασθαι ~ 219 θαῦμα). Cf. Romani (2008) 95–6.

435

344

C O M M E N TA RY 3 3 4 – 3 6 4 διέπρεπεν ἐν κονίηισιν ~ 345 κόνις ἀνέφαινε). The section about the tracks thus reaches closure as that source of evidence disappears (353 ἄ-φρασ-τος) to be replaced by the human witness (354 ἐφράσατο). The apparent continuity of evidence is a very tendentious report of the aporia we witnessed at 185–212. At this juncture-point, Apollo restates the essentials of his denunciation: 355 echoes 340 ἐμὰς βοῦς … ἐλαύνων and 342 εὐθὺ Πύλονδ’ ἐλάων. Apollo makes it possible for his audience to believe that the man reported Hermes’ passing confidently, and knew and said where Hermes was heading (all of which are false); Apollo omits the Onchestan’s location, which would cast doubt on his knowledge of Hermes’ destination. Much as he makes of his human witness, Apollo cannot bring him to Olympus to testify (see Introduction §1.2 for unavailable witnesses in early rhetoric), and returns swiftly to his narrative of Hermes’ activities, skirting round the limited progress made by his augury (213–15) and zig-zag to Pylos then Kyllene; here, he will not repeat the self-deprecating acknowledgement in 303* that his prophetic skill proved insufficient. The rhetoric intensifies from 356 to the end. 357 introduces an imaginative and possibly ad hoc word (διαπυρπαλάμησεν). 358 brings a simile (νυκτὶ ἐοικώς): in the Iliad and Odyssey similes are comparatively rare in direct speeches, but one traditional context for them is in rebukes, flyting and contemptuous speech – verbal power-plays; this type of simile recurs in Demosthenes’ invective.436 Apollo’s tendentiousness about the cave’s Tartarean murk is set into relief by the narrator’s presentation (145–53, 237–52), in which the cave was a rich ‘temple’; in terms of similes, Hermes entered it like a mist, and looked like an innocent baby when Apollo arrived.437 The reference to the eagle drives home this tendentiousness, while elevating the significance of Apollo’s search for Hermes. The eagle hunts Hermes as if he were a hare, a snake, or better a tortoise in an unprepossessing hole; Zeus’s avian royal counterpart and messenger hunts for the pretender to be Zeus’s messenger, lurking in a dangerously Tartarean cave (358–9) like the greatest threats to Zeus himself.438 Next, the external audience can judge that 361 is invented, and a

436

E.g. the assembly are like puppet-makers (4.26); Aeschines is like an insulter in the Dionysia procession (Cor. 122), spasms (198), and a wind (308). As a wind, he is both windbag and one to cause a storm – the latter being the most prominent image of the oration. See also Cor. 242. 437 Compare the contrasting similes for Patroclus offered by the narrator and Achilles at Il. 16.3–10. 438 Eagles and Zeus: RE i.373–5; Thompson (1936) 3–4, also mentioning its epi­thet διάκτορε. Chapman (1624) brought out the implication by translating ‘your owne Eagle’. For eagles and tortoises see e.g. Aeschylus TrGF test. M, Babr. 2.115.

345

C O M M E N TA RY 3 3 4 – 3 6 4 cluster of assonances builds for the finale (361 -γας -γαζ, 361–2 ἀλεγύνων … ἀπηλεγέως, 362 αὐτ- αὐτ- … ἀ- ἀ-). To end, Apollo decides to cite 263–4 directly, and not to continue with further narration of 265–312 or a separate epilogue. This neglects the potential material in Hermes’ oath-offer (274–6) and gestures (278–80, 295–7), but Apollo’s decision can still be interpreted in terms of his rhetorical goals. Hermes’ denial appears as blunt as possible; by citing him rather than para­ phrasing him, Apollo draws attention to the form of the denial as a piece of δολοφροσύνη. On the page, this comes down to the childish syntax, which simulates innocence; in performance, Apollo was perhaps made to put on a wheedling tone for these lines (and a desire for mimetic delivery could also explain Apollo’s invention of the eye-rubbing at 361); see Thomas (2010) 211–12 for this feature of childish intonation. The citation fulfils a major function of the epilogue according to rhetorical theory, namely to sway the audience’s sympathies by appealing to the character of one or both sides in the case (Lausberg 1998: 204–8): Hermes is brazen, Apollo a likably entertaining speaker. Furthermore, Apollo makes Hermes listen to his own words quoted for ridicule in a new context, immediately before he must start his speech.439 The narrative has many rhetorical figures (342–3 assonance and figura etymologica; 346–9 tricolon crescendo with dramatic suspense; 361–2, discussed above). Other phrases show bold or imaginative language (e.g. 348–9*, 357 διαπυρπαλάμησεν, 361 ὠμοργάζε, ἀλεγύνων*; Schulze 1868: 34). This contributes to enargeia, as do touches like the nominal phrase with ἄρ’ in 342*, and the simile in 358. The whole is clearly articulated, by verbal echoes linking one thought with the last (e.g. 352–3 στίβον … στίβος, 353–4 ἄφραστος … ἐφράσατο, 361–2 ἀλεγύνων … ἀπηλεγέως), and in 342–54 also by the more complicated parallelisms summarized above. Fundamental to the διήγησις as a whole is Apollo’s decision to focus on Hermes’ actions, and to pass over events before the theft and his own actions, but nevertheless to present his interpretation of phenomena with straightforward chronology and absolute confidence, to the point of almost omitting any argumentation. This rhetorical façade means that Apollo does not show the μῆτις to foreclose Hermes’ predictable response of arguing from εἰκός (for ‘foreclosing’ μῆτις, see 346 ἀμήχανος*). His general strategy risks making the narrative disjointed, particularly around 360, where he does not mention his 439

Mimetic delivery (ὑπόκρισις) was supposedly avoided by Pericles (Plu. P ­ eric. 5) and disavowed by Quintilian (Inst. 11.3.88–9). Our knowledge about Classical delivery, other than the importance attached to it, is very scant: Steinbrink (1992) 43–52. For the rhetoric behind characters’ citations of one another in Il., see De Jong (1987) 171–9.

346

C O M M E N TA RY 3 6 5 – 3 6 7 arrival at the cave. The speech is the poet’s product too, and Apollo’s confidence is undercut particularly by his ignorance about two cows having been killed (356–7*). While Apollo presents the phenomena as extraordinary to imply how Hermes is outrageous and dangerous, the poet’s broader narrative concern is to make even Apollo’s character-assassination a paradoxical ve­­­ hicle for Hermes to gain κῦδος among the other Olympians – his ἔργα (343*) are impressive and he is ἀγαυός (see 337, 352 πολύν*).

365–7 365 ἤτοι ὅ γ’ ὣς εἰπὼν κατ’ ἄρ’ ἕζετο Φοῖβος Ἀπόλλων: Barnes’s correction ἤτοι ὅ γ’ for Ω’s ἤτοι ἄρ’ brings the text towards a formula (ἤτοι ὅ γ’ ὣς εἰπὼν κατ’ ἄρ’ ἕζετο, τοῖσι δ’ ἀνέστη: Il. x5, Od. x1), which was unlikely to have been varied so as to introduce doubled ἄρα within the same clause. The only parallel in Epos for that feature would be Od. 16.213 ὣς ἄρα φωνήσας κατ’ ἄρ’ ἕζετο, but that arises from juxtaposition of two formulas.440 The corruption may have been a minuscule slip, or an Antizipationsfehler. For ὅ γε … , which betrays the movement of ὁ from demonstrative to article, compare 424, Il. 13.53–4, H.Aphr. 128–9, H.Ap. 316–17, 347–8. 366 Ἑρμῆς δ’ αὖθ’ ἑτέρωθεν ἀμειβόμενος ἔπος ηὔδα/Ἑρμῆς δ’ ἄλλον μῦθον ἐν ἀθανάτοισιν ἔειπεν: Both variants may well be ancient. 366(a) seems superior: αὖθ’ ἑτέρωθεν is more precise than 366(b) ἄλλον, and ἄλλον μῦθον is potentially stilted after 363 ἄλλου μῦθον (Hollander 1891: 637). 367 δείξατο δ’ ἐς Κρονίωνα: Hermes probably gestures towards Zeus as a sign of greeting and/or of starting to speak to him; the gesture seems not to have been pointing (LSJ, Vergados) but holding one’s arm out to the addressee with index and middle fingers extended, ring and little fingers tucked in: Boegehold (1999) 25–6. Related usage involving δείκνυμι ἐς includes the intransitive active for indicating something while speaking (e.g. Hdt. 4.150.3); a similar use of ἐνδεικνύμενος εἰς in Ps.-Plato Amatores 133b, and the transitive active ‘direct words towards (with a gesture)’ in Philostratus VS 541–2 ἀγωνιστοῦ ... τὸ ‘Ὦ Ζεῦ’ ἐς τὴν γῆν δείξαντος, τὸ δὲ ‘Καὶ γᾶ’ ἐς τὸν οὐρανὸν ἀνασχόντος, … ὁ Πολέμων ἐξέωσεν αὐτὸν τῶν ἄθλων εἰπὼν ‘οὗτος τῆι χειρὶ ἐσολοίκιζεν’. Such gestures are at the root of the connection of δείκνυμι with δεικανάω and δειδίσκομαι (‘salute’, including with hands, e.g. A.R. 1.884; see Forssman 1978).

For κατ’ ἄρ’ ἕζετο see Hsch. κ.1668 κατ’ ἔρ’ ἔαι: καθίσαι· Πάφιοι (similarly κ.1670–1), Ruijgh (1971) 433 n. 76, LfgrE s.v. ἕζεσθαι II 3aγ.

440

347

C O M M E N TA RY 3 6 8 – 3 6 9 θεῶν σημάντορα πάντων: In the Hesiodic corpus accusative θεῶν πατέρ’ ἠδὲ καὶ ἀνδρῶν (Hes. Th. x3) alternates with dative θεῶν σημάντορι πάντων (fr. 5.3, Sc. 56). Possibly the hymnist was modifying a traditional system because in this context ἄνδρες are less relevant.

368–86 368 Ζεῦ πάτερ: Hermes starts by staking an important claim to his lineage. ἤτοι ἐγώ σοι ἀληθείην ἀγορεύσω: ἀληθείη is already used in Od. both of not forgetting (ἀ-ληθ-) details in a narrative, and of not introducing fabrications: Cole (1983) 10–12. Hermes simultaneously makes an orator’s standard claim to conscientiousness and sets up his focus on filling in details which Apollo neglected.441 ἐγώ could for the moment be taken as asseverative (n. 344), but does turn out to reinforce this opposition of Hermes’ ἀλήθεια with Apollo’s account (369*). M’s ἀγορεύσω (parallels in 561, Dorotheus p. 405.18) would specifically rebut Apollo’s characterization of Hermes’ mode of speaking in 362 ἀπηλεγέως ἀγόρευεν. Ψ’s καταλέξω gives a formula (Il. 24.407, Od. x6), and so is the lectio facilior. Accented σοί would inappropriately imply that Hermes had lied to Apollo. For enclitic σοι rather than τοι see Od. 3.359, 11.381, H.Aphr. 275; the ancient distinction of σοί vs τοι (La Roche 1866: 349–51) was too strict. Here the poet may be avoiding τοι … τοι. 369 νημερτής τε γάρ εἰμι: Hermes cheekily appropriates an adjective which, as the narrator showed at 315*, really applies to Apollo (cf. 495* κερδαλέον). This strategy confirms that 368 ἐγώ was setting up a contrast between Hermes’ account and Apollo’s less than true one, and tackles Apollo’s presentation of Hermes’ ἦθος in his proem head on. As an opening gambit this corresponds to Zeus casting Apollo as the brigand (330), and hence shows appreciation of Zeus’s humour, inspires empathy, and suggests that he is a true son of his father, as he has just claimed (368). καὶ οὐκ οἶδα ψεύδεσθαι: Görgemanns (1976: 117) compared the proemic move of Socrates at Pl. Ap. 17b, among others: ἐπειδὰν μηδ’ ὁπωστιοῦν φαίνωμαι δεινὸς λέγειν ... For lying well as a matter of know-how, compare e.g. Hes. Th. 27 ἴδμεν ψεύδεα πολλὰ λέγειν ἐτύμοισιν ὁμοῖα. Despite the proverb οἶνος καὶ παῖδες ἀληθεῖς (Paus. Att. ο.10), Zeus will perceive that the baby Hermes has already learnt to lie expertly (390*).

441

Vergados gives useful parallels for liars in particular starting with a claim to truthfulness.

348

C O M M E N TA RY 3 7 0 – 3 7 2 370 ἦλθεν ἐς ἡμετέρου διζήμενος εἰλίποδας βοῦς: The beginning of Hermes’ narrative matches in form, and so answers effectively, the beginning of Apollo’s (340–1 κλέψας ἐκ λειμῶνος ἐμὰς βοῦς ὤιχετ’ ἐλαύνων | ἑσπέριος): ‘came to’ vs ‘went from’; participle with cows as object; and in 371 morning vs evening. Both gods indicate their opponent merely with a verb-ending, but whereas at 340–1 this flowed from the preceding words, here there is a jarring avoidance of Apollo’s name. Gemoll found this way of plunging into the account naive; Allen and Sikes found it aggressively discourteous. For the construction ἐς ἡμετέρου compare Od. 2.55, 7.301 (+ ΣHP), 17.534, Archil. fr. 196a.4 (ἐν), Hdt. 1.35.4, 7.8δ.1, Ar. Lys. 1063 εἰς ἐμοῦ. Hermes’ use of the plural may suggest Maia’s presence and vulnerability, or it could characterize his speech-style: see 267* ἡμετέρης. 371 ἠελίοιο νέον ἐπιτελλομένοιο: Glotz (1904: 207) plausibly suggested that this is not descriptive detail, but entails a charge that Apollo’s search fell outside acceptable visiting-hours; compare 246–7*, 372* for other restrictions on house-searches. The slight metrical anomaly (νέονˉ) is caused by modification of νέον καταδυομένοιο (197). Similarly ἅμ’ ἠελίωι ἀνιόντι at Il. 18.136, Od. 23.362 (with hiatus) < ἅμα δ’ ἠελίωι καταδύντι. 372 οὐδὲ θεῶν μακάρων ἄγε μάρτυρας οὐδὲ κατόπτας: Hermes’ point is not that Apollo cannot cite neutral corroboration for the results of his search: he has not pretended there were any. Rather, he ignored proper procedure in a way which gave him the liberty to throw his weight around. For attendants at a house-search, see the three witnesses at Hipponax fr. 79.17 (probably – the gist is unclear), the ‘followers’ in Isaeus 6.41–2, and the guards who are to watch over any sealed containers in Pl. Lg. 12 954b; Macrobius Sat. 1.6.30 reports an amusing anecdote set in a Roman context (c.150 bce) of a house-search involving custodes. The parties in attendance would be able to give evidence in subsequent litigation if the stolen goods were found, or if anything underhand occurred: see Isaeus 3.19 ὅταν μὲν ἐπὶ προδήλους πράξεις ἴωμεν ἃς δεῖ μετὰ μαρτύρων γενέσθαι, τοὺς οἰκειοτάτους καὶ οἷς ἂν τυγχάνωμεν χρώμενοι μάλιστα, τούτους παραλαμβάνειν εἰώθαμεν. The normal sense of κατόπται is ‘spies’, so they may be witnesses of the theft brought along to swear that the house-search had a realistic chance of finding something; or they may be official ‘observers’ of the search, to check that it is conducted properly.442 The

Much later, a night-policeman’s watch is described using κατοπτ- in POxy. 1033.13 (392 ce). The financial κατόπται of Boeotia are less pertinent, pace Feyel (1946). For

442

349

C O M M E N TA RY 3 7 3 – 3 7 5 witnesses and observers should have been not only present but divine – an implicit dig at Apollo’s merely human witness: Abramowicz (1937) 77. 373 ἀναγκαίης ὕπο πολλῆς: ἀνάγκη can be an unalterable objective ‘necessity’ or, as here, ‘compulsion, duress’ imposed by another individual. For quantification of the latter, compare πολλὴ ἀνάγκη in S. Tr. 295, E. Hec. 396. Apollo’s procedural irregularities spring from an intention to intimidate. The presentation of force majeure at D. 21.78–80 is somewhat similar, though the legal situation is different: Meidias and Thrasylochus burst into the home of a younger man, who is sitting indoors with female family-members, and speak threateningly (see MacDowell 1990; 295–9). On the accent of ὕπο with noun preceding and adjective following, see Hdn. Gr. i.482–3. 374 πολλὰ δέ μ’ ἠπείλησε: Hermes creates a chiasmus in 373–4: μηνύειν ἐκέλευεν ... πολλῆς, | πολλὰ δέ ... ἠπείλησε. Like πολλῆς, πολλά here expresses degree of intensity more than frequency; this is often the case when πολλά qualifies verbs of begging, praying and expressing emotions (see LfgrE s.v. πολύς I.3; Snell 1953: 18). It helps explain how an aorist can follow, as in Il. 1.351, H Dem. 439. βαλεῖν ἐς Τάρταρον: Another detail designed to stir Zeus’s indignation: this threat of tartarization appeared to be an arrogation of Zeus’s powers (256*). 375 οὕνεχ’ ὃ μὲν τέρεν ἄνθος ἔχει φιλοκυδέος ἥβης: The basic sense is clear: Apollo abused his advantage in age and strength. But the connotations of each word, and their synergy, reward careful analysis. Apollo’s connection to ἥβη (on which see 55–6*) focuses on ephebes. Classical representations of Apollo fix his physical form at the stage of a beardless youth; he undergoes various typical rites of passage in myth (as cowherd to Admetus and Laomedon, and killer of the Python), he had patronage of adolescents in various cults, and so on.443 This age is multiply suited to the traditional ‘ἄνθος of youth’ metaphor. The main connotations of ἄνθος besides ‘flower’ are ‘acme’ and ‘surface growth with colour-change’: youth (for humans) is ephemeral like a flower, was considered to mark an acme of energy and beauty, and in boys was accompanied by sprouting of the role of witnesses in general see Papakonstantinou (2007) 96–100 with further references. Radermacher’s discussion of our line remains useful. 443 For art see LIMC II(1).316–18; compare e.g. H.Ap. 449–50 ἀνέρι εἰδόμενος αἰζηῶι τε κρατερῶι τε | πρωθήβηι, Call. H Ap. 36–7. In the Septeria which ‘re-enacted’ his killing of the Python he was played by a κόρος (Plu. Def. Or. 411a). Graf (2009: 103–29) offers a convenient survey.

350

C O M M E N TA RY 3 7 5 facial hair and a darkening of the skin as they spent more time outdoors.444 The metaphor is part of a broader pattern of imagery relating the sap in plants to human vital fluids.445 This analogy helps to elucidate how τέρεν connects its principal referents, namely: (i) flowers and leaves plucked in their prime; (ii) the ‘flower’ of youth, male or female; (iii) youths and young adults, normally female, with particular reference to their skin; (iv) tender meat; (v) liquids, notably tears, but also blood and water in Emped. 31 B100. The semantic hub is ‘full of vital fluid’. τέρην skin is often moist, with water or unguent. In the Odyssey, soft and well-hydrated skin ‘melts’ to release tears (8.522, 19.204); the vital fluid from which tears derive is specifically αἰών at 5.151–60, while αἰονάω ‘wash, anoint’ (A. fr. 425, Hp.) suggests that unguents were seen as a means of replenishing the skin’s αἰών. The connection to vital fluids also explains the frequent connotation of vulnerability in τέρην, since wounds release other vital fluids through the skin.446 Hermes uses the ‘soft flower of youth’ metaphor to emphasize Apollo’s acme of strength; compare

444

For ‘soft flower of youth’ see Hes. Th. 988 (male), fr. 132 (female), prob­ably Alc. fr. 397, Pi. N. 5.6 (male), A. Su. 998 (female), etc. Mimn. fr. 1.6–7 pointedly makes the ‘flower of ἥβη’ last right up to old age. For ἄνθος connected to the tanning of male adolescents or to their first facial hair see Od. 11.319–20, Solon fr. 27.5–6, Pi. O. 1.67, Alcmaeon in Arist. HA 7.1 581a14–16 ἡ τρίχωσις τῆς ἥβης ἄρχεται, καθάπερ καὶ τὰ φυτὰ τὰ μέλλοντα φέρειν τὸ σπέρμα ἀνθεῖ.   I take the semantic heart of ἄνθος to be ‘flower’, against alternative proposals (e.g. Stanford 1936: 155–7 ‘that which rises to the surface’; Aitchison 1963 ‘upward growth’; Mader in LfgrE s.v. ‘das, was den Höhepunkt des Wachstums anzeigt und verkörpert’). The only sure cognate (Sanskrit ­andhas) is also vegetal. Almost all ‘surface growth’ ἄνθη involve a colour-change (hair, spume, oxidized crusts, skin rashes, the ‘nap’ of textiles – a sense documented by Borthwick 1976; even the smoke at Pi. N. 9.23 is part of a broader vegetal metaphor), while some ἄνθη involve colour-change without surface growth (e.g. Thgn. 452, PV 23, Hdt. 1.98.6, Pl. Phd. 100d). Aitchison and Mader cited ἄνθεα ποίης against the primacy of the sense ‘flower’, but the phrase probably means ‘grassland flowers’. 445 See e.g. Archil. fr. 188.1 οὐκέθ’ ὁμῶς θάλλεις ἁπαλὸν χρόα· κάρφεται γὰρ ἤδη, A. Ag. 76–82, Arist. Prob. 1.17 861a28–30. 446 Similarly Irwin (1974) 53–6. The connection to vital juices is less clear with ἁπαλός (273*). For (i) see e.g. Il. 13.180, Od. 9.449, Sapph./Alc. fr. 16, Ibycus PMG 315, E. El. 778, A.R. 3.898; the only plant which is formulaically τέρην is pennyroyal (Hes. fr. 70.21, H.Dem. 209), which was used particularly for its oil. For (ii) see n. 444, Il. 4.237, 13.553, 14.406 (all male and vulnerable); Anacreon PMG 375 transfers it to pipes accompanying youths. Female instances include Archil. fr. 196a.6 > Hipponax fr. 119, Thgn. 261. For the focus on skin in (iii) see Hes. Th. 5, Op. 522 (both moist skin), A.R. 1.1238 (moist mouth), 3.1204 (moist); 4.871 (anointed), Cratinus fr. 335, E. Med. 905 (moist with tears), Cyc. 515 (moisture in the context). For (iv) see E. fr. 467. For the tears of the weak in (v) see formulaic τέρεν κατὰ δάκρυ χέουσα/δάκρυον εἴβει+ (x4),

351

C O M M E N TA RY 3 7 6 – 3 7 8 e.g. Il. 13.484 ἥβης ἄνθος, ὅ τε κράτος ἐστὶ μέγιστον, or Pi. N. 5.4–6, where a victor in the boys’ pancration is οὔπω γένυσι φαίνων τερείνας ματέρ’ οἰνάνθας ὀπώραν. Both ἥβη and τέρην extend into more general descriptions of privileged states of luxury and opulence: see LSJ s.v. ἥβη I.2, and e.g. SEG 30.161 πλτον ... τέρν’ (Athens, c.350 bce). φιλοκυδέος points in the same two directions: κῦδος is especially linked to various kinds of success and (unlike νίκη or κράτος) to the exultant psychological state which they induce (see e.g. κυδιῶ, ὑπερκύδαντες, κύδεϊ γαίων). For a youth the word therefore suggests ambition for military or athletic success, but also a potentially arrogant glee.447 Hermes’ point is that Apollo replaced an ephebe’s proper ambitions with an improper exercise of privilege and strength – picking on a baby. 376 αὐτὰρ ἐγὼ χθιζὸς γενόμην· τὰ δέ τ’ οἶδε καὶ αὐτός: See 273 χθὲς γενόμην, which is how Apollo οἶδε καὶ αὐτός. Hermes begins a smooth transition from narrative (370–4) through an analysis of Apollo’s ἦθος based on age (375–6) to arguments for his own innocence, reordered compared to 265–73 so as to start with his age (376–80). τ’ is metrically unnecessary, because οἶδε had once had a digamma, and dubious, because τε in δέ τε normally marks a general truth. However, there are probably enough exceptions to resist Barnes’s deletion. See Ruijgh (1971) 695–9, 897–8, 912, though his explanation from formulaic modification is unconvincing. 377 οὔ τι βοῶν ἐλατῆρι, κραταιῶι φωτί, ἐοικώς: Now Hermes reprises the start of his earlier εἰκός-argument (265 οὔτε βοῶν ἐλατῆρι, κραταιῶι φωτί, ἔοικα), but with slight variations in effect. The last few lines have depicted Apollo as κραταιός (see also 386* κρατερῶι). And by using ἐοικώς Hermes can introduce the stronger οὔ τι, and reappropriate appearances for his defence by responding as precisely as possible to Apollo’s slur μελαίνηι νυκτὶ ἐοικώς (358*). 378–9 πείθεο· καὶ γὰρ ἐμεῖο πατὴρ φίλος εὔχεαι εἶναι· | ὡς οὐκ οἴκαδ’ ἔλασσα βόας: The parenthetic γάρ-clause throws rhetorical emphasis onto πείθεο by isolating it, and onto ὡς … βόας by raising suspense during the A.R. 3.461. For crying, anointing and αἰών see Onians (1954) 200–5, 209–11. For the presentation of baths as invigoratingly rehydrating see Ginouvès (1962) 418–24. 447 φιλοκυδής occurs in literature only at 481 and here, where it modifies formulaic ἐρ κυδέος + ἥβης (Il. 11.225, Hes. Th. 988, perhaps originally of Hebe herself). Schmitt (1856: 155) already read the modification in terms of characterizing Apollo’s ambition.

352

C O M M E N TA RY 3 7 9 hyperbaton. Hermes uses three tactics to appeal to Zeus’s trust, even while he knows that Zeus knows that Apollo is basically right. First, he substitutes the point at issue with οὐκ οἴκαδ’ ἔλασσα βόας, which is true and a clever piece of misdirection, but irrelevant to Apollo’s charge. Secondly, he reasons that Zeus’s implicit acknowledgement of Hermes as a son (332) makes him a πατὴρ φίλος, and both family relationships and φιλία should rest on mutual trust; he thereby also gives a broad hint at the duty to help a φίλος in difficulty (e.g. during a prosecution) regardless of the merits of the case.449 Thirdly, καὶ … ἐμεῖο go together, as the emphatic ἐμεῖο shows, to give ‘you are my father too’ (cf. Denniston 108): Hermes guilt-trips Zeus into pity by implying that he is being more fatherly to Apollo. However, εὔχομαι far more often occurs when sons are averring their relationship to notable (fore)fathers than vice versa (see LfgrE s.v. 2d, Gemoll); the other exception is Polyphemus’ claims at Od. 9.519, 529, which characterize his scorn for the gods. Hence Hermes’ presentation of Zeus’s pride in his son is probably more cheeky than earnest. πείθω ὡς + indicative is a rare construction until the early fourth century; until then, the verb normally governs an infinitive or consecutive clause. The three other fifth-century instances (S. OR 555, Hdt. 9.53.4, Th. 6.60.3) all contain the idea ‘persuade that X is necessary’, the first step away from πείθω + infinitive. The construction can tentatively be added to the relatively late linguistic features in Introduction §1.1. 448

379 ὣς ὄλβιος εἴην: With this accentuation, the three words form a parenthetic wish in an asseveration: ‘as surely as I would like to prosper’. Compare Ar. Thesm. 469–70 καὐτὴ γὰρ ἔγωγ’ (οὕτως ὀναίμην τῶν τέκνων) | μισῶ τὸν ἄνδρα, KG ii.494–5. The construction was identified here by Dartona (1537), the accent added to ὡς by Châteillon. The wish contains a joke, in that it was precisely in order to prosper that Hermes stole the cattle; see 24 and 171 for Hermes’ ungodlike concern for material wealth. One can also write ὡς and understand a purpose clause with optative (GH ii.223): ‘in order to be prosperous’. The manuscript tradition is unreliable on different accentuations of ὡς (Devine and Stephens 1994; 356–61, Probert 2006: 138–42). But it is less of a coincidence for the wishing idiom to be mistakable for a purpose clause than vice versa.

Compare Fraenkel (1965) 30–49 on hyperbaton around vocatives. γάρ-clauses are among the most common type of Greek parenthesis: Schwyzer (1939) 15–18. Compare e.g. Il. 2.484–8 for the position between an initial imperative and the subordinate clause dependent on it. 449 For φίλος applied to family-members see Robinson (1990). The role of mutual trust and honesty in ancient friendship are emphasized by the formula πιστὸς ἑταῖρος. Cf. Hes. Op. 707–9, Donlan (1985) 230–4, Herman (1987) s.v. ‘trust’. 448

353

C O M M E N TA RY 3 8 0 – 3 8 2 380 οὐδ’ ὑπὲρ οὐδὸν ἔβην: Another misleading true statement, like οὐκ οἴκαδ’ ἔλασσα βόας. With οὐδέ as simply ‘nor’ Hermes can still be referring to his return home, when he did not ‘step’ indoors (146). However, the echo of 23 οὐδὸν ὑπερβαίνων points the audience’s attention to the possibility that Zeus would understand οὐδέ ‘nor even’, marking an escalated (false) claim that Hermes did not even leave the cave. τόδε δ’ ἀτρεκέως ἀγορεύω: Hermes adds ἀτρέκεια to the ἀλήθεια and νημέρτεια promised in 368–9, in what may sound like ‘protesting too much’. The series suggests that τόδε refers to the whole speech, not only to 379–80, so that the clause is a further thing for Zeus to believe and should be preceded by a comma.450 For the external audience, Hermes’ words also oppose the charge he laid against Apollo in 272: τὸ δ’ ἀπρεπέως ἀγορεύεις. Apollo will echo them in 459*. 381–2 Ἠέλιον δὲ μάλ’ αἰδέομαι καὶ δαίμονας ἄλλους, | καὶ σὲ φιλέω, καὶ τοῦτον ὀπίζομαι: Hermes splits his audience into three groups, and articulates his social ties to each with a different verb. There is a strong break after ὀπίζομαι, and the usual punctuation with a full stop before Ἠέλιον leaves these clauses poorly connected to their surroundings. They follow naturally in the series of clauses subordinate to 378 πείθεο … ὡς: Zeus should believe Hermes because he believes that Hermes feels appropriate social obligations. To the gods in general Hermes claims to feel αἰδώς, a constraining desire to avoid others’ disapproval (see 5*). This is less likely to derive from respect for one’s social superiors (since Hermes is battling against his inferior status), or general reverence for gods (even human religious αἰδώς is normally specific to contexts of supplication, xenia and oaths, despite some exceptions such as Thgn. 1179, A. Pe. 809–10). Rather, Hermes’ αἰδώς corresponds best to a human’s feeling of αἰδώς towards any fellow-citizens and their social norms (e.g. Il. 6.442); in other words, he bravely presupposes that he belongs in divine society. Maia accused Hermes of ἀναίδεια at 156, and prominent forums for demonstrating this characteristic were stealing, greed and deception (LfgrE s.vv. ἀναιδείη, ἀναιδής). Hermes does not go so far as to claim αἰδώς towards Apollo; however, the focus on Helios – who is best placed to detect lies (cf. Od. 8.270–302, H Dem. 38–87) – is itself an ἀναιδής piece of misdirection, since Hermes stole the cows at sunset (68) and hid them so that Helios is in the dark. As Ilgen commented, ‘Iocus vero simul inest, si audimus Mercurium dicere, se Solem revereri: nam nocte furtum peregerat, utpote qui esset νυκτὸς μελαίνης ἑταῖρος; vid. vs. 290.’ For τόδε referring to the contents of a recent or current speech see e.g. Il. 24.300 τόδ’ ἐφιεμένηι, Od. 12.450 τάδε μυθολογεύω. Hermann wrote τὸ δέ τ’ on the grounds of ‘epicorum usus’, but δέ τ’ applied to a one-off proposition would be unusual (376*).

450

354

C O M M E N TA RY 3 8 3 In 382 Hermes gives special attention to the two other protagonists of this scene. His φιλία for Zeus constrains him to honesty – the converse of Zeus feeling trust as his πατὴρ φίλος (378–9*). Apollo is referred to simply with τοῦτον (so in 385), which often marks interpersonal distance, in the form of contempt (e.g. Il. 14.471, Od. 20.380), fear (e.g. Il. 18.257, Od. 17.566, 22.78) or hostility (e.g. Il. 9.118). Towards Apollo, Hermes feels not φιλία but ὄπις, an inhibiting consideration that a stronger party (esp. a god) will look on, become angry and perhaps force atonement.451 This gives a reason not to rob Apollo or lie about it now, while continuing to represent him as a particularly threatening, violent character. The opposition of Hermes’ φιλότης towards Zeus and ὄπις towards Apollo sets up his eventual switch to φιλότης with Apollo too (507–8*). Hermes also implies that he does not feel ὄπις or αἰδώς before Zeus, the god of oaths, in whose sight he is about to commit perjury (383–4). 381 is the only line in Herm. without a third-foot caesura; compare Eumelus fr. 3.2 Ἠελίου τε καὶ Ἀντιόπης, τότε δ’ ἄνδιχα χώρην. For lines of this shape see Lehrs (1860) on Il. and Od., and Eberhard (1886) 15 on H.Hom., wrongly including H Dem. 203. 383 μέγαν δ’ ἐπιδέξομαι ὅρκον: Since 384* belongs closely with οὐκ αἴτιός εἰμι, the lemma should be treated as a parenthesis. I propose ἐπιδέξομαι based on the consideration that the manuscript variation (ἐπιδεύομαι M: ἐπιδαίομαι Ψ) is best derived from ἐπιδέ*ομαι. The contour of the sentence (‘You already know … but I shall …’) suits ἐπι- in the sense ‘in addition’. For δέχομαι + ὅρκον ‘accept an oath-challenge’ see D. 39.3–4, Ps.-D. 40.10–11, 33.13–14; for ἐπιδέχομαι ‘accept/receive in addition’ see LSJ s.v. I and e.g. Ps.-Arist. Mir. 844a20 (of ἄργυρος), IG XII.9 234 (Eretria, c.100; of ὠφελία). Hermes would be inventing such a challenge from Apollo, recalling and even outdoing his cheekiness at 274 (‘If you like, I will swear’). Of other approaches, Barnes’s ἐπιδώσομαι seems the best: δίδωμι ὅρκον means ‘I offer an oath’ at e.g. E. IT 735, X. Cyr. 5.1.22; cf. Il. 22.254 θεοὺς ἐπιδώμεθα ‘let us swear by the gods’.452 With ἐπιδήσομαι one could compare E. Med. 161–3 μεγάλοις ὅρκοις ἐνδησαμένα τὸν … πόσιν.

451

For this sense see e.g. Pi. O. 2.6 with Rakoczy (1996) 67–9; it is the normal implication of ὀπίζομαι. The noun’s commoner sense is the (angry) look of a god. Od. 13.148 offers a parallel for ὀπίζομαι of one god’s attitude to a more powerful god (Poseidon to Zeus). 452 ἐπιδώομαι would be more prone to corruption, but singular hortatory subjunctives generally express compliance with a preceding imperative: Barrett (1964) 268. There are no parallels for Führer’s suggestion (LfgrE ii.202; cf. Pötscher 2004) that δαίομαι could be used of oath-sacrifices, which are anyway irrelevant. ἐπ μαίομαι

355

C O M M E N TA RY 3 8 4 384 οὐ μὰ τάδ’ ἀθανάτων εὐκόσμητα προθύραια: μά-interjections function independently only when answering ‘No!’ to a preceding clause. Since 385–6 make sense as a separate thought, 384 reinforces something that precedes, namely οὐκ αἴτιός εἰμι. For the order, compare Ar. V. 26 οὐδὲν γὰρ ἔσται δεινόν, οὐ μὰ τοὺς θεούς, Pl. 551 ἀλλ’ οὐχ οὑμὸς τοῦτο πέπονθεν βίος, οὐ μὰ Δί’, οὐδέ γε μέλλει.453 Hermes turns his claim not to be responsible into an oath. οὐκ αἴτιος is rather vague: Hermes may admit that he did what Apollo alleges, but claim that no crime attaches to it; as in other negations of αἴτιος (Il. 19.68, A. Ag. 1505–8, D.S. 13.31.2), he may even not very plausibly be passing the buck to a higher authority under which he acted – and after 382 οἶσθα καὶ αὐτός the point would be that Zeus should know that Hermes’ actions were just part of Zeus’s plan. But probably the line comes across as a more straightforward falsehood than his careful oath to Apollo at 275–7*. In Athenian terms, Hermes offers an oath without πρόκλησις (a mutually agreed suggestion to proceed using an oath) – something which, at least in the fourth century, carried no legal force and was represented as an easily abused rhetorical ploy. At D. 54.38–41 Ariston swears an oath he had offered at the preliminary hearing, then immediately warns the jury not to trust the oaths that Conon will swear.454 Moreover, children did not generally take binding oaths (274*). The ὅρκος promised in 383 turns out to be elaborately formal, with two very rare words at its heart. Though elsewhere προθύραιος is adjectival, here προθύραια is a substantive probably modelled on προπύλαια, first attested in Hdt. 2.63 and IG I3 462 (437/6) with reference to the monumentalized architectural form of an entrance-way. This entrance-porch to Zeus’s palace is εὐκόσμητα, which I take to mean ‘splendidly adorned’. The πρόθυρα of houses and sanctuaries are frequently described in terms of decoration or architectural attractiveness, for example Pindar O. 6.1–4 χρυσέας ὑποστάσαντες εὐτειχεῖ προθύρωι θαλάμου | κίονας ὡς ὅτε θαητὸν μέγαρον | πάξομεν· ἀρχομένου δ’ ἔργου πρόσωπον | χρὴ θέμεν (Van Herwerden 1882: 1–2), ⏑ ἐπαιδέομ’ (Shackle 1920: 100) and ἐπιδήομαι (Agar 1928: 36) are semantically implausible; Humbert’s ἐπιθήσομαι is palaeographically distant. 453 The normal form of a negative oath is ‘(οὐ) μά οὐ ’. For postponement of μά see e.g. E. Hipp. 306, Ar. Nu. 817; ναὶ μά is postponed in positive oaths at e.g. Anan. fr. 4, Eup. frr. 79, 84, 270. 454 On the status of litigants’ oaths in Greece I recommend Mirhady (1991), Thür (1996), Gagarin (2007). Papakonstantinou (2007: 106–7) argued that perjury was part and parcel of the reality of litigation, not a heinous crime which it is unthinkable that Hermes should commit; see also e.g. D. 49.65, Arist. Rhet. 1 1377a7–b12. Ovid perhaps had our passage in mind when his Mercury is reminded of the cattle-theft after a prayer focusing on the pardoning of perjury (Fasti 5.681–92).

356

C O M M E N TA RY 3 8 4 τηλαυγές, N. 5.53 (of Aiakos’ temple, adorned with victory-wreaths), IG I3 953 προθύρου κόσμον (in a sanctuary, probably two wreaths), E. Alc. 101 (of a house, for mourning), Clearch. fr. 24 (of temples, and the house of a loved one), Theodoridas 17 G–P, Posidippus 142 (with statues), Heraclides Criticus fr. 1.8 (of Tanagran houses), D.Chrys. 32.33 (of a temple during Mysteries).455 We may expect Hermes to swear by Zeus’s head, as suggested at 274*. It is more common for both gods and humans to swear by gods, who can exact atonement for perjury. When they swear by an object it is generally of clear value to the swearer and to gods, and often relates to the content or audience of the oath.456 The principal value of Zeus’s porch for Hermes is twofold: it is the threshold of the Olympian community, which he is currently asserting his right to enter; and a representative of a location where in the fifth century he would share control with Apollo, in that both herms and images of Apollo Agyieus were prominent kinds of porch ornamentation. For Herm.’s connection of Hermes to thresholds see Introduction §3.3.6, and 15* for a hint of his title προπύλαιος. On herms see Rückert (1998), LIMC V(1).295–306, 374–8 with further literature. Finds so far suggest that they spread from Attica after c.515, reaching at least Siphnos, Eretria, Arcadia, Clazomenae, Larissa, Iolkos and Rhodes by c.450.457 They are situated in porches at e.g. Th. 6.27, ‘Anacr.’ FGE 14 ἀγλαΐην προθύροις | Ἑρμῆι τε κρείοντι, ‘Simon.’ FGE 44 (corrupt). For Apollo Agyieus, see LIMC s.v. and e.g. Ar. V. 875 ὦ δέσποτ’ ἄναξ γεῖτον Ἀγυιεῦ, τοῦ ’μοῦ προθύρου προπύλαιε, Men. fr. 884 μαρτύρομαι ... τὸν Ἀπόλλω τουτονί | καὶ τὰς θύρας. At Call. H.Art. 142–3, Hermes and Apollo meet Artemis at the porch to Zeus’s palace, presumably as προπύλαιοι (Spanheim 1697: ii.231–3); I doubt whether Callimachus had our line in mind. Van Den Berg (2001: 262 n. 14) suggested that the ‘porch of the gods’ in Procl. De mal. subst. 14.15–18 may be modelled on our line, but it arises from allegorical interpretations of Pl. Phlb. 64c ‘porch of the good’: Erler (1978) 54 n. 2.

455

This connection tells against Ilgen’s less tangible interpretation of the porch as the sky adorned with stars (compare the gates of cloud at Il. 5.749–50), though this does bring out the cosmic aspect of εὐκόσμητα well. For πρόθυρον in Epos as the space outside the door see n. 213. The old restoration of προθ̣υ̣[ρα]ί̣οις in an imperial-era Egyptian inscription, noted by Vergados, is unlikely: Bernand (1969) 500. 456 See also 460*, or Hera swearing fidelity to Zeus by their marriage-bed at Il. 15.39. Torrance, in Sommerstein and Torrance (2014) 111–31, surveys objects by which oaths are sworn, with a focus on their potential to take on autonomous quasi-divine power. 457 This picture may of course change: the Sounion herm (Trianti 1977: 119) extended our understanding of herms in Attica backwards by several decades.

357

C O M M E N TA RY 3 8 5 – 3 8 6 385 καί ποτ’ ἐγὼ τούτωι τείσω, ποτὶ νηλέα φώρην: M’s ποτὶ ... φώρην is superior to Ψ’s ποτε ... φωνήν. Both ποτί and φώρην are difficiliores, Ψ’s repeated ποτε within a single clause lacks close parallels, 136 offers a parallel for Ψ changing the obscure φωρ- into φων-, and Hsch. φ.1110 lists the form φώρην, extant only here. φωρά normally means ‘theft’, but the sense ‘detection’ (here specifically of a house-search, as in LSJ s.v. φωράω 1) is given by Hesychius and used in Achilles Tatius 7.11.1, Oenomaus’ title Γοήτων φώρα, and D.L. 1.96.458 The use of an ambiguous word in its less common sense neatly allows Hermes to allude to Zeus’s joke (330) that Apollo has stolen Hermes: cf. Agar (1928) 36. For ποτί ‘in proportion to, at the price of ’ see LSJ s.v. πρός C III.4. The sense ‘in addition’ (LSJ s.v. D) is ruled out by word-order, while προστίνω is extremely rare. ποτί means that τίνω is used absolutely, as in e.g. Solon fr. 13.29. LSJ s.v. τίνω gives the epigraphic evidence for favouring the spelling τείσω. The manuscripts, as often, have τίσω. In Apollo’s view, Hermes owes him at least fifty cows. Hermes concedes ‘One day I will pay this person back …’.459 But he then twists his words into a threat: ‘… in proportion to his brutal ransack’, i.e. by ransacking Apollo’s ‘house’, most naturally taken as Delphi (see 177–81, 523). This threat continues the accusation of bullying from 371–6; νηλής is highly emotive in Epos, generally used of the absence of pity of a person or external force (bronze, fate, old age, etc.) as it causes physical harm.460 To threaten retribution before an arbitrator may sound misguided, but by packaging it in ἐξ ἀπροσδοκήτου word-order, with play on the two senses of φώρα, Hermes takes the edge off his aggression – a tactic also tried by Apollo at 362–3*. 386 καὶ κρατερῶι περ ἐόντι: The echo of 377 κραταιῶι brings home Hermes’ implicit point that it is Apollo who is like a cattle-rustler. σὺ δ’ ὁπλοτέροισιν ἄρηγε: ὁπλότερος ‘younger’ < ὅπλα + contrastive -τερος. The derivation suggests the original sense ‘young men who can work and fight’ as opposed to the elderly: Del Freo (1994). The superlative is found at Il. 9.58. Hermes ends by making a resounding contrast with 385: I versus you; future versus present; harm the strong versus help the weak; specific versus general (note the imperfective imperative; the dominant usage of ἀρήγω in Epos, though not later, is for habitual divine support: LfgrE s.v.). For the stem 458

Hsch. claims that the two senses had different accents; I follow this without confidence. For φώρα ‘detection’, Schneider (1856: 289) made the analogy φώρ: φωράω: φώρα ~ θήρ: θηράω: θήρα. 459 For καί ‘actually’ see 291–2*. For τούτωι see 381–2*. 460 The extension to pitiless actions could be called hypallage, or compared with more flexible adverbial uses such as Anacr. PMG 417.2 νηλεῶς φεύγεις.

358

C O M M E N TA RY 3 6 8 – 3 8 6 ἀρηγ- applied to legal support see e.g. Il. 18.502, 23.574 ἐπ’ ἀρωγῆι (opposed to ‘unbiased’), A. Su. 726 ἀρωγοί, Ag. 40–7, Eu. 486. Helping the young per se is not a prominent principle of Zeus’s behaviour or Greek law. However, it is the relevant application of two broader principles to such an instance of (alleged) bullying. The first, seen in Hesiod Op. and numerous Athenian sources, is that legal procedures are designed to restrain social power-relations and promote ‘fair’ distribution, while unconstrained demonstrations of power are framed in terms of thinking oneself above the law.461 The second is that Zeus characteristically protects other vulnerable groups, particularly suppliants and guests. 368–86 Hermes’ defence speech before Zeus: Considering the constraints of space and content, Hermes’ defence is a reasonable adaptation of the four parts enjoined by later theory for a forensic speech, including several of their major functions: Görgemanns (1976) 117–19. Lines 368–9 are a proem focused on ἦθος, in response to Apollo’s focus on Hermes’ ἦθος. Hermes’ youth puts him in a weak position: compare Diomedes at Il. 14.110– 12 αἴ κ’ ἐθέλητε | πείθεσθαι, καὶ μή τι κότωι ἀγάσησθε ἕκαστος | οὕνεκα δὴ γενεῆφι νεώτατός εἰμι. Hermes promises ἀλήθεια (368*), a primary aspect of which is a complete, rounded account. His narrative (370–7) does fill in a gap in Apollo’s, namely a description of the latter’s actions: he matches Apollo’s tactic of focusing on the opponent’s activities rather than those of the speaker. The strategy, combined with skilful selectivity, allows Hermes to make a reasonable complaint about Apollo’s heavy-handedness, and to avoid telling lies about himself. The aim is an overtly emotional appeal to pity the younger victim against the teenage bully, combined with a substantive claim about Apollo failing to follow procedure. Hermes also seamlessly works in the argument from εἰκός at the centre of his whole speech (376–7*), but – with an emphasis corresponding to Apollo’s – he is more interested in characterizing Apollo than in argumentation, and he does not expand.462 The remainder of Hermes’ speech corresponds in some degree to the epilogue enjoined by later theory (Lausberg 1998: 205–8). He recapitulates the main contention of his narrative (379–80), and continues to play on the

461

See e.g. Hes. Op. 219–24, Solon fr. 3.32–9, E. Su. 435–7, Th. 1.77.3, 2.37.3, Pl. R. 1 338c, Grg. 483c–d, and throughout D. 21 (e.g. sections 66, 76, 123–5, 188, 210). Ober (2005) observed that the conceptualization and implementation of equality before the law may have been quite different in an oligarchy. But Hesiod suggests that Hermes’ basic implications would have made sense to audience-members from a range of states. 462 Heiden (2010: esp. 420–1) also emphasized the malleability of ‘facts’ and the greater importance of interpersonal relationships in Hermes’ speech.

359

C O M M E N TA RY 3 8 7 audience’s sympathies (esp. 381–2* ὀπίζομαι, 385* νηλέα, 386* ὁπλοτέροισιν ἄρηγε). The oath is inserted with enough syntactic disruption to suggest an artless level of emotional investment; the entire section is framed with reference to the audience, including specific appeals to Zeus’s relationship with Hermes (378, 382), and the implication of belonging when he claims to feel αἰδώς towards the other gods (381–2*). This is a tactic not used by Apollo; it corresponds to Hermes’ immediate sensitivity to the response of an audience at 270–1. It is this approximation to the classical form and functions of an epilogue that is hardest to parallel in other speeches in Epos (Introduction §1.2). Hermes’ style is fairly unadorned (with the exception of 373–4 πολλῆς | πολλά), and he keeps to the natural pauses of the hexameter. I am not as certain as Van Nortwick (1975: 93–5) that this sounds specifically childish, in the absence of the clearer markers which Apollo has ridiculed (363–4). When we compare Apollo’s aims and strategies in 334–64*, there is no neat opposition of amateurish, blustering Apollo and cutting-edge, virtuoso Hermes. Apollo is more virtuosic with words, Hermes better at working his audience.

387–96 387 ἐπιλλίζων: ἰλλός and its derivatives apply to a range of eye-gestures. The less ambiguous instances fall into three broad categories: (i) not looking straight, (ii) staring unresponsively, and (iii) lowering an eyelid. The action therefore recalls 278–9* where, after his previous protestation (cited on Olympus by Apollo at 363–4 and partly reprised by himself in 377), Hermes looked from side to side and blinked or winked repeatedly. ἐπιλλίζω may cover or at least suggest both actions simultaneously. As at 278–9, the tone of these gestures seems to be either seductively (and suspiciously) cute or embarrassed. The latter could betray Hermes’ guilt, or could be a front to demonstrate his αἰδώς before the other gods (381): cf. A.R. 4.1315–16, where Jason looks away from the Libyan heroines out of αἰδώς. Regarding sense (i), στραβός, διεστραμμένος ‘squinting’ and the like are dominant glosses on ἰλλός, ἔπιλλος and ἰλλαίνω from Herodian onwards; so the transferred use at anon. SH 1026 ἰλλοὶ δὲ πόδες, and Eustathius’ unpleasant advice δύναται δέ τις καὶ ἰλλῶι ἀνδρί, ὅπερ ἐστὶ στραβῶι, οὐκ ἀφυῶς ἐπισκῶψαι τὸ ‘ἑτέρωσε βάλλειν ὄμματα’ (Comm. Od. ii.120). Compounds of ἰλλός also refer to looking aside without a squint.463 This explains why

463

See Poll. 2.52 (ἰλλώπτειν), 2.54 (ἀνίλλωμα, upward), Ap.S. p. 73.27 (downward), Phot. π.414 (παριλλαίνουσαν). Eust. Comm. Il. ii.693 relates ἰλλίζω to δενδίλλω ‘look to and fro’. Compare παραβλῶψ ‘looking aside’ > ‘squinting’.

360

C O M M E N TA RY 3 8 7 ἰλλώπτω often signals teasing or mockery, which is a well-attested implication of sideways and downward looks.464 A squint eye does little work, and so (ii) looks unresponsive. At Ar. Th. 846 the kinsman becomes ἰλλός by prolonged gazing, probably ‘unable to focus’. According to Nic. Ther. 162–3 the asp ‘can constantly be seen fixedly ἐπιλλίζουσα with sleepy eye’, probably ‘staring unresponsively’ (after all, snakes have no eyelids).465 οἱ ὀφθαλμοὶ ἰλλαίνονται is a symptom of tetanus at Hp. Morb. 3.12; the disease makes moving the eyeball and eyelid harder. However, another ocular symptom of tetanus is (iii) half-closure of the eyelid through contraction of the orbis ocularis, and squinting often co-occurs with ptosis (drooping eyelid). Galen UP pp. 805–6 interprets ἴλλωσις as καμπύλον βλέφαρον, which half-closes the eye in the Hippocratic corpus (Prog. 2.33, Coac. 214.). Compare Ael. Dion. ι.8 explaining ἐνιλλώπτειν as τὸ ὑγροῖς (‘oozing desire’: Onians 1954: 202–3) καὶ ἐπιμύουσι καὶ κατακεκλασμένοις (‘effeminate’) ὀφθαλμοῖς ἐμβλέπειν. Σ Ar. Eq. 292a equates ἰλλωπεῖν with rapid blinking. Finally, two passages connect the stem simultaneously to rapid eye-shifts and fluttering the eyelids in seductive or lustful looks: Clement Paed. 3.11.70 explains ἐνιλλώπτειν as διὰ τῶν ὀφθαλμῶν βλεφαρίζειν, which like κλαδαραὶ ὄψεις is nothing but διὰ τῶν ὀφθαλμῶν μοιχεύειν. Similarly, Adamantius Phys. 1.23 makes ἰλλώπτειν an action of the eyeball (shifting glances?), combined by ἀνδρόγυνοι and μοιχικοί with blinking or half-closing the eyelids.466 This suggests that in our passage ἐπιλλίζω might describe simultaneous movements of the eyeballs and eyelids.

See A. Eu. 113 (ἐγκατιλλώπτω), Plu. Mor. 51c–d (οἷον ἐπιλλώπτουσαν ἐξ ὀφρύος … παρρησίαν of some flatterers), Poll. 2.52 (κατιλλώπτειν glossed as καταβλέπειν ἐπὶ χλευασμῶι; cf. Clem. Paed. 3.11.73), Hsch. κ.1808 (κατιλλαίνω related to μυκτηρίζω, ‘turn one’s nose up at’ i.e. ‘look down one’s nose at’; cf. Paus. Att. on ἰλλώπτω and ἐπιλλῶ). The exact point at Od. 18.11–12 μοι ἐπιλλίζουσιν ἅπαντες, | ἑλκέμεναι δὲ κέλονται, ἐγὼ δ’ αἰσχύνομαι ἔμπης is unclear, but Apollonius borrowed the word as ‘taunt’ (1.486, 4.389, and 3.791, where Σ glosses ‘ἐπικλίνειν the eyes in mockery’). Sideways looks also convey mockery or scorn at e.g. Ar. Nu. 362–3, Ael. Dion. σ.19 (perhaps taking σιλλοῦν from σείω + ἴλλους, as implied in Poll. 2.54). 465 Sophron fr. 156 ἰλλοτέρα τᾶγ κορωνᾶν could mean ‘more beady-eyed than crows’ or (as in (i)) ‘more crooked than hooks’. Glosses on γλοιάζω include Hsch. γ.655 κατιλλώπτειν, Phot. γ.140 ἐπιλλώπτειν, Galen Ling. exol. p. 91 παραδίνειν τοὺς ὀφθαλμούς. I interpret ‘have tired eyes’, as shown by blinking and oscillation of the pupils. 466 For the lustful connotation see also Philemon fr. 125 (a man who κατιλλώπτει a woman from behind) and anon. epig. 36.3 G–P θῆλυ κατιλλώπτοντι Πριήπῳ. These may be related to gaze-shifts, staring or drooping/fluttering eyelids. 464

361

C O M M E N TA RY 3 8 8 – 3 9 1 388 καὶ τὸ σπάργανον εἶχεν ἐπ’ ὠλένηι οὐδ’ ἀπέβαλλεν: Epos rarely extends attributive discourse (i.e. the words immediately before and after a speech, attributing it to the speaker) into a description of the speaker’s clothing and posture. The unusual detail and the duplication in εἶχεν … οὐδ’ ἀπέβαλλεν (compare 243*) point attention to the lyre’s position, and to the possibility that Hermes could have revealed it at this point. Olympus would be an impressive stage for his first public performance. But he need not reveal his best trick yet, and continues to control his swaddling to his advantage (as at 151–2, 237–42, 306): he can already see Zeus’s response. 389 Ζεὺς δὲ μέγ’ ἐξεγέλασσεν ἰδὼν κακομηδέα παῖδα: Whereas Apollo reached a chuckle at 281* from a position of anger, Zeus was initially well disposed and can laugh heartily. κακομηδέα resembles 282 δολοφραδές, and the laugher’s ability to see through Hermes’ tricks is again part of the reason for laughing. 390 εὖ καὶ ἐπισταμένως ἀρνεόμενον: ἐπίστασθαι describes a practical ability, often an acquired ‘skill’. Several craft metaphors for skilled speech have IndoEuropean heritage. Epic examples include the formula πυκινὸν ἔπος, Il. 14.92 ἐπίσταιτο ... ἄρτια βάζειν (of persuasive speech), and Od. 11.367–8 of a narrative (see 25*, 447*). By the fifth century, the analysis of rhetoric as a learnable skill (τέχνη) was becoming much more fine-grained, and an audience may have related ἐπισταμένως to topics such as the structure of Hermes’ speech (see 368–86*). In any case, Zeus’s perception of Hermes’ intellectual ability at speaking gives the lie to his opening claim (369 οὐκ οἶδα ψεύδεσθαι), and the juxtaposition of εὖ with 389 κακο-μηδ-έα points up the non-alignment of eloquence and honesty. ἀμφὶ βόεσσιν: For ἀμφί + dative giving the topic of a speech see e.g. Od. 4.151. The legal setting combines this with the usage for an object of dispute (GH ii.87); compare IC IV 72 i.18 (Gortyn, before 450) for a dispute over a slave, ἀνπὶ δόλοι. 391 ἀμφοτέρους δ’ ἐκέλευσεν ὁμόφρονα θυμὸν ἔχοντας: For the external audience, ἀμφοτέρους … ὁμόφρονα θυμὸν ἔχοντες echoes the start of the scene, where the two gods arrived ἀμφὶς θυμὸν ἔχοντες (315) to where justice awaited ἀμφοτέροισι (324); the rehabilitation of the stem ἀμφ- is neat. We may also recall the use of ὁμόφρονες in 195, where Apollo was imagining ideal dogs who normally worked together for the benefit of the herd, as he and Hermes are to do now. For ὁμόφρονα θυμὸν ἔχοντες see also Il. 22.263, H Dem. 434 (with Introduction §1.3), Thgn. 81.

362

C O M M E N TA RY 3 9 2 – 3 9 5 392 Ἑρμῆν δὲ διάκτορον ἡγεμονεύειν: Zeus brings fulfilment to Apollo’s ‘prophecy’ that Hermes would lead him to the cows (303*; cf. 394*), and completes the progress of Hermes as leader since 259*; for the use of one of Hermes’ standard epithets see below, 387–96*. Here the juxtaposition with ἡγεμονεύειν implies the connection of διάκτορος to διάγω and the interpretation ‘guide’. Related ancient interpretations are the more specific ‘messenger’ and the less specific ‘attendant’, and the epithet seems normally to be contextually pertinent in Epos, where Hermes’ appearances centre on these functions.467 Of the eight uses in Il., Hermes is a guide for six (in book 24) and probably a herald at 2.103. Of the ten uses in Od., Hermes is a messenger for six, a guide of souls at 24.99, a patron of attendants at 15.319. 393 ἐπ’ ἀβλαβίηισι νόοιο: Zeus ordered Hermes to indicate the place ‘in a condition of [LSJ s.v. ἐπί B I.1i] planning no harm’; Allen and Sikes paraphrased ‘bona fide’. The phrase evokes classical legal formulas such as ἀβλαβῶς καὶ ἀδόλως for how the parties to alliance agreements should treat each other: see IG I3 s.v. ἀβλαβ-, Th. 5.18, SEG 29.405b.6 (Olympia, fourth century), TAM ii.1183.3 (Phaselis, fourth century). An alternative interpretation of ἀβλαβίη νόοιο could be a state where a mind (either objective or possessive genitive) suffers no βλάβη, but the idea that Hermes might be punished with madness is far-fetched when the talk has been, if at all, of tartarization.468 394 ὅππηι δαὖτ’: δαὖτε = δή + αὖτε. αὖτε marks surprise in the indirect question (see LfgrE s.v. 2b, where the emphasis on indignation is too narrow) while δή strengthens ὅππηι (Denniston 219): cf. Od. 10.281 πῆι δαὖτ’, Archil. fr. 88 πῆι δηὖτ’. For Epic δαὖτε versus lyric δηὖτε see Wackernagel (1916) 9. βοῶν ἴφθιμα κάρηνα: The echo of 302 reinforces the relationship of 392* ἡγεμονεύειν and 303 ἡγεμονεύεις. For the formula see 94*. 395 νεῦσεν δὲ Κρονίδης: Zeus’s nod here, to confirm a spoken order, is hard to parallel exactly. On the one hand he regularly bows his head to confirm requests, assertions and promises about the future; on the other νεύω denotes gestures which function as orders, but generally in contexts where spoken 467

Apparent exceptions: Il. 21.497, Od. 8.335–8, Herm. 514. The sense ‘attendant’ is opposed to ‘guide’ at Plu. Mor. 777b: λόγος qua reason is a gift of Hermes ἡγεμών while λόγος qua speech is διάκτορος and instrumental. 468 For mental impairment compare e.g. Thgn. 704 βλάπτουσα νόοιο, E. Hipp. 511 οὔτ’ ἐπὶ βλάβηι φρενῶν. For ἐπί see LSJ s.v. B III.3 ‘on terms of ’, as in A. Ag. 1024 ἐπ’ ἀβλαβείαι. Radermacher considered sense B III.2 ‘with a view to’, as in ἐπὶ βλάβηι ‘with intent to harm’; ἀβλαβίη then has to mean ‘avoiding impairment’.

363

C O M M E N TA RY 3 9 6 orders are excluded. If Herm. was designed for performance at fifth-century Olympia (Introduction §2), it is tantalizing that Pheidias’ statue of Zeus there (c.430) was allegedly inspired by Zeus’s nod at Il. 1.524–8 (Schwabl 1976: 28 n. 8). ἐπεπείθετο δ’ ἀγλαὸς Ἑρμῆς: For Hermes’ reasons for accepting see 387– 96* below. The application of ἀγλαός to a proper name is rare, and develops from its application to τέκνον and υἱός, where it implies parental pride (314*); here, the extent of that pride is very much at stake. The earliest other use of ἀγλαός with a proper name is Pi. I. 8.27 (478 bce). 469

396 ἔπειθε Διὸς νόος αἰγιόχοιο: πείθω normally has a personal subject. However, the expression resembles an active counterpart of πείθομαι βουλῆι (Il. 12.109, 241); Zeus’s mind/plan can also ‘rouse’ other gods (Il. 15.242, Od. 24.164). For Διὸς νόος see 10* and for αἰγιόχοιο see 183*. 387–96 Zeus’s mediation: The hymnist gives Zeus’s words in indirect discourse – generally a rare form in Epos, but favoured (as here) for orders which are carried out without further ado.470 This keeps the focus on the two antagonists and the opposition of their extended direct speeches, which are enclosed by a pair of more succinct interventions by Zeus. While reconciliation was a primary goal of ancient Greek arbitration (as emphasized in Roebuck 2001), Zeus’s instructions simply tell Hermes and Apollo to have a ὁμόφρων θυμός. This demonstrates his insight into each god’s motivation: the evident irony of asking (while laughing) for Hermes to ζητεύειν, as well as to ἡγεμονεύειν | καὶ δεῖξαι τὸν χῶρον where he hid the cows, soothes Apollo’s indignation with more humour and an acknowledgement (less acerbic than at 330–2*) that Zeus appreciates what really happened. Meanwhile, he sees that Hermes will leave satisfied so long as his overtures to join Olympus as Zeus’s son (esp. 378*) are not rejected outright. And indeed, his laugh demonstrates acceptance, and he has publicly recognized Hermes as a guide (392 διάκτορον) and herald (331); Hermes can leave confident that his impressive début will help his underlying goal of acquiring divine status. Fletcher (2008: 23) suggested that by instructing Hermes only to ‘show’ the cows, Zeus also leaves the door open for Hermes to negotiate to 469

Zeus’s confirmatory nods: Il. 1.524–8, 8.246, 17.209, Hy. 1.D7, Sappho fr. 44a.7–8, Call. H.Art. 28, 39–40, etc.; see Schwabl (1976). νεύω of secretive or silent directives: Il. 9.223, Od. 9.468, 12.194, Hy. 7.9, Th. 1.134.1, etc.; see Lateiner (1995) 79–80. 470 See De Jong (1987) 114–18 on the Iliad. Nünlist (2007: 58 n. 20) gathered passages of indirect discourse for H.Hom.; Faulkner (2015) noted the rarity with which Zeus speaks directly within these poems.

364

C O M M E N TA RY 3 9 7 – 3 9 9 keep them. Zeus’s arbitration is ‘easy’ at this stage (396): little do we suspect that by line 403 Apollo’s anger will be revived. These lines round off the scene on Olympus with framing elements recalling the lines leading up to its start: 391* ~ 315 + 324; 392*+394* ~ 302–3; see also 397* ~ 323.

397–408 397 τὼ δ’ ἄμφω σπεύδοντε Διὸς περικαλλέα τέκνα: The word ἄμφω and the nuance of eagerness in σπεύδοντε mark the gods’ response to Zeus’s instructions, which began with 391* ἀμφοτέρους … ὁμόφρονα θυμὸν ἔχοντας. The sense in that line of a unity transcending the discord of 315–24 is enhanced now by Διὸς περικαλλέα τέκνα, which occurred in 323. This phrase is unique to Herm. (also 504), and all three occurrences track the brothers’ movements relative to Olympus; the hymnist seems to have modified formulaic περικαλλέα κούρην| (Il. 16.85; a patronym precedes at Hes. fr. 193.11) to create a marked phrase as the ground for structurally significant intratexts (as in 89*; see Introduction. §3.2). 398 ἐς Πύλον ἠμαθόεντα: Dependent either on σπεύδοντε (‘hurrying to’) or on ἷξον (in which case cf. A. Pe. 492–3 ἔς τε Μακεδόνων | χώραν ἀφικόμεσθ’ ἐπ’ Ἀξιοῦ πόρον). For the location of ‘Pylos’ see 216*; the epithet ‘sandy’ must originally have applied to Messenian Pylos and the remarkable beach at Voidokilia, but it is used in the context of ‘Pylos’ reaching the Alpheus already at Il. 11.712. ἐπ’ Ἀλφειοῦ πόρον ἷξον: πόρος can be either ‘crossing-place’ or ‘course’. Thryon/Thryoessa is called Ἀλφειοῖο πόρος in the first sense at Il. 2.591–2, H.Ap. 423–4; despite the formula, a different crossing-point can perfectly well be intended here, e.g. the ford servicing Olympia (Pritchett 1980: 267). Alternatively, the ‘course of Alpheus’ is mentioned in Pindar O. 10.48 τιμάσαις πόρον Ἀλφεοῦ, and probably at 1.92, 2.13 (all, naturally, in the vicinity of Olympia). Either way, choosing to describe the Alpheus in this way emphasizes again Hermes’ control of passage/crossing (see Introduction §3.3.6). The scribe of Θ wrote 397 σπεύδοντο ... 398 ἠμαθόεντα δ᾽ ἐπ’ …, where the odd placement of δ’ suggests his motivation was to remove the hiatus. For hiatus at the caesura compare 148 ἄντρου ¦ ἐξίκετο, GH i.90–1. 399 See 103–4* for the αὔλιον ὑψιμέλαθρον and for the earlier implication that the land constitutes ἀγροί, i.e. is subject to human management; also 262* ἀγραύλους.

365

C O M M E N TA RY 4 0 0 – 4 0 2 400 ἡχοῦ: So Fick (1896: 272), comparing IG VII 235.16 (Oropos, c.380) ἡχοι ‘where’. I suspect Hesychius η.1022 ἤχου: ἐνθάδε should be emended to ἡχοῦ: ἔνθα. The word derives from the same relative stem as epic ἧχι, with the ending of πανταχοῦ, ἀλλαχοῦ, πολλαχοῦ (Schwyzer i.624, 630). Since both morphemes were widespread individually, the word’s dialect alignment is unclear. Of the manuscripts, M ὅχου and D ἧχ’ οὗ adapted to the required meaning; Ψ had either ἧχ’ οὐ or ἦχ’ οὐ. τὰ χρήματ’ ἀτάλλετο νυκτὸς ἐν ὥρηι: ἀτάλλετο is passive (the middle does not occur). The paradosis simplified to unmetrical χρήματ’ ἀτιτάλλετο, since the reduplicated verb is frequently used of rearing domesticated animals. For χρήματα used of livestock qua property see e.g. Od. 2.203 χρήματα … βεβρώσεται. Unlike ἀτιτάλλω, ἀτάλλω retains the main connotations of ἀταλός (youth, innocence, gambolling), which here characterize Hermes’ care of the cows in terms of his own vivacious self-rearing. See e.g. the contrast of youth and old age in Pindar fr. 214 γλυκεῖά οἱ καρδίαν | ἀτάλλοισα γηροτρόφος συναορεῖ | Ἐλπίς, cited at Pl. R. 1 330e–1a in a contrast to childhood nightmares; also S. Aj. 559 with Finglass (2011), Comanus fr. 6 ἀτάλλειν κυρίως ἐστὶν τὸ ἐκ νηπίου τρέφειν, Hsch. α.8003 ἀτάλματα: παίγνια. The line-end νυκτὸς ἐν ὥρηι recalls 67 (the start of Hermes’ cattle-rustling), but when juxtaposed with ἀτάλλετο it adds a further touch of characterization to Hermes’ cow-herding: their midnight feast at 105–7 was an anomaly necessitated by Hermes having extracted them from their routine on Pieria.

401 παρὰ λάϊνον ἄντρον: For Ψ’s παρά, M offers ἐς. The picture left by 103–41 is that the landscape has a steading with a man-made roof, containing the cows and the cooked meat (103–4*, 106, 134), and a cave, rather than a rock, containing the hides (124*; see 404*). Here, Hermes goes either ‘past’ or ‘into’ a cave and fetches the cows. The latter reading demands an abrupt reinterpretation of the αὔλιον ὑψιμέλαθρον as equal to the cave. The former reading simply involves equating this cave with the one where the skins are, which Hermes has every reason to bypass. M’s reading perhaps derives from a gloss: from late antiquity on, παρά + acc. occasionally means ‘into’. The combination λάϊνον ἄντρον recurs in Theocritus 7.149 (for the poem’s use of Herm. see 450*), Mesomedes 7.1. 402 ἐς φάος ἐξήλαυνε: So Hermann. Ω had εἰς φῶς, but |ἐς φάος V- at 258, and the uncontracted form is far better attested in Epos where either is admissible. The lemma recalls 12 ἔς τε φόως ἄγαγεν* at Hermes’ birth, suggesting

366

C O M M E N TA RY 4 0 3 – 4 0 4 that for the surviving cows this is a symbolic ‘rebirth’ after their separation from the pristine landscape of Pieria, the introduction of their two colleagues to death, and their concealment. For the initiatory scheme see 493–4*. βοῶν ἴφθιμα κάρηνα: See 94*. The formula echoes two more recent uses: 394, continuing the correspondence 397* ~ 391, which shows Hermes’ ­obedience to Zeus’s instructions; and 302, where Apollo foresaw that Hermes would lead him to the cows: Fernández Delgado (1998) 8. 403 ἀπάτερθεν ἰδὼν ἐνόησε βοείας: I take M’s ἀπάνευθεν to be a simplification to a more common synonym. ἀπάτερθεν may suggest that Apollo spots the skins ‘from a distance’ (so the closest parallels: Thgn. 1059 ἀπάτερθεν ὁρῶντι, Q.S. 5.312 ἀπάτερθεν ἴδωνται); or that his action of looking is done ‘at a distance’ from that of Hermes (see Il. 2.587, 18.217).471 Either is compatible with the interpretation that Apollo spots the skins in a cave as Hermes walks past it to the steading (401). Last time Apollo was near by, he ‘noticed’ (218 εἰσενόησεν) the footprints, and spoke in bemusement, as he is about to do here. The parallelism re­­ inforces the sense that the skins are ‘signs’ available to be deciphered (125–6*), though in this case Hermes did not deliberately leave them visible. 404 πέτρηι ἐν ἠλιβάτωι: Ψ had ἐπ’ instead of ἐν (M strangely has γαίη κατ’), but ἐνὶ πέτρηι was transmitted at 124*. Barnes altered 124 to match 404, but I prefer the opposite course for three reasons. (i) For the skins to be within a cave at 124 is more coherent with Hermes’ care in clearing up after his butchery; (ii) the audience needs to be able to connect the cave of 401* to something, and πέτρη is a better candidate than αὔλιον ὑψιμέλαθρον; (iii) ἠλίβατος (whose exact sense is unclear) often applies to cliffs, suggesting that the poet understood πέτρη in the sense ‘rockface’ here, rather than ‘boulder’ (Pindar O. 6.64 in fact calls the Hill of Kronos at Olympia πέτραν ἀλίβατον). The fact that Apollodorus’ version does involve Hermes hiding the cows in a cave (Bibl. 3.112 τὰς μὲν λοιπὰς εἰς σπήλαιον ἀπέκρυψε, δύο δὲ καταθύσας τὰς μὲν βύρσας πέτραις καθήλωσε …) does not seem to me adequate evidence for his text of Herm., since he made numerous small normalizations to the story (Introduction §4.4). In this instance he is simplifying the partial domestication of the landscape (see 103–4*, which contrast with the strong connotation of wildness here in ἠλίβατος).

Allen and Halliday interpreted ‘looking aside’, and cited Od. 17.304 νόσφιν ἰδὼν ἀπομόρξατο δάκρυ. There I interpret ‘saw and privately wiped away a tear’ (cf. Il. 17.408 νόσφιν ἀκούων ‘hearing apart, in private’).

471

367

C O M M E N TA RY 4 0 5 – 4 0 6 τάχα δ’ εἴρετο: So M. Ψ’s ἤρετο is the Attic back-formation from ἐρέσθαι to ἠρόμην; the same error arose in various manuscripts of Il. 1.513. 405 πῶς ἐδύνω, δολομῆτα, δύω βόε δειροτομῆσαι: Apollo immediately juxtaposes power (ἐδύνω) and cunning (δολομῆτα), a conceptual pair which will remain important in the following lines.472 His question implies – as 406 will clarify – a discrepancy between the power needed to skin two cows and Hermes’ physique, and hence acknowledges the persuasiveness of Hermes’ own εἰκός-argument (265). The juxtaposition of δολομῆτα then suggests that not only the cattle-rustling but the slaying too involved clever technique which enhanced Hermes’ strength. In fact, the audience may recall line 117, where one of only two other uses of the stem of δύναμαι (the other being 175) described precisely Hermes’ physical strength as he tugged the cows towards their death. δειροτομέω was an emotive word (see especially Il. 21.89, 555), not just because of its sacrificial connotations (Kitts 2005: 156–70) but because δειρή already has a strong connotation of vulnerability throughout Epos. Although Apollo senses some μῆτις in the killing of the cows, he alleges the normal sacrificial method of throat-slitting, whereas at 119* we had the characteristic variant of Hermes ‘boring into’ the cows – τορῆσαι not τομῆσαι. For the sound-patterning ἐδύνω δολομῆτα δύω... δειροτομῆσαι, see Introduction §3.2. 406 ὧδε νεογνὸς ἐὼν καὶ νήπιος: The tone of νήπιος is sarcastic: Apollo realizes full well that the normal, alliterative inference from ‘newborn’ to νήπιος does not work in Hermes’ case, with respect either to mental capacity (contrast δολομῆτα) or to physical passivity (contrast ἐδύνω ... δειροτομῆσαι), and he had to face Hermes’ un-babylike pretence to be babyish earlier (237– 41, 273). For other pointed uses of νήπιος see 145–54*, 163–5*, 210*. 406–7 αὐτὸς ἐγώ γε | θαυμαίνω κατόπισθε τὸ σὸν κράτος: Apollo continues to engage ironically with Hermes’ defence ‘I am νήπιος; you are κρατερός’. The implication of ‘even I myself ’ is that Apollo accepts the label κρατερός, but only when emphasizing Hermes’ κράτος too. The noun means ‘power’, rather than the normal epic sense ‘superiority’, because of the influence of the wider range of κρατερός (cf. Il. 7.142 δόλωι, οὔ τι κράτεΐ γε). κατόπισθε locates not the verb but κράτος in time, in predicative position: 472

Kahn commented ‘C’est dans l’écart dolomètis–kratos que s’inscrit l’ensemble du passage’ (1978: 95). For forms in -μήτης and -μητις, see 13*. For ἐδύνω rather than ἐδύνασο, compare Il. 15.21 ἐκρέμω (emendable to ἐκρέμα’); models from the sigmatic aorist middle were available from Archil. fr. 172.1 ἐφράσω on.

368

C O M M E N TA RY 4 0 7 – 4 0 9 compare Thgn. 280 μηδεμίαν κατόπισθ’ ἁζόμενον νέμεσιν. Hermes’ past feat is inexplicable enough, so the thought of his future strength is stupefying; see the reference to his growth in 408. Apollo calls this response θαῦμα – a stem which will recur frequently in this scene.473 As in 343* ἀγάσσασθαι, the disapproving aspect of marvel is present (see also LSJ s.vv. θαυμάζω, θαυμαστός), so that the word effectively negotiates the mixed sentiments of surprise, irony and genuine anger. 407–8 οὐδέ τί σε χρή | μακρὸν ἀέξεσθαι: I take μακρόν to be a predicative adjective, ‘to grow tall’, or possibly (employing a fifth-century meaning) ‘to grow big’; such a collocation with αὐξ- is surprisingly rare (e.g. Arist. Or. 36.36 τὸ μήκιστον αὐξηθῆι; cf. Arist. HA 5 555a20 αὐξανόμενα δὲ ταῦτα ... μακρὰ γίνεται). West took μακρόν as adverbial and temporal (‘for a long time’; cf. 280*), but there seem not to be convincing parallels until late antiquity. LfgrE s.v. suggests ‘a lot’, despite showing that in Epos adverbial μακρόν and μακρά are confined to the spatial contexts ‘take long steps’ and ‘sound over a long distance’. The formula οὐδέ τί σε χρή+| (494*; Il. x9, Od. x9, H Dem. 82) elsewhere has a moral implication – ‘You should not be doing/intending that’. Apollo wrily ‘assumes’ that Hermes is responsible for his growth, as he was for controlling his body in order to feign innocence (239–40). 408 Κυλλήνιε Μαιάδος υἱέ: A vocative counterpart to Κυλλήνιος Ἀργεϊφόντης+|. Hipponax fr. 35 ἐρέω γὰρ οὕτω· ‘Κυλλήνιε Μαιάδος Ἑρμῆ’ might refer to an otherwise lost epic tag, though the bare genitive to express parenthood is unusual (145*). For repeated vocatives as a possible feature of conversation with children see 30–8*, 439*, Thomas (2010) 209–11.

409–13 409–10 ὥς ἄρ’ ἔφη, καὶ χερσὶ περίστρεφε καρτερὰ δεσμά | ἄγνου: Apollo’s urgency may be suggested by the omission of the usual participial clause which follows the seven other instances of ὣς ἄρ’ ἔφη καί in Epos, though this pattern is discontinued in Apollonius. χερσὶ περίστρεφε is ambiguous. It could mean that Apollo plaited ropes with his own hands (more likely than ‘for Hermes’ hands’); cf. 411* and Hp. Epid. 6.3.18 τὰ κλήματα τῆι χειρὶ ... στρέφων. Or he could be winding the ropes around Hermes with his hands; cf. Plu. Fac. Lun. 931a ὁ ἥλιος ... See 414, 440, 443, 455, and Introduction n. 101 for other referen­ces. θαυμαίνω occurs at Od. 8.108 θαυμανέοντες, where ΣBEH compared ὀνομαίνω ~ ὀνομάζω.

473

369

C O M M E N TA RY 4 0 9 – 4 1 0 κύκλον ... περιστρέφει περὶ τὴν σελήνην. Or he could be winding them around Hermes’ wrists; cf. Il. 15.19 περὶ χερσὶ δὲ δεσμὸν ἴηλα for the idea that χερσί may depend on περι- here. On any account, the thematic context encourages us to envisage Apollo actually binding Hermes.474 This brings at least partial fulfilment of Maia’s prediction that Apollo would apply ἀμήχανα δεσμὰ περὶ πλευρῆισιν (157). This correspondence, along with the imperfective aspect of περίστρεφε and plural δεσμά, and the intertextual relationships with H.Ap. and Hy. 7 (discussed below, 409–13*), all suggest that Apollo binds more than Hermes’ wrists; even if χερσί is taken as belonging to Hermes, it can mean ‘forearms’. The ropes are made of ἄγνος (Vitex agnus-castus, also called λύγος). Greek texts often mention this plant as growing near water (here, the Alpheus), and being a convenient source of cord in the wild (e.g. Il. 11.105, Od. 9.427 with ἐϋστρεφέεσσι, Longus 2.13.3 with στρέψαντες; compare λυγίζω ~ στρέφω). At S. fr. 25 Theseus ‘softens and spins’ λύγοι to make rope-chains (σειραῖα δεσμά) which are strong enough to drag the bull of Marathon. However, it is offensive and ineffective to bind Dionysus with λύγοι in Hy. 7.13, or Hera’s statue at Samos in the aition for the Tonaia.475 After all, as the helmsman observes, Dionysus is a θεὸς καρτερός (Hy. 7.17–18) requiring particularly καρτερά chains. Instances of gods being bound successfully do require stronger materials, and repeatedly mobilize the language of κράτος. At Il. 5.385–91 the Aloadai bind Ares in a bronze jar: κρατερός is a recurrent term (385, 386; also 392), but Hermes is able to extricate Ares stealthily. In the fragmentary Hy. 1, Hephaestus trapped Hera in an unfathomable chair. In the prologue of Prometheus Bound, Kratos himself helps to restrain Prometheus.476 At Od. 8.266–366 Hephaestus forges special chains to bind Ares and Aphrodite: they are κρατερά (8.336, 360), but also repeatedly described as a technical δόλος (276, 281–2, 297), which outdoes Ares’ powerful physique (329–32). Herm. responds to both motifs (binding a god, binding with vitex), and to the contest of κράτος and cunning which they repeatedly involve. There may be special engagement with Od. 8, considering the combination of κράτος, δόλος and bonds with the cameo of Apollo and Hermes (8.334–43).

474

Contrast e.g. Allen and Halliday, who thought the ropes were for the cows. Apollo may have bound Hermes in pre-existing versions of the story: see below, 409–13*. 475 Paus. 8.23.5, Menodotus FGrH 541 F1. Cf. Artemis Lygodesma’s statue in Paus. 3.16.7–11. These cult-myths of binding are well interpreted by Meuli (1975) with Graf (1979) 214–16; Von Staden (1993), esp. 46–51. 476 Kahn (1978: 97) suggestively related this to the parallels between Hermes and Prometheus in Herm. (Introduction §1.3); whether Kratos’ role in the myth was normal at Herm.’s date is unclear.

370

C O M M E N TA RY 4 1 0 – 4 1 2 410–11 ταὶ δ’ ὑπὸ ποσσὶ κατὰ χθονὸς αἶψα φύοντο | αὐτόθεν: The narrative is abrupt and elliptical, focalizing Apollo’s shock at Hermes’ instantan­ eous escapology. The ropes are at one moment in or around hands, at the next sprouting underfoot (Eitrem 1906: 275), and the agent of this change is not apparent. ταί follows δεσμὰ ἄγνου as if they had been called ἄγνοι (feminine as in Call.(?) fr. 756) – a constructio ad sensum which, though of a standard type, exacerbates the abruptness.477 Apollo’s ropes do not prove nearly as ἀμήχανα as Maia threatened (157). Hermes’ ability to release them from the inside complements his earlier manipulation of his swaddling, and his manufacture of sandals: see below, 409–13*. For the rerooting of plants Eitrem (1909: 333–4) compared the miraculous rerooting of Heracles’ club when in contact with the statue of Hermes Polygios at Trozen (perhaps < πολυ-λύγιος; Paus. 2.31.10); see also Radermacher 145–6, Thompson (1955–8) F960.1, 971. A slightly more general trope is nature efflorescing in response to divine presence, as at Il. 14.347– 9, E. Ph. 651–4; for an example in H.Ap. see again below, 409–13*. 411 ἐμβολάδην ἐστραμμέναι ἀλλήληισιν: ἐμβολάδην is a hapax whose general sense is secured by several cognates to do with grafting: see LSJ s.vv. ἐμβολάς, ἐμβολάδιον, ἐμβάλλω I.8; Gp. 10.77.4 ἔμβολον.478 ἐμ- governs the dative ἀλλήληισιν, as probably at E. Ba. 591 κίοσιν ἔμβολα. Hermes keeps the shoots winding around one another, so that they retain a sign of their brief life as a rope: ἐστραμμέναι picks up 409 περίστρεφε (particularly if that is taken as ‘plaited’), and ἐμβολάδην combines the way new fibres are successively inserted as a rope is plaited with the inosculation of neighbouring trees as they grow. The form of the regrowing shoots is thus described so as to hover indeterminately between a natural phenomenon and a rope, with even a hint of the agricultural technicality of grafting trees: this contributes to our sense of Apollo’s confusion. Though the particularities of the description are thus sufficiently motivated, Richardson may well be right in speculating that an original audience could have taken these lines as aetiological for an unusual feature of the landscape by the Alpheus, as at 124–6, 135–6, 415–16*. 412 ῥεῖά τε καὶ πάσηισιν ἐπ’ ἀγραύλοισι βόεσσιν: For the epic motif of divine ease in acting, see 396, 417–18*, LfgrE s.vv. ῥεῖα 2, ῥηΐδιος 2. For an adverb and a prepositional phrase coupled by τε καί see e.g. Th. 3.78.1 κακῶς

477

For such constructions see KG i.54–5. There is no compelling reason to posit a lacuna, as Baumeister did. 478 Clarke noted the likely sonic influence of Il. 21.364 παντόθεν ἀμβολάδην.

371

C O M M E N TA RY 4 1 3 τε καὶ κατ’ ὀλίγους προσπίπτοντες, Isae. 2.2 προσηκόντως τε καὶ κατὰ τοὺς νόμους, X. Mem. 1.2.10 ἀκινδύνως τε καὶ μετὰ φιλίας. The shoots grow rapidly over the cows, in a defiant display of Hermes’ active binding abilities and his claim to retain control over the animals; possible nuances within the latter are the use of ropes during the rustling of livestock (implied at E. Cyc. 225), and the lore according to which vitex could be used to protect controlled reproduction.479 Baudy’s interpretation that the shoots regrow as a fence (1989: 9) is incompatible with ἐπ’. 413 Ἑρμέω βουλῆισι κλεψίφρονος: For the traditional use of βουλῆισι and διὰ βουλάς + a god in the genitive, see LfgrE s.v. βουλή 1dα. LSJ translates κλεψίφρων as ‘with cozening mind’, but the natural interpretation is ‘who cheats people’s minds’. For κλέπτω νοῦν/φρένας see Il. 14.217, Hes. Th. 613, A. Ch. 854, E. Tro. 682, etc.; for this type of compound see Risch (1974) 191–3. Gregory of Nazianzus uses Θεοῦ κλεψίφρονι βουλῆι in a poem steeped in early epic phraseology (Poem. Arc. 8.13; use of Herm. is noted in Moreschini and Sykes 1997); the word is used elsewhere only of the planet Mercury’s influence in Ps.-Manetho Apotel. 1.93. In Nonnus D. 8.127 Hermes tries to use κλεψινόοις ὀάροισι. 409–413 κλεψίφρονος. The rope trick: With Apollo’s discovery of a further aspect of Hermes’ crime, there is a renewed prospect of him haling Hermes back to Olympus (ἀπαγωγή-style: 290*). This involves a hasty attempt to bind Hermes, which seriously underestimates his abilities. He renders it ineffective with abrupt, apparently magical, and irksome nonchalance. While sudden divine release from bonds is a topos (e.g. Od. 14.348–9 δεσμὸν μὲν ἀνέγναμψαν θεοὶ αὐτοί | ῥηϊδίως), the way in which a god performs it can be characteristic, as when at Hy. 7.12–15, 35–42 the pirates’ chains (again vitex) spring away from an effortless and smiling Dionysus, and subsequently vines and ivy grow over the mast and yard-arm while the sea becomes literally wine-dark.480 Hermes’ modus operandi involves him unbinding the shoots from himself but not from one another, making them reroot in the ground and grow rapidly into ropes around the cows. The main characterizing features are therefore mastery of binding, and management of the natural world (both plants and livestock). The former complements how Hermes wove shoots together to create shoes and wound his swaddling around himself (151–3, 237–42, 305–6). Indeed, at 79–85 Hermes pulled up ῥῖπες, plaited them, and bound them ὑπὸ 479

See the concern about reproduction at 493–5*, and Baudy (1989) and Von Staden (1993) for the lore. 480 How to account for this intertext (allusion, shared traditional motif, etc.) remains unclear. A light-hearted connection between Apollo and Etruscan pirates is certainly possible. On the contrast of the two scenes see Kahn (1978) 113–16.

372

C O M M E N TA RY 4 1 3 – 4 1 5 ποσσί, whereas here he unbinds vitex stems (which were used as ῥῖπες: Von Staden 1993) from himself but keeps them plaited, and replants them ὑπὸ ποσσί.481 Regarding Hermes’ swaddling, we recall Maia’s threat at 157 that Apollo would replace it with δεσμὰ περὶ πλευρῆισιν. When the vitex reroots, there are notable echoes of the intertwining bushes of Od. 5.476–81 (ἐξ ὁμόθεν πεφυῶτας ~ φύοντο αὐτόθεν here; πυκνοί | ἀλλήλοισιν ἔφυν ἐπαμοιβάδις), which complement the allusion to 5.488–91 in the description of Hermes wrapping himself in his swaddling at Herm. 237–42*. Hermes’ marvel is clear one-upmanship against Apollo, binding his cows when Apollo had tried to bind him. This sense is enhanced by two intratextual connections to the Hymn to Apollo (Introduction §1.3). Apollo is ‘bound’ in swaddling at H.Ap. 121–9, but bursts through the bands, leaving debris whereas Hermes skilfully recycles; Delos flourishes when Apollo first walks at H.Ap. 135–9, and here Hermes engineers the ‘plants flourishing’ motif to constrain Apollo’s animals in his own manifestation of divine power. The speed with which he escapes seems to surprise Apollo (410–11*), and may have overturned the audience’s expectations of the story. Yalouris (1953: 170–3) observed that hand-binding occurs in the possible depiction of Hermes and Apollo on the early sixth-century Corinthian crater Louvre E633; as noted on 155–61*, such an incident led to the etymologization of lyra from lytron (‘payment for release’), probably already in fifth-century sources.

413–25 413–14 αὐτὰρ Ἀπόλλων | θαύμασεν ἀθρήσας. τότε δὴ κρατὺς Ἀργεϊφόντης: Radermacher noted the opposition of Ἑρμέω and Ἀπόλλων at either end of 413. As recently as 403–7 Apollo was subject to catching sight of something unexpected and feeling θαῦμα. But this time Apollo seems to have run out of practical ways to express any frustration (a feature of epic θαῦμα discussed in Kahn 1978: 106–7), so that our attention reverts immediately back to Hermes. In this to-and-fro, the shift from κλεψίφρων to κρατύς returns us to the question of where κράτος lies (405, 409–10). 415–16 χῶρον ὑποβλήδην ἐσκέψατο, πῦρ ἀμαρύσσον | ἐγκρύψαι μεμαώς: I shall discuss the sense of πῦρ ἀμαρύσσον | ἐγκρύψαι first, then the difficulties surrounding ὑποβλήδην. The verb ἐγκρύπτω differs from plain κρύπτω and is frequently applied to firebrands buried in ash: Od. 5.488 (a passage in the hymnist’s mind: 237–8*, 409–13*), Ar. Av. 841, Arist. Juv. 470a16, Sotad. The sandals were θαυματὰ ἔργα, and Apollo will feel θαῦμα here too (414).

481

373

C O M M E N TA RY 4 1 5 – 4 1 6 Com. fr. 1.29, Thphr. Ign. 19. One should therefore not separate it from πῦρ, as Baumeister did by placing a lacuna after 415. The manuscripts have ἀμαρύσσων, in which case the fire is in Hermes’ eyes and his aim is to hide it within himself, presumably because it is psychologically revealing; the phrasing would be chosen to recall his actual covering over of embers in the same location at 140.482 However, ‘spotting/examining the place’ is not a way to conceal one’s looks. Lohsee’s emendation ἀμαρύσσον (1872: 48) removes the doubling of nominative participles and restores a logical connection: Hermes fears that Apollo, who was angry at seeing the skins, will be angrier still if he can see any remains from the fire, so he checks the place.483 The actual state of the fire would not determine Hermes’ nervousness, but in my view the embers are still visible (416*). One might object that this belittles Hermes’ attempt to cover them up at 140*. However, one might account for that, and for the implication of ἐγκρύψαι that the fire is to be preserved by being covered, through authorial motivation, if this is a further topographical aetiology – probably the start of an ash-altar (such as those of Zeus and Ge at Olympia: Introduction §2.2). ὑποβλήδην here does not match the word’s other uses. The only prior occurrence (Il. 1.292) modifies ἠμείβετο and was generally taken as ‘interrupting’ (e.g. Dionysius of Sidon in Ap.S. pp. 156–7) or ‘replying’ (Hsch. υ.575); later uses are mostly replies, often with a proposal (ὑποβολή), and occasionally interrupting.484 ὑποβάλλω is used of looking at Plu. Curios. 522a, but of looking up under the brows; such looks, with the head tilted down, can be supplicatory (e.g. Crinagoras 50.3 GP) or confrontational (Cairns 2005) but neither attitude is pertinent here. This suggests an emendation, subject to the need to explain how such a rare word entered the tradition by error. Passow’s ὑποβλέβδην (1823: 949) entails the unusual sense ‘look furtively’ for ὑποβλέπω (as in Plu. Curios. 521b); no other adverbs like βλέβδην seem to be found. One might try ὑποκλέβδην, resting on the similarity of minuscule κ and β; κλέβδην is noted in Et.Gen. α.832 and Σ D.T.

For ἐγκρύπτω ‘hide within (sc. oneself)’ see D.S. 10 fr. 26 ὁ φθόνος ἐγκρυπτόμενος, ἐπειδὴ καιρὸν ἔλαβεν, ἅθρους ἐξερράγη. For the nuances of fiery looks see 45* (including the line-end at Hes. Th. 826–7 ὄσσε ... πῦρ ἀμάρυσσεν), 278*; on any interpretation the present passage recalls Hermes’ scheming during those. 483 Adorjáni (2013a) denied that Hermes has a reason to hide the embers, and proceeded to an interpretation of the textual problem which involves Hermes’ looking guilty in order to tempt Apollo into enquiring why, so that Hermes can reveal the lyre. 484 A.R. 1.699 may mean ‘interrupting’ the din. 3.400, 1119, Q. S. 2.147, Colluth. 147 are non-interrupting replies with a proposal. Dios ap. Stob. 4.21a.17 is a reply without proposal or interruption. Quite separate is Ps.-Manetho Apotel. 6.292 ‘suppositiously’. See also Rengakos (1994) 147–8. 482

374

C O M M E N TA RY 4 1 6 – 4 1 7 p. 96. Alternatively παρβλήδην could suggest a furtive sidelong glance (see 279*, A.R. 3.444–5, LSJ s.v. παραβάλλω IV), though παραβλήδην is not found in this sense; the corruption to ὑπο- would require little more than an ink-blotch. 416 Λητοῦς δ’ ἐρικυδέος υἱόν | < … >: Here, unlike after 415, we do need a lacuna. A mention of the lyre is needed to accompany 418 λαβών, which is secured by the parallel phrasing at 499.485 This could perhaps be achieved by assuming a one-line lacuna between 418 ἐόντα and λαβών, with a sense such as ‘For he revealed the lyre’, or by Hermann’s lacuna after 418. However, if I am correct that 415–16 refer to the smouldering traces of Hermes’ sacrificial fire, this fire and Hermes’ eagerness to keep it concealed can hardly be raised then dropped so abruptly. Hence a lacuna after 416 seems more likely. The context suggests the following drift, perhaps across two lines: ‘But the son of Leto did see the embers, was angry, and began to intervene physically again. However Hermes, revealing the lyre under his swaddling, …’ I suspect that 416 actually ended with nominative υἱός, and so continued the rapid alternation of subject in 413–14; after the lacuna arose, the ending would have been adapted to the grammar of 417. Hermes’ swaddling may well have been mentioned, and for the last time as he moves away from his mother’s sphere towards an Olympian position (cf. Vergados 2011: 23). 417–18 ῥεῖα μάλ’ ἐπρήϋνεν Ἑκηβόλον ὡς ἔθελ’ αὐτός, | καὶ κρατερόν περ ἐόντα: Against Apollo’s physical power (see κρατερόν), and the belligerent nuance of ἑκηβόλος (218*), Hermes is once again ‘easily’ able to cope ‘as he himself wanted’; ὡς ἔθελ’ αὐτός also overturns an assumption that someone may be constrained at Od. 20.136 (Odysseus the beggar drinking). The clear echo of ῥεῖα from 412 constructs a parallelism between this passage and that first response to Apollo’s anger and κράτος, the vitex-marvel. For the power of music against anger see e.g. Chamaeleon fr. 4 εἴ ποτε συνέβαινεν χαλεπαίνειν αὐτὸν [Cleinias the Pythagorean] δι’ ὀργήν, ἀναλαμβάνων τὴν λύραν ἐκιθάριζεν· πρὸς δὲ τοὺς ἐπιζητοῦντας τὴν αἰτίαν ἔλεγεν ‘πραΰνομαι’, Aristoxenus fr. 122 (lyre-music off-setting the heat of wine at a symposium), A.R. 1.492–5 (Orpheus sings just as a fight is being 485

M has λαβὼν δ’ ἐπ᾽ ἀριστερὰ λύρην, where unmetrical λύρην must be an intrusive annotation. It is not a palaeographically plausible correction for λαβών, even without 499 to guide us; and Kämmerer (1815) pointed out that λύρης would be normal with πειρήτιζε. I take λύρην as a gloss on the object which had to be supplied in an already corrupt text. Schneidewin’s suggestion (1848: 683), that ἄθυρμα ss. λύρην became -α χειρός ss. λύρην in Ω by intrusion of χειρός from 419, seems more ingenious than plausible.

375

C O M M E N TA RY 4 1 8 – 4 2 1 averted: Introduction §4.2), Ath. 14 627e τὴν αὐθάδειαν πραΰνει (sc. ἡ μουσική), Aelian NA 12.44 (Indian lute calming elephants), ΣbT Il. 9.186–91. Richardson suggested that the pacifying power of Hermes’ performance is specifically related to Hes. Th. 98–103, the classic statement of a bard’s power to soothe cares. 418 λαβὼν δ’ ἐπ’ ἀριστερὰ χειρός: Text: 416*. Construction of χειρός: n. 204. 419–20 πλήκτρωι ἐπειρήτιζε κατὰ μέλος, ἣ δ’ ὑπὸ χειρός | σμερδαλέον κονάβησε: = 53–4*. Here there seems at first to be a touch of Apollo’s focalization in σμερδαλέον; but the subsequent words immediately complicate this impression of fear. Unlike in Od. 17.542, where σμερδαλέον κονάβησε, γέλασσε δέ ... separates Penelope’s response to Telemachus’ sneeze from the foreboding which it brings for the suitors’ future, here σμερδαλέον acts more as a foil, raising an idea of a fearsome clash of arms only to overwrite it immediately with a spirit of reconciliation. 420–1 γέλασσε δὲ Φοῖβος Ἀπόλλων | γηθήσας: The fourth laugh of the hymn. Like 281, this laugh signals the start of the end of a phase of Apollo’s anger. γηθέω specifies precisely this replacement of tension with delight.486 However, the laughs also form a chiasmus: as 389* was related to 281 (Zeus and Apollo laughing at Hermes’ protestations), 420 is related to 29: Apollo’s laugh at the lyre confirms the visionary nature of Hermes’ laugh at the tortoise as something which would be an effective bargaining tool with his brother. 421–2 ἐρατὴ δὲ διὰ φρένας ἤλυθ’ ἰωή | θεσπεσίης ἐνοπῆς: ἐρατή initially seems to be a fairly traditional description of the sound as ‘lovely’ (see LfgrE s.v. ἐρατός 3, and the Muse Erato). It will turn out, however, to be the first indication of Apollo’s remarkably eroticized response to the lyre, which recalls and clarifies Hermes’ imaginative address to the tortoise in 31* φυὴν ἐρόεσσα.487 Such language bridges Apollo’s musical desires (for further music,

486

See the initial summary in LfgrE s.v.: ‘ein Freudegefühl das auf Abwesenheit oder Behebung von Angst/Spannung, bzw. auf Erfüllung einer Hoffnung beruht.’ Similarly at Il. 6.212 Diomedes γηθεῖ to find that he will not have to fight his xenos Glaucus, without any indication that he had conceived such a hope beforehand. The verb recurs in Herm. at 499* γηθήσας. 487 The connection of 420–1* with 29–30 lays the groundwork for an audience to appreciate this connection. For the erotics of Hermes’ lyre see Introduction §3.4, and Kloss (1994) 79–82 for its non-traditional nature.

376

C O M M E N TA RY 4 2 2 – 4 2 3 to acquire the lyre, to learn citharody) and the sexual desire which might be aroused by the lyre-music of a hetaira at a symposium. ἔρος is the latest incapacitating factor which Hermes aims against Apollo’s anger, following humour and astonishment. The hymnist modifies the formula περὶ φρένας ἤλυθ’ ἰωή (Il. 10.139; Od. 17.261 of lyre-music) so that the sound pierces Apollo, as Hermes loves to pierce (Introduction §3.3.6); in Epos, διελθεῖν normally refers to weapons.488 The lemma therefore continues the ambiguous representation in 420–1 where the lyre’s sound was σμερδαλέον but caused laughter and delight: it is the delightful instrument of the world of Eros, but has an irrational frightening facet derived from its newness and conquering power. So too ἰωή applies to flames and winds as well as to sounds (see Danek 1988: 91–7), the main senses of θεσπέσιος are ‘godinspired’ (esp. of music) and ‘preternatural’ of the din of war (see Dieu 2013), and ἐνοπή can be musical (e.g. 512, Il. 10.13, E. Ion 882) but often describes the clamour of war. There may be a specific recollection (by the poet, the audience and Apollo) of the other epic instance of the collocation θεσπεσίη ἐνοπή, of the Python’s death-rattle at H.Ap. 360: the lyre is also the ‘voice-in-death’ of an animal (38) which is ‘taking on’ Apollo, but a lovely one (Introduction §1.3). 422 καί μιν γλυκὺς ἵμερος ἥιρει: The mixture of eroticism and power continues: unqualified γλυκὺς ἵμερος otherwise refers to sexual desire in Epos, and even when qualified the formula can colour a desire as a displacement of a sexual one, as at Il. 3.139 (Helen’s nostalgia supplants desire for Paris), Od. 22.500 (Odysseus’ desire for weeping when his reunion with Penelope is being delayed). Such desire gradually (imperfectively) ‘seized’ Apollo with its power.489 If my reconstruction at 416*–17 is right, Apollo is seized when about to seize Hermes. ‘The enchantment of Hermes’ music binds Apollo more tightly than any chains’ (Clay 1989: 138). 423 θυμὸν ἀκούαζοντα: West’s θυμόν (1966a: 149) is an accusative of respect appended, as often happens with body-parts, to the object of ἥιρει. Ω᾽s θυμῶι would be instrumental with ἀκουάζοντα. Since the θυμός is not normally a sense-organ, the juxtaposition with ἀκουάζοντα would indicate an interaction For ‘piercing’ sounds cf. the conventional metaphors of τορός and ὀξύς, and the more lively ones of e.g. A. Th. 563 ἱκνεῖται λόγος διὰ στήθεων, Ch. 380–1 διαμπερὲς οὖς ἵκεθ’ ἅπερ τε βέλος, 451 δι’ ὤτων δὲ συντέτραινε μῦθον. For desire being incapacitating see further 434*. 489 Similarly Od. 8.304 χόλος... ἥιρει| (anger increasingly took hold of Heph­aestus as he summoned the other gods). Contrast the more sudden presentation at H Aphr. 57 ἵμερος εἷλεν|. 488

377

C O M M E N TA RY 4 2 4 – 4 2 5 of sound and emotion. Similarly at Od. 1.353 (ἐπιτολμάτω κραδίη καὶ θυμὸς ἀκούειν), where Penelope’s emotions are particularly stirred by Phemius’ song. Here the effect on Apollo’s θυμός is very different, and without better parallels I prefer to emend. Either way, one can contrast Apollo’s holistic response here and in 434 with the way Hermes’ θυμός dampened his ἵμερος at the smell of meat (131–3); for other indications to compare the two passages see 425–33* (at n. 501), 436*, 454*. λύρηι δ’ ἐρατὸν κιθαρίζων: For the instrumental dative, cf. Il. 18.569–70 φόρμιγγι λιγείηι | ἱμερόεν κιθάριζε.490 λύρη only occurs here in Epos, but already at Archil. fr. 93a; indeed, cf. λυρασταί in Mycenaean Thebes (Av 106 ru-ra-ta-e). On the variety of words for ‘lyre’ see 17*. D’Orville (n.d.) f. 253 entertainingly but misguidedly commented ‘vox ἐρατὸν facit inamabilem sonum totiens repetita’. The repetition (also in 426) marks thematic importance, and words repeated within a short compass are characteristic of this section: Introduction §3.2. 424–5 στῆ ῥ’ ὅ γε θαρσήσας ἐπ’ ἀριστερὰ Μαιάδος υἱός | Φοίβου Ἀπόλλωνος: θαρσήσας shows that, though the lyre turned out to calm Apollo easily, Hermes was still nervous until he saw the laugh of 420. He then moves within reach of Apollo, whose power he previously had to avoid (e.g. 235–7, 409–13), while the poet juxtaposes the gods’ formulas – neighbours still separ­ ated by a line-break.491 For the point of ἐπ’ ἀριστερά see 454*.

425–33 425 τάχα δὲ λιγέως κιθαρίζων: λιγύς suggests a clearly audible tone, and applies formulaically to the φόρμιγξ (e.g. Il. 18.569–70, cited on 423*). It is often attributed to female voices, but also to Nestor, so that the issue is probably not one of pitch (Il. 1.248; Kaimio 1974: 33). The adverb is rare, and suggests the intriguing possibility of influence on Theodoros II Laskaris Epist. 74 (to his teacher Georgios Akropolites, c. 1250) ἡμεῖς δ’ ὡς Ἑρμοῦ τέκνα κιθαρωιδοῦντες λιγέως ἅμα καὶ ἄιδοντες. Laskaris certainly alludes to Il., Od. and Batr., particularly in Epist. 51.

For the textual problem of whether H.Ap. 515 ended ἐρατὸν κιθαρίζων (so M), see Introduction n. 27. 491 Compare the juxtapositions in 239*, 314*, 318*, 327*, and contrast 413–14*. If books of hexameter hymns were ever written without line-breaks, Μαιάδος υἱός would ‘stand on the left of ’ Φοίβου Ἀπόλλωνος. 490

378

C O M M E N TA RY 4 2 6 – 4 2 8 426 γήρυετ’ ἀμβολάδην: Varying formulaic αὐτὰρ ὃ φορμίζων ἀνεβάλλετο καλὸν ἀείδειν (Od. 1.155, 8.266); cf. 17.262 ἀνὰ γάρ σφισι βάλλετ’ ἀείδειν, just after ἤλυθ’ ἰωή, which we had in Herm. 421. Those passages, taken on their own, are compatible with ἀναβάλλομαι simply meaning ‘begin’, but ἀμβολάδην more technically means ‘in a manner involving an ἀναβολή’, a proemic section. Of extant uses, Hsch. α.3518 ἀμβολάδην: ἀναβολῆι χρώμενος seems best suited to our passage.492 ἐρατὴ δέ οἱ ἕσπετο φωνή: Either ‘lovely was his voice which followed (the lyre)’ (cf. Il. 18.572, Pindar P. 1.3) or ‘lovely was the voice which accompanied him’ (LfgrE s.v. ἕπομαι 4b). The latter will be supported by 440 σοί ... ἕσπετο θαυματὰ ἔργα, referring to lyre-music. 427–8 κραίνων ἀθανάτους τε θεοὺς καὶ Γαῖαν ἐρεμνήν | ὡς τὰ πρῶτα γένοντο καὶ ὡς λάχε μοῖραν ἕκαστος: κραίνειν elsewhere, including in Herm. (531, 559), means to ‘bring about fulfilment of ’ (e.g. Il. 5.508) or more often to ‘ordain fulfilment of, ratify’ something for another: Benveniste (1969) ii.35– 42. This sense has been thought impossible here since the gods pre-exist and are not brought to fulfilment by Hermes’ song. Indeed, it is for this reason that our line is the most likely source of Hesychius’ equations of κραίνω and τιμάω (κ.3922–4). But rather than inventing a separate meaning, we should attempt to see the stretch of normal usage as marked and significant.493 The stem relates to prophecy (e.g. 559, A. Ag. 1255, E. Su. 139, Ion 464), and its use here sets up the hymnist’s maverick alignment of the power of Hermes’ music with the seemingly rather different efficacy of Apollo’s utterances. See further 425–33* below, Introduction §3.4. The epithet ἐρεμνήν elsewhere in Epos and Hellenistic literature has full semantic force when applied to the land: it is used of caves at Hes. Th. 334, earth qua limit of Underworld at Od. 24.106 and A.R. 3.642, sunset at 3.1191, and shadowy islands at Simias fr. 1.7. This hints that Hermes presented For the dithyrambic ἀναβολή see Comotti (1989). Egan (2006) documented the musical uses of ἀναβολή etc., but his interpretation that non-dithyrambic ones basically refer to resuming song is barely suitable in the passages in Od., Herm. and Pindar, and only persuasive for Nonnus. Châteillon first approached the correct interpretation here with exorsus; cf. exordior of a hexameter proemic hymn at Cic. Resp. 1.36. 493 Some have argued that the sense here and in Od. 19.567 ἔτυμα κραίνουσιν and Emped. 31 B111.5 is ‘declare with authority’, separate from any sense of fulfilment (Allen and Halliday, Zumbach 1955: 46–9). But the dreams in Od. 19 will be fulfilled, and Empedocles’ speaker is promising to bring about special powers for his addressee. Detienne claimed that Hermes’ song ‘makes [the gods] real’ here (1996: 70–5, citation from 71) but did not explain how they lacked reality before. 492

379

C O M M E N TA RY 4 2 9 – 4 3 1 Gaia in terms of the dark landscapes of his birth and theft. According to Hesiod, at least, Gaia is Hermes’ father’s grandmother and mother’s greatgrandmother. On the level of the hymnist’s performance, Gaia’s prominence could be motivated by her cult at Olympia (see Introduction §2.2). Our lines resemble Hes. Th. 108–12 εἴπατε δ’ ὡς τὰ πρῶτα θεοὶ καὶ γαῖα γένοντο ... καὶ ὡς τιμὰς διέλοντο, with the other cosmic features of 109–10, alongside which Gaia generally appears, omitted so as to focus on her theo­ gonic role as παμμήτειρα (as in Hy. 30). Moreover, the closest epic parallel for λάχε μοῖραν is Th. 203–4 λέλογχε | μοῖραν.494 429–30 Μνημοσύνην μὲν πρῶτα θεῶν ἐγέραιρεν ἀοιδῆι | μητέρα Μουσάων· ἣ γὰρ λάχε Μαιάδος υἱόν: We were primed in 426 to identify a proemic hymn; the repetition here and in 432 of 60 ἐγέραιρε connects the hymnic quality of this part of the second song with Hermes’ earlier hymn-like performance (52–61*). Hermes turns out to address his hymn to Mnemosyne – a marked choice. Line 430 suggests two preliminary reasons for it (discussed further below, 425–33*): Mnemosyne is the mother of the Muses, so Hermes needs her assistance qua experientially limited performer of divine history; then, picking up 428 ὡς λάχε μοῖραν, comes the claim (endorsed by the hymnist or channelled through Hermes’ indirect speech?) that Hermes is part of Mnemosyne’s allotted sphere of influence. For a god ‘receiving’ someone through an allotment of τιμαί compare Call. H.Ap. 43 κεῖνος ὀϊστευτὴν ἔλαχ’ ἀνέρα, κεῖνος ἀοιδόν, Pi. O. 8.15–16 ὔμμε δ’ ἐκλάρωσεν πότμος | Ζηνί. Here and at 432 Διὸς ἀγλαὸς υἱός Hermes is referred to by means of his parentage – a usage energized by the context of a theogonic song; in this line there is also an elegant balance with μήτερα Μουσάων. 431 κατὰ πρέσβιν: ‘by seniority’ – an ordering principle which may correlate largely with age (as perhaps at Pl. Lg. 9 855d, 11 924c), but need not (see e.g. πρεσβύτατος of Poseidon at Od. 14.142). Hesiod’s Theogony contains traces of both principles, and others: the Muses’ theogony at Th. 11–21 starts with Zeus and Hera, before descending (in power but not age) to Gaia, Okeanos, Night, and the rest; by contrast, 886–944 treats Zeus’s family in largely chronologic­ ­al order. See further 425–33* below. For the uses of πρεσβ- in Greek hymnography see Keyssner (1932) 72–5. καὶ ὡς γεγάασιν ἕκαστος: The perfect tense indicates that Hermes described the gods’ innate characteristics, the aorist in 428 that he narrated their births. 494

These parallels will be discussed further in 425–33* below, along with the connection of λάχε μοῖραν to 128–9. For ὡς τὰ πρῶτα when a narrative about the gods begins at the beginning see also Od. 8.286, H.Ap. 214.

380

C O M M E N TA RY 4 3 3 433 πάντ’ ἐνέπων κατὰ κόσμον: For κατὰ κόσμον in an aesthetic judgement about poetry see Od. 8.489. Early sources suggest the semantic hub of κόσμος was a complex of pleasingly ordered parts, and that applications to poetry may touch on its ornamental language, its elevation of a topic, or the IndoEuropean metaphor of poetic ‘construction’.495 Here the intratext to 255–6 (see 438*) suggests that κατὰ κόσμον could also be taken as ‘matching the cosmos’, i.e. accurate to the theogonic topic. ὑπωλένιον: Ω had ἐπωλένιον, but ὑπωλένιον in the same context at 510; neither word is used elsewhere about lyre-playing. While ἐπώλενιον probably corresponded better to what medieval scribes envisaged (and is thus the lectio facilior), art shows that lyres normally hung by a strap from the lower left forearm, which was raised so that the hand could manipulate individual strings (West 1992: 65) – exactly as ὑπωλένιον implies. I therefore prefer ὑπωλένιον even though Hermes has not explicitly added a strap to his instrument. 425 τάχα–433 Hermes’ Theogony: Whereas Hermes’ first song (52–61*) tailed off as soon as he had exhausted its domestic content, his second topic is more interesting for Apollo, and allows Hermes to express his acquaintance with the broader Olympian family and his μοῖρα within it (428). These lines also offer a deeper reflection on the properties of hymnic and theogonic poetry. Jaillard’s interpretation is particularly stimulating: (2007) 199–205, 223–5; (2012) 292. Hermes sings of Mnemosyne, then of how the gods and Gaia were born, of how they were allotted prerogatives, and of their inborn characters. This is a theogony with a proemic hymn which engages with Hesiod’s Theogony in particular (cf. Introduction §1.3). The concept of a pacifying performance may allude to Th. 98–103 (see 417–18*); 427–8* ~ Hes. Th. 108–12 and 203–4 λέλογχε | μοῖραν. The possible meanings of 431* κατὰ πρέσβιν can best be imagined by considering the multiple ordering-principles of the Theogony. Shelmerdine (1984: 207) noted that the emphasis on an allotment of prerogatives suits Hesiod’s trajectory (see Th. 855, immediately before the shift to cataloguing Zeus’s wives) but may not have been normal in other theogonies. Harrell (1991: 327–9) added that Hesiod narrates a series of divine power-struggles

On κόσμος see esp. Kerschensteiner (1962) 5–10. Poetry ornamental: Hy. 7.59, Solon fr. 1.2, Parmenides 28 B8.52 (paradoxically derogatory), Pindar fr. 31 (the Muses κοσμήσουσιν Zeus’s works), E. fr. 773.44. Poetic construction: e.g. Pindar fr. 194, Democritus 68 B21, Aristophanes Ra. 1005; Od. 8.489 is juxtaposed with 492 ἵππου κόσμον, and refers to Demodocus’ song at 8.75–82, which seems from the summary to have a non-sequential narrative, like Od. itself. By contrast, οὐ κατὰ κόσμον of speech seems to refer to its being ‘out of order’ socially (Il. 2.213–14 Thersites, Od. 8.179 Euryalos, 14.363, 20.181, Il. Parv. fr. 2).

495

381

C O M M E N TA RY 4 2 5 – 4 3 3 which contrast with the resolution to be achieved by Hermes and Apollo, not least by means of this appealing theogonic song. Finally, Hermes’ theogony is headed by a proem to Mnemosyne, while Hesiod’s begins with a ‘Homeric’ hymn to her daughters, a point given emphasis when she is here (non-formulaically) called ‘mother of the Muses’ (430). Herm. thereby constructs a ‘future reflexive’ allusion: Hesiod’s genealogy is a generative source for the hymnist’s, who however uses genealogy itself to place Hermes’ genealogy before Hesiod’s.496 The hymnist retains from Hesiod’s proem the self-aware focus on the source of his ability (Th. 22–34, 94–101) and a mise-en-abyme structure (inset hymns and theogonies at Th. 3–4, 11–21, 36–52, 66–75); the self-reflexivity of Hermes’ first song also primes us to look for self-reflexivity here. Much as Pindar speaks in a proem to the lyre of how προοιμίων ἀμβολὰς τεύχηις (P. 1.4), the composer of a proemic hymn to Hermes speaks of him performing an ἀναβολή to Mnemosyne.497 For the hymnist, this selection of Mnemosyne not only contributes a connection to Hesiod, but gains relevance from the formulaic use of μνήσομαι by the composers of H.Hom. to describe their work as performers of divine history (580*; cf. H.Ap. 1, 150, 160). As an aspect of Hermes’ self-presentation, however, the justification of starting with Mnemosyne because she was allotted tutelage of him obviously does not hold water. While heralds and messengers did need accurate memories, Herm. foregrounds instead the in­genuity which affords Hermes knowledge without memories.498 Hermes’ proemic choice therefore comes across as tendentious self-presentation. It allays Apollo’s fear about his independent-mindedness, much as the references to a definitive allotment of roles (428) and to the gods’ inborn characters (431) paper thinly but diplomatically over Hermes’ ability to alter Apollo’s character during the hymn (Introduction §3.4) while creating new prerogatives. Hermes may also choose Mnemosyne to lay a precocious claim to an educative function while he advertises for Apollo’s first lyre-lesson (to be paid for in kine). Hermes and Mnemosyne were certainly found together in later school-rooms; in the classical period, lyre-playing and literacy were two of the main lessons, with writing becoming increasingly prominent and

496

Barchiesi (1993) introduced the term ‘future reflexive’. For genealogy as a symbol of allusivity see Introduction §1.3 on H.Ap. and H.Pan. For the affiliation of the proem of Theogony with H.Hom. see Introduction n. 51. 497 I discuss whether Herm. was actually proemic in Introduction §2.3. 498 See IG I3 776 (early fifth century) ℎερμεί[αι τόδ’] ἄ̣γαλμα [διδὸς] χάριν ἐν̣[θάδ’ ἔ] θεκεν | Οἰν̣[…]ς κέρυχς, μ̣[νεμ]οσύνες ℎέ̣[νεκα. Hermes granted his son Aithalides, the Argonauts’ herald, an infallible memory (A.R. 1.640–9).

382

C O M M E N TA RY 4 2 5 – 4 3 3 increasingly connected to Hermes.499 This reading of Hermes’ choice of proemic addressee suggests that not only Hermes and Hesiod but the hymnist too uses his proem to construct (tendentiously) a relationship of reciprocity with a tutelary musical deity. After the proem, Hermes proceeds to a theogony focusing on the gods’ births, characters and prerogatives, organized κατὰ πρέσβιν. While the latter is flexible (431*), it may well suggest that Hermes’ song ends with something like a (p)rewrite of H.Ap., then with Hermes himself as the most junior yet in some sense climactic god.500 Hermes’ representation of the gods’ allotted μοῖραι recalls 128–9* μοίρας κληροπαλεῖς, but is more nuanced than the symbolism of his butchery, in that it is hierarchical rather than egalitarian, and includes more figures (e.g. Gaia) than a cultic Dodecatheon.501 Most importantly, Hermes’ putative description of his own prerogatives preempts the reconciliation with Apollo which the song is itself moving forward. This is one area in which Hermes’ representation of his position on Olympus could be said to have a quasi-prophetic efficacy in bringing about future changes (427* κραίνων). Hermes’ song, including an account of his own μοῖρα, advertises lyre-playing to Apollo and thus causes the swap of the lyre for prerogatives, i.e. it brings about the final state of his μοῖρα by describing it (Clay 1989: 138). The process lies at the intersection of two of the facets of κατὰ κόσμον noted on 433*, namely intelligent construction of a poetic representation, and matching the cosmic order.

499

For memory and writing see e.g. Mnemosyne holding a scroll on LIMC s.v. 1 (Attic, c.450), or PV 789 ἐγγράφου σὺ μνήμοσιν δέλτοις φρενῶν. For writing and Hermes see 75–8*, and e.g. Plato Phdr. 275a, where Thoth – equated with Hermes already in Hdt. 2.138 – is rebuked for inventing letters, ‘a medicine not [sc. as commonly assumed] for memory but for reminding’. A lekythos of c.470 by the Cartellino Painter depicts a boy reading Ἑρμῆν ἀείδω (not necessarily Hy. 18) from a scroll, suggesting both the use of hymns as school-texts and an early educational function for Hermes: Sotheby’s New York, 8 December 2011 lot 1, discussed in Beazley (1948). There is a school-text papyrus of Hy. 1 from c.100 bce: Cribiore (1996) 233. Later, Arrian Cyn. 35 speaks of educators typically praying to the Muses, Apollo, Hermes and Memory. On Hermes’ choice of Mnemosyne cf. Romani (2011) 130–2. 500 See Shelmerdine (1984) 207 n. 22, Baudy (1989) 12. I say ‘in some sense climactic’, since Greek catalogues do not always end climactically by any means. Clay (1989: 139) asserted that Hermes’ birth brings the Olympian order to completion. May Pan not be offended! 501 Burkert (1984: 841) interpreted the parallelism between the second song and the butchery as suggesting that theogonic myth is the verbal justification for sacrificial cult. Vergados 6 n. 10 usefully compared the metaphorical feast of hymns in Pindar fr. 52f.127–8 οὔ σε παιηόνων ἄδορπον εὐνάξομεν.

383

C O M M E N TA RY 4 3 4 As with Hermes’ proem, we can interpret the surprising ‘ratificatory’ quality ascribed to his theogony as a self-reflexive hint, in this case to the efficacy of hymnic performance. Greek hymns combine representation and per­ locution by using narrative (often mentioning the addressee’s benevolence) to elicit further benevolence, in which case the narrative’s truth becomes self-fulfilling. The comparison of Hermes’ song and butchery as media for representing the divine order would extend this consideration to Herm.’s own probable setting at an event including sacrifices at Olympia (Introduction §2). The implication that Hermes juxtaposes a tendentious version of Apollo’s birth and prerogatives with his own replicates Herm.’s self-promoting use of the Hymn to Apollo (Introduction §1.3). This matches the point made above about Hermes’ choice of Mnemosyne in his proem, namely that hymnists can construct their addressees so as to contribute to their self-presentation: see Introduction §3.4 on what Hermes’ lyre-playing implies about Herm.’s claim to a more playful and neoteric brand of hymnody, and the role of the allusions to the Hymn to Apollo in this.

434–62 434 τὸν δ’ ἔρος ἐν στήθεσσιν ἀμήχανος αἴνυτο θυμόν: The shift from ἐρατός to ἔρος intensifies the theme started in 421–2*, since the noun applies to music considerably less often than the adjective in archaic and classical sources.502 The ‘sexiness’ of Hermes’ theogony is counter-intuitive genre-bending which characterizes him as a musician: see further on 454*. Before, Apollo was πολυμήχανος (319) against Hermes’ μῆτις, and threatened to apply ἀμήχανα bonds (157, 257). The bonds (410–13) were outdone by Hermes, who then revealed a new μηχανή (cf. 436 μηχανιῶτα) which transmutes Apollo’s ἀμηχανία from frustration to desire (see 447*; Kahn 1978: 112–13). ἀμηχανία is regularly associated with desire, e.g. Il. 15.14 of Hera with reference to her deception of Zeus, Hes. Th. 589 ~ Op. 83 of Pandora, Sappho fr. 130, E. Hipp. 598. Apollo’s loss of control as an agent is emphasized by αἴνυτο θυμόν, which modifies the formula for death ἐκ δ’ αἴνυτο θυμόν| (Iliad x4), and so once more is considerably more intense than the apparently similar 422 ἵμερος ἥιρει.

At Il. 13.636–9 Menelaus ridicules the Trojan ἔρος of μολπή (among other things). E. fr. 192, where Amphion declares that time (i.e. practice), inspiration, and ἔρως ὑμνωιδίας are the sources of his song, expresses an idiosyncratic level of investment. ἵμερος for music is also unusual; at Od. 23.144–5 it applies to music for a fake wedding.

502

384

C O M M E N TA RY 4 3 5 – 4 3 6 435 καί μιν φωνήσας ἔπεα πτερόεντα προσηύδα: A common formulaic line (Il. x17, Od. x29, H Dem. 320); spoken words are probably ‘winged’ because they travel swiftly through the air – this at least was a common ancient interpretation: see LfgrE s.v. πτερόεις with references to the large body of scholarship on this point. 436 βουφόνε, μηχανιῶτα, πονεόμενε δαιτὸς ἑταῖρε: Apollo’s full-line address encapsulates his responses since 403: disapproval at cow-killing; wonder at lyre-engineering; recognition that these actions relate Hermes to dining and after-dinner entertainments, respectively. As often, the series of vocatives suggests the speaker’s difficulty in describing his addressee. But whereas such series are generally either derogatory (e.g. Il. 1.149, 3.39, 13.824) or respectful (Wendel 1929: 49–51), Hermes forces Apollo to blend the two. The vocatives not only respond to Hermes’ past actions but suggest epithets for general future activities, because of Apollo’s foresight, and the principle that Greek vocatives describing an activity normally have a general scope (190*). Apollo thus appears to recognize Hermes’ future as a patron of sacrificial officiants, craftsmen and domestic slaves. In βουφόνε, as in 405* δειροτομῆσαι, Apollo’s tone is likely to bring out the murderous aspect of the sacrificial, as in Burkert’s analysis of the Attic Bouphonia/Dipolieia (1983: 136–43). However, literary parallels for βουφονdo not make so much of this (Il. 7.466 |βουφόνεον, PV 531, Call. Aet. fr. 67.6), and counsel against connecting Herm.’s use specifically to the Bouphonia.503 Since the word makes a narrower point than μηχανιῶτα I take it as a separate noun. μηχανιῶτα is a hapax: Apollo engineers a new word to describe Hermes’ engineering (of the cows, ropes, lyre, situation). For the importance of the stem μηχαν- see 434*. For the light-hearted ending -ιῶτα see 294–6*. The echo of 301 σπαργανιῶτα underlines the focus on binding among Hermes’ schemes (Introduction §3.3.6). Dining has close formulaic ties to πονέομαι and especially πένομαι, so that πονεόμενε δαιτὸς ἑταῖρε belong together.504 Besides correcting νυκτὸς ἑταῖρε (290), as Apollo connects his greater knowledge about Hermes’ nocturnal activities (the butchery) to the invention of the lyre, the phrase figures Hermes simultaneously as a domestic servant who prepares the meat

503

Contrast Nobili: ‘rimanda chiaramente al rituale ateniese’ (2011: 167). Contrast e.g. Ernesti, Càssola. δαῖτα πένοντο| is formulaic, and see Il. 24.444 περὶ δόρπα φυλακτῆρες πονέοντο, Od. 17.258 παρὰ μὲν κρειῶν μοῖραν θέσαν οἳ πονέοντο, A.R. 2.263 πεπονήατο δαῖτα, LSJ s.v. πονέω B II.2b.

504

385

C O M M E N TA RY 4 3 7 at δαῖτες and a ἑταῖρος who participates (one of the πόσιος καὶ βρώσιος ... ἑταῖροι of Theognis 115) – a figure still balanced between the brat whom Apollo first perceived and the peer he will recognize presently. Hermes straddles the status-boundaries which commensality at the time often patrolled. This is underscored by the echo of 31, where Hermes foresaw in the tortoise a δαιτὸς ἑταίρη, both the lyre as a prestigious accompaniment to dining and the courtesan (already in our minds because of Apollo’s erotic response), who is not on a social level with the ἑταῖροι. This finesse is lost with Matthiae’s emendation ἑταίρην (1800: 291). 437 πεντήκοντα βοῶν ἀντάξια ταῦτα μέμηλεν: Apollo enters enthusiastically into a new Hermetic spirit of exchange, dropping a broad hint that he will cancel the debt of cattle if he gets the lyre and its music. Citing the number ‘fifty’ shows that his recent specific rage at the two cows killed has subsided, while the indirectness of the request may signal a residual but reduced wariness (Lohsee 1872: 8); indeed, there is some humour in his being able to cite the lyre’s value in terms of cows, the epic ‘currency’ for luxury items. The claim that in the tortoise Hermes acquired ὄλβος (24) is coming true. Apollo’s suggestion of swapping property to resolve a case of theft is couched in terms which could also signal a different and even more coopera­ tive sort of exchange – that of gifts, whose transfer carries an extra symbolism of good-will.505 For the idea of a song producing a gift ‘of corresponding value’ cf. Theoc. 17.114 δωτίναν ἀντάξιον ... τέχνας.506 This nuance extends

505

These are only the first of various forms of exchange which the two gods will think through in what follows: see further 462 (gift-exchange), 465 (sharing), 477 (exchange of services), 490–5. 506 The complement of this is the idea that the lyre gifted to Archilochus by the Muses is a τιμὴ ἀξία for the cow they take from him. The relationship of Herm. to the (undatable) story of Archilochus’ poetic initiation on the Mnesiepes inscription (Archil. T4 in Tarditi (1968) = SEG 15.517 A ii.22–55) is based on several shared motifs, albeit rearranged: Hermes and Archilochus are young and bantering; they set off by moonlight for meadows to fetch a cow for exchange; Apollo prophesies their musical fame. Meanwhile Apollo and Archilochus meet musical gods in the early morning, who organize a ‘worthy’ exchange of cow(s) for a lyra. See Kambylis (1963) 134–6, 141–50, Müller (1985), Jaillard (2007) 216–20, Vergados (2011a) 88–90. Thalmann (1984: 155) observed that the background of poetic initiation tropes may contribute to the startling sense in our scene that Hermes is initiating Apollo, of all people, into lyre-playing.   A remarkable parallel to Herm.’s phrase, from a period where we know little about its reception, is when Michael Psellos calls his Epist. 171 μυρίων ἀλόγων ἀντάξιος – an exchange of a form of literature for a number of animals; immediately thereafter he calls his addressee his ‘brother’, and he is asking for one of the mules Hermes ac-

386

C O M M E N TA RY 4 3 8 – 4 3 9 our interpretation of Hermes’ second song as a reflection on the reciprocity of hymnist and dedicatee (425–33*). The song simultaneously appeals to its immediate audience and elicits a gift from them; this is the kind of response, both emotional and financial, that Herm. itself might merit (cf. Eitrem 1906: 276, Nobili 2011: 198). The hymnist draws out another Hermetic side to hymn­­­ography by showing that it belongs in a world of transaction. ταῦτα μέμηλεν: Ω had μέμηλας. The sense ‘You have come up with’ (corresponding to present middle μέλομαι) governs a genitive, although the similar ἐπιμελέομαι can take an accusative pronoun (LSJ s.v. 1). I prefer to assume a misread ligature (c = -ας for < = -εν) so as to restore a normal construction. Either way, Apollo responds to and wittily completes Hermes’ claims in 266–7* about the babyish things which μέμηλε to him, while the lyre lay hidden; the allusion to Od. 8.248–9 noted there is now completed with the addition of music. 438 ἡσυχίως καὶ ἔπειτα διακρινέεσθαι ὀΐω: Significantly overturning 255 τάχα νῶϊ διοισόμεθ’ οὐ κατὰ κόσμον. ἡσυχίως is ‘calmly’, like ἡσύχως in e.g. A. Su. 724. It could hint specifically at the ἡσυχία of after-dinner lyre-playing (e.g. Pi. N. 9.48 ἡσυχία δὲ φιλεῖ μὲν συμπόσιον): see 454*. For καὶ ἔπειτα see 291–2*. 439 νῦν δ’ ἄγε μοι τόδε εἰπέ, πολύτροπε Μαιάδος υἱέ: The hymnist modifies the Homeric formula ἀλλ’ ἄγε μοι τόδε εἰπέ (Il. x6, Od. x13), to emphasize the contrast καὶ ἔπειτα ... νῦν δ’. In most cases (not Il. 11.819, 24.197), the formula is extended with καὶ ἀτρεκέως κατάλεξον, but here Apollo slips in a further vocative: in fact his speech contains an unusual number of addresses (436, here, 446, 455; cf. 408), which is a plausible feature of the register in which Greeks spoke to young children.507 The choice of πολύτροπε here (contrast 446 Διὸς καὶ Μαιάδος υἱέ) recalls 13 πολύτροπον, so that Hermes’ characteristics from the introduction are seen to be realized. For the ambivalent tone see 13*; Antisthenes (referenced there) mentioned a type of ‘polytropy’, musical modulation, which could be relevant here on a late view of the hymn’s date, around the middle of the fifth century when composers like Melanippides and Phrynis seem to have brought bolder and more frequent modulations into musical life, at least in Athens.508 The point

quires at Herm. 568. Psellos’ statement (Paneg. 4.59) that he does not need an opening invocation to Apollo, Hermes, or the Heliconian Muses perhaps alludes to the first two Homeric Hymns in the Ψ family and the proem of the Theogony. 507 Thomas (2010) 209–11. Eitrem (1906: 277) attractively observed that Apollo’s vocatives in 439–55 (‘son of Maia’, ‘son of Zeus and Maia’, ‘son of Zeus’) track Hermes’ general trajectory from Kyllene and his mother’s world to Olympus and his father’s. 508 See e.g. Hagel (2000) 81–7; D’Angour (2006: 273) argued for an early date for Melanippides.

387

C O M M E N TA RY 4 4 0 – 4 4 2 would not be that Hermes’ lyre-playing featured modulation, but that Apollo expresses his versatility in a way suited to his musicality. 440 ἢ σοί γ’ ἐκ γενετῆς τάδ’ ἅμ’ ἕσπετο θαυματὰ ἔργα: θαυματὰ ἔργα here refers to citharody (clarified in 442). The phrase was used of Hermes’ sandals (80), which already bore some resemblance to the lyre in their manufacture (78–86*), and also recalls the promise of Hermes’ ‘revelation of κλυτὰ ἔργα’ (16; this echo is reinforced by 439* πολύτροπε ~ 13). For θαῦμα at the lyre cf. 443, 455; it supersedes Apollo’s θαῦμα at 407, 414. Apollo asks if citharody ‘accompanied’ (see 426*) Hermes from birth.509 His conscious emphasis in σοί γε is that nobody else was involved. But the emphasis also suggests a comparison with other gods’ abilities, which is ironic: by showing his adherence to the received wisdom that gods have innate talents, most pertinently expressed in his own claim to the lyre as birthright in H.Ap. 131, Apollo draws attention both to his failure to comprehend Hermes’ inventive bricolage and to Herm.’s rewriting of how the lyre was invented.510 It was not actually Apollo’s birthright; he still imagines it might be Hermes’, but wrongly. For Apollo’s understanding of musicality see also 441–2*, 447–8*. 441–2 ἦέ τις ἀθανάτων ἠὲ θνητῶν ἀνθρώπων | δῶρον ἀγαυὸν ἔδωκε καὶ ἔφρασε θέσπιν ἀοιδήν;: Apollo’s second idea, that gods receive powers as gifts, is a thought writ large in the final movement of Herm.; see also Hes. Th. 885 (Zeus dispensing powers), H.Aphr. 29 (Zeus to Hestia). The Muses frequently bestow musicianship on humans, as in the initiations of Hesiod (Hes. Th. 22–34), Demodocus (Od. 8.44, 64) or Archilochus (SEG 15.517 A ii.26–38). Here Hermes’ musicianship could have been the gift of a god or a human. The latter could arise just from the ‘polar’ rhetoric discussed on 144*, but there were competing traditions that the aulos was invented by Athena or Hyagnis (sources in 452*). Either way, θέσπις emphasizes the possibility of a divine gift of musical inspiration (so at Od. 8.498, Hes. Th. 32; cf. fr. 310 θέσπιον), and it has a strong connotation of prophecy (cf. θεσπίζω) which will be taken further in 443* ὄσσαν and subsequently.511 The adjective thereby constitutes a response to Hermes’ claim to be under the patronage of Mnemosyne in his musical activities (429–30).

M’s ἐκ γενετῆς fits the usage of Il. 24.535 and Od. 18.6 exactly, whereas the epic sense of Ψ’s ἐκ γενεῆς is ‘from one’s descent’ (Il. 10.68). 510 See Görgemanns (1976) 123. For another innate divine quality see e.g. Athena’s militarism at Hy. 28.4–6. Pindar makes a claim that his music is an example of innate talent at O. 2.83–8. 511 See Fernández Delgado (2007a) on our line, and its influence on S. Ichn. 250 ἐγήρυσε θέσπιν αὐδά[ν. At Hes. Th. 31–3 the adjective immediately precedes Hesiod’s claim 509

388

C O M M E N TA RY 4 4 3 – 4 4 4 ἀγαυός is normally reserved for people; its application here to a gift is marked, and may forge an intratext with the word’s other use at 342*, where the sandal-traces were ἀγαυοῦ δαίμονος ἔργα.512 Again Hermes and his invention share features (n. 55) – in this case the mixed feelings of wonder and disapproval they inspire in Apollo. 443 θαυμασίην γὰρ τήνδε νεήφατον ὄσσαν ἀκούω: θαυμασίην returns us to the ‘headline’ response of 440 θαυματὰ ἔργα*, but ὄσσα adds a distinctive touch to 442 θέσπιν ἀοιδήν. As often, the hymnist controls several nuances at once. Homeric ὄσσα is an informative rumour from an unknown source, already ascribed in Il. 2.93 and Od. 1.282–3 to Zeus; this gives rise later to the sense ‘prophetic voice’ (e.g. Pindar O. 6.62 of Apollo; Neophron TrGF 15 F1 of Delphic Apollo; A.R. 1.1087–94 of a bird in augury; cf. ὀττεύομαι ‘take an utterance as ominous’). Meanwhile Hesiod uses ὄσσα of the musical voice of the Muses and Graces (Th. 10, 43, 65–7). And thirdly, Hes. Th. 830–5 uses ὄσσα of the ‘language’ of the various animals which make up Typhoeus’ heads. Apollo’s phrasing therefore marries the divinely inspired song of 442* θέσπιν ἀοιδήν to the bestial materials of Hermes’ lyre, and draws out the prophetic nuance of θέσπις to continue what will turn out to be a significant contrast between Hermes’ lyric voice and Apollo’s mantic one (cf. 427 κραίνων, Introduction §3.4). νεήφατος complements the range of ὄσσα. The second half may relate to φημί, in which case the adjective means ‘new-speaking’ and, being a hapax, may well itself be newly spoken. The contrast with παλαίφατα θέσφατα (Od. 9.507, 13.172) emphasizes the paradox of Apollo himself facing a new type of prophetic voice. On the other hand, νεήφατος could mean ‘newly killed’ (~ *gwhen-, as in ἀρηΐφατος). This recalls the paradox of an animal gaining a voice in death (25*), and hence Hesiod’s use of ὄσσα meaning ‘animal voice’.513 Sophocles responds to this nexus of ideas at Ichn. 329, where the satyrs speak of Hermes’ lyre-playing as the tortoise’s numinous ὀμφή (Introduction §1.3). 444–5 ἣν οὔ πώ ποτέ φημι δαήμεναι οὔτέ τιν’ ἀνδρῶν | οὔτέ τιν’ ἀθανάτων οἳ Ὀλύμπια δώματ’ ἔχουσιν: The lines forcefully reject Apollo’s to sing of past, present and future; for the connection of θέσπις to prophecy see also Dieu (2013). 512 For ἀγαυός of things see also Il. Parv. fr. 29 (leaves), Pindar fr. 52k.36 (singing), and a more common application to stars (Aratus onwards). 513 There are less good parallels for a potential relationship to φαίνω, though Pindar has πρόφατος ‘shown forth’ (O. 8.16). The first morpheme in νεήφατος is comparable to νεηγενής: see Schwyzer i.438–9. For contrasting discussion of ὄσσα see Collins (1999), esp. 245–6.

389

C O M M E N TA RY 4 4 6 – 4 4 7 prior acquaintance with the sound of lyres of all sorts (pace Martin 2010; cf. 451– 2, where they are absent from what the Muses know). The reprise of the godsand-men pairing also shows Apollo basically overruling his idea from 441–2 that some god or man might have bestowed a pre-existing lyre on Hermes. δαήμεναι initiates a cluster between here and 565 of ten uses of δαῆναι (compare Il. x8, Od. x17) plus two of its causative counterpart διδάσκω (484, 556). The verbs’ uses will map out the scope of Hermes’ and Apollo’s experi­ ential knowledge as they each hope to learn a new type of utterance (lyric poetry, prophecy) from the other. The field of knowledge in both case is a practical skill, as often with this stem (LfgrE s.v. δαῆναι 1b). The formula οἳ Ὀλύμπια δώματ᾽ ἔχουσιν, taken literally, allows that some divinity living elsewhere might have taught Hermes: Apollo will return to such a thought in 447* τίς Μοῦσα and 450 Μούσηισιν Ὀλυμπιάδεσσιν. 446 νόσφι σέθεν: Apollo now categorizes Hermes among the Olympians: Hermes’ plan is reaching fruition. Compare 537*. φιλῆτα: For the word, see 66–7*. For Apollo’s liberal and double-edged use of vocatives see 408, 439*. Here the tone has the ambivalence of 292, where Apollo described Hermes as the arch-φιλήτης shortly after smiling at his lies; 214 was more straightforwardly negative. 447 τίς τέχνη: After initially presenting musicality as genetic or giftwrapped (440–2), Apollo expressed some puzzlement at the second possibility (444–6), and in doing so invoked the idea of learning (δαήμεναι). He now starts to expand on three elements of learning music. The idea of poetry as a craft has deep Indo-European roots, though the earliest extant usages of τέχνη itself in literary and musical contexts are in Pindar (fr. 52k.39, P. 12.6; μουσική sc. τέχνη is first extant at O. 1.15; cf. Choerilus fr. 2.4). Apollo seems to envisage the ‘craft’ of the lyre in terms of playing technique, and Hermes’ answer will refer to someone τέχνηι ... δεδαημένος at 483. This makes a telling disjunction from Hermes’ actual ‘crafting’ of the instrument (25* τεκτήνατ’) – a sphere in which Apollo seems to take no interest. For Herm.’s other uses of ‘craft’, see 76*. τίς Μοῦσα ἀμηχανέων μελεδωνῶν: Μοῦσα is sufficiently personified to contrast with the Muses in line 450 (Kloss 1994: 176–7), even if an oral audience would not understand this immediately.514 Apollo conceives of meeting the relevant Muse as a way to learn a certain type of music (cf. 441–2*); one implication already in 445 is that it cannot be the Olympian Muses with

514

Contrast Häussler (1974) 1, who considered this the first ‘sicher metonymischer Gebrauch’ of μοῦσα.

390

C O M M E N TA RY 4 4 7 whom Apollo is familiar, but he appears to countenance there being some tenth Muse. Apollo’s thinking since 441–2* has led him to a rather tongue-incheek rejection of Hermes’ claim of a relationship to ‘Mnemosyne, mother of the Muses’ – i.e. the conventional Olympian ones – as his patron (429–30). The ‘Muse of unmanageable cares’ is probably the personification of music about desire; both the idea of Muses having specializations within μουσική (Murray 2004), and the idea of personifying individual styles as an idiosyncratic ‘Muse’ (e.g. Ar. Ra. 1304–6), take off in the classical period. Apollo thus perceives behind Hermes’ theogonic song a very different lyric genre whose content is μελεδῶναι. These are most likely not general human worries (Mimn. fr. 6, H.Ap. 532, Thgn. 185, 883, etc.) but erotic anguish, since this common nuance of the stem complements the heavy eroticization of Apollo’s response to the lyre; in particular 434 ἔρος ... ἀμήχανος is irresistibly recalled by ἀμηχανέων here, and μέλω will describe Apollo’s current pre­ occupation in 453.515 For erotic monody as an underlying presence in Herm. see Introduction §3.4. Different analyses of the genitive place the focus on slightly different parts of this general picture, and it would be artificial to pick one and exclude the others. The closest parallel seems to be Clio as Μοῦσα ... ἱστορίης in anon. AP 9.505f, i.e. the patron and inspiration of a certain form of μουσική. Construing the genitive as one of origin places slightly more weight on the idea that erotic monody is a response to the experience of desire. An objective genitive (cf. Od. 23.362–3 φάτις … μνηστήρων, ‘talk about the suitors’) or one expressing belonging make the more limited point that there is a genre of lyric (which can be personified) particular to erotic contents. I have not seen cogent parallels for taking Μοῦσα in a sense like ‘inspirer’, where the cares become the thing inspired, i.e. relate specifically to the experience of the audience (as of Apollo, or of Charmides at X. Smp. 3.1). Even less likely is the common interpretation that the Muse acts against cares, since the nouns governing such a genitive contain the notion of being a protection (37 ἔχμα) or at least a solution (E. Hipp. 716 εὕρημα συμφορᾶς, X. Mem. 3.8.3 πυρετοῦ ἀγαθόν).516

515

Compare e.g. Penelope’s yearning for Odysseus at Od. 19.517; Hes. Op. 65 of Pandora, with the etymological play γυιο-βόρους μελ-εδωνας; A.R. 3.4 μελεδήματα of Erato, Phanocles fr. 1.5. Kloss (1994: 176–7) claimed that μελεδῶναι here could mean ‘songs’, but the support for this (Hsch. μ.685 μεληδών: ὠιδή) remains mysterious. 516 The final interpretation is mentioned in e.g. Ernesti; Allen (1897), comparing Hes. Th. 55 ἄμπαυμά τε μερμηράων; Eitrem (1906) 277; Kaimio (1974) 36–7, comparing ἐπωιδή + genitive. Ilgen interpreted ‘musicianship characterized by cares’ (to the performer), but the genitive hardly ever gives a characteristic in Greek (Smyth 1920: §1320). Peponi (2012: 107–14) rightly insisted on the thematic importance of ἀμήχανος and the stem μελ- here, but presupposed a ‘descriptive genitive’ where-

391

C O M M E N TA RY 4 4 8 ἀμηχανέων should be taken as feminine genitive plural of ἀμήχανος, not a form of the otherwise unattested ἀμηχανής. Allen and Sikes list various α-privatives with three-termination forms in Epos, e.g. Od. 9.503 ἀεικελίην. I have tentatively written μελεδωνῶν from μελεδώνη (cf. Agar 1929: 19), not μελεδώνων from μελεδών. If one leaves aside ambiguous μελεδωνας| and μελεδωνων, the early forms are from μελεδώνη (Hippocratic ‘therapy’; plural ‘cares’ at e.g. Sappho fr. 37, Ps.-Theoc. 21.5, Phanocles fr. 1.5) and μεληδών (Simonides fr. 21.6, A.R. 3.812). μελεδών is not found for certain until Galen and Herodian, and need not be the source of congeners such as μελεδαίνω or μελεδωνός.517 448 τίς τρίβος: Some scholars have assumed that the word, uniquely, denotes ‘practice’, like τριβή.518 But a metaphorical ‘worn track’ has at least five pertinent implications. For the metaphor of a ‘path’ allowing progress in a structured intellectual pursuit such as music compare the Hippocratic VM 2 ἰητρικῆι δὲ πάντα πάλαι ὑπάρχει, καὶ ἀρχὴ καὶ ὁδὸς εὑρημένη, καθ’ ἣν ... τὰ λοιπὰ εὑρεθήσεται, ἤν τις ἱκανός τε ἐὼν καὶ τὰ εὑρημένα εἰδώς, ἐκ τουτέων ὁρμώμενος ζητέηι, LSJ s.v. μέθοδος II.2 (Plato on). The song-as-path metaphor has Indo-European roots.519 The ‘path’ of musicianship may also be understood as a metaphor for a career choice (see further 448–9*). Among words for ‘path’, τρίβος suggests the grind and smoothness which result from repeated travelling. This focus on the pressure which creates a track further connects Hermes’ citharody to his sandals (cf. 440*), since the latter imprinted a track (marked through the anomalous use of διατρίβω in 348*) and led Apollo effectively to ask τίς τρίβος before at 218–26. Finally, the grind of daily practice enters as a natural implication of the metaphor of a worn track (rather than its standalone meaning), supported in context by the similarity to τριβή. This emphasis on practice is quite different from earlier poetic representations of the origins of human musicianship. In epic this derives from a divine gift bestowed at birth (natural talent) or in a meeting with a musical god (a Dichterweihe); the importance of apprenticeship

by Apollo expresses his response as a listener; she also claimed a delightful side to μελεδῶναι which I do not find in the sources. 517 For μεληδών ~ ἀλγηδών, ἀχθηδών see Chantraine (1933) 361, and with μελεδών cf. σηπεδών, τηκεδών, etc. 518 E.g. Allen and Sikes, Kaimio (1974) 36 n.6, Görgemanns (1976) 125 n. 34. See LSJ s.vv. τριβή II, τρίβω III.4. By contrast Assaël (2001: 14–15) equated τρίβος with οἴμη and interpreted the ‘path of song’ with no reference to practice. 519 West (2007) 42–4. Compare 451* οἷμος, Taillardat (1965) 434 n. 1. Callimachus is instructed, by contrast, to travel κελεύθους | [ἀτρίπτο]υς (Aet. fr. 1.27–8), rather than the hackneyed highroad.

392

C O M M E N TA RY 4 4 8 – 4 4 9 and practice is barely alluded to, the rule-proving exceptions being when Phemius mentions his lack of a human teacher, and Odysseus emphasizes that Demodocus (σέ γε) must have been taught by a god.520 Poets such as Theognis and Pindar mention the value-adding role of teaching, but still emphasize that natural talent is a prerequisite (Marrou 1956: 40) – a ‘eugenic’ argument which appealed to their principal audiences. It is among the Presocratics that we find the first other analyses of mental improvement into τέχνη and practice (Protagoras 80 B10), sometimes alongside other factors (memory and σοφία in Anaxagoras 59 B21b). 448–9 ἀτρεκέως γὰρ ἅμα τρία πάντα πάρεστιν | εὐφροσύνην καὶ ἔρωτα καὶ ἥδυμον ὕπνον ἑλέσθαι: The point of ἅμα τρία πάντα (‘all three together’, sc. ‘with your music’ or ‘with each other’) is that Hermes’ music presents a new opportunity for combining pleasant states which Apollo knows separ­ ately (see also 454*). Since sleep and desire are not precisely simultaneous, there is a touch of exuberant hyperbole. Even so, I doubt whether ἀτρεκέως qualifies ἅμα, and construing it closely with τρία or ἑλέσθαι produces a false emphasis: I take it as lending asseveration to the whole sentence. This usage is probably first extant here, though Il. 5.208 ἀτρεκὲς αἷμ’ ἔσσευα is difficult; see e.g. Call. Lav. Pall. 5.137 ἔρχεται ... ἀτρεκές and, with the adverb fronted as here, SH 982.12, Q.S. 7.689, Nonnus D. 34.50. The three elements which Apollo identifies form a bridge between his response to lyre-music and the sympotic and specifically erotic/courtesanly qualities of Hermetic music which surfaced earlier in Herm. (esp. 28*, 31*, 54–6*, 434*, 447). εὐφροσύνη ‘cheerfulness’ suits Apollo’s pacification at 417, but has a strong connection to the symposium (e.g. Od. 9.6, Solon fr. 4.10, Xenophanes fr. 1.4, Panyassis fr. 16.19). Passages which combine εὐφροσύνη with sex, symposia and music include Solon fr. 26, Anacreon fr. eleg. 2.4, X. Smp. 8.21. Hesiod Th. 909–10 lists the Graces as Εὐφροσύνη, Θαλίη and Ἀγλαΐη, whose eyes ooze eros-appeal: all three sisters’ names can mean ‘party’, and all will occur in this section of Herm.521 Apollo’s own response of ἔρος (434) will also be repeated at future symposia. The reference to sleep could be taken as an extension of the pacifying quality of the lyre – as at Pindar P. 1.6–12 Apollo’s lyre soothes Zeus’s eagle and Ares to the point of sleep.522 It also shifts us from audience-response towards performance-setting,

520

Od. 22.347–8, 8.488. For Herm.’s representation of musical practice, and its datability or otherwise, see 483–9, Introduction §1.2. 521 Cf. 451–2, 454, 476. For Muse-names see also 452*: one might hear ‘Erato’ in ἔρωτα here. 522 Matthiae (1800: 292) made this comparison, adding that ‘soporific’ was no longer a positive term in concert reviews.

393

C O M M E N TA RY 4 5 0 and the implication in ἅμα that the setting for Hermetic lyric leads straight to sleep places it at night, and confirms its sympotic nature. Sleep limits the symposium at e.g. Od. 11.331, 373–4, Thgn. 469–70, Pl. Smp. 223b–d; later sources specify that Hermes himself received the final libation (n. 22). ἑλέσθαι occurs formulaically of ‘taking’ sleep and meals, and sporadically also of drink. However, the verb is anomalous with ἔρως, which is in fact often the subject of αἱρέω.523 Apollo thereby presents the lyre as a marvellous opportunity to have some agency in desire, perhaps covering up how he has been assailed by ἀμήχανος ἔρος (434) and ‘seized’ by ἵμερος (421). The nearest parallel in Epos to πάρεστιν ... ἑλέσθαι is Hes. Op. 287 ἐστιν ἑλέσθαι, in the influential contrast between the paths of vice and virtue. The most famous moralizing interpretation of Hesiod’s lines is Prodicus’ speech about Heracles’ choice (T84 = X. Mem. 2.1.21–34; cf. West 1978: 109, 229). Since we cannot be sure of the relative date of Herm. and Prodicus’ speech, nor of what Prodicus owed to earlier sources, we can at least try out a ‘Prodican’ interpretation of Apollo’s contrast of two paths: the Hermetic τρίβος (448) is well worn like Hesiod’s path of vice (Op. 287–8) and leads like Prodicus’ path of Kakia to the pleasures of food and drink, sex and sleep; meanwhile, οἷμος is used of the opposing paths in Herm. 451 and Hes. Op. 290. However, Prodicus’ Kakia is dismissive of the εὐφροσύναι offered by Virtue, the ‘welcome collaborator for craftsmen’ whose path leads among other things to the production of commemorative hymns (X. Mem. 2.1.32–3). On this approach, Apollo would take Hermes’ hymn-plus-theogony to deconstruct a Prodican reading of Hesiod’s passage, since it puts craft, hymns and εὐφροσύνη to the service of hedonism. This is suggestively in line with the overall presentation of Hermetic hymnody as crossing ordinary musical categories: Introduction §3.4. Suhle (1878: 24) noted that earlier poets would probably have said ἔρον καὶ νήδυμον. The form ἔρωτα is not found elsewhere in hexameter until Parmenides 28 B13, albeit ἔρον only occurs in ἐξ ἔρον ἕντο+|. But forms of ἔρως, -τος are metrically secure from Archil. fr. 191.1 in other genres. Janko (1982: 135) compares χρώς, ἱδρώς. On νήδυμος see n. 307. 450 καὶ γὰρ ἐγώ: ‘For in fact I’ – not ‘For I too’, since Apollo thinks Hermes must be affiliated with some other Muse. Despite this, Theocritus may well have had our line in mind at 7.37 καὶ γὰρ ἐγὼ Μοισᾶν καπυρὸν στόμα: Rudoni (2015). Μούσηισιν Ὀλυμπιάδεσσιν ὀπηδός: For the ‘Olympian’ Muses see e.g. Il. 2.491, Hes. Th. 25. For Apollo’s contrast between them and Hermes’ potential new Muse see 445*, 447*. In accordance with ὀπηδέω (LfgrE s.v. 1–2), ὀπηδός LfgrE s.v. αἱρέω I B 1b; for drink see Od. 11.584, 21.294; for ‘snatching sleep’ cf. Th. 2.75.3 ὕπνον τε καὶ σῖτον αἱρεῖσθαι, 3.49.3. Cf. Il. 16.282 φιλότητα ἑλέσθαι.

523

394

C O M M E N TA RY 4 5 1 refers not only to ‘following’ but also to ‘attending, offering (divine) support to’. The latter is more suited to Apollo’s general position as the Muses’ director or leader (e.g. H.Ap. 186–203, LIMC s.v. Apollon II.C), though the word’s roots in a kinetic metaphor neatly suit the path-imagery in 448 and 451. ὀπηδός is a hapax in Epos.524 This makes it tempting to trace Herm. as the source of Apollonius’ one use of the noun (3.595), in a context rich in connections to our hymn: Introduction §4.2. Pontani (2002) noted the likely influence of the lemma on Poliziano Epig. 11.6. 451 τῆισι χοροί τε μέλουσι: Apollo’s use of μέλω contrasts with 437, where the verb described Hermes’ occupation with the lyre; see 453*. There may be an underlying pun on μέλος ‘tune’, as at H.Ap. 188 μέλει κίθαρις καὶ ἀοιδή and a number of similar passages (E. HF 763–4 χοροὶ καὶ θαλίαι μέλουσι, Ar. Lys. 1305 τᾶι [= Sparta] σιῶν χοροὶ μέλοντι, Pl. Lg. 8 835e θυσίαι δὲ καὶ ἑορταὶ καὶ χοροὶ πᾶσι μέλουσι, Triph. 436 ἡμῖν δὲ χοροὶ θαλίαι τε μέλονται). An allusion to the Muse named Terpsi-chore gets some support in 452*. καὶ ἀγλαὸς οἷμος ἀοιδῆς: For ἀγλαός see 60–1*. In musical contexts it suggests elation, slightly different from the εὐφροσύνη induced by Hermes’ music (449). Often the splendour is of a choral performance, which is of particular relevance to the Muses.525 But the scope can be wider, e.g. when one of the Graces is called Aglaia (Hes. Th. 909); cf. 476*. Ψ’s οἶμος varies with ὕμνος (Mxm). Both usages can be parallelled, but οἷμος is preferable because of its perfect fit with 448 τρίβος.526 The ‘path’ where Apollo ‘attends on’ the Olympian Muses’ art is set against that of Hermes and his Muse; both nouns are masculine counterparts to words with musical application (τριβή ‘practice’, οἴμη ‘song’).527 Unlike τρίβος, οἷμος alludes to the genre of hymnic προ-οίμιον which the overall poem exemplifies. This again suggests that Hermetic song, as exemplified in Herm., stands apart from but complements what Apollo knows. οἷμος also seems to have been read by Callimachus: his Zeus makes Apollo patron of those who understand the

Hes. fr. 26.10 and Pany. fr. 16.13 have συνοπηδός, and Choerilus(?) fr. 23.5 has ὀπαδοί. The earlier attested noun is ὀπάων. Meier-Brügger (1991) derived ὀπαδός < ὀπαδέω < unattested adverb ὀπαδόν. 525 So of ἀγλαΐα at e.g. Sc. 272, 284–5, Bacchylides fr. 4.56, Pindar O. 1.14–15, 14.13–15, P. 1.2, N. 1.13, frr. 75.7, 148 (the dancer Apollo ἀγλαΐας ἀνάσσων), 199.3; cf. O. 3.5–6 φωνὰν ... ἀγλαόκωμον. 526 Pagliaro (1953: 34–40) analysed οἷμος and its musical applications. On the breathing see LSJ s.v.; it was restored to our line by West. The variant ὕμνος would have arisen during transmission as a sonic error. 527 For τρίβος and τριβή see 448*. Maslov (2012: 197–201) has discussed the connection perceived by ancient poets between οἷμος and οἴμη. 524

395

C O M M E N TA RY 4 5 2 οἷμοι of the λύρη (H.Zeus 78; Introduction §4.1) – a conjunction of two words which are extant in Epos only in Herm., in an alternative story about how Apollo acquired patronage of the lyre. 452 καὶ μολπὴ τεθαλυῖα: μολπή in Epos generally combines dancing (451 χοροί) with singing, often with a religious focus. The adjective may refer both to ‘vigorous’ singing (compare the formula θαλερὴ δέ οἱ ἔσχετο φωνή, e.g. Il. 17.696) and to the social ‘flourishing’ which μολπή reinforces.528 The phrase evokes two of the standard Muse-names, Melpomene and Thaleia, and the Grace ‘Thalia’ (Hes. Th. 77–8, 909); cf. 448–9* on Euphrosyne and Erato, 451* on Terpsichore and Aglaia, and the uses of θαλ- in 454, 480. Hermes responds to this mention of μολπή in 476–8. καὶ ἱμερόεις βρόμος αὐλῶν: Apollo ends with the raciest music the Olympian Muses currently manage, preparing to cap this in 453–4 with what Hermes has managed (cf. 449 ἔρωτα). For ἱμερόεις compare Theognis 532 αὐλῶν φθεγγομένων ἱμερόεσσαν ὄπα, Pindar fr. 140b.17 αὐλῶν ἐκίνησ’ ἐρατὸν μέλος. The epithet may be linked in Theognis to the aulos’ use in symposia by courtesans and for erotic monody.529 βρόμος suggests a loud, continuing, unarticulated sound; this is often threatening (e.g. of fire, waves, winds, military clamour). As such it sometimes adds to the wildness of piping (Hy. 14.3 βρόμος αὐλῶν| in Magna Mater cults, E. Ba. 160–1 λωτὸς ... βρέμηι), though musical βρόμος need not be so dangerous (e.g. Pindar N. 11.7 λύρα ... βρέμεται at banquets on Tenedos, Athenaeus Paean 12 λωτὸς βρέμων). Apollo’s phrase fills in the musical accompaniment to the Muses’ other activities in 451–2. The implication that the auloi had priority over the lyre, both in general and in Apollo’s personal relationship to music, is incompatible with normal views that the aulos was invented by Athena in the time of Perseus (e.g. Pindar P. 12), or invented by Hyagnis and imported to Greece later (e.g. Marm. Par. 10). This reinforces the boldness with which Herm. has denied Apollo a primary relationship with the lyre as his birthright.530

LfgrE ii.963.63 takes θαλερός when used of the voice and of tears to mean (approximately) ‘gushing’. This is too runny for other applications (fat at Od. 8.476, hair at Il. 17.439, plants at e.g. Hes. Op. 173), though I accept that the vitality denoted by θαλ(λ)- is opposed to desiccation. 529 For visual humour involving the juxtaposition of auloi and penises around 500 bce see e.g. CVA Castle Ashby pl. 36.1 (Ashby Painter), Würzburg 2 pl. 8.2 (Cleophrades Painter), and the red-figure plate by Epiktetos, Paris, Cabinet des Médailles 509 (with Mitchell 2009: 167–9). 530 See 444, and Thalmann (1984) 154. For the relative social standing of auloi and lyres in the fifth century see Wilson (1999), Stewart (forthcoming). 528

396

C O M M E N TA RY 4 5 3 – 4 5 4 453 ἀλλ’ οὔ πώ τί μοι ὧδε μετὰ φρεσὶν ἄλλο μέλησεν: The echo of 444 οὔ πω underscores Hermes’ innovation, but the focus now is not on the player’s knowledge but on the novelty of the hearer’s response. μέλω has become a keyword of the passage (437, 447 μελεδωνῶν, 451), and in this context may evoke the musical use of the stem in μελέτη, ‘training, practice’ (e.g. Pindar O. 14.18). The line implements a traditional structure, ‘, μετὰ φρεσί, μελ-’. There is a sequence of such phrases at Il. 18.463 (this line recurs in Od. x3), 19.29, 213, 343. Peponi (2012: 104–6, 113) made a suggestive connection also to Iliad 3.442 and 14.315 οὐ γάρ πώ ποτέ μ’ ὧδε ... ἔρος, though this ‘never so much before’ structure is also applied to quite different contexts (Il. 3.169–70, Od. 3.221, 4.141, 17.587, 19.350). 454 οἷα νέων θαλίηισ’ ἐνδέξια ἔργα πέλονται: οἷα is used flexibly, as if ‘like these things’ rather than ‘in this way’ had preceded; cf., with the order of clauses reversed, Il. 22.317–19 οἷος δ’ ἀστὴρ εἶσι ... ὣς αἰχμῆς ἀπέλαμπ’. Comparison of our line with Od. 10.222–3 οἷα θεάων ... ἔργα πέλονται, H Dem. 140 οἷα γυναικὸς ἀφήλικος ἔργα τέτυκται suggests that νέων specifies ἔργα as well as the θαλίαι. The present πέλονται implies that Hermes’ playing resembles a type of youthful performance which already exists but does not yet employ the lyre. The line recalls 55–6* ἠΰτε κοῦροι | ἡβηταὶ θαλίηισι παραίβολα κερτομέουσιν, when Hermes first played. Here the tone of the performance is replaced by the rule that it goes ἐνδέξια, from left to right.531 ἐνδέξιος and ἐπιδέξιος mark a traditional hierarchy at symposia, used to order the pouring of wine (Il. 1.597), making toasts (in various places according to Critias 88 B1.7, 6.6, 6.23; Eupolis fr. 354), trying a contest (Od. 21.141–2; Rhesus 363–4, perhaps referring to kottabos; toasting becomes competitive in Dionysius Chalcus frr. 1, 4) and even begging (Od. 17.365). Anaxandrides fr. 1 τὸν ἐπιδέξι’ ... λέγειν ἐπὶ τῶι πίνοντι has speeches of praise going to the right; in Plato’s Symposium the guests praise Eros in that order (177d), which continues even after Alcibiades’ entrance (214c, 222e). Hsch. τ.796 says that a lyre (as in Eupolis fr. 395) or myrtle-branch (as in Aristophanes fr. 444) was passed to the right and each guest sang when they took it; cf. Dicaearchus fr. 88, Polybius 4.20.10. From the audience’s perspective, Apollo’s comparison of lyre-playing to ἐνδέξια ἔργα marks his insight that performing snatches of lyric would become one of the kinds of left-to-right sympotic contest. The performance-rule puts a limit on the size of the θαλίαι we imagine here, and would suit the types of capping-poetry discussed on 55–6 particularly well, though also other 531

Matthiae (1800: 292–3) identified the sense. The translation ‘skilful’ has had a long life from Dartona (1537) to DGE s.v. II.2, but it has no external support.

397

C O M M E N TA RY 4 5 5 – 4 5 6 ­genres. It matches the situation in Herm., to the extent that Apollo is a youth, Hermes is precocious enough to be one (especially in his cattle-theft), and Hermes has stood on Apollo’s left (424) before playing. Apollo construes that action as an offer, which he welcomes, to hand on the lyre. Nevertheless, the generic disjunction between Hermes’ (solo) hymn-plus-theogony and sympotic contests based on multiple short performances marks Apollo’s simile as expressing a paradoxical insight, endorsed by the narrator’s very similar comparison at 55–6. I argue in Introduction §3.4 that the paradox points to Herm.’s intertextual ‘capping’ of the Hymn to Apollo, and the tone in which that is intended. 532

455 θαυμάζω, Διὸς υἱέ, τάδ’ ὡς ἐρατὸν κιθαρίζεις: The line rounds off Apollo’s response to the lyre and prepares for a transition to his promises in 456–62. The structural node is marked by another mid-speech vocative (439*) and by the repetition of the key motifs of θαῦμα and ἔρος (and specifically 423 ἐρατὸν κιθαρίζων). Hermes will parrot Διὸς υἱέ at 468*; despite the frequency of Διὸς υἱός in other cases, the form of address only occurs previously at Od. 8.335, where Apollo addresses Hermes (Ἑρμεία Διὸς υἱέ), then again in the classical period at Hy. 15.9, B. 17.20. 456 νῦν δ’: The repetition of this transitional phrase from 439 suggests that Apollo has slightly lost sight of his purpose when he used it first: his wonder there led him to vocalize various answers (with 444–6 correcting 440–2) to his question, and now he continues with a new thought. ἐπεὶ οὖν ὀλίγος περ ἐὼν κλυτὰ μήδεα οἶδας: Apollo perhaps recalls his initial focalization of Hermes’ inferiority at 245 παῖδ’ ὀλίγον, and acknowledges that he had been hasty; similarly Polyphemus at Od. 9.515–16, νῦν δέ μ’ ἐὼν ὀλίγος τε καὶ οὐτιδανὸς καὶ ἄκικυς | ὀφθαλμοῦ ἀλάωσεν. μήδεα are practical ‘ideas’, and a formulaic object of οἶδα at line-end. However, μήδεα are not elsewhere κλυτά, except Machaon’s ability to respond to medical emergencies at Q.S. 7.15 ἐπεὶ κλυτὰ μήδεα ἤιδη, which may or may not be a recollection of our line. The choice of adjective reminds us that Hermes’ ideas bring about κλυτὰ ἔργα (16): Apollo refers primarily to the lyre, an ἔργον which ἐμήδετο κύδιμος Ἑρμῆς (46); there is probably also a contrast with the rhetorical imagination which made Zeus see Hermes as κακομηδής (389). Hermes’ ideas are themselves ‘renowned’ (as the present hymn proves) but also ‘confer renown’ on him (as Apollo will confirm in 458). 532

The implication in Ar. V. 1219–47 is that skolia, as the name may suggest if it has to do with σκολιός, could also take less fixed orders: Bdelycleon might have to respond either to Cleon or to Theoros.

398

C O M M E N TA RY 4 5 7 – 4 5 8 Ψ’s οἶδας (also 467*) is the lectio difficilior, and M’s οἶσθα (used in 382) fits that manuscript’s tendency to simplify poetic forms. The poet perhaps wished μήδεαι ἀοιδάς to be audible behind the words; the one Homeric use of οἶδας rhymes with ἀοιδός (Od. 1.336–8). Ψ’s scribe, after finishing 456, recalled that it had started νῦν, as does 458. He therefore skipped straight to 459. 457 ἵζε πέπον: It is unclear on the level of characterization exactly why Apollo bids Hermes sit, and whether Hermes obeys. The instruction echoes 282–303, where Apollo called Hermes πέπον (282*), foretold his leadership of thieves (291–2), dropped Hermes and sat down himself (298–9), then predicted Hermes’ ‘leadership’ as a guide. Here, and especially in 461* ἡγεμόν’ ἕσσω, we have Apollo’s reformed predictions. Hermes evaluated sitting negatively at 172 (cf. 284) as a sign of immobility. Apollo’s request marks progress from trying to enforce Hermes’ immobility to asking him for self-restraint, before 461 and 468 bring references to positive kinds of sitting. καὶ θυμὸν ἐπαίνει πρεσβυτέροισιν: The text is probably defensible as ‘for the benefit of your elders, express assent to their impulse’. For the syntax compare S. Ph. 1398 ἃ δ’ ἤινεσάς μοι ‘The things you assented to for my benefit’; the nuance of benefit in the dative can be weak, little more even than possession (344*). In πρεσβυτέροισιν Apollo is only thinking of his own θυμός, which he cagily does not state, but which Hermes recognizes as the impulse to learn the lyre (cf. 475, also with θυμός). This caginess and the reference to a respectable plurality (πρεσβύτεροι) show Apollo mitigating the sense that his words from here to 462 betray greed for a personal quid pro quo.533 His request may also recall κατὰ πρέσβιν (431) to imply that Hermes has already shown one sign of honouring his seniors. Ruhnken instead changed θυμόν to μῦθον (1781: 46), comparing Il. 2.335 μῦθον ἐπαινήσαντες Ὀδυσσῆος. Schneidewin’s θυμὸν ἴαινε (1848: 686) is comparable to Od. 15.379 θυμὸν ἀεὶ δμώεσσιν ἰαίνει. θυμῶι (Gemoll) does not on its own mean ‘wholeheartedly’ (contrast Pi. P. 9.95–6 αἰνεῖν καὶ τὸν ἐχθρὸν παντὶ θυμῶι, A. Eu. 737–8 τὸ δ’ ἄρσεν αἰνῶ ... ἅπαντι θυμῶι), and Apollo wants some open expression of assent, so ‘in your heart’ is not enough. 458 νῦν: Either ‘presently’ or ‘given present circumstances’, i.e. the revelation of the lyre as such a strong bargaining tool. τοι: I take this as ‘for you’, and σοί τ’ αὐτῶι καὶ μητρί in 459* as an addition of specifics, following a traditional type; compare Od. 12.451–2 ἤδη γάρ τοι

533

For plurals and mitigation see Brown and Levinson (1987) 197, 203. Van Herwerden (1876: 72) proposed πρεσβυτέροιο, noting that the ending may have been corrupted by its proximity to σι at the end of 458.

399

C O M M E N TA RY 4 5 9 – 4 6 0 χθιζὸς ἐμυθεόμην ..., σοί τε καὶ ἰφθίμωι ἀλόχωι. Radermacher and Vergados took τοι as the particle, expressing certainty. κλέος ἔσται ἐν ἀθανάτοισι θεοῖσιν: Apollo’s prediction again recalls 16, where Hermes’ deeds were set to be κλυτά among the gods, and 291–2, where Apollo predicted γέρας among the gods including the title (κεκλήσεαι) ‘leader of brigands’ (cf. 456 κλυτὰ μήδεα, 457 ἵζε πέπον* for these intratexts). In all three cases Hermes will attain κλέος within the pantheon – an unusual conception, particularly with the noun κλέος itself, but one well suited to the particular context of this source, where a god born in secret (and his bashful mother: cf. 6) is not yet familiar.534 459 σοί τ’ αὐτῶι καὶ μητρί: The language is reminiscent of decrees conferring honours on ‘X ... αὐτὸν καὶ ἐκγόνους’, e.g. Peek (1972) no. 10 (470s). The prediction somewhat compensates for the unrealistic way Apollo has ignored Maia since 244. However, she has been gently kept in mind through metronymics (e.g. 301, 408). Her fame is proved by the start of Herm. and her epithet ἐρικυδής (89); Hermes’ first song suggested she was a topic awaiting an audience, and at 166–72 he insisted that neither he nor Maia should languish in obscurity. Shapiro (2019) emphasizes that Maia makes occasional appearances on Olympus in painting, as on the Francois Vase. τὸ δ’ ἀτρεκέως ἀγορεύσω: Echoing 380*, with a bantering tone since Hermes’ narrative and oath there were not as exact as Apollo would have liked. ἀτρεκέως ἀγορεύσω| is formulaic, normally of promising to tell a narrative precisely. However, the sense of ἀτρεκέως moved easily towards ‘with certainty’ (already in Il. 15.53, and dominantly in Herodotus’ usage), so that the formula can also apply to individual facts as here. 460 ναὶ μὰ τόδε κρανάϊνον ἀκόντιον: The javelin is abruptly introduced, and spears are not the most obvious accoutrement of Apollo: see Paus. 3.19.2–3, at Amyclae; LIMC s.v. Apollon 322 (orientalizing, late seventh century). Gemoll compared Od. 14.531, where Eumaeus takes a javelin as he leaves: maybe Apollo picked one up that morning. But objects held during Greek oaths generally have clear symbolic pertinence.535 The main contexts for javelins – w ­ arfare, LfgrE ii.1439.9 wrongly suggested that κλέος in our line means merely ‘honour’. Gods are more often (περι)κλυτός or the object of κλείω. As in those cases, the other application of the noun κλέος to a god in Epos also implies κλέος among a human audience; suggestively, it involves Apollo making a prediction about his own reputation (H Ap. 276, 381; see Introduction §1.3). 535 For objects by which oaths are sworn see also on 384*. Humans swear by a spear in a sign of extreme belligerence (A. Th. 529, A.R. 1.466). Examples with clear symbolism include Il. 14.271 (Hera holds land and sea to symbolize her ‘contact’ with 534

400

C O M M E N TA RY 4 6 1 hunting, and pentathlon – might interact with pentathletes competing by the Alpheus, or one’s sense that Apollo ‘hunted’ Hermes. But I suspect the symbolism lies more in κρανάϊνον and a pun on κραίνω.536 Javelins and spears were one of the main Greco-Roman uses of the hard, unsnappable wood of European cornel (Cornus mas).537 The javelin is a rod with the etymological potential to κραίνειν, and as such stands somewhere between the material form of the laurel bough and sceptre with which Apollo is sometimes depicted (for the latter see e.g. LIMC s.v., nos. 469, 1087), and the functional power invested in a sceptre (see Achilles at Il. 1.234, the only other instance of ναὶ μά in Epos) and in the staff which Apollo promises Hermes at 531–2*. Both κρανέϊνος and κρανάϊνος are ancient spellings, the latter being slightly closer to the transmitted κράναιον.538 For the lengthened iota see Il. 23.806 ἐνδίνων, Matro fr. 1.79 with Olson and Sens (1999). 461 κυδρὸν ἐν ἀθανάτοισι καὶ ὄλβιον ἡγεμόν’ ἕσσω: Ω’s ἡγεμονεύσω would almost certainly require a genitive or dative, despite the personal passive at Th. 3.61.2 οὐκ ἠξίουν ... ἡγεμονεύεσθαι (see KG i.124). Agar proposed ἡγεμόν’ ἕσσω (1896: 389), Tyrrell εἵσω (1894: 47). Either is palaeographically simple, and with κυδρόν evokes the formula καθέζετο κύδεϊ γαίων for triumphant gods; either suits the thematic interest in whether sitting is positive or negative (457*, 468*). However, both lack parallels as the future of ἵζω: καθ-έσω occurs in Eupolis fr. 99.110 but Agar’s doubled sigma has no classical parallel in this verb (cf. GH i.178–9), while Tyrrell’s εἵσ- is only attested later in the middle (A.R. 2.807). The alternatives ἡγεμόν’ οἴσω or ἄξω, or ἡγεμονώσω, therefore merit consideration. Humbert’s ἥσω gives a less good sense.

the Underworld gods by whom she swears), 23.585 (Antilochus swears that he did not cheat in the chariot race while holding his riding equipment), Ar. Lys. 194–239 (the women touch the κύλιξ φιλοτησία as they swear that if they transgress they will become teetotal). 536 Perhaps there is also a connection to Apollo Κάρνειος, as at Paus. 3.13.5. For the wordplay compare the gate of horn (κεράων ... κραίνουσι) through which prophetic dreams pass at Od. 19.566–7. 537 Xenophon favours cornel javelins for cavalry (Eq. 12.12; cf. HG 3.4.14, Cyr. 7.1.2). Theophrastus’ comment that Cornus sanguinea is no good for ἀκόντια presupposes such a use of Cornus mas (HP 3.12.1, noting the latter’s strength). κρανεία came to mean ‘spear’: Lyc. Al. 553, AP 6.122–3 (Nicias and Anyte), Ael. Arist. 3.290. Cornel also produced sturdy cudgels and bows. The Romans made similar use of it: see TLL s.v. cornus col. 975.2–22, Plu. Rom. 20.6. 538 With Ilgen’s κρανέϊνον compare ID 504.18 (c.280). κρανάϊνος is found in BGU 1253.4, PHeid. ii.217 (both second century bce), ID 2634.2 (Imperial); Chantraine (DÉLG s.v. κράνον) suggested it is analogical to ἐλάϊνος.

401

C O M M E N TA RY 4 6 2 Fletcher (2008: 28–30) gathered parallels for gods’ oaths fixing the prerogatives of the swearer or another god. In Apollo’s mouth his (not very believable?) arrogation of the ability to make Hermes a rich and glorious leader among the gods contrasts point for point with his arrogation of the ability to make Hermes a leader languishing among men in Tartarus (259*), as well as his prediction of Hermes the leader at 303, which is now being refined (457*). For Hermes’ desire for wealth through cult see 171 (opposed to sitting), 379. For ἐν ἀθανάτοισι, repeated for emphasis so soon after 458, see 170*. For κῦδος see 375*, and 477*, where Hermes will respond to this aspect of Apollo’s promise. 462 δώσω τ’ ἀγλαὰ δῶρα καὶ ἐς τέλος οὐκ ἀπατήσω: Like the preceding oath (459*), Apollo’s final declaration of non-deception toys with Hermes’ earlier use of deceptive oaths. ἐς τέλος means ‘for ever’, like διὰ τέλους, rather than ‘in the long term’ (as at e.g. Plb. 1.20.7). Such promises of eternal frankness are found in diplomatic formulas used in forging alliances and confirming oaths, not just in the realm of personal reconciliations. At the time of Herm. Olympia was an important forum for such diplomatic business, which was very much Hermes’ arena. See e.g. SEG 22.336 (Olympia, c.525) ἀρμόχθεν ... ἐπὶ φιλότατι πιστᾶι κ’ ἀδόλοι ἀείδιον, IG I3 53.11–12 (433/2) [hίνα ι τὰ πάντα πι]στὰ καὶ ἄδολα καὶ h[απλᾶ παρ’ Ἀθεναίον ἐς ἀί]διον Ῥεγίνοις, and s.v. ἄδολος, ἄνευ δόλου καὶ ἀπάτης (Hdt. x3, IK Iasos x3), εἰς τὸν ἅπαντα χρόνον in Isoc. Paneg. 128, X. HG 6.5.41. For gift-giving at treaties see e.g. Hdt. 1.69–70 (Croesus and the Spartans). For the phrase ἀγλαὰ δῶρα cf. 470*. Apollo has shifted his negotiating position since the beginning of the speech (437), where he suggested that the lyre could be payment in lieu of a debt, and the idea of song-performance eliciting a gift was secondary. Now he is offering gifts and intangible favours which will cement the gods’ bonding, besides allowing Hermes to keep the cows. For Apollonius’ likely use of this line see Introduction §4.2. 434–62 Apollo asks about the lyre: Apollo at first articulates his response, characterized by the narrator primarily in terms of desire, in terms of wonder with a tinge of disapproval (436, 440, 442–3, 455). He does subsequently acknowledge the erotics of lyric (447, 455) and, by way of contrast, of pre-existing kinds of music (esp. 452, and the allusions to the names of Muses and Graces in 448–52). But the talk in 449 of using the lyre to ‘take’ desire distracts attention from his status as a victim of desire. Apollo’s wonder is directed not only towards Hermes’ precocity (as it was recently at 407, 414), but towards the appeal of the lyre – something which evidently reflects well on the hymnist’s own inheritance of lyric hymnography. Apollo is so unfamiliar with lyres, in this polemical version of events, that

402

C O M M E N TA RY 4 3 4 – 4 6 2 he wonders whether Hermes’ instrument involves a whole new Muse, since the Muses he knows are not nearly so exciting (447, 450–3). Lines 440–6 represent Apollo’s bafflement about how Hermes could have acquired the lyre (his last thought, that no Olympian knows of it, corrects the second, that it could be a gift from an immortal); this stream of consciousness cedes abruptly at 447 to a volley of short direct questions.539 Nevertheless, Apollo’s musical credentials are admitted in the form of some immediate insights into Hermes’ music. Line 436 already makes a connection to the symposium, which recurs in the mention of playing the lyre before sleep (449) and the paradoxical comparison to sympotic rounds (454); the connection has been endorsed by a similar comparison from the primary narrator (55–6). In particular, I suggested that Apollo’s contrast of the ‘paths’ of music known to him and to Hermes may present the latter’s recent song as deconstructing any neat opposition between the lascivious and pious in music – between erotic monody and theogony (448–9*). These insights also have a rhetorical force in implying that Apollo will be a worthy student of the lyre. He uses a range of politeness strategies to express his enthusiasm. Line 437 is an indirect request, and accepts the need for negotiation, which is one of Hermes’ realms; similarly in 458–62 the terms of the transaction have shifted from payment in lieu of a debt to a gift-­exchange accompanied by intangible benefits, but Apollo’s use of the language of decrees and formal treaties still plays to Hermes’ future strengths. In 457, Apollo depersonalizes to conceal how invested he is in Hermes’ response. The peppering of vocatives assert Apollo’s desire to engage with Hermes, and in 459 he jokes about Hermes’ previous false claim to ἀτρέκεια. Apollo suggests in 446 that Hermes belongs on Olympus, and in 458 that he can make him and Maia famous there. His speech marks Hermes’ progress towards fulfilment through three main intratextual patterns. First, Apollo corrects his earlier reactions to Hermes at 245, 254–9 and 282–303: Hermes no longer seems ὀλίγος (456); a reconciliation is likely (438); Apollo will help Hermes to be a leader above ground (461), with κλέος not just from nocturnal brigandage (436, 457–8). Secondly, Apollo echoes earlier descriptions of Hermes by the narrator in 13–16: he is πολύτροπος (439); his lyre brings sleep (449; cf. 14–15), and it belongs in the series of κλυτὰ ἔργα (440, 456, 458), particularly alongside Hermes’ sandals, like which it proposes an unfamiliar ‘path’ (440, 448). Thirdly, all this reflects Hermes’ ambitions as expressed in 166–72, focusing particularly on locating a productive path of τέχνη (448), and the status attached to sitting (457, 461). 539

Apollo even betrays his lack of creativity in struggling to imagine how Hermes could have imagined and built such a thing (440; 447 on τέχνη).

403

C O M M E N TA RY 4 6 3 – 4 6 5

463–95 463 τὸν δ’ Ἑρμῆς μύθοισιν ἀμείβετο κερδαλέοισιν: Echoing 260*, but since Apollo’s speech was in marked contrast to his words at 254–9 (see esp. 438, 461), the nuance of κερδαλέοισιν has shifted from cunning defence towards driving home a good bargain. Cf. 494–5*. 464 εἰρωτᾶις μ’, Ἑκάεργε, περιφραδές: περιφραδής is a rare word, whose only epic use describes roasting ‘attentively’. However at S. Ant. 348 the word describes humankind as ‘very intelligent’. Our recollection of Herm. 282, where Apollo called Hermes δολοφραδές, helps this shade into ‘cunning’ (cf. Q.S. 5.143, of Odysseus contrasted with Ajax). Apollo’s questions betrayed some musical insight (434–62*; but not, as LfgrE s.v. περιφραδής claimed, technical knowledge), but also subtlety in that they implied a request to learn, which Hermes recognizes (475). The relationship of περιφραδ- to ‘periphrastic’ appears much later. For other replies which begin by commenting on the speaker’s question see Od. 3.80 (with εἴρεαι), 5.97, 9.364 (with εἰρωτᾶις), X. Cyr. 1.6.10. The digamma of Ἑκάεργε is neglected also at Il. 22.15. αὐτὰρ ἐγὼ σοί: Matching 460 ἐγώ σε| + promise of favour. This means that σοι probably carries the emphasis, and so accent, which it generally has in Epos (though not at 368*). 465 τέχνης ἡμετέρης ἐπιβήμεναι οὔ τι μεγαίρω: Speakers of the formula οὔ τι μεγαίρω| present themselves as in control of a situation where they might easily take offence at the magnitude of an imposition. Hermes is finally in a position to respond to Apollo’s politeness (457*; the indirectness of 437–8) with a gleeful show of magnanimity.540 The fact that the subject of μεγαίρω is often a god gently underlines that Hermes’ control of the situation is based on proper recognition being accorded to his divinity. What he offers expresses a further kind of transaction to the ones Apollo envisaged in 437 and 461–2, with an emphasis on sharing: Apollo may ‘embark on’, rather than take over, a craft; ἡμετέρης either modestly underplays Hermes’ ownership (267*) or literally insists that he will retain a share. Hermes responds to Apollo’s question τίς τέχνη (447), while the metaphor of embarking picks up on τίς τρίβος (448) as well as Apollo’s promise to make Hermes a ‘leader’ (461). For the external

The way the pragmatics of μεγαίρω work vary widely according to the situation. For example, at Od. 2.235 Telemachus claims not to μεγαίρειν the suitors’ impositions, when he has exaggerated them, since the gods will punish them: this betrays hopelessness and a real lack of agency. At 8.206 Odysseus does not μεγαίρει a further contest: he will brashly ‘take them all on’.

540

404

C O M M E N TA RY 4 6 6 – 4 6 8 audience it clearly recalls 166*, where Hermes insisted to his mother that he would ‘embark on the best craft’, with a hint that this might be herding or trickery, then that he would ‘embark on’ Apolline piety (173; this idea will recur presently in 470*). Hermes has engineered a situation where the two gods can embark together on symmetrical courses. 466 σήμερον εἰδήσεις: Vying for immediacy of favours with Apollo’s νῦν … κλέος ἔσται (458). The lapidary form resembles Hdt. 7.234.2 τὸ δὲ θέλεις ἐκμαθεῖν, εἰδήσεις. ἐθέλω δέ τοι ἤπιος εἶναι: So Zeus when he tells other gods not to worry at Il. 8.40, 22.184. Both ἐθέλω (e.g. Il. 1.116, 9.674) and ἤπιος (e.g. Od. 2.47, 230) are frequently used in Epos when those in a position to be unkind choose to be benevolent. Whether or not this is a specific allusion to Zeus’s words, therefore, Hermes continues his show of magnanimity, and of the power it implies (cf. 465*). 467 βουλῆι καὶ μύθοισι: The formula is used elsewhere to describe men’s valued intellectual qualities (Il. 4.323 of Nestor, Od. 16.420 of Antinous; 13.298 of Odysseus, where Athena emphasizes his excellence in deceptive use of them). By making the formula depend on ἐθέλω ... ἤπιος εἶναι, the hymnist brings to mind behind Hermes’ offer his previous strategy of attacking Apollo with both schemes and rhetoric (including his words’ disjunction from what he really thinks). On his sincerity see also below, 463–95*. σὺ δὲ φρεσὶ πάντ’ εὖ οἶδας: The logic of this shift to the topic of Apollo’s knowledge is again shifty. The phrase implies that Hermes need not explain something – perhaps lyre-playing (cf. οἶδας ~ 466 εἰδήσεις, 474–6), perhaps his kindly intentions. However it was delivered, Hermes’ statement appears shallow after his proofs that he can outwit Apollo and invent something the latter has never encountered. But rather than sarcasm, this is better seen as flattery to get what he wants (cf. 463 κερδαλέοισιν) and a means to bring up the subject of prophecy (468–73). οἶδας| continues Hermes’ strategy of ‘matching’ Apollo’s words of admiration (464, 466): compare 456, which πάντα here tops. 468 πρῶτος γάρ, Διὸς υἱέ, μετ’ ἀθανάτοισι θαάσσεις: γάρ introduces an explanation which gets back to Apollo’s mental abilities at 471.541 But at first the flattery is more general. Apollo has the ‘first’ seat, i.e. the one on the immediate right of Zeus – another of Hermes’ reorderings of Olympus to suit his

Radermacher ignored this and instead put σὺ δὲ φρεσὶ πάντ’ εὖ οἶδας in parentheses.

541

405

C O M M E N TA RY 4 6 9 – 4 7 0 needs. By expressing Apollo’s position in terms of seating, Hermes keeps alive Apollo’s promise of an honorific seat at 461* (if the text is right), and reference to seating arrangements at the banqueting context for Hermes’ lyre (454*). The external audience recalls behind μετ’ ἀθανάτοισι θαάσσεις Hermes’ earlier contrast at 170–2, where he scorned κατὰ δῶμα ἄντρωι ἐν ἠερόεντι θαασσέμεν (172*) compared to μετ’ ἀθανάτοις ὀαρίζειν. Hermes’ apparent deference is somewhat undermined by his decision to call Apollo Διὸς υἱέ, a term which also applies to himself (cf. 378*), and which Apollo used in 455*. 542

469 ἠΰς τε κρατερός τε: Modifying the Iliadic ἠΰς τε μέγας τε|, which is used to describe manliness; ἠΰς in Epos applies to immortals only here and at Hes. Th. 817 (Briareus). The inclusion of κρατερός is disingenuous, since Hermes has used Apollo’s power and his apparent lack of it to defend himself (265, 377, 386, 418), and then given evidence of his own strength (407, and his epithet κρατύς). φιλεῖ δέ σε μητίετα Ζεύς: Cf. Il. 2.197 φιλεῖ δέ ἑ [a king] μητίετα Ζεύς – ­perhaps the phrasing was formulaic. Hermes brings up the possibility of φιλία between gods (cf. 382, 507*), and presents Zeus as having μῆτις, rather like himself. 470 ἐκ πάσης ὁσίης: Recalling, like 465 ἐπιβήμεναι and 468 θαάσσεις, Hermes’ envy of Apollo’s status, which he hoped to match by adopting Apollo’s ὁσίη (173*). I interpreted the noun in these two passages to amount to the avoidance of offending the gods, and to betray a down-to-earth viewpoint about religious propriety. ἐκ (LSJ s.v. III.6) indicates the grounds for friendship, as at e.g. Od. 15.196–7, X. Hipp. 6.2, Ps.-Pl. Epist. 6 333e. Here Apollo’s proper behaviour is quantified by πᾶς ‘complete’: this again seems like flattery when read against 370–5 and other sources (see e.g. Felson 2012 on his brashness in H.Ap.). ἔπορεν δέ τοι ἀγλαὰ δῶρα: Hermes’ hint that Apollo should show him φιλία (469 φιλεῖ) and gifts, as Zeus does to Apollo, becomes clearer, especially with the echo of Apollo’s promise of ἀγλαὰ δῶρα from 462. The phrasing is

542

For the seat to the right of the host as an honour, see Il. 24.100 (Athena then Thetis beside Zeus), Od. 7.169–71 (beside Alcinous), 21.141–2, 245–68 (contest proceeds to right, and Eurymachus and Antinous are last in line), Pi. fr. 146 (Athena δεξιὰν κατὰ χεῖρα πατρός | ἡμένη), Pl. Smp. 175c, 194a (Agathon saves the space to his right on his couch for Socrates, who is therefore last to speak), Luc. Smp. 9–11 (seat of honour is to the right of host). Other types of ‘sitting first’, e.g. in the first rank of an assembly (Il. 19.50, Od. 2.14, Ar. Ach. 25, etc.), seem less pertinent models for a small Olympian gathering. For Hermes ordering Olympus see 128–9*, 425–33*.

406

C O M M E N TA RY 4 7 1 – 4 7 2 formulaic (cf. Hes. Th. 412 of Zeus’s honours to Hecate, Od. 16.230, etc.) but deployed with precise effect. 471–2 καὶ τιμάς· σέ γέ φασι δαήμεναι ἐκ Διὸς ὀμφῆς | μαντείας, Ἑκάεργε: We begin with the text. What matters to Hermes is that Apollo has received certain τιμαί, not that (like others) he knows the gods’ τιμαί. The word-order and the absence of μαντεῖος from Epos (whereas the noun μαντείη occurs a further three times in Herm.) mean that τιμὰς μαντείας cannot go together. Hence one must, with M against Ψ, punctuate after τιμάς. The common pairing δῶρα καὶ τιμαί is the divine equivalent of the complementary relationship familiar especially from the Iliad between tangible awards and esteem; cf. H Dem. 327–8, and among humans Pl. R. 2 361c, X. Oec. 4.7, Cyr. 2.2.20, etc. The clause starting at σέ explains the nature of Apollo’s special τιμαί, and it is Ψ’s γε rather than M’s δέ which expresses this.543 Hermes refers to sources (φασι), which he will contrast with personal ex­­ perience in 473. But he has no obvious sources for Apollo’s prophetic knowledge, and he appeared to know something of Delphi already at 178. He is concealing his degree of personal interest in Apollo’s prerogative, opening it up for discussion rather than asking for a share in it directly (see 532–3*, and compare 457*). Especially in Epos, ὀμφή is the voice with which a god instructs humans – whether through prophetic interpretation of signs or another medium. It is particularly associated with Zeus (here, 532; Il. 2.41 Agamemnon’s dream; 8.250 and Hes. fr. 150.12 Zeus Panomphaios; also Pi. fr. 70b.29) and Delphic Apollo (Herm. 543–5, Thgn. 808, E. Ion 908, S. OC 102). ἐκ Διὸς ὀμφῆς shows that μαντείας refers to prophetic utterances (as in e.g. Tyrt. fr. 4.2) rather than Apollo’s prophetic practices (as it does in 547, 556). That Apollo is the προφήτης of Zeus’s plans is widely attested, notably as one of Apollo’s first declarations in the Hymn to Apollo, where he eventually founds his oracle at Delphi (132 χρήσω δ’ ἀνθρώποισι Διὸς νημερτέα βουλήν).544 But Hermes is also picking up on one nuance of Apollo’s use of ὄσσα, the object of the last occurrence of the key-word δαήμεναι (443–4*). Apollo presented the

The construal of τιμάς as the object of δαήμεναι seems to be inherited, since it probably caused the insertion after 472 μαντείας of τ’. (M’s θ’ corrects this before a rough breathing, but is incompatible with its readings in 471.) This construal was more sustainable with γε than δέ, which would be misplaced. M’s branch attempted to fix not only τ’ but also the punctuation, and in this process γε had good cause to turn into a connective δέ. 544 For Herm.’s meticulous use of H.Ap. 131–2 see Introduction §1.3. At A. Eu. 616–18, Apollo declares οὐπώποτ’ εἶπον μαντικοῖσιν ἐν θρόνοις ... ὃ μὴ κελεύσαι Ζεύς. Σ S. OC 793 mentions further parallels in Aeschylus, Sophocles and Aristophanes. 543

407

C O M M E N TA RY 4 7 2 – 4 7 3 sound of the lyre as a divine ‘voice’ bestowed on Hermes to pass on to others. This plays into Hermes’ hands, since a lyre-for-prophecy swap was probably already on his mind in 172–81 when he focused on Delphi among Apollo’s τιμαί. He accepts the κληδών and the opportunity to connect lyre-playing and prophetic knowledge. Ἑκάεργε is Hermes’ third vocative (see 464, 468), and three more will follow. For multiple vocatives as a feature of conversation involving children see 439*. The first three are equally spaced during the first part of Hermes’ speech. All six focus attention on key elements of the negotiation: 464 on Apollo’s implicit request; 468, here and 477 on what he has to offer; 490 on what Hermes will give and 492 on what he will gain. 472 Διὸς πάρα θέσφατα πάντα: These words should be punctuated as a parenthesis to avoid ἐκ Διὸς ... Διὸς πάρα in the same clause. (See Matthiae 1800: 295, though he preferred to change πάρα to δ’ ἄρα.) For the phrasing cf. Hes. fr. 193.8 Ζηνὸς πάρα θέσφατα. θέσφατα are divine declarations about the future, in principle distinguishable from the various processes that reveal them to mortals (see e.g. 534), but occasionally equated with oracles (e.g. Od. 9.507, 11.151). The emphatic position of πάντα, and the verbless syntax, show that Hermes is attempting to generalize his previous comment: Zeus’s voice is the source not only of the Apolline oracles Hermes has in mind, but of all θέσφατα. Whether Zeus’s declarations constitute or publish fate (see Dietrich 1965), Hermes’ suggestion is that Zeus has unique access to fate and that his words have unique authority. Apollo will later mention a process of revelation outside Zeus’s control (557*), but Hermes’ comment should be evaluated more in terms of his rhetorical goals, namely: to focus on how the two brothers’ relationships with their father are different; to flatter Apollo, since it is impressive that his oracles depend on Zeus; to demonstrate his own insight into and aptitude for a possible role in this most significant form of prophecy. 473 καὶ νῦν αὐτὸς ἐγώ γε παναφνειὸν δεδάηκα: Scholars have exercised much ingenuity on the textual problem. νῦν αὐτὸς ἐγώ … δεδάηκα contrasts with 471 σέ γέ φασι δαήμεναι: ‘I know’/‘you know’/‘I have just found out’/‘they say’. ἀφνειόν picks up 470 δῶρα as μαντείας did τιμάς, chiastically; the echo of 171 suits the nexus of intratexts between that speech and this passage (see 465*, 468*). Of the variants at the start of the line, καί is unproblematic whereas x’s τῶν is awkward: Hermes has not learnt that Apollo is ‘wealthy in them [oracles]’, which is anyway a strained construction (compare Pindar N. 1.15 πολίων ἀφνεαῖς of Sicily?); nor can one take τῶν νῦν together, since the asyndeton would be inappropriate. By contrast Hermes has encountered Apollo’s wealth, in the form of the cows and the

408

C O M M E N TA RY 4 7 4 – 4 7 5 promise to make Hermes ὄλβιος with gifts (461–2). The corruption lies in the syllables before ἀφνειόν, where Ω’s -γε παῖδ’ does not scan. I have tentatively conjectured γε παναφνειόν, the latter not an attested word but comparable to παμπλούσιος and πανόλβιος. This resembles Evelyn-White’s μάλ’ ἀφνειόν but posits a simpler corruption. Vergados accepted Allen’s παῖ ἀφνειόν (1897: 265–6). While there are some parallels for addressing an elder sibling as παῖ (Dickey 1996: 69), coming from an infant this still seems too flippant for the prevailing tone as I see it, namely flattery with unflattering undercurrents. Groddeck’s παῖ Δαφναῖον (1786: 92) was worse, since Hermes has not learnt that Apollo is Daphnaios; West’s παραὶ δαφνέων (1966a: 150) imported a different species of invasive laurel. Hermann also changed γε to σε, but the latter is easily understood here from 471. Allen and Sikes invented the Aeolicism πεδάφνειον ‘quickly’ (cf. Bothe’s τάδ’ αἰφνίδιον). Shackle (1915: 162–3) proposed τῶν ... ἔγωγ’ ἔτι παῖς ἄφενος, after sensible discussion. Radermacher’s τῶν ... πέρ’ ἀφνειόν keeps τῶν at the expense of an anomalous elision and hyperbaton of περί. In Pötscher’s παῖς γ’ for γε παῖδ’ (2004a) it is hard to explain why γε got transposed. 474 σοὶ δ’ αὐτάγρετόν ἐστι δαήμεναι ὅττι μενοινᾶις: The repetition of this line at 489 is unusual and so arouses suspicion, but it is pertinent in both places, the flow is not greatly improved or harmed by removing either, and it is tricky to account for duplication of a line in the paradosis. I shall interpret the effect with both lines present. Hermes returns to 467* πάντ’ εὖ οἶδας, after explaining those words with Apollo’s status and specialist knowledge, and rephrases his point using the theme-word δαήμεναι, which has occurred twice in the interim. This verb now unites the learning of oracles to the learning of lyric, a central idea in the remainder of the hymn. Apollo will not need practical tuition in the lyre since he can independently pick up any subject; natural talent is not, in other words, a necessary prerequisite for this kind of music (as Apollo first thought: 440), but is a sufficient one in Apollo’s case. αὐτάγρετος may continue the back-handedness of Hermes’ speech, by pointing to his own unconstrained ability to lay hold of Apollo’s cows, and to do things – make sandal-tracks, hide cows – which Apollo cannot comprehend.545 475 ἀλλ’ ἐπεὶ οὖν τοι θυμὸς ἐπιθύει κιθαρίζειν: Càssola placed a paragraph break before the line, but 475–6 apply the generalization of 474 to the specific case of lyre-playing, and so belong closely with it. For Apollo’s θυμός in the For the latter point see Abramowicz (1937) 79. αὐτάγρετος was already at Od. 16.148 preferred to αὐθαίρετος. An etymological relationship to ἄγρα, ἀγρεύω ‘I capture’, etc. is unclear, but a synchronic one is likely: see Beekes (2010) s.v. ἄγρα.

545

409

C O M M E N TA RY 4 7 6 – 4 7 7 sense of ‘impulse’ to play the lyre see 457*, to which Hermes responds here. Metaphorical ἰθύω describes desire also at e.g. Od. 11.591, 22.408, Archilochus fr. 196a.3 θυμὸς ἰθύει (with assonance). Here it extends a series of musicas-path metaphors: cf. 452 οἷμος (with further intratexts in 476*), 448 τρίβος, 465 ἐπιβήμεναι. 476 μέλπεο καὶ κιθάριζε καὶ ἀγλαΐας ἀλέγυνε: The three imperatives echo Apollo’s enthusiasm for the Muses’ ἀγλαὸς οἷμος ἀοιδῆς, μολπή and αὐλοί from 451–2, with the last replaced now by the lyre; the following lines will confirm that Hermes’ description of the lyre aims to fit it into the realm of the Muses, which Apollo already knows. κιθάριζε, so soon after 475 ἐπιθύει κιθαρίζειν, reaffirms the claim of 474 that Apollo’s desires for knowledge are instantly achieved. As often, the third member of the tricolon is longer. ἀγλαΐας ἀλέγυνε reciprocates Apollo’s offer of ἀγλαὰ δῶρα (462), with an assonance paralleled in H.Aphr. 11 ἀγλαὰ ἔργ’ ἀλεγύνειν (of Athena). Choral performance has a prominent place in ἀγλαΐη (451*), but various passages emphasize the ἀγλαΐη of a feast (e.g. Od. 15.78, Pany. fr. 16.14), which was a formulaic context for the verb ἀλεγύνω (cf. 361*), as well as the forum suggested for Hermes’ lyre-playing (55–6, 449, 454). This pair of connotations helps Hermes in 480–1 to present his lyre as embracing both choral and sympotic performance. 477 δέγμενος ἐξ ἐμέθεν, σὺ δέ μοι, φίλε, κῦδος ὄπαζε: The offer involving δέγμενος will be fulfilled at 496–8 ἐδέξατο ... ἔδεκτο. The hymnist seems to have interpreted the obsolete athematic forms as aorist, though they are more frequently imperfective (GH i.296). Hermes uses an emphatic to-and-fro of pronouns (ἐμέθεν, σύ, μοι) to suggest both the interpersonal bond and the reciprocity of the exchange of favours.546 While his offer clearly recalls Apollo’s proposal at 461 of κῦδος as a reward for (implicitly) the lyre, the structure of the sentence prompts one to take ἀγλαΐας as the object of δέγμενος. Framing the swap in terms of ἀγλαΐαι and κῦδος pushes it further away from Apollo’s initial conception of a transaction of objects (n. 505), and makes the reciprocity seem natural thanks to traditional phrases such as Il. 6.509–10, Od. 15.78 κῦδός τε καὶ ἀγλαΐη, A.R. 1.287 ἀγλαΐην καὶ κῦδος. However, two elements jar in the offer. φίλε without the support of ὦ or a noun is strikingly rare, and a few passages suggest it could take on a

West’s δ’ ἐμοί seems to lay it on too thick: σὺ δέ μοι φιλ- is transmitted in similar requests at Thgn. 99 and Ar. Eccl. 970 (and without φιλ- at e.g. Od. 9.365, Thgn. 4).

546

410

C O M M E N TA RY 4 7 8 throwaway or casual tone.547 Hermes at 382 pointedly excluded Apollo from being his φίλος, and nothing has explicitly overthrown this. φίλε therefore constitutes a potentially offensive reading of the gods’ current relationship. Moreover, elsewhere in Epos κῦδος is bestowed on victors by a god or by one of their human conquests (Il. 15.644, 22.57). Can Hermes ask for κῦδος from his opponent without evoking a sense that he may be ‘defeating’ Apollo?548 478 εὐμόλπει μετὰ χερσὶν ἔχων λιγύφωνον ἑταίρην: The asyndeton and echo of μέλπεο signal the start of an expansion on what Hermes has just said: in fact, 478–9 expands on μέλπεο, 480–2 on ἀγλαΐας, 482–8 on κιθάριζε, and 489–92 on the swap in line 477. The lyre is personified as a ἑταίρα one can embrace. Hermes instantly connected the tortoise through the lyre to a sympotic courtesan (28*, 31* including ἑταίρη|); the idea now recurs, to complement Apollo’s highly eroticized response to the lyre since 421, and his identification of its sympotic potential.549 Unlike Apollo’s Muses, whose focus on group performance was emphasized in 451–2, this courtesan has the potential to join in the ἐνδέξια ἔργα on her own.550 In λιγύφωνος, λιγυ- is apt for both the metaphor’s tenor and its vehicle, whereas -φωνος is more distinctly animate.551 An audience may also have connected the adjective to λιγύφθογγος, restricted in Epos to κήρυκες and so relevant to Hermes. Apollo has called Hermes ‘ἑταῖρε of feasting’ (436), and we shall see a further connection of instrument and maker in 479*. Finally, in the context of the negotiation between the two gods, the courtesan imagery now gains the extra significance that her services come at a price.

A.R. 4.1073, 1086 has both bare vocative φίλος (also rare) and φίλε, during pillow-talk. Alcmaeon in Stes. fr. 93 is pretending to be casual in front of Adrastus, though already viewing him with suspicion. In Call. H.Del. 213 the brevity is part of Leto’s panting rhythm while in labour. But other passages are less marked: e.g. Thgn. 99, Euenus fr. 9, Theoc. 5.73, Call. 32 G–P, and the trope in epitaphs of addressing the onlooker as φίλε (e.g. Posidippus 103–4, and numerous inscriptions). Plural φίλοι is used much more freely as a bare vocative. 548 For this train of thought see also Kahn (1978) 160–1, Clay (1989) 141, Jaillard (2007) 78–80. 549 449*, 454*. For innuendos about ‘embracing’ one’s instrument from more recent contexts, see Zecher and Solie cited in n. 50. 550 This is the more notable since Greeks often thought of legendary female musicians operating in groups, but of male soloists (think of the Muses and Sirens, or the latter’s battle with Orpheus at A.R. 4.891–909). 551 λιγύς and lyre-playing: 425, LfgrE s.v. 1c. λιγύς and female song: e.g. Od. 10.254, 12.44, Th. 275, Sc. 206, Hy. 14.2, H.Pan 19, 24, and frequently of threnody. φωνή is rare of musical instruments: LSJ s.v. I.4; for the conceit that a tortoise acquires a φωνή in death see 25*. 547

411

C O M M E N TA RY 4 7 9 – 4 8 0 εὐμόλπει: The hapax is fashioned so as to initiate a sequence of four instances of εὐ(-) over the next five lines. The verb is based on εὔμολπος, itself rare except as a name; Paul the Silentiary uses it of Apollo at AP 9.396.6; a hymnist at Delphi may have gone by Claudius Εὔμολπος (FD III(1) 210). The more famous Eumolpos was the supposed son of Musaeus and recipient of his Ὑποθῆκαι (Musaeus T17–18), father-figure of the Eumolpidae, who among other things controlled exegesis of Eleusinian rituals (see Hitch 2011: 135–6, Daverio Rocchi 1996). A wide Greek audience at the time of Herm. might therefore have connected εὐμολπ- to both poetry and theological learning. By reading Hermes’ instruction as ‘be Eumolpos’ in a sympotic setting, we obtain a third paradoxical combination of sympotic poetry and narratives about the gods (see 55–6, 454, Introduction §3.4); a reference to the Eumolpidae could suggest collaboration with the Kerykes, whom Hermes fathered at Eleusis (E. fr. 370.113–14, Androtion FGrH 324 F1). 479 καλὰ καὶ εὖ κατὰ κόσμον ἐπισταμένην ἀγορεύειν: Barnes’s correction ἐπισταμένην is very suitable, so that the lyre’s personification continues through to ἀγορεύειν, which reveals that her words as well as her songs are worth listening to. Ω’s ἐπισταμένως was probably based on such phrases as 390 εὖ καὶ ἐπισταμένως.552 That intratext is indeed important in confirming the hint in 478* λιγύφωνον that the lyre has similar rhetorical gifts to its inventor, while κατὰ κόσμον recalls a quality of Hermes’ second song (433). The qualities picked out by καλὰ … ἀγορεύειν and εὖ κατὰ κόσμον are poised between social propriety of content, the aesthetic pleasure of the listener, rhetorical skill, and poetic construction. ἐπισταμένην grounds them in the technical competences of an ἐπιστήμη. The implication that the lyre’s talk is ethically valuable adds a new point which, while suited to much sympotic elegy, paves the way for the lyre-courtesan to leave the symposium in 481 (χορόν) and to morph into a new kind of speaker in 482–4*. 480 εὔκηλος μὲν ἔπειτα φέρειν: I prefer a stop before this clause, with φέρειν standing for an imperative (see 92–3*; the construction often follows an imperative and has a nominative predicate). One could alternatively place a comma after 479, and construe φέρειν as epexegetic after εὔκηλος, but this produces a straggling sentence. Either way, ἔπειτα means ‘hereafter’ (e.g. Il. 15.49, 20.136) rather than ‘then’, since 480–1 specify the settings for Apollo to hold his companion (478–9), rather than adding new ones.

Velareus (1528) canendi arte imbutus appears to have translated ἐπιστάμενος. If that was an attempted correction (one accepted by Vergados), it scans only with difficulty.

552

412

C O M M E N TA RY 4 8 1 μέν throws emphasis on εὔκηλος.553 Such emphasis suits the assonantal pattern |εὐμόλπει ... |εὔκηλος ... |εὐφροσύνην (see also 479 εὖ), and indeed εὔκηλος is a deformation of ἕκηλος designed to emphasize its positive quality (see Beekes 2010 s.v. ἕκηλος). ἕκηλος often describes peace of mind at a feast, both in Epos and in the patchwork of traditional dicta about hospitality in Pherecrates fr. 162, where it co-occurs with δαῖτα θάλειαν. But here Hermes is simultaneously foregrounding the calming nature of lyre-music over its potentially damaging desirability (421–2, 434, 447), and contrasting Apollo’s future with a past in which he was not always εὔκηλος, including within Herm. itself (Agar 1929: 21). ἐς δαῖτα θάλειαν: The first place where Apollo might take the lyrecourtesan clearly recalls the θαλίαι envisaged for it at 56 and 454; the lyre is δαιτὶ συνήορος ... θαλείηι at Od. 8.99. Cf. the connection of music and the adjective τεθαλυῖα in 452*. Hermes thus sticks closely to Apollo’s descriptions of both the lyre and the Muses’ activities, confirming the former and making the lyre a natural fit for the latter (see below, 463–95*). More immediately the phrase signals the start of an elucidation of the instruction ἀγλαΐας ἀλέγυνε (476*), where ἀλέγυνε alluded to contexts of commensality, and ἀγλαΐη referred to splendour and elation both at feasts and in contexts of choruses and κῦδος, which appear in 481. We see Apollo performing to the lyre at gods’ feasts at e.g. Il. 1.601–3. 481 καὶ χορὸν ἱμερόεντα: The formulaic epithet is always contextually pertinent: Il. 18.603 of mixed youths and girls; Od. 18.194 of the Graces; Hy. 6.13 of the gods including the bejewelled Horai; Sc. 280 of a girls’ hymenaeal chorus; and, with modification, Hes. Th. 7–8 of the Muses’ dancing-ground, near Himeros and the Graces (63). As with 480 ἐς δαῖτα θάλειαν, a formula is enriched by its context. ἱμερόεις suits the courtesan metaphor, and brings her sex-appeal from the private cuddle of 478 into public and potentially ritual genres (where, strictly speaking, the more resonant kithara would have been the norm). Both ἱμερόεις and χορός contribute to Hermes’ echoes hereabouts of the Muses’ traditional activities in 451–2. For Apollo’s lyre-playing accompanying the Muses’ chorus and dancing by other gods see e.g. H.Ap. 186–206. καὶ ἐς φιλοκυδέα κῶμον: This phrase moves us away from natural contexts of Apollo’s own lyric performances, though Hermes’ instruction has significant implications for his own conception of lyre-playing, as well as proleptic relevance for human lyre-players. κῶμος refers to a range of celebratory contexts for young male singing, of which the most frequently documented are (a) an advanced stage of sympotic revelry, whether based indoors For this use of μέν see Denniston 359–61, Il. 1.514, Od. 4.100 πάντας μὲν …, etc. Ilgen’s μιν produces two asyndeta in a row without conferring any benefits.

553

413

C O M M E N TA RY 4 8 2 or moving from house to house, and (b) the celebration of an athletic victory.554 The former is a place for hubristic insults (a link to κωμ-ωιδεῖν must have been available long before comedy was institutionalized at Athens), the latter for en-com-ium. The word therefore encapsulates Hermes’ ability, seen in his s­econd song, to marry praise-poetry with sympotic banter. The lyre’s sympotic role has been hinted at repeatedly, and lyres are found in sympotic κῶμοι at e.g. E. Cyc. 40 (the barbitos), Lissarrague (1990) figs. 2, 104; to this extent, Hermes casts Apollo as a rather undignified drunken komast. However, despite the political positioning which we discern in the symposia of (say) Alcaeus or Theognis, φιλοκυδέα distinctly shifts the balance of this κῶμος in favour of its epinician sense, as one of the vehicles (another being Hermes’ herald) for expressing the κῦδος of victory; for the kithara’s epinician credentials see Pindar O. 1.17–18, 2.1 and frequently.555 The adjective clearly recalls its only other attestation, in line 375, where Hermes grounded Apollo’s bullying tactics in φιλοκυδὴς ἥβη: the hubristic youth (a typical association of the post-sympotic κῶμος) is converted into a musician participating in epinicians, the model perhaps of lyric poets at Apollo’s Pythian Games. Pindar similarly keeps a Pythian κῶμος poised between revelry and Apollonian performance when describing it at Pythian 5.22–3 as τόνδε κῶμον ἀνέρων, Ἀπολλώνιον ἄθυρμα. However, the epinician context points not only to Delphi but also to the performance of Herm. at Olympia (in my view: Introduction §2), and to Apollo’s offer, recently accepted, to give Hermes κῦδος in return for the lyre (461, 477*). The latter line perhaps contained the cheeky implication that Apollo is to bring Hermes κῦδος by acknowledging that Hermes has outdone him – an acknowledgement which our line would situate in the performance of a song of praise on the lyre. 482 εὐφροσύνην νυκτός τε καὶ ἤματος: Hermes now affirms Apollo’s impression at 449 that the lyre was special in offering εὐφροσύνη and ὕπνος together. Of the three contexts in 480–1, the δαίς and κῶμος point more to evening performances, χορός to the day. But ‘night and day’ here points to more than the lyre’s constant utility: earlier in the hymn, Hermes operated against Apollo at night, and dawn brought Apollo (perhaps already being aligned with the sun)

554

For other contexts see e.g. Budelmann (2012) 186, Agócs (2012), with further bibliography. Agócs observed the shifting nature of Pindar’s komastic vocabulary – at times referring to the present performance as a κῶμος, at times presenting it metaphorically in terms of a sympotic κῶμος, at times contrasting it with a κῶμος in the evening of the victory, and so on. 555 Klein (1980) 135 referred κῶμος here only to the symposium. The scribe of p wrote φιλομειδέα (an adjective used in Epos only of Aphrodite and a nymph): this may be rooted in confusion of minuscule κ/μ.

414

C O M M E N TA RY 4 8 3 – 4 8 4 into the narrative (182–9*). The cosmic opposition night/day now expresses not the gods’ opposition, but their shared interest in the lyre. 483 τέχνηι καὶ σοφίηι δεδαημένος ἐξερεείνηι: Apollo treated citharody as an object of δαήμεναι or learnable practical knowledge (444*, after the false start of 440–2), then asked about its τέχνη, Μοῦσα and τρίβος (447–8). Hermes acknowledges the importance of τέχνη first. σοφίη, though a Homeric hapax (Il. 15.412), seems to have covered many forms of intelligence by Herm.’s likely date. For the application to lyre-playing see ‘Hesiod’ fr. 306, calling Linus παντοίης σοφίης δεδαηκότα, and GDRK 16.20 Μουσάων σοφίης δεδαημένος (third century ce; after a hymn to Hermes which shows familiarity with our story); at 511 σοφία will be explicitly restricted to an individual instrument.556 (ἐξ)ερεείνηι (already in 252, 313) makes a cluster of five appearances from here to the end of the hymn. In the terms of the lyre’s personification, the musician eliciting a sound from an instrument becomes someone questioning with intelligence – a new touch developed in the next line. 484 φθεγγομένη παντοῖα νόωι χαρίεντα διδάσκει: φθεγγομένη and διδάσκει continue the personification; for the former, cf. E. Ion 881–2 ἑπταφθόγγου  ... κιθάρας. In 483 the δεδαημένος performer asked a skilful question of the instrument; here he learns (δα-) more, through the answer. While 479 (καλά, κατὰ κόσμον) already suggested the propriety of the lyre’s talk, the role of teacher is a significant development for the courtesan-lyre, marked by the fact that what she teaches offers ‘all sorts of χάρις’, not just for the senses (and Hermes’ previous lines have echoed Apollo’s allusions to all three Graces – Aglaia, Thalia and Euphrosyne; cf. 449–54), but also for the intellect.557 As a teacher, the lyre moves towards two other female figures. One is the Muse, who answers questions from the epic performer at e.g. Il. 1.8, 2.484–7, teaches at e.g. Hes. Th. 22, and has been engaged in providing Hermes with χάρις since the hymnist invoked her in Herm. 1; Sappho (fr. 118) also questioned her χέλυς like a Muse and received an answer. In this reading, Hermes is responding to the second of Apollo’s questions from 447–8 (in between more obvious answers to the other two): his erotic lyric

The primary relationship of σοφία to (a) manual crafts and (b) music in epic and lyric is explored particularly by Maier (1970) 13–90, with our line on p. 22. For the common combination of τέχνη and σοφία see e.g. IG I3 766 (500–480). Later, Maximus Init. 439–40 τέχνης ... δεδαηκότας ... ἢ σοφίης, 452 τέχνην ἢ σοφίην δεδαήμεναι could derive from our line, despite the different syntax. 557 The νόος is rarely attuned to pleasures of the senses in Epos, though parallels exist such as Hes. Th. 51, where the Muses τέρπουσι νόον Διός. For νόος and χάρις see LfgrE iii.423.41–4. 556

415

C O M M E N TA RY 4 8 5 – 4 8 6 does not require one to make the acquaintance of a separate Muse, and so the instrument itself can encroach on that role. However, this knowledgeable respondent also resembles a prophetess, a point Apollo will show awareness of when he echoes these lines in 543–9*; he has already laid the groundwork in 442* θέσπις and 443* ὄσσα for the idea that the lyre’s voice is prophetic, as has the narrator in 427 κραίνων (425–33*). Hermes is fostering such an interpretation of his lyre–courtesan–Muse, to further his ambition of swapping it for a prophetic power of his own. As with the initial leap from tortoise to lyre, bargaining-chip and courtesan (24–31), the hymnist challenges us to keep up with Hermes’ ways of imagining the lyre (see Introduction §3.4). 485–6 ῥεῖα συνηθείηισιν ἀθυρομένη μαλακῆισιν, | ἐργασίην φεύγουσα δυήπαθον: Ten metrical feet in just seven dense words, of which δυήπαθον is a hapax, and συνηθείηισιν and ἐργασίην are hapaxes in epic.558 One of Hermes’ main emphases here is on ease, continuing the point made in 480 εὔκηλος. The lyre is an object of τὸ ἀθύρειν, as it (and the tortoise) was an ἄθυρμα at 32, 40, 52. The verb suggests leisured play, often childish (as suits Hermes), often musical (e.g. H.Pan 15, Pl. Lg. 7 796b, A.R. 3.949), and sometimes erotic (both musical and erotic at anon. AP 9.505d and possibly Erinna SH 401.20). The last two connotations are mobilized here for the lyre–courtesan metaphor. All three deflate contexts like the φιλοκυδὴς κῶμος (481) in a similar way to Pindar I. (3/4).57, where Homer is said to have left praise-poetry for future generations to ἀθύρειν. With the verb, ῥεῖα suggests not only technical facility (cf. dancing ῥεῖα at Il. 18.600, Od. 8.376), but also being ‘at ease’ (see esp. Hy. 20.5–7, where living ῥηϊδίως εὔκηλος all year round is the result of Hephaestus’ τέχνη and δαῆναι; also Od. 1.160 in the context of sympotic music). This easy toying comes from συνηθείηισι ... μαλακῆισιν. The noun refers to longstanding personal acquaintance (with people or habits), so that Hermes now addresses the third of Apollo’s questions from 447–8, the τρίβος of making progress with the lyre over time. Beside the courtesan metaphor, it is likely that συνήθεια also refers euphemistically to sexual relations, though the parallels are much later.559 After all, the eroticism of the passage is reinforced by the tactility of μαλ(θ)ακός, which is often eroticized in marked uses – such as here when describing an abstract noun. See e.g. Pi. fr. 122.8 (of the season

It is tempting to read δυηπαθέ’ (Zumbach 1955: 23), but s-stem n ­ euters spor­­­­adically formed 2nd-declension compound adjectives: see Lobeck (1837) 243–6 on e.g. Il. 16.428 ἀγκυλόχειλαι. 559 LSJ s.v. I.1b. At X. Cyr. 6.1.31 the variation συνουσίας (y) vs συνηθείας (z) may well date back to imperial times or earlier: Persson (1915) 165–7. 558

416

C O M M E N TA RY 4 8 7 for sex), A. Ag. 742 (of Helen’s look), Ar. Nu. 979 (of the voice when flirting), X. Mem. 3.11.10 (the hetaira Theodote’s kissing). The lyre-courtesan for her part shirks ἐργασίη which is δυήπαθος.560 The contrast in register between high-flown adjective and prosaic noun contributes to the humour. As a teacher the lyre avoids manual labour, and has been aligned with leisure. Equally, if one takes δυήπαθος as a predicate specifying one kind of work, then the lyre prefers a gentle touch to a painful one, while as a courtesan her ἐργασία could be represented as a cheerful contrast to the normative female ἐργασία of chaste weaving (X. Oec. 7.21, Pl. Pol. 280e, and frequently). For prostitutes’ ἐργασία see e.g. D. Cor. 129, Aeschin. 1.124, Ps.-D. Neaer. 36, 113, Ps.-Luc. AP 6.17, ἐργάτις in Archil. fr. 208, Call. Ia. 8 fr. 222 (of Simonides’ Muse), and especially Nicarchus(?) 2 G–P, where Nikarete abandons Athena’s ‘hungry’ ἔργα for Aphrodite’s ἐργασία, which involves music, κῶμοι and θαλίαι. These lines bring to a climax Hermes’ conceit of the lyre as courtesan, and they carry ambivalent connotations: the lyre, while alluring, is offered in a way which poses a challenge to Apollo’s vaunted virility and propriety. Pindar can present his work as μαλακός, as part of a luxurious κῶμος (e.g. O. 2.90, N. 4.4), but sources more often emphasize the effeteness of being μαλακός (e.g. Il. 17.588 and X. Oec. 1.19, where the application to ἀργία resembles the lyre’s shunning of ἐργασίη). μαλακός could be used to denigrate an excessive focus on musical education (E. Su. 883), or decadent scales in several music theorists (LSJ s.v. III.e). One’s initial surprise that Apollo might associate with a hetaira is increased when her job is described in a way which opposes it to the ‘proper’ hard work of a woman. 487 νῆϊς ἐὼν τὸ πρῶτον ἐπιζαφελῶς ἐρεείνηι: The second player is νῆϊς, both ‘ignorant’ (e.g. H Dem. 256) to suit the questioner-metaphor, and technically ‘inexpert’ (e.g. Od. 8.179; the opposite of Choerilus fr. 2.1 ἴδρις ἀοιδῆς); the single word is thus opposed to τέχνηι καὶ σοφίηι and δεδαημένος from 483. That phrase, and 485 συνηθείηισιν, implied familiarity acquired over time, opposed here by τὸ πρῶτον. And ἐπιζαφελῶς opposes μαλακῆισιν: the stem ἐπιζαφελ- otherwise describes anger (Il. 9.516, 525, Od. 6.330, A.R. 4.1672), and while ancient scholars had to seek out the sense through etymological speculations, the hymnist seems to connect the word sonically to περιζαμενῶς (495, also of anger).561 The image of questioning ‘angrily’ seems suitable

For a similar collocation, see Antipater of Sidon 57.1 G–P δυηπαθὲς ἐργάτα, of an ant. 561 For etymologies see Apion fr. 36 ‘angry’, Ap.S. p. 73 ‘huge’ (ζα- + ὀφελλ-), Porph. QH 1.79 ‘durable’ (~ ἀφελεῖν?), Hsch. ζ.74 ‘very rough’ (ζα- + φελλεύς?), Suda ζ.22 ‘very 560

417

C O M M E N TA RY 4 8 8 in metaphorical terms for rough technique, though perhaps the hymnist intended a sense such as ‘aggressively’. For the pattern of following ἐξερεείνηι with the simplex ἐρεείνηι, see Radermacher 158, Diggle (1994) 84 n. 64. 488 μὰψ αὔτως κεν ἔπειτα μετήορά τε θρυλίζοι: μάψ developed from ‘greedily’ to ‘unrestrainedly’ to ‘ungroundedly’; for the last, which makes it a counterpart for μετήορα, see e.g. Il. 15.40 (oath), 20.348 (boast, also with αὔτως), Od. 14.365 (lie). Scepticism of philosophical speculation about the heavens meant that cognates of μετήορος, at least by the late fifth century, could describe ‘head-in-the-clouds’ or groundless thought and discourse.562 The lyre’s teaching is finally described by the extremely rare verb θρυλίζω.563 This contains the often derogatory tone of θρυλέω ‘drone on’, but it can hardly be a coincidence that θρυλιγμός has a specific musical sense at D.H. Comp. 11.8, of a wrong note (probably a squeak) produced by a professional aulete’s embouchure failing.564 Striking a string with unwonted force deforms its normal quality, as noted in Ps.-Arist. Aud. 803a30. Further evidence of a musical sense may exist in Hdn. Gr. i.443, where ‘θρυλίζω, λυρίζω’ stands at the head of a sequence of musical verbs in -ίζω. The teacher-lyre thus responds to clumsy and angry fumbling with an out-of-tune drone (θρυλ-), and words which lack substance (μὰψ αὔτως ... μετήορά τε: for the irregular combination cf. 412*). All this makes an effective opposition to 484 παντοῖα νόωι χαρίεντα. Görgemanns (1976: 124–8) read lines 486–8 as attacking an identifiable (more or less Pindaric) lofty style contemporary to Herm., but the vagueness of the description encourages one to focus more on the terms laid out within the hymn. The words selected here to describe the unsatisfactory content of brazen’ (ζα- + ἀφελής). Such speculation probably underlies Nicander’s description of a fire as ζάφελος (Al. 556, ‘growing’?). 562 Ar. Pax 92 μετεωροκοπεῑς, Nu. 333 μετεωροφένακας, 360 μετεωροσοφιστῶν, E. fr. 913.4 μετεωρολόγων, Hp. AWP 2 μετεωρολόγα, Pl. R. 6 489c μετεωρολέσχας. 563 Ω’s θρυαλίζοι rests on a majuscule error in θρυλλίζοι. θρυλ-words are often transmitted in manuscripts with λλ, but see e.g. Posidipp. 87.3 πολυθρύλατον (papyrus of late third century), Phld. Mathes. vi.11 (first-­century papyrus), CLESard. 6.64 (early second century ce), Sappho in Obbink (2014) 37 (papyrus of c.250 ce). 564 σομφὸν (or ἀσύμφωνον) ἐμπνεύσας ἢ μὴ πιέσας τὸ στόμα θρυλιγμὸν ἢ τὴν καλουμένην ἐκμέλειαν ηὔλησε, in a context explicitly about an audience’s natural sense of pitch. The passage was borrowed by Porphyry, whose manuscripts more plausibly read θρυλισμός (In Ptol. Harm. p. 20.2). Vergados neatly suggested that Hermes is rubbing in the fact that Apollo so far knows only about auloi, though our evidence about the word is too slender to be sure. EM 456.56 (> Ps.-Zon. θ.1056) glosses θρυλίζω with ψιθυρίζω, from Herm., Theodoret PG lxxxi 1981 (a paraphrase of κατελάλησαν), or elsewhere.

418

C O M M E N TA RY 4 8 9 – 4 9 1 the lyre’s response make the lyre match her inventor’s behaviour at 293–8. There Apollo, after angrily questioning Hermes, was answered with a fart; this was μετήορον in that Hermes was ‘in the air’ (295 ἀειρόμενος μετὰ χερσίν) and ‘inflated’ (see e.g. Hp. Epid. 4.41 ὁ ἀπὸ τῶν φυσέων μετεωριζόμενος), and in that his fart was an airborne οἰωνός. This connection makes it relevant that τρυλίζω is used both of stomach-noises (Hp. Aff. int. 6) and of a bird-call (the quail’s: Pollux 5.89).565 The controlled reference back to Hermes’ fart points again to the similarity of Hermes and his inventions (Introduction n. 108), and insinuates that Apollo has been prone to the kind of behaviour which leads to bad lyre-playing, though now he has learnt a more gentle and playful way to question Hermes productively (cf. 464 εἰρωτᾶις ... περιφραδές). Apollo will seem to acknowledge the point at 546*. 489 σοὶ δ’ αὐτάγρετόν ἐστι δαήμεναι ὅττι μενοινᾶις: The repetition of 474* returns to the thought of 474–7, that Apollo’s epistemological advantages mean he will not need lessons in lyre technique – now specified in terms of the risk of a beginner offending the lyre as in 486–8. Hence they can proceed directly to an exchange. 490 ἀγλαὲ κοῦρε: A ‘declension’ of ἀγλαὸς υἱός| (cf. 500*) to the vocative. It is repeated in anon. De viribus herbarum 129; see 14* for the possibility of direct influence. For Hermes’ liberal use of vocatives to focus Apollo’s attention on the negotiation at hand, see 471–2*. 491–2 ἡμεῖς δ’ αὖτ’ ὄρεός τε καὶ ἱπποβότου πεδίοιο | βουσὶ νομούς, Ἑκάεργε, νομεύσομεν ἀγραύλοισιν: The balance with 490 ἐγὼ δώσω is made clear by δ’ αὖτε and the second vocative Ἑκάεργε.566 However, rather than continuing with a direct request, Hermes raises the outcome of the deal. This can be read as negative politeness, fostering a sense of cooperation and mitigating any sense of imposing on Apollo. Hermes also expresses the outcome with ἡμεῖς. Rather than being a plural-for-singular, characteristic of him, this can be taken as a true plural signalling the two gods’ future collaboration in the field of herding (cf. 465* ἡμετέρης vs 267*). After all, these forty-eight cows are only a section of the herd (74), and an ancient audience knew that Hermes’ role in cow-herding was no greater than Apollo’s. For Hermes’ weak connection to cow-herding see in general Introduction n. 79; as Vergados Tichy (1983: 162) connected θρυλίζω and τρυλίζω, but in purely phonetic terms qua onomatopoeias. Heiden (2010: 418) touched on the intratext between our passage and 293–8. 566 Hermes used this address in 464 and, with a focus on Apollo’s anger matched by 493–5, at 307. 565

419

C O M M E N TA RY 4 9 3 – 4 9 4 pointed out, the repeated νομ- in our line brings out the applicability to cows of the epithet νόμιος (found in e.g. Ar. Th. 976, Cornutus ND 16.17, GDRK 16.7, IG II2 3977), which probably more naturally referred to smaller stock. For νομεύω + acc. of place + dat. of animal, compare e.g. ὄρος ... νέμεται αἰξί (X. An. 4.6.17).567 Hermes’ pasturage of these cows on both mountain and plain matches the topography of Pieria, as well as of his cattle-rustling (hills at 70, 95; plains at 88, 96). But the merismus hill/plain also expresses proleptically the range of Hermes’ prerogatives over human herding, including transhumance, in contrast to when Apollo used it to express the range of his thefts (286–7*). Hermes’ sphere of power spreads from the mountain where it started – a place he could oppose to the habitat of ἄγραυλοι cows (262*) – to the plains, whose wealthier aristocracy is implied here by the rearing of horses (ἱπποβότου). At the same time, however, Hermes cedes to Apollo the more urban spaces of lyre-playing (480–1). 493 ἔνθεν ἅλις: The hymnist used this collocation in 179, where ἔνθεν was locative. Here, after the durable future state described by νομεύσομεν, ‘thence’ is more causal (as at e.g. A. Eu. 689) than temporal. 493–4 τέξουσι βόες ταύροισι μιγεῖσαι | μίγδην θηλείας τε καὶ ἄρσενας: At 192–4*, Apollo emphasized that the stolen cows were πάσας θηλείας and (I argued) invented a bull so as to add real-world verisimilitude and not overawe the Onchestan. Hermes now promises to make the mixing of sexes a reality, at the level of both the parents and the offspring.568 Introducing the cows to sexual reproduction with bulls from elsewhere will be the final stage of the rite of passage which transforms their apparently timeless situation in Pieria (72*) in the direction of the life of cows in men’s herds: the first stage was removing them physically from Pieria; the second imposing a sacrifice-like death on two of them, concealing the others, and then binding them with vitex (which had uses in fertility rituals: 409–13*). Hermes’ focus will be reproduction, not other details of productivity such as milking or ploughing. 494–5 οὐδέ τί σε χρή | κερδαλέον περ ἐόντα περιζαμενῶς κεχολῶσθαι: Hermes pretends to presuppose that Apollo was angry at losing out economically when he found two of the cows dead (echoing Apollo’s words

Conceivably the duplication of νομ- points to wordplay on the musical νόμοι which Hermes is exchanging for pastoral νομοί (cf. similar wordplay at H.Ap. 20 νομοὶ βεβλήατ᾽ ἀοιδῆς). 568 Note the emphasis on μιγεῖσαι | μίγδην. The adverb is elsewhere extant first in Pindar fr. 52m.7, though μίγδα is Homeric. For Herm.’s love of adverbs in -δην see 210*. 567

420

C O M M E N TA RY 4 6 3 – 4 9 5 at that moment: 407 οὐδέ τί σε χρή), and that he is still stubbornly angry (perfect tense); περιζαμενῶς presents a deterioration of relations since 307 ζαμενέστατε, with an unusual double-intensive form (also used of the north wind at Hes. fr. 204.126). In fact, Apollo’s anger has already visibly abated (417) because he thought the lyre was worth fifty cows (437), over which he appears prepared to cede control. Moreover, Hermes draws Apollo towards him by presenting him as κερδαλέος, but applies the word in the derogatory sense ‘profiteering’, and so denies Apollo the acumen implied when Hermes’ own words were described as κερδαλέοι (463).569 Despite giving up these cows, and despite their having to give up their pristine life (493–4*), Apollo is misguided to worry about losses, and stands to profit through the cows’ future reproduction – an idea which implies that he will in some sense retain ownership. The strategy of playful misrepresentation brings the speech’s cheeky undertones to a head, and the repetition of κερδαλέος as a framing device adds a sense of closure. 463–95 Hermes and the art of the deal: After expressions of goodwill and apparent flattery (464–77), Hermes passes to a presentation of his lyre (478–89), then looks to his future prerogative as a herder (490–5), defining this role in opposition to Apollo’s earlier idea of him as a wide-ranging brigand (491–2 ~ 286–7) and emphasizing reproduction and profit as key elements of his approach (493). He both acknowledges Apollo’s strategy of avoiding direct requests (464–5) and adopts it himself, in the way he jumps at 491–2 straight to imagining his future care of cows, and in 469–72, where he introduces several key factors underlying a potential exchange (φιλία among the gods; exchanges of divine gifts both material and honorific; the role of Zeus) and simultaneously drops a broad hint that he is interested in Apollo’s prophetic skill. This is only one of an array of politeness strategies, including compliments (esp. 468–74); the focus on the possibility of sharing skills and intangible advantages rather than exchanging objects (465, 477, 491–2, 494–5); a relentless focus in 464–80 on Apollo in the form of vocatives (x4: see on 471–2*) and second-person pronouns (x9) and verbs (x11); and detailed engagement with the terms of Apollo’s enquiries and offers. 465 τέχνης ... ἐπιβήμεναι recalls 447–8 τίς τέχνη ... τίς τρίβος;. Hermes circles back to those questions and

For κερδαλέος ‘profiteering’, Cozzo (1988: 39 n. 7) compared ­Aristotle EE 1221a23–4 κερδαλέος δὲ ὁ πανταχόθεν πλεονεκτικός; cf. Il. 1.149 κερδαλεόφρον. The connection to 463 was well discussed by Ludwich (1886a: 446), and Zeus had made a similar charge at 330–2; see also 549*. Apollo did have the epiklesis κερδῶιος in Hellenistic Larisa (e.g. FD III(4) 355.47).

569

421

C O M M E N TA RY 4 6 3 – 4 9 5 τίς Μοῦσα in 483–5, with a Muse-like lyre responding to one who plays with τέχνη and practice. 482 εὐφροσύνην νυκτός confirms Apollo’s point from 449 that the lyre offers εὐφροσύνην ... καὶ ὕπνον, while the sympotic focus in 478 and 480 confirms his observation at 454. The imperatives in 476, and 482 χορὸν ἱμερόεντα, recall the Muses’ traditional activities from 451–2, and so assure Apollo that the lyre can be integrated into them, besides completing Hermes’ own set of allusions to the names of the Graces (Euphrosyne, Thalia, Aglaia; cf. 484 παντοῖα ... χαρίεντα), to match the allusions made by Apollo. 467 πάντ’ εὖ οἶδας outdoes 456 κλυτὰ μήδεα οἶδας as a compliment, as 466 σήμερον εἰδήσεις outdoes 458 νῦν ... κλέος ἔσται as an offer of quick results. The reference to Apollo’s θυμός in 475 calls out his intervening reference to Hermes’ ‘elders’ θυμός’ in 457*. 464 ἐγὼ σοί and 477 κῦδος ὄπαζε recall and accept the offer of 460–1 ἐγώ σε | κυδρὸν ... ἕσσω. However, this last parallel is also an example of the back-handedness of Hermes’ politeness. Both here and in the φιλοκυδὴς κῶμος of 481 there is an implication that Hermes’ κῦδος entails a victory over Apollo. Similarly 465 οὔ τι μεγαίρω, 466 ἐθέλω ... ἤπιος εἶναι and 477* φίλε presumptuously express his advantageous position in the negotiation. The idea of Hermes’ kindness in word and thought at 466–7 points up that he has so far used these to run rings around Apollo’s power and knowledge, which he praises disingenuously in 469 and 474. The references to embarking on a skill, sitting, wealth and ὁσίη in 465–73 (cf. 492 cow-herding ~ 166 βουκολέων) all recall for the external audience Hermes’ ulterior motive, namely his envious ambition to match Apollo as expressed in 166–73. As a parting shot, 495 κερδαλέον περ ἐόντα cheekily makes Apollo out to be even more mercenary than Hermes, whose speech was introduced as κερδαλέος in 463. Hermes’ presentation of the lyre evidently offers the effaced narrator of early epic a chance to reflect on his own art (Vergados 14). But it is also tailored provocatively for the occasion. The return of the lyre-as-courtesan image in 478 and 485–6 emphasizes its erotic force, its sympotic role, and the question of payment, but also a dialogic quality brought out again in the new image in 483 of her answering questions. 479 κατὰ κόσμον and 484 νόωι χαρίεντα bring an ethical and didactic dimension to this courtesan-lyre, pulling her paradoxically towards the virgin Pythia in order to adumbrate the idea of exchanging her for a share in prophetic prerogatives. (Hermes here eagerly seizes on Apollo’s reference to the lyre’s θέσπις ἀοιδή and ὄσσα in 442–3. Hermes also makes the connection by focusing on δαῆναι as a kind of learning uniting his knowledge of the lyre to Apollo’s of prophecy: 474*.) However, the focus on the player’s ease in 480 εὔκηλος and 485 ῥεῖα tips over in 485–6 μαλάκηισιν ... δυήπαθον into a delicacy which poses some challenge to Apollo’s upstanding masculinity. This challenge comes from a lyre which is specifically like its maker: not only can it combine theogonic exposition with

422

C O M M E N TA RY 4 9 6 – 4 9 7 sympotic allure (478*; Hermes has done this in both his songs) and encomium with drunken revel (481*), but it shares a number of features of Hermes, including a clear and clever voice (478–9) which conveys epinician glory (481), and being a ‘companion of the banquet’ (478 ~ 436). Indeed, one can play the lyre badly much as Apollo has so far played Hermes badly: rough firsttime handling leads only to a squeak, or fart (488* ~ 293–6).

496–507 496 ὣς εἰπὼν ὤρεξ’, ὃ δ’ ἐδέξατο Φοῖβος Ἀπόλλων: The object of ὤρεξ’ is easily understood from 490. The Shuvalov Painter depicted a tortoiseshell lyre being handed over between Apollo and a naked youth (Geneva MAH 20334, perhaps c.425; Lezzi-Hafter 1976 S49, pl. 110d). This could depict or at least allude to our handover. Sarti (1992) found that Apollo is firmly associated with the κιθάρα in sixth-century Attic vase-painting, but increasingly with the λύρα in the fifth century. 497 Ἑρμῆι δ’ ἐγγυάλιξεν ἑκὼν μάστιγα φαεινήν: Though ἐγγυαλίζω ‘I hand’ was certainly not always ceremonial (see e.g. Od. 16.66, Hes. Th. 485), it tends to have divine subjects and may have been coloured synchronically by its similarity to ἐγγυάω. A sense of formality suits the moment of exchange towards which Hermes’ plans have been building. After ἐγγυάλιξεν, the manuscripts’ ἔχων is banal, whereas Martin’s ἑκών (1605: f. 37) produces a collocation found in A.R. 1.245 and 2.55–6, and clarifies Apollo’s attitude, which Hermes’ final words left open. D’Orville’s ἔχειν (n.d.: f. 255) was probably less likely than ἑκών to be corrupted. Apollo’s cattle-whip materializes suddenly as it is needed, like his jav­ elin (460) and Hermes’ staff (209). Cows were generally guided either by a long-handled whip (e.g. S. Aj. 1253–4; Boardman 1993: 72 fig. 67) or a stick, often tipped; some passages conflate the two.570 Whips are formulaically ‘shiny’, perhaps because their handles were embellished with metal, or because the leather strap was dyed like the bed-straps at Od. 23.201 ἱμάντα βοὸς φοίνικι φαεινόν (see Handschur 1970: 90–1). The items exchanged hint at the gods’ contrasting ways of operating: Apollo offers an instrument for controlling a subordinate, potentially with pain; Hermes offers one whose dialogic charm he has foregrounded (478–88).

For cattle-prods see 209 εἶχε δὲ ῥάβδον*. Slippage between μάστιξ and κέντρον for horses is found at Il. 23.383–7. Io is ‘herded by a whip’ at PV 682 though goading metaphors dominate the passage.

570

423

C O M M E N TA RY 4 9 8 – 5 0 1 498–9 βουκολίας τ’ ἐπέτελλεν, ἔδεκτο δὲ Μαιάδος υἱός | γηθήσας: ἐπιτέλλω covers a range of directives, from divine decrees (e.g. Od. 1.327), through commands to a subordinate, to fatherly instructions which hearers might well carry out of their own accord (e.g. Il. 9.179, 252). The gloss ‘I assign … as a task to’ covers a number of passages including this one. The verb casts Hermes as the younger and lesser partner, which may be patronizing since |βουκολίας echoes 167 |βουκολέων (assuming this emendation is right), where Hermes needed no instruction to pun on the topic.571 But this hint is not laboured: the poet chose not to give Apollo’s speech directly, emphasizes the balanced exchange through 496 ἐδέξατο ... 498 ἔδεκτο, and shows Hermes’ delight at the fulfilment of his hopes, which complements Apollo’s delight when he heard the lyre (421 |γηθήσας*). For the athematic form ἔδεκτο, see 477* δέγμενος. 499–501 κίθαριν δὲ λαβὼν ἐπ’ ἀριστερὰ χειρός | Λητοῦς ἀγλαὸς υἱός, ἄναξ ἑκάεργος Ἀπόλλων, | πλήκτρωι ἐπειρήτιζε κατὰ μέλος: As γηθήσας recalled 421, lines 499 and 501 recall 418–19 (and 501 also echoes 53, Hermes’ first strum).572 These connections imply the brothers’ like-mindedness and the lyre’s passage between them, and they frame the section of the hymn devoted to Apollo’s introduction to the lyre. The relationship of the lyre to herding has been realized in the course of this section, beautifully mirrored in the way the phrasing πλήκτρωι ἐπειρήτιζε κατὰ μέλος is only at this point revealed to sound suspiciously as if it referred to poking a cow’s limb (μέλος) with a prod (cf. esp. Agathias’ use of βού-πληκτρον ἄκαιναν, AP 6.41.3). 501–2 ἣ δ’ ὑπένερθεν | ἱμερόεν κονάβησε, θεὸς δ’ ὑπὸ καλὸν ἄεισεν: There is a cluster of textual problems, continuing into 503*. M has ὑπὸ νέρθεν | σμερδαλέον … ὑπὸ καλὸν ἄεισεν. Ψ had ὑπὸ καλόν | ἱμερόεν ... ὑπὸ μέλος ἄεισεν, where μέλος does not scan.573 By contrast ὑπὸ καλόν ἄειδεν is formulaic and has occurred in the corresponding lines 54 and 420. I infer that the deviation in 501 was caused by a scribe writing ὑπὸ καλόν by eyeskip to the line below, then adding something like νέρθεν as a correction. Ψ wrongly accepted καλόν in 501 and altered 502. M’s ὑπὸ νέρθεν is itself pleonastic and lacks

571

For the uses of this rare noun see DGE. Here and in A.R. 1.627, the plural suggests that the cowherd’s task is repeated or multifaceted. We cannot say whether the sense in Hsch. β.904 βουκολίη· κακολογία was available at Herm.’s date. 572 Line 500 is interposed to effect the change of subject, and is expansively given over entirely to this function. Λητοῦς ἀγλαὸς υἱός occurs only here and at 314. 573 Not even the spelling μέλλος (xΓ) can rescue it. For the variant’s appearance in a debate between Marullus and Poliziano see Thomas (2016) 286.

424

C O M M E N TA RY 5 0 3 satisfactory parallels: I propose ὑπένερθεν, through which the hymnist varied the ὑπὸ χειρός which ended the parallel lines 53 and 419. By the principle of utrum in alterum abiturum erat, ἱμερόεν should be retained as a significant variation of the earlier passages, whereas M’s σμερδαλέον marks an attempt to level them. Line 502 was therefore similar to Iliad 18.570 ἱμερόεν κιθάριζε, λίνον δ’ ὑπὸ καλὸν ἄειδεν, Sc. 202 ἱμερόεν κιθάριζε Διὸς καὶ Λητοῦς υἱός, and H.Ap. 185, where Apollo processes to Delphi and then (as here) to Olympus playing the lyre, which χρυσέου ὑπὸ πλήκτρου καναχὴν ἔχει ἱμερόεσσαν. Whereas σμερδαλέον suggested Hermes’ ‘aggressive’ intentions for the instrument in his conflict with Apollo, and the unfamiliarity of the new sound, ἱμερόεν shows that in the new spirit of reconciliation Apollo’s aesthetic response has won out, and its formularity suggests a normalization of the strange eroticism of Hermes’ lyre. The aorist ἄεισεν bundles this song of Apollo into an event over and done with before line 503, and so separate from the lyre-playing implied in 506. On any reading, the content of Apollo’s singing is treated curtly as the narrative presents a flurry of action; 498 ἐπέτελλεν similarly avoided further direct speech. But Sicherl (1960) identified other instances of minuscule σ/δ confusion in the tradition of the corpus of hymns (including here), and Ilgen’s more open-ended ἄειδεν ‘began to sing’ remains a possibility.574 503 καί ῥα βόας μὲν ἔπειτα ποτὶ ζάθεον λειμῶνα: M has (i) καί ῥα (ii) βόας μὲν ἔπειτα (iii) κατὰ ζάθεον λειμῶνα ἐτραπέτην, αὐτοὶ δέ ... against Ψ’s (i) ἔνθα (ii) βόες μὲν ἔπειτα (iii) ποτὶ κτλ. These three variations can be treated independently; I shall identify which variant in each pair is easier, proceeding from greater to lesser confidence that we should accept them. In (ii), βόας is certainly easier. ἐτραπέτην would normally be dual, and more often than not transitive; the context provides two agents whose oversight of cattle has just been under discussion.575 A scribe perhaps introduced βόες to contrast with αὐτοί, though in fact the structure μέν ... αὐτὸς+ δέ … fairly often entails a shift of focus but not of grammatical subject: see, of many examples, 509–11, Il. 1.485–7, 7.49–50, Od. 9.416–17, 11.20–1, Sc. 332–3. In (iii), ποτί is easier than κατά: there is no reason to think that the cows have scattered ‘throughout’ the Alphean meadow, ποτί provides a very natural complement to the verb (cf. Od. 9.315 πρὸς ὄρος τρέπε μῆλα), and ζάθεος – an

574

Parallels are not decisive: 54 and Il. 18.570 (above) call for imperfective presentation; conversely at Od. 21.411 ἣ δ’ ὑπὸ καλὸν ἄεισε, Odysseus’ bowstring is plucked once, and suits the aorist. 575 For the possibility of dual verbs standing for plurals, see GH ii.27–9, Richardson on H.Ap. 456. On the verb’s transitivity see Bonanno (1970–2).

425

C O M M E N TA RY 5 0 4 – 5 0 7 epithet of λειμών only here – suits one beside Olympus better than the one by the Alpheus. In (i), 401 ἔνθ’ Ἑρμῆς μὲν ἔπειτα (and cf. e.g. Il. 18.450 with αὐτός, 23.818 with Αἴας) makes ἔνθα the lectio facilior; perhaps the first few letters of the line were damaged in Ω. 504–5 Διὸς περικαλλέα τέκνα | ἄψορροι πρὸς Ὄλυμπον ἀγάννιφον: Διὸς περικαλλέα τέκνα occurred also at 323* and 397*; all are scene-changes where the gods travel together to or from Olympus, and the phrase increasingly implies the gods’ similarity and basis for cooperation (cf. 506–7* on ἄμφω). Ὄλυμπον ἀγάννιφον also echoes their first arrival (325). ἄψορροι flags up the intratextual structure. 505–6 ἐρρώσαντο | τερπόμενοι φόρμιγγι: ῥώομαι seems originally to have meant ‘go fluidly back and forth’; it applies to the legs of someone hurrying (Od. 23.3), but also to dancing (e.g. Hes. Th. 8), which would contribute to our image of the gods’ musical delight. τερπόμενοι introduces τέρψις into the poem, a more normal epic response to music than Apollo’s previous lust.576 The lyre loses some of its Hermetic sexiness – and the poem some of its linguistic virtuosity – as Apollo takes over. Cf. 501–2* on ἱμερόεν, 565*. 506–7 χάρη δ’ ἄρα μητίετα Ζεύς, | ἄμφω δ’ ἐς φιλότητα συνήγαγε: χαίρω appears in Epos in a limited range of situations, here joy at a return (LfgrE s.v. B 1aδ, though it is wrongly assigned to the miscellaneous group 1aθ). Zeus and χάρις return, with structural import, at 574–5*. Lines 504–5* prompted us to contrast the gods’ previous arrival on and departure from Olympus, and Zeus now appears once again as both the father patching up a family quarrel and the coordinator of global peace. ἄμφω contributes to this, by continuing the shift from 315 ἀμφίς to 390–1* ἀμφὶ ... ἀμφοτέρους and 397 ἄμφω, where the cognate words chart the brothers’ opposition and unification. The hymn’s only other use of μητίετα Ζεύς| came at 469*. There Hermes mentioned Zeus’s φιλία towards Apollo and the gifts and honours consequent upon it, while fishing for φιλία, gifts and honours from Apollo. Here Zeus is revealed to be the background agent of that process. In the human realm he manipulates battles to cause φιλία at e.g. Il. 3.323, 4.16, 84, Od. 24.476. 496–507 συνήγαγε. Hastening to a false ending: The narrator has adopted a leisurely pace since around 420, but speeds up noticeably here. The long-awaited exchange, and the gods’ pleasure in their new acquisitions,

576

For epic responses to song see e.g. Thomas (2014) with references. For the phrasing, see Il. 9.186 τερπόμενον φόρμιγγι λιγείηι|, Thgn. 791 |τερποίμην φόρμιγγι.

426

C O M M E N TA RY 5 0 7 – 5 0 8 are outlined succinctly; Apollo’s words about herding and his first lyrics are mentioned but not quoted; the journey back to Olympus and Zeus takes just a few lines, and we get no detail yet about how Zeus brings Hermes and Apollo to friendship. This marked acceleration contributes to a sense that the essential points of the narrative are over. Apollo has laughed (420), foreseen ‘peace’ (438) and promised gifts and candour (458–62); Hermes has claimed kindness (466) and, albeit rather presumptuously, called Apollo φίλε (477); now they exchange the key items which we have been focusing on since line 17–18, and both gods herd and enjoy lyre-music together (504–6). However, the closure turns out to be false.577 The structural key to what follows is the stem φιλ-, which occurs in three pairs at 507–8, 524–5 and 574–5 to articulate the move towards a more secure relationship (Introduction §3.2). The stem is notoriously difficult to translate (Konstan 1997: 28–78): here the focus is on ‘friendship’ as an affective relationship, but also on the ritual actions which formalized such relationships in ancient Greece. The gods have recognized their close social bond and achieved some w ­ illingness to cooperate. But Apollo’s emotional investment in the relationship has so far been rooted mainly in his desire for the lyre, and he will reveal his residual suspicion of Hermes in 514. Meanwhile Hermes has hinted at his ambition for κῦδος through prophecy (464–77), and Apollo is yet to answer. Democritus said ὁμοφροσύνη φιλίην ποιεῖ (68 B186): in what follows, the gods’ ὁμοφροσύνη motivates their progress to φιλία, through further gifts which are outward signs of their relationship (509, 527), and through an exchange of oaths.

507–23 507–8 καὶ τὸ μὲν Ἑρμῆς | Λητοΐδην ἐφίλησε διαμπερές: For the use of τὸ μέν (Ψ) and τὰ μέν (M) to mean ‘on the one hand’ see LSJ s.v. ὁ VIII.4, Slater (1969) s.v. ὁ B.5. I see no parallel in early epic, but Pindar’s uses suggest that one need not emend.578 The phrasing prepares for a corresponding phrase such as ‘And Apollo in turn showed affection for Hermes’. The μέν is 577

This false ending repeatedly led analyst critics, from Groddeck (1786) 36–8 on, to question the authenticity of the lines from around here to around 578. Ludwich (1886a: 449) already pointed to the intratext 574–5 ~ 506 as a reason to doubt that. Graefe (1963: 516–22) showed that the alleged stylistic inconcinnities are trivial. 578 Matthiae (1805) suggested καί ῥ’ ὁ. In Od. 2.46 τὸ μέν ‘first, ...’ starts an expansion of χρεῖος ... δοιά. At 11.363 and 18.227 τό can be taken as an inner accusative (‘make this conjecture’, ‘feel this anger’). Christensen (2010: 555 n. 63) proposed that initial τὸ μέν in these two passages and 7.239, 19.104, 19.509 could be a ‘polite “hedge” to soften the utterance’. To my mind the evidence is insufficient.

427

C O M M E N TA RY 5 0 8 – 5 0 9 t­herefore better not taken with 511 αὐτὸς δ’ αὖθ’. And even when a viable corresponding phrase arrives in 513 καὶ τότε Λητοΐδης Ἑρμῆν (Denniston 374 gives parallels for μέν ... καί), we find that Apollo’s affection cannot be stated as straightforwardly as Hermes’, but has still to be won gradually – all the way to 575*, which does clearly balance the present clause. In epic the aorist of φιλέω is characteristic of acts of hospitality or (as here) of a god conceiving affection.579 After such an aorist, διαμπερές means ‘for ever’: cf. Il. 10.89, 16.618, Hes. Th. 402. Considering that piercing was Hermes’ modus operandi as a φιλήτης (175–8, 284, Introduction §3.3.6), there could be wordplay on the etymology of διαμπερές (< διαναπείρω) as he now becomes (with further wordplay) a φιλητής. Compare the more obvious jeer at Il. 16.617–18 τάχα κέν σε ... ἔγχος ἐμὸν κατέπαυσε διαμπερές. 508–9 ὡς ἔτι καὶ νῦν | σήματ’, ἐπεί: The transmitted letters are defensible if correctly punctuated. Since neither σήματα nor σήματι offers smooth grammar either in the ἐπεί-clause or with ἐφίλησε, one must understand ἔστι with it. This verb is commonly omitted in the idioms τεκμήριον δέ and σημεῖον δέ, and cf. E. fr. 322 ἓν δέ μοι τεκμήριον, Pl. Crit. 111c ἧς καὶ νῦν ἔτι φανερὰ τεκμήρια. The last parallel suggests that one should take ὡς ... σήματ᾽ together, meaning ‘as (there are) signs (of) even now’, rather than leaving σήματ’ sc. ἔστι in asyndeton (Càssola). The text returns to the language of aetiology from 125* (ἔτι νῦν) and 136 (σῆμα). Hermes’ gift of the lyre and consequent invention of panpipes as a replacement (509–12) is identified as the cause of a lasting musical alignment which the audience could still recognize in representations of the two gods. The hymn thereby affords what was probably an unconventional way to make sense of that alignment. Of emendations, Radermacher’s σῆμα δ’, punctuated parenthetically, seems the most plausible. It would be a poetic equivalent of the phrases with σημεῖον/τεκμήριον already mentioned (the latter occurs in verse at E. fr. 898.5). A bolder emendation would be σῆμά τ’ ἐπῆν·, with the last word at some point mistaken for ἐπήν and ‘corrected’ to ἐπεί. This still seems more palatable than introducing another lacuna (West). 509 κίθαριν μὲν Ἑκηβόλωι ἐγγυάλιξεν: The repetition from 497, where ἐγγυάλιξεν was used of Apollo’s gift to Hermes, again expresses the reciprocity of the reconciliation. Apollo receives the lyre and is called Ἑκηβόλος,

579

I find fourteen other instances of the latter context. The verb generally denotes an internal change: ‘in the heart’ accompanies it at Il. 9.117, H.Ap. 138. Thus Il. 13.430 φίλησε, of Hippodameia’s parents loving her, is exceptional.

428

C O M M E N TA RY 5 1 0 – 5 1 1 implying his ownership of a bow (218*). This restores him, finally, to the characterization of H.Ap. 131, which the hymn has so far undermined (Introduction §1.3). 510 ἱμερτήν: For musical uses cf. Solon fr. 13.52, Pindar O. 6.7, Phanocles fr. 1.21, and perhaps IG I3 833b.2 (c.480 bce). As in 502 ἱμερόεν*, the lyre’s allure is still there even if the idea of it as a courtesan has been abandoned with the handover to Apollo. δεδαὼς ὃς ὑπωλένιον κιθάριζεν: Ψ had ἱμερτὴν δεδαώς, ὅ δ’ ... (M skips the line). However, forms of δαῆναι are used thematically hereabouts, and this one is perfectly suited to confirming Apollo’s talent for lyre-playing (cf. 474 = 489, 483), whereas it is a strange verb to apply to Hermes’ gift-­ giving. Ernesti moved the comma before δεδαώς, but Hermann rightly complained that, while δέ can be postponed (Denniston 187–9), ὃ+ δέ cannot. I infer a minuscule ὃσ > ὃδ’ confusion (cf. 502* ἄεισεν), already mentioned in Zimmermann (1920) 238.580 For ὑπωλένιον κιθαριζ- see 433*: Hermes’ skill then is now matched and supplanted by that of Apollo. 511–12 αὐτὸς δ’ αὖθ’ ἑτέρης σοφίης ἐκμάσσατο τέχνην· | συρίγγων ἐνοπὴν ποιήσατο τηλόθ’ ἀκουστήν: αὐτὸς δ’ αὖθ’ is the counterpart to 509 κίθαριν μέν (+ 510 κιθάριζεν), introducing a second musical instrument. For the combination of music, σοφία and τέχνη see 483*, though here τέχνη is more oriented to the manufacturer’s craft than to the player’s technique. Now that Hermes has acquired his herdsman role, he invents the herdsman’s favoured instrument (so already Il. 18.526); Rudoni (2015) has in fact argued that our line was construed as a foundational moment for the bucolic genre by Theocritus as he composed Idyll 7. Hermes creates not just the syrinx but its ἐνοπή. While the phrasing is old (cf. Il. 10.13 συρίγγων τ’ ἐνοπήν), the echo of 422 ἐνοπῆς suggests that his second instrument will be a suitable substitute for the first;581 in 47*, Hermes’ construction of the lyre’s internal props sounded remarkably like the construction of a graduated syrinx. The pipe’s sound also resonates with Hermes’ whistling in 280 μάκρ’ ἀποσυρίζων, where μακρά corresponds to τηλόθ’ ἀκουστήν here. 580

For a predicative adjective in a relative clause preceding the relative pronoun, see 530*, Il. 2.125 ἐφέστιοι, Pl. Phdr. 238a4 ἐκπρεπής. Ludwich’s δεδαὼς δ’ ὅ γ’ (1886a: 443) is apt but further from the paradosis. 581 We saw on 422 that |θεσπεσίης ἐνοπῆς finds a uniquely close parallel in the ἐνοπή of the dying Python at H.Ap. 360. Our passage could contribute to the intertext since, according to Str. 9.3.10, the citharists’ Pythian Nome ended with a section called σύριγγες which imitated that sound (cf. hypothesis a to Pindar’s Pythians).

429

C O M M E N TA RY 5 1 3 – 5 1 5 ἐκμάσσατο is from ἐκμαίομαι, a hapax as virtuosically invented as Hermes’ piping itself. It links the thought of 108* πυρὸς δ᾽ ἐπεμαίετο τέχνην (for ἐκwith verbs of successful seeking, cf. ἐξευρίσκω and later ἐκζητέω) and the sound of ἐκμάσσομαι ‘mould exactly’ (suited to the wax binding of the pipes). The verb thus reprises the wordplay in 294 σὺν δ’ ἄρα φρασσάμενος ~ συμφράσσομαι.582 The later myth of Pan and Syrinx has obscured whether or not the hymn’s aetiology would have seemed remarkable. Before and even shortly after Pan’s spread from Arcadia around 500, a strong connection between Hermes, herding and the syrinx is plausible. Haas (1985: 50–1) found four sixth-century images of Hermes playing the syrinx – more (for what it is worth) than any other god. Artistic representations then fade out, but in literary sources Hermes still plays it to lull Argus, Io’s ‘herdsman’ (e.g. S. Inachus fr. 269c.7); and Euphorion asserted that Hermes invented the one-reed pipe (fr. 203). Our description is markedly rapid. Perhaps the poet thought two descriptions of instrument-making would be tiresome, wanted to maintain a bias towards the lyre, or wanted to communicate the effortlessness of Hermes’ creativity (cf. 410–13). 513 καὶ τότε Λητοΐδης Ἑρμῆν: Balancing 507–8*, though (as I observed there) we now find out that Apollo is lagging behind Hermes in terms of readiness for φιλότης. 514 διάκτορε ποικιλομῆτα: Compare 392* and 155* respectively. ποικιλομῆτα here is a contextually relevant replacement of δῶτορ ἐάων, which follows διάκτορε at Od. 8.335 and Hy. 18.12. The lemma perhaps influenced Orph.H. 28.3 ποικιλόβουλε διάκτορε. 515 μή μοι ἅμα κλέψηις κίθαριν καὶ καμπύλα τόξα: The juxtaposition of κίθαρις and καμπύλα τόξα occurs elsewhere only at H.Ap. 131–2, where Apollo’s first words are to declare his interest in the lyre, the bow and prophecy. For Herm.’s systematic use of this part of H.Ap. see Introduction §1.3. Before answering Hermes’ implicit request for a share in prophecy, Apollo secures the lyre and bow. In other sources, his kithara and composite bow are frequently presented as a complementary pair of stringed objects. Already in Od. 21.406–11 Odysseus strings his bow and plucks it like a lyre-player, at a feast in honour of Apollo. Apollo carries both lyre and bow in e.g. LIMC s.v.

See 294–6*. Barnes considered either ἐκμάσσετο or ἐκμάξατο, which simplify out the wordplay. Ruhnken (1781: 49) compared Ps.-Manetho Apotel. 4.285 ὧν [entertainers] ὁ βίος χλεύηι τέχνην ἀπεμάξαθ’ ἑτοίμην.

582

430

C O M M E N TA RY 5 1 6 – 5 1 7 Apollon 382 (Berlin Painter, c.480); Sabinus 1 GP offers a rather different take on the trope. See in general Monbrun (2007). The reference to stealing the bow carries a further allusion. In Alcaeus’ Hymn to Hermes, Hermes stole the cows, then when Apollo found him stole at least Apollo’s quiver and arrows; this made Apollo laugh and so brought about the reconciliation (fr. 308 with Cairns 1983). Having avoided this version of the myth, the hymnist alludes to its existence (as a tradition, or in Alcaeus’ poem specifically) as a future story that Apollo fears. καμπύλα shows that τόξα means ‘bow’, not (as elsewhere) ‘arrows’; in Luc. DDeor. 11 Hermes steals both. I prefer M’s ἅμα κλέψης to Ψ’s ἀνακλέψης – a simple minuscule copying error.583 The mental image of Hermes stealing Apollo’s bow and lyre simultaneously is a hyperbole which complements Apollo’s deflation of his reasons for fear in 516–17*. In terms of politeness theory, Apollo mitigates the face-threat of expressing his (genuine) suspicion and requesting an oath from Hermes, by self-deprecatingly maximizing the difference between the content and grounds of his fear.584 516–17 τιμὴν γὰρ πὰρ Ζηνὸς ἔχεις ἐπαμοίβιμα ἔργα | θήσειν ἀνθρώποισι κατὰ χθόνα πουλυβότειραν: I follow Wolf in thinking that the hapax ἐπαμοίβιμα is the most likely common ancestor of Ψ’s ἐπαμοίβια and M’s ἐπ’ ἀμοίβημα. Compounds of ἐπαμοιβ- generally have relevant nuances of multiple alternations, often swift – of interchange or interlacing (cf. ἐπαμοιβή, ἐπαμοιβαδίς, ἐπημοιβός); an exception is the one-off exchange of armour at Il. 6.230. For the pertinent applications of τίθημι see LSJ s.v. A.III–VII.585 For -ιμος as a suffix to first-declension nouns compare e.g. ἀγώγιμος, ἀοίδιμος. The alternative is not to worry about the source of the mu in M’s reading and to accept Ψ’s, which would be a hapax modelled on formulaic πολεμήϊα ἔργα, like Il. 2.614 θαλάσσια ἔργα or Od. 11.246 φιλοτήσια ἔργα. The euphemistic periphrasis in the latter would be matched in our passage. τιμή from Zeus is prominently placed. At 172* Hermes implied that his competitiveness towards Apollo stemmed from a desire for it. Zeus indirectly declared Hermes’ prerogatives as herald and guide at 331 and 393. Then Hermes made Zeus the source of Apollo’s main τιμή (prophecy) at 469–71. Apollo uses those powers here to present Hermes’ prerogative as granted

For ἀνακλέπτω (‘steal by lifting’ rather than ‘steal back’) see LSJ s.v., adding IGASMG iv.30 (Paestum, c.515). 584 For the concept of a ‘face-threat’, see Brown and Levinson (1987). 585 By contrast, Ψ’s tmesis would amount to a variation on ἀμοιβὴν ἐπιτιθέναι, used of divine punishment at Hes. Op. 334. 583

431

C O M M E N TA RY 5 1 8 – 5 1 9 (present ἔχεις) but not yet exercised (future θήσειν).586 This is because of the human focus of Apollo’s words (ἀνθρώποισι κατὰ χθόνα πουλυβότειραν).587 In fact, the periphrasis ἐπαμοίβι(μ)α ἔργα prompts us to think not just of commercial exchange, but of other forms of interchange which we have already seen Hermes exercise among the gods: the swapped hooves, the interlaced sandals, gift-exchange and so on (see Introduction §3.3.6). The context (fear of theft) shows that Apollo is presenting patronage of trade as something inseparable from theft – though this is to ignore the importance of law, against Apollo’s own earlier thinking; in particular, at 291–2 he foresaw that Hermes would be invoked as ἀρχὸς φιλητέων. Finally, Hermes’ control of trade (or even theft) in the human sphere is not a sufficient reason to fear that he will steal Apollo’s possessions. Hence Apollo gives various signals that he is manipulating his attitude in these lines. As mentioned on 515*, the pattern underlying this manipulation, that Apollo humorously maximizes the difference between his fear and the reason for his fear, can be read as a conscious politeness strategy.588 518–19 ἀλλ’ εἴ μοι τλαίης γε θεῶν μέγαν ὅρκον ὀμόσσαι, | ἢ κεφαλῆι νεύσας ἢ ἐπὶ Στυγὸς ὄβριμον ὕδωρ: Where gods swear a ‘great oath of the gods’ and it is specified, it is an oath by the Styx, sometimes including Gaia and Ouranos.589 In 519 the Styx is indeed mentioned; it is ὄβριμος – both physically powerful, like the ὄβριμον ὕδωρ of mountain streams in spate at Il. 4.453, and threatening (cf. Hes. Th. 839 ἐβρόντησε ... ὄβριμον, A. Ag. 1411 μῖσος ὄβριμον). ἐπί + accusative is not a normal way to describe an oath either invoking Styx or uttered over a libation of water from Styx, but the usage can be defended as an adaptation of ἐπόμνυμι + accusative (e.g. Hes. Th. 793).590 The formulation with εἰ ... τλαίης γε acknowledges that such an oath is a 586

For expressions of power + future infinitive, KG i.185 compared Th. 3.28.1, S. Ph. 1394, X. Oec. 12.12 (often emended). For the phrasing compare Od. 11.302 τιμὴν πρὸς Ζηνὸς ἔχοντες. 587 The base formula was ἐπὶ χθονὶ πουλυβοτείρηι (x10). The hymnist perhaps wished to avoid a second ἐπ- and a second dative. 588 Among interesting older comments, D’Orville (n.d.: f. 256) took the contrast to 291–2 to indicate Apollo’s friendlier disposition here; Ilgen is worth consulting; and Lohsee (1872: 5–6) discussed the humorous slippage between trade and theft, for which Brown (1947: 79–83) provided useful parallels. 589 Od. 5.178–86, Hes. Th. 400, 784 (used in cases of νεῖκος among the gods), H.Dem. 259, H.Ap. 79–85. The content is not stated at Od. 10.299~343. θεῶν ὅρκος can also mean a human oath by the gods, e.g. Od. 2.377, E. Hipp. 657. 590 Since ἐπί + gen. occurs in the right sense (LSJ s.v. A I.2e) one might consider emending to ἐπὶ Στυγὸς ὀβριμοΰδρου. The compound would be a hapax (cf. ὀλιγόϋδρος, πολύϋδρος for the form), generated from ὄβριμον ὕδωρ| and normalized back to it in the course of transmission.

432

C O M M E N TA RY 5 2 0 – 5 2 2 big thing to ask, i.e. demonstrates negative politeness, compared to the more common formula εἰ δ’/ἀλλ’ ἄγε μοι ... ὄμοσσον. The politer form is found at Od. 5.178 = 10.343 (Odysseus to Calypso and Circe), but also at H.Ap. 79 when Delos asked Leto for assurance that Apollo’s first oracle-temple would be founded on her. Leto’s oath that Apollo would have a sanctuary there (87–8) falls short of Delos’ request, just as Hermes’ response here falls short of Apollo’s (522–3*; Introduction §1.3). Besides the request for an awesome oath by the Styx, Apollo offers Hermes the option of a promissory nod. Zeus’s nod is an irrevocable μέγιστον τέκμωρ (cf. 395*, Il. 1.525–6), as are those of Fate or a Fate (PMG 1017, Pindar O. 7.64–8, Bacchylides 17.24). But regarding other gods, Callimachus can claim that only Athena of Zeus’s daughters has been granted a binding nod (Lav. Pall. 131–3); Apollo’s own nods at Herm. 524 and 536 will be reinforced by a σύμβολον and by pledges and an oath. Radermacher, Abramowicz (1937: 80) and others took it as a sign of naivety that Apollo trusts Hermes enough to offer him a mere nod, and Ludwich (1890: 17) emended away the issue by changing νεύσας ἤ to νεύσειας. I find it more plausible to see Apollo’s request as testing the water: to what extent will Hermes live up to φιλότης by rejecting the opportunity to swear deceptively (as at 274–7, 383–4) and to take advantage of the loophole? (See on 522–3*.) 520 πάντ’ ἂν ἐμῶι θυμῶι κεχαρισμένα καὶ φίλα ἔρδοις: Apollo implies that Hermes is showing him some but not yet all (emphatic |πάντ’) signs of χάρις and φιλία. This pair of terms occurred in 506–7, which located the χάρις in Zeus, and will come back again in 575, where he has transferred it to Apollo and Hermes. The phrasing is similar to formulaic (τῶι) ἐμῶι κεχαρισμένε θυμῶι, Il. 24.661 ὧδέ κέ μοι ῥέζων Ἀχιλεῦ κεχαρισμένα θείης and Od. 15.360 φίλα ἔρδοι|. 521 ὑποσχόμενος κατένευσεν: Cf. formulaic ὑπέσχετο καὶ κατένευσεν. The form of ὑπίσχομαι recalls Hermes’ shifty oath at 275*, and prompts us to compare whether this promise is more reliable. The unusual use of κατανεύω to introduce a negative clause is perhaps designed to foster connections to 524 and 536, where it recurs (cf. much later IGUR iii.1155.75). 522 μή ποτ’ ἀποκλέψειν ὅσ’ Ἑκηβόλος ἐκτεάτισται: Hermes appears to go far beyond an assurance about stealing ‘the bow and the lyre together’ (515); his choice of ἑκηβόλος may signal engagement with the issue of the bow (218*). But the wording of his promise is once again tricky: μὴ ... ὅσα ‘not… all’ leaves it open to Hermes to steal some items, and the perfect tense ἐκτεάτισται avoids any commitment about Apollo’s future acquisitions.

433

C O M M E N TA RY 5 2 3 523 μηδέ ποτ’ ἐμπελάσειν πυκινῶι δόμωι: μηδέ is ‘nor even’ rather than ‘and not’, since not approaching is a much stronger promise than not burgling. For the tradition of using πυκινὸς δόμος in contexts of a home being prone to or safe from entry see Il. 10.267, 12.301, Od. 6.134. Apollo may well think that Hermes is rectifying the slipperiness of 522, if one assumes that πυκινῶι δόμωι appeared unambiguous to him, i.e. that at the time of the oath Apollo had one storehouse, namely the Delphic one, which Hermes threatened to burgle at 178 (with μέγαν δόμον, echoed here) and, more obliquely but in front of Apollo, at 385. That Delphi was Apollo’s first temple is affirmed throughout the Hymn to Apollo.591 However, like 522 ἐκτέατισται, Hermes’ words do not commit him to anything about the other temple adyta which Apollo would go on to get between the time of the oath and the time of Herm.’s performance. 507 καὶ–523 δόμωι. Hermes’ φιλία for Apollo: Hermes’ gift of the lyre at first seems a sufficient sign of his friendship, and he is able to make up for his loss briskly by creating the syrinx (511–12). Indeed, nearly the full range of his earlier inventiveness is echoed one way or another in these lines, to emphasize that his skills are still there even if now they will not be directed against Apollo: 508 ἐφίλησε διαμπερές suggests Hermes’ previous threatened and actual use of piercing while being a φιλήτης, 508–9 ἔτι καὶ νῦν σήματ’ the aetiologies arising from his butchery, 512 his whistling as a paralinguistic strategy at 280; 511 ἐκμάσσατο τέχνην suggested the invention of fire, and even the wordplay which introduced Hermes’ other paralinguistic trick of farting. However, when we reach 513*, we do not get the expected statement of Apollo’s reciprocal φιλία. He needs further reassurance if Hermes is to complete πάντα ... κεχαρισμένα καὶ φίλα (520). In itself, an exchange of promises was a normal enough part of the rituals for formalizing friendship (see 526– 7*), but the narrator has set this proposal up as a sign of Apollo’s continuing caution.592 He is embarrassed that he is about to refuse Hermes a direct share in his prophetic powers (Franke 55), and concerned about how Hermes will react. But he adopts various strategies for toning down his call for a formal promise, and hence displays the very good-will which he is trying to secure in return: these strategies include hyperbole in what he claims to fear (515), a joke about Hermes’ capacity for theft (516–17), the politer formula for 591

There are no identified dedications to Hermes at Delphi before the fourth century (see Jacquemin 1999), but Apollo allows him to approach the temple at E. Ion 29–40 and possibly at A. Eu. 89. I doubt Herm.’s audience would have factored such stories into their sense of whether 523 is a sustainable promise. 592 As when the cows were concealed, the hymn assumes for narrative interest that Apollo cannot simply foresee a predetermined future.

434

C O M M E N TA RY 5 2 3 – 5 2 4 requesting an oath, and inclusion of the option of a promissory nod despite the slipperiness of Hermes’ previous oaths (518–19). Hermes’ promises again sound good but are uncomfortably vague, even at this apparent moment of benevolent reconciliation: while 522 (‘I will not steal all your past acquisitions’) is obviously slippery, Hermes presents it as a foil to an apparently more mature and stronger promise in 523 (‘I’ll never even approach the adyton at Delphi’), which seems to satisfy Apollo.593 Two readings seem possible. In the first, Apollo is naive for allowing Hermes a ‘mere’ promissory nod in lieu of an oath, and again in not seeing that even 523 says nothing about the portfolio of sanctuaries beyond the ‘original’ adyton of Delphi which he would go on to acquire. The second reading is that Apollo deliberately gives Hermes the option of a nod, does perceive that Hermes is still willing to cheat him, and uses his testing of Hermes’ attitude to decide how effusive to be in the further offers in the following lines. A live performance may have guided the audience’s interpretation.

523–40 523–4 αὐτὰρ Ἀπόλλων | Λητοΐδης: For enjambement involving a patronymic see e.g. Il. 4.489–90, Sc. 478–9. 524 κατένευσεν ἐπ’ ἀρθμῶι καὶ φιλότητι: Apollo has a nod with negative content to match that of Hermes in 521. To avoid the phrase being made redundant by 525, ἐπί … φιλότητι should be interpreted as ‘on condition of (mutual, future) affection’, not ‘in (Apollo’s present) condition of affection’. The language of interstate diplomacy resurfaces: cf. SEG 22.236 (Sybaris, late sixth century) ἀρμόχθεν οἰ Συβαρῖται κ’ οἰ σύνμαχοι κ’ οἰ Σερδαῖοι ἐπὶ φιλότατι πιστᾶι κἀδόλοι ἀείδιον.594 For the combination ἀρθμ-/φιλ- see also Il. 7.302, Thgn. 326, ‘Thgn.’ 1312 ἄρθμιος ἠδὲ φίλος, PV 191 εἰς ἀρθμὸν ἐμοὶ καὶ φιλότητα, Call. fr. 497a ἀρθμὸν δ’ ἀμφοτέροις καὶ φιλίην ἔταμες. Here ἀρθμός brings its sense of peace but also a possible musical wordplay in its derivation from ἀραρίσκω: it is used of magadis-tuning in Telestes PMG 808; cf. Plu. Mor. 1030a. 593

Fletcher (2008: esp. 25, 41) has discussed whether Hermes’ nod here marks a maturation compared with his previous oaths. 594 Cf. 312*, 462* for other phrases from diplomacy. ἐπὶ φιλότατι occurs repeatedly in another document from Magna Graecia, SEG 49.1264.2 (Gela, 475–50). The tablet (on which see West 1999, Eidinow 2007: 160–7) is difficult, but if one translates ἐπὶ φιλότατι τᾶι [Ε]ὐνίϟο· μεδέν’ [Ε]ὐνίϟο σπευδ[αι]ότερον μεν μεδὲ φιντίονα … ‘[Apellis prays], on condition of Eunikos’ affection, that nobody be more significant or beloved [to the audience of a choral contest?] than Eunikos …’, the phrasing is similar to Apollo’s.

435

C O M M E N TA RY 5 2 5 – 5 2 6 525 φίλτερον ἄλλον ἐν ἀθανάτοισιν: φίλτερον forms, with 524 φιλότητι, the second of three pairs of φιλ-words which structure 503–75 (Introduction §3.2). For another ‘dearest friend’ promise, Matthiae (1800: 305) compared Callimachus 25 G–P. For Hermes’ concern about the divine community signalled by phrases such as ἐν ἀθανάτοισιν, see 170*. 526 Lacuna: If 526 follows 525 directly, two problems arise, which in my view justify a lacuna: (i) ἄνδρα Διὸς γόνον becomes a significant subcategory of τινα ... ἐν ἀθανάτοισιν; (ii) there is an abrupt switch in mid-line into extensive direct speech. One could defend (i) by saying that some men were immortalized, such as Zeus’s great-great-grandson Ganymede or son Heracles, and that these cases do strain language (e.g. S. Ph. 727–8 ἀνὴρ θεοῖς | πλάθη θεός of Heracles, Lobo PMG 506 θεὸν ἄνδρα Λίνον). However, these exceptions seem a very odd basis for Apollo’s antithesis, even granting the possible thematic importance of Heracles (see below). As for problem (ii), speeches in hexameter do occasionally start in mid-line (first at Il. 23.855 ἀνώγει | τοξεύειν· “ὃς μέν κε ...”), but not without a verb of speech to mark the transition from indirect to direct discourse.595 In fact, in early epic switches from indirect to direct discourse are limited to a few instructions; the direct speech is fairly brief, and the effect is urgent.596 None of those properties is replicated here. Readers happy to overlook problem (i) but not (ii) may, like Allen (1896), posit a lacuna after 526 (and perhaps then prefer ἐκ δ’ ἐτέλειον; cf. Il. 9.493 γόνον ἐξετέλειον|, Od. 4.7). I suppose a lacuna before it of at least two lines, containing a formulaic speech-introduction and the start of a construction requiring μή. Perhaps the last missing line itself began with μή, like 525, and the omission was caused by eyeskip. The end of Apollo’s speech, which might have given us a hint of the contents here, also falls in a lacuna of several lines (568*). If the two lacunas represent a tear affecting the recto and verso of a single page, it would originally have had 43 lines + the length of each lacuna. A column length of 45–7 lines was certainly viable in late antique codices of hexameter: see Turner (1977) table 16, nos. 94, 100a, 102, 105, 124, 132a, 162a, 167, 171a, 192–3, 199a, 236, 240, 266.

595

See e.g. Call. H.Art. 29, H.Del. 150, 162, 212, CA epic. adesp. 4.1, Q.S. 4.408, 7.650 (speech-within-speech), 12.254 (where Campbell 1981a compares Antimachus fr. 84, Call. Aet. fr. 1.7), 14.602; for Nonnus see Highet 1974 189–90. 596 See Il. 4.301–2 (indirect) + 303–9 (direct); 15.346–7 + 348–51 (Janko 1992 cited Il. 9.684 and Od. 15.424, which have too little indirect discourse to count as parallels); 23.854–5 + 855–8; Od. 1.36–9 + 40–1. The urgent tone is recognized in On the Sublime 27, as Gregory Hutchinson observed to me.

436

C O M M E N TA RY 5 2 6 – 5 2 7 μήτε θεὸν μήτ’ ἄνδρα Διὸς γόνον: With the assumption of a lacuna, Apollo may (for example) be specifying relatives whom he has good reason to treat as φίλοι. The most famous human son of Zeus – Heracles – has a special significance for the external audience (Introduction §2.2). Like Hermes, he will quarrel with Apollo and threaten to steal from Delphi, but their reconciliation is a good deal less secure (LIMC V(1).133–4). 526–7 ἐκ δὲ τέλειον | σύμβολον ἀθανάτων ποιήσομαι ἠδ’ ἅμα πάντων: The transmitted reading is defensible. σύμβολον (ἐκ)ποιήσομαι might in principle mean (i) ‘I shall interpret … as a sign’, (ii) ‘I shall create a token of agreement’ or, as I prefer, (iii) ‘I shall conclude a covenant’.597 Of these, (i) – no matter what object one understands from the lacuna – does not suit the emphasis in ἐκ- (with lengthy hyperbaton: examples in KG i.532, 535–6) and τέλειον (LSJ s.v. I.1b) on achieving full ratification. Gauthier (1972: 62–89, esp. 69–70) favoured sense (ii), and dated to c.400 the development of σύμβολα from tallies, proving a financial or guest-friend agreement, to a bilateral judicial agreement. However, both the singular and the genitives (indicating that this σύμβολον belongs to and will be recognized by everyone) are then difficult, since tallies were fundamentally paired and particular. I therefore prefer the interpretation ‘covenant’. Though I have not found singular σύμβολον in this sense until FD III(1) 486 i.A7 (treaty of Delphi and Pellene, 280s), συνθήκη is already found for συνθῆκαι in similar contexts in Thucydides, and συμβόλαιον in Isocrates Euth. 2. Herman (1987: 46–68) has discussed the recognized procedure for concluding such agreements after hostility: mediation or benefaction (both, in our case) leads to an exchange of pledges, oaths, gifts and formal declarations of friendship. Hermes has run out of gifts and made his oath; Apollo has made an oath declaring friendship (523–5), but still has to negotiate a response to Hermes’ interest in prophecy. His phrasing raises the possibility that after his speech (in the lacuna at 568*) he made the most widespread gesture to encapsulate such pledges in art, namely a handshake. This form of σύμβολον belongs in the worlds of diplomacy and commerce – Hermes’ worlds. More specifically, the fact that it belongs to ‘all’ and not just immortals suits Apollo’s emphasis on how Hermes will organize exchanges among humans (516–17). The wording τέλειον σύμβολον simultaneously suggests a ‘fulfilled omen’. Apollo has just brought to fulfilment the σύμβολον which Hermes identified (as his first word) in the tortoise (30*);

597

(i): e.g. Archil. fr. 218, D. 15.4, Plu. Phoc. 4. (ii): e.g. IG I3 34.11 (c.425?), II2 141.19 (376/5); see Gauthier (1972) 81–4. (iii): e.g. Ps.-D. 7.9–13 (public), PLille I 4 ii.18 (217 bce; private). Feyel (1946: 21) also preferred (iii) here.

437

C O M M E N TA RY 5 2 8 – 5 2 9 our line clarifies that Hermes’ ‘omen’ became the ground of his ‘pact’ with Apollo.598 528 πιστὸν ἐμῶι θυμῶι καὶ τίμιον: In 520, ἐμῶι θυμῶι expressed where Hermes’ show of χάρις and φιλία would affect Apollo; here the words mark the reciprocal movement of loyalty and trust flowing from Apollo’s θυμός into the pact. For exchange of πίστεις as an important part of ritualized demonstrations of friendship in Greece see Herman (1987) 49–50. 529–30 ὄλβου καὶ πλούτου δώσω περικαλλέα ῥάβδον | χρυσείην: The opening description of this staff calls to mind Hermes’ epithet χρυσό(ρ)ραπις (539*). He regularly has a golden staff (besides the passages mentioned below see Hipponax fr. 79.7–9), which is associated with various functions. (a) Explicitly and formulaically, it controls sleep (Il. 24.343–4, Od. 5.47–8, 24.2–4). (b) It is used to herd the souls of the dead at Od. 24.5 (see 210* on herding staffs), while (c) at its other mentions in early epic and Pindar, Hermes is acting as divine messenger and deliveryman.599 (d) Alcaeus’ Hymn to Hermes seems to have involved the idea that the gods’ friendship was sealed by exchanging the lyre for a golden ‘prophetic staff’, explaining the epithet χρυσόρραπις.600 Perhaps this staff’s prophetic nature was derived from (a)–(b) through Hermes’ role in bringing prophetic dreams to sleepers (14*). In any case, our passage probably evoked the Alcaean version of the myth (as in 515*): the golden staff is again a pledge of Apollo’s friendship and an aetiological σῆμα of Hermes’ epithet χρυσόρραπις, but its significance for his prophetic powers is raised only to be substantially restricted in 531–2*. Besides these four functions, Apollo’s gift explicitly creates wealth and prosperity – probably a symbolic reading of its goldenness. Some passages suggest that Hermes creates wealth for others, e.g. Hy. 29.8 χρυσόρραπι δῶτορ ἐάων, Cornutus ND 16.3, Epictetus 3.20.12. But here the focus is more on Hermes’ stated interest in becoming wealthy himself (171, 379, 473), since Apollo is echoing his promise in 461–2, ὄλβιον

With my interpretation, τέλειον σύμβολον ἀθανάτων preserves a neat syntactic match for 518 θεῶν μέγαν ὅρκον. Richardson proposed the emendation πάντως to escape the problem of explaining why Apollo is t­ hinking of humans here; Vergados accepted this with the further emendation ἀθάνατον earlier in the line. 599 For Hermes’ staffs, see De Waele (1927). By contrast, the magic wands of Circe and Athena in Od. are used in physical transformations (10.238; 13.429, 16.172). The echo of 208 φέριστε in 533* may support the connection between Hermes’ earlier ῥάβδος and our passage. 600 ΣD Il. 15.256. There is some doubt whether all of this is from Alcaeus, but see Cairns (1983). For the prophet’s golden staff, cf. that of Chryses at Il. 1.15. 598

438

C O M M E N TA RY 5 3 0 ἡγεμόν’ ἕσσω, | δώσω τ’ ἀγλαὰ δῶρα. For a god, having prerogatives leads to cult-honours and so indirectly to wealth, as Ilgen observed. The phrase, and 532 ἐπικραίνουσα, influenced the author (Soterichus in the fifth century ce, or Andronicus in the fourth?) of On the Foundation of Hermopolis 4–7, 14–17: Zeus/God gave Hermes a ῥάβδον χρυσέην διακοσμήτειραν ... ἵνα κρήνειεν the order to create the world and a city fit for it; he uses it to create φιλότης where there had been strife among the elements (cf. Agosti 2016: 231–4, Livrea 2002). Hermes also receives a staff from Apollo at this point in Hyginus Ast. 2.4 (interpreting it as the kerykeion, as do hypothesis a to Pindar’s Pythians, Servius on A. 4.242); in Pherecydes FGrH 3 F131 he received staff (a) from Apollo while the latter was herding Admetus’ cows, but the exact narrative context is lost. 530 τριπέτηλον, ἀκήριον ἥ σε φυλάξει: Preller (1846: 514–15) saw that ἀκήριον must go with σε rather than with what precedes. Compare Od. 20.47 διαμπερὲς ἥ σε φυλάσσω, and 510* for a predicative adjective fronted in a relative clause. The staff which will keep Hermes ‘free from the Keres’ is his kerykeion, which is by far his most common staff in art from the Chigi olpe (c. 640) on, and which conferred diplomatic inviolability on his heralds.601 A reference to it here helps Apollo fulfil his promise from 461–2 to make Hermes a ‘leader’. ἀ-κήρ-ιον in this context probably constitutes a wordplay on κήρ-υξ. However, the kerykeion is not the same as the golden staff of 529, and the normal distinctive 8- or ȣ-shaped top is pointedly not τριπέτηλον, even if with some generosity a few kerykeia in art could be considered ‘threeleaved’.602 The poet wilfully hinders us from a simple identification of what is a hybrid of Hermes’ different staffs. The three leaves point in two new directions. Divine sceptres are regularly topped with three leaves, though not often carried by Apollo or Hermes.603 And secondly there is a double connection to the ῥάβδος which Hermes carried as a cattle-prod at 210*: it may well have been a leaves-and-all improvisation like the sandals (84), and Hermes

601

Mosley (1973) 81–9. In Epos, heralds have staffs symbolizing their authority to speak; later, they carry the kerykeion or caduceus with its more specialized form. For kerykeia see Hirzel (1907) 73–82, Crome (1938–9), Hornbostel and Hornbostel (1988). 602 LIMC VIII s.v. Kerykeion, esp. the ψ-shaped type F, on e.g. LIMC s.v. Hermes 526b, 646–7. Contrast the approach of Anghelina (2017). The  Cor­­­inthian pinax where Hermes’ kerykeion once had triple leaves in each of its rings (LIMC s.v. Amphitrite 14, Chittenden 1947 pl. 20a) remains a one-off, perhaps motivated by the prominence of Poseidon’s trident in those dedications. τριπέτηλος does not otherwise apply to staffs. 603 See e.g. LIMC s.vv. Aphrodite 42, Hera 149, Zeus 67. Siebert (1985: 278) emphasized the lack of correlations between sceptre-type and bearer.

439

C O M M E N TA RY 5 3 1 – 5 3 2 is depicted with a cattle-prod with three leaf-shaped metal prongs by the Theseus Painter.604 See further on 531–2* at end. 531–2 πάντας ἐπικραίνουσα θεμοὺς ἐπέων τε καὶ ἔργων | τῶν ἀγαθῶν, ὅσα φημὶ δαήμεναι ἐκ Διὸς ὀμφῆς: Ludwich (1908) found the correction θεμούς, based on Hesychius θ.255 θεμούς· διαθέσεις, παραινέσεις, for which no other source is known. The manuscripts’ θεούς leaves no construction for the following genitives, and is a difficult object of ῥάβδος ἐπικραίνουσα (contrast 427–8). The meaning of θεμός is unclear, but Hesychius’ παραινέσεις may well be right for our passage. ὑπό-θε-σις, which is presumably cognate, means both ‘basis’ (cf. θέμε-θλα < θεμός) and ‘suggestion, advice’ (cf. the name Them-andros?).605 Od. 4.163 ἤ τι ἔπος ὑποθήεαι ἠέ τι ἔργον therefore justifies construing ἐπέων τε καὶ ἔργων as objective genitives.606 If θεμός did mean roughly ‘advice’, it contains wordplay, since Apollo’s advice came in the form of θέμιστες. 532 echoes both 471 φασι δαήμεναι, where Hermes pretended to second-hand knowledge about prophecy, and 444 φημι δαήμεναι, where Apollo commented on Hermes’ lyric ability. These are the key skills now being negotiated; see further 537*. ἐπικραίνουσα recalls the marked idea in 427* that Hermes’ lyre had the power to κραίνειν. This was the start of its alignment with an oracular voice, which justifies Hermes’ hints in 468–73 that a share in Apollo’s prophetic skills could fairly come into their bargain. At 460*, Apollo swore on a κρανάϊνον javelin, which probably alluded to the power of staffs to κραίνειν, and this is in our mind because 529–30* echoed the terms of his promise at 461–2 to make Hermes prosperous, and toyed with Alcaeus’ interpretation of Hermes’ golden staff as prophetic. In 530* the staff morphed into the kerykeion, and now that too is connected through ἐπικραίνουσα θεμούς to prophecy. Hermes and heralds will use it to add their official ratification to Apollo’s oracular pronouncements: centres such as Delphi often advised (e.g. military, religious, ‘colonial’) policies which had to be implemented with the agreement of other states through heralds. For example in 435 bce the Epidamnians consulted Delphi on whether to hand control of Epidamnos to Corinth, were advised to do so, and sent a delegation to Corinth to transact the business.607

604

LIMC s.v. Hermes 246 (c. 500 bce). D’Amicis et al. (1994) 228 offer a clear image. For -μός attached to a verb-stem see Buck and Petersen (1945) 184. Uncertainty remains about how θεμόω attained the sense ‘impel’ in Od. 9.486, or what Themotheos (Eretria, third century) thought his name meant. 606 The hymnist preferred to ignore the digamma in ἔργον than to write ἐπέων ἰδὲ ἔργων. 607 Th. 1.25.1–2. For topics addressed to the Delphic oracle, see Fontenrose (1978) 39–41, 244–67. Jaillard (2007: 234) similarly related Herm. 531–2 to the vicarious authority wielded by heralds. 605

440

C O M M E N TA RY 5 3 3 This interpretation clarifies τῶν ἀγαθῶν, which is not neuter (Delphic advice was notoriously liable not to be ‘good’, for example through the legend of Croesus, and Apollo will confirm as much at 541*) but masculine, referring to the high status rather than the morality of the people who worked with heralds to consult Apollo and implement his advice.608 Apollo is effectively offering Hermes a socially prestigious forum for collaborating with him. However, his emphasis in πάντας and ὅσα is false advertising, since on no account of the line is Hermes’ staff involved with ‘all’ prophecies. Nor, conversely, was the staff-bearing heralds’ power of performative speech limited to such a collaboration with Apollo; for example, they announced athletic victories. Apollo in these four lines has offered Hermes an object which encapsulates a range of his connections to staffs with different functions, in line with the general tendency of the final part of Herm. to fit in as many of Hermes’ roles as possible (Introduction 3.3.6). The staff is identifiable as the famous one which controls sleep; it is like a prophet’s staff, but not as straightforwardly as Alcaeus said; it brings wealth for Hermes and perhaps also for those he favours; its three ‘leaves’ perhaps resemble a divine sceptre, or a herder’s rustic staff or κέντρον such as Hermes carried at 210; it confers the inviolability of the kerykeion, and hints at the complementary role of oracles and heralds in major transactions. In all this, Apollo appears to be fulfilling his promise to make Hermes prosperous and a leader (460–2); it is the first of two roles in prophecy (cf. 550–66) which he will offer in lieu of the Zeus-derived prophetic knowledge which Hermes was interested in at 468–73. 533 μαντείην δέ, φέριστε διοτρεφές, ἣν ἐρεείνεις: After allowing Hermes a limited role in implementing prophecies, Apollo smoothly approaches the central point of Hermes’ comments from 469–72. As at 208, the complimentary φέριστε mitigates a face-threat, namely Apollo’s refusal to let Hermes share his knowledge of Zeus’s mind. διοτρεφές asserts Zeus’s care for Hermes, to compensate for this exclusion. ἣν ἐρεείνεις limits μαντείην: Hermes will later be allowed another prophetic skill, just not the Zeus-derived knowledge he ‘asked about’. The verb excited analyst critics, since Hermes did not ask Apollo direct questions about prophecy. Ilgen’s rebuttal has not been bettered: ‘nec interrogat directe, nec, ut communicetur secum, petit diserte, sed velut aliud agens, per circuitum in eius mentionem incidit; ex quo astutia eius maxime elucet. Idem vero fecerat Apollo; non expressis verbis petierat a Mercurio citharam … 608

Low-status consultants, such as Aesop (Fontenrose 1978: Q108), appear only in later sources. See also n. 627.

441

C O M M E N TA RY 5 3 4 – 5 3 6 At enim Mercurius, callidus et gnavus verborum Apollinis interpres, facile intellexit, quid laudes eius sibi vellent.’609 Lines 468–73 signal that Apollo’s prophetic arts are interesting, desirable, a reason for Hermes to be nice to him, and a model of divine gift-giving (cf. 470* ἀγλαὰ δῶρα). Pragmatically this amounts to a broad hint, which Hermes has primed Apollo to notice by commenting, correctly, that Apollo’s own questions about the lyre had an ulterior motive (464*; cf. the hints at 437, 457). Apollo’s words in 534 imply that he did perceive Hermes’ words as a request for, not just an enquiry about, prophecy. But at first he uses the politeness strategy of minimizing the force of Hermes’ hint. 534 οὔτε σὲ θέσφατόν ἐστι δαήμεναι: θέσφατον (also in 540) and δαήμεναι engage directly with Hermes’ words from 471–2*. 534–5 οὔτέ τιν’ ἄλλον | ἀθανάτων: More politeness: Hermes is not alone in being excluded. 535 τὰ γὰρ οἶδε Διὸς νόος: Matthiae suggested τά, which would refer  to the contents of the μαντείη mentioned in 533.610 The γάρ-clause expands on the exclusion of ‘other’ gods by describing the access of Zeus and Apollo to the special knowledge; appeals to Zeus’s knowledge traditionally contrast it with other people’s inability to foresee future possibilities.611 Ω’s reading τό is much less easy to take in the loose way suggested for τά. It could refer to the concept of a certain sort of μαντείη, or to the preceding proposition that nobody else may learn it. But Zeus does not ‘know’ such prophecy (his knowledge constitutes its basis), and his knowledge of the proposition does little to explain (cf. γάρ) its truth. 536 πιστωθεὶς κατένευσα καὶ ὤμοσα καρτερὸν ὅρκον: Apollo’s loyalty to his former nod and oath is a sign of the value he places on the similar gestures just completed at 521–5, and of the ὁσίη (470) through which he maintains Zeus’s favour. This interpretation of Apollo’s underlying point suggests that

609

Ilgen 469. Similarly Sowa (1984) 195, Fernández Delgado (1990) 217–18, Adorjáni (2012). For analyst arguments see e.g. Matthiae (1800) 40–8, Hermann lxxiv, Arfelli (1905) 406–7, Brown (1947) 143. 610 The lack of a clear antecedent could have contributed to the corruption, but for the minuscule confusion of ο/α cf. 155 τόδε*. 611 Cf. Il. 3.308, Od. 14.119, 15.523, West (2014) 8. In the first two passages the human speaker believes that the other gods will also have access to fate, which Apollo is here denying.

442

C O M M E N TA RY 5 3 7 – 5 3 8 πιστωθείς is better taken to indicate that he gave Zeus pledges, rather than that he received them from Zeus.612 In the absence of further detail the aorist was probably not temporal (indicating that Apollo made a separate pledge prior to swearing), but coincident, so that the oath was the pledge: cf. Th. 4.88.1 πιστώσαντες ὅρκοις ‘having extracted a guarantee by oaths’, IG II2 5007 ὅρκον πιστόσασθε (purportedly dating to 510s). 537–8 μή τινα νόσφιν ἐμεῖο θεῶν αἰειγενετάων | ἄλλόν γ’ εἴσεσθαι Ζηνὸς πυκινόφρονα βουλήν: The use of νόσφι in the secondary sense ‘except’, with an emphatic pronoun, a contrast with any other god, and a verb of knowing, echoes the hymn’s other use of νόσφι at 444–6 οὔ ... δαήμεναι ... τιν’ ἀθανάτων … νόσφι σέθεν. (Cf. 532* for a priming echo of 444 φημι δαήμεναι.) The passages bring out the shared relationship of Hermes’ musical knowledge and Apollo’s prophetic knowledge to δαήμεναι, and so their potential exchangeability. But, crucially, only Hermes’ knowledge can be shared. Zeus’s plan (see 10*) is here πυκινόφρονα, a double-edged adjective. It could indicate (politely) a similarity between Zeus and Hermes (ὃς ἐπὶ φρεσὶ πευκαλίμηισι κέκασται, Il. 20.35) as ‘planners’.613 Apollo has changed the epithet he preferred at H.Ap. 132 (and 252, 292), where he wished to prophesy Διὸς νημερτέα βουλήν. In its detailed engagement with that passage, Herm. appropriated νημέρτεια for Hermes, if only to show his impudence (369*); in what follows, Apollo will project a more Hermetic and less ‘unerring’ vision of his oracles (541–9*). On the other hand, etymologically the metaphor in π(ε)υκ- is of stitching together what we might call a ‘watertight’ plan, and the uses of πυκάζω at Hes. Op. 542 and 793 suggest that the metaphor was still perceptible. Such a plan is unassailable to outsiders, and specifically recalls 523 πυκινῶι δόμωι, of Apollo’s temple: Zeus’s plan is therefore as impenetrable to Hermes as Delphi, now that Hermes has sworn not to approach it. After μή τινα νόσφιν ἐμεῖο …, ἄλλόν γ’ at line-start is redundant. Apollo is rhetorically preparing to say that Hermes must fall in line, but pleonasm is also a recurrent feature of Greek oaths: e.g. Il. 1.88–90 οὔ τις ἐμεῦ ζῶντος καὶ ἐπὶ χθονὶ δερκομένοιο | σοὶ κοίληις παρὰ νηυσὶ βαρείας χεῖρας ἐποίσει | συμπάντων Δαναῶν, οὐδ’ ἢν Ἀγαμέμνονα εἴπηις, IG I3 1453 § 10 (addition to

The passive can mean both: cf. Od. 15.436 ὅρκωι πιστωθῆναι ‘give guarantee by an oath’ and E. IA 66 ἐπιστώθησαν versus Od. 21.218 πιστωθῆτον ‘receive a guarantee’ (of Odysseus’ identity) and S. OC 1039 πιστωθείς. These pledges are absent from H.Ap.: Vergados 8. 613 For πυκ- and mental activity (almost always planning or advising), see L ­ fgrE s.v. πυκ(ι)νός B 3b, πυκιμήδης, and formulaic φρεσὶ πευκαλίμηισιν. 612

443

C O M M E N TA RY 5 3 9 – 5 4 0 councillor’s oath c.448) ἐάν τις κόπτηι νόμισ[μα] ἀργυρίο ἐν τῆσι πό[λεσι] καὶ μὴ χρῆται νομ[ίσμασιν τοῖς] Ἀθη[να]ίων ἢ σταθμοῖς ἢ μέτ[ροις ἀλλὰ ξενικοῖς νομίσμασι]ν καὶ σταθμοῖς καὶ [μ]έτροις, Demosthenes 24.151 (alleged juror’s oath) ἀκροάσομαι τοῦ τε κατηγόρου καὶ τοῦ ἀπολογουμένου ὁμοίως ἀμφοῖν. For the accentuation ἄλλόν γ’ see West (1998) xviii. 539 κασίγνητε χρυσόρραπι: Both vocatives continue the politeness with which Apollo blunts his refusal of Hermes’ wish. Following 529–32, χρυσόρραπι underscores that Hermes will get a supporting role in bringing Zeus’s plan, disseminated by Apollo, to completion. Vocative κασίγνητος, while fairly conventional in later prose (Dickey 1996: 88), is no platitude in Epos: four of the other five contexts are emotionally fraught (Il. 4.155, 5.359, 18.52, 21.308), while at Od. 4.810 Penelope appeals to the dream of her sister immediately after her plaints about being abandoned by Odysseus and Telemachus. This is the rhetorical climax of Apollo acknowledging his kinship with Hermes more and more clearly through his vocatives (see 455*; Greene 2005: 347). For κασίγνητος used of half-siblings, see LfgrE s.v. G. μή με κέλευε: Line 533* understated the point of Hermes’ comments about prophecy, as something he ‘asked about’ rather than indirectly asked for. Now Apollo overstates them as continual (imperfective aspect) instructions. The general context of politeness suggests that the exaggeration is jovial, as in 515*. 540 θέσφατα πιφαύσκειν ὅσα μήδεται εὐρύοπα Ζεύς: ὅσα contributes to the exaggeration seen in 539* κέλευε: Apollo suggests that Hermes is asking for a wholesale revelation, beyond the normal piecemeal revelations of prophecy. The contrast with πυκινόφρονα (537–8*) activates the root meaning of πιφαύσκω, ‘bring to light’.614 μήδεται continues the politeness strategy seen there of aligning Zeus with Hermes, since the stem μηδ- applied to Hermes at 46, 389, 456. The context of Zeus having foresight whose dissemination is restricted favours the interpretation of εὐρύοπα as ‘far-seeing’ rather than ‘far-speaking’; both interpretations are found in antiquity (the latter at e.g. PMG 1008, Ap.S. p. 79, and implied in Pindar P. 6.24 βαρύοπα). The epithet, originally an accusative (Debrunner 1927: 188–90), was taken as indeclinable like μητίετα, with which it makes a formulaic system.

For πιφαύσκω and prophecy see e.g. Od. 2.162, A. Ch. 279, Eu. 620. φαίνω is itself used of prophecy, e.g. Hdt. 1.159.2, A.R. 2.315; on Ps.-Manetho Apotel. 2.208 see 544*.

614

444

C O M M E N TA RY 5 4 1 523 αὐτάρ–540. Apollo’s φιλία for Hermes: After Hermes’ promises, Apollo’s side of the bargain follows. The transition is marked by a pair of φιλ- words (524–5) as Apollo swears, in language coloured by Hermes’ future expertise in diplomacy and commerce (524, 526–7), to a special friendship. This is sealed with further gifts, of which the first is a hybrid of a divine sceptre, a cattle-prod, Hermes’ normal golden staff, and his kerykeion (529– 32). While Apollo implies that he is fulfilling his promise at 461–2 to make Hermes prosperous, he knows that Hermes has a further interest in prophecy which he is not allowed to satisfy, and that Hermes is still willing to be tricky (507–23*). Apollo is concerned to mitigate the risks of how Hermes might react to disappointment. The first is to insist that the staff (primarily in the guise of a kerykeion) will provide a prestigious way for Hermes to collaborate with Apollo’s oracles – ratifying the diplomatic business which they prompt. Only later does he tackle Hermes’ desire for prophecy more directly (533–40), with a range of politeness strategies: deflating that desire (533), humorously overinflating it (539–40), acknowledging how Hermes set up prophetic knowledge and knowledge of the lyre as substitutable (534*; 537–8); explaining his refusal in terms of a binding pledge which simultaneously implies how seriously he will take his current pact with Hermes (536); finally being open about their close kinship (539). The politeness will continue in a new vein in 541–9 as Apollo imprints Hermes’ character into his own oracular pronouncements, before the further gift of a minor prophetic role follows in 550–66.

541–9 541 ἀνθρώπων δ’ ἄλλον δηλήσομαι, ἄλλον ὀνήσω: Apollo switches with |ἀνθρώπων from the divine focus of 534–40. He then with lapidary succinctness contrasts unfortunate and fortunate consultants, which he will expand chiastically in 543–5 versus 546–9.615 The future tenses do not imply that Apollo has never acted in such an ambivalent way before; their emphasis engages the attention of audience-members, who are examples of the future ἄνθρωποι affected. Our two main sources for the verb δηλέομαι (Epos, Herodotus) both apply it to pillaging, whose relevance is clear in 542*. Elsewhere in epic δηλέομαι involves direct physical injury when people are the object, and

For similar phrasing cf. Il. 12.267–8 ἄλλον μειλιχίοις, ἄλλον στερεοῖς ἐπέεσσιν | νείκεον, 18.536–7 ἄλλον ζωὸν ἔχοντα νεούτατον, ἄλλον ἄουτον, | ἄλλον τεθνηῶτα, Od. 4.692 ἄλλόν κ’ ἐχθαίρηισι βροτῶν, ἄλλόν κε φιλοίη.

615

445

C O M M E N TA RY 5 4 2 – 5 4 3 this background probably increases the verb’s harshness here, even if in Herodotus the causation is sometimes less direct.616 So far from bringing the clarity promised by his birthplace (Δῆλος ~ 540 πιφαύσκει), Apollo can bring δήλησις.617 The word-choice is relevant to Herm.’s engagement with the Hymn to Apollo, where Apollo’s beneficial foundation at Delphi is premised on killing off a δήλημα βροτοῖσιν, the Python (H.Ap. 364). Our hymnist hints that Pytho got more than its name from the beast (Introduction §1.3). Gods were often said to have power both to help and to harm, e.g. Od. 16.212 ἠμὲν κυδῆναι … ἠδὲ κακῶσαι, Hes. Th. 442–7, Op. 5–6; a similar antithesis applies to Hermes at 577–8. Seers were subject to a good deal of scepticism in ancient sources (for epic passages see LfgrE iii.32.41–4); it is rarer to blame the god for misleading you (see 545*), and rarer still to represent the god joking about it. 542 πολλὰ περιτροπέων ἀμεγάρτων φῦλ’ ἀνθρώπων: Od. 9.465 |πολλὰ περιτροπέοντες, A.R. 2.143 περιτροπάδην, and περιελαύνω in several passages refer to rounding animals up in order to steal them. Apollo thereby activates the subsidiary sense ‘pillage’ in 541 δηλήσομαι, and presents his oracular self as a rustler – i.e. as Hermetic. (He does not realize the particular interaction we may perceive with Hermes’ combination of herding and cheating in 167* βουκολέων.) Hermetic too is the startling reimagination of θεωρία to major oracular sites such as Delphi as a process of corraling a lesser species from many locations in order to fleece them, or at least to exploit their economic potential. Again contrast the Hymn to Apollo, particularly 537–8, where the περικλυτὰ φῦλ’ ἀνθρώπων will bring gifts to Delphi and be welcomed in return (Introduction §1.3). That model of healthy χάρις-exchange is broken here when Apollo replaces the περικλυτὰ φῦλα with ones made up of ἀμεγάρτων people. The adjective allows for pity (e.g. Od. 17.219 ἀμέγαρτε συβῶτα); Ap.S. p. 25 glosses ὃ οὐκ ἄν τις ζηλώσειεν διὰ χαλεπότητα.618 543 καὶ μὲν ἐμῆς ὀμφῆς ἀπονήσεται: καὶ μέν signals an expansion of the ‘headline’ in 541 (cf. Denniston 390). For ὀμφή see 471*. The emphasis on its ὄνησις here and in 541 and 545 perhaps recalls the appearances of that stem in 30–8 as a concern characteristic of Hermes and fulfilled by the lyre,

616

E.g. at 9.63.2 soldiers are killed as a consequence of their clothing weighing them down. Apollo’s (less direct) mode of harming consultants will become clearer in 542 περιτροπέων, 545 ἀπατήσω. 617 For Δῆλος and δῆλος see e.g. Call. H.Del. 52–3. 618 See below, 541–9*. Ancient scholars sometimes construed ἀμέγαρτος as ‘unstinting’ (Orion p. 22, ΣDT Il. 2.420b); this does not suit early usage, but cf. Opp. H. 4.412 πολλὴν ... ἄγρην ... ἠδ’ ἀμέγαρτον.

446

C O M M E N TA RY 5 4 4 whose value in terms of prophetic gifts is now being negotiated. Cf., slightly differently, Reece (1997) 38. 544 φωνῆι τ’ ἠδὲ ποτῆισι: Ψ had φωνῆ καὶ πτερύγεσσιν. These are prob­ ably ancient variants, since both fit augury practices and neither is a plausible concordance interpolation, or gloss on or correction of the other. ποτῆισι elsewhere is confined to the singular, and nearly all other uses derive from Od. 5.337.619 The exception may provide a terminus ante quem for our variant: in Ps.Manetho Apotel. 2.207–8 the planet Mercury creates good prophets, μακάρων ἄπο θέσφατα θνητοῖς | φαίνοντα [cf. Herm. 540 θέσφατα πιφαύσκειν] πτηνῶν τε ποτῆς ἄπο θεσπίζοντας. Mercury’s involvement makes Ps.-Manetho’s use of Herm. suitable.620 For datives expressing the cause of motion compare E. Ph. 1043 ἔβα Πυθίαις ἀποστολαῖσιν, Schwyzer ii.167. Auspices principally motivated oracular consultations by causing the initial decision to consult or by confirming the correctness of such a decision.621 Overall the paradosis mentions three aspects of auspices: bird-cries, flight-paths (M) and wing-movements (Ψ). Cries are mentioned as a major part of bird-omens at e.g. Il. 12.207, Pi. fr. 285, X. An. 6.1.23, Arr. An. 1.25.6, Plu. Soll. An. 975a ἐπί τε κίνησιν ἐπί τε φωνὰς καὶ γηρύματα καὶ σχήματα (the modes of bird-signs); they are crucial when sight is ruled out, e.g. for Tiresias (A. Th. 25, S. Ant. 1001–4, OR 31), or at night (Il. 10.276, or the screech-owl’s interventions). Flight-paths are also a standard feature of augury, particularly the basic mapping left vs right ~ u ­ nfavour­able vs favourable. Wing-motions are essential to the ominous quality of the eagles’ flight at Od. 2.148–51, and ‘wings’ are a source of or synecdoche for augury at E. Ion 377 πτερά (secl. Badham), S. OC 97 πτερόν, Call. Lav. Pall. 124 πτέρυγες (in another contrast between ‘good’ and ‘vain’ bird-observations), Aet. fr. 43.66 πτερύγεσσι, Hec. fr. 47.9. The core principle of our only early ‘technical’ text 619

The line was only present in some Alexandrian copies, but influenced Alex. Aet. fr. 5.5 (probably), Aratus 278, scholia and lexicographers. 620 As in 413* and n. 639. Hermes was a patron deity for astrologers, and Dorotheus of Sidon’s work, an important source for Ps.-Manetho, presented itself as written from ‘Dorotheus the Egyptian to his son Hermes’. The influence of Mercury or of its zodiacal ‘house’ (Gemini and Virgo) at a person’s birth repeatedly in Apotel. 2 spells the presence or absence of Hermetic character-traits. See also 566*. 621 The similar phrasing at Hes. fr. 240.10–11, where a favoured consultant at Dodona ἐξερεείνηι | δῶρα φέρων τ’ ἔλθηισι σὺν οἰωνοῖς ἀγαθοῖσιν, probably suggests the latter. In IMagn. 17 (third century) Magnesian exiles are told in one prophecy to await a bird-omen (crows turning white) before returning to consult about their homecoming. For other kinds of omen preliminary to oracular consultation see the Athenian and Tetrapolis Πυθαΐδες, which set out only after lightning and extispicy respectively (Boethius 1918: 1–12, Σ S. OC 1047); cf. Fontenrose (1978) 418, 442.

447

C O M M E N TA RY 5 4 5 – 5 4 6 about augury (IK Ephesos 1678a, late sixth century) employs a combination of these factors: a bird flying right to left is favourable unless it raises its left wing (above horizontal?); conversely a bird flying left to right is unfavourable unless it raises its right wing.622 τεληέντων οἰωνῶν: Probably not referring to a particular set of species as a fourth aspect of augury: a range of birds could be auspicious, but not even the eagle had to be so (see further 546*). τελήεις is an epicism whose normal reference is to hecatombs; here it recalls the description of the eagle in omens as τελειότατον πετεηνῶν (Il. 8.247, 24.315; cf. Posidippus 27.4). For related words expressing the idea that oracles will come to fruition see e.g. Tyrt. fr. 4.2 τελέεντ’ ἔπεα of a Delphic oracle, Pi. P. 5.62 ὄφρα μὴ ... ἀτελὴς γένοιτο μαντεύμασιν (of Apollo), IG I3 1163.41 φράζεσθαι λογίν πιστὸν ἔθεκε τέλος (446 bce). 545 οὐδ’ ἀπατήσω: Apollo’s promise of non-deceit to the favoured visitors replicates his promise to Hermes in 462, ἐς τέλος οὐκ ἀπατήσω. He implicitly specifies what he meant in 541 δηλήσομαι. Oracles ‘deceived’ through a misleading ambiguity whose natural interpretation was incorrect.623 Croesus charges Delphi with ἐξαπατᾶν at Hdt. 1.90.2, though the Pythia then defends herself. At Luc. DDeor. 18.1, Hera comments that Leto’s son ἐξαπατᾶι ... ἐπαμφοτερίζοντα πρὸς ἑκάτερον τῆς ἐρωτήσεως ἀποκρινόμενος. In Plutarch, Theon figures that such ἀμφιβολίαι were taken as ‘composed like refuges to duck into if one’s prophecy slipped up’, and relates their versification to the charlatanism of oracle-mongers (Pyth. Or. 407b–c). 546 ὃς δέ κε μαψιλόγοισι πιθήσας οἰωνοῖσιν: The unusual description of squawks as λόγοι suggests that for Apollo the cries of birds are as comprehensible as language, unlike for the humans who misplace trust in them. Other seers are also said to understand the language of birds, such as Melampous and Tiresias at Apollodorus Bibl. 1.97, 3.70. For the idea that not all birds are ominous see e.g. Od. 2.181–2 ὄρνιθες δέ τε πολλοὶ ὑπ᾽ αὐγὰς ἠελίοιο | φοιτῶσ’, οὐδέ τε πάντες ἐναίσιμοι (though Eurymachus is wrong in that instance), Call. Lav. Pall. 123–4 γνωσεῖται δ’ ὄρνιχας, ὃς αἴσιος οἵ τε πέτονται | ἤλιθα. 622

For bibliography on augury see 213*. Birds’ flight, wings and sound could also be used in forecasting the weather: Ael. NA 7.7 < Arist. fr. 270 §21 Gigon. 623 Fontenrose (1978: esp. 235–6, 437–8) argued that Delphic obscurity was just a trope of literature and post eventum story-telling about oracles. Parker (1985: 301–2) countered that a totally demystified oracle would have been psychologically and politically unsatisfactory. For stories of misleading oracles see e.g. Hdt. 1.53.3, 66.2, 165–7, 6.80, 9.33.2, Th. 1.126.4. Many others produce bafflement rather than misdirection.

448

C O M M E N TA RY 5 4 7 – 5 4 9 Apollo’s antithesis (ὅς τις ἄν vs ὃς δέ κε) recalls Hermes’ contrast between the lyre’s successful and unsuccessful questioners (482 ὅς τις ἄν vs 486 ὃς δέ κεν). μαψιλόγοισι, besides opposing φωνῆι … τεληέντων in 544, recalls Hermes’ application of μάψ to language in 488*, and copies his metaphors about the lyre’s human voice (479 ἀγορεύειν, 484 φθεγγομένη, 488 θρυλίζοι). See further 541–9* below. 547–8 μαντείην ἐθέληισι παρὲκ νόον ἐξερεείνειν | ἡμετέρην, νοέειν δὲ θεῶν πλέον αἰὲν ἐόντων: παρὲκ νόον could mean ‘senselessly’ (as in Il. 20.133 μὴ χαλέπαινε παρὲκ νόον) or ‘beyond (my) intention/expectation’ (as in Call. Hec. fr. 8 παρὲκ νόον εἰλήλουθας); see also Faulkner (2008) on H.Aphr. 36. The latter sense is compatible with a wider tonal range including e.g. pity at human fallibility, though such a generous interpretation is soon quashed. The continued imitation of Hermes’ pair of lyre-players is visible in ἐξερεείνειν (~ 483, 487), and suggests that the bad consultants here have similar mental shortcomings to the bad player, both in ignorance and lack of humility (487 νῆϊς, ἐπιζαφελῶς).624 The sequence παρὲκ νόον … νοέειν δὲ θεῶν πλέον sarcastically adds that the consultants’ aspirations for νοεῖν are paradoxically ‘senseless’, or involve pitting one’s νοῦς against Apollo’s. Either way, they are hubristic and such a consultant needs rather to ‘know himself ’, as in the injunction inscribed at Delphi at least by the time of Plato’s Charmides (164d), and possibly by Herm.’s date.625 The prominence of ἡμετέρην (line-start with hyperbaton, following on from ἐμῆς in both 543 and 545) asserts that Apollo’s knowledge, which is special even among the gods, cannot be approached presumptuously, however things might be in other oracular shrines. The swift shift from human error (546) to hubris (548) is evidently hyperbolic: not only was it quite possible for humans to make innocent mistakes about augury, but no consultant went with a view to ‘knowing more than the gods’. 549 φήμ’: The parenthetic usage is very rare, except in reporting conversation, and so marked. Cf. Plato Grg. 526c ἐνίοτε δ’ ἄλλην [sc. ψυχὴν ὁ Ῥαδάμανθυς] εἰσιδὼν ὁσίως βεβιωκυῖαν καὶ μετ’ ἀληθείας, ἀνδρὸς ἰδιώτου ἢ 624

For the further connection between the bad consultant and Hermes, see below on 541–9*. 625 PV 309 γίγνωσκε σαυτόν + 327 μηδ’ ἄγαν … (Okeanos to Prometheus) perhaps already alludes to the pair of Delphic injunctions γνῶθι σαυτόν and μηδὲν ἄγαν. See Tränkle (1985) 20. Vergados compared our line with Ion’s warning to Creusa against trying to force Apollo into revealing what he does not want (E. Ion 373–80); but what irritates Apollo here is posing questions (whether awkward or not) based on insignificant auspices.

449

C O M M E N TA RY 5 4 1 – 5 4 9 ἄλλου τινός, μάλιστα μέν, ἔγωγέ φημι, ὦ Καλλίκλεις, φιλοσόφου …, Herodas 4.50 μαρτύρομαι ... μαρτύρομαι, φήμ’, Hp. Cord. 9 διὰ τοῦτο δέ, φημί, καὶ φλεβία μὲν ἐργάζεται τὴν ἀναπνοὴν ἐς τὴν ἀριστερὴν κοιλίην, ἀρτηρίη δ’ ἐς τὴν ἄλλην (Hellenistic in date). Greek probably inherited clause-initial φᾶμι, but φημί was the normal accent in the historical period: Wackernagel (1877) 467. ἁλίην ὁδὸν εἶσιν, ἐγὼ δέ κε δῶρα δεχοίμην: The man’s pilgrimage will be in vain, like the response given to the bad lyre-player (488 αὔτως). Apollo gets gifts, and the situation contrasts with 545 οὐδ’ ἀπατήσω: this consultant will get a misleading response rather than be excluded from one altogether. Similarly Delphi in the sixth century had supposedly received Croesus’ gifts and promptly given an ambiguous response about invading Persia (Hdt. 1.53); cf. 542* περιτροπέων, with the implication that Apollo will corral humans for economic gain. Although one could not simply purchase benevolence from a Greek god (as Croesus found out), promoting future favour (χάρις) was one of the underlying principles of dedicatory practices. Apollo plays along with the jokes Zeus and Hermes made about his acquisitiveness at 330–2, 494–5 (Graefe 1963: 517), while blaming human error for any disappointment. Fickle gift-exchange was also associated with wooing women (e.g. Od. 18.286–7, Sappho fr. 1.22, Ar. Thesm. 345–6), so that Apollo is simultan­ eously acknowledging Hermes’ implication at 484* that his lyre resembled the Pythia as well as a courtesan. 541–9. Reimagining the oracles: Having refused to give Hermes a share in Zeus-derived prophecy, Apollo mitigates this disappointment by describing the behaviour of his oracles in terms which have a great deal to do with Hermes, and which confirm that Hermes’ idea of exchanging the lyre for prophecy was reasonable.626 Apollo is like Hermes in that he demonstrates the ability to reimagine oracles, in his focus on ὄνησις and economic gain (541, 543, 545, 549), and in his self-presentation at the start as ‘rustling up’ visitors (542). To complement Hermes’ signs which were unambiguous yet misleading (the hoof-tracks), and meaningful but too open-ended to be interpretable (the sandal-prints; the fart and sneeze), Apollo’s oracle involves an auspice where the sign may or may not be meaningful, followed by a prophecy which may be misleadingly ambiguous (see also 550–66*). His area of expertise is also pointedly like that of Hermes at 478–88, in the similar antithetical division 626

Matthiae uncomprehendingly commented on 541 ‘ista sententia nescio an melius in Mercurium cadat’ (1800: 309). Schneidewin (1848: 696) began the process of interpreting the similarities. Eitrem (1906: 280) did so in terms of politeness; see also Sowa (1984) 196.

450

C O M M E N TA RY 5 4 1 – 5 4 9 of felicitous and infelicitous cases expressed by indefinite relative clauses (543 ὅς τις ἄν/546 ὃς δέ κε ~ 482 ὅς τις ἄν/486 ὃς δέ κεν), similar metaphors of animals ‘speaking’ (544, 546; ~ the tortoise-lyre at 479, 484, 488), and similar lexis (546 μαψιλόγοισι ~ 488 μάψ; 547 ἐξερεείνειν ~ 483, 487; 549 ἁλίην ~ 488 αὔτως). The metaphorical aspect of this, in particular, shows awareness that Hermes in 484* was aligning the lyre with female respondents including the Pythia. Heiden (2010: 417) added that Apollo describes oracular consultation as a personal rather than an institutionalized interaction, which increases the alignment with Hermes’ dialogic portrayal of lyre-playing.627 Finally, Apollo responds in kind to Hermes’ gambit of describing the bad player in terms which suit Apollo’s previous behaviour, where ignorant and rough questioning produces an ‘ungrounded’ wrong note (463–95*). At first, the promise that Apollo ‘will not deceive’ in the positive scenario (545) recalls his promise to Hermes at 462. But then the negative scenario is phrased as a warning that a vain-speaking οἰωνός could inspire overconfidence and the thirst to know more than the other gods: Hermes’ fart was an οἰωνός (295) after which Apollo said θάρσει (301), eventually leading to Hermes’ hints that he would like a share in the highest form of prophecy (533 |μαντείην ... ἐρεείνεις| ~ 547). However, 549 disentangles Hermes’ behaviour from that of the bad oracle-consultant: his journeys around Greece have not been in vain, and he receives more gifts from Apollo than he offers. While all this is calculated to appeal to Hermes’ sensibility, it is very disconcerting for the external audience – in a way that does not replicate at all the setup at 478–88. The audience’s involvement is clear already in 541 from the future orientation of Apollo’s words, the bluntness (which returns in 549), the harshness of δηλήσομαι, and the possible wordplays present in it. Lines 546–8 slide with casual ruthlessness from human error in misunderstanding an auspice to hubristic offence, off-loading responsibility for the breakdown of χάρις-exchange upon ‘unenviable’ and herd-like worshippers (542, 549). The effect is particularly striking in a genre of hymn which gives prominence to its appeal to χάρις-exchange (579*). Though Apollo avoids specifying that he is talking about Delphi, this was his most prominent oracle at any viable date for Herm., and may be imagined to be the only one at the time of the narrative; in any case it is in our mind from 178, 523*. What we have here is a far less appealing oracular Apollo than the one who founded Delphi in the Hymn to Apollo, with which 541 δηλήσομαι and 542 ἀμεγάρτων may interact specifically; see Davies (1979) 47–8. 627

For the limited evidence for private individual consultations at Delphi see Eidinow (2007) 45–53.

451

C O M M E N TA RY 5 5 0 – 5 5 1

550–66 550–66 Preliminary comment: The following passage is a puzzle – not just because of our partial knowledge of Greek cults, but because Apollo is deliberately continuing to lay claim to Hermetic credentials by showing mastery of linguistic indirectness. My commentary here is grounded in a cognitive approach, and it is worth mentioning at the outset some of the main starting-points. The interpretative journey through a puzzle, from a preliminary (largely hypothetical) model of what is being referred to and why, through various failed ideas towards potential solutions, is integral to the puzzle’s significance as an utterance; indeed, even stating up-front that this is a riddle spoils the effect, though not as much as stating a solution. On our hermeneutic journey through a riddle, we tend to focus on one model (parts of which may be tagged as vague) until some form of incompatibility appears; at that point, a range of alternative replacement models are activated and evaluated in parallel. I shall use the terms ‘referent-hypothesis’ and ‘function-hypoth­esis’ to disentangle some possible ancient models of what our passage is alluding to and why. The treatment of Hermes’ staff (529–32*) has primed the audience to some extent to admit a solution which creatively amalgamates several cultural items at once. As usual, I return at the end to a summary discussion.628 550 ἄλλο δέ τοι ἐρέω: This formula introduces warnings (both hostile and friendly) in the Iliad x8/9 (21.94 being the exception) and at Od. 11.454, 15.27, 16.299, but also vaguer transitions (e.g. Od. 19.236, 495, 570). Its traditional referentiality may therefore raise a tentative function-hypothesis 1, that Apollo is going to warn Hermes of something else. 550–1 Μαίης ἐρικυδέος υἱέ | καὶ Διὸς αἰγιόχοιο, θεῶν ἐριούνιε, δαῖμον: This is an unusually elaborate address, and Hermann thought of deleting 551 altogether. But the echo ἐρικυδέος ... ἐριούνιε (in the same metrical position) enhances Apollo’s praise – a function of his shorter vocatives identified in

628

In constructing an approach here I have tried to combine evidence from cognitive science on a range of topics; I thank Sandra Kotzor for some bibliographical orientation. Traxler (2011: 187–240) gives a helpful survey of priming effects and of the ‘event-indexing model’ of how we process discourse and update our understanding of it. Dutke and von Hecker (2011) discussed the extent to which mental models admit elements of vagueness, Rukavina and Daneman (1996) the existence of individual strategies for dealing with incompatibility depending on one’s broader hermeneutic intuitions. Kounios and Beeman (2009) reviewed the evidence for parallel activation of solutions to verbal puzzles. Coulson, Urbach and Kutas (2006) discussed the processing of punchlines.

452

C O M M E N TA RY 5 5 2 – 5 5 3 533*, 539*. For the sense of ἐριούνιε see 3*. In context here it self-deprecatingly contrasts with the more ambivalent ὄνησις offered by Apollo’s oracles in 541–5. Since θεέ is generally avoided, one could take θεῶν … δαῖμον together as an intensive ‘god of gods’, but that hyperbole reaches a level of absurdity unsuited to Apollo’s rhetoric. It is better to assume that δαῖμον retains a distinctive nuance (cf. 343*), and my preference is to place a comma before it. θεῶν ἐριούνιε then corresponds to the structure of δῖα θεάων (Schwyzer ii.115– 16). In epigraphic poetry, δαῖμον is used particularly in contexts of healing and the Underworld; the latter includes addresses to Hermes the psychopomp in IosPE I2 226 (Olbia, Hellenistic), IG XII.4 2257 (Cos, first century bce). 552–3 σεμναὶ γάρ τινές εἰσι κασίγνηται γεγαυῖαι, | παρθένοι: The description begins in resolutely anthropomorphic terms (κασίγνηται, παρθένοι) which suggest reference-hypothesis 1: that the creatures are one of the small female groups which throng Greek religion, such as the Fates, Graces, Seasons, Muses, numerous groups of nymphs, and some of heroines.629 (Herm.’s general focus on the divine sphere probably favours this over any human group.) A group of unmarried females would constrast with Hermes’ courtesan-lyre (function-hypothesis 2). The only such groups described elsewhere as σεμναί appear to be the Semnai and Erinyes at Athens, and the Graces (e.g. Pi. O. 14.8, fr. 95, E. Hel. 1341), who are not normally virgins. But in any case, other words of the σεβ- stem can be applied to nymphs freely, e.g. A. Eu. 22 σέβω δὲ νύμφας of Corycian nymphs. For Hermes and nymphs see Od. 14.435, H.Aphr. 262–3, Semonides fr. 20, Ar. Th. 977–8, LIMC s.v. Hermes 347–57. Without punctuation, κασίγνηται is part of the subject and εἰσι ... γεγαυῖαι make up the verb (KG i.38–9). One could equally well place a comma after εἰσι, so that κασίγνηται becomes a subject complement, as in A.R. 3.535–6 κασιγνήτη γεγαυῖα,| μήτηρ (perhaps borrowing from our line), Ps.-Plato Epinomis 986a δυνάμεις ... γεγονυίας ἀδελφάς. A particularly suggestive parallel for the former construal is Theodectas TrGF 72 F4 εἰσὶ κασίγνηται διτταί ..., starting a riddle about night and day. While the choice of a kinship metaphor is particularly loaded in an Oedipus play, circumlocutions of the type ‘dust, sister of mud’ (A. Ag. 494–5) had a longer history and raise the possibility of function-hypothesis 3, that we are in a fairly conventional type of riddle.

629

See e.g. Larson (2001) 7, ‘Female pluralities, particularly triads, are ubiquitous in Greek mythology and cult’. Many groups are παρθένοι: see Larson (2001) s.v. ‘virginity’. Where M has σεμναί, Ψ had μοῖραι, which does not match the subsequent description (at Parnassus, nothing to do with Zeus, given to Hermes …). The reading probably started as an ill-judged med­ieval gloss on ‘three reverend sisters’. If so, there is no reason to base emendations on it (e.g. Hermann’s Θριαί), or to cast doubt on σεμναί (e.g. Feyel 1946 invented a group of Σμῆναι).

453

C O M M E N TA RY 5 5 3 – 5 5 4 553 ἀγαλλόμεναι πτερύγεσσιν: Compare Il. 2.462 ἀγαλλόμενα (v.l. -αι) πτερύγεσσιν| (geese, cranes, swans), Mosch. Eur. 59 ὄρνις ἀγαλλόμενος πτερύγων πολυανθέϊ χροιῆι (peacock), Orph.H. 6.2 χρυσέαισιν ἀγαλλόμενον πτερύγεσσιν| (Protogonos). The wings force us to modify reference-­ hypothesis 1. A small modification would lead to the supposition of a set of rapacious winged female deities, such as Sirens, Harpies or Erinyes (see Kenner 1939: 87–8, Verdon (2007) 104–5); recall function-hypothesis 1, that the passage will be a warning. However, the revisualization could also elicit a more radical change of model, for example (reference-hypothesis 2) that the anthropomorphic terms from before were metaphorical, but the wings are literal, so that the ‘maidens’ are birds – sizeable ones, to judge from ἀγαλλόμεναι. As in function-hypothesis 2, the anthropomorphism would match Hermes’ tortoise–lyre–courtesan, and large birds (in augury) had their own connections to Hermes’ lyre in 541–9*. Or, in reference-hypothesis 3, the wings are also metaphorical and we have a set of objects with (say) handles. Either interpretative shift lends support to function-hypothesis 3, that Apollo may be engaging in misdirection and riddling language. 554 τρεῖς: The deities of reference-hypothesis 1 are often triads, whether literally or through iconographical convention. The classic discussion of trios in Greek religion is Usener (1903); for triads of nymphs in reliefs see Larson (2001) 259. κατὰ δὲ κρατὸς πεπαλαγμέναι ἄλφιτα λευκά: This phrase leaves no doubt that we are dealing with a riddle (function-hypothesis 3). Up to πεπαλαγμέναι, whose normal sense is ‘spattered’ (unpleasantly, unintentionally), the phrase reinforces the current form of reference-hypothesis 1, that Apollo is describing rapacious winged goddesses. ἄλφιτα forces re-evaluation both semantically and grammatically. The phrasing turns out to modify formulaic (ἐπὶ δ’) ἄλφιτα λευκὰ παλύνειν+| ‘to sprinkle pale barley (over)’ (Il. 11.640, Od. 10.520, 14.77, variant in Il. 18.560); παλύνω ~ πάλη ‘powder’, so ἄλφιτα probably refers to barley-meal, not groats. Apollo substitutes πεπαλαγμέναι for πεπαλυμμέναι (which is unattested), but changes it from passive to middle (‘having sprinkled … on themselves’, not otherwise attested). The participle introduces mantic wordplay: πεπάλαχθε is found in the sense ‘shake lots for’ at Il. 7.171 (see West 1998 on the confusion here and at Od. 9.331 with πεπαλέσθαι), and divination by barley did exist.630 The Of course alphitomancy is not actually denoted, since it conflicts with κατὰ κρατός. ἀλφιτομάντεις are mentioned in Poll. 7.188, Ael. NA 8.5 (from hearsay), Hsch. α.3337 (feminine); Oenomaus ap. Euseb. Praep. 5.25 envisages consultation of an ἀλευρομαντεῖον by Spartans during the Persian War. Alphitomancy may or may not have been a type of koskinomancy, which may or may not be replicated in the

630

454

C O M M E N TA RY 5 5 4 connection reinforces function-hypothesis 2, that Apollo is making a connection between these creatures and Delphic οἰωνοί (as well as Hermes’ lyre). What of our reference-hypotheses? The middle participle, if taken literally, suggests that the creatures have the ability to sprinkle barley, as in the fully humanoid hypothesis 1. But there is no credible evidence for humans applying powdered barley to themselves; the ghoul Alphito might have done so, but she is mentioned only in Chrysippus fr. 313 and is not a triad.631 Aristophanes’ use of ἀλφιτόχρως κεφαλή (fr. 553) suggests that either hypothesis 1 (goddesses) or 2 (birds) could have been updated to involve white/grey hair or head-plumage, but in either case one will need to elaborate a reason why Apollo expresses the idea with the particular (culinary, ritual?) metaphor of sprinkling barley. With reference-hypothesis 3 (objects with handles), the barley-flour can be taken literally to be covering used kitchen-implements (see n. 631). However, the lemma also seeds reference-hypothesis 4, that the creatures are bees. The key is not that barley-sprinkles bear some resemblance to pollen on a bee’s head (Ilgen’s delightful idea), but rather a practice mentioned in Arist. HA 8(9) 627b18–20: γινώσκουσι δέ τινες τῶν μελιττουργῶν τὰς ἑαυτῶν ἐν τῶι νομῶι ἄλευρα καταπάσαντες. Now παρθένοι matches the bee’s reputation for chastity, and σεμναί their numerous links to reverence and priestesses.632 The implication of size in ἀγαλλόμεναι turns into a diversionary tactic and bathetic joke, albeit one which helped forge the mental connection to οἰωνοί (in function-hypothesis 2). τρεῖς, however, is too small a number for floured bees. The discrepancy suggests a bridge towards the initial form of reference-hypothesis 1: three nymphs, taken to manifest themselves physically as a floured colony of bees (or three). The numerous connections of bees to nymphs help (cf. νύμφη ‘bee-larva’, σειρήν ‘solitary bee’); the dual nature would be comparable to that of dryads and naiads.633 It seems unlikely that all ancient audience-members would readily have identified such a trio.

modern practice of shaking beans, grains and pebbles through a sieve and observing patterns in those which remain (Arnott 1978). For a different technique of divination with barley see Plu. Alex. 26.9. 631 Pye (1810) related the lemma to hair-powder, which he had grown up with. In Hermippus fr. 25 ὥσπερ αἱ κανηφόροι | λευκοῖσιν ἀλφίτοισιν ἐντετριμμένοις, the participle’s case may well be an error of assimilation: the similarities with Ar. Eccl. 732–4 (see Ussher 1973) imply that floured kitchen-implements were being compared to κανηφόροι wearing white lead cosmetics. 632 Chastity: e.g. Semonides fr. 7.83–93, Arist. HA 5 553a18–21. Reverence and priestesses: Frazer (1898) iv.223, Ransome (1937), Davies and Kathirithamby (1986) 69–70. 633 The bee-deity on orientalizing jewellery from Camiros may be Artemis (BM 1860.0404.123–4 with Cook 1895: 11). A gem with a woman-headed bee(?) is less clear

455

C O M M E N TA RY 5 5 5 – 5 5 7 555 οἰκία ναιετάουσιν: οἰκίον can be used of the dwellings of gods, humans and animals, and so does not help narrow down our model of what is being discussed. ὑπὸ πτυχὶ Παρνησοῖο: The phrase occurs elsewhere at H.Ap. 269, where Telphousa advises Apollo to found his oracle in such a spot (see Vergados 56). The location suggests that we try to update reference-hypothesis 1 to a set of divinities local to Parnassus. One might think of nymphs of local springs (listed at Larson 2001: 147–8), or those of the Corycian cave, or the Thriai, who were associated with Athena, Apollo and prophecy by pebbles.634 But since all of these – to our knowledge – lack wings and connections to bees, Hermes and either barley or grey hair, this hypothesis can only offer part of an amalgamated solution. By contrast, both birds and bees (reference-hypotheses 2 and 4) have a strong imaginative presence in the area, not least as the builders of the second temple in one version of Delphic history (Sourvinou-Inwood 1979: 238–44). Inscriptions, papyri and manuscripts all show variation in the spelling of Parnassus, between -σσός (M) and -σός (Ψ). 556–7 μαντείης ἀπάνευθε διδάσκαλοι ἣν ἐπὶ βουσίν | παῖς ἔτ’ ἐὼν μελέτησα· πατὴρ δ’ ἐμὸς οὐκ ἀλέγυνεν: The first words finally show that function-hypothesis 1 (warning) was misdirection, and clarify that Apollo is offering Hermes a substitute form of prophecy. This also makes sense of the hints which contributed to function-hypothesis 2 (significant contrast with tortoise–lyre–courtesan and with Apollo’s οἰωνοί). διδάσκαλοι specifically recalls Hermes’ description of the lyre as a female teacher (484). ἀπάνευθε is both spatially ‘separate’ from the Delphic sanctuary, and didactically ‘separate’ (Carnegie 1908 H9, Ransome 1937: 103). Muses become bees at Philostr. Im. 2.8.6. For the Brisai of Euboea and formerly Keos, who teach bee-keeping, see Arist. fr. 511 Rose, Ps.-Opp. C. 4.275; also Call. Aet. fr. 75.54–7 with Harder (2012), Σ Pi. P. 4.106c. The bees which fed Zeus were interpreted as a heroine by Didymus (Lact. Inst. 1.22). In Orph.H. 51.7 nymphs are βομβήτριαι, though Fayant (2014) prefers to connect this to the hymn’s emphasis on water. Cf. Scheinberg (1979), Larson (2001) 84–7. 634 For the Corycian cult of Pan and the Nymphs see Amandry (1981–4), Larson (1995). Hermes has very little presence there; his epithet κωρυκιῶτα in Orph.H. 28.8 (from long after the Parnassian cave lapsed) may refer to the other Corycian cave in Cilicia. For Thriai see Pherecydes FGrH 3 F49 with Fowler (2013) 81–3, Philochorus FGrH 328 F195 with Jacoby’s comments, Call. Hec. fr. 74.9. According to Philochorus they were a trio of nymphs on Parnassus who had nursed Apollo and invented prophecy by pebbles. In Call. H.Ap. 45 θριαὶ καὶ μάντιες belong to Apollo, whereas in Znb. 5.75 the line πολλοὶ θριοβόλοι, παῦροι δέ τε μάντιες ἄνδρες is explained by a story that pebble-prophets had been Apollo’s upstart rivals. For Apollod. Bibl. 3.115, see below on 550–66*.

456

C O M M E N TA RY 5 5 8 from its chain of communication from Zeus to Apollo to humans. This and 557 πατὴρ δ’ ἐμὸς οὐκ ἀλέγυνεν underline that Apollo’s current offer is compatible with both Hermes’ vow not to approach Delphi (523) and Apollo’s not to share Zeus’s plans (533–40).635 The clause ἣν ἐπὶ βουσίν | παῖς ἔτ’ ἐὼν μελέτησα makes Hermes the rightful successor to this form of divination now that Apollo has passed on his cattle-whip (497), while underlining that he comes behind Apollo in age, herdsmanship and prophecy.636 The aorist tense presents Apollo’s practice of this form of prophecy as a past phase viewed from some distance. The use of the stem μελ- recalls its frequent use when Apollo was finding out about Hermes’ invention of the lyre (437, 453*), but the vicinity of διδάσκαλοι and παῖς also activates the educational usage of μελετάω (compare e.g. Isoc. Panath. 212 παῖδες ὄντες ἐμελέτησαν, Pl. Lg. 1 634b, Galen MM 1.2). This casts the creatures as kourotrophic, a common role among nymphs (above, n. 95) which reinforces reference-hypothesis 4 – particularly since bees have independent connections with prophetic inspiration.637 Of the local deities which seem a partial fit in the current version of reference-hypothesis 1, the Thriai were prophetic and taught Apollo. Whether the Corycian nymphs gain similar reinforcement is unclear, since it is disputed whether the myriad ἀστράγαλοι found in their cave relate to prophecy: Larson (1995: 347– 50) was in favour, Amandry (1984: 376–8) had doubts. For prophecy using ἀστράγαλοι, see Halliday (1913) 206–15. 558 ἐντεῦθεν δἤπειτα ποτώμεναι ἄλλοτε ἄλληι: The line is perhaps influenced, down to its weakened use of ἔπειτα to introduce the next point in a description, by Od. 13.106 ἔνθα δ’ ἔπειτα τιθαίβωσσουσι μέλισσαι. The transmitted ἐπ’ ἄλληι at line-end requires a feminine noun; with Schneidewin’s emendation (1848: 698) compare e.g. Od. 4.236 (> Thgn. 157), Hes. Op. 713, the ends of Solon frr. 13, 15, Xenophanes fr. 26.

For non-spatial ἀπάνευθε see Od. 10.524 ‘sacrifice a sheep separately for Tiresias’. ἀλέγυνεν ‘see to its provision’ (MΘ < Ω: ἀλέγεινεν p by itacism) makes the point that Zeus is not the source better than Hermann’s popular conjecture ἀλέγιζεν. The fact that the prophecy will turn out to be based on eating (559–60) further supports ἀλέγυνεν: cf. 361*, 476*. 636 Why Apollo had taken his herd from Pieria to Parnassus is probably not a question worth probing. 637 Waszink (1974) argued persuasively that the widespread conception that bees converted heaven-sent dew into honey led to their role as mediators of inspiration to prophets (e.g. Pindar O. 6, Iamus) and poets; I take the ‘Delphic bee’ (= Pythia at P. 4.60) to be a passing metaphor grounded in this idea. The swarming of bees is also ominous, e.g. at Hdt. 5.114, Artemidorus 2.21. 635

457

C O M M E N TA RY 5 5 9 – 5 6 0 The creatures’ ‘flitting’ reinforces reference-hypothesis 4 and function-hypothesis 2: these are bees with a significant connection to the οἰωνοί of 543–9, whose ποταί were important (at least in M’s reading at 544). For the verb and bees see Il. 2.90 αἳ μέν τ’ ἔνθα ἅλις πεποτήαται, αἳ δέ τε ἔνθα, A.R. 1.881–2 ἄλλοτ’ ἐπ’ ἄλλον | καρπὸν ἀμέργουσιν πεποτημέναι and more generally Pindar P. 10.53–4 ἄωτος ὕμνων | ἐπ’ ἄλλοτ’ ἄλλον ὥτε μέλισσα θύνει λόγον. By contrast, reference-hypotheses 3 (object, perhaps sieves) and 1 (Thriai or Corycian nymphs) are both suppressed in the face of literal flying. 559 κηρία βόσκονται καί τε κραίνουσιν ἕκαστα: For ‘eating [sc. honey from a] honeycomb’ see e.g. Hipponax fr. 26a.3, Ar. Thesm. 506, X. An. 4.8.20. This decisively supports reference-hypothesis 4 (bee-nymphs), since the Greeks believed that bees and related insects eat honey.638 If the honeycomb functions as food for the bees, it can also be a τραπέζωμα to the nymphs whose presence lies behind them. Indeed, honeycomb offerings are attested in various small cults from the fourth century on, for Helios and Mnemosyne and for the Moirai at Piraeus (IG II2 4962, 4791), Mnemosyne in the Agora (Agora 18.603), Despoina at Lykosura (IPArk. 34, third century), and most pertinently for Hermes and Kourotrophos at the Samian Pythagoreion (IG XII.6 1.260, fourth century). Comatas offers to bring Pan eight bowls of honeycombs at Theoc. 5.59; cf. Zonas 6.7 GP. The choice of βόσκονται assimilates the bees to the cows which Hermes has just acquired (cf. 72 βοσκόμεναι). This is one way to prepare for the bees’ transferral to Hermes. Another is the echo in κηρία ... κραίνουσιν of 530–1 ἀκήριος ... ἐπικραίνουσα, which described Hermes’ other prophecy-related gift, his staff. It is relevant to function-hypothesis 2 that Hermes’ lyre also had the power to κραίνειν at 427 (cf. 531–2*). For καί + generalizing τε, see Ruijgh (1971) 746. 560 αἳ δ’ ὅτε μὲν θυίωσιν: For ὅτε + subjunctive expressing indefiniteness without ἄν, see GH ii.256. M’s θυίωσιν is preferable to Θ’s θυίσωσιν (< Ψ), since the aorist is not otherwise found in Epos. The verb, expressing fast and irregular movement, is comparable to Pindar’s use of θύνω for bees (cited on 558*). It may have particular resonance near Delphi: fourth-century accounts mention a workplace called Thyiai west of the town (e.g. CID ii.46b7 with discussion), Hdt. 7.178 mentions a place called Thyia with an eponymous heroine, and later the Thyiades were Attic bacchants who travelled to Parnassus

E.g. X. Oec. 17.14 drones eat the τροφή deposited by workers; Arist. HA 1 488a16–8, 8(9) 623b17–26 worker-bees are caught eating honey when one removes the comb, 626a2 beekeepers leave out honey for a colony to feed on over winter.

638

458

C O M M E N TA RY 5 6 1 – 5 6 3 (Paus. 10.4.3, cf. 10.6.3). However, the stem generally had a strong relationship not to Hermes or Apollo but to Dionysus (e.g. Alcman PMG 63, A. Th. 498, A.R. 1.636, Thy-one = Semele). ἐδηδυῖαι μέλι χλωρόν: For the formation of ἐδηδυῖαι see GH i.426. χλωρός denotes a pair of qualities of young grass (χλόη) – a pale yellow-to-green colour and moistness. In μέλι χλωρόν| (also at Od. 10.234), as in formulaic χλωρὸν δέος, both qualities are pertinent. 561 ἐθέλουσιν ἀληθείην ἀγορεύειν: Compare M’s reading in 368, ἀληθείην ἀγορεύσω: the phrasing is not known to be formulaic, and Apollo may be adapting his description to echo Hermes’ arrogation there of a prophetic level of νημέρτεια. The verb brings speech into the personification, as for the lyre and bird-omens (544 φωνῆι, 546* μαψιλόγοισι), and therefore re­­ inforces function-hypothesis 2. The parallelism suggests that the bees’ inarticulate sound (as well as their flight) is converted into an articulate message by the skill or otherwise of the interpreter. This mechanism suppresses reference-hypothesis 1 (that the creatures are local Delphic deities), since the Thriai prophesy through pebbles and the Corycian nymphs if at all through ἀστράγαλοι. The lemma might recall the Muses’ famous comment at Hesiod Th. 28, ἴδμεν δ’ εὖτ’ ἐθέλωμεν ἀληθέα γηρύσασθαι. According to one tradition Hesiod’s induction occurred near the river Melisseeis (‘Frequented by bees’; Σ Nic. Th. 11); both Muses and bees offer poetic inspiration (n. 637), Muses can come as a trio, and they haunt Parnassus. Their relevance may again be to enhance function-hypothesis 2, since Hermes presented his lyre as somewhat Muse-like at 484*. 562 ἢν δ’ ἀπονοσφισθῶσι θεῶν ἡδεῖαν ἐδωδήν: This qualification to the initial claim (559 κραίνουσιν ἕκαστα) reintroduces the pattern of felicitous then unfelicitous consultations familiar from 482–8 and 543–9. For ἀπονοσφισθῶσι LSJ’s ‘are deprived of ’ is possible, but νοσφισθείς means ‘having forsaken’ at e.g. Od. 11.73, Archil. fr. 173 and H Dem. 92 νοσφισθεῖσα θεῶν ἀγορήν. With the latter interpretation, as with the bird-omens, consultants need not offend the deities to receive a deceptive response. For honey/mead as the gods’ food – the inherited belief on which their taste for nectar supervened – see the references to τραπεζώματα in 559*, Call. H.Zeus 49 for the particular case of the infant Zeus eating honeycomb, Roscher (1884–90) 282, Ransome (1937) 133–8. 563 πειρῶνται δἤπειτα παρὲξ ὁδὸν ἡγεμονεύειν/ψεύδονται δἤπειτα δι’ ἀλλήλων δονέουσαι: What I have labelled 563(a) (Ω) and 563(b) (xm) share too much to be independent, but too little to rest on copying errors. Like other

459

C O M M E N TA RY 5 6 4 – 5 6 5 doublets in H.Hom., they suggest that our paradosis derives from a collation of two copies which displayed the occasional variations typical of Greek epic performance traditions (Introduction §5.1). In 563(a), παρὲξ ὁδὸν ἡγεμονεύειν attributes to the creatures some of Hermes’ seducing guidance (Kahn 1978: 133), and may recall 547 παρέκ in the corresponding description of the misled auspice-taker. In 563(b), δονέω means as normal ‘set vibrating, drive to and fro, agitate’; the composer envisaged a swarm of more than three. I see no decisive evidence for a sense ‘buzz’ (LSJ, DGE s.vv. δονέω, δόνησις) here or in Choerilus fr. 3, where φῦλ’ ἐδονεῖτο like bees. For the active used intransitively see Antiphanes fr. 183 ἐπεδόνει (of cheese topping), Aretaeus SD 2.1.3. The rarity of this usage might have led to recomposition, i.e. to 563(a), but is insufficient reason to emend the manuscripts’ δενέουσαι differently (δινέουσαι, δύνουσαι, διιοῦσαι?). 564 τάς τοι ἔπειτα δίδωμι: Inferential ἔπειτα introduces the payoff of the preceding description: compare Od. 5.73. The word seems stuck in Apollo’s mind (558, 563(a/b)). σὺ δ’ ἀτρεκέως ἐρεείνων: For ἀτρεκέως questioning see also Hp. Mul. 21, 62. The adverb is more common with verbs of knowing and declaring. With the ancient derivation from τρέχω (Alexander in Et.Gen. α.1374, ‘unsurpassably’) or the modern one from τρέπω (‘untwistingly’), it would integrate the idea of accurate, intelligent questioning (as in 483 versus 487) with the spatialized contrast of paths to knowledge which lead straight or astray/in circles, which we have seen for auspices in 547–9 (a vain journey, παρὲκ νόον ἐξερεείνειν) and for the bees in 563 (παρὲξ ὁδὸν ἡγεμονεύειν/δονέουσαι). 565 σὴν αὐτοῦ φρένα τέρπε: The same phrase occurs in an uncertain context in Mimnermus fr. 7 and a musical one in Ps.-Theoc. 27.14. For possessive + αὐτοῦ see 92–3*. The middle combination φρένα τέρπεσθαι ‘take pleasure in one’s senses’ occurs in several musical settings: see e.g. Il. 1.474, 9.186, B. 16.7; the gist at Ichn. 357 τέρπου φρένα is too unclear to know if Sophocles was inspired by our line (see Introduction §1.3). By presenting this mode of divination in musical language, Apollo supports function-hypothesis 2, that there is a significant connection to the musical question-and-answer of 484. 565–6 καὶ εἰ βροτὸν ἄνδρα δαήσαις | πολλάκι σῆς ὀμφῆς ἐπακούσεται: The end of 565 is transmitted ἄνδρ’ ἀδαῆ (M) or ἄνδρα δαείης (Ψ). But δαείης would mean ‘learn’ (ἐδάην), not ‘teach’ (ἔδαον), and M’s unmetrical reading raises the suspicion that Ω was damaged. My δαήσαις is inspired by Hsch. δ.68 δαῆσαι: διδάξαι. For the optative ending see GH i.465. The counterpart to Hermes’ personal, quasi-aesthetic pleasure is the possibility of helping a human, by continuing the chain of teaching from

460

C O M M E N TA RY 5 6 6 the bee-maidens to Apollo (556 δι-δά-σκαλοι) to Hermes to a worshipper (δα-ήσαις). This teaching would grant Hermes an ὀμφή to match 543/545 ἐμῆς ὀμφῆς (same metrical position; ἐπακούσεται echoes ἀπονήσεται there) and 532 Διὸς ὀμφῆς. A literal interpretation of σῆς and διδάσκαλοι would have the creatures teach Hermes how to interpret their sound and flight, after which Hermes will voice prophecies to his human favourite. But Apollo has shaped his description to match the lyre as teacher, and a more plausible, less literal scenario is that σῆς denotes ownership of the creatures’ prophetic ‘voice’, which Hermes teaches the human to decode independently. Hermes’ teaching seems a more unalloyed force for human benefit than Apollo’s treatment of oracle-visitors. |πολλάκι recalls and overturns the word’s other use (283) in Apollo’s statement of how harmful Hermes will be to humans.639 However, Hermes’ teaching is only phrased as a possibility. Hence an audience-member who has been baffled by 550–63 may legitimately infer that Hermes’ beneficence had not been shared widely. 566 αἴ κε τύχηισιν: Most coherently taken as ‘if he hits the mark [sc. in his question]’. Compare Il. 5.279, 7.243 αἴ κε τύχωμι|