Etymology: An Old Discipline in New Contexts 9788074222634 [PDF]

Zitiervorschau

Etymology: An Old Discipline in New Contexts

etymologie.indb 1

21.1.2014 11:36:37

etymologie.indb 2

21.1.2014 11:36:38

Etymology: An Old Discipline in New Contexts

Edited by Bohumil Vykypěl and Vít Boček

Nakladatelství Lidové noviny Praha 2013

etymologie.indb 3

21.1.2014 11:36:38

The present volume was prepared with the support of a grant from the Czech Science Foundation “Etymological Dictionary of the Old Church Slavonic Language: Summing up a Generation Project” (Nr. 13-17435S). Edited by Bohumil Vykypěl and Vít Boček Reviewed by Radoslav Večerka and Ľubor Králik Studia etymologica Brunensia 16 Eds. Ilona Janyšková & Helena Karlíková © editors ISBN 978-80-7422-263-4

etymologie.indb 4

21.1.2014 11:36:38

Table of contents

Table of contents / 5 By way of introduction / 7

Bohumil Vykypěl (Brno) Grammaticalization and etymology / 9 Vít Boček (Brno) Etymology and language contact studies: Some notes on mutual relationships / 13 Jan Bičovský (Praha) Blind spots of the comparative method / 23 Václav Blažek (Brno) Glottochronology, its principles and results / 41 Ivo Vasiljev (České Budějovice) What can etymology do for a better understanding of the Vietnamese language / 67 Helena Karlíková (Brno) Etymologische Rekonstruktion mit Unterstützung durch die kognitive Linguistik? / 75 Марина М. Валенцова (Москва) Этнолингвистический комментарий к этимологии слов мара и упырь / 85

etymologie.indb 5

21.1.2014 11:36:39

etymologie.indb 6

21.1.2014 11:36:39

By way of introduction

As is well known, the roots of etymology reach deep into the past: much as the content of this linguistic discipline was not always conceived in the same manner, its origins can nevertheless be found as early as Classical Antiquity of Greece and ancient India. However, the formation of etymology as a full-fledged scholarly discipline began much later, namely at the beginning of the 19th century – together with the formation of comparative linguistics. Since then etymology has undergone intensive internal development; the way of its work has changed; a whole series of new theories and methods such as the laryngeal theory or the theory of the Indo-European root, glottochronology or Greenbergian multilateral comparison have been developed many of which must be counted with in the etymological research; preferences of general comparative approaches have changed and so on. At the same time the “linguistic surroundings” of etymology have naturally been developing, too. Most recently, in the last 20–30 years, we could witness intensive research particularly in fields such as grammaticalization and lexicalization, language contact, cognitive linguistics etc. Even this must be taken into account in the concrete etymological research. For example, the grammaticalization and lexicalization research is relevant with respect to the etymological question of native “anomalous” word-formation. The research in the area of language contact, in turn, offers the notion of borrowability as a probabilistic criterion for one of the crucial etymological problems, namely the identification of loanwords, and more generally, modern contact linguistics describes the mechanisms of language contact in a new way, which is especially relevant for the “cultural” aspect of the etymological research. Furthermore, the notion of semantic maps formulated in the frame of cognitive linguistics can serve as an instrument of prediction of the semantic development dealt with intensively by etymologists. Finally, that sort of language typology which works with the notion of language type can explain the degree in which different ways of word-formation are used in individual languages according to the degree in

etymologie.indb 7

21.1.2014 11:36:39

which different language types dominate in these languages. The present volume is not meant to substitute the desirable treatise on the relationship of the venerable old discipline of etymology and the dynamic young trends of present-day linguistics and on the question of what possibilities of utilizing findings of these trends there are for the etymological practice; it primarily wishes to point out simply the need for such a treatise. The papers presented in this volume are then small contributions to some aspects of the sketched problems. Bohumil Vykypěl

etymologie.indb 8

21.1.2014 11:36:39

bohumil vykypěl grammaticalization and etymology

9 Vykypěl

1 Fashionable grammaticalization What would be worth a treatise from the sociology of science is the question of which topics are frequent in present-day linguistics and why they are so. In this, the prehistoric definition of fashionableness in science would surely be useful that was formulated by Vilém Mathesius (1915: 474): fashionable is a work “written without inner interest and undertaken merely because others have dealt with the problem in question and have shown which schema can be used. By this, the topic concerned becomes fashionable, for an original work is imitated in a mass and mechanical way.” One of such fashionable present-day topics is, as it seems, also the so-called grammaticalization. Nevertheless, in spite of their fashionableness in the sense of Mathesius, questions of grammaticalization are, as I believe, indeed interesting and have also some less treated aspects. In what follows I wish to write briefly on one of these aspects. 2 Grammaticalization for etymology: two examples One of the important components of grammaticalization research is, as everyone knows, uncovering of the origin of grammatical elements including grammatical words. This task is common to progressive grammaticalists and conservative etymologists without both “camps” apparently knowing this clearly. On the grammaticalist side, to my knowledge, only Lehmann (2004: 172–3) has implicitly stressed the need for a good etymological reasoning in grammaticalization argumentations. On the etymological side, in turn, not the newest introduction to etymology (Durkin 2009) has considered the research on grammaticalization in a systematic manner. Yet, it is obvious that both sides can enrich each other. It seems at first sight that etymologists are those who can give more, as they offer a large amount of the concrete material for more general considerations of grammaticalists. However, exactly the general, or possibly even universalistic-implicational, frame of present-day research on grammaticalization may offer to the etymologists some instruments for their own research. Two examples can be given. The present paper was written with the support of a grant from the Czech Science Foundation (Nr. 13-17435S).

etymologie.indb 9

21.1.2014 11:36:39

10 Vykypěl

If we limit ourselves to the semantic aspect of etymological work, one instructive point relating to what was said should be mentioned in the theoretical perspective. The leading etymologist Eva Havlová sketched a very suggestive parallel between the practically concrete work of etymologists as the preparation for the formulation of theoretically general Lautgesetze in the 19th century and the present-day work of the same etymologists as the preparation for a future formulation of “Bedeutungsgesetze” (cf. Havlová 2012: 53). In this respect, etymologists can, then, be referred to Regularity in Semantic Change by Elisabeth C. Traugott and Richard B. Dasher, resulting considerably from research on grammaticalization (cf. Traugott–Dasher 2001): much as this work is, in principle, a provisional probe it represents a possible base to be further developed by etymologists. Even more important is the second example as it concerns one very substantial general instrument for the etymologists’ work, namely the semantic parallels. It was also Eva Havlová who, as early as 1965, stressed “the need for a lexicon of semantic changes” (cf. Havlová 2012: 11–2). Although this need exists up to the present day, etymologists have – thanks to grammaticalists and their effort – at least the World Lexicon of Grammaticalization by Bernd Heine and Tania Kuteva (2002) that, in fact, represents a partial lexicon of semantic changes, limited to semantic changes leading into the area of grammar. 3 Etymology for grammaticalization: two examples The World Lexicon of Grammaticalization has without doubt many problems. Nevertheless, it is a usable and useful work which may be amended and complemented in various ways. Two small Slavistic complementations will follow – as, in turn, a contribution of etymologists for grammaticalists. 3.1 As is well-known, in Northern Slavonic languages, imperfective verbs form the future tense with a construction containing the future tense of the verb ‘to be’ and the infinitive; cf. Czech and Russian examples: Czech Budu

pracovat.

Russian Búdu

rabótat’.

be-1pers.fut ‘I will work’

etymologie.indb 10

work-inf

21.1.2014 11:36:40

From the point of view of possible grammaticalization pathways, the auxiliary verb occurring in this construction is of interest. Heine and Kuteva list the Russian form of the Northern Slavonic future construction under the grammaticalization pathway copula > future (Heine–Kuteva 2002: 96). However, this is imprecise, since both in Russian and in other Northern Slavonic languages the auxiliary in question serves as copula only in the future tense. It is more appropriate to categorize this construction under another pathway formulated by Heine and Kuteva, namely change-of-state (‘become’) > future (Heine–Kuteva 2002: 64–5). Interestingly enough, Heine and Kuteva refer in this latter lemma of their lexicon to Dahl (2000), who mentions in passing precisely the Northern Slavonic future form of imperfective verbs as a possible example of the pathway from ‘become’ to the future tense (cf. Dahl 2000: 359–60); the authors seem thus not to have read his paper to the very end. As regards Dahl himself, he is not entirely sure that the assumption is correct that the auxiliary of the Northern Slavonic construction meant originally ‘to become’. In fact, this assumption is wide-spread among etymologists (cf. Berneker 1908–13: 79, Sadnik–Aitzetmüller 1975: 94, Kopečný 1980: 114) and seems to be quite well-founded, since in Old Church Slavonic, Old Czech as well as Old Polish, byti ‘to be’ meant also ‘to become, to come about, to happen’ (cf. Gebauer 1903: 126ff, Urbańczyk 1953–55: 182ff, Kurz 1966: 152ff, Koch 1990: 716). 3.2 Heine and Kuteva register various grammaticalization (or lexicalization) pathways leading from numerals meaning ‘one’. One of them which does not seem to be very wide-spread according to the data offered by the authors is one > other (cf. Heine–Kuteva 2002: 223). Nevertheless, we can add another example: Common Slavonic *inъ ‘other’. Although the details are not entirely clear, most etymologists presuppose that this word developed from the Indo-European numeral *oinos ‘one’. The original meaning ‘one’ was maintained in some Slavonic compounds such as Old Church Slavonic inorogъ ‘unicorn’ and derivatives such as Old Church Slavonic inokъ ‘monk’ (cf. Kopečný 1980: 313ff, ESJS 1989: 244–5 with references). Heine and Kuteva comment on this grammaticalization pathway as follows: “More research is required on the contextual conditions leading to this grammaticalization.” (Heine–Kuteva 2002: 223) As regards the Slavonic example, an interpretation of these conditions has traditionally been offered: etymologists assume that the meaning ‘other’ arose in contrastive contexts when inъ ‘one’ was used in two clauses referring to two possibilities to be chosen, and the second inъ referred thus to the second possibility, that is, to another possibility (cf. Machek 1968: 227, Kopečný 1980: 318, ESJS 1989: 245).

etymologie.indb 11

11 Vykypěl

21.1.2014 11:36:40

12

References

Vykypěl Berneker, E. 1908–13. Slavisches etymologisches Wörterbuch. i. Heidelberg: Carl Winter. Dahl, Ö. 2000. Verbs of becoming as future copulas. In: Tense and Aspect in the Languages of Europe. Ed. by Ö. Dahl. Berlin – New York: Mouton de Gruyter, 351–61. Durkin, P. 2009. The Oxford Guide to Etymology. Oxford: Oxford University Press. ESJS 1989 = Etymologický slovník jazyka staroslověnského. Ed. by E. Havlová et al. Praha: Academia, Brno: Tribun EU 1989ff. Gebauer, J. 1903. Slovník staročeský. i. Praha: Unie. Havlová, E. 2012. O metodách etymologické praxe. In: Methods of Etymological Practice. Ed. by B. Vykypěl and V. Boček. Praha: Nakladatelství Lidové noviny, 9–66. Heine, B. – Kuteva, T. 2002. World Lexicon of Grammaticalization. Cambridge: Cambridge University Press. Kopečný, F. (ed.) 1980. Etymologický slovník slovanských jazyků. Slova gramatická a zájmena. ii. Praha: Academia. Koch, C. 1990. Das morphologische System des altkirchenslavischen Verbums. München: Wilhelm Fink. Kurz, J. (ed.) 1966. Slovník jazyka staroslověnského. i. Praha: Academia. Lehmann, C. 2004. Theory and method in grammaticalization. Zeitschrift für Germanistische Linguistik 32: 152–87. Machek, V. 1968. Etymologický slovník jazyka českého. Praha: Academia. Mathesius, V. 1915. Několik poznámek k mým studiím syntaktickým a jejich kritice ve Vědě České. Časopis pro moderní filologii 4: 473–5. Sadnik, L. – Aitzetmüller, R. 1975. Vergleichendes Wörterbuch der slavischen Sprachen. i. Wiesbaden: Otto Harrassowitz. Traugott, E. C. – Dasher, R. B. 2001. Regularity in Semantic Change. Cambridge: Cambridge University Press. Urbańczyk, S. (ed.) 1953–55. Słownik staropolski. i. Warszawa: PAN.

etymologie.indb 12

21.1.2014 11:36:40

vít boček etymology and language contact studies: some notes on mutual relationships

13 Boček

It is a truism that the two linguistic disciplines mentioned in the title of the present paper, i.e. etymology and language contact studies, have, first and foremost, one topic in common, namely the study of loanwords. In what follows I will offer some general thoughts concerning various, more or less arbitrarily chosen, aspects of this topic in order to show that: first, in one aspect the practice of etymologists can strongly influence and direct the work of language contact researchers (1); second, in another aspect, by contrast, etymology can profit from the theory of language contact studies (2); third, that there is also an area where both disciplines can mutually and constantly cooperate (3). 1 One of the central topics in language contact studies is the relationship between lexical and non-lexical, i.e. phonological and grammatical, external influences. In the – now already classical – model by Sarah G. Thomason (cf. Thomason–Kaufman 1988, Thomason 2001, 2008, 2009, 2010), as well as in the – still rather alternative – model by Frans van Coetsem (cf. Coetsem 1988, 2000), a distinction has been made between two fundamental types of language contact, based precisely on different language material that is primarily transferred from one language to another: in the first type of contact (termed borrowing) it is the lexical, whereas in the second (termed shift or imperfect learning in Thomason’s model and imposition in Coetsem’s model) it is the non-lexical material. It seems that it is a general opinion among language contact researchers that lexical contact influences are much easier to detect than the non-lexical ones. At the first glance such evaluation is quite understandable. I would, however, like to emphasize here that sometimes it has been exaggerated. In recent works by Thomason we can read passages in which this has happened in an unfortunate way, cf.: “[E]stablishing that words have been borrowed is much less likely to present major difficulties for a historical linguist” (Thomason 2008: 49); “I focus on structural interference here because identifying lexical transfer is, by comparison, child’s play” (Thomason 2009: 361); “[G]iven a reasonable amount of luck, loanwords will declare their origin, in which case no further effort is needed” (Thomason 2010: 34). Such statements may perhaps also represent a distilled form of generally little interest in problems of lexical borrowings from the side of language contact The present paper was written with the support of a grant from the Czech Science Foundation (Nr. 13-17435S).

etymologie.indb 13

21.1.2014 11:36:41

14 Boček

etymologie.indb 14

studies. An exception is the topic of loan phonology, i.e. the question of how foreign sounds are adapted into the recipient language (cf. Calabrese–Wetzels 2009). However, the study of loan phonology represents rather a second step in the investigation, which is possible only after a layer of words has been determined as loans. The first step, the determination proper, as a traditional domain of etymology, is apparently not so attractive for contactologists, cf. already Jasanoff (1989: 623): “Simple lexical influence is the commonest way for one language to influence another, but it is also the most superficial and in some respects the least interesting.” In my opinion, such a view of the situation is given by the perspective of a historical linguist who usually only secondarily takes advantage of the work, made by etymologists, for his own interpretations (the most typically, the ascertained existence of the loanwords and their concrete forms are used to establish a more exact dating of a sound change). However, the etymological analysis proper and the final unfolding whether a given lexical element is or is not a loanword are often very difficult and complicated tasks. Etymological dictionaries are full of entries in which different etymologies are given for one and the same word, and very often there are competing explanations as to the foreign or native origin. In many cases the two explanations are very well substantiated (phonologically as well as semantically), and no further corroborative evidence can be added to anyone of them. For example, many Proto-Slavonic words are problematic from the etymological point of view, and so it is very difficult to establish the extent of the Proto-Slavonic contact with other language groups. In this respect, a very instructive case is represented by the problem of Iranian loanwords in Slavonic. There is a tradition of two competing approaches. The first one comes from Meillet (1926) and sees the lexical correspondences in both branches prevalently as the common inheritance from Indo-European, whereas the second, based on Rozwadowski (1915), tends to explain these correspondences rather as a result of secondary Iranian influences on Slavonic (cf. Zaliznjak 1962: 30; Reczek 1987: 13). According to this, some authors speak of almost no Iranian influence on Slavonic at all (cf. Burrow 1955, 22–3), whereas others, on the contrary, assume some tens of items of Iranian origin in the lexicon of Slavonic (cf. Loma 2000). The existence of lexical elements of foreign origin (yes or no) and their number (if yes, how many), is, nevertheless, a very important factor in determining the type of language contact that occurred between the two languages in question. The results of etymological analyses are, therefore, essential for the better understanding of sociolinguistic aspects of the contact because the distinction between borrowing (primarily lexical elements are transferred) and shift/imposition (primarily non-lexical elements are transferred) reflects

21.1.2014 11:36:42

differences in the degree of bilingualism in both ethnic groups in contact, and above all the means by which the elements are transferred: in the case of borrowing the direction is from the non-native (Thomason) or non-dominant (Coetsem) language to the native/dominant one, whereas the opposite holds true in the case of shift/imposition. It seems that the above-mentioned careless look of contactologists on lexical borrowings suits better for contact situations of historically documented languages in which we have better support from the empirical evidence, and so the determination of borrowings is easier. On the contrary, the more past contact situations we are dealing with, the more problematic are the statements about the lexical borrowings. The lesson from etymology to language contact studies can, therefore, be formulated as follows: in language prehistory the difficulties in determining lexical borrowings can be as great as in detecting non-lexical influences.

15 Boček

2 In traditionally oriented etymological works, a relatively simple algorithm is used to distinguish between the native and the borrowed words. If a word can be successfully explained from the system of the language in which it is found, i.e. if the word conforms to word-formation rules of the given language, then it is a native word. If, on the contrary, the opposite holds true, that is, the word is not transparent from the point of view of word-formation, then it is of foreign origin, or – more exactly – the probability increases that it is of foreign origin (cf. Vykypěl 2006: 427–435 with further references). Analogically, in traditional historical linguistics (both the historical phonology and the historical grammar), those phonological and grammatical changes which can successfully be explained from the system of the given language, i.e. for which internal, structural causes and motivations can be found, are distinguished from the phonological and grammatical changes which cannot convincingly be explicated from the system, i.e. for which the internal motivations are missing or not sufficiently clear. In the second case the changes are understood as the results of external influences, i.e. as the phenomena of foreign origin. Thus, in both disciplines, in etymology as well as in historical linguistics, a sharp distinction is made between two kinds of explanations, the internal one and the external one, and the linguistic phenomena are seen as either native or contact-induced. It was the language contact studies that has questioned this dichotomy and come with the concept of multiple causation of language change. According to it, internal and external explanations are not mutually exclusive categories. One and the same language change can be motivated internally and externally

etymologie.indb 15

21.1.2014 11:36:42

16 Boček

etymologie.indb 16

at the same time. Such a view was proposed for instance in the papers of Dorian (1993) or Romaine (1995) and represents also one of the core principles of Thomason’s model (cf. Thomason–Kaufman 1988: 57, 59, 61, 63, 114; Thomason 2001: 62, 94; 2010: 31); for a more recent discussion of the topic, see Aikhenvald (2006: 9–10) or Chamoreau–Léglise (2012: 7–13). It must be added that when the sharp delimitation between internally and externally motivated changes is questioned, it does not mean that the difference is completely abandoned. Of course, in historical linguistics there has appeared and still appear models which do not need to work with this dichotomy at all because they either see all language changes as internally motivated only (cf. Lass 1980), or, on the contrary, insist that every language change has its external causation (cf. Milroy 1997, 1999). A more traditional solution is to choose one of the explanations as primary and the other as secondary. From this perspective, the new position of language contact studies with its concept of multiple causation of language change is virtually an intermediary one. Here, the dichotomy is neither rejected to the benefit of one of its extreme points, nor is it completely obscured by saying that every change has internal as well as external motivations. It may rather be said that the dichotomy is transcended and transformed into a trichotomy. There are now three categories of changes: 1) changes that can successfully be explained from language-internal dispositions only, 2) changes that would not occur without any contact influence of another language, i.e. that have been triggered externally, and 3) changes that can be understood as the results of an interplay of two equipollent factors: internal dispositions and external influences. A question arises as to whether it is possible to transpose this trichotomy, constructed for structural (phonological and grammatical) changes, also to the lexical elements. It would mean that in a language we can find words whose origin is native and foreign at the same time. To my knowledge, only Johanna Laakso (2001) has dealt with this problem. Her remarks are cautious and diligent and may be recommended to the attention of etymologists. She is inclined to accept this trichotomical solution as useful even if it has some heavy consequences. For example, the introduction of the third category of language changes and – in the case of lexicon – of the words without clearly specific needs does not comply with Occam’s razor, which is, otherwise, strongly advocated in Thomason’s model (cf. Thomason–Kaufman 1988: 142; Thomason 1998: 158; 2001: 175). Like in the case of structural changes, in the lexicon too we are faced with the problem of how to recognize which of the three categories an element belongs to. Laakso declares this only as a future and very difficult task: “It is possible that many alleged borrowings simultaneously represent a conti-

21.1.2014 11:36:42

nuity of native linguistic elements or constructions, but detecting them is hard work. No rose garden for historical contact linguistics, but blood, sweat and tears” (Laakso 2001: 209). As for the empirical aspect of the question, Laakso has mentioned only some Finnic examples of words that have several fully convincing etymologies, some of them native, others foreign. Note that in this respect we are very close to the topic presented in the first part of the present paper. At this occasion I would, however, like to mention other phenomena which can, in my opinion, be connected with the question of multiple origin of the words. Again, examples can be put forward from the area of Iranian-Slavonic contacts. The first phenomenon is semantic calques, i.e. words that are formally of native origin, whereas semantically of foreign origin. The most typical example is Proto-Slavonic *bogъ ‘god’ prevalently connected with PIE root *bhag- ‘to divide, share, award’, but simultaneously expected to be semantically influenced by cognate Iranian forms (cf. Avestan baγa-, Old Persian baga ‘lord, god’; cf. Havlová et al. 1990: 71). The second kind of phenomena is in a sense an opposite of the semantic calques: the meaning is of native origin, but the form can be influenced by foreign language cognates. Thus the words are not “true” borrowings; they contain, nevertheless, some foreign formal elements. Gołąb proposed such a solution for thirteen Proto-Slavonic words with initial x-. According to the author, these words were “iranized” and represent “phonemic readjustment to the Iranian system” (Gołąb 1973: 133). True, both changes in semantic calques and in phonemic readjustments are believed to be secondary, so in traditional understanding the words would belong to the category of native words. However, for reconstructed words the exact dating of the processes is not possible. Thus, the phenomena under consideration could possibly be interpreted, for instance, as shared innovations with their centre in the area of the most intense contact of the two languages, where bilingualism is so strong that it cannot be decided whether a relatively numerous group of speakers have a language A or, on the contrary, a language B as their native/dominant language. From this central area, innovations can spread in both directions to “ordinary”, monolingual speakers of the languages A and B.

17 Boček

3 An area where etymology and language contact studies can constantly enrich each other is loanword typology. Extensive research on lexical borrowings, undertaken within the frame of Loanword Typology Project, brought not only a vast bulk of empirical data from 41 languages in the form of the World Loanword Database (for online access, see http://wold.livingsources. org/), but also important generalizations (cf. Haspelmath–Tadmor 2009b, especially the introductory chapters). One of the central questions has been

etymologie.indb 17

21.1.2014 11:36:43

18 Boček

etymologie.indb 18

which categories of words are universally more prone to borrowing and which are, on the contrary, more resistant to it. The findings can be classified according to various criteria, in particular formal ones (differences in parts of speech) and semantic ones (differences in various semantic domains). In the first case, the results are neither surprising (borrowing scale is nouns > adjectives, adverbs > verbs), nor very meaningful, because the motivations for resistance can be very different in individual (types of ) languages (cf. Matras 2012: 650). In the second case, the situation is perhaps more interesting. Even though some of the semantic fields established by the authors are not very well substantiated and loaded (e.g., the category termed possession contains verbs expressing possession, such as “to own” or “to have”, and expressions associated with money, such as “coin” or “beggar”), the majority of them have a clear delimitation and homogeneous content. Among categories that are more prone to borrowing are above all religion and belief, clothing and grooming, the house, and law. The domains with the lowest rates of borrowing are sense and perception, spatial relations, the body, and kinship. From these latter results, more concrete data on resistance have been extracted, and a list of one hundred least borrowable (most stable) words, or, strictly speaking, most stable meanings, has been compiled. This so-called Leipzig-Jakarta 100-item list of basic vocabulary is now considered to be the best among all attempts up to now to establish a universally valid index of basic lexical elements (cf. the most famous Swadesh list used mainly in lexicostatistics). It includes body parts (“mouth”, “ear”, “arm”, etc.), universally present natural phenomena (“water”, “night”, “star”, etc.), generic animals terms (“fish”, “bird”, etc.), generic actions (“to go”, “to come”, “to drink”), basic properties (“big”, “new”), singular pronouns (“I”, “you”) or interrogatives (“where”, “which”), cf. Tadmor (2009: 69–71) for the full list. One of the basic questions is how these findings can be useful for further research in various branches of linguistics. Mostly three areas are usually mentioned: general linguistics, sociolinguistics, and historical linguistics (cf. Haspelmath 2009, Tadmor 2009, Haspelmath–Tadmor 2009a, Matras 2012). Above all, the last is believed to be the most important because the list of words that are the most likely to be inherited provides us with a new tool to measure genetic affiliation. It must be added that this holds true mainly for those who ascribe the decisive role for the determination of genetic relationships to formal correspondences between lexical elements, to a lesser degree for those who take into account formal correspondences in both lexical and grammatical elements, and, finally, to a minimal degree for those who primarily deal with the correspondences in grammatical elements only. For our present thoughts,

21.1.2014 11:36:43

however, another aspect is also important. We can see again that, similarly as in the case of the topic from the first part of the present paper, the usefulness of the results is prevalently perceived in the sense of the possibility of secondary working with the data. Curiously enough, etymology as the closest discipline to historical linguistics is not mentioned as an area that can profit from the generalizations of the Loanword Typology Project. In my opinion, however, the primary usefulness of the results of the project consists in the possibility to take them into account directly in the course of etymological analysis, i.e. by etymologizing concrete words. When two competing etymologies are at hand, the fact that the meaning of the analyzed word belongs to the list of basic vocabulary or, on the contrary, to a domain with high rate of borrowability, can be used as a hint that the word is of the native or, on the contrary, of the foreign origin, respectively. Thus, the borrowability of a particular word (meaning) can be used as an additional criterion in etymological work, being perhaps equal to other secondary criteria, such as, for instance, semantic parallels. I would like to emphasize that the profit can go also in the opposite direction, from etymology to loanword typology. It seems that also in the loanword typology the findings of the etymologists are taken more or less for granted and effortless. Nevertheless, they are not so self-evident. There are hundreds of subtle etymological analyses of individual words behind the compendious chapters by the authors of the Loanword Typology Project and this kind of work still continues in the vast number of minute etymological papers. The new empirical findings of etymologists as well as their partial generalizations can be useful for even more general considerations of loanword typologists. For instance, etymological research can shed some light on the problems such as what are the specific conditions for borrowing of the words which otherwise have very low rate of borrowability, belonging perhaps even to the basic vocabulary list. For example, the word for “dog” ranks the 84th place in the Leipzig-Jakarta list of basic vocabulary. It can, nevertheless, be borrowed as well, like in Russian where the word sobaka “dog” comes from Iranian (cf. Vasmer 1955: 684). A motivation for this is sometimes seen in taboo reasons (cf. Večerka et al. 2006: 184). Thus, one of the traditional topics of etymology, the problem of taboo, can enrich the area of loanword typology. In sum, the cooperation between etymology and loanword typology can be seen as mutually advantageous, the former discipline contributing mostly by the empirical data, the latter by the theoretical principles. In this sense, it parallels the mutual relationship between etymology and historical linguistics.

etymologie.indb 19

19 Boček

21.1.2014 11:36:44

20 Boček

4 The purpose of the present paper was above all to show that the results of the study of loanwords from the perspectives of etymology and language contact studies can significantly enrich both sides. It seems, however, that so far both disciplines were not very straightforward in doing this. Curiously enough, one can have an impression that more advances have been made by etymologists (cf. Bree 2008, Vaan 2008, Durkin 2009: 132–78, Strathmann 2011, Dejkova 2012, and especially Mailhammer 2013). References Aikhenvald, A. Y. 2006. Grammars in Contact: A Cross-Linguistic Perspective. In: Grammars in Contact: A Cross-Linguistic Typology. Ed. by A. Y. Aikhenvald and R. M. W. Dixon. Oxford: Oxford University Press, 1–66. Bree, C. van 2008. Substrate Words. In: Yesterday’s Words: Contemporary, Current and Future Lexicography. Ed. by M. Mooijaart and M. J. van der Wal. Cambridge: Cambridge Scholars Publishing, 138–147. Burrow, T. 1955. The Sanskrit Language. London: Faber and Faber. Calabrese, A. – Wetzels, W. L. (eds.) 2009. Loan Phonology. Amsterdam–Philadelphia: John Benjamins. Chamoreau, C. – Léglise, I. 2012. A multi-model approach to contact-induced language change. In: Dynamics of Contact-Induced Language Change. Ed. by C. Chamoreau and I. Léglise. Berlin–Boston: De Gruyter Mouton, 1–15. Coetsem, F. van 1988. Loan Phonology and the Two Transfer Types in Language Contact. Dordrecht: Foris Publications. Coetsem, F. van 2000. A General and Unified Theory of the Transmission Process in Language Contact. Heidelberg: Universitätsverlag C. Winter. Dejkova, X. 2012. Lingvističeskaja kontaktologija i etimologičeskaja praktika (dopolnenije k „Bolgarskomu etimologičeskomu slovarju“). In: Theory and Empiricism in Slavonic Diachronic Linguistics. Ed. by I. Janyšková and H. Karlíková. Praha: Nakladatelství Lidové noviny, 309–19. Dorian, N. C. 1993. Internally and externally motivated change in language contact settings: doubts about dichotomy. In: Historical Linguistics: Problems and Perspectives. Ed. by Ch. Jones. London: Longman, 131–55. Durkin, P. 2009. The Oxford Guide to Etymology. Oxford: Oxford University Press. Goląb, Z. 1973. The initial x- in Common Slavic. A contribution to prehistorical Slavic-Iranian contacts. In: American Contributions to the Seventh International Congress of Slavists. Warsaw, August 21–27, 1973. Vol. i. Linguistics and Poetics. Ed. by L. Matejka. The Hague: Mouton, 128–56. Haspelmath, M. 2009. Lexical borrowing: Concepts and issues. In: Haspelmath–Tadmor 2009b, 35–54. Haspelmath, M. – Tadmor, U. 2009a. The Loanword Typology project and the World Loanword Database. In: Haspelmath–Tadmor 2009b, 1–34. Haspelmath, M. – Tadmor, U. (eds.) 2009b. Loanwords in the World’s Languages: A comparative handbook. Berlin: De Gruyter Mouton. Havlová, E. et al. 1990. Etymologický slovník jazyka staroslověnského. 2. blagъ – dělo. Praha: Academia. Jasanoff, J. H. 1989. Review of Thomason–Kaufman 1988. Language 65: 623–8. Laakso, J. 2001. Native or borrowed, or both – is it possible to have many mothers? Word 52/2, 197–212.

etymologie.indb 20

21.1.2014 11:36:44

Lass, R. 1980: On Explaining Language Change. Cambridge: Cambridge University Press. Loma, A. 2000. Skythische Lehnwörter im Slavischen. Versuch einer Problemstellung. In: Studia etymologica Brunensia 1. Ed. by I. Janyšková and H. Karlíková. Praha: Nakladatelství Lidové noviny, 333–50. Mailhammer, R. 2013. Towards a framework of contact etymology. In: Lexical and Structural Etymology: Beyond Word Histories. Ed. by R. Mailhammer. Boston–Berlin: De Gruyter Mouton, 9–31. Matras, Y. 2012. Review of Haspelmath–Tadmor 2009b. Language 88: 647–52. Meillet, A. 1926. Le vocabulaire slave et le vocabulaire indo-iranien. Revue des Études Slaves 6: 165–74. Milroy, J. 1997. Internal vs external motivations for linguistic change. Multilingua 16: 311–23. Milroy, J. 1999. Toward a speaker-based account of language change. In: Language Change: Advances in Historical Sociolinguistics. Ed. by E. H. Jahr. Berlin: de Gruyter, 21–36. Reczek, J. 1987. Najstarsze słowiańsko-irańskie stosunki językowe. Kraków: Uniwersytet Jagielloński. Romaine, S. 1995. Internal vs. external factors in socio-historical explanations of change: a fruitless dichotomy? In: Proceedings of the Twenty-First Annual Meeting of the Berkeley Linguistics Society: General Session and Parasession on Historical Issues in Sociolinguistics – Social Issues in Historical Linguistics, February 17–20, 1995. Ed. by J. Ahlers. Berkeley, 478–90. Rozwadowski, J. 1915. Stosunki leksykalne między językami słowiańskimi a irańskimi. Rocznik Orientalistyczny 1: 95–110. Strathmann, P. 2011. Über Wortentlehnung und Lautsubstitution – Versuch einiger Klärungen und Explikationen. Indogermanische Forschungen 116: 4–28. Tadmor, U. 2009. Loanwords in the world’s languages: Findings and results. In: Haspelmath– Tadmor 2009b, 55–75. Thomason, S. G. 1998. On reconstructing past contact situations. In: The Life of Language. Papers in Linguistics in Honor of William Bright. Ed. by J. H. Hill, P. J. Mistry and L. Campbell. Berlin – New York: Mouton de Gruyter, 153–68. Thomason, S. G. 2001. Language Contact: An Introduction. Edinburgh: Edinburgh University Press. Thomason, S. G. 2008. Social and linguistic factors as predictors of contact-induced change. Journal of Language Contact – THEMA 2, 42–55. Thomason, S. G. 2009. Why Universals Versus Contact-Induced Change? In: Vernacular Universals and Language Contacts. Evidence from Varieties of English and Beyond. Ed. by M. Filppula, J. Klemola and H. Paulasto. New York – Abingdon: Routledge, 349–64. Thomason, S. G. 2010. Contact Explanations in Linguistics. In: The Handbook of Language Contact. Ed. by R. Hickey. Malden: Wiley–Blackwell, 31–47. Thomason, S. G. – Kaufman, T. 1988. Language Contact, Creolization, and Genetic Linguistics. Berkeley – Los Angeles – Oxford: University of California Press. Vaan, M. de 2008. On Wanderwörter and Substrate Words in Etymological Research. In: Yesterday’s Words. Contemporary, Current and Future Lexicography. Ed. by M. Mooijaart and M. J. van der Wal. Cambridge: Cambridge Scholars Publishing, 199–207. Vasmer, M. 1955. Russisches etymologisches Wörterbuch. Zweiter Band. L–Ssuda. Heidelberg: Carl Winter. Večerka, R. et al. 2006. K pramenům slov. Uvedení do etymologie. Praha: Nakladatelství Lidové noviny. Vykypěl, B. 2006. Der Sinn der Etymologie oder Die Etymologie als historische Hilfswissenschaft. In: Studia etymologica Brunensia 3. Ed. by I. Janyšková and H. Karlíková. Praha: Nakladatelství Lidové noviny, 427–41. Zaliznjak, A. A. 1962. Problemy slavjano-iranskich jazykovych otnošenij drevnejšego perioda. Voprosy slavjanskogo jazykoznanija 6: 28–45.

etymologie.indb 21

21 Boček

21.1.2014 11:36:45

etymologie.indb 22

21.1.2014 11:36:46

jan bičovský blind spots of the comparative method

23 Bičovský

The nature of sound change and the way it is formalized in historical linguistics both contribute to a number of predictable cognitive biases in the reconstruction and result in “blind spots” that are neither an argument contra the validity of a reconstructed language nor the method itself but have never-the-less influence over the procedure and since they are largely of intuitive character it is vital for the discipline to explore exactly where the blind spots may lie and what explanations and directions they may hide from our perspective in order to be able to reach them in some alternative way. Sound-law has a number of weaknesses that balance its strengths: it is achronological and susceptible to oversimplification. The blind spots that result from these features are the topic of this article.

In the ongoing debate on the adequacy or otherwise of the comparative method (CM) for reconstructing the linguistic past¹ most, if not all, attention has focused on the theory of language change and its aspects crucial for the functioning of the CM: regularity, naturalness, motivation, social factors, among others. This is true of the theoretical works as much as of basic introductions and textbooks. Here is not the place to recapitulate the arguments pro and contra the adequacy of our methods – my aim in this contribution to the etymological enterprise is different. Apart from the limitations of material, the fact that historical chance has a great influence on the degree to which CM can yield any interesting results, and the biases of judgement stemming from the tradition of the respective discipline, individual ontology of the “scientific” mind and such, there is a problem which has to do with reasoning and intuitions in general. The phenomenon under study (sound change) allows for a rather simple representation of (classes of) phonemes and “evolves from” or “evolves into” rules, formalizable into diagrams (tree structures) or formulae analogous to the rules of generative phonology or simple mathematical equations. By their nature as code-signs, and their ability to be analysed as composed of binary features, phonemes in fact lend themselves rather easily to mental operation of the same order as mathematical symbols or chess-pieces. A good etymology is like a nice game of sudoku: if there are three possible candidates for this or that slot, see if there is a 3 or a d in that slot over there, if there is, write 4 or t, if there isn’t, check the slot two steps to the right and one to the top: if there is a 9 or a f, write 6 or þ and there you go: problem solved. Sudoku, of course, 1

etymologie.indb 23

For critical assessments see especially Anttila (1989), Durie & Ross (1996), Lass (1980; 1993), Hoenigswald (1990), Rankin (2003), Harrison (2003), Hale (2007) among many others.

21.1.2014 11:36:46

24 Bičovský

etymologie.indb 24

guarantees that there is a solution and that the data is sufficient to reach the solution. But the mechanism is very similar. Even before the theoretical backing to that concept was first offered by the Neogrammarian school in the late 19th century, sound-law had been the basic tool of the comparative method (both external and internal) whereby the results of systemic and contextual changes in the nature and number of features, elements and contrasts, such as the several possible types of mergers or splits in related languages, are stated and further cognates identified and chance resemblances rejected. I will return to this in more detail below, but let us agree (or disagree, if so you wish) right from the beginning: though sound-laws capture results of historical processes and can be ordered in a relative chronology, by their very nature they are achronological and they are necessarily stored as such in the brain. Also, though they may describe events that may be “manifested” in thousands of lexemes, in order to be useful at all sound-laws have to consist of only a limited and – importantly, in order to by mentally manageable – small number of elements and sub-rules. Once we understand that sound-law is in fact a rule of substitution of one element in structure a for another in a related structure b, it immediately becomes clear what an ideal rule should be, or what should ideally be the simplest and most manageable rule to which all more complex rules can be broken down: A rule of search-and-replace one item for another, without further conditions, in all instances, regardless of the context/structure. That such rules can rarely be postulated at a longer evolutionary distance is self-evident. But this is the condition for the manageability of the rule. As to its storability and retrievability into the working memory, it should ideally be one where the equation is further strengthened by some other property, such as phonetic proximity, or a striking example (an unsuspected relation), and perhaps others. Thus a rule like Proto-Indo-European *l > Iranian r is an ideal companion of a reconstructing linguist, simple, fool-proof, efficient, while Proto-Indo-European *dw- becoming Armenian erk- is more likely to be memorable for the opposite reason: for being so grossly counter-intuitive. One should not wish that relations between languages are all of this type: what a boring world we would be living in! And one should be grateful that the properties of language change in all probability make this impossible. The paths of sound change are not isolated strands of replication – phonological change is largely, though not exclusively, the result of their intersecting. That is why one-to-one equations are rare and bidirectional even rarer: there is always a rivulet breaking off or flowing into the stream somewhere in between.

21.1.2014 11:36:47

To sum up – there are limits on the number of active mental operations any brain can perform at any one time. The complexity of the sound-laws required to be considered simultaneously is a limiting factor on the number. Complex rules can of course be broken down to more simple ones, so one could in principle speak about the number rather than complexity of rules. More complex rules can not only be broken down to simple one-step rules, they can also usually be simplified by ignoring some of the more rarely fulfilled conditions or some less relevant intermediate steps. In other words, only parts of the rules are activated. This is not a problem of the method, or of the rules. It is not even a problem of the brain: it is simply what brains are and how they work. One more requirement we have on any such mental tool is that its application be as broad and universal as possible (this looms in the background of what I see as the main problems in the three case studies discussed below). In terms of Occam’s razor we want as few phonological and grammatical items and rules in the reconstructed stage as necessary to explain the data. Though the physical relations between the systems have always existed, their formulation is a matter of discovery and not only is discovered with only a few, perhaps even two related forms, it clearly cannot be discovered on the basis of all the relevant forms at once. But once it has been discovered and formulated, in its application to newer data, there are but two possible scenarios: either data fits the rule, check, or it doesn’t, explain (sub-rules, language contact, what you will). It is a very logical misconception to consider sound-laws in the first case as explanatory: they are, indeed, but in a very uninformative way. One could answer that the phonemes f and t in a language L are what they are because the proto-L *xw and *d changed regularly that way by a Lautgesetz, if asked about the etymology of a word like fitim, but it is tantamount to saying that birds fly because they crawled out of the ocean (which of course is TRUE under a certain perspective). Yet statements such as “this happened because of the Grassmann’s law” are not uncommon occurrence, though what they say about the historical facts is analogical to saying that “the war in Gaul happened because of Caesar’s De bello Gallico”. They tell us more about why we know something to be true in single instances than why it really happened, in general or otherwise. So there should be an “if ”: if this is a case of the law, then fitim comes from Proto-L *xw…d and is relatable to words of similar structure and meaning in other related languages and in Proto-L. Do we in fact ever answer the question in this way? No. (Ok, rarely.) It is rather the other way round: if it is an f…t, as we know it is, and the form would be relatable to words in other languages through the law in question, it must be a case of the law. Is this

etymologie.indb 25

25 Bičovský

21.1.2014 11:36:47

26 Bičovský

etymologie.indb 26

reasoning flawed? Of course it is. But is this a possible way for intuitive abduction? I believe it is. I will return to this reasoning in the third case study below. Sound-laws stored in our brains are retrieved into the working memory for different purposes, at different occasions and on different levels of conscious and rational thinking. Take for instance that one may entertain the idea that Sanskrit kupya- and English whiff are related, because both mean ‘gorilla’. For a trained specialist, it does not require much conscious effort to realise that on the formal level, the two are at least close enough to make it worthwhile to proceed to the next step, look into their paradigms, earliest attestations, corpora, etymological dictionaries: in short, anything we would consider the proper rigorous approach to proving, as much as anything can be proved in our field, that indeed they have to be related. Is it possible that we have overlooked in this quick scrutiny a detail that immediately shows that these two are NOT related? Perhaps: that remains to be shown in the following steps. But we will hardly consider the effort wasted chasing ghosts: there was enough reason to suspect that indeed further evidence would complement our intuition, not least because it has routinely done so. But are we all that sure that the same kind of blunder, a simple trick of mind, an abduction or a short-cut in intuitive reasoning, is not part of the process whereby the original hypothesis were to be proved or rejected? Would we avoid the wrong turn in the maze if we were not in a hurry? Aye, there’s the rub. I think there are reasons to believe that indeed, such biases are likely to be built into the kind of reasoning that is also used in linguistic reconstruction, both intuitive and conscious, and at least some of the same wrong turns will be taken in both cases or rather alternative corridors ignored. Let us return to whiff: this time assume the word whiff has been encountered for the first time, say a gloss in an obscure Uqbar manuscript. Before checking for possible cognates in a dictionary or writing down the list of all the possible regular reflexes that could exist in all the relevant languages, what we do is browse the memory for the lexemes in the separate lexicons already stored in the brain. How do we do that? Through the eyes of sound-laws: seeing the historical sources behind the attested form with the focus on this or that segment as the occasion and cognates require. The brain probably starts checking great amount of data and coming up with the first likely candidates. Surely not with all of them, surely not always first with the most likely ones. Many simple operations, all of them of the same kind and all therefore susceptible to the same slips of reasoning are performed simultaneously. The idea that what we experience as “finding” or “remembering” is really the act thereof is an illusion. The candidates are first sifted intuitively and unconsciously and we are only served for further

21.1.2014 11:36:47

analysis whatever gets caught in the sieve. (Just to complete the picture, consider a third example: say one were told that English whiff is “clearly cognate” with Sanskrit hukya. A specialist says “no” almost immediately (the wh – h correspondence is wrong) but the why comes only afterwards. Again, one decides before one reasons to defend the decision.) The velocity of the retrieval of possible cognates (and cognates of cognates) that is customary with specialists is immense.² That the brain can be trained to perform miraculously complex operations is without doubt (and that the capacities of some of the great minds in historical linguistics are on a par with chess-masters or nuclear physicists should be obvious once one realises fully the complexity of thinking required to perform these tasks half consciously at great speed). But the rational control and double check that are the crucial components of the scientific method can hardly be employed at this velocity, since say between English wh and Sanskrit h there are many steps. Possible alternatives multiply with each additional segment in the sequence, involving all levels of sound change, systemic as well as contextual. Using the more complex sound-laws (and those that earn a name, such as Bartholomae’s or Winter’s are as a rule complex) as a step-by-step procedure would be too time-consuming. In reality, much more is remembered than generated in this procedure, so that what happens in evaluating the wh – h correspondence is a simple case of compare-and-reject since the brain rather retrieves a relation of English wh to Sanskrit k as a single-step rule and provides a set of reference items like what ~ kad to facilitate and support the retrieval. In this multi-dimensional sudoku, wrong turns and bad decisions are routinely encountered and routinely overcome. How many simple steps of this kind does one make in a single day? How many had to be taken before an etymological dictionary has been produced? Imagine that to solve the relation of kupya- to whiff you may have to bring into the equation Tocharian B käp, Czech čip, Hittite ku-a-pa-aš and Old Irish ceicht or some such delightful company. That is routine. Relations between any two may be problematic; one uncertainty balancing another, experience tends to favour this over that, all that is necessarily part of the business. All of them have paradigms with their own internal problems, three infinitives to feed, a mortgage on the genitive and a dying compound form that needs a walk round the folk-song third-verse now and then. Time is precious. No one has enough of it to take every mental step in every single sound-law: we cannot 2

etymologie.indb 27

27 Bičovský

I am of course prepared to admit that I might be the slowest historical linguist ever – someone has to be.

21.1.2014 11:36:48

28 Bičovský

but rely on the laws and our ability to connect the languages compared. So the question is – can we rely on our intuitive use of the tool enough to trust our results? There has been enough critique in the last decades that would answer in the negative condemning reconstructions as mere illusions and artefacts of the arbitrary workings of the method. But the new findings and deeper understanding of known data surely proves the opposite. In Indo-European linguistics, as anywhere else in science, whenever a hypothesis is proposed, it can either be rejected as not disprovable or put to test. There are, however, hypotheses that even the best routine and the most meticulous analyses cannot prove or reject: those that are not produced and never put to test, explanations that are never attempted and compared with alternatives. It is in our interest to test all the possible hypotheses – but do we take time to think whether we are able to generate all the possible ways the data fit together? There are blind spots that we are likely to encounter time and again in this routine: the roads not taken, the equally valid reformulations of the same facts, the very different conclusions one may draw from the same data considered in a different order. There are a good number of such blind spots I think I have identified, but in this short contribution I only hope to do two things – to draw attention to the problem as such and to give three examples as to how they may relate to the reconstruction of the history of one particular language family: Indo-European. The three blind spots (I AM aware of the Three Blind Mice, yes) all have to do with the time dimension of sound change vs. sound-law. One for all, all for one: Slavic *xIn principle there is no limit on what time-span is captured by a simple formula like *q > *h or *dw > erk-: it may have taken but a few generations, it may have been millennia. Moreover, the starting and ending points are arbitrarily dictated by historical chance: any two subsequent stages in a change like *q > *k > *kj > *c > *tś > *ts > *s > *θ > *h would capture with equal preciseness relations between possible cognates, but it is usually the case that only some (at least two) are attested and the stages attested are of course those that require the sound-laws for their analysis. The sequence in which the data arrives and is processed in the brain constraints the pathways of explanation immediately available for pursuit. What was the sequence for the initial Slavic *x-? First, it had to be noticed as a particular problem against the more general background of the ruki rule as the

etymologie.indb 28

21.1.2014 11:36:48

source of *x: that was already done by Holger Pedersen, in his seminal article on Slavic s (Pedersen 1895). His explanation is still valid: “Das slavische ch ist meiner Ansicht nach aus š entstanden”³ (Pedersen 1895: 74), and so are the conditions (progressive assimilation to a rather uneven set of phonemes). He also identified the phenomenon in Lithuanian (and lack thereof in Latvian) and parallels in Iranian and Armenian, thus establishing it as a feature of the satəm area. Examples are legion, in lexical morphemes as in grammatical (for *rs: *wrso- > *wršo- > *wirs̨a- >> *vьrxъ ‘hill’; for *us: *h₂éusos > *aušos > *aus̨as >> *uxo ‘ear’; for *Ks: (post-Proto-Indo-European) *rēksom > *rēkšom > *rēks̨Vm >> *rěxъ ‘he, she, it said’; for *is: loc. pl. *-oisu > *-oišu > *-ais̨u >> *-ěxъ). As to initial x-, apart from the *xod- etymology (attributing the shift to the effect of a ruki-preverb for which see Bičovský 2013) and obvious loans like *xlěbъ from Germanic, he offered no explanation, concluding that these *x- have to have some other source. But the quest for the origin of *x-, and not anything else, started with Pedersen’s formulation. The problem could be stated as follows: Slavic *x comes from an earlier *š and this *š in turn is the result of the ruki rule. Therefore initial *x cannot come from ruki, unless from *ks, itself not a very frequent Proto-Indo-European initial cluster. But an insidious trap was hidden in this formulation as a sort of implication or reformulation: *x- cannot come from a *š, therefore anything else BUT a *š should be examined as a possible source. There are two approaches: either to seek a unique source for most of the etyma in a phonetically close element (segments or sequences that would already have features such as [velar] or [continuous]) or to find various ways of turning the etymological correlates in other languages to x- somehow. Almost a musketeer “all for one or one for all” dilemma. “One for all” runs into the difficulty of creating a set of Slavic words that may in cases look suspiciously close to semantically close words in other languages without having any genetic relation. Slavic *xoldъ may have come from a *k’h₂ to match Baltic š-, or PIE *ǵH- or *gH- to please Latin gelidus and Germanic *kald-, but why it did not end up rather as a **soldъ 3

etymologie.indb 29

29 Bičovský

Though more than a hundred years later e.g. Voyles and Barrack (2009: 404) write: “/s/ becomes /x/ if immediately preceded by any of the segments /k/, /r/, /u(:)/, /i(:)/ and if followed by a consonant”, only noting without comment that “Shevelov (1964:127) suggests that the historical sequence of this change might have been /s/ > /š/ > /x/” which is in fact what Pedersen wrote. An unlucky statement of the change in a textbook, to be sure, since it would rather help the student to realise under what conditions the change does not take place which is of course more interesting in comparison to other Indo-European branches with similar changes. As for initial *x-, they do not consider any other source but preverbs and loanwords. It is a pity that Collinge’s (1985) treatment of the ruki rule is somewhat inadequate.

21.1.2014 11:36:49

30 Bičovský

or **zoldъ or *goldъ is not clear. The “all for one” approach implies that one can do the same trick of χιποίησις not once, but as many times as many there are the cognate groups that require this change for Slavic. As a “one for all (or most)” approach, a number of scholars agree that the sk- group of cognates is most promising and devise different ways this may have turned into a *x at the cost of losing some promising links and setting up new etyma. A Prakrit-like change to *kh- with *s first becoming a glottal fricative and then reanalysed as pre-aspiration and in turn changed into the more natural post-aspiration has been mentioned. Another account (Rejzek 2010) would have the *sk⁴ change into *xk and later *xx- without any valid typological parallel or phonetic reason. The “one for all” approach within ruki was adopted by Bańkowski (2000), who would have the sequence sk- metathesised to *ks early enough to feed ruki. The strength of that explanation (ruki rule) is overbalanced by the unnaturalness of the metathesis, which is unparalleled in other groups of *sT- in Slavic (and elsewhere) and of course unprovable. In a way, I have to include here my attempt in Bičovský (2008), as a “one for all” solution to which I return later. The “all for one” approach – to try to turn whatever reflex of x- there was in the clearest cognate to a *x- – was taken by Shevelov (1964: 131ff ), who allowed for no fewer than six Proto-Indo-European or Balto-Slavic sources (cognate groups) for the same phoneme in initial position, besides the obvious seventh: loanwords. These are: 1. Indo-European *kc-, 2. Indo-European *ks-, 3. *s- with prefixes, 4. *s- in non-prefixed words, 5. *k-, 6. *sk-. Of these, numbers 1, 4, 5, 6 he takes for affective changes, 2 and 3 as “not necessarily affective” – since these can also be results of the ruki rule. His trick was affective change and his premise that the *s̨ > *x change has already taken place and *x was a phoneme in the language. Affective change is naturally never a very strong explanation, but in the absence of alternatives it may have appeared as the only solution. For cognates that would point to an initial *k’ or *k, the disappearing ghost of a laryngeal was proposed, invisible outside Slavic but turning *k’H into a *kh and later into a *x in Slavic (last and on a more systemic way by Šefčík 2013, who goes to a great length to argue for such a scenario reorganising the whole subsystem of stops in Slavic to explain 3 cases of uncertain etymology). 4

etymologie.indb 30

Note the suspicious rarity of Slavic initial *sk- that would unequivocally point to Proto-Indo-European *sk-. Derksen (2008) has two sk- roots and two šč- ones (with larger, or more inclusive dictionary the numbers grow, but the proportion to other clusters remains). There is no lack of sp- or st-, not it is the case that *sk- clusters would have been scarce in Proto-Indo-European or Baltic. Though not the type of blind spot I discuss here, this is in fact a case of failure to notice a gap. Most Slavic sk- and šč- is of prefixal extraction.

21.1.2014 11:36:49

What is clear enough from these approaches is that reconstructing linguists see this as an either/or question: initial *x- either cannot have anything to do with ruki, or must have everything to do with it. We might have been trying hard to explain a fact that already was explained. Where is then the blind spot? What trick has the achronological nature of the sound-law played? To use the metaphor I introduced earlier, we have completely discounted the possible tributaries. Unlike the Grimm’s law, for which see below, ruki rule can be proved to have rubbed shoulders with other changes, at least with those that define Slavic as different from Baltic, with whom the first stage of the change is shared. The ruki rule is a complicated change (as to context) but most of all it captures (for Proto-Indo-European to Slavic) a very long sequence: *s > *š > *s̨ > *x and possibly as much as 30 centuries. Had it been formulated in separate steps: 1. *s > *š; 2. *š > *s̨; 3. *s̨ > *x, there would have been a chance of noticing that what in 1. is the cause, in 2. and 3. is the context of further changes for reasons of 1., but is no longer the cause of the change. In other words, while it is correct to state that in the vicinity of *r *u *k *i the Proto-Indo-European *s has allophones that coalesce into a later *š, it is misleading and uninformative to say that in the vicinity of Proto-Slavic *r *u *k *i, the *š evolves into *s̨. Could any of the cognate clusters be tributaries to the original ruki rule? Only one: the minority of cases that link it to Sanskrit kṣ, such as *xud- to kṣudrá- ‘poor’. Could they have been tributaries to a post-ruki x? That is what scholars try to find out – and mostly fail, since good tributaries on the phonetic level are missing and those that offer the slightest hope of such a change in Proto-Indo-European, the velars and laryngeals, evolve differently. I argued in Bičovský (2008) that instead of looking for possible sources of Proto-Slavic *x- we should look for sources of Early Proto-Slavic *š- (not from *sj or *xj), though I forced the explanation in one single direction, that is *sk either original Proto-Indo-European, or s-mobile Siebs’ law variants of non-Slavic Indo-European cognates such as *skoldh for *xoldъ, though of course there might have been more than one source for the *š- that I have not thought about, giving the elusive s-mobile too much power over the process. But the question is not whether my account is correct (of course it may not be) but whether the reason why this direction of research has been ignored is not the direct result of the way we pose (or inherit) the question and use the sound-law to answer it.

31 Bičovský

Sound-shift under operation Donald Ringe (2006: 93) remarks about the Grimm’s law: “It is true that no sound change can be shown to have occurred between any of the components

etymologie.indb 31

21.1.2014 11:36:50

32 Bičovský

etymologie.indb 32

of Grimm’s Law”, before he goes into the discussion of the reconstructable stages of the Proto-Germanic consonant shift, e.g. for reasons of impending merger, the changes of *t > *þ and *d > *t must have occurred in this order. Less can be said about the *dh > *d. The first step appears to have been a fricative/ approximant *ð but this change could have taken change even before Grimm (anyway, for the moment, this particular stage of the change is not relevant here). Since the three stop series remain distinct after the change, they must have been distinct throughout the whole change, regardless of the phonetic details: that much we know for certain. But the phonetic details matter, since they have implication for typology, to say the least. By Ringe, as by others, the transition from stops to fricatives is described as a single-step change (unless some crucial details are omitted from the discussion for the sake of clarity: Ringe’s book is after all concerned with English; again, Collinge 1985 is not too explicit about the law, mostly ignoring the question of the intermediate steps and stating the law only through comparison with Sanskrit on p. 63, without any comments on the limiting factors etc. throughout the passage on Grimm). But there are two things that typologist may have to say on that point. First: the change is strange and without good parallels, and more importantly, the system that was created that way is even more of a monstrosity. As for the first: True, voiceless stops have evolved into voiceless fricatives in Goidelic (cf. Old Irish athir ‘father’ from pre-Irish *atīr). But they have not done so unconditionally – they lenited in medial position (such as Old Irish athir < pater). We also do have examples of a shift of a single member of the series (such as *p > *φ in Celtic). But for the whole series to shift systematically, it is most likely that some more general feature was introduced. Voyles (1992: 36) is correct: “The 1st S[ound]S[hift] clearly did not occur in a single step”, though he does not give any reasons. (As was already Fourquet 1948.) The other, more serious problem, is a problem of chronology. Assuming the direct transition of Proto-Germanic voiceless stops to fricatives, there would have to be a stage in Proto-Germanic with no voiceless stops in most of the key positions (in sC they would have to be analysed by the speaker as underlyingly either fricatives or voiced stops as well). That would make Proto-Germanic a strange bird indeed. To my knowledge there is no language that would NOT have voiceless stops (aspirated, glottalized, what you will). But that seems to be what is required of Proto-Germanic. I would argue not that there could have (as Voyles and Fourquet) but that there must have been an intermediate step, which would guarantee that the typological oddity would not happen. On typological grounds again, the best

21.1.2014 11:36:50

candidate is a series of voiceless aspirates: these at least are known to have shifted to fricatives in a number of languages (the prime examples being Greek and Italic, where the Proto-Indo-European voiced aspirates first devoiced and then at different times and with slightly different outcomes turned into fricatives). Also within Indo-European, Armenian shifted unvoiced stops to aspirates, a fact pointed out already by Fourquet, and though in other details Germanic and Armenian may part company, this is still good comparative evidence. That aspiration was also behind the second, Old High German shift is likewise probable. Aspiration as the first step in the change seems to be the most logical conclusion. What would have followed? Judging again by the situation in later Germanic languages, such as Modern German, the voiced series would lose this no longer distinctive feature and the series would next be *ph *th *kh *kwh vs. *b̥ *d̥ *g̥ *g̥w. For the relative chronology the only certain fact concerning the fate of the PIE *bh *dh *gh *gwh in Proto-Germanic is that they did not merge with *b̥ *d̥ *g̥ *g̥w, but become fricatives (or more precisely approximants) in most environments. Again, the motivating factor on this shift would have been the aspiration or some such modulation in Voice Onset Time. I think it is safe to say that such a change could not have taken less than – as an absolute minimum – a hundred years. It changed the shape of Germanic to such a degree, that it seems unlikely within the framework of natural language change to be achievable in a very short time. Now a system like *ph *b̥ *β~*b may have been in existence for as long as there are no direct attestations of Germanic, or at least until the workings of the Verner’s law merged original PIE *p and *bh in pretonic medial position.⁵ For the possible loanwords from and to Proto-Germanic and Proto-Slavic, Grimm’s law is taken as the marking point. If the word is clearly Germanic and yet has the (from the Proto-Indo-European perspective) stops expected for Slavic, it must have been borrowed before the Grimm’s law when the phonology of these elements in both languages was close enough (Slavic would have borrowed voiced aspirates as plain voiced stops which is exactly how they 5

etymologie.indb 33

33 Bičovský

Even the first attestations of Germanic in Latin may not necessarily reflect spirants, but more graeca, aspirates. The Latin rendering of names like Chatti with , which is usually taken as evidence for fricative pronunciation has a disadvantage of explaining why the fricative h was not seen as a natural counterpart to x. Note that the one Germanic tribe with a transparently Indo-European etymology, Teutones (apparently *þeuðaniz or similar in post-Grimm Proto-Germanic), were clearly not *theuthones or *seuzones in Latin and Greek sources in the 1st century BCE. Celtic mediation is suspected but unprovable. Greek mediation would on the other hand support the aspirated stage and lex Grimm unfinished.

21.1.2014 11:36:51

34 Bičovský

would have ended up having come from Proto-Indo-European). After Grimm, a little adjustment has to be done: Slavic did not have voiceless fricatives *f and *þ and as to *x or the sequence *xw, we cannot be sure. It is taken for granted that by the time *χlaibaz was borrowed as (later) Slavic xlěbъ, the *x was in the system. But consider some of the Germanic borrowings in Slavic proposed and the mismatch of stops they display: Germanic *fulką ‘folk, army’ ~ Slavic *pъlkъ Germanic *fetilą ‘fetter, latch’ ~ Slavic *petъlja Germanic *flōkaną ‘weep’ ~ Slavic *plakati (if genuine) It is not at this point all-important whether they have or have not been borrowed, before or after Grimm (*plakati looks suspicious to most). But in case they are taken as possible borrowings, these loanwords must be post-Grimm (there is no inner Slavic change that would turn the Pre-Grimm *d into *t). Pre-Grimm *p…g and *p…d would have been borrowed as such into Slavic. For the post-Grimm era, only a slight adjustment has to be done: Germanic *f must be accommodated into the system as Slavic *p. As much as this adjustment is natural and expectable, it is by no means the only alternative and is by no means necessary. As a comparandum, note the following loans from Latin or (through) Old High German, into Czech: firmare ⇒ firmen ⇒ biřmovat ‘to confirm’, fasianus ⇒ Old High German fasant ⇒ bažant ‘pheasant’, Old High German farwa ⇒ barva ‘colour’ with f borrowed as b rather than p. Still, *f > *p is a good explanation. But it may not be necessary. The arguments are more of a historical kind: intensive contact between Slavs and Germanic tribes are connected with the Germanic migrations in the few centuries before and after the beginning of our era. At the time of the migrations, Grimm’s and Verner’s law should have been finished. So that Chatti and Cherusci should have already referred to their war contingents as *folkō. But I am not going to argue that the form which has been loaned into Slavic must have been pre-Grimm. My point is that in similar cases, one should not overlook the possibility that the word was borrowed during the change since it may provide a smoother transition of the loan, therefore fewer steps in the explanation, i.e. a better explanation. Consider the following, equally valid equations: Germanic *phulg̥ ą ‘folk, army’ ~ Slavic *pъlkъ Germanic *phed̥ilą ‘fetter, latch’ ~ Slavic *petъlja Germanic *phlōg̥aną ‘weep’ ~ Slavic *plakati

etymologie.indb 34

21.1.2014 11:36:51

We have already seen that the first stage in Grimm’s law lead to an opposition similar to that of modern German: an aspirated tense stop vs. a devoiced lenis stop. Again, Czech (and other Slavic languages) with the inherited Proto-Slavic opposition intact (presumably) borrowed words like German Bändl ‘ribbon’ as pentle (b…d as p…t) or Middle High German behhari ⇒ pohár ‘beaker’ (examples are legion). The important point is that the speakers need not even recognise that there is any phonetic distance between the donor language and their own in this particular part of the system. They are not trying to find the best match for the sound that is presumably difficult to pronounce (as is the case with English θ for most other European languages).

35 Bičovský

Hiding a murder in a train disaster: Grimm’s law and its exceptions “PIE voiceless stops became PGmc. fricatives, provided that they were not immediately preceded by another obstruent (usually *s, but sometimes another stop).” (Ringe 2006: 95) “The stops *p *t *k remain unchanged after s and f, t, χ.” (Beekes 2011: 130–1) “PIE voiceless stops *p, *t, *k’ (+*k),*kw became the voiceless fricatives *f, *þ, *x (> *h), *xw (> *hw) when not preceded by an obstruent. The voiceless stops, however, remained unchanged after *s (cf. Go. steigan ‘climb’ beside Gk. στέιχω (stéikhō) ‘id.’) or when preceded by another stop (cf. Go. -hafts “having, having taken” beside Lat. captus ‘taken’).” (Jasanoff 2008: 196, 197). “Grimm I did not apply if the consonant was preceded by s … Additionally, if PIE had two voiceless stops in a row, only the first one underwent Grimm’s law⁶…” (Fortson 2010: 341) While the lesson to be learnt from the case of the initial *x- from ruki is not to ignore an explanation already available, the exceptions to Grimm’s law are quite the opposite and the metaphor of the train disaster is fitting. To suspect 6

etymologie.indb 35

All the underlying mine. – The conditions are stated in this fashion also in e.g. Meier-Brügger (2010: 286), Hock (1991: 38).

21.1.2014 11:36:52

36 Bičovský

etymologie.indb 36

a murderer in one of a hundred dead in a train accident requires a good reason to suspect the murder and perhaps a Poirot to prove it. Unlike train disasters, which hopefully apart from the victims of vis maior do not necessarily have murder(er)s involved, language change should always be suspected of such revolting behaviour. As I wrote above, the way we formulate a question and its answer are potentially misleading. It is usually the case when a sound-law is applied to a new item that the question is stated in the following manner (for the English word foot): “if it is a Germanic word, as we know it is by comparison with other Germanic languages, it can be reconstructed into Proto-Germanic as *fōtuz and if the f…t can be related to other Indo-European forms through the Grimm’s law, as it can, it is a case of the Grimm’s law.” Regularly, the change does not take place in clusters of a certain shape: the *sC and *Cs clusters with Proto-Indo-European unvoiced stops and CC clusters (there are only voiceless ones it appears). The well-known examples are Proto-Indo-European *steh₂ failing to yield *sþandan ‘stand’ or similar in Germanic, and Proto-Indo-European (H)ok’toh₁ ‘eight’ surfacing as *aχta in Proto-Germanic, not **aχþa, but of course there are dozens if not hundreds of others. We then have a general rule and a limitation on its applicability in two environments. Put into words, they read: after a spirant, voiceless stops do not evolve into fricatives, and in a sequence of two stops, only the first one undergoes Grimm’s law (thus Ringe, Fortson, and Jasanoff above, among others). The preceding fricative or stop block the tendency. A teleological explanation for both these environments is that the sequence that would have resulted (but never did) would consist of two fricatives and Proto-Germanic would not tolerate this type of clusters (for which further evidence is provided by the fact that there were no such sequences at that time, though see below…). But wait a minute: have we not seen that the sequence really was rather *k > *kh > *kx > *χ than simple *k > *χ? What then was the effect of the fricative? Which steps were blocked, which reversed? If, as is typical for e.g. English, the aspiration did not surface in the stop, the constraint on such sequences in PG would be a perfectly acceptable condition, but a different condition from “no FFs”. That being established, what shall we conclude about the CC clusters: did they really evolve into FF (fricative+fricative) through aspirates and then – since they violated a constraint – the second reversed to its original condition (instead of the first)? Voyles (ibid.) sees that as a possible sequence: “a possible development there might have been *nokt (ev.) > *nokhth [1st SS, stage 2] > *noxθ [ev.] > *naxt”, but then he gives no hint as to what he expects to happen in *sC clusters.

21.1.2014 11:36:53

We know that at least for the fate of the original *ks we have evidence for a shift to *χs, that is two fricatives, e.g. *h₂eǵsis ⇒ *aχsō “axis” without any reversal or dissimilation whatsoever (at least at the early post-Proto-Germanic stage). So that “no *FF” was not such a high ranking constraint to start with. Never the less, we have no other choice but to conclude that indeed, in CC clusters, there must have been an inter-stage of double fricatives resolved later to fricative-stop if we want to keep the constraint. If first Grimm’s law turns the CC into FC only, with blocking in the sequence, then we need another round of Grimm’s law to turn it to FF and then apply the constraint (that is the circularity in Beekes’s account: if Grimm does not operate after *f *þ *χ, we apparently need a pre-Grimm *f *þ *χ – which I am all for – but that is not what he has in mind or the change would have to happen twice). Unless, of course, we do not expect the change to be a direct one-step stop > fricative in the first member of the cluster, and by this point I think we can’t. In which case it can’t be Grimm’s law and has to be EARLIER and independent. And fortunately, such changes are so frequent cross-linguistically that to invoke such a process for Proto-Germanic is unproblematic and banal. A dissimilation of two stops into a FC is documented both historically (in a number of Romance dialects) and can be observed synchronically as a variation (as in Modern Greek). But the fact that Grimm’s law may have had no effect on such clusters whatsoever may also allow us to conclude, that the change may have been LATER than Grimm’s law. It would simply survive Grimm’s law unaffected and later dissimilate. As for the aspiration-to-affrication theory, it predicts that a sequence of Proto-Germanic voiceless stops should have yielded *ChCh in the first step (so also Voyles loc.cit.). As a rule, aspiration is phonetically realised only on the second member, thus CCh. Again, one would expect the constraint to work in the opposite direction: kth > kþ not xt! So apparently the same conditions that are employed to explain the two types of exceptions predict each other out. Summing up, if it walks like a Grimm, talks like a Grimm, even if it is wearing Grimm’s glasses, it can still be George Sand.

37 Bičovský

Conclusion I have tried in these few pages to exemplify some of the ways in which mental manipulation with linguistic formulae, such as sound-laws, is susceptible to logical fallacies, that these fallacies are likely to occur under the same circumstances and that even in retrospect one may wonder whether the potential for alternative explanation has been exhausted even in the most familiar and basic sound-laws and rules that have been part of the discipline for decades

etymologie.indb 37

21.1.2014 11:36:53

38 Bičovský

and repeatedly proved their usefulness. Perhaps there are or will be found completely novel ways of looking at sound change or of modelling the process, but it would be a shame if it turned out that some of the new interpretations could have been arrived at with the present state of knowledge and methodology. No explanation is complete if all the possible (and reasonably probable) alternatives have not been considered. The sound-law, being our eye on the sound change and sound relations, has its inbuilt blind spots. There is slim chance that those can be removed permanently, since they have to do with the structure of thinking, but we should be aware of what are likely to be the situations where some crucial evidence or logical possibilities are blocked from our view and never available for further analysis and see past them. References Anttila, Raimo 1989. Historical and Comparative Linguistics. Amsterdam – Philadelphia: John Benjamins. (Current Issues in Linguistic Theory, 6.) Bańkowski, Andrzej 2000. Some remarks on the origin of the Slavic velar spirant. In: Studia Indogermanica Lodziensia 3: 75–8. Beekes, Robert S. P. 2011. Comparative Indo-European Linguistics: An introduction. Second edition, revised and corrected by Michiel de Vaan. Amsterdam – Philadelphia: John Benjamins. Bičovský, Jan 2008. Initial x- in Slavic revisited. Chatreššar 2008, 23–45. Bičovský, Jan 2013. A note on the Slavic suppletion series *xoditi ~ *šьdlъ ~ *jьdǫ “to go”. Linguistica Brunensia 61: 129–34. Collinge, N. E. 1985. The Laws of Indo-European. Amsterdam – Philadelphia: John Benjamins. (Current Issues in Linguistic Theory, 35.) Derksen, Rick 2008. Etymological Dictionary of the Slavic Inherited Lexicon. Leiden – Boston: Brill. (Leiden Indo-European Etymological Dictionary Series, 4.) Durie, Mark & Ross, Malcolm (eds.) 1996. The Comparative Method Reviewed: Regularity and irregularity in language change. New York – Oxford: Oxford University Press. Fortson, Benjamin W. 2010. Indo-European Language and Culture: An introduction. Second edition. Chichester: Wiley-Blackwell. (Blackwell Textbooks in Linguistics, 19.) Fourquet, J. 1948. Les mutations consonantiques du germanique. Essai de position des problèmes. Paris: Les Belles Lettres. (Publications de la Faculté des Lettres de l’Université de Strasbourg, 111.) Hale, Mark 2007. Historical Linguistics: Theory and method. Malden etc.: Blackwell. (Blackwell Textbooks in Linguistics, 21.) Harrison, S. P. 2003. On the limits of comparative method. In: The Handbook of Historical Linguistics. Ed. by Brian D. Joseph & Richard D. Janda. Malden etc.: Blackwell, 213–43. Hock, Hans Heinrich 1991. Principles of Historical Linguistics. Second revised and updated edition. Berlin – New York: Mouton de Gruyter. Hoenigswald, Henry M. 1990. Is the “comparative” method general or family-specific? In: Linguistic Change and Reconstruction Methodology. Ed. by Philip Baldi. Berlin – New York: Mouton de Gruyter, 375–83. (Trends in Linguistics, Studies and Monographs, 45.) Jasanoff, J. J. 2008. Gothic. In: The Ancient languages of Europe. Ed. by Roger D. Woodard. Cambridge: Cambridge University Press, 189–214. Lass, Roger 1980. On Explaining Language Change. Cambridge: Cambridge University Press.

etymologie.indb 38

21.1.2014 11:36:54

Lass, Roger 1993. How real(ist) are reconstructions? In: Historical Linguistics: Problems and perspectives. Ed. by Charles Jones. London: Longman, 156–89. Meier-Brügger, Michael 2010. Indogermanische Sprachwissenschaft. 9. Aufl. Berlin – New York: Walter de Gruyter. Pedersen, Holger 1895. Das indogermanische s im Slavischen. Indogermanische Forschungen 5: 33–87. Rankin, Robert L. 2003. The comparative method. In: The Handbook of Historical Linguistics. Ed. by Brian D. Joseph & Richard D. Janda. Malden etc.: Blackwell, 183–212. Rejzek, Jiří 2008. The Proto-Slavic Word-Initial x-. Praha: Karolinum. (Acta Universitatis Carolinae, Philologica, Monographia, 150.) Ringe, Donald A. 2006. A Linguistic History of English, Vol. i: From Proto-Indo-European to Proto-Germanic. Oxford: Oxford University Press. Šefčík, Ondřej 2013. On the origins of x in Slavic. Linguistica Brunensia 61: 121–8. Shevelov, George Y. 1964. A Prehistory of Slavic: The historical phonology of Common Slavic. Heidelberg: Carl Winter. Voyles, Joseph 1992. Early Germanic Grammar: Pre-, Proto-, and Post-Germanic Languages. San Diego etc.: Academic Press, Inc. & Harcourt Brace Jovanovich, Publishers. Voyles, Joseph & Barrack, Charles 2009. An Introduction to Proto-Indo-European and the Early Indo-European Languages. Bloomington: Slavica Publishers.

etymologie.indb 39

39 Bičovský

21.1.2014 11:36:54

etymologie.indb 40

21.1.2014 11:36:55

václav blažek glottochronology, its principles and results

41 Blažek

‘Classical’ and ‘recalibrated’ glottochronology 1 The method called glottochronology represents an attempt to date the divergence of related languages in absolute chronology. Its author, Morris Swadesh, was inspired by the so called radiocarbon method, used for dating organic remnants. Here we can reiterate the main steps in the deduction of the method. In the first place there was the discovery of the radiocarbon isotope C¹⁴, existing in the atmosphere in the proportion 1 : 10¹² with the usual isotope C¹². Due to the food-chain this radioactive isotope occurs first in green plants and subsequently in biological tissues of animals. After the death of any living organism the disintegration of the radioactive isotope follows according to an exponential function. This exponential disintegration means that after a constant time period T (= half-time of disintegration) the concentration of the radioactive isotope is reduced by half, after 2T by a quarter, etc. On the basis of this phenomenon, W. F. Libby developed the radiocarbon method (1947), serving to determine the age of organic remnants younger than 50 millennia. The method was recently defined with more precision (e.g. the change of the half-time from 5568 to 5730 years; correlation with dendrochronology, etc.), but its basic idea remains. Since M. Swadesh borrowed the mathematical apparatus from Libby, it is useful to repeat it here. (1) ΔN(t) = -λ·N(t) · Δt

(2) dN(t) = -λ·N(t) · dt ∫ dN(t) = - ∫ λ · dt N(t)

... decrease ΔN from N radioactive nuclei in the time interval Δt, where λ is a constant of proportion ... approximation of discrete quantities by connected ones, allowing the integration ... leading to the solution

ln N(t) = -λ · t + C. After delogarithmization we reach N(t) = e-λ t ⁺ C = e-λ t · eC, where eC = K. So we can write N(t) = K· e-λ t. It remains to determine the function of the constant K. It is possible thanks to the initial conditions, i.e. in the time t = 0, when N(t) = N₀: The present study was prepared thanks to the grant of the Czech Science Foundation (GAČR), P406/12/0655. First presentation of results of the glottochronological study comparing old Germanic languages was realized together with Zuzana Malášková at the conference GeSuS held at Freiburg University in May 2013.

etymologie.indb 41

21.1.2014 11:36:55

42 Blažek

(3) N(t) = N₀ · e-λt, where N₀ represents the number of undisintegrated nuclei at the beginning of the process. From the equation (3), which is a standard solution of the differential equation (2), we deduce the significance of the half-time of disintegration T, defined as the time interval, in which the number of the undisintegrated nuclei decrease in 1/2: (4) N(T) = –21 N₀ –21 N₀ = N₀ · e-λT, after a reduction –21 = e-λT , after logarithmization ln –21 = -λT, i.e. ln 2 = λT, or (5) T = ln 2 λ The half-time of disintegration of the radioactive isotope C¹⁴ was empirically established as 5730 years. This allows one to determine the value of the constant of disintegration λ. For practical calculations it is helpful to use the formula derived from the definition of the half-time of disintegration. If the number of the undisintegrated nuclei decreases by 1/2 after every time period T, we get: 1 (6) N(t) = N₀ · (–)n 2 , where n means, how many periods T correspond with the age of the specimen. Hence N(t) = (–)n 1 , i.e. N₀ = 2n. Let us logarithmize it: 2 N₀ N(t) N₀ ln = ln 2n = n · ln 2 and we reach N(t) N₀ (7) n = N(t) ln 2 ln

From here we get the age of the specimen (8) t = n · T . 2 Around 1950 Libby’s radiocarbon method inspired Morris Swadesh, an American anthropologist and specialist in native American languages, to extend its application to the development of languages. His goal was the absolute dating of the time of divergence of related languages. Swadesh thought that the replacement of words in languages was determined by an exponential rule

etymologie.indb 42

21.1.2014 11:36:56

similar to the rate of disintegration of radioactive nuclei of isotope C¹⁴. To calculate the rate of this change he established a testing word-list, consisting first of 215, later of 200 semantic units, which had to be universal and immune from borrowing. Thanks to the cooperation of specialists in Sinology, Egyptology, classical philology, Romance and Germanic linguistics, he was able to determine the average constant of disintegration applied to one millennium, in 19.5% changes in the testing word-list, i.e. in the development of a given language, on average, 80.5% of the units of the basic lexicon should be preserved during this period (see Swadesh 1952). Naturally, this is only true if the constant actually is universal. In 1955 Swadesh published a new study, reflecting the first critical reactions (Swadesh 1955). He radically reduced and changed the testing word-list, now consisting of 100 semantic units. On the basis of the reduced ‘basic lexicon’, the constant of disintegration was changed to 14% per millennium, i.e. 86% of the lexical units should be preserved in the development of one language after one millennium. The elementary postulates may be formulated as follows: [1] In the lexicon of every natural language it is possible to determine the part which is more stable than others. This is commonly called the basic lexicon. [2] It is possible to define the set of meanings expressed in every language by words from the basic lexicon. This can be designated the basic testing list (BTL). The symbol N₀ will signify the number of distinct meanings included in the list. [3] The share r of the words from the basic testing list preserved after the constant period Δt, is constant; i.e. it depends only on the length of the time interval, not on a specific language or a choice of words. [4] All words representing the basic testing list have equal chances of being preserved during the same time interval. [5] The probability of any unit from the basic testing list being preserved does not depend on the probability of the corresponding unit being preserved in the basic testing list of any other language. To calculate the time elapsed between the existence of two languages A and B, where B is a descendant of A, Swadesh used the mathematical apparatus from the radiocarbon method. He began with equation (3): (9) N(t) = N₀ · e-λ t, where λ represents the analogy to the constant of disintegration in equation (3). This is defined exactly as the share of the words in the basic testing list which are replaced during one millennium. Hence:

43 Blažek

(10) N(t) = e-λ t, or ln N(t) = -λt. From here N₀ N₀

etymologie.indb 43

21.1.2014 11:36:56

44 Blažek

N(t) N₀ or ln ln c , where c = N(t) -λ -λ N₀

ln (11) t =

If the share r from the postulate (3) is also related to the period of one millennium, it will represent the constant which is complementary to λ, i.e. (12) r = 1 - λ. For the decrease of the words from BTS per millennium the equation ΔN = N₀ - N(t₁) = N₀ - N₀ ∙ e-λ · ¹ = N₀ (1 - e-λ) is valid. The same value must be reflected in the product N₀ · λ. From the comparison 1 - e-λ = λ = 1 - r (see 11) we reach (13) r = e-λ. The same result is accessible from the comparison of the right sides of the equations expressing the shares of the preserved words in the BTL per millennium: N = N₀ ∙ e-λ · ¹ & N = N₀ ∙ r. Consequently it is possible to rewrite the equation (10) by means of (13) in the form (14) c = rt , where t indicates the time in millennia. Regarding the postulate (5) the share c₂ of the preserved lexicon from the BTL in two related languages, i.e. the languages that developed from a common protolanguage, is equal to the square of the share of the words preserved in the individual development: (15) c₂ = (rt)² = r²t. Logarithmizing it, we express t: ln c₂ = ln r²t = 2t ln r. From here (16) t = ln c₂ or with respect to the equation (13) 2 ln r (17) t = ln c₂ , -2λ where c₂ means the share of commonly inherited pairs of the words in BTL in both of the languages analyzed. In the application of glottochronology the formulae (16) or (17) are used most frequently. For an illustration of the practical procedure let us estimate

etymologie.indb 44

21.1.2014 11:36:57

the time of divergence of German and French. In the BTL of both languages there are 33 pairs of commonly inherited words. Both lists are complete, which means that c₂ = 0.33. Applying this to the equations (16) or (17), we reach the time of divergence in millennia:

45 Blažek

t = ln 0,33 = -1,10866 = 3.675 2 ln 0,86 -0,30164 It is more advantageous to calculate a rich set of data with corresponding share of preservation of BTL for one language (c₁) or for two related languages (c₂) – see Table 1: c₁ 0.99 0.97 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 c₂ 0.97 0.94 0.90 0.81 0.72 0.64 0.56 0.49 0.42 0.36 0.30 0.25 0.20 0.16 0.12 0.09 0.06 0.04 0.02 0.01 t 0.03 0.20 0.35 0.70 1.10 1.50 1.90 2.40 2.90 3.40 4.00 4.60 5.30 6.10 7.00 8.00 9.30 10.7 13.0 15.3

The time of divergence for German and French is shown on line t, corresponding with c₂ = 0.33. This value may be located approximately between 3.40 and 4.00 millennia in table 1. Thus it is concretely possible to estimate the age of the common ancestor of German and French as 3700 BP or 1700 BC according to the methodology developed by Swadesh. The preceding steps are operative only with a pair of synchronous (contemporaneous) languages. It may also be necessary to estimate the divergence of nonsynchronous languages (i.e., if each of the compared languages was recorded at a different time). Let us designate as t₁ and t₂ the respective time spans from the disintegration of the common ancestor of the compared languages to the written records of each language. In this case the equation (16) can be modified as 2t = ln c₂, and further ln r (18) t₁ + t₂ = ln c . ln r Since t₁ and t₂ are usually unknown, and only their subtraction Δt₁₂ is at our disposal, it is possible to substitute the sum t₁ + t₂ with t₁ + t₁ + Δt₁₂ = 2t₁ + Δt₁₂, where t₁ is shorter from both intervals t₁, t₂. Hence for two nonsynchronically attested languages the final formula appears as follows: (19) t₁ = ln c - Δt₁₂ , where t₁ = min (t₁, t₂). 2 ln r 2

etymologie.indb 45

21.1.2014 11:36:58

46 Blažek

3 Swadesh’s glottochronology was welcomed by specialists studying languages without a lengthy literary history. On the other hand, the sharpest negative reaction was from specialists in the Indo-European languages. This was understandable, since some glottochronological estimates of the time-depth of Indo-European languages strongly disagreed with well-known historical facts. More interesting than aprioristic rejection was criticism of the specific premises, postulates, and conclusions, especially if the critics offered alternative solutions. The most remarkable modifications eliminating some of the weak points of the method were formulated by the Canadian Sheila Embleton (1986) and the Russian Sergei Starostin (1989, 1999). Both scholars agreed that the ‘classical glottochronology’ of Swadesh was mistaken in that the replacement of words was not distinguished from borrowing. For example, one such innovation was Russian glaz ‘eye’, which replaced common Slavic *oko. On the other hand, it is possible to identify a borrowing, probably of Iranian origin, in Russian sobaka ‘dog’, besides the less frequent pës, which reflects common Slavic *pьsъ ‘dog’. Starostin offered a simple solution: eliminate all borrowings before any calculation. Applying this procedure to the testing languages used for the estimation of the constant of disintegration λ, we reach a lower value of the constant and its significantly smaller dispersion (table 3). Starostin compared the proportions of inherited lexicon in histories of the same languages during various times of divergence, as related to millennial time spans, specifically in some Romance languages versus Vulgar Latin from the middle of the first millennium AD and versus early classical Latin from the time of Plautus, c. 200 BC. The values of c in table 2 are now calculated without loans; time is expressed in millennia: Table 2 language

c=

N(t) No

, t = 1.5

λ=

ln c -t

, t = 1.5

c=

N(t) -t

, t = 2.2

75/97 = 0.77

λ=

ln c -t

, t = 2.2

French

88/99 = 0.89

0.07

0.12

Spanish

90/98 = 0.92

0.06

79/97 = 0.80

0.10

Rumanian

87/96 = 0.91

0.06

76/95 = 0.80

0.10

For the differences between the results in the third and fifth columns Starostin n c , is not valid. finds the only explanation, the formula (11), implying λ = l____ -t The empirical figures from the table 2 confirm that the optimal approximanc tion is the function λ* = _λ = l____ (20). t -t² The preceding thoughts are based on the data in Table 3:

etymologie.indb 46

21.1.2014 11:36:59

language English

age t [millennia] λ after Swadesh λ without loans 1.3

0.14

German

1.2

Norwegian (Riksmål)

1.0

Icelandic French Spanish Rumanian

λ* = λ / t

0.10

0.08

0.08

0.05

0.04

0.20

0.05

0.05

1.0

0.06

0.06

0.06

1.5

0.09

0.07

0.05

1.5

0.07

0.06

0.04

1.5

0.09

0.06

0.04

Japanese

1.2

0.11

0.06

0.05

Chinese

2.6

0.10

0.10

0.04

47 Blažek

It is apparent that the dispersion of the ‘constant of disintegration’ λ according to Swadesh is very high, from 6 to 20%. After the elimination of borrowings, the dispersion of this value for the nine languages analyzed tapers to 5–10%. The interval will be yet narrower in the case where λ is a function of time. Abstracting specifically from the data on English, this value oscillates from 4 to 6%. These results led Starostin to the new value of the ‘constant of decrease’: λ = 0.05 per millennium. The situation of English is more complex, since it seems that its development is faster than is usual in other languages. This phenomenon is undoubtedly connected with the massive influence of Old Norse in the period 800–1100 and Old French in the following five centuries, causing, according to Starostin, certain pidgin-like features in English. But even the new value of λ = 5% does not prevent the tendency to reach a more recent date of divergence, especially in the case of longer time periods. Starostin seeks a solution in the following idea. It is empirically proven that individual words in the lexicon of every language, including BTL, are replaced unevenly. If the words in any language were to be ordered from least stable to most stable, the words with the lowest stability would be replaced most quickly, while the more stable words would have a longer life. This means that the speed of lexical change decreases over time. Summing up, „c“ is not a constant, but a function of time, c = c(t) and formula (9) should be modified as follows: (21) N(t) = No ∙ e -λ · c⁽t⁾ · t² for a development of one language, where c(t) = N(t) , and N₀ (22) N(t) = No ∙ e-²λ · √c(t)∙ t² for the divergence of two languages, developed from a common protolanguage.

etymologie.indb 47

21.1.2014 11:37:00

48 Blažek

From here it is possible to deduce for the time of development of one language (23), or for the time of divergence of two languages (24): (23) t = √ln c √ -λc (24) t = √ ln c √ -2λ√c The result is a transcendental function, since c = c(t). The easiest way of determining of the time of divergence for the empirically investigated values is offered in Table 4, calculated by Sergei Starostin: c₁ 0.99 0.97 0.95 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 c₂ 0.97 0.94 0.90 0.81 0.72 0.64 0.56 0.49 0.42 0.36 0.30 0.25 0.20 0.16 0.12 0.09 0.06 0.04 0.02 0.01 t 0.3

0.8 1.0

1.5

2.0

2.4

2.8

3.2

3.7

4.1

4.7

5.3

6.0

6.8

7.8

9.0

10.7 12.7 16.6 21.5

Now it is possible to return to the question of the time of divergence between German and French. In both languages there are 3 loans in the BTL and 33 common cognates. 3 Hence c₂ = = 3 = 0.351 = 35.1%. 100 - 3 - 3 94 The corresponding time of divergence is c. 4 220 years. Naturally, it is an exaggeration to conclude that two languages were separated in a single specified decade. It is better to use the formulation that their common protolanguage disintegrated in the 23rd cent. BC. 3.1 The situation of two asynchronically attested languages is solved by Starostin differently from Swadesh. Starostin’s strategy consists in projection of the historical data to the present level, and only after this synchronization the same approach as for living languages is applied to them. It is useful to demonstrate this procedure on specific idioms, for example, classical Latin (e.g. of Caesar: 1st century BC) and Gothic of Wulfila’s translation of the New Testament (4th century AD). The Latin corpus (i.e. the 100-word-list) is complete, while in the Gothic list 18 units are missing (if Crimean Gothic ada ‘egg’ is included). This means that there are 82 common semantic pairs from the BTL, and of these 39 cognates (i.e. etymologically related forms) inherited from a common protolanguage. The proportion 39/82 = 47,6%. A language recorded

etymologie.indb 48

21.1.2014 11:37:01

at the time interval Δt ago would preserve till the present c-times fewer words from BTL. For Latin recorded 20.5 centuries ago it is c. 0.845. If Gothic had existed till the present time, in its hypothetical descendant the share of the preserved BTL would be 0.892 (see table 4). The common protolanguage of Latin and Gothic projected into the present would preserve cLG · cL · cG = 0.476 · 0.842 · 0.892 = 0.357, i.e. 35.7% common words. We may mention that the result of the comparison of German and French gave the share 0.351. This means that the dating of the divergence of the representatives of modern Germanic and Romance languages is practically the same as the dating of the divergence of Latin and Gothic, the 23rd cent. BC. This seems quite natural, but for the ‘classical glottochronology’ it was an unattainable goal.

49 Blažek

Application of ‘recalibrated’ glottochronology to old Germanic languages Six old Germanic languages, namely Gothic, Old Norse, Old English, Old Frisian, Old Saxon, Old High German, are described enough to apply the method described above in §3. To supplement the words missing in the limited Old Saxon lexicon, the Middle Low German counterparts were used in five cases, once the Crimean Gothic equivalent was taken into account, when the Biblical Gothic item was missing (‘egg’). The following mutual percentages and corresponding time intervals of divergence for every pair of languages were obtained – Table 5: language / % // age of divergence

Old Norse

Old English

Old Frisian

Old Saxon

Old High German

Gothic

76/84 = 90.48 1030

75.5/84 = 89.88 1060

69/82 = 84.14 1370

75/84 = 89.29 1090

75.5/84 = 89.88 1060

95/100 = 95.00 730

85.5/95 = 90.00 1050

93/99 = 93.94 800

92.5/100 = 92.50 900

91.5/95 = 96.32 620

97.5/99 = 98.48 390

98/100 = 98.00 450

90/94 = 95.74 670

91/95 = 95.78 670

Old Norse

Old English

Old Frisian

Old Saxon

etymologie.indb 49

97.5/99 = 98.48 390

21.1.2014 11:37:02

50 Blažek

To construct a tree-diagram depicting mutual relations of compared languages, if they are asynchronically attested, it is necessary to determine the times of attestation. The following chronology was used: Gothic – AD 360, Old Norse – AD 1100, Old Frisian – AD 1300, Old Saxon – AD 900. In case of Old English and Old High German, two variant times A and B were tested, AD 800 for beginning and AD 900 for florescence of their literary traditions. We have measured the time interval of divergence Δ (Table 1). Knowing the times of attestation t₁ and t₂ in every language pair, we are able to date the partial protolanguages for every pair of languages (t₀) and finally to establish the most probable scenario of the development. We start by the following equations: 2 Δ = (t₁ - t₀) + (t₂ - t₀); thus Δ = (t₁ + t₂)/2 - t₀; from here t₀ = (t₁ + t₂)/2 – Δ Scenario A pair of languages

(t₁ + t₂)/2

Δ

(t₁ + t₂)/2 – Δ = t₀

Gothic – Old Norse

(360+1100)/2 = 730

1030

730 – 1030 = -300

Gothic – Old English

(360+800)/2 = 580

1060

580 – 1060 = -480

Gothic – Old Frisian

(360+1300)/2 = 830

1370

830 – 1370 = -540

Gothic – Old Saxon

(360+900)/2 = 630

1090

630 – 1090 = -460

Gothic – OHG

(360+800)/2 = 580

1060

580 – 1060 = -480

Old Norse – Old English

(1100+800)/2 = 950

730

950 – 730 = +220

Old Norse – Old Frisian

(1100+1300)/2 = 1200

1050

1200 – 1050 = +150

Old Norse – Old Saxon

(1100+900)/2 = 1000

800

1000 – 800 = +200

Old Norse – OHG

(1100+800)/2 = 950

900

950 – 900 = +50

Old English – Old Frisian

(800+1300)/2 = 1050

620

1050 – 620 = +430

Old English – Old Saxon

(800+900)/2 = 850

390

850 – 390 = +460

Old English – OHG

(800+800)/2 = 800

450

800 – 450 = +350

Old Frisian – Old Saxon

(1300+900)/2 = 1100

670

1100 – 670 = +430

Old Frisian – OHG

(1300+800)/2 = 1050

670

1050 – 670 = +380

Old Saxon – OHG

(900+800)/2 = 850

390

850 – 390 = +460

Results A 1. There is an apparent opposition of Gothic vs. Northwest Germanic. The average value (-300 + -480 + -540 + -460 + -480)/5 = -2260/5 = -452 ≈ 450 BC determines a probable dating of disintegration of the common Germanic protolanguage.

etymologie.indb 50

21.1.2014 11:37:02

2. The younger result of the Gothic-Nordic disintegration was probably caused by a secondary neighbourhood of ancestors of Goths (East Germans) and newly incoming ancestors of North Germanic populations in southern Scandinavia. Abstracting from this value, the average will be (-480 + -540 + -460 + -480)/4 = -1960/4 = -490, i.e. 490 BC. 3. The opposition of North vs. West Germanic is also apparent. The average (220 + 150 + 200 + 50)/4 = 620/4 = 154 ≈ 150 AD determines a probable dating of disintegration of the common Northwest Germanic protolanguage. 4. The data of disintegration within the West Germanic branch are concentrated in the time interval from 350 to 460 AD. The average value is (430 + 460 + 350 + 430 + 380 + 460)/6 = 2510/6 = 418.33 ≈ 420 AD. 5. If the closer relation between Old Saxon and Old High German expressed by the result 460 AD is ascribed to a mutual interference between two neighbouring literary languages, the closest relatives are Old Saxon and Old English (460 AD), preceded by Old Frisian (430 AD to both Old Saxon and Old English). The position of Old High German may be determined in two ways: (i) a simple average (350 + 380 + 460)/3 = 1190/3 = 396.66 ≈ 400 AD; (ii) abstracting from the result between Old Saxon and Old High German, the average is (350 + 380)/2 = 365 AD. These results determine the time, when the ancestor of Old High German separated from other West Germanic languages.

51 Blažek

Tree-diagram A

-500

-300

-100

+100

+300

+500 Goth.

Germanic -450 (-490*)

ON Northwest Germanic 150

430 West Germanic 460 400 (365*)

OFris. OEng. OSax. OHG

The asterisked data are calculated without the “deviating” pairs, Gothic-Old Norse (300 BC) and Old Saxon-Old High German (460 AD).

etymologie.indb 51

21.1.2014 11:37:03

52

Scenario B

Blažek pair of languages

(t₁ + t₂)/2

Δ

(t₁ + t₂)/2 - Δ = t₀

Gothic – Old Norse

(360+1100)/2 = 730

1030

730 - 1030 = -300

Gothic – Old English

(360+900)/2 = 630

1060

630 - 1060 = -430

Gothic – Old Frisian

(360+1300)/2 = 830

1370

830 - 1370 = -540

Gothic – Old Saxon

(360+900)/2 = 630

1090

630 - 1090 = -460

Gothic – OHG

(360+900)/2 = 630

1060

630 - 1060 = -430

Old Norse – Old English

(1100+900)/2 = 1000

730

1000 - 730 = +220

Old Norse – Old Frisian

(1100+1300)/2 = 1200

1050

1200 - 1050 = +150

Old Norse – Old Saxon

(1100+900)/2 = 1000

800

1000 - 800 = +200

Old Norse – OHG

(1100+900)/2 = 1000

900

1000 - 900 = +100

Old English – Old Frisian

(900+1300)/2 = 1100

620

1100 - 620 = +480

Old English – Old Saxon

(900+900)/2 = 900

390

900 - 390 = +510

Old English – OHG

(900+900)/2 = 900

450

900 - 450 = +450

Old Frisian – Old Saxon

(1300+900)/2 = 1100

670

1100 - 670 = +430

Old Frisian – OHG

(1300+900)/2 = 1100

670

1100 - 670 = +430

Old Saxon – OHG

(900+900)/2 = 900

390

900 - 390 = +510

Results B 1. There is an apparent opposition of Gothic vs. Northwest Germanic. The average value (-300 + -430 + -540 + -460 + -430)/5 = -2160/5 = -432 ≈ 430 BC determines a probable dating of disintegration of the common Germanic protolanguage. 2. The younger result of the Gothic-Nordic disintegration was probably caused by a secondary neighbourhood of ancestors of Goths (East Germans) and newly incoming ancestors of North Germanic populations in southern Scandinavia. Abstracting from this value, the average will be (-430 + -540 + -460 + -430)/4 = -1860/4 = -465, i.e. 465 BC 3. The opposition of North vs. West Germanic is also apparent. The average (270 + 150 + 200 + 100)/4 = 720/4 = 180 = 180 AD determines a probable dating of disintegration of the common Northwest Germanic protolanguage. 4. The data of disintegration within the West Germanic branch are concentrated in the time interval from 430 to 510 AD. The average value is (480 + 510 + 450 + 430 + 430 + 510)/6 = 2810/6 = 468.33 ≈ 470 AD.

etymologie.indb 52

21.1.2014 11:37:04

5. If the closer relation between Old Saxon and Old High German expressed by the result 510 AD is ascribed to a mutual interference between two neighbouring literary languages, the closest relatives are Old Saxon and Old English (510 AD). The results of Old High German indicate younger results in sum with Old Saxon and Old English (510 and 450 respectively), than Old Frisian with Old Saxon and Old English (430 and 480 respectively). It means that it should be Old High German which separated from the common dialect continuum first preceding Old Saxon and Old English, namely (510 + 450)/2 = 480 AD, preceded by Old Frisian: (430 + 480)/2 = 455 AD (see Tree-diagram B1). If the young result of Old High German-Old Saxon will not be taken into account, the relation Old High German to Old English, 450 AD, will be practically identical with Old Frisian in its relation to Old Saxon, Old English and Old High German: (480 + 430 + 430)/3 = 446.66 ≈ 450 AD (see Tree-diagram B2).

53 Blažek

Tree-diagram B1

-450

-250

-50

+150

+350

+550 Goth.

Germanic -430

ON OFris.

Northwest Germanic 180 West Germanic 450 510 480

OEng. OSax. OHG

Tree-diagram B2

-450

-250

-50

+150

+350

+550 Goth.

Germanic -430

ON OFris.

Northwest Germanic 180 West Germanic 450 510

OEng. OSax. OHG

etymologie.indb 53

21.1.2014 11:37:04

54 Blažek

etymologie.indb 54

Statistics I. Missing words Gothic: 3 bark, 10 bone, 13 claw, 27 feather, 30 fly (v.), 35 green, 48 liver, 50 louse, 61 nose, 69 round, 75 skin, 78 smoke, 83 swim, 84 tail, 93 warm, 100 yellow; Σ 16. The item 24 egg was reconstructed *addja on the basis of Crimean Gothic ada. Old Frisian: 1 ashes, 3 bark, 24 egg, 50 louse, 100 yellow; Σ 5. Old Saxon: 61 nose; Σ 1. The items 2 ashes, 3 bark, 4 belly, 30 fly, 50 louse, 84 tail missing in the Old Saxon lexical corpus were substituted by their Middle Low German counterparts. II. Non-cognates (Σ), plus apophonical and morphological variants (ΣΣ), and shares of common cognates (σ): Gothic + Old Norse: 11/2 breast, 14 cloud, 26/2 fat, 36 hair, 53 meat, 55 mountain, 68/2 root, 70 sand, 76 sleep, 89/2 tooth. Σ 6; ΣΣ 8; σ 76/84. Gothic + Old English: 11/2 breast, 14 cloud, 26/2 fat, 28/2 fire, 36 hair, 53 meat, 55 mountain, 70 sand, 73 seed, 89/2 tooth, 94/2 water. Σ6; ΣΣ 8.5; σ 75.5/84. Gothic + Old Frisian: 5 big, 14 cloud, 17 die, 20 dry, 26 fat, 28/2 fire, 36 hair, 53 meat, 55 mountain, 68/2 root, 70 sand, 73 seed, 86/2 this, 89/2 tooth, 94/2 water, 99 woman. Σ 11; ΣΣ 13; σ 69/82. Gothic + Old Saxon: 11/2 breast, 14 cloud, 26/2 fat, 28/2 fire, 36 hair, 53 meat, 55 mountain, 70 sand, 73 seed, 86/2 this, 89/2 tooth, 94/2 water. Σ 6; ΣΣ 9; σ 75/84. Gothic + Old High German: 14 cloud, 26/2 fat, 28/2 fire, 36 hair, 53 meat, 55 mountain, 70 sand, 73 seed, 86/2 this, 89/2 tooth, 94/2 water. Σ 6; ΣΣ 8.5; σ 75.5/84. Old Norse + Old English: 3 bark, 53 meat, 55 mountain, 68/2 root, 86/2 this, 94/2 water, 100/2 yellow. Σ 3; ΣΣ 5; σ 95/100. Old Norse + Old Frisian: 5 big, 14 cloud, 17 die, 20 dry, 53 meat, 55 mountain, 68/2 root, 69 round, 76 sleep, 86/2 this, 94/2 water. Σ 8; ΣΣ 9.5; σ 85.5/95. Old Norse + Old Saxon: 14 cloud, 53 meat, 55 mountain, 68/2 root, 76 sleep, 86/2 this, 94/2 water, 100/2 yellow. Σ 4; ΣΣ 6; σ 93/99. Old Norse + Old High German: 3 bark, 11/2 breast, 14 cloud, 53 meat, 55 mountain, 68/2 root, 76 sleep, 86/2 this, 94/2 water, 100/2 yellow. Σ 5; ΣΣ 7.5; σ 92.5/100. Old English + Old Frisian: 20 dry, 46 leaf, 68/2 root, 69 round. Σ 3; ΣΣ 3.5; σ 91.5/95. Old English + Old Saxon: 46 leaf, 98/2 who. Σ 1; ΣΣ 1.5; σ 97.5/99. Old English + Old High German: 46 leaf, 69/2 round, 98/2 who. Σ 1; ΣΣ 2; σ 98/100. Old Frisian + Old Saxon: 5 big, 20 dry, 68/2 root, 69 round, 98/2 who. Σ 3; ΣΣ 4; σ 90/94. Old Frisian + Old High German: 5 big, 20 dry, 68/2 root, 69 round, 98/2 who. Σ 3; ΣΣ 4; σ 91/95. Old Saxon + Old High German: 11/2 breast, 52 many. Σ 1; ΣΣ 1.5; σ 97.5/99.

21.1.2014 11:37:05

etymologie.indb 55

21.1.2014 11:37:05

all

ashes

bark

belly

big

bird

bite

black

blood

bone breast

1

2

3

4

5

6

7

8

9

10 11

English

*swartaz

b

c

*`ainan

*`reustan

*`rustiz

a

a

a₁

*`lōđan

*`lēwaz

a

a

*`ītanan

*`lakkaz < *bhleg- burn

a

*fuglaz

*`raidaz

a

d

*xrefaz

d

*stōr( j)az

*`ūkaz

c

c

*wam`ō

b

*grautaz

*kwiþuz/-az/-iz

a

b

*rentōn

b

*mekilaz

*`arkuz

a

a

*askōn

*allaz

a

a

brusts

bloþ

swarts

beitan

fugls

(braiþs broad)

mikils

wamba

qiþus

azgo

alls

Proto-Germanic Gothic

brjóst

bein

blóð

svartr

blár

(blakkr usky)

bíta

fugl

(breið broad)

stórr

(grautr porridge)

mikill, mykill

(búkr body)

wǫmb

kviðr

bǫrkr

aska

allo

Old Norse

Table 6: Lexical material of old Germanic languages

brēost

bán

blōd

sweart

(blǽwen blue)

blæc

bítan

fuġol

(brád broad)

(> stór great)

ġréat

micel, mycel

hrif b., womb

būc

wamb, womb

(cwiđ(a) womb)

rind(e)

asce/æsce

eal

Old English

brust

briast

bēn

blōd

swart

(blāu blue)

bīta

fugel

brēd big, bread

stōr big, high

grāt

hrif, href

būk/būch b., trunk

wamme

alle, al

Old Frisian

briost

bēn

blōd

swart

(blāo blue)

(blac ink)

bītan

fugal

(brēd broad)

(stōri famous)

grōt big, mighty

(mikil great, much)

Mwamme

rinda

Mborke f.

Masche

all

Old Saxon ( M= MLG)

brust

bein

bluot

swarz

(blāo blue)

(bla(c)h id.)

bīz(z)an

fogal

(breit broad)

(stuori big, strong)

(grōz big, strong)

mihhil

href

būh belly, body, stomach

wamba, wampa

(quiti vulva)

rinta

aska

all

Old High German (MHG)

etymologie.indb 56

21.1.2014 11:37:07

burn

claw

cloud

cold

come

die

dog

drink (v.)

dry

ear

earth

12

13

14

15

16

17

18

19

20

21

22

*xunđaz

*rakkōn

a

b

*mulđō(n)

* erþō

b

*auzōn

a

*gaistaz

a

*þurzuz

a

b

*đrenkanan

*sweltanan

a

*ster`anan

*kulđjōn

a₁

c

*kalđaz

a

b

*wulknan

d

*đewanan

*skeujan

c

*đawjanan

*melhman

b

a₁

*naglaz

b

a

*klō/ēwō

a

*kwemanan

*swelanan

b

a

*`rennanan

a

(mulda dust)

aírþa

auso

þaursus

drigkan

hunds

swiltan

diwan

qiman

kalds

milhma

(ganagljan to nail)

brinnan

brenna, brinna

mold

jǫrð

eyra

þurr

drekka

rakki

hundr

svelta

deyja

koma

kuldi

kaldr

ský

(OSw molin)

nagl

kló

(svæla to smoke)

beornan

erthe mulde e., dust

molde e., dust, sand

āre

gâst

drinka

hund

sterva

kuma, koma

kald

wolken

neil

klawe / klē

swela

berna

eorđ(e)

eāre

đyrre

drincan

ræc

hund

sweltan

steorfan

díeġan

cuman

ceald

wolcen

scéo

næġ(e)l

clawu

swelan

brinnan

ertha/erđa

ōra

thurri

drinkan

hund

sweltan

ster`an

dōian

cuman

Mkülde

kald

wolkan

(skio/skeo light cloud)

nagal

clāuua

Mswelen

molta e., dust

erda

ōra

durri

trinkan

(n. d. rache sheep’s dog)

hunt

(swelzan to burn)

sterban

touwen

comen, cumen, quemen

kalt

wolcan

nagal

klāwa

swilizōn

brinnan

etymologie.indb 57

21.1.2014 11:37:08

*funōn

*fuwer : fūr

c₁

a

green

hair

hand

35

36

37

give

good

33

34

a

full

32

*taglan/-az

*xanđuz

*munđō

b

b

*xēran

a

*grōniz

a

*gōđaz

*ge`anan

*fullaz

*fōtz, fōtuz

*fleuganan

a

a

b

a

*fiskaz

*bēlan

b

c

a

*ailiđaz

fly (v.)

fish

29

*feþrō

a

*smer-þra-

b₁

a

*smerwan/-ōn

b

foot

fire

28

*faitiđa

a₁

31

feather

27

*augōn

*faitaz

a

*ajjaz

a

30

eye

fat

25

26

a

egg

*matjanan

b

24

*etanan

a

eat

23

handus

tagl

goþs

giban

fulls

fotus

fisks

fon

smairþr

augo

*addi

matjan

itan

mund

hǫnd

(tagl hair, tail)

hár

grœnn

góðr

gefa

fullr

fótr

fljúga

fiskr

fúrr/fýrr

funi

bál

eldr

ǫðr

smør, smjǫr

feitr

auga

egg

(metja to lap)

eta

mund

hond

(tœgl tail)

hœr

grǽne

gód

ġiefan

ful(l)

fōt

fleōġan

fisc

fýr / fiur

bǽl

ael(e)d

feđer

smeoru

fǽt(t)

fǽt(t)

ēaġe

œġ

(metian to supply with food)

etan

hand

hēr

grēne

gōd

geva

fol, ful

fōt

fliāga

fisk

fiōr / fiūr

fethere

(smere tallow)

fatt, fet

āge

eta

hand

hār

grōni

gōd

ge`an

ful

munt

hant

(zagal tail)

hār

gruoni

guot

geban

fol

fuoz

fliogan

Mvlēgen fōt

fisc

fiur / fuir

federa

smero

feizzit

Mveiz

ouga

ei

ezzan

fisk

fiur / fuir

ēld

feđara

smero

Mveet

ōga

ei

etan

etymologie.indb 58

21.1.2014 11:37:09

man (vir)

51

leaf

46

louse

know

45

long

knee

44

49

kill

43

50

I

42

lie

horn

41

liver

heart

40

47

hear

39

48

head

38

pl. firar karl

b

lús

langr

lifr

liggja

lauf

blað

veit

vita

kunna

kné

drepa

*karlaz *kerlaz

laggs

ligan

ek deyða

*ferxwjaz

*lūsz/-iz

*langaz

*li`rō(n)

laufs

hjarta horn

a

-

-

-

*leg( j)anan

*`lađan

*lau`az/-an

a

b

-

*waita 1/3 sg.pr. wait

b₁

witan

*weitan inf.

b

kunnan

*kunnan

kniu

us-qiman

dauþjan

ik

haúrn

haírto

a

a

d

*knewan

*kwaljanan

b

c

*eka

*đauđjanan

a

*xurnan

a

a

a

*xertōn

heyra

hausjan

*xausjanan *xauzjanan

a

hofuð

*xa`uđan

haufuð

a₁

haubiþ

*xau`uđan *xau`uđin

a

pl. firihos Mkerle

Mlūs

lang

Mlever

liggian

(lōf foliage)

blad

wēt

witan

kunnan

kneo / knio

quellian

bi-dōdian

ic

horn

herta

hōrian

hōbid

pl. fíras (zerl servant)

long/lang

livere

lidz(i)a

(lāf foliage)

bled

wēt

wita

kunna

kniu / knē / knī

dēda

ek / ic

horn

herte/hirte

hēra/hōra

hā(ve)d

carl < ON ceorl

lús

lanġ

lifer

lieġan

léaf

blæd

wát

witan

cunnan

cnéo(w)

cwellan

dýdan

ic

horn

heorte

hieran/hýo

heáfod

háfod

karl

g.pl. fireo

lūs

lang

lebara

ligan/ liggen

(loub foliage)

blat

weiz

wizzan

kunnan

chniu

quellen

tōten

ih

horn

herza

hōren

houbit

etymologie.indb 59

21.1.2014 11:37:10

56

mouth

mountain a

55

*filzan

*munþaz

*mūlōn

*sna``ōn

c

b

c

munþs

faírguni

*ferg(w)unjan *ferg(w)unjō

a

b

*`ergan/-az

*tunglan

b

mena (bairgahei hill-country)

*ketwa/u-

*mēnōn

e

a

*memzan

d

moon

*līkan

c leik

*xulđan

b

mimz

*fleiskan

a

(múli snout)

munnr / muðr

all

(ǫrgyn Earth)

(bjarg rock)

tungl

máni

kjǫt

(lík body)

hold

mjǫk

*mekuz

c

mangr

(ǫl- very)

gumi

(verr husband)

mannr maðr

margr

manags

(filu much)

guma

wair

manna

b₁

54

*managaz/-igaz

b

meat

*felu

a

53

*gumōn

e

many

*wiraz

d

52

*mannōn ~ *mannz

c

múð

fierġen-

beorġ

móna

(líc id.)

(hold corpse)

flǽsc

maniġ

fe(a)la

ġuma

wer

manna monn mann

snabba

mūla

mūth

berch

mōna

(līk id.)

flēsk

manich

(snabbe beak)

(Mmūl n. snout)

mūth

berg

māno

(līk id.)

flēsk

manag/-ig

(filu/filo much)

gumo / gomo

felo/ful

wer

wer

man(n)

(breid-goma bridegroom)

man(n)

(mūla snout)

mund

(Fergunna)

berg

māno

(līh id.)

fleisc

(manag / menig much)

filu

gomo

wer

man(n)

etymologie.indb 60

21.1.2014 11:37:11

rain

road

root

round

67

68

69

b

person (homo)

64

red

*gumōn

a₁

one

63

65

*manniska- n. adj.

a

not

62

66

*mannōn ~ *mannz

a

nose

61

*wrōt-

*sena-walaz

*sena-waltaz

cf. ON kringla circle

a₂

a

a₁

b

*trenđaz

*wurtiz *wurti-waluz

c

*wegaz

a a₁

*rauđaz

*regnan /-az

*ainaz

*ne

*nasō

*naxtz

c

-

-

a

a

a

*neujaz

night

a

new

60

*swirhjan-

b

59

*xalsaz

a

neck

58

*namōn *namnan

a

name

57

waúrts

wigs

rauþs

rign

guma

mannisks

manna

ains

ne / ni

nahts

niujis

hals

namo pl. namna

kringlóttr

(sí-)valr

rót

(urt plant)

vegr

rauðr

regn

gumi

menskr

mannr maðr

einn

eigi, ekki

né

nǫs

nátt/nótt

nýr

svíri

háls

nafn

sinewealt

rót < ON

wyrt

weġ

reád

reġn

ġuma

mennisc

mann

án

ne

nasu

neaht/niht

níewe

swéora

heals

noma

trind

wortele

wei

rād

rein

(breidgomo groom)

manniska

mann / monn

ān/ēn/on

ni / ne

nose

nacht

nīe

hals

nama/noma

Msinewolt

sinuwel

wurt

weg

rōd

regin / regan

gumo

mennisko

man(n)

ēn

ni / ne

naht

niwi/nigi

hals

namo

sin(a)wel

wurz

weg

rōt

regan

gomo

me(n)isko

man(n)

ein

ni / ne

nasa

naht

niuwi

hals

namo

etymologie.indb 61

21.1.2014 11:37:12

sand

say

see

seed

sit

skin

sleep (v.)

small

70

71

72

73

74

75

76

77

*maþliđa dixi

c

svǫrðr horund

*felmaz

*skenþan

*swarđuz /-ō

b₁

c

d

sofa lítill

a

leitils

*swefanan

*lītilaz *lūtilaz *luttilaz

b

slepan

*xarundan

*slēpanan

e

a

skinn

all/fell

húð

*fellan

(faura-filli foreskin)

b

sitja

*xūđiz

sitan

sáð

a

*sēmōn

c₁

(manaseþs mankind)

frae/frjó

eigin

sjá

lýtel

swefan

slœpan

sweard / swearð

scinn < ON

film

fell

hýd

sitan

sǽd

seón

(mæþlan to speak)

secġan

segja (mál speech)

cweðan

sand

(mealmiht sandy)

kveða

sandr

(malmr ore)

*setjanan

*sēđan *sēđiz

c

fraiw

sai¹an

(maþlja I talk)

qiþan

malma

-

*aigenan

*fraiwan

a

b

*sexwanan

*sagjanan *sagēnan

b

-

*sanđaz /-an

*kweþanan

b

a

*malmōn/-az *melmaz

a

littich

lītik

slēpa

(swarde scalp)

(skene bark)

filmene

fell

hēd

sitta

sēd

sia

sedza / sega

quetta

sand/sond

luttil

slāpan

(swarde scalp)

fel

hūd

sittan

sāmo

sād

(Mīne awn)

sehan

(mathal lawcourt)

seggian

quenthan

sand

melm sand, dust

liuzil luzzil

slāfan

(swarde scalp)

(Mscint peel)

fel

hūt

sizzen

sāmo

sāt

sehan

sagēn

quedan

sant

(melm dust)

etymologie.indb 62

21.1.2014 11:37:13

smoke n.

stand

star

stone

sun

swim

tail

that

this

thou

tongue

teeth

tree

78

79

80

81

82

83

84

85

86

87

88

89

90

*raukiz/-az

*smeukanan v.

a

b

*sōwilō

a₁

*þat

*sa

*þa-sa

b

a

b

*tungōn

*tanþuz

*tunþuz

*`aumaz

a

a

a₁

a

*þū̆

*jainaz / jēnaz

a

a

*taglan/-az

*xalēn

b

c

*stertaz

*sunnōn

a

a

*xalluz/-az

b

*swimmanan

*stainaz

a

a

*sternōn

-

*stanđanan

*smēxiz

c

-

*smalaz

b

tunþus

tuggo

þu

sa

þata

jain

(tagl hair)

sauil

sunno

(hallus rock)

stains

staírno

standan

(*smals in smalista least)

tǫnn

tunga

þú

sá

þat

(h)inn

hali

tagl

stertr

svimma

sól

sunna

hallr R halaz

steinn

stjarna

standa

reykr

smár

(*smalr in smalmenni)

beám

tóð

tunge

ðū̆

ðes

sé / se

ðæt

ġeon

tæġl

steort

swimman

sunne

stán

steorra

standan

smoca n.

réc

smæl

bām

tōth

tunge / tonge

thū̆

this

thet

jen(e)

stert

swimma

sunne

stēn

stēra

stonda

rēk

smel

bōm

tand

tunga

thū̆

*these

that

Mjēne

Mstart

sunna

stēn

sterro

standan

rōk

smal

boum

zan(t)

zunga

dū̆

desēr

daz

jenēr

zagal

sterz

swimman

sunna

stein

stern / sterro

stantan

rouh

smāhi

smal

etymologie.indb 63

21.1.2014 11:37:14

two

walk/go

warm

water

we

what?

white

who?

woman

yellow

91

92

93

94

95

96

97

98

99

100

*līþanan

*gangan

*gēnan

a

b

c

*kwēniz

*wī`an

*gelwaz

*gulwaz

a

a₁

a₃

b

*xwar-jis (where-he)

a₁

*xwiz

a₂

*kwenōn

*xwaz

a

*xwītaz

a₁

*xwat

*wīz

a

a

a

b

(qens wife)

qino

(¹arjis which)

¹as

¹eits

¹ata

weis

gulr

víf

(kván wife)

kona

hverr

hver (RSw hwaR)

hvítr

hvat

vér

(á river)

vatn

*axwō

a a¹a

*watar *watōn *watnan wato

varmr

*warmaz

b

heitr

ganga

líða

tveir

tré

baðmr (OSw bagn)

*xaitaz

iddja -jedum we went

(at-)gaggan

ga-leiþan

twai

triu

bagms

a

d

*twai ~ *twōz

*trewan

b

a

*`agmaz ~ *`agnaz

a₁

ġeolu

wíf

cwén

cwéne

hwá

hwít

hwœt

wé

(eá stream)

wœter

wearm

hát

éode went

ġán

gangan

líðan

twǽġen

tréow

trē

wīf

hwā

hwīt

hwet

wi

(ā/ē id.)

water/ weter/-ir

warm

hēt

gān

gonga / gunga

(lītha to suffer)

twēn

gelu/-o

wīf

quān

quena

hwē

hwīt

hwat

wi / we

(aha id.)

watar

warm

hēt

gān

gangan

līthan

twēne, twō

trio/treo

gel(o)

wīb

quena

hwer

hwīt

hwaz

wir

(aha id.)

wazzar

warm

heiz

gēn

gangan

(līdan to leave)

zwēne, zwō

64 Blažek

Abbreviations: M Middle, N Norse, O Old, R Runic, Sw Swedish. Note: The bold letters indicate so-called secondary synonyms, differentiated from the primary synonyms especially on the basis of their frequencies in texts. For calculation the primary synonyms are preferred. In parentheses there are cognates of different meanings. The grey empty cells indicate that corresponding lexemes in compared languages are not attested. The underlined items indicate loans.

Basic sources of Germanic languages and reconstructions Proto-Germanic: Orel (2003), Kroonen (2013). Gothic: Lehmann (1986); Starostin (2012). Old Norse: Bergsland & Vogt (1962), Starostin (2012), de Vries (1961). Old English: Clark (1916), Holthausen (1963). Old Frisian: Boutkan & Siebinga (2005), Bremmer (2009), Holthausen (1925). Old Saxon: Berr (1971), Köbler (1993a). Old High German: Köbler (1993b). Basic sources to glottochronology Classical method: Swadesh (1952, 1955). Critics of the classical method: Bergslandt & Vogt (1962), Tischler (1973). Revised methods: Embleton (1986, 2000), Starostin (1989, 1999); Burlak & Starostin (2005); cf. also Novotná & Blažek (2007). Conclusion In the first part of the present contribution differences between ‘classical’ and ‘recalibrated’ variants of glottochronology were demonstrated. In the second part the latter variant was applied to six old Germanic languages. In Table 6 the lexical data are summarized. On the basis of etymological analysis they are arranged according to their reconstructed Germanic protoforms. For calculation only cognates with exactly corresponding meanings, which represent the primary synonyms, are taken in account. Only application of this strict approach gives relatively realistic and consistent results which are in agreement with historical facts or archaelogical data. Three scenarios of development are discussed, depending on the input chronological data. The final models depicted as tree-diagrams present in principle the same topology: the first split between East and Northwest Germanic in the 5th century BC, the second split between North and West Germanic in the 2nd century AD and small differences in details appear only in disintegration of the West Germanic branch during the 5th century AD.

etymologie.indb 64

21.1.2014 11:37:16

References

65 Blažek

Berr, Samuel 1971. An etymological glossary to the Old Saxon “Heliand”. Bern: Lang. Bergsland, Knut & Vogt, Hans 1962. On the Validity of Glottochronology. Current Anthropology 3/2: 115–53. Boutkan, Dirk & Siebinga, Sjoerd Michiel 2005. Old Frisian etymological dictionary. Leiden: Brill. Bremmer, Rolf H., Jr. 2009. An Introduction to Old Frisian: History, Grammar, Reader, Glossary. Amsterdam: Benjamins. Burlak, Svetlana A. & Starostin, Sergej A. 2005. Sravnitel’no-istoričeskoe jazykoznanie. Moskva: Academija. Clark, John R. 1916. A Concise Anglo-Saxon Dictionary. New York: Macmillan. http://archive.org/details/AConciseAnglo-saxonDictionary Embleton, Sheila 1986. Statistics in Historical Linguistics. Bochum: Brockmeyer. Embleton, Sheila 2000. Lexicostatistics / Glottochronology: from Swadesh to Sankoff to Starostin to future horizons. In: Time Depth in Historical Linguistics 1. Ed. by C. Renfrew, A. McMahon & L. Trask. Cambridge: McDonald Institute for Archaeological Research, 143–65. Holthausen, Ferdinand A. W. 1925. Altfriesisches Wörterbuch. Heidelberg: Winter. Holthausen, Ferdinand A. W. 1963. Altenglisches etymologisches Wörterbuch. 2. Aufl. Heidelberg: Winter. Köbler, Gerhard 1993a. Altsächsisches Wörterbuch. 3. Aufl. 2000ff. http://www.koeblergerhard.de/aswbhinw.html Köbler, Gerhard 1993b. Althochdeutsches Wörterbuch. 4. Aufl. 1993 http://www.koeblergerhard.de/ahdwbhin.html Kroonen, Guus 2013. Etymological Dictionary of Proto-Germanic. Leiden – Boston: Brill. Lehmann, Winfred P. 1986. A Gothic etymological dictionary. Leiden: Brill. Novotná, Petra & Blažek, Václav 2007. Glottochronology and its application to the Balto-Slavic languages. Baltistica 42: 185–210, 323–46. Orel, Vladimir 2003. A Handbook of Germanic Etymology. Leiden: Brill. Starostin, Sergej 1989. Sravnitel’no-istoričeskoe jazykoznanie i leksikostatistika. In: Lingvističeskaja rekonstrukcija i drevnejšaja istorija Vostoka. Materialy k diskussijam na Meždunarodnoj konferencii (Moskva, 29.v.–2.vi. 1989g.), i. Moskva: Institut vostokovedenija, 3–39. Starostin, Sergej 1999. Comparative-historical linguistics and lexicostatistics. In: Historical Linguistics & Lexicostatistics. Ed. by Vitaly Shevoroshkin & Paul Sidwell. Melbourne: Association for the History of Language, 3–50. (AHL Studies in the Science & History of Language, 3.) Starostin, Georgij 2012. Gothic & Old Norse 110-word-lists with etymological comments. http://starling.rinet.ru/cgi-bin/response.cgi?root=new100&morpho=0&basename= new100\ier\grm&limit=-1 Swadesh, Morris 1952. Lexico-statistic dating of prehistoric ethnic contacts. Proceedings of American Philosophical Society 96: 452–63. Swadesh, Morris 1955. Towards greater accuracy in lexicostatistic dating. International Journal of American Linguistics 21: 121–37. Tischler, Johann 1973. Glottochronologie und Lexikostatistik. Innsbruck: Kowatsch. (Innsbrucker Beiträge zur Sprachwissenschaft, 11.) Vries, Jan de 1961. Altnordisches etymologisches Wörterbuch. Leiden: Brill.

etymologie.indb 65

21.1.2014 11:37:16

etymologie.indb 66

21.1.2014 11:37:17

ivo vasiljev what can etymology do for a better understanding of the vietnamese language

67 Vasiljev

1 The reason for asking such a general question, i.e. how can etymology contribute to a better understanding of the Vietnamese language, are obviously the typological specifics of this language that put it on the very opposite of languages which have been providing traditional fields for etymological research (cf. Vasiljev 2009). In order to be able to give a preliminary answer to this question, it is, first of all, necessary to see whether and to what extent etymology, as it is usually understood, can be applicable to such a language. The main characteristics of the language type which is dominant in modern Vietnamese and which was called polysynthetic by Vladimír Skalička, is that grammatical meanings are expressed by word order as well as by specific uses of certain nouns, verbs or adjectives that in other contexts bear clear lexical meanings: so if the verbs cho ‘give’ or giúp ‘help’ are placed after another verb with lexical meaning, they will embrace a grammatical meaning, that of dative. So the sentence Đọc bài này cho tôi nghe nhé! ‘Read this article for me!’ will be structured in the following way: Read – article – this – give – I – listen – exhortative particle. Individual words in Vietnamese – unlike words in the Indo-European languages – do not include any grammatical meanings (such as case, number, tense, person etc.). Consequently, there is practically no morphology in Vietnamese. As a result, most words in Vietnamese are not only monosyllables, but they have no features in them distinguishing nouns from pronouns, adjectives or verbs (in other words, there are no obvious word classes) and they have no morphological structure. Accordingly, there is also no morphological word derivation in the narrower sense of the word. This is being compensated in a variety of ways. For example, the number of adjectives is restricted to words describing an inherent quality, while the rest of words expressed in other languages by adjectives such as wooden, metallic or earthen in English are formed through the use of noun adjuncts (or attributive nouns). Some complementary qualities are expressed through separate semantemes, e.g., the word ‘smash (something)’ is expressed by the phrase đánh vỡ (literally: strike broken). Word formation thus follows the same principle as syntax: new words are coined as several types of phrases built with otherwise fully independent, monosyllabic lexical items. As a result, their semantic structure is clear. Of

etymologie.indb 67

21.1.2014 11:37:17

68 Vasiljev

course, certain words are used as generic indicators, but they still retain the phonetic features of an independent word, though their meaning may be regarded more or less as grammatical. The rest of a compound word including such a transparent “word/particle” consists of one or two independent words with their own lexical meanings. For example the common word nhà, meaning house or family, is used as a formative element of names of professionals: Nhà văn nhà + literary text Nhà báo + newspaper Nhà sử học + historiography Nhà ngôn ngữ học + linguistics Nhà nghiên cứu tiếng Anh + study, language, English

= writer = journalist = historian = linguist = Anglicist¹

One can see the great transparency of these formations with hardly any potential for etymological research. 2 There is only one lexical structure that allows for occasional formations in which one of its two constituent words may have an uncertain meaning or be empty of any meaning at all in contemporary Vietnamese. This structure represents paratactic two-noun phrases the constituents of which are supposed to be similar in meaning. This phrase has a general meaning of ‘such and such objects as a whole, the sum of all objects of a given type’: Hoa quả Cây cỏ Cỏ cây Xe cộ Chợ búa Nhà cửa Chó má

blossom – fruit plant – grass grass – plant vehicle – sledges for wood market – ? house – door dog – dog in Thai

fruit in general flora flora vehicles in general markets in general houses in general dogs in general

Though in most cases the lexical meaning of both constituents of the phrases is transparent enough, in some cases (as búa or má) it cannot be explained within the context of modern standard Vietnamese; in some other cases (as cộ) the 1

etymologie.indb 68

These examples include polysyllabic words sử học ‘historiography’, ngôn ngữ học ‘linguistics’ and nghiên cứu ‘study’ that are imported from Chinese and do not follow some of the rules of Vietnamese grammar. They will be discussed later.

21.1.2014 11:37:18

meaning may not be known to many present-day native speakers. However, such cases are rather rare. Another common word formation pattern of certain interest from the point of view of etymology is partial reduplication of one-syllable words, usually adjectives or verbs. The meaning of these polysyllabic formations is usually some modification of the original meaning of the base word, while the second element as such carries no separate meaning. For instance: Mới Cứng Gần Xa Rộng Hẹp

new hard close far wide, broad narrow

mới mẻ cứng cỏi gần gũi xa xôi rộng rãi hẹp hỏi

69 Vasiljev

novel stubborn emotionally close far away large narrow-minded

One can see that the reduplications are rather variable in form, with little mutual resemblance. Yet, sometimes it is difficult to say, whether a disyllabic formation is a duplicated formation or a pattern where one adjective is added to another, both carrying a similar meaning, a pattern that is also very common, e.g., to ‘big’ plus lớn ‘big’, meaning ‘great’ (to lớn). Let us consider the following example: Sâu

deep

sâu sắc

very deep, or with deep in-sight

However, an alternative explanation is also possible, where sắc would be a separate word meaning ‘sharp’: ‘deep’ and ‘sharp’ would perfectly add up to more or less the same meaning (‘with deep and sharp in-sight’). Occasionally there are formations that are neither phrases of two independent words nor reduplicative formations, such as bù nhìn ‘puppet’. Here the possible meanings of the individual “monosyllabic words” do not add up to anything meaningful (bù ‘compensate’, nhìn ‘look at’). Can it be perhaps a loan word? Or let us consider the phrase chuột rút ‘cramp’: the first element means ‘mouse’, the second ‘pull out’. A thorough etymological research may certainly help. Such cases are, however, rare. 3 From what has been said, it would seem that within the scope of pure Vietnamese there is little room for etymology. Yet, the external relations of the language have still to be examined, the genealogical origins of the language in the first place.

etymologie.indb 69

21.1.2014 11:37:18

70 Vasiljev

etymologie.indb 70

According to most recent theory (cf. Phạm Đức Dương 2007: 29), the Vietnamese language must have developed from close contact of a language coming from the Austro-Asiatic language family that possessed initial consonant clusters and prefixed grammatical morphemes and was not a tonal language and another one from the Thai language family, monosyllabic, tonal and with no morphology. The contact of these two hypothetic languages over a rather long period of time resulted in a new language, a monosyllabic tonal language without consonant clusters and prefixes that came to be called Vietnamese. The former of these two languages must have belonged to a people of hunters and gatherers, while the latter was spoken by a sedentary people of rice growing farmers. Accordingly, some of the most basic vocabulary of the new language is traceable back to Austro-Asiatic languages spoken in the hilly areas of the continental hinterland of Northwest Vietnam and Laos stretching from there to Burma and the Northeast of India on the one hand and to the southern tip of the Malay Peninsula on the other. This process must have happened some three thousand years ago, at a time where no script was available to fix it. Moreover, the Austro-Asiatic family (sometimes called Mon-Khmer, which, however, is only one subgroup of languages within the Austro-Asiatic family) was not yet sufficiently studied. Thus only few isoglosses are known, e.g. Vietnamese nước – in Mường, an archaic form of Vietnamese spoken now by the Mường minority people in North Vietnam, pronounced as nạc (ươ/a being a regular correspondence of sounds seen in the history of Vietnamese or in Vietnamese dialects) – can be traced back to dak with the same meaning in some Austro-Asiatic languages. Words of Thai affiliation in most cases indicate sedentary village life based on rice growing agriculture (typically, Vietnamese gạo ‘rice’ and Thai khao with the same meaning). However, due to the absence of morphological frameworks, the non-existence of historical evidence and limited results of comparative studies, it is, in general, difficult to decide whether similar sound forms resulted from common origin, borrowing or accidental coincidence. Under these circumstances the room for etymological studies seems to be limited. Yet, since the 2nd century before the Christian era the historical scene of the country that had been home to the developments described above radically changed. The country was occupied by forces and administrators of successive Chinese dynasties and many descriptions of historical events, including some features of the country and its people, were given in literary works written in Chinese. As Chinese script is ideographic the interpretation of its use in specific cases is often difficult: should one interpret the meanings of written

21.1.2014 11:37:19

characters? Or have they been used as phonetic transcriptions of underlying alien words? And, in the latter case, which way of transcribing the sounds was used? Interpreting the meanings that the characters had in Chinese was the first and for centuries the most likely option. It often led to wild folk etymologies, though in the Middle Ages, these were circulated in learned circles. The best known example was the oldest name of the local people and country found in Chinese historical texts: the Giao Chỉ.² The Vietnamese students of Chinese classics took the characters at their face value: the first character meaning ‘to cross’ and the second ‘a toe’. Therefore, the people called by this name must have had ‘crossed toes’ of their feet. This interpretation was taken seriously by many educated people up to the mid-20th century. It is, however, more likely that these characters had been used regardless of their meaning just to catch some sounds similar to their ordinary pronunciation in the Chinese of the time. So far it is not known what those sounds might have been and whether they meant anything in the local language. A case in which a Chinese character was probably used for transcription of a local (“Vietnamese”) word was recently analyzed by Vietnamese historians leading to a plausible explanation of what must have been a Chinese ethnonym (exoethnonym) of the local people. In general, populations of various ethnic origin and speaking various languages that were settled to the South of the Yangtze River were summarily referred to as Việt or “Bách Việt” (Hundred Viet) by the Han (or Chinese) people originally occupying only the territories to the North of that river. When referring to specific local groups of these Viet populations the Chinese authors added attributive nouns, put in accordance with the rule of Chinese grammar before the main noun. The ancient Chinese texts referred to the people settled in the Red River Delta and its surroundings, that is, to the ancestors of the present Vietnamese people as Lac Viet. The meaning of the word Lạc was also used in various other phrases like Lạc dân ‘the Lạc people’, Lạc điền ‘the Lạc rice fields’ (Phan Huy Lê et al. 1985: 80), Lạc vương ‘the King Lạc’, Lạc tướng ‘the Lạc chieftain’ (Phan Huy Lê et al. 1985: 104f). The clue to the original meaning of the mysterious word Lạc may be, it was proposed, in the phrase lạc điền ‘the Lạc rice fields’ in which the word lạc may have been a transcription of an old Vietnamese word nạc meaning water. Growing rice on irrigated rice fields must, indeed, have been a conspicuous feature of the local landscape and people. The character Lạc that carried no contextually relevant meaning of its own could then easily be used in other phrases referring to that 2

etymologie.indb 71

71 Vasiljev

The standard Sino-Vietnamese transcription of the characters is used.

21.1.2014 11:37:20

72 Vasiljev

people, in which it became an ethnonym. This example, not the only one of its kind, shows one of the areas where etymology can contribute to a better understanding of Wörter und Sachen of Vietnam. 4 There is, however, still another area where etymological approaches are not only possible, but are also badly needed in order to improve the level of general understanding by the contemporary generations of Vietnamese people of the present state of their language. It is the realm of the so-called Sino-Vietnamese vocabulary. In East Asia Chinese was used for almost two thousand years as an international written language for both national and international communication. The pronunciation of Chinese characters was standardized about one thousand years ago. This standardized pronunciation underwent gradual changes in each country that adopted Chinese characters and written Chinese as language of their administration, education and works on history, philosophy and other areas of learning and literary creativity. In this way so-called Sino-Korean, Sino-Japanese and Sino-Vietnamese standard pronunciations of Chinese characters came into being (cf. Vasiljev 1964, 1969, Nguyễn Tài Cẩn 1979). Despite all their differences, they are much more conservative than the actual standard in Mandarin Chinese and are in many cases mutually identifiable from one language to the other. The most important thing is, however, that they are phonologically well adapted to the individual languages and modern terminology created in Japan and China since the last quarter of the 19th century could be easily imported and adopted on a massive scale also in Korea and Vietnam. As far as Vietnamese is concerned this vocabulary has some structural features that differ from ordinary Vietnamese syntax. Though Sino-Vietnamese monosyllables resemble those in original Vietnamese, they often are, as Emeneau (1951: 2) put it, “phonologically free, but syntactically not free”. This means that the Sino-Vietnamese monosyllables have all the features of a free Vietnamese word, but they are either part of polysyllabic units borrowed from Chinese and built according to the syntactic rules of the Chinese language (the main difference being the different position of adjectives and attributive nouns: before the head noun in Sino-Vietnamese as opposed to postposition of attributes in Vietnamese proper) or they have no obvious grammatical structure. The first group is represented by polysyllabic words like sử học ‘historiography’ and ngôn ngữ học ‘linguistics’ quoted above (fn. 1). In these the monosyllable học carries the meaning ‘learning’ or ‘science’ and can be considered to be a productive derivative “morpheme” which can be repeated

etymologie.indb 72

21.1.2014 11:37:20

in hundreds of terms all referring to a certain branch of learning. On the other hand the phrase học ngôn ngữ, built in accordance with a syntactic rule of the Vietnamese language proper, means ‘to learn language’, where học is a semanteme as any other verb. Also, the phrase can be freely modified, as for instance in học hai ngôn ngữ khác nhau ‘to learn two (mutually) different languages’. On the other hand such phrases as ngôn ngữ ‘language’ or nghiên cứu ‘to study’ are both built of elements neither of which serves as an independent word and they have no obvious grammatical structure, neither Vietnamese proper nor Sino-Vietnamese. The element ngữ, however, can serve as a productive derivative morpheme in such words as Việt ngữ ‘Vietnamese as a foreign language’. Moreover, because more or less productive derivative elements are “phonologically free” they can be, when necessary for some reason, transformed into free words, usually with a new specific meaning. For instance, the Sino-Vietnamese productive derivative element ngọai carrying the meaning ‘outer’, ‘related to the outside’ as in the words ngọai quốc ‘foreign country’, ngọai giao ‘foreign relations or diplomacy’, ngọai thương ‘foreign trade’, ngọai kiều ‘foreign national’, ngọai lai ‘imported from abroad or from a different language’, ngọai lệ ‘exception’ etc. has joined Vietnamese proper as an independent word meaning ‘produced abroad’ or ‘imported’ (as in vải ngọai ‘imported cloth’) or in the genealogical meaning ‘belonging to the mother’s side’ (as in bà ngọai ‘grandmother from the mother’s side’).

73 Vasiljev

5 However, Sino-Vietnamese loan words and morphemes are only one, though the most important part of the impact of Chinese upon Vietnamese throughout two millennia. A number of free words have been borrowed from Chinese dialects beyond the standardized Sino-Vietnamese forms. It was probably this way that Vietnamese borrowed the monosyllable ngoài ‘outer’ with a different tone, which is used as an independent word in normal Vietnamese syntactical contexts: e.g. nước ngoài ‘another country’ or ‘abroad’, a literal translation of the Sino-Vietnamese phrase ngọai quốc ‘foreign country’. Both phrases co-exist with differing connotations and in different syntactic contexts. The latter case is at the same time an example of a calque or a loan translation from the Sino-Vietnamese lexicon into Vietnamese proper. Passing through the Sino-Vietnamese lexicon Chinese words can easily slip into Vietnamese language. However, with the experience of massive borrowing in the 1950ies under huge economic and political influence of the People’s Republic of China upon socialist Vietnam it was soon noticed that words describing physical objects, unlike those applying to abstract notions and actions, are difficult to assimilate. Therefore, Sino-Vietnamese words like phi cơ borrowed from

etymologie.indb 73

21.1.2014 11:37:21

74 Vasiljev

Chinese (‘flying engine’, ‘aircraft’) started to be translated into corresponding Vietnamese words and built in accordance with Vietnamese syntactic rules and phi cơ was replaced with máy bay (‘engine [that] flies’). Another example of this sort is Sino-Vietnamese hỏa xa (‘fire vehicle’, i.e. ‘train’) that was replaced with xe lửa (‘vehicle [run by] fire’) or tàu hỏa (‘vessel [run by] fire’). It is worth of attention that both xe and hỏa in these phrases are independent words borrowed from Chinese, the former not being mediated by Sino-Vietnamese, while the latter was borrowed from Sino-Vietnamese. A curious example of how complex the borrowing process can be is the case of the Sino-Vietnamese word trường hơp ‘a case, an incidence’. It comes from the Japanese (Japanese proper) word baai ‘a place [where things] meet’ that is written with Chinese characters for ‘place’ and ‘meet’. When the Chinese borrowed this word they red the characters in Chinese, which accordingly became trường hợp in Sino-Vietnamese that could be re-interpreted as ‘a place [that is] convenient’, if only language worked this way (we do not accept the cognitive ethnolinguistic approach to the “inner form” which in our opinion is not supported by sufficient evidence based on linguistic data). 6 To conclude, although typologically the Vietnamese language seems not to provide an inviting terrain for classical etymology, in contact with a typologically similar language or languages it still offers many interesting opportunities for etymological research and interesting answers. References Emeneau, M. B. 1951. Studies in Vietnamese Grammar. Berkeley – Los Angeles: University of California Press. (University of California Publications in Linguistics, 8.) Nguyễn Tài Cẩn 1979. Nguồn gốc và quá trình hình thành cách đọc Hán Việt [The Origins of the Sino-Vietnamese Pronunciation and the Process of its Establishment]. Hanoi: NXB Khoa học xã hội. Phạm Đức Dương 2007. Bức tranh ngôn ngữ - văn hóa tộc người ở Việt Nam và Đông Nam Á [A Linguistic and Ethnocultural Image of Vietnam and Southeast Asia]. Hanoi: NXB Đại học Quốc gia. Phan Huy Lê, Trần Quốc Vượng, Hà Văn Tấn, Lương Ninh 1985. Lịch sử Việt Nam, tập 1 [History of Vietnam, vol. 1]. Hanoi: NXB Đại học và Trung học Chuyên nghiệp. Vasiljev, I. 1964. Příspěvky ke studiu sinovietnamské slovní zásoby [Contributions to the Study of the Sino-Vietnamese Vocabulary]. Doctoral thesis (unpublished). Vasiljev, I. 1969. Le lexique international dans les langues de l’Extrême Orient. Archiv orientální 37: 589–96. Vasiljev, I. 2009. The type of Vietnamese. In: Recherches fonctionnelles et structurales 2009. Ed. by B. Vykypěl and V. Boček. München: Lincom Europa, 65–100. (Travaux linguistiques de Brno, 5.)

etymologie.indb 74

21.1.2014 11:37:21

helena karlíková etymologische rekonstruktion mit unterstützung durch die kognitive linguistik?

75 Karlíková

Die etymologische Analyse des Wortes stellt bekanntlich ein relativ kompliziertes Problem dar, das – vereinfacht gesagt – nicht nur die Erklärung der jeweiligen Wortform im Kontext der verwandten Sprachgemeinschaft sowie der breiten historischen und sozio-kulturellen Zusammenhänge, sondern auch die Erklärung seiner Genesis und die Entwicklung der motivierenden Elemente einschließt. Es ist die Rekonstruktion der ursprünglichen Bedeutung des Wortes, die meistens viel schwieriger ist als die Rekonstruktion seiner Form. Bei der Rekonstruktion der Form kann man sich mehr oder weniger auf eine Reihe gut beschriebener und auf Richtigkeit überprüfter Regeln der phonologischen Entwicklung von prähistorischen Phasen zur gegenwärtigen Form des Wortes in der jeweiligen Sprache verlassen und so die einzelnen Segmente seiner internen morphophonologischen Struktur samt ihrer Funktion beschreiben. Bei der Rekonstruktion der Wortbedeutung sind jedoch solche allgemeinen in der diachronen Sprachwissenschaft anerkannten Regeln nicht vorhanden. Wie bekannt, behält nämlich die lexikalische Einheit während ihrer Existenz, während der sie des Öfteren eine dramatische Entwicklung durchläuft, ihre etymologische (ursprüngliche) Bedeutung nicht; sie gewinnt vielmehr fast in der Regel eine neue, sekundäre Bedeutung, ja sogar Bedeutungen. Das Ergebnis einer solchen Entwicklung sind diverse semantische Verschiebungen und Modifikationen, infolge derer ein Komplex von mehreren von einer gemeinsamen genetischen „Mutter“ stammenden Wörtern entsteht, die jedoch semantisch voneinander derart entfernt sind, dass Intuition und klassische morphophonologische Dekomposition für die etymologische Analyse allein nicht ausreichend sind. Franz Dornseiff schrieb zwar in seinem Aufsatz aus den dreißiger Jahren des 20. Jahrhunderts selbstbewusst: „Ich will gleich vorausschicken, dass ich das ‚Problem des Bedeutungswandels‘ für Schein halte“ (Dornseiff 1938, 120). In der Tat kann jedoch das Auflösen des Bedeutungswandels große Probleme bereiten, keinesfalls geht es bloß um Schein. Man kann dennoch sagen, dass die Erforschung des Bedeutungswandels und die Rekonstruktion der etymologisch primären Bedeutung heutzutage viel weiter fortgeschritten sind als in der ersten Hälfte des vorigen Jahrhunderts. Ein sehr effizientes Instrument des etymologischen Ansatzes ist ohne Zweifel die Berücksichtigung der Daten des gesamten Wortnestes. Durch den Die vorliegende Studie ist im Rahmen des Projekts Nr. 13-17435S Etymologisches Wörterbuch des Altkirchenslavischen: Abschluss des Generationsprojekts [Etymologický slovník jazyka staroslověnského: završení generačního projektu] entstanden, welches von der Forschungsagentur der Tschechischen Republik finanziell gefördert wird.

etymologie.indb 75

21.1.2014 11:37:22

76 Karlíková

Vergleich der verwandten Sprachdaten in ihrer Gesamtheit, in unterschiedlichen Zeitabschnitten, in allen verwandten Sprachen, die auf eine gemeinsame Wurzel zurückgehen (des Öfteren in unterschiedlichen Ablautstufen), der auf die phonologische, morphematische und semantische Struktur hinzielt, lässt sich ein relativ plastisches Bild der Entwicklungsprozesse innerhalb des jeweiligen Wortnestes ausarbeiten und auf dieser Basis die ursprüngliche etymologische Bedeutung des untersuchten Wortes zuverlässiger zu rekonstruieren. Ein gutes Beispiel eines Wortnestes, das formal sowie semantisch ziemlich unterschiedliche Wörter enthält, bietet das Set der in einigen slavischen Sprachen vorkommenden Ausdrücke, die dem altkirchenslavischen Verb prěkutiti ‘aufputzen, bunt ausschmücken’ entsprechen. Was die Wortform betrifft, ist dieses Verb offensichtlich die Kontinuante des urslavischen Verbs kutiti, dessen Entsprechungen lediglich in einigen Slavinen auftauchen: kirchenslavisch kutiti (Miklosich 1862–65, 325), slovenisch dialektal kutiti se (Pleteršnik 1894–95, i, 491), tschechisch kutit, slovakisch kutiť, polabisch t’autait/t’aitat, obersorbisch skućić, skućeć, polnisch dialektal skutnąć (Boryś–Popowska-Taborska 1994–2010, iii, 130), russisch кутúть. Die Bedeutungen dieser Verben sind unterschiedlich: in den westslavischen Sprachen ‘(vor allem für den Beobachter) unauff ällige Arbeit leisten’, im Kirchenslavischen ‘Ränke schmieden, anzetteln’, im Tschechischen und Slovakischen darüber hinaus auch ‘in Erde buddeln, graben’, in slovenischen Dialekten ‘sich neigen, sich verstecken’, im Russischen ‘schlemmen’. In einigen slavischen Sprachen gibt es auch vom iti-Verb derivierte deverbale Substantive: tschechisch skutek ‘Tat, Handlung’ (vgl. alttschechisch skutiti ‘tun, machen, verüben’, skutiti sě ‘geschehen’), slovakisch skutok mit derselben Bedeutung, polnisch skutek ‘Ergebnis, Folge’. Ausgehend von dem für die Analyse zusammengestellten Minikorpus von relevanten Ausdrücken scheint es notwendig, neben dem urslavischen Verb *kutiti auch das von ihm derivierte Iterativ *kutati bzw. seine sekundär nasalisierte Variante *kǫtati zu rekonstruieren, dessen Kontinuante lexikal-semantisch dem altkirchenslavischen Verb prěkutiti näher stehen: bulgarisch кътам, serbisch/kroatisch kùtati, tschechisch kutat, slovakisch kutať, polnisch dialektal kutać (Karłowicz 1900–11, ii, 539), russisch кýтать, ukrainisch кýтати, weißrussisch dialektal кýтаць (SBrH 2, 590). Auch diese Verben bieten ein breites Bedeutungsspektrum. In süd- und ostslavischen Sprachen ist die Bedeutung ‘verstecken, vermummen’ belegt, das sich mit der oben angegebenen altkirchenslavischen Bedeutung ‘aufputzen, bunt ausschmücken’ leicht in Zusammenhang bringen lässt. Denselben Motivationstyp stellt auch die in den weißrussischen Mundarten belegte Bedeutung ‘umzäunen’ dar.

etymologie.indb 76

21.1.2014 11:37:22

Im Tschechischen und Slovakischen taucht die Bedeutung ‘fangen, (geheim) vorbereiten, unauff ällige Arbeit leisten’ auf (ähnlich auch in ukrainischen Mundarten) und ‘graben, scharren’ (so auch in polnischen Mundarten). Das urslavische Verb *kǫtati ist in einigen Slavinen auch in der präfigierten Form belegt: vgl. z. B. altkirchenslavisch sъkǫtati ‘unterdrücken, beruhigen’, bulgarisch dialektal скъ́там, скýтам, скъ́твам, скýтвам ‘verstecken, eingraben, schmücken, zieren’, ‘verstecken, einpacken, verkleiden’ (SRNG 38, 198–200).¹ Die Mannigfaltigkeit der Bedeutungen der zu dieser Wortfamilie gehörenden Verben ist in diesem Falle derart groß, dass es nach etlichen Untersuchungen nicht möglich ist, sie von einer gemeinsamen indogermanischen Wurzel abzuleiten.² Dies ist auch der Grund, warum die Rekonstruktion der ursprünglichen Bedeutung nicht einfach ist. Aufgrund des Vergleichs der formalen und semantischen Struktur der diskutierten Wörter scheint *(s)keu-t-/*(s)kou-t- als ihre wahrscheinlichste indogermanische Wurzel zu sein. Von den verwandten Ausdrücken außerhalb des slavischen Sprachgebiets kann das altpreußische (mit der nasalisierten Wurzel) -kūnst ‘schützen, hüten’ genannt werden; die Verwandtschaft der anderen in diesem Zusammenhang diskutierten Wörter ist fraglich. Von der Bedeutung ‘verstecken, vermummen’ lassen sich im slavischen Sprachmaterial sowohl die Bedeutungen ‘verhüllen, anziehen’ sowie ‘schmücken (= mit bunten, schmuckhaften Dingen bedecken, verkleiden)’ (vgl. z. B. auch die bulgarischen dialektalen Ausdrücke скъ́там, скýтам, скъ́твам, скýтвам, die sowohl die Bedeutung ‘verstecken’ – darüber hinaus auch die davon ableitbare spezielle Bedeutung ‘beerdigen’ – als auch die Bedeutung ‘schmücken’ nachweisen) als auch die metonymische Bedeutung ‘den Zaun bauen’ ableiten. Aus diesem Blickwinkel betrachtet kann die Bedeutung ‘Ränke schmieden, anzetteln’ (das Kirchenslavische kutiti) als das Ergebnis des Bedeutungswandels verstanden werden, das durch das Schema 1 ausgedrückt wird.

77 Karlíková

Schema 1: ‘verstecken’ → ‘etwas geheim machen’ → ‘Ränke schmieden, anzetteln’ → allgemein: ‘tun’³ Bei der etymologischen Analyse einzelner Wörter, insbesondere bei der Erforschung der Entstehung der Motivierung und des Bedeutungswandels spielt die 1 2 3

etymologie.indb 77

Vgl. ausführlich dazu ESJS 15: 923. Vgl. ausführlich zu den einzelnen Analysen ESJS 12: 705–706, s. v. prěkutiti. Die Bedeutung ‘tun’ ist durch alttschechisch kutiti belegt, und zwar in der Kollokation kutiti zlost ‘Übel tun’ (Gebauer 1903–16, ii, 185).

21.1.2014 11:37:23

78 Karlíková

Suche nach entsprechenden semantischen Parallelen, d. h. die Suche nach dem analogischen semantischen Wandel, eine wichtige Rolle. Stellt sich heraus, dass der beschriebene Bedeutungswandel kein Einzelfall ist, vielmehr dass man ihn auch für andere Ausdrücke des jeweiligen semantischen Feldes voraussetzen kann, gewinnt die etymologische Analyse eine zuverlässigere Basis. Aus dieser Perspektive betrachtet, ist deshalb ein Vorteil, wenn eine konkrete lexikalische Einheit im Rahmen des ganzen semantischen Feldes etymologisch erforscht wird. Dieser methodologische Ansatz erlaubt parallele semantische Veränderungen besser zu untersuchen und ein semantisches Modell des jeweiligen semantischen Feldes aufzubauen. Als gutes Beispiel eines breiten und vielfältigen semantischen Feldes kann das Set der Ausdrücke dienen, das der Terminus verba affectuum bezeichnet. Typisch für dieses Set ist es, dass die Bedeutungen vieler hierher gezählter Ausdrücke sekundär sind, wobei der Ausgangspunkt des semantischen Wandels konkrete Vorgänge und Zustände sind. Die Motivierung des semantischen Wandels ist sehr unterschiedlich; ich beschränke meine Ausführungen auf eine einzige Motivierungsquelle, nämlich auf ‘brennen, in Flammen stehen’. Die indogermanische Wurzel *gȗ her- ‘heiß, warm’ ist im urslavischen Verb *gorěti ‘brennen’ enthalten und von derselben Wurzel ist auch das urslavische Substantiv *gor’e⁴ abgeleitet. Dieses Substantiv hat in einer Reihe slavischer Sprachen Kontinuanten, einschließlich des Altkirchenslavischen (gorje), und besitzt die Bedeutung ‘Weh, Leid, Trauer, Übel’, in einigen Slavinen (slovenisch gorjȇ, alttschechisch hoře, niedersorbisch gorje, pomoranisch gȯř) auch ‘Zorn, Ärger’; alttschechisch hoře weist auch die Bedeutung ‘Schreck’ auf.⁵ Zu erwähnen ist auch das alttschechische denominative Verb hořekovati ‘jammern, klagen’. Kontinuanten von urslavisch *gorěti weisen neben der ursprünglichen Bedeutung auch die übertragene Bedeutung ‘für etwas brennen, für etwas schwärmen’ auf. Im Slovinzischen ist auch die Bedeutung ‘sich ärgern’ belegt. Der Bedeutungswandel ‘brennen, in Flammen stehen’ → ‘Weh, Leid, Trauer’ hat Entsprechungen im Rahmen von Wortfamilien, die von bedeutungsähnlichen indogermanischen Wurzeln gebildet wurden. Zur indogermanischen Wurzel *pekȗ - ‘auf Feuer Essen zubereiten, essbar machen’ bezieht sich das urslavische Verb *pekti und seine Fortsetzungen in allen slavischen Sprachen mit der Bedeutung ‘backen’, in den ostslavischen Sprachen auch mit der Bedeutung ‘sorgen, sich kümmern’. Durch das Suffix -ělь ist von urslavisch 4 5

etymologie.indb 78

Vgl. für detaillierte etymologische Analyse ESJS 3, 190–191. Vokabulář webový [on-line], Version 0.8.0., zitiert am 4. 11. 2013. Erreichbar durch http:// vokabular.ujc.cas.cz.

21.1.2014 11:37:24

*pek- das deverbale Substantiv *pečalь abgeleitet, das in den süd- und ostslavischen Sprachen (bulgarisch, mazedonisch печал, serbisch/kroatisch pȅčal, pèčal, russisch, ukrainisch печáль) mit der Bedeutung ‘Trauer, Wehmut, Leid’ belegt ist; altkirchenslavisch pečalь bedeutet ‘Trübsal, Leiden’ (so im Mazedonischen, Serbischen, Kroatischen und Russischen) und ‘Sorge, Fürsorge’ (auch im Serbischen, Kroatischen und Russischen). Durch das Suffix -ja ist von derselben Wurzel das Substantiv *peča gebildet, mit der Bedeutung ‘Pflege, Fürsorge’ im Tschechischen (péče), Polnischen (piecza), Obersorbischen (pječa). Die zwei zuletzt genannten Wurzeln (*gȗ her- und *pekȗ -) unterlagen dem Bedeutungswandel ‘brennen, in Flammen stehen’ → ’Jammer, Leiden’ lediglich in den Slavinen. Man kann jedoch auch außerhalb der slavischen Sprachen semantische Parallelen feststellen, und zwar in den indoiranischen Sprachen. Das die indogermanische Wurzel *k’euk- ‘aufflammen, erglühen‘ enthaltende vedische Verb śóčati bedeutet ‘glüht, leuchtet, brennt’, sekundär auch ‘leidet’. Das deverbale Substantiv śóka- besitzt einerseits die auf die ursprüngliche Bedeutung der indogermanischen Wurzel ‘Glut, Flamme’ zurückgehende Bedeutung, andererseits auch die neue Bedeutung ‘Schmerz, Trauer’. Zu demselben Bedeutungswandel kam es auch im Neupersischen: Das Substantiv sōg bedeutet ‘Trauer, Kummer’ (vgl. Pokorny 1959–69, 1: 597, Rix 2001, 331). Wie aus dem aufgeführten Sprachmaterial ersichtlich, gehen die Bedeutungen ‘Trübsal, Leiden’ auf das Grundkonzept ‘Schmerz verursachen’, in diesem konkreten Falle ‘durch Brennen’ zurück. Dieselbe Motivierungsquelle ist im diskutierten semantischen Feld auch für weitere Ausdrücke zu suchen; wodurch sie sich unterscheiden, ist lediglich die Ursache des Schmerzes: (a) ‘stechen‘ (indogermanisch *gȗ elH- ‘stechen, quälen’ > litauisch gélti geliù ‘stechen, wehtun (physisch und psychisch)’, althochdeutsch quelan ‘leiden’, altkirchenslavisch žaliti ‘(weh)klagen, betrübt sein’ und die Entsprechungen mit derselben Bedeutung in anderen Slavinen); (b) ‘schneiden, reiben u. ä.’ (indogermanisch *(s)kerbh- ‘schneiden, scharf sein’ > altkirchenslavisch skrъbъ ‘Qual, Pein, Kummer, Trübsal’, skrъběti ‘sich quälen, sich grämen, betrübt sein’ und die Entsprechungen mit derselben Bedeutung in anderen Slavinen, litauisch skur̃bti ‘armselig sein’, dialektal auch ‘sich grämen, kränken’); (c) ‘schlagen, pauken’ (indogermanisch *pleH₂-k/g- ‘schlagen, sich an die Brust schlagen’ > altkirchenslavisch plakati (sę) ‘weinen, klagen’ und Entsprechungen mit derselben Bedeutung in allen Slavinen, lateinisch plangere ‘sich an die Brust schlagen, wehklagen’, plānctus „das mit Wehklagen verbundene Schlagen auf die Brust, auf Arme und Hüften, als Zeichen der

etymologie.indb 79

79 Karlíková

21.1.2014 11:37:24

80 Karlíková

Trauer, die laute Trauer, das Händeringen“⁶, gotisch Perfekt faíflokun ‘sie betrauerten’).⁷ Interimistische Generalisierung, die sich aus den erforschten semantischen Parallelen im Rahmen des Sets verba affectuum ableiten lässt, ist die Rekonstruktion des durch den Bedeutungswandel motivierten semantischen Modells, das das Schema 2 ausdrückt. Schema 2: ‘Schmerz verursachen’ → ‘Trübsal, Leiden’ Dieser Bedeutungswandel widerspiegelt offensichtlich Bestandteile alter Bestattungsrituale, die mit Selbstfolter zusammenhängen, wie z. B. Sich-auf-Brust-Schlagen als Zeichen des Trauerns, Sich-im-Gesicht-Verletzen, Sich-an-Körperteilen-Verletzen, Sich-mit-glühenden-Kohlestücken-Verbrennen usw. (Petleva 1992). Dieser geht von der konkreten Situation aus, in der die Ausdrücke entstanden sind, impliziert die ältesten Bräuche unserer Vorfahren und stellt so einen Bestandteil „des naiven Weltbildes“ dar, wie es von Apresjan (1974, 1994, 2009) beschrieben wird. Zugegeben, es ist möglich, dass wir es hier bloß mit einer Metapher zu tun haben, die auf der Verbindung des Leidens und des physischen Schmerzes beruht (Jakubowicz 2012, 179). Hinsichtlich der hypothetischen Ebene, auf der die Rekonstruktion der etymologisch primären Bedeutung und des nachfolgenden Bedeutungswandels geschieht, ist es jedoch äußerst schwierig, mit Sicherheit zu entscheiden, welche Lösung richtig ist. Die Vorstellung des sprachlichen Weltbildes wird vor allem durch zwei Grundmerkmale konstituiert, und zwar die menschliche Körperlichkeit⁸ und die Fähigkeit der Imagination. Aus dieser Annahme ergibt sich alles andere. Die eben skizzierte Vorstellung ist im Einklang mit der kognitiven Auffassung der Welt, in der – vereinfacht gesagt – die primäre Rolle der Metapher zugeschrieben wird (vgl. Vaňková 2001). Bekanntlich besteht der Sinn der kognitiven Linguistik (dieser linguistischen Orientierung) in der Bemühung, zu verstehen, wie Sprachphänomene durch menschliche Erkenntnis motiviert sind, und zu formulieren, welche die generell gültigen Universalien hierin sind. Man kann sagen, dass ähnliche Prinzipien auch für die etymologische Erforschung gelten. Die Grundannahme ist, dass die Bezeichnung eines Begriffes, einer 6 7 8

etymologie.indb 80

Zitiert nach Georges (1913–18, ii, 1728). Vgl. für ausführliche Erklärung der semantischen Motivationen Karlíková (1998). Wojtyła-Świerzowska (1998, 20) präzisiert dieses relativ allgemein formuliertes Merkmal als „ja – moje ciało – najbliższa udomowiona przestrzeń – mój ród, moja krew“.

21.1.2014 11:37:25

Erscheinung, eines Gegenstandes nicht in einem isolierten Raum entstanden ist, dass ihre Entstehung damit zusammenhing, auf welche Art und Weise ein Einzelwesen die Umwelt wahrnahm und bewerte, in welchem Milieu es sich befand. „Die uralte Motivierung, die sich die Etymologie zu enthüllen bemüht, zeigt uns den Menschen, der nicht die Dinge an sich, sondern nur in Beziehung auf ihn selbst, das heißt in konkreten Lebenssituationen, Arbeitsprozessen usw., wahrnimmt“ (Havlová 1997, 523–524). Der Etymologe versucht, mit Hilfe von konkreten, in die jeweiligen Sprachstrukturen verankerten Bildern allgemeinere Schlüsse zu formulieren und das Etymon zu rekonstruieren. Bei der Rekonstruktion der ursprünglichen Bedeutung des Etymons und der semantischen Veränderungen seiner Kontinuanten ist es – wie bereits betont wurde – immer wichtig, die analysierte lexikalische Einheit in den Kontext der Ausdrücke desselben semantischen Feldes einzuordnen. Im Gegensatz zur kognitiven Linguistik formuliert die Etymologie weder Universalien noch arbeitet sie mit qualifizierenden Schemata (englisch image schemas) wie innen × außen, zahlbar × nicht zahlbar, nahe × fern, oben × unten, klein × groß u. ä. Auf der anderen Seite gehört zum langfristigen Programm der etymologischen Forschung die Ausarbeitung der semantischen Parallelen, die universal für die etymologischen Analysen benutzt werden können. Nicht unwichtig ist die Tatsache, dass das Sprachmaterial, das kognitive Linguistik für die einzelnen semantischen Felder zusammenstellt, mit dem Sprachmaterial vergleichbar ist, das der etymologischen Analyse dient. Es lässt sich folglich sagen, dass die beiden linguistischen Disziplinen in dieser Hinsicht einander nahe stehen und sich gegenseitig durch ihre Forschungs(zwischen)ergebnisse bereichern können. Ein gutes Beispiel für die Darstellung, wie Motivierungsausgangspunkte in einem semantischen Feld für die etymologische Erforschung unklarer Wörter nützlich sein können, liefert das semantische Feld der Ausdrücke, die mit dem Konzept ‘groß’ assoziiert sind (Karlíková 2012). Was die Lexik sowie Motivierung der Bedeutung betrifft, ist dieses Feld relativ vielfältig. Das Alttschechische unterstützt diese Behauptung, denn es verfügt über eine große Menge von Adjektiven, die die Größe von (konkreten sowie abstrakten) Gegenständen bezeichnen. Die Herkunft einiger Ausdrücke ist unklar, z. B. die Adjektive holemý und horutný. Es gibt mehrere Ausgangsbasen der Motivierung der Bedeutung ‘groß’, eine davon ist durch das Schema 3 repräsentiert.

81 Karlíková

Schema 3: ‘stark, mächtig (sein)’ → ‘groß (sein)’ Hierher gehören u. a. die alttschechischen Adjektive mohútný, mohutný ‘groß, stark’, ‘mächtig’, im modernen Tschechischen ‘große Ausmaße habend, riesig,

etymologie.indb 81

21.1.2014 11:37:25

82 Karlíková

etymologie.indb 82

mächtig, schön gewachsen’ u. a. Bedeutungsähnlich ist slovakisch mohutný. Was die Wortbildung angeht, so sind entsprechende Adjektive auch in einigen anderen Slavinen zu finden. Die Bedeutung ‘sehr groß’ belegt jedoch lediglich weißrussisch могутны, in den anderen Sprachen behalten diese Adjektive im Wesentlichen die primäre Bedeutung ‘mächtig, stark’. Derivatologisch handelt es sich um die Form des alten Partizips praesentis activi *mogǫt- des urslavischen Verbs *mogti ‘können’ bzw. um seine adjektivisierte Form mit dem Suffix -nъ, vgl. kirchenslavisch mogǫtьnъ ‘potens’ (Miklosich 1862–65, 378). Die indogermanische Wurzel, die sich als *magh- ‘können, imstande sein’ rekonstruieren lässt, ist auch z. B. in litauisch magė́ti magù ‘gefallen’ enthalten (Rix 2001, 422). Im Gegensatz zum Adjektiv mohútný, mohutný, das eine relativ klare Herkunft hat, ist die Herkunft von alt- und mitteltschechisch holemý verborgen. Man kann Äquivalente auch in anderen slavischen Sprachen finden, in einigen davon sogar mit derselben Bedeutung: kirchenslavisch golěmъ (im Altkirchenslavischen gibt es lediglich das Adverb golěmo ‘viel’ belegt), bulgarisch голям, mazedonisch голем, serbisch/kroatisch gòlem, slovenisch dialektal hólməš (< *golěmьši in der Funktion des Elativs, vgl. Bezlaj 1973, 78), polnisch archaisch (seit dem 16. Jh.) golemy u. a. Die urslavische Ausgangsbasis lässt sich aufgrund des Vergleichs aller verwandten slavischen Wörter als *golěmъ rekonstruieren. Die interne morphologische Struktur des urslavischen Adjektivs golěmъ, namentlich das Suffix -m-, deutet darauf hin, dass es sich um die Form des Partizips preasentis passivi handelt (im modernen Tschechischen ist die ursprüngliche m-Partizipialform nur in einigen Adjektiven erhalten, z. B. známý, vědomý, kradmý). Im Litauischen entspricht ihr das Partizip praesentis passivi in der Form gãlimas ‘möglich’. Die auf den ersten Blick klare Situation ist jedoch dadurch kompliziert, dass in den slavischen Sprachen das Verb (*golěti), von dem das genannte Partizip abgeleitet sein soll, nicht belegt ist. Dies ist jedoch selbstverständlich kein Grund, um daraus den Schluss zu ziehen, dass das Verb nicht existiert hat, und die Tatsache, dass es im Litauischen das – dem nicht belegten urslavischen Verb – genau entsprechende Verb galė́ti ‘über etwas Macht bekommen’ gibt, liefert eine indirekte, jedoch robuste Unterstützung für die Annahme der Existenz dieses Verbs auch im Urslavischen. Auch Rix (2001, 185) schließt übrigens sein Vorkommen nicht aus. Die indogermanische Wurzel lässt sich als *gelH- ‘Macht bekommen über etwas’ rekonstruieren, die auch in irisch gal ‘Mut’ und kymrisch gallu ‘können, fähig sein’ enthalten ist (ESJS 3, 186). Für diese Analyse sprechen darüber hinaus die Ausdrücke mit parallelem Bedeutungswandel, die sich im Rahmen dieses semantischen Feldes finden lassen.

21.1.2014 11:37:25

Ein anderes alttschechisches Adjektiv mit der Bedeutung ‘(ganz) groß’ ist horutný. Dieses Adjektiv ist lediglich im Alt- und Mitteltschechischen belegt; im modernen Tschechisch ist es nicht erhalten geblieben. In alttschechischen Belegen taucht es im Kontext mit den Substantiven blázen ‘Narr’, bláznovstvie ‘Narrheit’, křivda ‘Unrecht’, vymyšlenie ‘Erdichtung’ auf. Im Mitteltschechischen findet man es in der Kollokation mit dem Substantiv klam ‘Täuschung’. Die Form dieses Adjektivs entspricht dem alten mit dem Suffix ǫt- derivierten Partizip (dieser Bildungsmuster ist im Rahmen der Ausdrücke mit der Bedeutung ‘groß’ nicht unüblich, vgl. das bereits genannte mohutný), das wahrscheinlich vom Verb *gorěti ‘brennen, flammen’ (< *gȗ her- ‘heiß’) abgeleitet ist. Die Motivationsbasis lässt sich durch das Schema 4 darstellen:

83 Karlíková

Schema 4: ‘für etwas sehr entflammt sein’ → ‘Aufregung, Emotionen erweckend (durch das Aussehen, Größe)’ → ‘groß sein’ Wie ersichtlich lassen sich im Rahmen dieses semantischen Feldes beobachten einerseits die parallelen semantischen Verschiebungen, die zur Entstehung der Bedeutung ‘groß’ führ(t)en, andererseits die parallelen Wortbildungsprozesse, mit denen die Entstehung dieser Bedeutung verbunden ist. Fazit: Die Antwort auf die am Anfang dieses Aufsatzes gestellte Frage ist nicht einfach. Zweifelsohne ist die kognitive Komponente im etymologischen Ansatz sehr wichtig, ja bei der Rekonstruktion der ursprünglichen Bedeutung und dem nachfolgenden Bedeutungswandel wohl unentbehrlich. Die obigen Ausführungen sollten zeigen, dass sich gemeinsame Momente finden lassen, in denen sich die Etymologie und die kognitive bzw. kognitiv-kulturologische Linguistik treffen. Das Herangehen, dessen sich die beiden Disziplinen bedienen, ist jedoch notwendigerweise unterschiedlich, denn die unterschiedlichen Forschungsmethoden dienen unterschiedlichen Forschungsvorhaben der beiden Disziplinen. Während in der kognitiven Linguistik bei der Beschreibung des sprachlichen Weltbildes die lexikalische Semantik (mit nicht geringem Gewicht auf den philosophisch-psychologischen Aspekten) im Mittelpunkt der Interesse steht, ist die Hauptaufgabe der Etymologie die Rekonstruktion des Etymons. Anders ausgedrückt: die Rekonstruktion der etymologisch primären Form bleibt mit der Rekonstruktion der primären Bedeutung und ihrer nachfolgenden Entwicklung untrennbar verbunden. Was gemeinsam ist, ist das Vorhaben, bei der Analyse des lexikalischen Materials die maximale Komplexität zu erreichen.

etymologie.indb 83

21.1.2014 11:37:26

84

Literaturverzeichnis

Karlíková Apresjan, Ju. D. [Апресян, Ю. Д.] 1974. Лексическая семантика. Синонимические средства языка. Москва. Apresjan, Ju. D. [Апресян, Ю. Д.] 1994. Naiwny obraz świata a leksykografia. Etnolingwistyka 6, 5–12. Apresjan, Ju. D. [Апресян, Ю. Д.] 2009. Исследования по семантике и лексикографии, i: Парадигматика. Москва. Bezlaj, F. 1973. Arhaizmi v koroških narečjih. In: Koroški kulturni dnevi. i. Maribor, 72–81. – zitiert nach F. Bezlaj: Zbrani jezikoslovni spisi. i. Hrsg. von M. Furlan. Ljubljana 2003, 494–500. Boryś, W. – Popowska-Taborska, H. 1994–2010. Słownik etymologiczny kaszubszczyzny. i–vi. Warszawa. Dornseiff, F. 1938. Das „Problem des Bedeutungswandels“. Zeitschrift für Deutsche Philologie 63, 119–138. ESJS = Etymologický slovník jazyka staroslověnského. Praha, Brno 1989–. Gebauer, J. 1903–16. Slovník staročeský. i–ii. Praha. Georges, K. E. 1913–18. Ausführliches lateinisch-deutsches Handwörterbuch. i–ii. 8. Aufl. von H. Georges. Leipzig. Havlová, E. 1997. Die komplexe Situation als Motivierungselement bei urslavischer Wortschöpfung. Slavia 66, 523–524. Jakubowicz, M. 2012. Badania etnolingwistyczne a etymologia. Etnolingwistyka 24, 173–181. Karlíková, H. 1998. Typy a původ sémantických změn výrazů pro pojmenování citových stavů a jejich projevů ve slovanských jazycích. Slavia 67, 49–56. Karlíková, H. 2012. Vyjádření pojmu „velký“ ve staré češtině a jeho ekvivalenty ve slovanských jazycích. In: Praslovanska dialektizacija v luči etimoloških raziskav. Ob stoti obletnici rojstva akademika Franceta Bezlaja. Zbornik referatov z mednarodnega znanstvenega simpozija v Ljubljani, 16.–18. septembra 2010. Hrsg. von M. Furlan und A. Šivic-Dular. Ljubljana, 125–133. Karłowicz, J. 1900–11. Słownik gwar polskich. i–vi. Kraków. Miklosich, F. 1862–65. Lexicon palaeoslovenico-graeco-latinum. Vindobonae. Petleva, I. P. [Петлева, И. П.] 1992. Этимологические заметки по славянской лексике. xvii. In: Этимология 1988–1990. Москва, 52–57. Pleteršnik, M. 1894–95. Slovensko-nemški slovar. i–ii. Ljubljana. Pokorny, J. 1959–69. Indogermanisches etymologisches Wörterbuch. i–ii. Bern – München. Rix, H. (Hrsg.) 2001. Lexikon der indogermanischen Verben. Die Wurzeln und ihre Primärstammbildungen. 2. Aufl. Wiesbaden. SBrH = Слоўнік беларускіх гаворак паўночна-заходняй Беларусі і яе пагранічча. i–v. Мiнск 1979–1986. SRNG = Словарь русских народных говоров. Москва – Ленинград, Санкт-Петербург 1965–. Vaňková, I. 2001. Obraz světa v mateřském jazyce. In: Obraz světa v jazyce. Hrsg. von I. Vaňková. Praha, 19–28. Wojtyła-Świerzowska, M. 1998. Kognitywizm w etymologii. Rocznik Slawistyczny 51, 17–30.

etymologie.indb 84

21.1.2014 11:37:26

марина м. валенцова этнолингвистический комментарий к этимологии слов мара и упырь

85 Валенцова

Собирая в поисках этимона как можно более полный список реализаций праславянского корня во всех славянских языках (а также генетически родственных балтийских, германских, иранских), уделяя особое внимание семантике отобранных слов, этимологи уже давно, сначала спонтанно, а потом целенаправленно начали обращать внимание на этнокультурный контекст исследуемых слов. Как подчеркивали в докладе на viii съезде славистов Н. И. и С. М. Толстые, «культурно-исторический контекст, в котором реально функционирует или функционировало слово, становится дополнительным источником, а в ряде случаев и решающим критерием реконструкции исходного значения слова и направления его семантической эволюции. Перспективность такого подхода в этимологии, в особенности для архаического пласта мифологической лексики, после ряда недавних работ не нуждается в доказательстве» (Толстой–Толстая 1978: 365). С тех пор появилось множество других работ, использующих подобный подход, в первую очередь, труды О. Н. Трубачева, В. Н. Топорова, Вяч. Вс. Иванова (см., например: Трубачев 2004, 2008, 2009, Иванов–Топоров 1974, Иванов 1983, Топоров 2004) и ряда других лингвистов. В свою очередь и этнолингвистика немыслима без этимологии. Занимаясь в числе прочего поиском «этимологии верований», мотивировок архаичных представлений, тех ментальных связей между вещами, которые стали культурными константами, – то есть реконструкцией прасмыслов, прамотивировок, прасимволов¹, этнолингвисты пользуются также этимологической реконструкцией праслова (праформы) и его празначения, поскольку, по словам О. Н. Трубачева, «… именно этимология уполномочена дописывать историю лексических значений и реконструировать дописьменный период этой истории …» (Трубачев 2004, 1: 124). Этимология привлекается для выяснения культурной семантики слов, символики реалий традиционной духовной культуры – на основе их названий и названий обрядовых действий, которые с ними совершаются. 1

etymologie.indb 85

О проблемах реконструкции древней славянской духовной культуры и семантической реконструкции культурной лексики см.: Толстой–Толстая (1978), Толстой (1979, 1985, 1989), Толстая (2008: 176–225) и др.

21.1.2014 11:37:27

86 Валенцова

Примером такого подхода могут служить работы Н. И. и С. М. Толстых по этнолингвистике (Толстой–Толстая 1993, Толстой 1997: 306–438, 1995: 364–372, 383–428, Толстая 2006, 2007, 2008: 50–174, 275–332, 347–466 и др.), исследования их коллег и учеников (Журавлев 2005, Плотникова 1994, 2004, 2006, Седакова 2003, Усачева 1977 и др.), других авторов, работающих в этом направлении (Хобзей 2002, Антропов 1998, 2004, 2013, Березович 2007, 2010, Зубов 1982, 2010, Климова 2008 и др.). К областям, где пересекаются интересы этимологии и этнолингвистики, относится мифология – одна из наиболее древних систем познания и оценки мира. Мифологическая лексика давно привлекает внимание ученых, в том числе с точки зрения этимологического анализа. Некоторые термины получили убедительную этимологию (например, raroh, rarach – см. Трубачев 1967, там же (с. 64–65) подробнee о результатах изучения этого слова В. Махеком)², этимология других слов до сих пор точно не установлена. Не существует единого мнения о том, следует ли различать корни mar- и mor-, или они восходят к одному этимону, и к какому именно (см. ЭССЯ 17: 206–207; ЭССЯ 19: 214, Dukova 1983: 30–36), есть две этимологии мифонима волколак, волкодлак³; многочисленны попытки этимологизировать название мифологического персонажа упырь/опырь/упир (см. об этом ниже); продолжают появляться этимологии имени з.-слав. домового и духа-обогатителя словац. škriatok, чеш. skřítek, пол. skrzytek⁴. В данной статье предложены этнолингвистические или этнокультурные комментарии к двум терминам народной демонологии, которые не имеют общепризнанной этимологии. Этнолингвистические данные могут быть использованы для аргументации в пользу той или иной этимологической версии упомянутых слов. 2

3

4

etymologie.indb 86

«Западнославянское rarogъ было, бесспорно, одним из таких элементов, никогда не известным прочим славянам, и мы трактуем его как праслав. диал. *rarogъ» (Трубачев 1967: 66). Традиционная этимология: vlkodlak < *vlk ‘волк’ + *dlaka ‘шкура’ – и выдвинутая и разработанная В. Н. Топоровым и В. В. Ивановым, предполагающая сложение слов со значением ‘волк’ + ‘медведь’ (см. Топоров 1984, Иванов 1975, 1983, 2007, 2008). По одной из версий слово происходит из нем. Schrat ‘злой лесной человечек’. Другая на основании лингвистического (зап.-морав. skřitkovať ‘копошиться, делать что-либо украдкой, как škřítek’, укр. ховати, виховувати ‘воспитывать’ и ‘скрывать, прятать’, словац. chovať ‘содержать, выкармливать, заботиться’) и этнографического (укр. хованец также высиживается из яйца черной курицы и т.п.) материала предполагает связь имени skřítek с чеш. krýt ‘выводить, заботиться’ и ‘скрывать’ (Зубов 1982: 138, 2010: 118). С точки зрения реализации значения ‘скрываться, укрываться’ показательно поверье, записанное Ф. Бартошем, что střítek – это чертик, который скрывается от грома (молнии) и прячется у человека за пазухой (skřítek, ten čertík, před hromem sa krýje a krýje, a fuk člověkovi pod pažu, a hrom ho zabije (цит. по: Зайцева 1975: 107), хотя здесь не исключено влияние «народной этимологии».

21.1.2014 11:37:27

1. Мара – мора

87 Валенцова

По данным ЭССЯ, *mara – «трудное слово. Наиболее известны две различные его интерпретации. Согласно одной, *mara (русск., польск.) толкуется как аблаут к *mora (см. цслав. мора ‘maga’ и другие слав. соответствия со значениями ‘кошмар, удушье; колдунья, ведьма’), далее – к др.-исл., др.-англ., др.-в.-нем. mara ‘кошмар, удушье’ (так I. Franck , против Solmsen ). Шустер-Шевц также считает *mara , интерпретируемое как *mōra к *mŏra, восходящим к гнезду *merti, *morъ и имеющим те же германские соответствия . Ф. Славский отвергает предлагаемую Младеновым версию относительно связи болг. марá ‘налягане на сън’ с морá (к праслав. *merti, *morъ), считая ее семантически неправдоподобной . Однако преобладающей в настоящее время, по-видимому, является другая гипотеза, согласно которой *mara считается производным с суф. *-rā- от того же корня *ma-, от которого образованы *mamъ (см.), *mamiti (см.); *manъ (см.), *maniti (см.) ( персонаж, вызывающий эту болезнь). Основная сема в значении конкретных персонажей – ‘смерть’⁵. Плодотворность привлечения этнографического материала при исследовании мифонимов можно проиллюстрировать примером описания в ЭССЯ однокоренного с mara/mora имени Марены/Морены, чучела, куклы, наряжаемой и уничтожаемой на масленицу или Ивана Купалу. Варианты ее имени объединены в одну статью *marěna I/*morěna; для них «наиболее вероятным представляется толкование в связи с слав. гнездом *mer- ‘умирать’, см. *merti, *mьrǫ, *morъ ‘умирать; смерть, мор’ (< и.-е. *mer-, merə- ‘умирать’), причем, очевидно, *morěna > *marěna (o > a, удлинение). Показательно, что в чешских говорах синонимами слов mařena, mořena служат smrť, smrtnica, smrtelnička, smrtolka, smrtka см. еще польск. диал. mór = marzana …» (ЭССЯ 17: 210). Далее говорится, что исходя из се5

etymologie.indb 88

В том числе для рус. кикиморы, которая «прядет» оставленную хозяйкой без благословения пряжу, но не прядет, а только рвет нитки, пачкает и оплевывает работу. В мифологическом смысле нить – символ человеческой жизни, полотно – символ жизненного пути, одежда – символ и заместитель в обрядах самого человека; поэтому поговорка «От кикиморы рубахи не дождешься» вполне вписывается в мифологическую символику смерти.

21.1.2014 11:37:29

мантики обряда, такое объяснение естественно, поскольку кукла представляет собой символ зимы и смерти. Учитывая результаты последних исследований М. Белетич и А. Ломы и этимологическую трактовку имени *marěna/morěna, есть основания принять и для слов *mara/*mora этимологию, возводящую их к единому этимону. Вместе с тем интересные данные дает география распространения терминов с корневыми гласными o и a⁶. Вырисовываются отчетливые ареалы распределения корневых реализаций. На восточнославянской территории в украинских, белорусских и большой части русских диалектов представлены почти исключительно реализации корня mar- (см., например, ЭССЯ 17: 205–206, Власова 1995: 235–240, Новичкова 1995: 359–361), mor- становится заметен лишь в русских говорах, главным образом, северных и восточных. На западнославянской территории преобладает корень mor- с вариантом дальнейшего сужения гласного – mur-/můr- (в зап.-словац., морав., чеш. и серболуж. диалектах) и «переходных» вариантов типа mŭor- в пол. говорах (ЭССЯ 19: 212–213, Budziszewska 1991: 17–18). В южной Славии, по данным ЭССЯ (19: 212) и др. источникам, представлен корень mor-, кроме болгарских диалектов, где распространены оба корня, а также mur- (также см. Дукова 1980, Dukova 1983: 30–36). На обобщенной и не претендующей на точность карте-схеме (см. ниже) показаны ареалы преимущественного распространения лексем с корнями mor-/mar- в разных значениях (картина станет еще более выразительной с учетом данных окружающих неславянских языков). Конечно, на территории преимущественного распространения одного корня фиксируются и слова с другим корнем. Например, в нижнелужицком зафиксировано также и mara, а в укр. мора (оба без указания места), корень mar- изредка встречается в чешском; в русском представлена также мора (твер., сарат., арх. енисей. обл. – СРНГ 18: 254) и достаточно широко 6

etymologie.indb 89

89 Валенцова

«Лингвистическая география сказала очень много нового и ценного для этимологии, выдвинула ряд положений, значение которых трудно переоценить: каждое слово представляет собой индивидуальный в известном смысле продукт истории развития и географического распределения форм; соотношение форм, которое застает в определенный момент лингвист-наблюдатель, как правило, чуждо какой бы то ни было случайности, но объяснимо в свете данных истории, культурных влияний и взаимоотношений форм между собой; границы диалектов и вообще языковых территорий имеют относительное значение и не могут служить препятствием для распространения общих слов и форм, отсюда следует важный вывод о необходимости изучения внешней взаимозависимости лингвистических систем» (Трубачев 2004, 1: 212).

21.1.2014 11:37:29

90 Валенцова

известно слово кикимора (никол., вологод., новгор., перм., уральск., олон., тобол., орл., рязан., сарат., сиб., яросл., сев.-двин., вят., твер. – СРНГ 13: 205). На основании доступного материала не всегда можно судить о частотности и  весе этих слов в каждой конкретной традиции. Часть таких «исключений» поддается объяснению с  учетом диахронии, например: mor- в северных и восточно-русских говорах – миграционными потоками заселения с юго-запада⁷; смешение обоих корней в болгарском можно объяснить миграцией на Балканы славян – носителей различных диалектов (давно известны украинско-болгарские и украинско-сербские изоглоссы⁸, в настоящее время исследования карпато-балканских схождений продолжают разрабатываться на разных языковых уровнях⁹). Переходная зона, разделяющая ареалы корней mar- и mor на территории современной Словакии располагается полосой с северо-востока на юго-запад. В восточной Словакии и на юге среднесловацкой области в качестве имени мифологического персонажа представлен термин mara¹⁰: в Земплине (Собранце, Гуменне – SSN 2: 188), на Спише (окр. Попрад – Michálek 1989: 283, SSN 2: 188), Верхнем Спише (окр. Стара Любовня – AT ÚEt, inv. č.790/А), в Замагурье (Blagoeva-Neumanová 1976: 115–116, собственные записи), в Гемере-Малогонте (GM: 371), Текове (окр. Кремница – SSN 2: 188); Погронье (окр. Зволен – SSN 2: 188). 7

См., напр.: Васильев (2013), где на основе анализа севернорусской топонимии и гидронимии обосновывается версия о направлении заселения Северно-Западной Руси из Галицко-Волынских земель через Полесье и Верхнее Поднепровье (с. 22–25, 28); «сравнительно поздняя русская топонимия Русского Северо-Запада» локализуется «в том числе на землях позднего русского заселения … вплоть до Урала и Сибири», Среднего Поволжья, юга Европейской России (с. 29). 8 Обзор истории и проблематики карпатско-южнославняских исследований до 1984 г. см. Нимчук (1988). 9 См., напр., сборники: КБДЛ, КБДЛ 2. 10 Для наших целей в качестве географических помет достаточно указание на крупный исторический регион; перечень лексем и источников не претендует на полноту и является, скорее, экземплификационным; подробнее о географии морa/марa см.: Валенцова (2013: 318–321).

etymologie.indb 90

21.1.2014 11:37:30

91 Валенцова

Для Западной и северной части Средней Словакии характерны термины mora и mura, напр.: mora, mura (Тренчин, Мыява, Кисуце – SSN 2: 188, AT ÚEt, č. 578В), nočňí mura (Загорье, Скалица – SSN 2: 188); mora (Поважье – окр. Поважска Быстрица – SSN 2: 188), mora (Кисуце, окр. Чадца – AT ÚEt, č. 578В), mora (Орава, окр. Дольный Кубин, окр. Трстена – SSN 2: 188, собственные записи), mora, mura (Липтов окр. Липт. Микулаш, окр. Ружомберок – SSN 2: 188, собственные записи), mora, zmora (Зволен, окр. Б. Быстрица – SSN 2: 188), mora, zmora (Верхнее Погронье – Horehronie: 330, AT ÚEt, inv.č. 763, 168, собственные записи), mora (Гемер-Малогонт – GM: 371), mora (Верхний Спиш, окр. Стара Любовня – AT ÚEt, inv.č. 790), umornica (Спиш – Michálek 1989: 283), mura (Новоград, окр. Лученец – SSN 2: 188). Смешение имен mara/mora наблюдается в переходном поясе, в частности, в Замагурье, на Верхнем Спише, Верхнем Погронье, в области Гемер-Малогонт. С таким распределением корней mor-/mar- на территории Словакии согласуется и карта, составленная В. Важным (Vážný 1955, mapa č. 3) для энтомологических названий с этими корнями: названия бабочек с корнями mor- и mur- (mora, mor, morka, zmora, saňimorka, ratimorská pani, múra, mura, mur (m.), kanimura) широко представлены в Западной (Загорье, Мыява, Поважье, Понитрие) и в Средней Словакии, особенно в ее северной части (Липтов, Погронье, Турьец), единичные названия – в Гемере (на среднесловацком юге). С корневым а зафиксировано только одно название бабочки – slepá mara – в трех пунктах: в р-не Банской Быстрицы, Зволена и Жарновицы (в стóлицах Зволенской и Тековской), то есть в «зоне

etymologie.indb 91

21.1.2014 11:37:32

92 Валенцова

смешения». В этом же переходном поясе зафиксировано и название бабочки mura (Спишске Теплице, Летановце – сев. часть Спиша). Проблема распределения терминов *mara и *mora на славянской территории пока остается нерешенной¹¹. 2. Упырь – вампир Этимологические словари славянских языков не предлагают какой-либо определенной версии этимологии слова упырь, перечисляя лишь существующие гипотезы.¹² Трудность этимологизации связана с большим количеством вариантов слова, в котором варьируют практически все фонемы, ср.: рус. упырь, упирь (Фасмер 1986–87, 4: 165); др-рус. упирь, опыръ, впыръ, упыръ (Преображенский 1910–14: 64), укр. упир, опир, опир’, опиер’, опер’, пiр, упир’, упiр, ўпир’, вопир (Хобзей 2002: 142, 145-147), блр. вýпар (ЭСБМ 2: 225), упiр, чеш. upír, слц. upír, пол. upiór, wypiór, ст.-пол. upierz, пол. диал. lupiór, wypiór (Machek 1968: 669), wąpierz (Baranowski 1981: 51), кашуб. łupi, łåṕi, oṕi (Sychta 3: 30, 330), болг. въпир, вепир, вапир, ляпир (БЕР 1: 117), вампыр, вепыр, в1пыр (Геров: 105), вапирин, въпер, вопер, випир, йепир, влепир(ин), лепир(ин), лампир, лемпир, лемптир, липир, липирь, липирин, упир (Dukova 1985: 39–40), с.-х. устар. upir, upirina. Существует более десятка этимологий термина упырь. Практически все исследователи этого слова сходятся в том, что оно образовано при помощи префикса; чаще всего префикс этимологизируют как ǫ-. Основные версии этимологии корня основаны на разных признаках, приписываемых упырю: *pyr- ‘огонь, жар’ и реконструкция первичного значения как «несожженный» (покойник); *per- ‘летать’ и сближение слов упырь и нетопырь, где префиксу у- (< ǫ-) приписывается отрицательное значение («не-птица»); *pěr- ‘втыкать, вонзать’ и попытка мотивировать название упыря его свойством «впиваться, пронзать» жертву, чтобы выпить 11 Возможно, следовало бы учесть также возможность влияния на распространение корней mar/mor- у славян и в Восточной Европе в целом буддийского образа Мары (санскр. и пали māra ‘убивающий, уничтожающий’), западно-европейской Мары, злого духа, воплощения ночного кошмара (англ. nightmare, фр. couchemar) (МНМ 2: 109–110) и греч. μώρρα ‘einer der Namen des kindertötenden Gespenstes Gello’ (Dukova 1983: 30). 12 См. Фасмер (1986–87, 4: 165), Преображенский (1910–14: 64), Аникин (2012: 42–45), Holub–Kopečný (1952: 403); Machek (1968: 669), Brückner (1927: 594), ЕСУМ 6: 38–39, ЭСБМ 8: 225, Skok (1971–74, 3: 564) и др. Й. Рейзек, приводя возможные праформы *ǫpyrь, *ǫpirь, *upirь и др., высказался определенно: «Неясное. Разнообразие форм и неясная мотивация наименования не дают возможность предложить убедительное объяснение» (Rejzek 2001: 692).

etymologie.indb 92

21.1.2014 11:37:33

ее кровь (хотя возможно было бы и обратное направление мотивирующего действия – пронзание колом тела самого вампира, т.е. нарушение целостности трупа, что препятствовало бы его «хождению»); предлагалась также версия заимствования корня из тюркских языков (башкир., татaр. убыр ‘злой дух, ведьма, демон’ и др.). Литература об упыре, и этимологическая, и этнографическая, огромна.¹³ Вместе с тем, признаки, на которых построены известные этимологии, не являются характерными только для данного персонажа. Сжигание трупа вампира как последнее и радикальное средство избавления от его вредоносного воздействия (после предшествующих: обсыпание маком, вкладывание в рот монеты или камня, вынос гроба через окно, пробивание груди колом, подрезание жил на ногах, переворачивание вниз лицом и т.п.), не специфично для борьбы с  его «хождениями», это универсальный способ уничтожения опасных или нежелательных реалий и явлений – см. статью Уничтожение ритуальное (СД 5: 371–373). Связь вампира с птицей вообще чрезвычайно редка: одно из таких немногочисленных свидетельств, приведенное ниже, показывает, что эта связь формальная и типологически принадлежит к другой сфере: это мотив опасности зрения (и голоса) «заложных» покойников и вообще нечистой силы¹⁴; требование сюжета «заставляет» вампира забираться на колокольню или превращаться в птицу, чтобы с высоты как можно дальше увидеть (или «покрыть» голосом) и тем самым убить больше людей. Мотив полета души ведьмы (или двоедушника, моры, покойника и под.) в виде птицы или бабочки¹⁵ не может быть использован в качестве семантического аргумента, поскольку направление мотивации в данном случае обратное: энтомологические названия сами мотивированы именами мифологических персонажей (см. Vážný 1955). Обширный корпус этнолингвистических данных дает возможность предложить еще одну гипотезу происхождения названия персонажа упырь, а именно: возможность возводить его к глаголу *piti ‘пить’ < *pōi-/ *pī- (Фасмер 1986–87, 3: 269) с суффиксом -rъ (аналогично, например, *žir< *ži-ti, *mir- < *mei-/mī; *mar- < *ma-niti, *ma-miti; или аналогично бел. павадыр, рус. поводырь, укр. поводирь < povoditi + yr’ (ЭСБМ 8: 87).

93 Валенцова

13 См., напр., СД 1: 281–286 (там лит.), Спировска–Вражиновски (1988), Плотникова (2004: 212–218, 634–645), Новичкова (1995, 547–551), EĽKS: 277–278 и др. 14 Ср.: “Najwięcej ludzi umiera, gdy za sprawą upiora zwonё same jidą Sy iii 330, iv 25 (por. głuż. zwony du ‘dzwonia’ Trof 455), ponieważ jak daleko słychać ich głos, tak daleko ludzie umierają” (Treder 1989: 85). 15 Ср., напр.: болг. мора, душа ожившего нечистого мертвеца, летает в виде птицы и бабочки, мучит спящих, высасывает кровь, как вампир (БМ: 219).

etymologie.indb 93

21.1.2014 11:37:33

94 Валенцова

etymologie.indb 94

Упырь – нечистый покойник, встающий из могилы и сосущий кровь, молоко у людей и животных, сок из деревьев, то есть жидкую субстанцию, являющуюся и реально, и символически их жизненной силой. Эта субстанция дает вампиру дополнительную жизнь, в то время как донор чахнет, сохнет и умирает. Ср. текст, записанный в восточной Словакии: Ked’ telo nezmeravie po smrti, mrtvý chodí dobytok piť. Vracajúci sa mŕtvy cicia krv dobytku, takže tento zdochne. Keď sa premení na vtáka, koľko dedín “zapatrí” (zazre), všetky sú jeho a spíja všetko (Vyšné Revište, okr. Sobrance – AT ÚEt č. 50, zap. J. Mjartan, 1949). В основе образа упыря лежит один из наиболее древних концептов мировосприятия: магические действия отбирания, кражи чужого «благополучия» (сюда входят: здоровье, богатство, включая урожай на полях и приплод скота, счастье, то есть любовь и удача, и сама жизнь) часто кодируется глаголами пить, выпивать, высасывать (см. СД 4: 56–57): ведьма тянет, высасывает (= отбирает) молоко у коров (в.-слав.), в том числе собирая до рассвета полотном росу с чужого поля, затем давая выпить своей корове; у выдаиваемой ведьмой коровы пересыхает вымя, она чахнет и погибает (рус. – Власова 1995: 73); ю.-слав. воздушный змей ламя высасывает воду из озер, вызывая засуху (болг.), у сербов дракон аджаjа, приводящий тучи и град, защищает свое село от града, если его поить молоком (Плотникова 2004: 670), он же выпивает урожай (макед. – там же: 231); молоко у рожениц пропадает, потому что его выпивают вредоносные демоны (ю.-слав., з.-слав.); болг. летающий змей может выпить жизненную силу девушки или женщины (ассоциирующуюся с кровью), отчего после его посещений она чахнет, сохнет, бледнеет и умирает; души умерших супругов могут приходить к оставшимся в живых и иссушать их до смерти; по южнославянским поверьям, души некрещеных детей (навjе, навће) выпивают кровь у новорожденных и кровь и молоко рожениц («ее выпили навjата» – там же: 241, 690) и под. Ср. также сохранившееся в языке выражение нежелательного или принудительного лишения человека силы: рус. выпить (высосать) всю кровь (все силы); пить глазами; выжать все соки, вплоть до поэтического есенинского «пускай ты выпита другим …». Точно так же вампир лишает людей и животных жизни, основной субстанцией которой является кровь (о концепте кровь – см. СД 2: 677–681). Истоки этих разнообразных народных верований коренятся в убеждении, что умершие вообще (и особенно умершие преждевременной и неестественной смертью – опойцы, висельники, утопленники) – страдают от жажды. Поэтому у славян в рамках похоронной и поминальной обрядности широко распространена практика оставлять для душ умерших

21.1.2014 11:37:34

стакан воды в доме, на окне, во дворе, на могиле, поливать могилу водой, вином (свидетельства многочисленны у восточных и южных славян (см. ПА ИСл РАН, Плотникова 2004: 606, 607, СД 4: 57, и др.), у западных славян они в определенной степени нивелированы церковными запретами. Упырь как «нечистый» покойник тоже принадлежит к жаждущим. Немаловажная семантическая составляющая образа упыря-вампира – это обладание бóльшим количеством жизненной силы еще при жизни, отчего оставшаяся, неистраченная, неизжитая ее часть сильнее и дольше связывает его с земным миром (и тем брутальнее его нападения на живых). Вампирические свойства приписывались людям со сверхъестественными способностями и особыми приметами (ведьма, колдун, лекарь, двоедушник; ребенок, родившийся с зубами или «в рубашке»); душам людей, не доживших свой век, поэтому не истративших всех данных им сил; покойникам, похороненным с нарушением правил, напр.: через тело переступил человек, перепрыгнуло животное, перелетела птица; над телом передали какой-либо предмет – с магической точки зрения, такому покойнику сами живые по ошибке «передали» жизненную силу, отчего он начинал «ходить», то есть жить. С этнокультурной и семантической точек зрения упырь – это тот, кто упивается (выпивает, отпивает, наливается, надувается) кровью (или другой жизненной субстанцией), отбирая ее у живых. Объяснение имеющихся славянских форм имени вампира фонетически и морфологически довольно трудно. Само многообразие форм можно было бы объяснить длительной историей существования слова, огромной территорией его распространения и смешением его с другими, близкими по звучанию корнями, имеющими с ним общие семы (напр., с корнем *pěr- ‘впиваться’, ‘пронзать’, ‘протыкать’, поскольку вампир впивается, вонзается в жертву),¹⁶ законами развития отдельных славянских языков, воздействием на них соседних неславянских языков (возможно, колебания и/ы в корне или отвердения конечного согласного в рус. и укр. упирь/ упырь/упыр связанно именно с влиянием тюрк. убыр) и т.п. Чередование в анлауте в славянских диалектах (упир/опир/въпир/ вапир/lepir/oṕi/u̯oṕi/u̯eṕi/roṕe и т.п.) можно объяснить фонетическим переходом ǫ- (если принять как начальную форму *ǫpirъ) в у- (упир), из

95 Валенцова

16 Напр., бел. вупар объясняют как контаминацию слов упыр ‘кровосос, вампир’ и ўпiрацца ‘упрямиться’ (ЭСБМ 8: 225). Ср. аналогичный случай в ряду мифологических имен: «Напр., намной – дух, который наваливается во сне и оставляет синяки на теле, что является предвестием несчастья (яросл.). Другой вариант этого персонажа – навной (калинин.) – позволяет установить связь со словом навь (навье) ‘мертвец, покойник’» (Черепанова 1983: 49).

etymologie.indb 95

21.1.2014 11:37:35

96 Валенцова

которой затем делабиализацией – в о-(пир); протетическим начальным v- (vǫpirъ) объясняется переход в вам-(пир), ва-(пир), въ-(пир) (на разных этапах действия фонетических законов), протетическим l- – формы lepir, u̯eṕi, u̯oṕi и т.д. Возможны и семантические причины изменения приставки; например, в словацком формы с префиксом u- регулярны: upiť sa ‘упиться’, upiť si ‘отпить чего’ (VSRS 5: 231), обычны они и для чешского: upití, n. – das Abtrinken = napití, opilosť, die Trunkenheit; upíjeti – odpiti, ab-, wegtrinken; opiti, betrinken, berauschen; pitím přemoci, Jemanden übertrinken. Jg. Sv. Lucie noci upije (od 13. 12. dne již neubývá), an Lutze thut der Tag stutze … (Kott: 383) (последняя календарная паремия хорошо известна и словакам). Безусловно, предложенная гипотеза требует основательной этимологической проработки, объяснения всех фонетических и др. изменений в различных наименованиях вампира. Формальных трудностей в ней не больше, чем в любой другой, предложенной ранее. Вместе с тем новая версия опирается на этнографические и культурные данные, мифологический контекст более широкого круга представлений о смерти и посмертной жизни. Литература Аникин, А. Е. 2012. Русский этимологический словарь. Вып. 6. Москва. Антропов, М. П. 1998. Белорусские этнолингвистические этюды: 1. колода / колодка. In: Слово и культура. Памяти Никиты Ильича Толстого. Том ii. Москва, 21–33. Антропов, М. П. 2004. Белорусские этнолингвистические этюды: 2. Вызывание дождя (акциональный код). In: Язык культуры: Семантика и грамматика. К 80-летию со дня рождения академика Никиты Ильича Толстого (1923–1996). Москва, 190–216. Антропов, М. П. 2013. Белорусские этнолингвистические этюды: 3. «Куст» (часть первая). In: Ethnolinguistica Slavica. К 90-летию академика Никиты Ильича Толстого. Москва, 162–187. Белетич, М. – Лома, А. 2012. Сон, смерть, судьба (наблюдения над прасл. *mora, *mara). In: Этнолингвистика. Ономастика. Этимология. Материалы ii Международной научной конференции. Екатеринбург, 8–10 сентября 2012 г. В 2 частях. Екатеринбург, 159–160. БЕР = Български етимологичен речник. София 1971–. Березович, Е. Л. 2007. Язык и традиционная культура. Этнолингвистические исследования. Москва. Березович, Е. Л. 2010. Русская топонимия в этнолингвистическом аспекте. Мифопоэтический образ пространства. Москва. БМ = Българска митология. Енциклопедичен речник. Составител А. Стойнев. София 1994. Валенцова, М. М. 2013. Словацкие «рефлексы» общеславянской мифологической лексики. In: Ethnolinguistica Slavica. К 90-летию академика Никиты Ильича Толстого. Москва, 316–343. Васильев, В. Л. 2013. Славянская колонизация Русского Северо-Запада в свете ономастики. In: Славянское языкознание. xv Международный съезд славистов. Минск, 21–27 августа 2013 г. Доклады российской делегации. Москва, 17–35.

etymologie.indb 96

21.1.2014 11:37:35

Власова, М. Н. 1995. Новая абевега русских суеверий. Иллюстрированный словарь. Санкт-Петербург. Дукова, У. 1980. Названия на демонични същества от общославянско mor- в българския език (Морá, Морáва, Мáра, Мáрен, Марóй, Марóк). In: Език и поетика на българския фолклор. София, 108–113. Геров = Геров, Н.: Рѣчник на блъгарскый языкъ. Чясть прьва. Пловдивъ 1895. ЕСУМ = Етимологiчний словник украïнськоï мови. Т. 1–6. Киïв 1982–2012. Журавлев, А. Ф. 2005. Язык и миф. Лингвистический комментарий к труду А. Н. Афанасьева «Поэтические воззрения славян на природу». Москва. Зайцева, Н. И. 1975. Мифологическая лексика в чешском и словацком языках. Дисс. канд. филологических наук. Минск. Зубов, Н. И. 1982. Древнерусская теонимия (проблема собственного и нарицательного). Дисс. канд. филологических наук. Одесса. Зубов, М. I. 2010. Кiлька етимологiчних зауважень до назв слов’янських мiфологiчних персонажiв. Мовознавство, 2010/2–3, 113–123. Иванов, В. В. – Топоров, В. Н. 1974. Исследования в области славянских древностей. Лексические и фразеологические вопросы реконструкции текстов. Москва. Иванов, В. В. 1975. Реконструкция индоевропейских слов и текстов, отражающих культ волка. Известия АН СССР, Серия литературы и языка 34/5, 399–408. Иванов, В. В. 1983. История славянских и балканских названий металлов. Москва. Иванов, В. В. 2007. Балканские имена «вурдалака» и их происхождение. In: Балканские чтения 9. Terra balkanica / Terra slavica / К юбилею Татьяны Владимировны Цивьян. Москва, 70–79. Иванов, В. В. 2008. Развитие *(-)tl-, (-)dl- > (-)kl-, (-)gl- в циркумбалтийской языковой зоне и диалектное членение праславянского. In: Славянское языкознание. xiv Международный съезд славистов. Охрид, 10–16 сентября 2008 г. Доклады российской делегации. Москва, 249–275. КБДЛ = Карпато-балканский диалектный ландшафт: Язык и культура. Москва 2008. КБДЛ 2 = Карпато-балканский диалектный ландшафт: Язык и культура 2009–2011. Вып. 2. Москва 2012. Климова, К. А. 2008. Новогреческая мифологическая лексика в сопоставлении с балканославянской. Дисс. канд. филологических наук. Москва. МНМ = Мифы народов мира. Энциклопедия. Т. 1, 2. Москва 1997. Нимчук, В. В. 1988. Карпатоукраинско-южнославянские языковые параллели и тождества (история и перспективы проблемы). In: Общеславянский лингвистический атлас 1984. Москва, 294–313. Новичкова, Т. А. 1995. Русский демонологический словарь. Санкт-Петербург. ПА ИСл РАН = Полесский архив Института славяноведения РАН, Москва. Плотникова, А. А. 1994. Слав. *viti в этнокультурном контексте. In: Балканские чтения–3. Лингво-этнокультурная история Балкан и Восточной Европы. Тезисы и материалы симпозиума. Москва, 100–103. Плотникова, А. А. 2004. Этнолингвистическая география Южной Славии. Москва. Плотникова, А. А. 2006. Семантические и культурные балканизмы в этнолингвистическом аспекте. In: Исследования по славянской диалектологии. Вып. 12. Ареальные аспекты изучения славянской лексики. Москва, 7–19. Преображенский, А. 1910–14. Этимологический словарь русского языка. Т. 1. Москва. СД = Славянские древности. Этнолингвистический словарь. Под общей ред. Н. И. Толстого. Москва, Т. 1. 1995. Т. 2. 1999. Т. 3. 2004. Т. 4. 2009. Т. 5. 2012. Седакова, И. А. 2003. Судьба сакральных слов в славянских языках (между язычеством и христианством). In: Славянское языкознание. xiii Международный съезд славистов. Доклады рос. делегации. Москва, 534–549.

etymologie.indb 97

97 Валенцова

21.1.2014 11:37:36

98 Валенцова

Спировска, Л. – Вражиновски, Т. 1988. Вампирите во македонските верувања и преданиjа. Скопjе. СРНГ = Словарь русских народных говоров. Москва, Ленинград, Санкт-Петербург 1965–. Толстая, С. М. 2006. Культурная семантика и этимология. In: Studia etymologica Brunensia 3. Praha, 415–428. Толстая, С. М. 2007. Почему слепой не видит? (к этимологии слав. *slěp-). In: Словенска eтимологиjа данас. Zборник симпозиjума одржаног од 5. до 10. сеnтембра 2006. године. Београд, 409–419. Толстая, С. М. 2008. Пространство слова. Лексическая семантика в общеславянской перспективе. Москва. Толстой, Н. И. 1979. Этногенетический аспект исследования древней славянской духовной культуры. In: Комплексные проблемы истории и культуры народов Центральной и Юго-восточной Европы. Итоги и перспективы исследований. Москва, 17–26. Толстой, Н. И. 1985. Язык и этнолингвистика. Общественные науки 1985/4, 147–160. Толстой, Н. И. 1989. Некоторые соображения о реконструкции славянской духовной культуры. In: Славянский и балканский фольклор. Реконструкция древней славянской духовной культуры. Источники и методы. Москва, 7–22. Толстой, Н. И. 1995. Язык и народная культура. Очерки по славянской мифологии и этнолингвистике. Москва. Толстой, Н. И. 1997. Избранные труды. Т. 1. Славянская лексикология и семасиология. Москва. Толстой, Н. И. – Толстая, С. М. 1978. К реконструкции древнеславянской духовной культуры. (Лингво-этнографический аспект). In: Славянское языкознание. viii Международный съезд славистов. Доклады советской делегации. Загреб – Любляна, сент. 1978 г. Москва, 364–385. Толстой, Н. И. – Толстая, С. М. 1993. Слово в обрядовом тексте (культурная семантика слав. vesel-). In: Славянское языкознание. xi Международный съезд славистов. Доклады российской делегации. Москва, 162–186. Топоров, В. Н. 1984. Прусский язык. Словарь. K–L. Москва. Топоров, В. Н. 2004. Исследования по этимологии и семантике. Т. 1. Теория и некоторые частные ее приложения. Москва. Трубачев, О. Н. 1967. Из славяно-иранских лексических отношений. In: Этимология 1965. Материалы и исследования по индоевропейским и другим языка. Москва, 3–81. Трубачев, О. Н. 2004, 2008, 2009. Труды по этимологии. Слово. История. Культура. Москва, Т. 1–2. 2004. Т. 3. 2008, Т. 4. 2009. Усачева, В. В. 1977. Об одной лексико-семантической параллели (на материале карпато-балканского обряда «полазник». In: Славянское и балканское языкознание. Карпато-восточнославянские параллели. Структура балканского текста. Москва, 21–76. Фасмер, М. 1986–87. Этимологический словарь русского языка. Перевод с немецкого и дополнения чл.-корр. АН СССР О. Н. Трубачева. Изд. второе. В 4-х томах. Москва. Хобзей, Н. 2002. Гуцульська міфологія. Етнолінгвістичний словник. Львів. Черепанова, О. А. 1983. Мифологическая лексика Русского Севера. Ленинград. ЭСБМ = Этымалагiчны слоўнiк беларускай мовы. Мінск 1978–. ЭССЯ = Этимологический словарь славянских языков. Москва 1974–. AT ÚEt – Archív textov Ústavu etnológie SAV, Bratislava. Baranowski, B. 1981. W kręgu upiorów i wilkołaków. Łódź. Blagoeva-Neumanová, T. 1976. Ľudová démonológia na Zamagurí v kontexte a koreláciách s démonologickými predstavami východných, južných a západných Slovanov. Rigorózna práca. Bratislava. Brückner, A. 1927. Słownik etymologiczny języka polskiego. Warszawa.

etymologie.indb 98

21.1.2014 11:37:36

Budziszewska, W. 1991. Polskie nazwy zmór i niektóre wierzenia z nimi związane. In: Studia z filologii polskiej i słowiańskiej 27, 17–23. Dukova, U. 1983. Die Bezeichnungen der Dämonen im Bulgarischen. Балканско езикознание / Linguistique balkanique 26/4, 6–46. Dukova, U. 1985. Die Bezeichnungen der Dämonen im Bulgarischen. iii. Entlehnungen. Балканско езикознание / Linguistique balkanique 28/2, 5–62. EĽKS = Encyklopédia ľudovej kultúry Slovenska. Diel 2. Bratislava 1995. GM = Ján Michálek a kol.: Gemer–Malohont. Národopisná monografia. [Martin, 2011.] Holub, J. – Kopečný, F. 1952. Etymologický slovník jazyka českého. Praha. Horehronie = Horehronie. Kultúra a spôsob života ľudu. Bratislava 1969. Kott = Kott, F. Š.: Česko-německý slovník zvláště grammaticko-fraseologický. Díl čtvrtý. T–Y. Praha 1884. Machek, V. 1968. Etymologický slovník jazyka českého. Praha. Michálek 1989 = Ján Michálek a kol.: Ľud hornádskej doliny (na území Popradského okresu). Košice 1989. Rejzek, J. 2001. Český etymologický slovník. Voznice. Skok, P. 1971–74. Etimologijski rječnik hrvatskoga ili srpskoga jezika. T. 1–4. Zagreb. SSN = Slovník slovenských nárečí. Bratislava 1994–. Sychta = Sychta, B.: Słownik gwar kaszubskich na tle kultury ludowej. T. 3. Wrocław – Warszawa – Kraków 1969. Treder, J. 1989. Frazeologia kaszubska a wierzenia i zwyczaje (na tle porównawczym). Wejherowo. Vážný, V. 1955. O jménech motýlů v slovenských nářečích. Bratislava. VSRS = Veľký slovensko-ruský slovník. Diel i–vi. Большой словацко-русский словарь. Т. i–vi. Bratislava 1979–1995.

etymologie.indb 99

99 Валенцова

21.1.2014 11:37:37

etymologie.indb 100

21.1.2014 11:37:38

etymologie.indb 101

21.1.2014 11:37:38

Studia etymologica Brunensia Edited by Ilona Janyšková & Helena Karlíková

1

Studia etymologica Brunensia 1. Sborník příspěvků z mezinárodní vědecké konference Etymologické symposion Brno 1999. Ed. I. Janyšková & H. Karlíková. Euroslavica, Praha 2000. 375 p.

2

Studia etymologica Brunensia 2. Sborník příspěvků z mezinárodní vědecké konference Etymologické symposion Brno 2002. Ed. I. Janyšková & H. Karlíková. Nakladatelství Lidové noviny, Praha 2003. 458 p.

3

Studia etymologica Brunensia 3. Sborník příspěvků z mezinárodní vědecké konference Etymologické symposion Brno 2005. Ed. I. Janyšková & H. Karlíková. Nakladatelství Lidové noviny, Praha 2006. 463 p.

4

Varia Slavica. Sborník příspěvků k 80. narozeninám Radoslava Večerky. Ed. I. Janyšková & H. Karlíková. Nakladatelství Lidové noviny, Praha 2008. 275 p.

5

Bohumil Vykypěl: Život a dílo Adolfa Erharta. Kapitola z dějin české vědy. Nakladatelství Lidové noviny, Praha 2008. 220 p.

6

Studia etymologica Brunensia 6. Sborník příspěvků z mezinárodní vědecké konference Etymologické symposion Brno 2008. Ed. I. Janyšková & H. Karlíková. Nakladatelství Lidové noviny, Praha 2009. 420 p.

7

Dobrodružství etymologie. Články Františka Kopečného z prostějovského časopisu Štafeta. Ed. A. Bičan & E. Havlová. Nakladatelství Lidové noviny, Praha 2009. 232 p.

8

Anton Matzenauer: Beiträge zur Kunde der altpreußischen Sprache. Ed. E. Havlová. Nakladatelství Lidové noviny, Praha 2009. 204 p.

9

Vít Boček: Studie k nejstarším romanismům ve slovanských jazycích. Nakladatelství Lidové noviny, Praha 2010. 180 p.

etymologie.indb 102

21.1.2014 11:37:38

10

Eva Havlová: České názvy savců. Historicko-etymologická studie. Nakladatelství Lidové noviny, Praha 2010. 272 p.

11

Bohumil Vykypěl: Studie k šlechtickým titulům v germánských, slovanských a baltských jazycích. Etymologie jako pomocná věda historická. Druhé, přepracované a rozšířené vydání i s dodatkem o šlechtických titulech v keltských jazycích. Nakladatelství Lidové noviny, Praha 2011. 332 p.

12

Václav Machek: Korespondence. i–ii. Ed. V. Boček & P. Malčík. Nakladatelství Lidové noviny, Praha 2011. 528 + 504 p.

13

Jana Villnow Komárková: Slovanská terminologie tkaní z pohledu etymologie. Na příkladě českého, chorvatského a srbského jazykového materiálu. Nakladatelství Lidové noviny, Praha 2011. 224 p.

14

Methods of Etymological Practice. Ed. B. Vykypěl & V. Boček. Nakladatelství Lidové noviny, Praha 2012. 120 p.

15

Theory and Empiricism in Slavonic Diachronic Linguistics. Ed. I. Janyšková & H. Karlíková. Nakladatelství Lidové noviny, Praha 2012. 492 p.

16

Etymology: An Old Discipline in New Contexts. Ed. B. Vykypěl & V. Boček. Nakladatelství Lidové noviny, Praha 2013. 104 p.

etymologie.indb 103

21.1.2014 11:37:38

Etymology: An Old Discipline in New Contexts Edited by Bohumil Vykypěl and Vít Boček Studia etymologica Brunensia 16 Eds. Ilona Janyšková and Helena Karlíková Published by the Etymological Department of the Institute of the Czech Language of the AS CR, v. v. i., in NLN, s. r. o., Nakladatelství Lidové noviny, Dykova 15, 101 00 Praha 10, www.nln.cz. Designed by David Březina (davi.cz) and Vít Boček. Typeset in Skolar PE by Vít Boček. ISBN 978-80-7422-263-4 Number of pages: 104 Praha 2013 Price (including VAT): 119 CZK

etymologie.indb 104

21.1.2014 11:37:39