Comparative Grammar of the Greek Language [PDF]

Zitiervorschau

r

THE STUDENTS' SERIES OF HISTORICAL AND COMPARATIVE GRAMMARS EDITED BY JOSEPH WRIGHT

COMPARATIVE GRAMMAR OF THE

GREEK LANGUAGE BY

JOSEPH WRIGHT PH.D., D.C.L,, LL.D., LITT.D.

FELLOW OF THE BRITISH ACADEMY PROFESSOR OF COMPARATIVE PHILOLOGY IN THE UNIVERSITY OF OXFORD

'

Nur das Beispiel fuhrt zum Licht; Reden thut es nicht

Vieles

'

HENRY FROWDE OXFORD UNIVERSITY PRESS LONDON, NEW YORK AND TORONTO 1912 [All rights reserved^

OXFORD

:

HORACE HART

PRINTER TO THE UNIVERSITY

PREFACE Grammar

In writing this

have followed as

I

Series,

my

far

Grammars

possible the plan adopted in the other

as

of the

object being to furnish students with a concise

account of the phonology, word-formation, and inflexions

As the book

of the language.

some more or

This

omitted. dialects

less

is

not intended for specialists

is

especially the case in

which have been preserved

as to render

of their phonology.

It

must

not,

full

account

however, be assumed

have been entirely omitted

that these dialects

contrary,

regard to those

such scanty fragments

in

impossible for us to give a

it

;

on the

have made considerable use of them

I

in the

phonology and elsewhere, wherever they have helped throw

light

dialects,

Much

^

important details have been intentionally

to

upon the development and history of the other

such as

Attic, Ionic, Doric, Aeolic.

of the time and labour spent on this

Grammar

has been taken up with selecting examples from the vast

amount of material which

I

had collected to illustrate the This selection was

sound-laws of the various dialects. necessary

if I

students for

planned,

was

whom

keep steadily

to

the Series of

otherwise

view the class of

would have been

it

produce a Comparative Greek

Grammar

times the size of the present one.

compression of the material,

I

In

far at

originally

easier

In spite of this great

all

Greek Philology.

255345

to

three

least

venture to think that

included within a modest compass 1

in

Grammars was

I

have

that the ordinary

Preface

vi

require

student will

know about the subject, and I who thoroughly masters the book

to

believe that the student

only have gained a comprehensive knowledge of

will not

Comparative Greek Grammar

in particular,

but will also

have acquired the elements of the Comparative Grammar of the Sanskrit, Latin, and Germanic languages.

Examples

have been more copiously used from these than from the other branches of the Indo-Germanic family of languages,

because

it

study this

who

can be safely inferred that the students

Grammar

will

already possess

a

practical

knowledge of one or more of them. This Grammar makes no pretence whatever of being an original

and exhaustive

of this kind there

on the

treatise

is practically

either of these features,

but

In a book

subject.

no scope

for a display of

have contrived

I

to

bring

within a comparatively small space a great deal of matter

which

who

will

be new to students, and especially to those

are unable to study the subject in works written in

foreign languages.

All that

I

have attempted

to

do

is

to

furnish our countrymen with a systematic and scientific

treatment of Comparative Greek

and

philological books

Grammar based upon

articles of the best

present day in the wide

field of

Comparative Philology.

Specialists in the subject will accordingly find is

new

in the

still

is

all

the other Indo-Germanic languages

and not a few points about which

a great divergence of opinion even

philologists.

that

innumerable points which have never been

satisfactorily explained,

there

little

book.

In Greek as in there are

the

workers of the

In

all

such cases

I

among

the best

have carefully considered

the various explanations which have been proposed, and

have given those with which

I

agreed without, as a

rule.

Preface stating

my

authority,

but where

vii

was unable

I

with any of the proposed explanations preferred to state that the

phenomenon

to

agree

have generally

I

in question

has

never been satisfactorily explained or that the explanation is

unknown, rather than burden the book with attempted

explanations with which

did not agree.

I

I

have generally

omitted to give the authorities for various statements

made

throughout the Grammar, except in special cases where I

thought

it

desirable to

refer

the student for further

information to the sources which deal more fully with the case in point. I

gratefully

acknowledge the help

have derived from

I

the learned books and articles by the splendid band of Philologists who have done so much to throw upon the history and philology of the various Indo-

German light

Germanic languages. list of the books and

On pp.

xiv-xvii will be found a select

articles

which

I

useful in the writing of this book, but a

have found most

mere place

in a list

would not adequately express my indebtedness to the works of Brugmann, Hirt, Gustav Meyer, Osthoff, Johannes Schmidt, and Wackernagel. In conclusion I wish express

to

my

sincere thanks to the Controller of the

University Press for his great kindness in complying with

my

wishes in regard to special type

;

for his valuable, collaboration in the

verborum

;

and

to

Mr. A. Davidson,

making of the index

lastly to the press-reader,

Mr.

W.

F. R.

Shilleto, for his invaluable help with the reading of the

proofs.

JOSEPH WRIGHT. Oxford, January^ 1912.

CONTENTS PAGES

INTRODUCTION

1-4

Indo-Germanic languages

Classification of the

The Greek

and their

dialects

(§ i).

classification (§ 2).

CHAPTER

I

Pronunciation and Accentuation

Vowels (§§4-18)

5-18

Consonants (§§ 19-27). Pitch and broken or acute and slurred or stress accent (§ 28) circumflex accent (§ 29) word-accent (§§ 30-4) sen;

'

;

'

'

'

;

;

tence-accent (§§ 35-40).

CHAPTER The

II

Primitive Indo-Germanic Vowel-sounds

The Indo-Germanic vowel-system

.

18-20

Table of the normal development of the prim. Indg. short and long vowels, short diphthongs, and short vocalic nasals and liquids in Greek, Sanskrit, Latin, Old Irish, Gothic, Old English, Lithuanian and Old Slavonic (§ 42).

CHAPTER

(§ 41).

III

The Greek Development of the Indo-Germanic Vowelsystem

The 46)

(§

a

21-49

:— a

short vowels

u

;

(§§ 50-1)

(§§ 47-8) ;

e

(§

52)

;

;

(§

43)

(§

53)

;

(§

(§ ;

54)

44)

i

(§ 45)

;

long vowels

:

;

(§ 55).

;

ei (§ 58)

;

The

oi (§ 59)

eu (§ 61) ou (§ 62). The long diphthongs General remarks on the short vocalic nasals and

60)

(§ 63).

(§

The

3 (§ 49).

short diphthongs :— ai (§§ 56-7)

au

e

;

;

liquids (§ 64)

;

;

short vocalic nasals

(§ 65)

;

short vocalic

Contents PAGES

The long

liquids (§§ 66-7).

The lengthening

(§ 68).

vocalic nasals

of short vowels

and liquids

The

(§ 69).

shortening of long vowels (§§ 70-1) quantitative metathesis (§ 72). Assimilation of vowels (§§ 73-4). Epen;

Prothesis

thesis (§§ 75-6).

(§

Anaptyxis

77).

(§ 78).

Vowel-contraction (§§ 79-80).

CHAPTER

IV

Ablaut

49-61

General remarks on ablaut (§§ 81-5). The weakening or loss of vowels (§§ 86-90). The lengthening of vowels 91-4).

(§§

bases

The

ablaut-series (§§ 95-6).

Dissyllabic

(§ 97).

CHAPTER V The

Primitive Indo-Germanic Consonants

Tableof the prim. Indg. consonants

(§ 98).

.

.

.

62-71

The normal

Indg. explosives in Greek, Germanic, Gothic, Sanskrit, Lithuanian and Old Slavonic :— the tenues (§ 100); the tenues aspiratae (§ 102) the the mediae (§ mediae aspiratae (§ 103). Consonantal sound-changes which took place during the prim. Indg. period

prim.

equivalents of the Latin, Old

Irish, prim.

)

;

;

(§§ 105-12).

CHAPTER The Greek Development

of the

VI I

ndo- Germanic Con-

sonant-system

71-111

of mediae aspiratae to tenues aspiratae De-aspiration of aspirates (§ 115). Assimilation

The change (§ 114).

of consonants (^"§ 116-17). General remarks on the semiGeneral vowels (§§ 1 18-19) (§§ 120-6); j (§§ 127-30). remarks on the liquids (§ 131 1(§§ 132-5); r(§§ 136-8). General remarks on the nasals (§ 139); (§§ 140-6) ;

w

)

;

m

(§§ 147-54)

b

(§^ 159-60)

t (§§

80).

;

;

,

r)

The labials :— bh (§§ 162-3). The

(§§ 155-6).

ph (§

161);

(§§ 157-8) dentals :—

164-70); d (§§ 171-4); th (§§ 175-6); dh (§§ 177Indg. palatals,

The normal equivalents of the prim.

a3

Contents pure velars and labialized velars in Greek, Latin, Old Irish, Germanic, Sanskrit, Lithuanian and Old Sla-

The palatals :— k (§§ 182-7) g (§§ 188gh (§§ 193-4). The pure velars:— qh (§ 200); gh (§ 201). The (§§ 195-6); g (§§ 197-9) labialized velars:— q"' (§§ 202-4); g' (§§ 205-7); q"h The spirants — s (§§ 212-23) (§ 208) g^h (§§ 209-10), sh, zh (§ 225) )?, \>h, d, dh (§ 226) j (§ 227). (§ 224) vonic

(§ i8i).

kh

91);

;

192);

(§

;

:

;

;

;

;

CHAPTER

VII

Sandhi

111-116

General remarks on sandhi (§ 228) initial sounds (§§ 231-2). (§§ 229-30)

sounds

final

;

CHAPTER

VIII

The Formation of Nouns and Adjectives

116- 138

.

.

General remarks (§ 233). Root-nouns (§ 234). Suffixes ending in a vowel :—ja- (§ 235) --, -a- (§ 236); (i)jo-, ;

•ew(i)jo• (§ 237)

•(i)ja•, -ejo-,

•ma2)

239)

{§

-ino-, -ina- (§ 243)

;

245)

C§

-meno-, -mena-

;

;

--,

•ra- (§ 248)

;

-;

(§

251)

;

(§

(§263);

••

-u-

d-en275)

—

(§

279) 282)

(§

The (s, + i;

o

§ 11.

tti

;

§ 12.

short diphthongs

+ v;

+

,.

a,

preserved

which changed short u

o

e,

in

v

ai, ei, ol

av, ev, ov

;

u

but the original

i,

to

ii,

6 (=Indg.

like the

i

in

was

quality

the second element of the u-diphthongs.

was nearly

= a,

vi

;

English five, as

as

,

58) had nearly the same sound as

ei, §

the ai in English stain until about the beginning of the

century

fifth

b.

c,

then became long close e

it

, , ^ ^,

.

in Attic,

Ionic and the milder Doric dialects, although the retained in writing, as

The

€i

was then used

;

to

ei

was

express the long close e which

arose from contraction and from compensation lengthening, as

,

0'€,

rpus, u.\ov from

Cret. ev9,

*^

0i'Aee, ^rpeje?,

from

*9,

*€-€'

€h =

;

;

this

ei

was written e on the oldest Attic inscriptions, whereas prim. Greek et was always written ei, § 13. 01 had the same sound as the oy in English boy, as oiSa,

^^, =

.

above) was a special Greek development and arose partly from the loss of an intervening consonant and partly from contraction, as vios from §14.

ui (see

L'i

*fiSva/a,

*/9,

loc. sing.

Hom.

,.

.,

had the same sound as the au in German Haus, § 15. and was nearly like the ou in standard English house, as §

1.

had approximately the same sound as

heard

in

house

(eus),

§ 17. fifth

ov

the

{=

mouse

(meus), as

Indg. ou,

century b.c,

it

, ^,

is

often

English dialect pronunciation of

southern

§

62)

= o+v

(see

Zev.

above) until the

then became long close

through the

Phonology

8

intermediate stage of long close

The

retained in writing.

=

Cret.

was always

8\^,

^,

= Dor.

from *8opfo^ ; from

on the oldest Attic § 18.

although the ov was

ov was then used to express the

which arose from contraction and comfrom voos,

later

pensation lengthening, as ;

,

this was written whereas prim. Greek ov ;

inscriptions,

written ov.

The

original long diphthongs ai, ei,

€,€€,; ,^,, became short before consonants already &c.

(§

thongs

18-21

, , ^ ,? *

long close &c.

[§§

63).

The second element

-, -,

• ceased

, \, )

to

au, eu,

Greek, as in

from

*€,

of the long final diph-

be pronounced in the second

probably much

century b.c, and in

;

in prim.

modern mode of writing these diphthongs

earlier.

as a,

u,

The (0ea,

only dates back to manuscripts of the

twelfth century.

B.

The plosives , ,

The Consonants.

voiceless explosives

§ 19.

,

,

,

the voiced ex-

had approxi-

mately the same sound-values as in English. consonants require special attention.

The remaining

§

and the

r,

liquid

the nasals

,

20. In the oldest period of the language y

", ?, ^,

positions a voiced explosive like the

ago, as ykvos,

g

in

was

in all

English go or but already at

an early period it became a voiced spirant in the popular dialect medially between vowels. The guttural nasal r) (=the in English think and the ng in sing) was expressed

by

1^

,

on the oldest inscriptions, but after the combinations qm in such words as had become (§ 155), it came to be expressed by y, as ayyeXoy,

, ? , , ^.

,

§ 21.

In the earliest historic period of the language

was a compound consonant

like the

zd

in

English blaz(e)d

Pronunciation

22-4!

,

, ^, .

^^

and arose from older dz by metathesis (§ 129, 8), as The dz must have become Zev9, zd before the disappeared in words like from

,.

*-8(:

69,

forms like in Attic probably became beside some time during the fourth century b. c. Some scholars was pronounced like the s (= z) in English assume that measure, pleasure already in the earliest period of the 153)

(§

cp. also

;

language.

had a strong

§ 22.

formed by trilling the point of It was voiceless initially and medially after , , and probably trill

the tongue against the gums. (written

p,

see

§ 215),

In other positions

after all other voiceless consonants.

,

, ^, ,,, it

was voiced

epvOpos, §

like the

^),

other positions, as

,

in

z)

before voiced explosives, as

6$,

(cp. §

--

voiceless

in

doubtful

,, , It is

-- and Attic, Boeotian, Thessalian and were pronounced in such words as Ion. beside Attic, &c.

129,

7).

th

in

like the

but

Xvkos.

the Ionic

Cretan

hard, bearing, bear, as

eap.

was voiced (=

23.

how

Scotch r

Some

--,

scholars assume that the

English thin or a kind of lisped

s,

••

was

whilst others

was the same as the sh in English she. were aspirated voiceless explosives like German paar, teil, kein and in the Anglo-

think that the sound §

24. 0,

, X

,,, ^, , € ' ' .

the p,

k

t,

in

Irish pronunciation of pair, tell, kill, as ;

, '.?

^,

\€,

re0oy,

;

became a spirant (= th in English thin) at an early period some dialects, and also became spirants later, but , , X must have been aspirated voiceless explosives at

in

the time de-aspiration

took place, cp.

^, ^,

and also when , , became aspirated before a following rough breathing, cp. :

,

ov,

(§

115)

;

Phonology §

25.

y\f

|,

, ^.

probably represented the combinations

(often written

?,

0?), as e^co,

€^,

25-8

[§§

«y,

?

The spiritus asper corresponded to the EngHsh It house, and was originally represented by H. disappeared in the prehistoric period in Lesbian, Elean, then the dialect of Gortyn, and the Ionic of Asia Minor. came to be used in Ionic to represent the e from older a was halved \-, ], and the At a later period the (§ 51). former was used for the spiritus asper and the latter for the lenis. From these fragments came the later signs and § 27. On F and 9 see § 120 and § 47, note 2. §

h

26.

'

in

'.

"

Accent. §

By

28.

accent

in

its

widest sense

is

meant

the

gradation of a word or word-group according to the degree

of stress or of pitch with which

Although

uttered.

strictly

its

various syllables are

speaking there are as

different degrees of accent in a

many

word or word-group as

there are syllables, yet for ordinary purposes

it

is

only

necessary to distinguish three degrees, the principal accent, the secondary accent, and the weak accent or as it is generally termed the absence of accent.

accent

is

The secondary

as a rule separated from the principal accent by

one intervening syllable. Indo-Germanic languages have parti}' pitch (musical) and partly stress (expiratory) accent, but one or other of the two systems of accentuation always predominates in each language, thus in Greek and Vedic the accent was predominantly pitch, whereas in the oldest periods of the Italic dialects, and the Keltic and Germanic The languages, the accent was predominantly stress. effect of this difference in the system of accentuation is clearly seen by the preservation of the vowels in unaccented syllables in the former languages and by the weakening or

at least

All

the

loss of

them

in

the latter.

In the early period of the

Accentuation

§ 29]

1

parent Indg. language, the stress accent must have been

more predominant than the pitch accent, because it is only upon this assumption that we are able to account for the origin of the various phenomena of quantitative ablaut 86-90).

(^§

now

It is

a generally accepted theory that at

a later period of the parent language the system of accentuation became predominantly pitch with which

was

probably connected the origin of qualitative ablaut (§ 83). This pitch accent was preserved in Greek and Vedic, but

became predominantly of nearly

predominantly stress the

stress again in the primitive period

It had also become Greek by about the beginning of see Kretschmer, Kuhn's Zeitschrift,

the other languages.

all

Christian

era,

in

XXX, pp. 591-600. §

29.

The

quality of the prim. Indg. syllable-accent

was

broken or acute and the slurred or The former was a rising and the latter circumflex. a rising-falling accent. Long vowels with the acute accent were bimoric and those with the circumflex trimoric. All original long vowels including the first element of long diphthongs had the acute accent. The circumflex accent was unoriginal and arose in prim. Indo-Germanic in the of two kinds, the

following

manner

'

:

•as from -a-es in the

'

'

{a)

From

nom.

pi.

'

the contraction of vowels, as

of a-stems, -os from -o-es in

the nom. pi. of o-stems, -oi from -o-ai in the dat. sing, of

€, ?

The circumflex also arose by ^eo), see § 79. from *r/)e/e9, vowel contraction within Greek itself, as T)yov, and cp.

*efapj09,

§

*/,

from

*ayKovja.

43

,^,.

See

The combinations

129,

§

3.

aaj, oaj, (aj,

/' became

ai, 01,

(through the intermediate stage ahj, &c.), and then

^, , */, ,

the second element of the diphthong partly disappeared in

*/ ,

/

and Ionic before a following 0, as Hom. from *^/, cp. aor. ; ; Skr. tdsya; Hom. ^^, from *e/ze(r/"o ; opt. Hom. Att. from */^, Skr. syam; Hom. from cp. reXa, from from *fi8vaja = Skr. vidusi. pf. part. fem. Hom. •ey

Attic

from

^

*€6/

;

;

See

§

129,

9.

Prothesis. §

assumed

77. It used to be

Greek before an scholars are

initial liquid,

now

that prothesis took place in

nasal or f

+

vowel, but most

was only developed

in the initial

combination Indg. r

vowel, as epe/3oy, Skr. rajah, Goth, riqis, darkness;

OE. read,

Skr. rudhirdh, Lat. ruber, Lat. rego.

Even

,

generally agreed that a prothetic vowel

these examples

in

7-ed

opky(u beside

;

not improbable

is

phase of ablaut which

that the initial vowel represents a

€

it

+

has not been preserved in these and similar words in the

Forms

other Indg. languages. beside Lat. lino

;

man

;

beside Skr. nar-,

beside

like

beside Lat.

beside Lat.

beside Skr. vati, he blotcs

;

Hom.

;

mulgeo

n5men

€{)

{)

;

beside

;

;

which were formerly regarded as containing a prothetic vowel, represent a different grade of ablaut. A-NAPTYXIS. §

78.

By

anaptyxis

is

meant the development of a vowel

between a liquid or nasal sonant.

Vowels of

+

a preceding or following con-

this kind are

found

in

the

old and

Phonology

44

[§

, , ?, ? . modern periods of most of examples occur

gether

the Indg. languages.

No

79

sure

Greek, but they are not altoon inscriptions, &c,, as Attic (Vase) (Hipfor

in classical

uncommon

;

ropovos (quoted by

ponax) for

Hesychius

for

Cp. beside Tarentum) beside ropvos, Lat. pocolum, poculum beside p5clum, French canif beside English knife, mod. northern dial, starak, stak beside Old English styric, styrc, calf; filam, marabl beside standard English film, marble.

Vowel-contraction. 79. In treating of vowel-contraction

§

distinguish three periods

:

(i)

it

is

necessary to

contractions which took place

already in the Indg. parent language,

(2)

those which took

place in primitive Greek, and were accordingly all

the

Greek

dialects,

and

(3)

common

to

those which took place in the

individual dialects.

To treat first

in detail the

question of vowel-contraction in the

period would be beyond the plan and scope of this

grammar, because the resultant long vowels or diphthongs were not merely common to Greek, but to all the Indg. languages. It will therefore be sufficient to give here only a few examples of such contractions e 4- e > e, as Skr. asam, from *esm, older *e (augment) + esm, / was. + a > in the dat. sing, of o-stems, as from *dhweso + ai, cp. the original ending preserved in infinitives like Skr. vidmane, to know Att. Cypr. of€ua, Skr. davane, to give. + e > in the nom. pi. of o-stems, as Skr. vrkah, Goth, wulfos, from older *wlqo + es, wolves, cp. :

,

^

^,

;

,

*18,

a + e > a she-wolves,

in the nom. pi. of a-stems, as Skr. vrkah, from *wlqa + es.

Voivel-coniradion

8ol

§

45

^,

a + a > a in the dat. sing, of a-stems, as cp. Goth. gihki, for a gift, Indg. -a + ai. Most of the contractions were due to the loss of intervocaHc s (through the intermediate stage h) and j in primitive Greek. After the loss of these consonants the combinations a, e, were contracted in certain cases -f 1 or in the prehistoric

period of place

f took

intervocalic

The

the dialects.

all

much

a

at

loss of

period

later

and

accordingly belongs to the history of the separate dialects

But the great majority of vowel-contractions took Greek became differentiated into the various separate dialects. Vowels were contracted in Attic more extensively than in any of the other dialects, although even in this dialect there were certain combinations which did not undergo contraction in all cases ea, where an intervocalic (i) The combinations eo, s or j had disappeared, remained in dissyllables, but underwent contraction in words of more than two syllables, as from 060y from *ef€ao9, but 'iap from *fiaap, but gen. but from *8€], I bind, from *fiaapo9 Sios from *8fijos but But when the same combinations arose after the loss of f they remained uncontracted in older Attic even in words of more than two syllables, as ui{f)o9, 122).

(§

place after primitive

,^ :

,

€,

*^,

^,

€{)\6^,

.

T]8k{f)o!iv•,

;

;

[) ^,

gen. r]8k[f)o^,

from *afos,

as,

until;

Kpe{f)a9,

f,

kvvi{f)a,

as

;

-,

ue{f)ap6i, eof,

{)8,

, (2) t,

§

^,

Cypr.

r]8e{f)a.

Dor.

-fjFos;

In like

oa remained uncontracted y^6{f)avos,

IlepiKXiovs, from *-K\ifeaos.

by

Hom.

etwy,

Hom.

manner the combinations ?, after the loss of

;

;

[),

,

, ,

gen.

The combinations e,€were contracted when preceded

but remained in other cases, as ace.

80.

-eay

;

Below

gen. is

?,

beside

aXids, beside

-.

given a classification of the contractions

arranged according

to the

nature of the

first

vowel.

.* ^

Phonology

46

>

a^- a

8

+€ >

*8;

in Att. Ion., but

€/3

from

a-fi

from

€€ >,

;

^

^,

(,

;

Ion., but

in Att.

from

from

in Dor., as

Att. Ion.

in

Horn,

+

> >

from

;

Dor. (Theocritus) kna^d from -ao

+

in

-. 77

from

au, as

>

the

all

from

*5?,

€. in

, , ^.

Indg. *dnsulos.

^,

as

dialects,

in Att. Ion., but

€, > ,

Dor.

^

Dor. Aeol. and Elean, as

Boeot. (fivaavres from -aoj^rey.

a+f

;

from *afi-

;

;

;

from

Dor.

;

;

a-\-o>

=

from o/oae. from *naFi9 yipai from ykpai.

Dor.

as

;

80

from

Horn. Ion.

in all the dialects, as

Horn.

[§

from

Dor. Locr., as subj. from Ion.

}. 6, ;

/?

, 5 ; ^, , ^.,{) , . //

from as subj. +; from + > , as rt + € {= e) > in Att., as

/

from

Ion.

*0-

from ae/pay. fiavo^ 6 (= prim. Gr. ei) > in Att. Ion., but ; in Dor., as ;

+

from *(^)6'5 indie, from opaei. a + oi > a>, as 5^ from

from

;

;

;

Dor.

//,

from

€,

from > , as + + > ;, as gen. ^poy from *fkaapo^•, e + e > ei (= long close e) in Att. Ion. and

€

(^

e

, ,, , 6€;

severe Dor. and severe Dor., as from from 0'€€ Lesb. severe Dor. from TpeiY, Lesb. = Skr. dsi; 76/ from + > €1, as eZ from yit'ei from ^yiveai. in Dor., and ev in Ion. + o > ov (=5, later u) in Att., from and Boeot., as Att. yipovs from *yep«ro9 ; in Aeolic

€

from *yei'eaa.

mild Dor., but

1

;

;

(,

*/€.

*6

^

Soi

§

Virau'lcoutraciioii

*.]^ €/i€oy

^ -^^

beside

;

from

;

47

;

, , ^ Ion. Boeot.

;

from /StAeoy older *j8eAe^oy.

+a > + ££ > 6 + >

e

677

>

i;

in Ion.

e

e?;

>

?;

in Ion., as dat.

T?

in

77

e €

€ €

+ +

>. >

j;

7;,

(§ 51),

as

from -e?;?. from

as

0^

in

Att.,

=

^; = Att.

the dialects,

all

via.

yei^ea.

from

as

^).

0, ?,

as

^ from

Dor.

beside

Ion.

;

,

^() =

Skr. bhdrase.

as

from

+ ai>7),a.s indie, mid. 0epi; from + € (= e, § 58) > € (= long close

e),

^^

*/€€/09, older *K\if^avos.

e (

e

^. , ^, , /. , ,, , ,- . ;

+ ti (= prim. Gr. ei) > + 01 > 01 in Att., as

> + + > ,

as

as Ji

as 0iAei~ from

ei,

.

from

from from Atfi, Att. -da' had

^ €,

from the

its

genitive, o

+a

>,

+e >

from

as

Indg.

*aidosm;

ou in Att. Ion. mild Dor., but

severe Dor,, as

in

^,

severe Dor. -are, from -oeTe

Att. eXarrofy, Ion.

;

from o-'rL>

-oey,

from

*KofiAo$•

ovis, Skr. dvih.

,

+ > Of in Att. Ion. and mild Dor., ;

+

Tj

+

>

i/ooy

/€9,

;

;

€),

from *ofi?

oly

but

in

=

Lat.

Aeol. and severe

in all

Boeot. Lesb. severe Dor.

severe Dor. -arrey, from

8€9,

> , as > , as subj. > in all the from

Dor.

. ,/ , ^ ^. . from

Dor., as

+ o +

from

;

Indg. *-oses.

as

OL,

from

+

irova

from

;

Ion.

from -;.

dialects, as

from

the dialects, as

from

-ooj/rey.

=

Phonology

48 o

+ ci (=

long close

from o

+ €c (=

prim. Gr.

beside Lesb.

+

>,

,

> ov, from

from

as

, 6€ , , from

as

oi,

6[.

^vvool.

;

from

*.

older

from

?

* ,

from aeXioy a+

+

L

>

,

-.

See

> >

•

Dor. Lesb.

;

323.

§

from

-aay, cp. § 323.

in

/,

*

Dor. and Aeol., gen.

older

+€ >

*.

from

as dat. Dor.

a,

>

in

=

from

;

+; >

V

+V > + et {=

^.

^

>

nom.

+>

Gr.

,

;

gen. Dor.

from

€,

, -rjFe?;

} from

Ion.

from

*^J7i7.

^ in Att. Ion. and mild Dor., as

^^. the from *€ ^ from *(. Ion.

+ ( (=prim.

-

from

^^ all the dialects, as Att. subj. e)

from

Att.

in all the dialects, as Att. subj.

long close

;

Dor. and Aeol.,

, ,

pi.

*€, *€. ?;

-.

Skr. tasam.

in all the dialects, as

ace.

olXlos

Dor. Boeot. ay, from Soy, until;

in all the dialects, as

yay from

+ +

;

72) in Att. Ion., but

^0?, Att. ecoy.

Dor.

gen.

+

(§

Hom.

Dor.

;

*.

^ ^

^. ^,

from D, as parepos, pTJrcpoi from from from

a,

> as

;

;

;

+ y > , as Att. (inscription) a'y = •, Dor. yd from + > , Att. Aay beside Hom. in Dor. and Aeol., as Dor. olXlos, Lesb. +€

>

80

^,

from

ov, as

>

ei)

inf.

89,

as

from

euroi

+

>

e)

oluoet?.

[§

from

ei)

>

; in

all

dialects, as Att.

}

;

in Att. Lesb., but

Lesb.

from

;

^^ =

Dor. and Boeot., as Att.

^, ^.

in

Dor. Boeot.

from

*TTp(iifaTo^

;

§§

0)

Ablaut

8r-2l

, as > , as

+e >

+


+

+ + +

;

49 from

Att.

?

?;

-^.

^ ^. / ^.; ^ -. from

plycovre^

-cooi/re?.

>, >, >

+6 =

from from

in all the dialects, as

from r;

^ .^ ^

-€.

from from -?;.

as subj.

as subj.

/aiyo)

in all the dialects,

from

as

subj.

from

prim. Gr. 6)

+ 01 > , as + Of > , as

>

,

as

from

from *-€/. from

CHAPTER

IV

ABLAUT § 81. Up to this point we have treated the Indg. vowels and their equivalents in the more important languages without any reference to the manner in which these vowels stand to each other in any one language. It now remains to illustrate and formulate the manner in which they stand to each other, or in other words to discuss the phenomenon of what is called ablaut or vowel gradation. And for this purpose we shall confine our examples almost entirely to Greek, partly because it is the language which concerns us most intimately in this book and partly because, having preserved the Indg. vowels more faithfully than any other

language,

it

is

best fitted to illustrate the various

phenomena

of ablaut.

By

vowel gradation is meant -such and accentual differences in the vocalic elements of groups of etymologically and morphologically related words as were caused by sound-laws which operated in the prim. Indg. language before it §

82.

ablaut

or

quantitative, qualitative

Phonology

50 became

^, : Such

differentiated into the separate languages.

7€-, -. ? --

root-syllables of

in the

are e.g. the differences eXinov,

83

[§

. €. . , -^ , , *, ^ , . :

pes

Dor.

:

Dor.

give.

:

€€

'•

;

:

:

:

:

:

:

:

from

:

:

According as the vowels which stand

83.

Lat.

:

:

:

:

§

:

:

:

Dor.

:

:

datus Skr. da-d-mdh, we Examples in other than Goth. wulf5s, wolves,

Lat.

:

are

root-syllables

pedem

Lat.

:

:

:

:

:

:

:

:

:

in ablaut

relation to each other differ in quality or in quantity only,

or both in quality and quantity,

we have what

called

is

qualitative, quantitative or qualitative-quantitative ablaut.

Qualitative ablaut only occurs in syllables which have

the strong grade of ablaut and

.-.

ablaut

(§

as

96),

ev

It

The

ablaut.

most

to

'ip

-.

:

Dor.

:

account for this phase of

interchange between e and

and between e have been so regulated that e originally the chief-accented syllable and in the next

and 5 seems stood in

difficult

in the e-series of

:

:

:

, ,

most part confined

for the

and of e

:

:

:

is

is

^ .,

to the interchange of e

€,

to

following syllable, as in :

^,

^-^, ۥ.

:

;

Quantitative and qualitative-quantitative ablaut mostly

,^ , -=

arose through the loss or weakening of vowels in un-

^ ,,

accented syllables, as :

The

=

Indg.

:

:

:

5.

-^

^, '4. :

Skr. i-mdh

:

e^mi, :

:

:

:

Dor.

Lat.

datus

=

Indg. a

•

:

:

a.

must have been more predominant than the pitch accent at the time quantitative ablaut came into stress accent

Ablaut

§§84-5] existence, because

it is

51

only upon this assumption that we

are able to account for the weakening and eventual loss of

vowels §

unaccented syllables.

in

84. Scholars are

See

28.

§

now generally agreed

that the factors

which brought about the phenomenon called ablaut were ot various kinds. Although the prime factor was doubtless the system of accentuation which prevailed at different periods in the parent Indg. language, there were also several other factors more or less connected with accent, such as vowel-contraction, lengthening of vowels by compensation for the loss of a vowel in the next syllable, rhythmical lengthening (see Wackernagel, Das Dehniingsgesetz der gricch. Composita),

numerous analogical forma-

mixing up of the various ablaut-series through

tions, the

the influence of analogy,

&c.

And

as

all

these vowel-

changes and probably many others connected with ablaut took place long before the separate languages came into existence,

it

practically impossible

is

determine their

to

chronological order or to be certain about the precise

nature of some of the vowel-changes.

In the following

account of ablaut certain more or less problematical details

have been omitted as being beyond the scope of this book. student who wishes to pursue the subject in greater detail should consult Brugmann's Grundnss, vol. i, second

The ed.,

pp.

482-505,

pp. 138-50

;

Hirt's

and Kurzc vergleichende Grammatik, Der indogcrmanische Ablaut and the

Handbuch der

excellent epitome in his

Formenlehre, pp. 84-105. § 85. From the examples given above

seen that ablaut root-syllables

words

not confined to what

but that

For

syllables.

is

into root-bases

--

it

:

occurs

--?,

Dor.

it

-

••

and

Skr. bhari'tra-m, arm, ev

also

practical purposes

is

griech. Laut-

(§§ is

82-3)

2

be

equally in

other

-• -:

convenient to divide

ev :

will

generally called

suffix-bases, as in :

it

und

---

:

:

Lat. feri-mus (older

Phonology

52

[§

86

OHG.

bera-mes. In the following paragraphs we simply bases or ablaut-bases, and Bases or ablaut-bases are suffix-bases simply suffixes. The monosyllabic mostly monosyllabic or dissyllabic. bases are called heavy or light according as they contain •mos), shall

call

root-bases

-

a long or a short vowel, as *dhe•,

••,

Dor.

;

*es•,

*ei•

in

*•,

*bha•

-, -.

in

--,

The

dis-

heavy when the first syllable contains a short vowel and the second syllable a long vowel, and light when both syllables contain a short vowel, as *pele•, fill, *gen6•, know, *peta•, y7>', see § 458; *leiqe•, The bases underwent numerous leave, in XciVe-re vowel-changes owing to the operation of various sound-laws which took place in the prim. Indg. period. The more important of these changes were syllabic bases

are called

:

.

:

I,

The Weakening or Loss of Vowels.

§ 86. Vowels were weakened or disappeared in syllables which did not have the chief accent of the word. Such syllables are said to have the weak grade of ablaut. The weak grade is subdivided into weak grade i (wg. i) and weak grade 2 (wg. 2) according as the syllable in which it

occurs originally had the secondary accent or was unaccented.

In the former case short vowels merely became

reduced in quality (generally written e, o, a) and long vowels became reduced in quality and quantity (generally written , § 49), whereas in the latter case both short and long vowels disappeared through the intermediate stage of reduced vowels.

At a

later period in the parent Indg.

language the reduced short vowels full

quality again

and thus

strong grade vowels

fell

e, o, a.

e,

o,

a

regained their

together with the original

When

the vowel e entirely

em, en, el, er) the second element of the diphthong became vocalic or re-

disappeared in diphthongs

(ei,

eu,

Weakeuhig or Loss of

«7-9]

§§

inained consonantal according as

sonant or a vowel

in the

53

was followed by a con-

it

next syllable.

Long vowels were reduced

87.

§

Voivels

(= Aryan

to a

but a in

i

the other languges) or disappeared in the heavy ablaut-

*^

49), as ^ero? for

series

(§

Indg.

*dhatos

•, ;

=

-,

Lat.

Skr. bhavi-tum, Indg.

*bhewa•.

Skr.

(he gods,

:

da-dh-mah

sthitah,

Lat. stare,

*bhew3-tum, sing,

:

-

-dita

:

-,

be

to

da-dha-mi,

base

:

--, deva-ttah, given by

sing, da-da-mi,

with -ttah from older *-d-tos beside Lat. datus,

€- from

beside

The

88.

pi.

:

Skr.

Skr.

;

Dor.

:

Dor.

da-d-mdh

payrjvai

Aiyycu,

note), Lat.

(§

:

§

49, note), Skr. hitdh, :

for

Lat. status, Indg. '^states

pi.

(§

--, Xayapos datus, 86^ *6 49, --, d5ntim ', :

*e-bhw9t

:

base *bhewa•.

element of the long diphthongs

first

au was reduced

The

ei, 5i, ai,

9u then became contracted to i, before a following consonant already in the prim. Indg. period. But as the second element of long eu, 5u,

to

9.

9i,

diphthongs often disappeared in the parent language

we

,

thus have the ablaut relation

the earliest historic period of

i

all

dhitdh, pp., sucked, Lat. filius felare,

:

,,

ayati, he gives root

:

:

inf.

:

:

i

Skr. miirah,

i

and

and

u, as

(§

e, 5,

:

63)

a

in

he sucked, Lat.

beside Skr. pay-

patum,

drink

duU,

from

Skr.

mulam,

''

, -

stupid

;

:

(§

70),

Lat.

became unaccented they were :

-uai,

:

,

a and

the languages, as Skr.

, 6 -, . When

m5rus.

shortened to

§

drink

:

e, o,

:

Skr. ildhar, udder

:

-^

to

:

evpi

:

(-,

Lat. di-rutus

:

:

?,

89. In the light ablaut-series the short vowels e, o, a

were reduced

to voiceless (?)

the intermediate stage of

e, o,

a

or disappeared ^through

At a later period in the parent Indg. language the reduced vowels regained their full

quality again

e,

and thus

o,

a.

fell

together with the original

Phonology

54 strong grade vowels

^^,

as

e, o, a,

[§

Indg.

.

gen. sing. Lat. pedis, Skr. padah, Indg. *pedes

from

*oq•

:

-^,

;

Indg.

*aktos

Greek there are no sure examples of the light ablaut-series. is

^, .

:

*- ' :

When the vowel

In in the

:

*--

*.

from

-

^,

:

=

:

from

a

loss of o,

which stands in ablaut relation Examples of the loss of e are Skr. from Indg. *s-mes *es-mi, e5oy from *€9,

:

s-mdh ds-mi, / am Lat. pedem, '^ from

pi.

;

should also be noted that the above

It

not the same as the

to e, as in

:

90

*peqtos,

:

:

e entirely disappeared in the diphthongs

ei, eu, em, en, el, er the second element of the diphthong became vocalic or remained consonantal according as it was followed by a consonant or a vowel in the next syllable, as

( € , *^-,, ^ ' -, * ( . , ^ -., ,-;-': -, ., i'/zei/

=

:

:

from

el?

:

imdh

Skr.

xe(f ),

:

*€9

:

emi,

from

:

Hom.

from

yeveTO,

Skr. pitrsu

i-mah, we go

*ef•aa

veo-yvoi

:

;

:

•

:

:

:

;

:

:

;

Skr. y-inti, they go

:

from

;

^-,

:

'iXtnov

Xtneii/,

{)9,

=

Indg. *j-enti

:

:

*i-mes

:

:

;

:

yeroy

;

;

k-

:

:

§

90.

bases

(§

The combinations em9,

They became

(i)

ena, el3, era

, , ,, , ,

97) a threefold

development

when

had

in

the

heavy

Greek.

prim.

in

element

first

pa (see § 68) when had the secondary accent, (2) the last element had the secondary accent and the first pa when neither element disappeared, and (3) the first nor the last element had the secondary accent and (e)J9, (e)w3 became i, u, as (Hesych.) from Lat. latus Dor.

*

,,,

:

;

:

.

from

:

;

*

;

:

, .

, , -, •-€^ ^?; • ^,, Lengthening of Vowels

91-2]

§§

€0•8^

Dor.

:

:

:

:

-^.

bhavitum,

/o

Z>^,

Lat. (g)natus

;

Lat. lana.

Dor.

re-

;

:

:

:

base *weje•

;

e-0u

:

Skr.

;

inf.

base *bhewa•.

The Lengthening

2.

55

of Vowels.

kinds of vowel lengthening took place in the prim. Indg. period, as lengthening by compensation for the loss of a syllable, contraction of vowels and rhythmical § 91. Several

lengthening. See Streitberg, Indogermanische Forschungen, pp. 305-416.

iii,

§

With

92.

quantitative ablaut

is

connected the prim.

Indg. lengthening of vowels by compensation for the loss

The vowels

of a syllable.

thus lengthened have what

is

The vowels

in

called the lengthened grade of ablaut

nearly

all

the examples which have this lengthening belong

And

the lengthened vowels e, and Ig. 2. short accented vowel in an originally open syllable

to the e-series of ablaut.

are respectively called {a)

(Ig.).

A

became lengthened

if

Ig. i

the following syllable

This occurs especially

appeared.

nouns, as Lat. pes. Dor.

s'from

in the

entirely dis-

nom. singular of

prim. Indg. *pets, *pOts,

8,

, , , , , , ,? , , .

*p6des or -os, *podes or -os, beside ace. pedem, from prim. Indg. *p9tere Indg. *pedm, *podm; Lat. Indg. *p3term; and similarly beside :

ferus,

:

:

:

base *kered•, cp.

Note.

— Also

{b)

The

e

when a

€

was

:

:

short vowel disappeared after a long

from an original form *dhwesaso

also lengthened in prim.

singular of

Hom.

:

:

:

vowel, as in gen.

active

:

the s-aorist,

as

:

nom.

Indg. in the

*leksm from

older

Phonology

56 *legesm,

cp.

Lat.

lexi

:

pres.

legit

Lat.

;

[§§

93-5

vexi,

Skr.

a-vaksam pres. vehit, vahati. The s-aorist in Greek was a new formation with the vowel from the present, as :

^-^,

e-Ae|a,

but Skr. i-raiksam, see

§

507.

augment with a following vowel, as in rja (§ 79), Skr. asam, Indg. *esm from older *e-esm ^a for Skr. ayam, Indg. *ejm from older *e-ejin Dor, ajami. Skr. ajam pres. §

93. Contraction of the

;

,

*,

The

',

;

the dat.

nom. plural of a-stems -oi from -o-ai in singular of o>stems -os from -o-es in the nom. in the

;

;

plural of o-stems, see 94. Rhythmical

§

79.

lengthening

in

the

first

elements of

, ^, ^^-, '^

compounds and before

suffixes so as to avoid a long suc-

cession of short vowels, as

Hom.

as -as

contraction of case-endings with the stem,

from -a-es

§

,

:

:

:

Up6^,

erepos.

Ablaut-Series. §

95.

The vowels vary

within certain series of related

The parent Indg. language vowels called ablaut-series. had six such series, three light and three heavy, viz.

Ablaut-Series

§ 96]

57

lengthened grade have already been explained ceding paragraphs.

And some

indication of the probable

origin of the difference between strong grade

grade 2 has been given

in the pre-

much

in § 83, but

i

and strong

still

remains

obscure about the origin of these qualitative differences.

The is

first

found

syllable

in

ablaut-series

is

by

far the

of dissyllabic heavy

syllable of dissyllabic light

grade of this series occurs

in

alwaj's in the

It

first

and in the second and nearly always in the bases. And one or other

bases

syllable of dissyllabic light bases first

most important.

many monosyllables and

nearly

all

suffixes.

The

second and third series are exceedingly rare. Apart from a few monosyllabic heavy bases the fourth, fifth, and sixth

heavy and even here the number of examples is not very great. The e in the fourth series often came to be regarded as a formative element in prim. Greek and was then extended by analogy to bases to which it did not series only occur in the second syllable of dissyllabic

bases,

originally belong, see §§ 458, 506. § 96. Many examples of the various grades of ablaut have been given in the preceding paragraphs. In this and the following paragraph are given examples of the

various ablaut-series, and of their application to dissyllabic light

and heavy bases.

58 sg. I.

(€-€

Phonology

96

Dissyllabic

97]

The

III. sg. 2.

sg. I.

Wg.

Ig. 2.

a Lat.

I.

wg.

2.

a

amb-ages

The

IV. sg.^i.

59

a-series,

Ig.^ I.

^

4

Bases

-a/CToy

e-series.

wg.

sg. 2.

e

wg.

I.

2.

3

'*

^eroy for *0aroy

Skr. d4-dha-mi i

-

--V.

for

The

o-series,

wg.

sg. 2.

3

^oro? for Lat.

da-dh-mdh

hitah

donum,

wg.

I.

—

*?

Skr.

2.

da-d-mah

Lat. datus,

VI. The a-series. sg.

I.

wg.

Sg^2.

I.

^ '^,
aL.

r)

of representing these nasals in

which is common on inscriptions. They Greek was by came to be represented by after the combinations gn, gm as in Att. had become assimilated to rjn, (§ 189),

yyvoaL,

.

y6s = ytf;voaL,

?;$.

It

has already been

pointed out that these nasals only occurred in the parent

language before their corresponding explosives and underin the different languages all changes in the place of

went

articulation in

,

•qveyKa,

I

common

bore, Skr.

with these explosives

an^sa, he

Lat. ango, cp. Goth,

,

obtained, cp. Lat.

aggwus,

(§

narroiv.

,

139).

nanciscor :

:

;

ayKOiv,

O.Lat. ancus, cp. Skr. arjkah, hook, OE. din%Q\, fish-hook 7T€VTi,

Lesb.

pimp, Goth,

Skr. pdnca, Lat. quinque,

fimf, Lith. penki, five

cp. Skr. rincanti, they leave.

;

O.Welsh

Lat. linquo,

The Labials

156-9]

§§

156.

The

,, , §

preceding vowel before Ion.

' Xy

=

Indg. ghj

129,

(§

Ion.

;

89

nasal disappeared with lengthening of the

Att.

Att. ^arrooj/,

,

from prim. Greek from *ay\jov, cp.

Att.

as Ion.

7),

from

^^'.

from

*€//

also disappeared, but without lengthening of the pre-

It

from

ceding vowel, before *K\ayyjoi

plango

:

from prim. Gr. yj, as clango from *nXayyj(u gen. adXntyyo^.

Lat.

;

;

§

In O.Ir.

,

medially, as

Greek

before

TToAiiy,

;

;

and generally also

Skr. pat (gen. padah), Lat. pes,

Skr. pra, Lat.

,,

Skr. puruh, O.Ir.

Lat. plecto

,

initially

Skr. pitar-, Lat. pater, O.Ir. athir,

,?,

fdsder, faiher

foot]

Labials.

disappeared

it

vowels) remained in

, ^,

Lat.

Skr. Lat. Lith. O.Slav, p, Germanic f, initially and medially between

(=

157. Indg.

OE. OE.

:

:

The

b.

Ion.

;

;

il,

pro-,

Goth,

O.Slav, pro-,

filu,

much,

many

;

Skr. sapta, Lat. septem, Goth,

,

Skr, sarpami, Lat. serpo, / creep OE. ofer, over, Skr. upari, above Skr. svapnah, Lat. somnus from *swepnos, O.Ir. suan, sleep, OE. swefan, sleep Lat. caper.

sibun, seven

§

-

before

, ;' ;

;

became

158.

became

,

,^, as

as

(see

:

bystealih

:

§ 117).

:

•-

(nom. 1.), f/ie day after the feast, where the weak form of *^Q6.;foot, cp. Skr. upa-bd4•, stamping, ;

is

;

Lat. s-uper,

See

trampling.

§

p)

107.

,

159. Indg. b

manic

§

(=

remained

medially, as

Skr. Lat. O.Ir. Lith. O.Slav, b, Gerin

Greek

initially

and generally also

blustering, Skr. huk-ka.ra.h, (he roaring

,,,

Phonology

90

of a

Lat. biicina, trumpet,

lion,

bellow

;

I

libo

O.Slav, bucati,

roar,

to

Skr. barbarah, stammering

foreign,

dance, Skr. bal-baliti, he whirls

ambu,

Lat. imber, cp. Skr.

;

160-3

[§§

;

^,

water.

It

Lat.

should

be noted that b was a rare sound in the parent Indg-

,, ,

language.

On

160.

§

as in as

the change of

see

:

§

before voiceless sounds,

to

/3

see

:

§

became

106.

,

117.

ph

, ,

ph was one of the It was preserved language. § 161.

prim. Latin

fell

it

rarest sounds in the parent in Sanskrit

and Greek, but

in

together with original bh, and in prim.

Germanic and the Baltic-Slavonic languages with I crack, crackle, Skr. sphurjati, Skr. he cracks Skr. sphyah, wedge Keltic,

original p.

;

phalakam,/oo/stoo/, see

§

?,

;

102.

bh §

bh (= Skr. bh,

162. Indg.

Germanic

dially,

voiceless

member of a brathir, OE.

, ,

,

in

O.Ir. berim,

Greek, as

OE.

here, O.Slav, bera,

§

nail,

bro))or, brother;

163.

On

OHG.

,

the change of

tion of

,

:

as in

,,

^,

^^,

nebul, mist;

jambhah, became as Skr.

/ bear;

,

Skr. bhratar-, Lat. frater, O.Ir.

bru, Lith. bruvis, eyebrow; Lat. nebula,

Lat. f initially and b meand Baltic-Slavonic b) became Skr. bharami, Lat. fero,

,

b, b, Keltic

OE.

Skr. ndbhas-,

Lat.

tooth,

,^,

Skr. bhruh,

cloud,

umbilicus

;

OE. camb, comb. :

see

§ 117.

before voiceless sounds, as in

to

see

:

§

109. see

On

§ 115.

the de-aspira-

The Dentals

64-6]

§§

91

The Dentals. t

(=

164. Indg. t

§

Germanic

Skr. Lat. Lith. O.Slav,

O.Ir.

t,

th,

t,

Indg. combinations pt, kt, qt,

\, d, but t in the

generally remained in Greek initially and medially, as

st)

OE.

rdviuy Lat. tendo, stretches

)jenne,

/

stretch,

Skr. tanoti, he

Lat. tenuis, O.Ir. tana, Lith. tenvas,

;

OE. )jynne, 6, Skr.

thin; raros, Skr. tatdh, Lat. tentus, stretched;

OE. ]?aet, the, that; rpeh, OE. J)ri, O.Slav, trije,

,, ^,

tdd, Lat. is-tud,

Skr. trdyah,

Lat. tres, O.Ir. tri, Lat.

tremo

OE.

faeder, /rt//!^r;

^,

Skr. bharatha, Lat. (imper.)

ferte, Goth, bairijj, O.Slav, berete,

vetus

three;

Skr. pitdr-, Lat. pater, O.Ir. athir,

;

ye bear;

eros,

Lat.

Skr. srutah, Lat. in-clutus, renowned, O.Ir.

;

,

^^ , , *\ * *, , , ^,; * .€ *' * ?,

cloth, renown

.

esti, he is

;

;

eari, Skr. asti, Lat. est, Goth, ist, Lith.

Goth, hliftus,

thief;

OE. eahta, Hom.

asta, Lat. octd, O.Ir. ocht, §

§

165.

7 became ,

166. Prim.

as

Greek

Skr. astau,

eight.

from

:

109) became a double spirant

(§

Most is uncertain. was ]?])(= th in Engl, thin) or a kind Before and after consonants, and finally it

the precise pronunciation of which

scholars assume that of lisped ss.

became

through the intermediate stage

Hom. *widswos

dialects, as

older

it

Jaos, Att.

;

, -,

from

;

Att.

;

:

;

in

'^,

eVepcrafrom

*69

;

Cret.

*€^ ^,

-, -,

the dialects, as dat. :

;

pi.

beside stem

-, -, -,

Medially after long vowels and diphthongs

all

nom.

:

yiyas, SiSovy,

veOTt]T-,

the

all

from *fiTafo^, from from from

far/^os-

Cret.

Att.

:

:

?, Cret.

'4

from

from

;

:

it

became

in

from

Phonology

167-9

, , , , ,

92

it became and

Medially between short vowels

,

in Att. Ion.,

Cret. (Cret. also ),

[§§

Horn,

dialects, as aor.

in Boeot.

in the other

Cret.

Att.

:

loc.

Skr. patsu, beside nom.

Horn,

pi.

^, ;

, , */, ^ ,, , , 167. Indg.

§

same

the

became

tj

*navTJa

from

;

129,

*6/

168. Initial

§

-,

as

tw- became

,

from from

;

Att.

6.

later to

§

168, as

Hom. from

Thess. Cret.

Lesb.

Lesb.

§

Greek and then had

ts in prim.

further development as the ts in

Att. Boeot.

-

which

Skr. tva, tvam,

-- in Att. and Boeot.,

as Att.

€€9,

Boeot.

and

fhcc.

;

Hom. See

*tOtJos.

was

simplified

Medial -tw-

^,

-- in the other dialects,

^,

became

Att.

Skr. padah.

Hom.

, ', , .

Skr.

catvarah,/o«r. For further examples see § 124, 3. § 169. Ti remained initially and also in the combination Medially it partly as and partly remained. The reason for this twobecame

,

fold

P.

treatment has never been satisfactorily

Kuhn's

Kretschmer

Zeitschrift, vol.

explained.

565-91—

xxx, pp.

after investigating the subject in great detail, arrives at the

remained medially when the

was when the accent was on the when the was unaccented. penultima, but it became On the other hand Brugmann Grundriss, vol. p. 662 following result

tl

:

l

accented, and also finally

* ,.

l

i,

*

assumes that the became consonantal before vowels, as in from gen. Ion. from and then the was levelled out into the nom. and ace. sing. l

;

A careful examination of the material by Kretschmer shows that both explanations leave a large residuum of unexplained forms, even when due allowance is made for a considerable number of analogical The ti«stems and likewise the adjectives in formations. collected

^' have ,

as

,,,

;

,

170-5]

§§

The Dentals

^.

(^,

The pi.

of oo-verbs have

• in Dor. and Boeotian, as Att.

but

beside Dor.

became

170.

§ 171.

manic

t)

before a following

from *K€ut-t6^

;

uttarah,

-

, , ^. , ^ ^,

verbs and the third pers.

§

93

See

latter.

(=

Indg. d

§

of

pres. third pers. sing,

,

:

-

in Att. Ion.,

as

:

Skr.

;

110.

Skr. Lat. O.Ir. Lith. O.Slav, d, Ger-

,,

generally remained initially and medially in Greek,

as SeKa, Skr. dasa, Lat. decern, O.Ir. deich, Goth, taihun,

OE. tien, Lith dezimt•, fen I give, O.Slav, dati, to give; domus, O.Slav, domu, house; ;

Skr. dddami, Lat. do,

Skr.

/zoy,

damah,

Lat.

Skr. dvau, dva, Lat.

,,

,

duo, O.Ir. dau, d5, Lith. du, OE. twa, two; Skr. ddmi, Lat. edo, OE. ete,I eat; e^oy, Skr. sadas-, si-rt/, Lat. sedere, OE. sittan, to sit; Lat. gen. cordis,

OE. heorte, heart; veda, OE. wat, he knows, Lat. videre, to see; O.Ir. cride, Lith. szirdis,

padam,

€, ace.

§

,

172.

;

sella,

from *sedla, OE.

,

became

as

setl, seat.

Hom.

*f6^, quontodo. §

from

=

*•

Lat.

older

^.

became , as Zeuy, cp. Skr. dyauh, sky padyah, on foot. See § 129, 8. became before a following voiceless dental, as See § 110.

173. Indg. dj

^, §

Skr.

*•,

Lat. pedem, OE. fbt,/oot. became as from bandage worn by runners on the ankle Lac. iXka

Skr.

,

Skr.

174. :

;

:

th §

175. th

was

was preserved it

fell

a rare sound in the parent language.

It

Sanskrit and Greek, but in prim. Latin together with original dh, and in prim. Keltic, in

,

Phonology

94

,

[§§

Germanic and the Baltic-Slavonic languages with t.

or mould

Skr. vettha, thou knowest to

bake

§

176. Indg. sth

?, -? =

suffix

,

became

?,

,

Skr. prthuh, broad;

in,

Skr. manthati, he twists, shakes, see

/ stand;

;

as

§

176-81 original

a platter battle-din,

102. Skr. tisthami,

Skr. sturah, strong; superlative

pillar,

Skr. -isthah, see

§

102.

dh §

177. Indg.

dh (=

before and after d, Keltic

became

r,

,

Skr. dh, Lat.

before

and

1

f initially,

u

after

and Baltic-Slavonic languages

voiceless

in

Greek, as

b medially

(w), in other cases d,

Germanic

d, d)

to suck,

Skr.

dhayanti, O.Ir. denait, they suck, Lat. felare, Goth, daddjan, to suckle courage, passion, Skr. dhumdh, OE. duru, pi. Lat. fumus, Lith. pi. dumai, smoke; Skr. dadhami, I put, Lat. fores, Lith. durys, door place, OE. dad, deed, Lith. deti, O.Slav, deti, to lay, cp. wine, Skr. madhu, O.Ir. mid, OE. Lat. facio, feci; ;

^,

;

,

,,

,

,

Skr. medus, mead, honey I burn, Skr. rudhirah, Lat. ruber, OE. read, red; edhas-, fire-wood, Lat. aedes, sanctuary, originally, fre-

medu, meodu,

Lith.

;

,,

OE. ad, funeral OE. iider, udder;

place, hearth,

Lat. uber,

=

178. Indg. dhj became

madhyah, §

§

,,

as

On

dh appears :

as

^,

Skr. udhar,

, ?,

Lat. fido

middle, see § 129, 6.

179. Indg.

180.

;

Skr. srudhi, hear thou.

, §

pile

;

,

imperative Skr.

before voiceless dentals, as see

the de-aspiration of

§

110.

,

as in

see

§

115.

The Palatals and the Velar Gutturals. § 181. In treating the history of these consonants in the

various languages

it

is

convenient to divide the Indg.

family of languages into two great groups according to the

§

8

The Palatals and Velar Gutturals

95

development which these sounds underwent in The palatal explosives k, kh, g and gh usually appear in Greek, Italic (Latin, Oscan, Umbrian), Keltic and the Germanic languages as explosives or as sounds which are directly developed from explosives, whereas in the Aryan, Armenian, Albanian and BalticSlavonic languages they usually appear as spirants. The former group is generally called the centum- and the latter the satsm-group of languages, where Latin centum and Zend satam represent the original Indg. word *)imtom, hundred. This twofold development of the palatals is probably due to dialectal differences which already existed in the Indg. different

the two groups.

parent language.

In addition to the palatals the parent

language also had two kinds of velars,

pure velars and The pure the palatals in the centumviz.

labialized velars or velars with lip-rounding.

velars

fell

together with

languages, but were kept apart in the satsm-languages.

On

the other

hand the pure velars

fell

together with the

labialized velars in the satam-languages, but

were kept

apart in the centum-languages.

The

following table contains the normal development

of the palatals and velar gutturals in the various languages. The labialized velars are here indicated by ^. In other parts of this grammar the ^ is almost always omitted as

being unnecessary, because the Greek words themselves generally indicate whether they originally contained a pure velar or a labialized velar.

96

Phonology

[§

'8i

The Palatals

182-7]

The Palatals.

I.

§

182. Indg.

k (=

O.Slav,

sz,

Lith.

,,

,

Lat. O.Ir.

c,

remained

s)

generally also medially, as

OE. hund,

97

Germanic h g, 5, Skr. s, in Greek initially and Lat. centum, O.Ir. cet, ;

, ?,

Skr. satam, Lith. szimtas, hundred)

OE. heorte, Lith. Goth, hlifa, / siea/

Lat. cor, O.Ir, cride,

heart;

,

Lat. clepo,

;

Lat. in-clutus, Skr. srutah, renowned, O.Ir. cloth,

Lith.

Lat. canis, O.Ir. cu,

,

OE. hund,

szu,gen. szuns, dog, hound; dasa, Lith. deszimt-,

Lat. vicus,

village;

,

oIkos,

;

folKos,

Lat.

Lat. 5cior. §

, §

:

;

,

,

;

;

decem, Goth, Skr.

Skr. vesah, house,

OE. eahta,

?,

, ^,

kw

as became See § 124, 184. Medial kj became

183.

Skr. dsvah, horse.

Lat.

fen

oct5,

astau, asta, Lith. asztuni, eight;

renown

Skr. sva, gen, sunah,

taihun, Skr.

daddrsa, he has seen

szirdis,

Skr.

Skr. asuh, quick,

Lat. equus,

OE. eoh,

2.

in Att.

and

in Ion., as :

root *pak•.

were simplified to , and initially, as Att. from '^].(.. See § 129, 7. before a liquid or nasal, as § 185. ks became from stallion-ass Lat, mulus Ao|oy, Lat. luxus from Lat. aranea from from Lat. luna from *louksna from See § 218.

., and

Ion.

*?, ;

;

;

§

186. ;

. §

;

187.

** . * *€ * . ^ *, :

:

;

:

:

;

k disappeared from {rom*

became See § 107.

;

before sk, as :

eoiKa

;

from

:

from

:

before voiced sounds, as

:

Phonology

98

[§§

188-92

g § z,

(=

188. Indg. g

Lat. O.Ir. g,

Germanic k, Skr. j, Lith. initially and generally

O.Slav, z) remained in Greek as ykvos,

medially,

also

Lat.

,

genus, Goth, kuni, Skr.

jdnas-, race, generation, O.Ir. gein, birth

;

y

Lat.

OE. ceosan, to choose, Skr. jusate, he tastes yovv, Lat. genu, OE. cneo, Skr. janu, ^«^^ Lat. (g)n5tus, O.Ir. gnath, Skr. jnatah, knoivn, OE. cnawan, gusto, /

taste,

;

O.Slav, znati, Skr. kjr&h,

djami, /

weorc,

drive-,

ay ,

acre;

Lat.

kyoi,

189.

came

The §

awe See

ic,

/;

milk, Skr. mrjati, he wipes, rubs.

as in yiyvmaKOi,

yv,

This explains

why the

guttural

155).

At

on inscriptions in Ion. already in the about the end of the fourth century b. c.

fifth

be expressed by

to

was

t/v

in

simplified to

v,

Greek as

(§

yu, yuoa.

u occurs in Att.

190. gj became , as from of, Skr. ydjate, he honours

*/, I

;

§

129,

became :

from

stand in

*ap-nayj(u.

8.

§ 191.

XiXcKTai

OE. OE.

'ipyov,

M.Ir. bligim,

yv,

a later period

and

OE.

Lat.

;

yyvoa, arvyvo^, y6s. nasal

Lat. ago, O.Ir. agim, Skr.

ego,

mulgeo, iy, melzu, / Medial y became ^,

ivork

meolce, Lith. §

aypSs, Lat. ager, Goth, akrs,

knoiv;

to

field,

;

/?,

iy.

before voiceless consonants, as

See

§

^,

106.

kh §

kh was one

192.

was generally preserved

Indg. language.

It

in prim. Latin

fell

it

of the rarest sounds in the parent in

Greek, but

together with original gh, and in the

prim. Keltic, Germanic and Baltic-Slavonic languages with original k.

become

It

is

doubtful what simple

in Sanskrit,

because

combination skh which medially.

^,

it

kh would have

only occurs in the original

became ch initially and cch / split, Skr. chindtti from

Lat. scindo,

The Pure Velars

193-5]

§§

*skhinatti, he

OE. scadan,

splits,

Skr. chydti, he

slits,

see

divide

to

,

;

102.

§

99 /

slit,

gh §

gh (=

193. Indg.

h medially between vowels and

Lat.

also initially before vowels except u, f initially before u,

g before and

,

Skr. h, Lith.

O.Slav,

z)

,

Skr. himah, snow; zema, O.Slav, zima,

, Lith.

O.Ir. g, Germanic g, 5, became voiceless in Greek, as

after consonants,

z,

on

the

hiems, O.Ir, gaim,

Lat.

heman,

winter, Skr.

in winter;

ground, Lat. humus, Lith. zeme, O.Slav.

zemlja, earth, ground;

fundo, OE. geote,

,

Lat.

x^{F)%

I pour, Skr. hutkh, poured, sacrificed; OE. g5s, Skr. hasah, goose;

^,

Lat. (h)anser,

Lat. lingo, O.Ir.

/

ligim, Skr. rehmi, lehmi, Lith. leziu, O.Slav, liza,

Pamph. /^€,

Lat. veho,

OE. wege,

vezu, O.Slav, veza, / bear, carry, move

, . I press

tight,

,

lick

Skr. vahami, Lith. ;

Lat. ango,

OE. enge, narrow;

.

Skr. ahas-, need, distress,

fore-arm, Skr. bahuh, arm.

§

ghj became

194.

Ion.

:

2.

in Att.

^,

and

Ion., as Att.

in

See

129,

§

7.

The Pure Velars. q

195. Indg. q (= Lat. O.Ir. but c before i, and a Indg. §

k

=

before palatal vowels)

,

became

generally also medially, as

karkatah, cmo fruit, Lat.

;

Germanic h in

Greek

^,

I

celer,

quick;

drive,

,

k

initially

but c

and

Lat. cancer, Skr.

Skr. kalika.,fower-bud

carpo, I pluck, pick,

4,

;

k, O.Slav,

OE.

;

haerfest, autumn, Skr.

krpanah, sword, Lith. kerpu, / shear; I call, call out, Lith. kalbk, speech; beautiful;

g, 3, Skr.

, , , c,

e, Lith.

?,

Lat. calo,

Skr. kalyanah,

Skr. kalah, black, cp.

Lat. caligo

Skr. kalayati, kalayati, he drives,

coUis,

Lat.

2

hill,

Lith.

;

Lat.

kalnas,

Phonology

loo mountain

;

kravis•,

raw

,

OE. hra(w),

Kpias, flesh,

gore;

blood,

Skr,

carrion,

corpse,

O.Lat. ancus, Skr. arjkas-,

oiyKos,

anka,

knot

loop,

Skr. yuktah, Lith. junktas, yoked ;

. , young man. 196. qj became

§

196-200

meat, Lat. cruor, O.Ir. cru, Lith. kraujas,

bend, hollow, Lith.

boy,

[§§

Ion.

;

^,

and See

in Att.

Lat. junctus,

/^, Skr. maryakah,

root *plaq•.

in

Ion., as Att.

129,

§

7.

9 §

g (= Lat. O.Ir.

197. Indg.

before

i,

and a

palatal vowels)

=

Indg.

became

gerve, crane; ay as,

ay

in

medially, as ykpavos,

also

,

assembly,

grus,

Lat.

OE. cran,

I assemble, Lat. grex,

herd, flock,

gramah, croivd (vyov, yugam, O.Slav, igo, yoke;

O.Ir. graig, herd of horses, Skr.

jugum, Goth, juk, Skr.

Lat.

,

aTeyo?, reyoy,

house

:

OE.

Lith.

])aec,

Lat. tego,

;

stogas,

*yj ya,

from § 198. gj became , as Skr. tigmah, pointed, sharp, Lat. in-stigo. §

199.

,

became

augu, I groiu,

O.Ir.

roof,

teg,

I cover, Skr. sthdgati, sthagayati,

he hides, conceals.

iy;

Lith.

Skr. agas•, offence, crime, sin;

guilt,

ydp,

Germanic k, Skr. g but j before g but Greek initially and generally g,

Lith. g, O.Slav,

e,

:

See

§

before voiceless consonants, as :

8.

:

augeo, Goth, duka, Lith.

Lat.

increase, add.

cp.

^

129,

See

§

106.

qh 200. qh was of rare occurrence in the parent language. became in Greek, kh but ch before i and a = Indg. e

§

It

in Sanskrit

;

in prim. Lat.

it fell

,

together with original gh,

gh, in prim. Keltic and Germanic with original k, q, in Lith. and O.Slav, with original q. I laugh, Skr.

kakhati, he laughs Koyxos, Lat. congius measure), Skr. sarjkhah, muscle, see § 102. ;

[a

small liquid

The Labialized

'

Velaria

9h 201. Indg.

§

h

Lat.

e,

gh (=

initially

Skr.

gh

but

before and

h before i, and a

=

Indg.

medially between vowels,

g before and after consonants, O.Ir. g, Germanic g, 5, Lith. g but before palatal vowels) became voiceless

g, O.Slav,

?,

, ,

\8,

brass, Lith.gelezis, iron X in Greek, as I lay hold of, Lat. pre-hendo, / seize, Goth, bi-gita, I find,

O.Slav, gadaja, I giiess, divine; lang, Skr. dirghah, long ;

?,

OE.

ligan, O.Slav, lezati, mist, Skr,

stige,

/

meghah,

to lie

cloud;

dozvn

^, ;

;

Lat. longus,

O.Ir.lige,