195 35 144MB
English Pages 582 [593] Year 1888
E L E M E N T S OF
THE
COMPARATIVE OF
GRAMMAR
THE
I S DO-GERMANIC LANGUAGES.
Da m u s s sieh manches Rätsel lösen. Doch m a n c h e s Rätsel knüpft sich a a c h . Goethe's Faust, Part I.
ELEMENTS OF
THE
COMPARATIVE GRAMMAR OF
THE
INDO-GERMANIC LANGUAGES. A CONCISE OF OF S A N S K R I T , OLD GREEK,
EXPOSITION
THE
HISTORY
O L D I R A N I A N ( A V R S T I C AND LATIN,
UMBRIAX-SAMNITIC,
GERMAN, L I T H U A N I A N
OR,D P E R S I A N ) , OLD IRISH,
AND O L D
OLD
GOTHIC,
AHMKNIAN, OLD
HIGH
BULGARIAN
HY
KARL BRUGMANN, PROFESSOR OF COMPARATIVE P H I L O L O G Y IN THE UNIVERSITY OF
VOLUME I: I N T R O D U C T I O N
A N D
P H O N O L O G Y .
T K A N B L A T E U F R O M TUP) G E R M A N BV
JOSEPH W R I G H T , PH. I).
STKASSBURGr. KARL
J. T R Ü B N E R . LONDON.
TRÜBNER & CO.
1888.
LEIPZIG.
PREFACE TO THE ORIGINAL EDITION. ose who have impartially followed the development of comparative philology in the last twenty years will be aware of the great progress it lias made in the interval. In both the scope and the nature of its work it lias shewn all the elasticity and creative vigour of a science that is still young in spite of its seventy years. That its diverse and scattered details need to be once again brought together under one systematic arrangement will hardly be doubted by any competent judge. The first edition of Schleicher's excellent Compendium* appeared in 1861, and was since twice published in a revised form by the author, the second time shortly before his death. When it was edited finally in 1876 with very few changes indeed, the two editors were already of opinion that it required complete remodelling (see the preface p. IX). Thus 1 feel sure that the 'Elements', the first volume of which is now before the reader, meets a real need. Time and competent criticism will decide whether it has solved the problem in any degree satisfactorily and whether it can claim to be of the same service to our science and its students as was Schleicher's book in its time. My task was attended by all manner of difficulties and I am fully conscious that in some respects it is inadequately discharged. I trust nevertheless that the result may be of some service, at least for the present. I take this opportunity of drawing especial attention to one or two points in the plan and execution of my work. It was not my object to collect all the various views, often widely divergent, which have been from time to time put forward by authorities, on questions relating to the histoiy of the Indo-Germanic languages. As a rule I give only the views that I consider right or at least probable, after submitting them to repeated tests. A mere regard to the size of the book necessitated this course. The most important authorities will be found summarily enumerated in the introduction and at the beginning of the separate sections. To mention in each case who has treated the subject, and who was the first author of the view I have accepted, seemed to me unnecessary to the purpose of the book, and excluded by the small space at my disposal. I took what was good wherever I was sure I had found i t ; all parties among the different methods and schools of linguistic science are fairly represented in these pages. Hence though I may per-
VI
Preface to the Original Edition.
haps have overlooked much that is useful, I still hope that approximately at any rate I have attained the aim I had set before me: to exhibit the present state of our knowledge in a concise form, giving prominence to all the more important points. At the same time I have not confined myself to stating such results as seem certain to stand for all future time. I have spoken of many problems that are still unsolved, in order to give the reader not only a survey of what has been already accomplished, but also a glimpse of the work that still remains to be done. This was indispensable in view of the position in which the scicnce of the Indg. languages at present stands. Thanks to the discovery of many fresh sources of information, and still more, I think, to the fertile combination, which the past ten years have brought about between minute investigation on the one hand and the philosophy of language on the other (the character of which I have endeavoured to sketch in my essay, 'Zum heutigen Stand der Sprachwissenschaft' p. 38 fl\), new problems to be solved have of late arisen on all sides, and that in such numbers that many decades of work will still be needed to master them, so far as we can reckon that they are to be solved at all. 1 ) Had I silently passcd over all such unsettled questions in the history of the Indg. languages, the picture of the whole subject which I tried to draw would have been marred by many grievous blanks. But further I hoped to incite the reader to independent investigation, by referring to much that has not yet passed the stage of problem and hypothesis. In doing so I have taken care that what is uncertain should not preponderate over what is certain, and should everywhere be clearly distinguished from it. And if, instead of merely marking the phenomenon in question as unexplained, I have often hazarded a conjecture, at all events in something like the direction in which the solution of the riddle is to be sought, despite those who resent every 'probably' or 'perhaps' in scientific works, I may quote Goethe's words: 'The opinions you venture, are like the pawns you move forward on the chess-board; they may be taken, but they have introduced a game that will be won*. The conjectures I have admitted into this book should always be regarded only as challenges to more minute investigation. As regards the arrangement of the material my aim is to let the different branches of language and the separate languages appear each as a unit complete in itself on the common background of the Indo-Germanic primitive community, yet in such a way that each single phenomenon appears separated as little as possible from the kindred examples in other languages. My method of exposition takes about the middle course between that of Bopp's Comparative Grammar on the one hand, which may 1) As regards exceptions and irregularities it has now become the effort of all scholars to aim at seeking for the reason of the exception, not occasionally only, but in every case, and systematically, and we consider our duty to science undischarged until we have found the answer to the 'why'. How many phenomena of language, which once were thought perfectly clear, so that a final judgment' could be pronounced upon them, have by this means become unanswered problems!
P r e f a c e to the Original Edition.
vn
be described as a connected picture sinking what belongs to t h e separate languages in the Indo-Germanic w h o l e , and that of Schleicher's Compendium, on the other, which does little more than string t o g e t h e r It seeks to unite the excellences of a number of separate grammars. both. This of course could not be done without some sacrifice of system. In t h e phonology I was obliged to add a series of chapters on combinatory sound-change a f t e r tracing the development of the primitive IndoGermanic sounds one by one. H e r e I have had now and then to repeat in sum what had already been mentioned or to expand and supplement what b e f o r e h a d been merely hinted at. I p r e f e r to submit to this incongruity of a r r a n g e m e n t , which Schleicher has avoided, than to disregard consciously t h e considerable didactic advantage that it a f f o r d s ; and in this point at least I hope to h a v e on my side the academical t e a c h e r s who lecture on comparative g r a m m a r . In the transcription I h a v e been as far as possible conservative. T h e r e a d e r will easily perceive why I was obliged to discard one or two symbols widely in use, e. g. t h e representation of the Sanskrit palatal and cerebral sibilants by g and s (or sh). I ought p e r h a p s in general to h a v e paid even more attention than I h a v e done to t h e rule that the same sound in different languages should be r e p r e s e n t e d by t h e same s i g n : e. g. Av. 3, not y, on account of the Germanic (Ags.) 3 ; Gothic w, not d, on account of t h e Old H i g h German w. And in several other points t h e notation might certainly be improved, at least if it be t h o u g h t desirable to sacrifice what is widely or universally in use in favour of less usual methods. As r e g a r d s t h e Aryan languages and Armenian especially, I gladly take this opportunity of expressing a wish, which, I think, most IndoGermanic scholars share with me, that t h e Congress of Oriental scholars ') should sanction by its authority as soon as it is practicable to do so, some suitable system of transcription for these languages, which should h a v e r e g a r d to t h e needs both of special philologists and of comparative scholars, so t h a t the vagaries of t h e prevailing usage in t h e m a t t e r may be a t last brought within bounds. T h a t t h e question of transcription is not yet ready for a final settlement is no valid objection. My work has been least independent in dealing with t h e p h o n o l o g y of Iranian and Armenian. H e r e I could generally connect it very closely with the excellent works o f B a r t h o l o m a e and H i i b s c h m a n n , especially with t h e former's H a n d b o o k of t h e Old Iranian Dialects, a n d t h e latter's Armenian Studies. I h a v e only deviated in a very few points from Hiibschmann's t r e a t m e n t of t h e history of t h e Armenian sounds. In the Keltic branch many will look for a more detailed account of t h e British dialects. I fully admit that comparative philology must devote more attention to these dialects than has h i t h e r t o been t h e case. But my knowledge in this d e p a r t m e n t is too small for me to v e n t u r e to set a n y ') T h e a t t e m p t m a d e in t h e fifth Congress (see ' V e r h a n d l u n g e n des fünften internationalen Orientalisten-Congresses', Berlin 1881, p. 89) has unfortunately led to no result.
Yin
Preface to the Original Edition.
dialect of the British group side by side with Irish. It was my intention two years ago to make myself so far familiar with Cymric as to be able to treat the chief features at least of this language along with the rest. My removal to Freiburg however at that time cut me off almost entirely from all Keltological literature, and I was thus obliged to abandon my design. As regards Irish also the want of a library') sufficiently furnished with philological literature caused me difficulties in many ways. I am hcnce all the more thankful to Prof. T h u r n e y s e n of Jena [now in Freiburg], who not only sent me information on many points of Keltic philology, but also at my request undertook the trouble of revising the proof-sheets relating to Keltic. Botli have been a help to my work in many details, in far more places than has been indicated by a reference to his name. In particular it was such help only that made it possible in all cases to distinguish the British forms correctly according to dialect and period. It may also be remarked that almost all the addenda relating to Keltic p. 565 ff. [now incorporated with the text of this translation] were due to communications from Thurneysen. If I add lastly that my treatment of some of the phenomena of Keltic philology is not in harmony with the views of this scholar, it is in order to prevent misunderstanding, that he may not in any way be made responsible for mistakes that might be found in the Keltic parts of this book. I have to thank Dr. H o l t h a u s e n of Heidelberg [now in Halle] for some notices relating to Germanic, which he placed at my disposal after an inspection of several proof- and the finally revised sheets and of which I have for the most part made use. I have not been able consistently to take account of the scientific literature which has appeared this year. I especially lament that the second edition of G. Meyer's Greek Grammar, Braune's Old High Gorman Grammar and the second edition of Leskien's Hand-book of the Old Bulgarian Language did not appear until after my printing had begun. To facilitate reference, I shall add an index of words at the end of the whole work. Its continuation and conclusion shall follow as quickly as time and strength permit. F r e i b u r g i. B., July 1886. KARL BRUGMANN. ') It is only since the founding of the chair for comparative philology that the University library here has systematically purchased works on this subject (apart of course from the classical languages and Germanic), so that the necessary material can only be procured very gradually.
TRANSLATOR'S PREFACE. W h en Prof. Brugmann and Mr. Triibner proposed to me, two years ago, while I was still a student in Heidelberg, that I should translate the 'Grundriss dor vergleicheiulen Grammatik der indo-germanischen Sprachen' into English as soon as it appeared, I gladly accepted the proposal, in the hope that I should thus be rendering valuable service both to English and American students of philology, especially to the former who would otherwise very possibly, if they did not happen to know German, have to remain an indefinite length of time without being able to enter into a systematic and s c i e n t i f i c study of languages, based on firm and rigid principles. If the work of translating the original has taken me longer than I at first anticipated it would, this is almost entirely due to the povertystricken state of our language as regards current philological technical terms. Some of the terms employed by me, e. g. slurred and broken accent, intersonantal, initiality, strong-grade etc., will no doubt sound strange at first reading. But those, who have either been trained at a German university or are familiar with German philological literature, will readily confess how difficult it sometimes is to find a very exact and appropriate English equivalent for many of the German grammatical technicalities. Time will of course show whether I have in each case hit upon the best term, but I have everywhere tried rather to represent as cxactly as I could the meaning of the author than to produce an elegant English paraphrase, which could preserve neither the brevity nor the rigidly scientific form of the original. Last spring Prof. Brugmann gave me in Leipzig a list of corrections, consisting partly of misprints not given at the end of the original work, and partly of a few slight mistakes which were mentioned in the reviews of the work. These as also the corrections and emendations at the end of the German edition have been worked into the body of the translation. I also received from the author a short time ago one or two other oorreotions of the original which will be found at the end of the present edition.
X
Translator's Preface.
At the end of the volume will be found a short list of those abbreviations whioh would be likely to cause the reader any difficulty. In conclusion I have to express my sincere gratitude to Prof. Brugmann and Mr. P. Giles, Fellow of Gonville and Cajus College Cambridge, for the great assistance they have rendered me by helping in the reading of the proof-sheets. J. WRIGHT. L o n d o n , November 29, 1887.
CONTENTS OF THE FIRST VOLUME. INTRODUCTION. DEFINITION
OF
GUAGES,
TIIK AND
SCIENCE T11E
OF
DIVISION
TH10 OF
Tage
IN DO-GERMANIC THE
LAN-
INDG. FAMILY
OF
LANGUAGES INTO ITS VARIOUS BRANCHES
1
Definition of the s c i e n c e of the I n d g . l a n g u a g e s ( § 1 )
1
P r i m i t i v e home of the I n d g . people ( § 2 )
2
Differentiation of dialects ( § 3 )
2
1. T h e A r y a n b r a n c h ( § 4 )
4
2. T h e Armenian b r a n c h ( § 5J
5
3. T h e G r e e k branch ( § 6 )
6
4. T h e Albanian b r a n c h ( § 7 )
7
5. T h e I t a l i c b r a n c h ( § 8 )
8
6. T h e K e l t i c b r a n c h ( § 9 )
9
7. T h e Germanic b r a n c h ( § 1 0 )
10
8. T h e B a l t i c - S l a v o n i c b r a n c h ( § 1 1 )
11
S k e t c h of the Indg. p h a s e s of development to be discussed in this book ( § 1 2 ) ON
THE
STRUCTURE
12 OF
THE
INDG. LANGUAGES
IN
GENERAL.
ROOT AND SUFFIXES
13
D e v e l o p m e n t of word-formation and inflexion from composition (§ 13)
13
M e a n i n g of the hyphens usual in t h e analysis of words (§ 14)
.
.
16
.
.
19
PHONOLOGY. THE SOUNDS OF THE INDG. PRIMITIVE LANGUAGE ( § 15) PHONETIC ELUCIDATIONS
.
20
1. V o i c e d and voiceless sounds ( § 1 6 )
20
2. Sonorous and noised sounds ( § 17)
21
3. S o n a n t s and consonants ( § 1 8 )
. 2 1
xir
Contents of the first volume. Page
ON THE PRONUNCIATION OF THE LETTERS
The different kinds of alphabets and transcription (§ 19) Sanskrit ( § 2 0 ) , Iranian ( § 2 1 ) Armenian ( § 2 2 ) Umbrian-Samnitic ( § 2 3 ) . , , . . . . , Old Irish ( § 2 4 ) Gothic ( § 2 5 ) Lithuanian ( § 2 6 ) Old Bulgarian ( § 2 7 )
22
. . . , . . . .
22 23 24 25 26 27 28 28 30
HISTORY OF THE SEPARATE PRIM. INDG. SOUNDS. THE
VOWELS.
A. T h e v o w e l s u s s o n a n t s The prim. Indg. vowel sounds (§ 28) Indg. I Prim. Indg. ( § 2 9 ) Aryan (§ 30) Armenian (§ 31) Greek ( § 3 2 ) Italic ( § 3 3 ) O.Irish ( § 3 4 ) Germanic (§ 35) Baltic-Slavonic ( § 3 6 ) Indg. l Prim. Indg. ( § 3 7 ) Aryan ( § 3 8 ) Armenian (§ 39) Greek ( § 4 0 ) Italic ( § 4 1 ) O.Irish ( § 4 2 ) ; Germanic (§ 43) Baltic-Slavonic ( § 4 4 ) Indg. it Prim. Indg. ( § 4 5 ) Aryan ( § 4 6 ) Armenian (§ 47) Greek ( § 4 8 ) Italic ( § 4 9 ) O.Irish ( § 5 0 ) Germanic ( § 5 1 ) Baltic-Slavonic ( § 5 2 ) Indg. u Prim. Indg. ( § 5 3 ) Aryan (§ 54)
30 30 30 32 32 32 32 33 34 35 35 37 37 38 38 38 38 39 39 39 40 40 40 40 41 41 42 43 43 45 45 45
Contents of the first volume.
XIII Page
Armenian (§ 55) Greek ( § 5 6 ) Italic ( § 5 7 ) O.Irish ( § 5 8 ) Germanic ( § 5 9 ) Baltic-Slavonic ( § 6 0 )
45 45 46 46 47 47
Indg. e Prim. Indg. ( § 6 1 ) Aryan ( § 6 2 ) Armenian ( § 6 3 ) Greek ( § 6 4 ) Italic ( § 6 5 ) O.Irish ( § 6 6 ) Germanic ( § 6 7 ) Baltic-Slavonic ( § 6 8 )
47 47 49 60 50 51 54 56 59
Indg. Prim. Indg. ( § 6 9 ) Aryan ( § 7 0 ) Armenian ( § 7 1 ) Greek ( § 7 2 ) Italic ( § 7 3 ) O.Irish ( § 7 4 ) Germanic ( § 7 5 ) Baltic-Slavonic ( § 7 6 )
61 61 62 62 63 63 64 64 65
Indg. Prim. Indg. ( § 7 7 ) Aryan ( § 7 8 ) Armenian ( § 7 9 ) Greek ( § 8 0 ) Italic ( § 8 1 ) O.Irish ( § 8 2 ) Germanic ( § 8 3 ) Baltic-Slavonic ( § 8 4 )
66 66 68 70 71 72 76 77 79
Indg. Prim. Indg. ( § 8 5 ) Aryan ( § 8 6 ) Armenian (§ 87) Greek ( § 8 8 ) Italic ( § 8 9 ) O.Irish ( § 9 0 ) Germanic ( § 9 1 ) Baltic-Slavonic ( § 9 2 )
83 83 83 84 84 84 85 85 86
Indg. a Prim. Indg. ( § 9 3 ) Aryan ( § 9 4 )
87 87 88
Contents of t h e first volume.
XIV
Page
Armenian ( § 9 5 ) Greek ( § 9 6 ) Italic ( § 9 7 ) O.Irish ( § 9 8 ) Germanic ( § 9 9 ) Baltic-Slavonic (§ 100)
89 90 91 93 94 94
Indg. a
96
Prim. I n d g . (§ 101) Aryan (§ 102) Armenian (§ 103) Greek (§ 104) Italic (§ 105) 0.1rish (§ 106) Germanic ( § 1 0 7 ) Baltic-Slavonic (§ 108)
•
.
.
.
96 96 97 97 98 99 99 99
Indg. a (§ 109—110)
100
Primitive I n d g . vowel contraction General r e m a r k s (§ 111) 1. Contraction of a-, e-, o-vowels with a-, e-, o-vowels (§ 112 — 115) II. Contraction of a-, e-, o-vowels with i- and ((-vowels (§ 116)
105 105
B. T h e
vowels
as c o n s o n a n t s
General r e m a r k s ( § 1 1 7 ) Indg. i Prim. I n d g . (§ 1 1 8 - 1 2 2 ) A r y a n (§ 123—127) Armenian ( § 128) Greek (§ 129—132) Italic (§ 133—136) O.Irish (§ 137—140) Germanic (§ 141 — 144) Baltic-Slavonic (§ 145—148) Loss of i in the prim. Indg. period (§ 149 — 150) Indg. y, Prim. I n d g . (§ 151—156) A r y a n (§ 1 5 7 - 1 6 1 ) Armenian ( § 1 6 2 ) Greek (§ 163—167) Italic (§ 1 6 8 - 1 7 2 ) O.Irish (§ 1 7 3 - 1 7 6 ) Germanic (§ 177—181) Baltic-Slavonic (§ 182—186) Loss of y, in the prim. Indg. period (§ 187 — 188)
106 108 109 109 110 110 113 117 117 120 124 125 130 135 137 137 140 145 145 149 153 155 159 162
Contents of the first volume.
XV Page
NASALS.
A. T h e n a s a l s a s c o n s o n a n t s Prim. Indg. period (§ 189-196) Aryan (§ 197-200) Armenian (§ 201—202) Greek (§ 203 -205) Italic (§ 206—209) 0.1.ish (§ 210—212) Germanic (§ 2 1 3 - 2 1 5 ) Baltic-Slavonic (§ 216-219) Nasal changes in the prim. Indg. period (§ 220 — 221)
162
. . . .
B. T h e n a s a l s a s s o n a n t s
162 166 169 169 173 177 180 183 188 190
Preliminary remarks (§ 222) 1. The short sonant nasals (§ 223—252) General remarks (§ 223) Prim. Indg. period (§ 224 227) Aryan (§ 228—231) Armenian (§ 232) Greek (§ 233-236) Italic (§ 237-241) O.Irish (§ 242 - 243) Germanic (§ 244—247) Baltic-Slavonic (§ 2 4 8 - 2 5 2 ) 2. The long sonant nasals (§ 253)
190 190 190 191 194 197 197 199 200 201 204 206
LIQUIDS.
A. T h e l i q u i d s a s c o n s o n a n t s Prim. Indg. period (§ 2 5 4 - 2 5 6 ) Aryan (§ 257-261) Armenian (§ 262-263) Greek (§ 264—266) Italic (§ 267-271) 0.1rish (§ 272-274) Germanic (§ 275-277) Baltic-Slavonic (§ 278—281) Changes of liquids in the prim. Indg. period (§ 282) B. T h e l i q u i d s a s s o n a n t s Preliminary remarks (§ 283) 1. The short sonant liquids (§ 284—305) General remarks (§ 284) Prim. Indg. period (§ 2 8 5 - 2 8 7 ) Aryan (§ 288—290) Armenian (§ 291) Greek (§ 292—294)
207
. . . .
207 210 213 214 216 219 220 222 225 226 226 226 226 227 229 232 233
Contents of the first volume.
XVI
Page
Italic (§ 2 9 5 - 2 9 7 ) O.Irish (§ 298) Germanic (§ 2 9 9 - 3 0 1 ) Baltic-Slavonic (§ 302-305) 2. The long sonant liquids (§ 306) ABLAUT (VOWEL
235 236 237 239 241
GRADATION).
The nature of ablaut and the different grades of ablaut in general (8 3 0 7 - 3 1 0 )
244
1. 2. 3. 4. 5. 6.
247 254 255 255 256 258
The The The The The The
e-seriea e-series a-series «-series "-series o-seriee
(§ 311—314) (§ 315) (§316) (§ 317) (§ 318) (§ 319)
EXPLOSIVES.
General remarks: The nature of explosives (§ 320), place of articulation (§ 321), form of articulation (!) 322), system of the Indg. explosives i§ 323)
261
A. T h e e x p l o s i v e s culation
262
according
t o t h e i r p l a c e of
arti-
1. The labial explosives Indg. prim, period (§ 324—326) Aryan (§ 327-329) Armenian (§ 3 3 0 - 3 3 2 ) Greek (§ 333—335) Italic (§ 3 3 6 - 3 3 8 ) O.Irish (§ 339-341) Germanic (§ 342-344) Baltic-Slavonic (§ 345—347)
262 262 264 265 266 267 268 270 271
2. The dental explosives Prim. Indg. period (§ 348—350) Aryan (§ 351—359) Armenian (§ 3 6 0 - 3 6 2 ) Greek (§ 363—365) Italic (§ 366—370) O.Irish (§ 371—373) Germanic (§ 374—376) Baltic-Slavonic (§ 377—379)
272 272 273 275 277 278 282 283 285
3. The palatal explosives General remarks (§ 380) Prim. Indg. period (§ 381—383)
286 286 288
C o n t e n t s of t h e first volume.
XVII I'nge
a.
T h e l a n g u a g e - g r o u p with explosives G r e e k (§ 3 8 4 - 3 8 6 ) I t a l i c (§ 387—389) O.Trish (§ 390 - 392) G e r m a n i c (§ 3 9 3 - 3 9 5 )
289 289 290 292 293
b. T h e l a n g u a g e - g r o u p with s p i r a n t s or a f f r i c a t a e A r y a n (§ 3 9 6 - 4 0 7 ) A r m e n i a n ( § 408—410) Albanian ( § 4 1 1 ) Baltic-Slavonic (§ 4 1 2 - 4 1 6 ) 4
. . . .
295 295 300 302 302
The velar explosives G e n e r a l r e m a r k s ( § 417) P r i m . I n d g . p e r i o d (§ 4 1 8 - 4 2 3 )
305 305 306
a. T h e l a n g u a g e - g r o u p with l a b i a l i s a t i o n P r e l i m i n a r y r e m a r k on t h e i r r e g u l a r a p p e a r a n c e of l a b i a lisation (§ 424' G r e e k ( § 425—429) I t a l i c (§ 4 3 0 - 4 3 3 ) O . I r i s h (§ 434—438) G e r m a n i c (§ 4 3 9 - 4 4 4 ) . .
309
b.
T h e l a n g u a g e - g r o u p w i t h o u t labialisation A r y a n (§ 445—454) A r m e n i a n ( § 455—457) A l b a n i a n (§ 458—460) B a l t i c - S l a v o n i c (§ 461—465) R e t r o s p e c t of t h e h i s t o r y of t h e v e l a r explosives (§ 466) . I n t e r c h a n g e of t h e p a l a t a l a n d v e l a r explosives (§ 467) . B. T h e
explosives
after
their
form
. .
309 311 318 323 326 331 331 336 337 337 341 342
o f a r t i c u l a t i o n.
P r i m . I n d g . p e r i o d (§ 4 6 8 - 4 6 9 1 Aryan (470-482) Armenian (§ 4 8 3 - 4 8 5 ) G r e e k (§ 4 8 6 - 4 9 7 ) I t a l i c ( § 498 - 5 1 1 ) O.Irish (§ 5 1 2 - 5 2 6 ) G e r m a n i c ($ 527 - 541) B a l t i c - S l a v o n i c (§ 5 4 2 - 5 5 1 ) T h e combination m e d i a a s p . 4- t, (§ 552) T h e I n d g . t e n u e s a s p i r a t a e (S 553)
344 347 357 359 365 375 381 396 403 405
THE SPIRANTS. G e n e r a l r e m a r k s on t h e p r i m . I n d g . s p i r a n t s (§ 554) I n d g . s. P r i m . I n d g . p e r i o d ( § 555)
407 408 *
xvill
Contents of t h e first volume. Page
Aryan (§ 556—558)
410
Armenian (§ 5 5 9 - 5 6 2 ) Greek (§ 5 6 3 - 5 6 6 )
.
416
.
418
Italic (§ 5 6 7 - 5 7 2 )
424
O.Irish (§ 5 7 3 - 5 7 6 )
430
Germanic ( 5 7 7 — 5 8 3 )
432
Baltic-Slavonic (§ 584—588J P r i m . I n d g . c h a n g e s of s ( § 4 8 9 ) . . .
438 .
445
I n d g . z. P r i m . I n d g . p e r i o d (§ 5 9 0 )
447
Aryan (§ 591)
447
Armenian (§ 592J
. . . .
Greek (§ 593)
449 449
Italic (§ 594)
450
O . I r i s h (§ 5 9 5 )
451
Germanic (§ 596)
451
Baltic-Slavonic (§ 597)
452
Indg. j (§ 598) OTHER
452
COMBINATORY
SOUND-CHANGES.
Preliminary remarks (§ 599) CONTRACTION (HIATUS,
453
ELISION).
General remarks (§ 600)
453
Aryan (§ 601) .
455
.
.
Armenian (§ 602)
455
G r e e k (§ 6 0 3 )
455
Italic (§ 604)
457
O.Irish (§ 605)
458
Germanic (§ 606)
459
B a l t i c - S l a v o n i c (§ 6 0 7 J
460
SHORTENING OF LONG VOWELS. General remarks (§ 608)
461
Aryan (§ 609)
461
Armenian (§ 610J
461
Greek ( § 6 1 1 )
461
Italic ( § 612)
462
O.Irish (§ 613)
.464
Germanic (§ 614)
464
Baltic-Slavonic ( § 615)
465
LENTHENING OF SHORT Aryan ( § 6 1 6 )
VOWELS. 465
XIX Page
Armenian ( § 6 1 7 ) Greek (§ 618) Italic ( § 6 1 9 ) O.Irish (§ 620) Germanic (§ 621) . . . Baltic-Slavonic (§ 622) ANAPTYXIS
FROM
. . .
CONSONANTS.
General remarks (§ 623) Aryan (§ 624) Armenian (§ 625) Greek (§ 626) Italic (§ 627) Germanic (§ 628) Baltic-Slavonic (§ 629) VOWEL
.
.
468 .469 470 470 470 472 472
ABSORPTION.
General remarks (§ 630) Aryan (§ 631) Armenian (§ 632) . . . Italic (§ 633) O.Irish (§ 634) Germanic (§ 635) Baltic-Slavonic (§ 636) PALATALISATION
AND
LABIALISATION
473 473 473 473 474 475 .477 WITH
EPENTHKSIS.
General remarks (§ 637) Aryan (§ 638) Greek (§ 639) O.Irish (§ 640) Germanic (§ 641) Baltic-Slavonic (§ 642) LOSS
466 466 466 467 467 467
OF
SANDHI
SYLLABLE
BY
DISSIMILATION ( § 6 4 3 )
477 478 478 479 481 481 481
(SATZPHONETIK).
Sandhi in general (§ 644) Prim. Indg. period (§ 645) . Aryan (§ 64^—650) Armenian (§'-651) Greek (§ 652-654) Italic (§ 655-656) O.Irish (§ 657—658) Germanic (§ 659—662) Baltic-Slavonic (§ 663—666)
483 488 490 497 498 501 506 512 521
XX
Contents of the first volume. Page
ACCENTUATION. Accentuation in general (§ 667—668) Prim. Indg. period (§ 669 - 6 7 1 ) Aryan (§ 6 7 2 - 6 7 4 ) Armenian (§ 675) Greek (§ 6 7 6 - 6 7 8 ) Italic (§ 6 7 9 - 6 8 3 ) O.Irish (§ 684—685) Germanic (§ 6 8 6 - 6 8 9 ) BaltU-Slavonic (§ 6 9 0 - 6 9 2 ) ADDITIONS AND CORRECTIONS
527 532 537 539 540 545 550 552 556 562
INTRODUCTION. DEFINITION MANIC
OP
THE
LANGUAGES,
INDO-GERMANIC
SCIENCE AND
OP
THE
FAMILY
THE
INDO-GER-
DIVISION
INTO
ITS
OP
THE
VARIOUS
BRANCHES. § 1.
The science of the Indg. languages forms, like Indg.
Mythology, a section of Indg. 'Philology', which has to investigate
i. e. of that science,
the intellectual development of the
Indg. peoples from the time before their separation up to the present day.
Its method accordingly is historical and its task
is to investigate the whole development of the Indg. languages from the time when they were still one language down to the present day.
Its unity is in no sense
broken by the
results
furnished by the specialists in Sanskrit, Ancient and Modern Greek, Latin and the Romance languages, Keltic etc., for the sciences of the Indian languages, of Ancient and Mod. Greek etc. are integral components of the grand whole formed by that of the Indo-Germanic. It is true that the so called comparative science of language has hitherto been
almost
exclusively confined to the
older
periods of the Indg. languages, but this is due to the division of labour which was involved
in the method by which
alone
progress could be made, as well as to the limitations of human strength.
Probably the same division of labour will still be
necessary, but it implies no real opposition between the different parts of the science. B r u g m a n n , Elements.
1
2
Primitive home. Differentiation of dialects.
§ 1-3.
Compare the author's inaugural address 'Sprachwissenschaft und Philologie' (in his 'Zum heutigen Stand der Sprachwissenschaft', Strassburg 1885, p. 3 ff.). § 2. W e are not at present in a position to determine finally what was the primitive home of the Indg. tribes. Only so much is clear, that these tribes in pre-historic times must have been far less widely diffused than they were at the beginning of the historic era. It was formerly usual to place this primitive home somewhere in Asia, whereas at present scholars lean rather to the opinion that the Asiatic members of the stock passed over from Europe. Cp. O. S c h r a d e r 'Sprachvergleichung und Urgeschichte', Jena 1883, p. 442 ff. § 3. It is impossible to suppose that a language should have gone through a long course of development, and be spoken by a people of any considerable numbers, without a certain amount of dialectical variation; and hence we cannot look upon the speech of the Indogermans, even while they still occupied a comparatively small territory and maintained a fairly close degree of intercourse with one another, as bearing in any strict sense, a uniform character. Local differences had no doubt already arisen, though actual instances of this have hardly as yet been established with perfect certainty. One such I have maintained, though only as a conjecture, in §§ 380, 417. In historical times there appear a multitude of dialects, diverging in a greater or less degree, the whole field of which can hardly be included in any one survey. W e may take for granted that the differentiation of dialects about the year 2000 B. C. had gone so far, that a number of communities existed side by side, which could no longer, or only with difficulty understand one another. The historic record of the various individual developments begins at very different periods. E. g. the Indian development is known to us from about the year 1500 B. C. onwards, the Latin from about 300 B. C., the Irish since the eighth century of the Christian era (with the exception of the Ogam inscriptions, cp. § 9), and the Lithuanian from the middle of the sixteenth century.
3 A number of separate developments, which sprung from the primitive Indg. language, have perished without leaving any traces of their existence. Of others we have only very scanty fragments l e f t , on which it is scarcely, if at all, possible to found a g r a m m a r , as of P h r y g i a n , Macedonian, Messapian, Gallic, and Burgundian. The remainder have come down to us with a more abundant supply of material. The dialects belonging to this last class, are arranged into eight groups (branches of language): 1. A r y a n , 2. Armenian, ¿3. Greek, 4. Albanian, 5. Italic, 6. Keltic, 7. Germanic, 8. BalticSlavonic. Each group is distinguished by the fact that its individual members show in common a considerable number of changes in sound, inflexion, syntax, and vocabulary; e. g. the shifting of the prim. Indg. explosives, discussed in § 5*27 ff. (what is known as the first sound-shifting), is one of the numerous characteristics of the close relation existing between the Germanic dialects. Of the dialects, which have come down to us only in scanty fragments, there are a certain number which can be assigned with certainty to one or other of these eight groups, e. g. the Burgundian to the Germanic, and the Gallic to the Keltic g r o u p , whilst others, e. g. the Macedonian, seem to belong to none of them. It is in itself very possible that some of these eight chief members stand to each other in a closer relationship, and form a single group historically distinct from the rest. All attempts, however, to establish such a closer relationship, have hitherto proved futile. The Italo-Keltic hypothesis has perhaps the best prospect of attaining a greater degree of probability in the future. Cp. J o h a n n e s S c h m i d t Die Verwandtschaftsverhältnisse der indogermanischen Sprachen, Weimar 1872. L e s k i e n Die Declination im Slavisch-Litauischen und Germanischen, Leipzig 1876,introduction. D e l b r ü c k Einleitung in das Sprachstudium, Edition 2. 1884, p. 131 ff. C. S c h r ä d e r op. cit. p. 66 ff. The Author, Zur F r a g e nach den Verwandtschaftsverhältnissen l*
4
Aryan.
Indian.
der indogermanischen Sprachen, in Techmer's Internat. Zeitschr. für allg. Sprachwissenschaft I 226 if. § 4. (1) The Aryan branch consists of the Indian and Iranian groups. The oldest and most archaic dialect of the I n d i a n group is the language of the Yeda, the oldest portions of which (the hymns of the Rigveda) possibly go back as far as 1500 B.C. x ). Next comes Sanskrit in the stricter sense of the word (also called classical Sanskrit), the continuation of a dialect now lost, which existed side by side with the Yedic and differed but slightly from it in the formation of its sounds and inflexions. It became separated from the popular development as the literary language, and was stereotyped in a fixed and purely artificial form. The vulgar language, called Prakrit, as early as the third century B.C. differed very materially from Sanskrit and at that time was divided into at least three chief dialects. Through the influence of Buddhism, Prakrit was also raised to a literary language (Pali). From the popular dialects of Prakrit have descended the numerous modern Indian languages and dialects, Hindi (Hindustani), Bengali, Uriya, Maharatti, Guzerati, Sindhi, Penjabi and others. In this work we shall only deal with the Yedic language and classical Sankrit. Cp. B o p p Kritische Grammatik der Sanskrita-Sprache in kürzerer Fassung, Edit. 3., Berlin 1863. B e n f e y Vollständige Grammatik der Sanskritsprache, Leipzig 1852. The same Author's Kurze Sanskrit-Grammatik, Leipzig 1855. W h i t n e y Sanskrit Grammar, Leipzig 1879 (with supplement I, Grammatisches aus dem Mahabharata, by A. H o l t z m a n n , 1884; supplement II, The roots, verb-forms, and primary derivatives of the Sanskrit language, by Whitney, 1885). The oldest recorded dialects of the I r a n i a n group are O l d P e r s i a n (West Iranian), the language of the Persian 1) Cp. A. K a e g i Der Rigveda, die älteste Literatur der Inder, Edition 2, Leipzig 1881.
§ 4-5.
Iranian.
Armenian.
5
cuneiform inscripitions dating from about 520 to 350 B.C., and A v e a t i c, also called Zend and Old Bactrian (East Iranian), the language of the Avesta, the sacred book of the Zoroastrians, which has come down to us with corruptions of many kinds, deviating considerably from its original form. The various portions of the work were composed at very different periods'). Some parts of the Avesta, including 17 hymns (gapa f.), are written in a peculiar dialect, which is more archaic than the language of the other parts, and as distinct from which the latter is called Later Avestic or Zend in the stricter sense. Not one of the Modern Iranian languages is a direct continuation of Old Persian or Avestic. The Modern Persian dialects (Gilani etc.), Kurdic and probably also Ossetian (spoken in the neighbourhood of the Caucasus) are more closely related to the former, and the language of Afghanistan (Pastu) to the latter. W e shall only deal with the two Old Iranian languages. Cp. S p i e g e l Die altpersischen Keilinschriften, Edition 2., Leipzig 1881. J u s t i Handbuch der Zendsprache, Leipzig 1864. B a r t h o l o m a e Handbuch der altiranischen Dialekte, Leipzig 1883. § 5. (2) A r m e n i a n , which de Lagarde and Friedr. Miiller assigned to the Iranian group, but Hiibschmann (Kuhn's Ztschr. X X I I I 5 ff., 400 ff.) has proved to be an independent member of the Indg. family of languages, is known to us since the fifth century of our era. The literary language (O.Armenian), which had then become fixed, remained in this usage without any material changes down to modern times, and is separated by a great interval from the modern dialects. Historical investigations have hitherto been almost exclusively confined to O.Armenian. The term Armenian will always be used in the sense of O.Armenian. Cp. P e t e r m a n n , Grammatica linguae Armeniacae, Berlin 1837. The same Author, Brevis linguae armeniacae grammatica, litteratura, chrestomathia cum glossario, ed. I I , Berlin 1872. 1) On the difficult question as to the age of the Avesta cp. E d u a r d M e y e r Geschichte des Alterthums I (1884) p. 501 ff.
6
Greek.
H ü b s c h m a n n Armenische Stadien I, Leipzig 1883 (of special importance for the phonology). § %. (3) G r e e k had strongly marked variations in different localities long before Homer's time and in the historic era appears split up into numerous dialects. These may be classified in the following manner: 1. I o n i c A t t i c , a. The district of Ionia, b. Attica. 2. D o r i c , a. Laconia with Tarentum and Heraclea, b. Messenia, c. Argolis and Aegina, d. Corinth with Corcyra, e. Megara with Byzantium, f. the Peloponnesian colonies of Sicily, g. Crete, h. Thera and Melos together with Cyrene, i. Rhodes with Gela and Acragas, k. the other Dor. islands in the Aegean, as Carpathus, Astypalaea etc. 3. N o r t h W e s t G r e e k , a. Phocis, b. Locris, c. Aetolia, d. Acarnania, e. Phthiotis and the district of the Aenianes, f. Epirus, g. probably also Achaia. 4. A e o l i c , a. Lesbos and Aeolian Asia Minor, b. North Thessaly, c. Boeotia. 5. E l e a n (belonging to North West Greek?). 6. A r c a d i a n C y p r i a n 1 ) . 7. P a m p h y l i a n . These dialects are found in their purest form on inscriptions. The literary language, especially that of the poets, is in many respects artificially constructed; even the language of the Homeric poems is in great measure an artificial dialect. Towards the end of the fifth century B. C. there was formed on the basis of the Attic dialect a literary language common to all Greeks, which almost entirely excluded the use of the other dialects from the later prose literature of antiquity. Hence it is that we have to gather our knowledge of most of the non-Attic local dialects either entirely, or almost entirely, from inscriptions. The language of mediaevel Greek literature is an artificial mixture of ancient Greek with forms of the then spoken popular language in varying degrees of modification. Modern Greek 1) Collitz (Die Verwandtschaftsverhältnisse der griech. Dialekte 1885) places this dialect in close relation with Aeolic and puts forward the hypothesis that 'the Thessalian dialect in point of language forms the transition-stage from Boeotian to Lesbian, from Lesbian to Cyprian-Arcadian and from Cyprian-Arcadian to Boeotian' (p. 9).
§ 6-7.
Greek.
Albanian.
7
exhibits a rich development of dialects; as yet little has been done towards fixing the historical relation of the separate dialects to ancient Greek 1 ). W e shall confine our attention to the development of the O.Gr. dialects. Cp. A h r ens De Graecae linguae dialectis, I. De dialectis Aeolicis, Göttingen 1839, II. De dialocto Dorica, Göttingen 1843. R. M e i s t e r Die griechischen Dialekte, I. Asiatisch-Aolisch. Böotisch, Thessaliscb, Göttingen 1882. G. C u r t i u s Grundzüge der griechischen Etymologie, Edit. 5., Leipzig 1879. R. K ü h n e r Ausführliche Grammatik der griechischen Sprache, 2 vols., Edit. 2., Hannover 1869—1870. G. M e y e r Griechische Grammatik, Leipzig 1880 (Edit. 2. 1886). D e l b r ü c k Die Grundlagen der griechischen Syntax, Halle 1879. The Author, Griechische Grammatik, in Iw. Müller's Handbuch der klassischen Altertumswissenschaft II (1885) p. 1 — 126. § 7. (4) A l b a n i a n , the language of ancient Illyria, has only been known to us from monuments of any extent since the seventeenth century. The historical treatment of this language, which is beset with manifold difficulties — the greatest of which lies in clearly separating pure Albanian words from those borrowed from the Greek, Latin, Romance, Slavonic and Turkish languages —, is still in its infancy. Cp. G u s t a v M e y e r Albanesische Studien, Wien I 1883, I I 1884 (in which the whole of the older literature has been collected together) and 'Der Einfluss des Lateinischen auf die alban. Formenlehre' in the Miscellanea di Filologia, dedicata alia memoria dei professori Caix e Canello p. 103 if. 2 ). W e shall only take this language into account in discussing 1) Considerations introductory to the study of the relation of Middle and Modern Greek to O.Greek and the researches on the subject are given by K r u m b a c h e r Beiträge zu einer Geschichte der griech. Sprache, in Kuhn's Ztschr. XXY1I 481 ff. 2) The treatise of the same scholar 'Über Sprache und Literatur der Albanesen' in his 'Essays und Studien zur Sprachgeschichte und Volkskunde', Berlin 1885, p. 49 ff. is strongly to be recommended as an introduction to this subject.
8
Italic.
Latin.
Umbro-Samnite.
§ 7-8.
the development of the Indg. palatal and guttural explosives § 411 and § 458—460. § 8. (5) The I t a l i c branch consists of Latin on the one hand and of the Umbrian-Samnitic dialects on the other. L a t i n , with which the little known dialect of Falerii was closely related, is known to us from about 300 B . C. onwards. So long as the language was confined to Latium, there existed no dialectical differences of any importance. The contrast between the popular and the literary language, which had already arisen at the beginning of the archaic period of literature (from Livius Andronicus to Cicero), became still sharper in the classical period, and the further development of the former is almost entirely lost to our observation until the Middle Ages, when the popular Latin of the various provinces of the Roman empire meets us in a form more or less changed and with a rich development of dialects (Romance languages: Portuguese, Spanish, Catalanian, Provençal, French, Italian, Raetoromanic and Roumanian) W e shall only consider the development of the Latin of antiquity. Cp. Cor s sen Über Aussprache, Yocalismus und Betonung der lateinischen Sprache, 2 vols., Leipzig 1858. 1859, edit. 2., 1868. 1870. R . K ü h n e r Ausführliche Grammatik der lateinischen Sprache, 2 vols., Hannover 1877. 1879. F . S t o l z and J . Gr. S c h m a l z Lateinische Grammatik, in Iw. Müller's Handbuch der klass. Altertumsw. I I (1885) p. 127—364. The U m b r i a n - S a m n i t i c dialects are known to a certain extent through inscriptions, which for the most part belong to the last centuries before our era, and through words quoted by Roman writers. W e are best acquainted with Umbrian ( B r é a l Les tables Eugubines, Paris 1875, B ü c h e l e r Umbrica, Bonn 1883) and Oscan ( Z v e t a i e f f Sylloge inscriptionumOscarum, Petersburg-Leipzig 1878). Of the Yolscian, Picentine, Sabine, 1) Cp. B u d i n s z k y Die Ausbreitung der lat. Sprache über Italien und die Provinzen des römischen Reiches, Berlin 1881, G r ö b e r in the Archiv für lat. Lexikographie I 35 ff., 204 ff.
Aequiculan, dialects
9
Keltic.
§ 8-9.
Vestinian,
we have only
Marsian, very
Pelignian
scanty
and
Marrucinian
remains ( Z v e t a i e f f
scriptiones Italiae Mediae dialecticae, Leipzig 1884).
In-
All these
dialects were forced into the background at an early period by the intrusion of Latin.
The Sabines, who received citizenship
in 267 B. C., seem to have been the first to become romanised. The slowest to give way was Oscan, which in the mountains did not perhaps become fully
extinct for centuries after the
Christian era. Cp. further B r u p p a c h e r
Osk. Lautlehre,
E n d e r i s Yersuch einer Formenlehre
Zurich
1869,
der osk. Sprache, Zurich
1871. § 9.
(6)
The K e l t i c
languages fall into three groups:
Gallic, Britannic and Gaelic 1 ). W e know something of G a l l i c through Keltic names and words quoted by Greek and Latin authors, through inscriptions and coins.
But the interpretation of the forms, mostly proper
names, is in most cases so uncertain that from these remains linguistic research has hitherto gained comparatively little. B r i t a n n i c broke up into Cymric (or Welsh), Cornish and B a s Breton (or Armorican); the connexion between the last two is especially close.
We
are acquainted
with Cymric and B a s
Breton from the eighth or ninth century onward, at first through glosses; the oldest records of Cornish are somewhat later.
The
last Darned dialect became extinct at the end of the eighteenth or beginning of the nineteenth century, both the others are still living. Irish-Gaelic, Scotch-Gaelic (which is also known as Gaelic in the stricter sense),
and Manx (spoken
form the G a e l i c division.
in the Isle of Man)
All three of these are still living.
The first two languages seem hardly to have differed from each other in the ninth century.
The
oldest monuments are the
O.Gaelic Ogam inscriptions (Ogam is the native name for the 1) The first two languages have often been classed together as a special group, but without sufficient reason. See T h u r n e y s e n Keltoromanisches 1884 p. 7 ff.
10
Keltic. Germanio.
§ 9-10.
Gaelic runes), which possibly date as far back as about 500 A.D. The literary record of Irish begins in the eighth century, at first with glosses, and then from the year 1100 onward we have many extensive manuscripts which contain sagas, ecclesiastical literature etc. Scotch-Gaelic literature begins somewhat later than that of Irish and in the older period is closely connected with Irish. Manx is only known to us during the last few centuries'). W e shall consider chiefly Old Irish. Cp. J. C. Z e u s s Grammatica Celtica (e monumentis vetustis tam Hibernicae linguae quam Britannicarum dialectorum Cambricae Cornicae Aremoricae comparatis Gallicae priscae reliquiis construxit J. C. Z.), Edit. 2. by H. E b e l , Berlin 1871. W i n d i s c h Kurzgefasste irische Grammatik, Leipzig 1879. § 10. (7) The G e r m a n i c branch is divided into Gothic, Norse, and West Germanic. G o t h i c , the most archaic language of the Germanic group, is chiefly accessible to us through the biblical translation of the West Gothic bishop Ulfilas (311—381 A.D.). The language perished with tho Gothic nation. (The East Goths, who dwelt in the Crimea, retained their language down to modern times; but of this branch of Gothic we possess only a few isolated words.) N o r s e (or S c a n d i n a v i a n ) , which in certain special points closely coincides with Gothic 2 ) and is therefore by some writers classed together with it as East Germanic as opposed to West Germanic, down to the Yiking period (800 —1000 A. D.) was practically a single language, but later on became broken up into four languages, Icelandic and Norwegian (West Norse) on the one hand, Swedish and Danish (East Norse) on the other. The earliest records are runic inscriptions, the oldest of which 1) "Windisch's article «Keltische Sprachen» in Ersch und Gruber's A. Encykl. d. W. u. K. second section, XXXV p. 132 ff., serves as a good introduction to the Keltic languages generally, the monuments of these languages and Keltological literature. 2) The most important of these coincidences are Goth, ddj = O.Icel. gyj from i (§ 142) and Goth, ggv — O.Icel. ggv from # (§ 179). Cp. Braune in Paul-Braune's Beitr. IX 546 f.
§ 10-11.
11
appear to belong to the fifth century. By O.Norse was, in former times, generally understood O.West Norse only; the term is now more properly applied to the whole development of the Scandinavian languages up to the sixteenth century. The oldest representatives of W e s t - G e r m a n i c are AngloSaxon, the continuation of which is Mod.English; 0 . F r i s i a n ; O.Saxon, now Low German; O L o w Franconian, now Dutch, F l e m i s h , and the language of the German Lower Franconia; and O.High German, the present Upper- und Middle German dialects. The oldest records of these languages date from the eighth or ninth century, tho oldest OHG. record belonging to about the period 740 - 745; but with Frisian we arc acquainted only since tho fourteenth century. W e shall limit ourselves chiefly to Gothic and Old High German. Cp. J. G r i m m Deutsche Grammatik, I 2 , II 2 , new reprint, Berlin 1870. 1878, III, I V Göttingen 1831, 1837. Rumpelt Deutsche Grammatik, mit Rücksicht auf vergleichende Sprachforschung, I (Lautlehre) Berlin 1860. A. I l o l t z m a n n Altdeutsche Grammatik I , 1 (Die specielle L a u t l e h r e ) , 2 (Vergleichung der deutschen L a u t e unter einander), Leipzig 1870.1875. W . S c h e r e r Zur Geschichte der deutschen Sprache, edit. 2., Berlin 1878. L e o M e y e r Die gothische Sprache, ihre L a u t gcstaltung insbesondere im Verhältniss zum Altindischen, Griechischen und Lateinischen, Berlin 1869. W . B r a u n e Gotische Grammatik, edit. 2., Halle 1882. L. W i m m e r Altnordische Grammatik, translated from the Danish by E. Sievers, Halle 1871. A. N o r e e n Altnordische Grammatik I (Altisl. und Altnorw. Gramm, unter Berücksichtigung des Urnordischen), Halle 1884. E. S i e v e r s Angelsächsische G r a m m a t i k , Halle 1882 (2. ed. 1886). W . B r a u n e Althochdeutsche Grammatik, Halle 1886. O . B e h a g h e l Die deutsche Sprache, Leipzig und P r a g 1886. § 11. (7) The B a l t i c - S l a v o n i c branch. The Baltic division consists of Prussian, Lithuanian, and Lettic. The former died out in the seventeenth century and we only possess very few imperfectly recorded specimens dating from the fifteenth and
12
Baltic-Slavonic.
§ 11-12.
sixteenth centuries. The two latter are still living and their oldest records belong to the sixteenth .century. W e confine ourselves chiefly to Lithuanian. Cp. N e s s e l m a n n Die Sprache der alten Preussen, Berlin 1845. The same author, Thesaurus linguae Prussicae, Berlin 1873. S c h l e i c h e r Litauische Grammatik, Prag 1856. K u r c h a t Grammatik der littauischen Sprache, Halle 1876. B i e l e n s t e i n Die lettische Sprache, nach ihren Lauten und Formen erklärend und vergleichend dargestellt, 2 Theile, Berlin 1863. 1864. The same author, Lettische Grammatik, Mitau 1863. The S l a v o n i c languages fall into a South-Eastern and a Western group. To te former belong Russian (a. Great Russian and White Russian, b. Little Russian), Bulgarian and Illyrian (a. Servian and Croatian, b. Slovenian), to the latter Czech (Czech in the stricter sense, Moravian and Slovakian), Sorabian or WeDdish (Upper- and Lower Sorabian) and Lechish (Polish and Polabian or Elbe-Slavonian). All these languages, with the exception of Polabian, are still living. The most archaic, and for Indg. grammar the most important, is the language in which the Slavonic apostles Cyril and Methodius (ninth century) wrote. In our investigations we accordingly consider this language of the first importance. It is called O.Slovenian by Miklosich, but by Schleicher, Schaffarik and others with more justness O.Bulgarian. It became the ecclesiastical language of the Greek church, and that form of it, which was modified through the influence of other Slavonic languages, especially of Russian, is generally known by the name of Church Slavonic. Cp. M i k l o s i c h Vergleichende Grammatik der slavischen Sprachen 4 vols. Vienna, I 2 1879, I I 1875, I I I 2 1876, I V 1874. S c h l e i c h e r Die Formenlehre der kirchenslavischen Sprache erklärend und vergleichend dargestellt, Bonn 1852. Leskien Handbuch der altbulgarischen (altkirchenslawischen) Sprache, Weimar 1871, edit. 2., 1886. § 12. In accordance with what has been said in §§ 4—11, we shall speak exclusively or at all events especially of the oldest periods of those languages whose developments are
13
§ 12—13.
known to us through records extending over a great length of time. The term 'science of the Tndg. languages', as has been already stated in § 1, does not demand such a restriction. This division of labour is wholly due to the course which the science has hitherto taken, and is fully justified by the present state of the science. B o p p in his Vergleichende Grammatik (3 vols., edit. 3, Berlin 1868—1871) and S c h l e i c h e r in his Compendium der vergleichenden Grammatik der indogermanischen Sprachen edit. 4, Weimar 1876) also limit themselves similarly as we have done. W h e n we speak of primitive Indg. forms; of prim. Ar.; of prim. Gr. or of prim. Ital. etc., we generally mean those forms which were in use towards the close of the primitive period of these languages. But we also often mean such forms as belonged to an earlier period of this stage, and which had already undergone a change towards its termination. For instance, we therefore speak equally well of p r i m . G e r m . *fcmx° and of p r i m . Germ. (Goth, faha, s. § 214), of p r i m . G r . *norai and p r i m . Gr.
*7ioam ( A t t . nnai,
s. § 4 9 0 ) , a s of L a t .
sequontur
and L a t . secuntur (§ 431). Forms, put down by us as prim. Ind., prim. Ar. etc., are therefore not to be indiscriminately regarded as belonging to the same period. Again, if w e , for example, uniformly write not h, in prim. Germ, forms where we put h for Gothic words, it must not be implied that y had not in certain cases already become h in prim. Germanic (cp. § 529). ON T H E S T R U C T U R E O F T H E INDO - G E R M A N I C L A N G U A G E S IN G E N E R A L . ROOT A N D S U F F I X E S . ' ) § 13. The Indg. languages belong to the inflexional class. The inflexion of words has not existed from the very beginning, but has been gradually developed and perfected. 1) "With what follows compare D e l b r ü c k Einleitung in das Sprachstudium2 (Leipzig 1884) p. 61 ff. and P a u l Principien der Sprachgeschichte (Halle 1880) p. 154 ff.
14
Root-period.
Word-formation.
W e have to presuppose a period in which suffixal elements were not yet attached to words. T h e word-forms of this period are called roots, and the space of time prior to inflexion, is called the root-period. It dates much further back than that stage of development whose word-forms we are able to deduce by a comparison of the separate Indg. groups of languages. This stage is usually simply called the Indg. parent-language. W h a t we understand by word-formation and inflexion arose by composition, that is, by the following process: a group of words which formed a syntactical complex was fused into a unity, in which the whole was in some way isolated in relation to its elements 1 ). This word-fusion from the beginning onwards occurred in the same way, just as afterwards, in the age of separate languages (partly even in historical periods) the final members of compounds became suffixes, e. g. Goth -k in mik (Mod. HG. mich) from prehistoric *me ge (cp. Gr. tutyt beside f,uf), French -ment in fièrement from fera mente, MHG. and Mod.HG. -heit in schœnheit, originally meaning 'schöne Beschaffenheit* (beautiful state or condition), where heit is still an independent word in MUG. and O H G . , Mod.Irish -mhar in buadhmhar 'victorious' = O.Ir. mär m,Or Cymr. mawr 'magnus' (Zimmer Kelt. Studien I I p. 22 ff). The formation of suffixes is not a work which belongs to any special prehistoric period and which was concluded at any definite point of time. But when once this process had b e g u n , it was performed anew through all periods of the history of the Indg. languages and will probably also be repeated again ia the f u t u r e , so long as our languages continue to develop.
1) "With respect both to meaning and form either the whole may pass through a development in which the separate parts do not participate when used independently, or vice versa the separate parts may pass through a development in which the whole does not participate, or it may happen that the separate parts cease to be employed independently, while they are preserved in combination, or lastly the mode of combination may vanish from living use and only remain preserved in certain set forms". Paul in the above work p. 165.
§ 13.
Suffixes, and their origin.
Words without suffixes.
15
Suffixes are divided into word-forming suffixes or inflexional suffixes in the stricter sense, to which on the one hand belong case endings (e. g. -m in Indg. *ekuom — Lat. equom, Skr. d&vam), and on the other, personal endings (e. g. -mi in Indg. *eimi = Gr. ilfu, Skr. end), and stem-forming suffixes (e. g. -ter- in Indg. *p9teres = Gr. nart'otg, Skr. pitaras, -sko- in Indg. *gi%t,sk6nti = Gr. puoY.ovrt fiaay.ovm, Skr. gdchanti). It is impossible to draw a sharp line between the two species of suffixes, since many an element, which was originally only stem - forming, has come to be treated on the same level with word-forming suffixes. Cp. e. g. Lat. legimini, whose final part mint, felt by the Romans as a personal ending (like -mur etc.), contains the stem-forming suffix -meno- (cp. Gr. \tyof.ui>oi); properly legimini estis = Xsyo/.isvot lent. The derivation as well as the original value of those suffixal elements, which in the Indg. prim, period were already no longer felt to be the members of a compositum, is for the most part very doubtful. Of the conjectures which have been put forth regarding these suffixes, there are only a few to which one can allow a certain amount of probability. To these belongs especially the hypothesis, that a part of the personal endings were originally independent substantival pronouns, cp. the -in of the 1. sg. impf. Indg. *ebherom (Skr. dbharam, Gr. trpfoov) with the pronominal stem *-me (Lat. me, Goth. mi-k). There are in the Indg. languages many words of which we can not prove that they either contain or ever did contain a suffix. These, therefore, apparently represent the form they had in the pre-inflexional period, e. g. Gr. /.IF, Goth. mi(-K) cpf. "me; Skr. nu, Gr. w\ Lat. nu(-dius), O.Ir. nu no, OHG. nu no, Lit. nii{-gi) 'now'. In other cases, root-words had disappeared in those compositions which we call inflected words. W e must guard against fancying that, towards the close of the primitive ppriod, or even later, elements like es, which we abstract as root from such forms as Indog. "esti, (Gr. ftfn, Skr. dsti), had an independent existence and a meaning without any definite syntactical relation.
On the use of hyphens.
§ 14. According to the analogy of such forms as Fr. visà-vis, rouge-gorge, it has become usual to indicate component parts of inflected words by hyphens, e. g. Gr. u-/ut, nu-réç-sg, -g 'poison*
(nyog 'frost*: Lat. frigus
ni-vio 'I drink', imper. ?u-6i : Skr. pl-ta-s
(§ 5 7 0 ) .
'drunk, having drunk'
pl-tl-s 'draught (of liquids)', O.Bulg. pi-ti 'to drink' pi-vo
(neut.)
'potus, beer'. § 41. in-cli-no,
Italic.
L a t . vivos,
ac-cli-nis
'leads on, unites with'. 'sis', Osc. fefac-i-d
Osc. b i v u s nom. pi. 'vivi'.
: Gr. xli-vr)
Skr.
xXl-rv-g,
L a t . s-T-mus,
Lat.
abhi-Sri-na-ti
Umbr. 2. pers. sing,
sir
'fecerit' opt. of redupl. perf. (cp. Skr. 3. pers.
sing. opt. pf. mid. va-vft-i-td
from vart-
'roll,
turn');
the -i-
(-£-) in the Umbr. and Osc. forms, as well as in L a t . sim (O.Lat. pers. pi.
siem sies),
has been
taken
over from the
sis
1. and 2 .
Osc. l i - m u 'famem' : Gr. Xl-fio-g.
In L a t i n ei was also written for this t, e. g. veivos,
faxseis.
It was not, however, pronounced as a diphthong, but as a long (probably open) i. feido
The old diphthong ei in deico (cp. Ssix-vv-fxi),
(cp. nsiOut) etc. had become I, but the spelling ei for this T
was still retained.
A t this period the sign ei began to be used
for the old i which had fallen together with ei (cp. also late Greek spelling si for original i, as noXsirrji).
the
Cf. § 65.
§ 41—44.
Indg. l in O.Ir. Germ. Balt.-Slay.
39
The i in the U m b r . and Osc. words sir and b i v u s , l i m u (quoted above) is to be pronounced long. This quantity can be directly proved in Umbr. p e r s n i h m u persnihimu,
i. e. persmmu
(§ 2 3 ) 'precamino', depon. 3. sing, imper. from a verbal stem persnl-,
cp. Lat. f ini-re.
It also follows from the spelling ei (on
tables in Latin alphabet) in sei beside sir si =
Lat. sis, in veiro
beside viro 'viros' (cp. Skr. vira-s § 38), probably also in pers-ei beside pers-i 'quid', the ending of which was a demonstrative particle corresponding to the -I in Gr. ovroa-i.
This spelling
ei is to be put on a like footing with the ei in Lat. veivos and similar forms (see above), and points to an open pronunciation of the i.
The same quality of the «-sound follows for Oscan
from fusid 'foret, esset' (opt. of the s-aorist, cp. Skr. 3. sing, mid. mq-s-T-td from
man- 'think'), assuming
that
the
vowel
in the end syllable had not been shortened and then become i. § 42. Iwor.
Old Irish.
1% Cymr. lliw 'color, splendor':
It is less certain that the I in crithicl emax =
Lat.
Indg. T:
1
Skr. kri-nd-ti 'buys'. For the treatment of i in unaccented syllables see §§ 613. 634. 657,
§ 43.
2. 6.
Germanic.
Goth, skeirs (st. skei-ra-~) O.Sax. skir
O.Icel. skirr clear, bright', prim. Germ. *ski-ra-, related to Gr. axia 'shadow' or to L a t . de-sci-sco 'I withdraw, depart from'. Goth, freidjan frida
'to spare', OHG. vrlten 'to foster, pamper', O.Icel.
'to adorn, decorate': cp. Skr. pri-td-s
'dear, pleased, joy-
ful, satisfied'. Goth, sv-ein OHG. OS. swin Ags. swm O.Icel. svin 'pig': O.Bulg. sv-inu Lat. su-inu-s
'pig's (adj.), belonging to a
pig'. 1. pers. pi. opt. pf. Goth. vSs-ei-ma OHG. war-i-m
O.Sax.
war-i-n, O.Icel. vcer-i-m 'we might be': cp. Skr. 3. sing. opt. pf. mid. vavft-i-td (from vart-~). It must be observed that ei in Goth, always signifies the monophthong l (§ 25). § 44.
Baltic-Slavonic.
Lith. vy-ti
O.Bulg.
wind, turn' : Gr. l-xta 'willow, salix' fr. *fZ-rsa.
vi-ti 'to
Lith
dvynil
(dual) 'twins': L a t binT. O.Bulg. griva 'mane': Skr. grwa- 'nape'. Observe that y is the Lith. symbol, for I (§ 26).
40
Prim. Indg. u.
Indg. u in Ar. Armen.
§ 45—47.
I n d g . u. § 45. I n d g . *jug-6-m 'yoke : Skr. yugd-m, Gr. £vyo-v, L a t . jugu-m, Goth, juk, O.Bulg. igo fr. *iugo. Indg. weak stemform *ku-n- 'dog' (strong stemform *ku-on-, Skr. acc. sing. ¿van-am): Skr. gen. sing. Sii-n-as, Gr. gen. sing, xv-v-og, O.Cymric nom. pi. ciin cwn, Goth, hun-d-s, Lith. gen. sing, szu-n-s fr. *szu-n-es (§ 664, 2). Indg. *nu now': Skt. mi, Gr. vv, Lat. nu-diu-s, O.Ir. nu no, OHG. nu no, Lith. nii-gi 'now then'. W e a k nominal suffixes Indg. -u-, -tu-, -nu-: Skr. svad-u-s, Gr. i]d'-v-g 'sweet', Lat. fruc-tu-s, O.Ir. fiuss fiss 'knowledge' from *uissu-s, Goth, su-nu-s Lith. sii-nu-s O.Bulg. sy-nu (from *sy-nu-s, § 665, 4) 'son'. W e a k form of the suffix of the pf. part. act. Indg. -us-: Skr. vid-ui-T Gr. Idvia from '*ftd-vo-iu (§ 131) fem. 'knowing', Goth, ber-us-jos 'parents' (properly 'those who have brought forth'), Lith. vilk-us-i O.Bulg. vluk-usi fem., to Lith. velku O.Bulg. vlekn 'I draw'. § 46. A r y a n . Skr. putra- Avest. pupra- O.Pers. pusa(§ 261) 'son': Osc. p u k l o - 'son'. Skr. upa Avest. upa O.Pers. upa (read upa) prep, 'to': Gr. vno, Lat. s-ub. Skr. su- Avest. hu- O.Pers. u- prefix = 'well, good, very': O.Ir. su- so-. Suffix of the loc. pi. Ar. -su: Skr. aive-su Avest. aspae-'m 'in equis', 0 . P e r s . mUdai-suv-a 'in Medis' (-a is postposition): O.Lith. knygo-su (Mod. knygo-se is formed by analogy) 'in libro' from knygos pi. 'liber', O.Bulg. rqka-chu 'in manibus' from rqka 'manus'. In Sanskrit ru before a following u seems to have passed into % (dissimilation). Present stem Sf-mi- ('hear') from *Sni-nu-, cp. Avest. 2. pers. sing. opt. srunu-ya, (Skr. Spiu-yus), O.Ir. 1. pers. sing. pres. clu-ni-m, rt. Men-, bhf-kuti-s 'distortion of the eye-brows' beside bhru-bhanga-s bhru-bhatdga-s the same. Cp. Osthoff Morph. Unt. I V . 215 f. § 47. A r m e n i a n , dustr 'daughter': Skr. duhitarGr. dvyaryp Goth, dauhtar (§ 51) Lith. dukte O.Bulg. dusti 'daughter'. surb 'clean': Skr. iubhra-s 'clear, bright', nu 'daughter-in-law': Skr. snusa, Gr. wo-g O H G . snur O.Bulg. snucha 'daughter-in-law'.
§ 47—49.
Indg. u in Greek and Italic.
41
u in non-final syllables was dropped: dster gen. of dustr, srboy gen. of surb, verb srbem 'I clean'. Cp. § 632. § 48. G r e e k , xkv-ro-g 'celebrated': Skr. ¿ru-td-s 'heard, celebrated', Lat. in-clu-tu-s, O.Ir. cloth from *cla-to-s (§ 50) 'celebrated', OHG. Hlud-rich Hlud-olf Hlot-hari (§ 51). VTTHQ VTISQ 'over': Skr. updri 'above, over', Lat. s-uper, Goth, ufar 'over'. "Weak present suffix -mi- in ru-w-rui: Skr. ta-nu-te, from root ten- 'stretch', Lat. mi-nu-d. In prim. Gr. and probably also for some time after the introduction of the letter Y, u was spoken by all Greeks. At an early period one part of the dialects changed u into u which in the christian era was further changed into i. Another part, the Boeotians, Laconians, and probably also the Arcadians, Cyprians, Pamphylians, Chalcidians and Lesbians, retained u until far into historical times, e. g. Bceot. hyovoo-g (OD — u) = Att. hyvQo-g. Meanwhile the vowel u in Bceot. underwent a modification which cannot be precisely defined, after r, 6, J, v, i, as is seen by the spelling tov instead of ov, e. g. riohov-'$svog = Att. Jlokv-'StvoQ; cp. Osc. iu = u § 49. The spiritus asper in vthuo (Skr. uddn- 'water'), vno vneg (Skr. upa upari) etc. seems to point to a transition from initial u- to iu in prim. Greek. Cp. ifing: Skr. yusma-. It is not improbable that v became < through the dissimilating influence of a following v, as mvvto-g from *Ttv-w-ro-g, Skr. pu-nd-mi 'I clean, clear up'. § 49. I t a l i c . Lat. rubro- (nom. ruber) Umbr. r u f r o - 'red': Skr. rudhira-s Gr. egvOgo-g O.Bulg. rudru 'red'. Lat. duo Umbr. neut. t u v a 'two': Skr. ved. duvd Gr. Sva 'two'. Osc. p u - k l o Pelig. Mars, pu-clo- 'son' Lat. pu-tu-s 'boy': Skr. pu-tra-s 'son. In Latin, before labials and perhaps also before I, a frequent sound, lying between u und i, made its appearance for Indg. u. libet, older lubet: Skr. lubh- 'to feel a strong desire', fimu-s fimu-m, probably related to Gr. Bv-fio-v. limpa (limpidus), older form
42
Indg. u in Italic and O.Ir.
lumpa
§ 40—50.
to Osc. D i u m p a i s 'Nymphis', prim. Italic *dumpa-1).
relation between silua silva and Gr. i'ly
The
is uncertain (cp. Osthoff
Morph. U n t . I Y 158 f., W . Meyer Kuhn's Ztschr. X X V I I I 1 6 3 f.). This i occurs lacri-ma, festos
especially
older form lacru-ma, beside lacu-bus
to the
often for the -u- of «-(eM-)stems, e. g. laci-bus
manu-festus.
flucti-bus
flucti-vagus
Compare
the change
same u\ ¿-sound (§ 9 7 ) in unaccented
mani-
syllables
of a before
the same consonants. R e m . It has been conjectured (Osthoff in the above named work pages 125, 158) that the i in libet,, fimus, limpu, silua arose from u phonetically only in unaccented syllables, in such combinations as qui lubet, cdnlubet, in fiimo (cp. ilico from *in slocij). But should not the historical relation of loebertatem leiber to Osc. L u v f r e i s 'Liberi' Falisc. loferta 'liberta' be such that in Lat. *loub- ( = prim. Italic *louf-, § 65 page 52) the consonantal u before b underwent precisely the same change as the sonantal u of lubet, so that *lofib first became *loib, then loeb, and lastly lib-? In that case the change of lubet (with u in the chief accented syllable) to libet could scarcely be questioned. For the change of loeb- to lib- cp. § 8 1 rem. 2. I cannot resolve to separate Lat. liber liberta from Osc. L i i v f r e i s and Falisc. loferta, although I do not fail to recognise the overwhelming difficulties which present themselves (s. Danielsson in Pauli's Ital. Stud. 4, 156 if.). T h e spelling with o shows that u in Umbr. had, in certain cases or everywhere, an open sound: sopam Bummum', trifo
'tribum'.
'suppam, supinam',
somo
F o r Umbr. i- from iu- see § 133.
I n Osc., u after d, t, n, s underwent a change which represented by iu. D i u m p a i s 'Nymphis', Lat. lumpa e i t i u v a m beside eituam 'pecuniam' (suffix -u(u)a-). ( a borrowed w o r d ? ) . Lat. numeru-s. § 50. su-td-s sru-td-s
(see a b o v e ) ; tiurri'turrim'
N i u m s i e i s Nivitrfdir^t; gen. 'Numisii',
S i u t t i i s 'Suttius'.
0. Irish,
Gr. vlv-g
suth
vlo'-g 'son',
Cp. Bceot. IOV =
to
ov § 48.
neut. "birth, fruit' fr. *su-tu:
Skr.
sruth
Skr.
neut. 'river' fr. *sru-tu:
Gr. pv-ro-g 'flowing', Lith. sru-ta, 'dung-water',
'I hear': Skr. ¿ru-
was
clu-ni-m
Gr. xAi>-, rt. fcleu-.
1) The spellings of the Lat. word as lumpha, limpha, lympha are due to the supposed derivation from vvfipt], which even gave rise to nymphaticus ( L o w e Archiv. f. Lat. Lexikogr. I 25).
§ 50-52.
Indg. u in Germ. Balt.-Slav.
43
Compare also Gall. Druid-es, O.Ir. drui (pi. druid) perhaps to, Skr. dru- Gr. Jou- 'wood'; Gall, catu- combat' in Catu-rlg-es etc. (O.Ir. cath, gen. catha): OHG. hadu- 'war'. u became o through the assimilating influence of a or o in the following syllable (cp. the similar change of i to e § 34). cloth from *clu-to-s: Gr. y.Xv-to-q etc. (§ 48). srotho srotha gen. of smth. bond bonn 'solea' fr. *bundo-s: Lat. fundii-s Skr. budh-nd-s 'bottom' (§ 469, 8). Compare West Germanic and Norse o from u § 51. u became o in monosyllabic words, nu no — Skr. nu etc. (§ 45). sit- so- 'well' = Skr. su-. du- do- 'bad' = Skr. dus- Gr. J r s Goth. tuz-. In this case too it is probable that the change from u to o was caused by the vowels of the following syllables. For the fate of u in other than chief accented syllables cp. §§ 634. 657. § 61. G e r m a n i c . 1. pers. pi. pf. Goth, ana-budum 'we bade', OHG. butum O.Icel. budom 'we offered': Skr. bu-budh-imd 'we watched, observed', Gr. ns-jivorcu nvH-iafiai, Lith. bud-inti 'to rouse', O.Bulg. bud-iti 'to wake', from rt. bheudh- 'wake, observe'. Goth, ga-kusts (st. ga-kusti-) OHG. knst 'test, choice': Skr. justi-s 'show of affection, favour'; kustu-s 'trial': Lat. gustu-s 'taste' from rt. geus- 'choose, ysvtaOai'. Before h and r Indg. u appears in Goth, as au. auhsa (st. atihsin-) 'ox': Skr. uksd (st. uks&n-') 'bull, ox', daur (st. daiira-) neut. 'door': Skr. dur- (concerning the d- see § 480) Arm. durn Gr. dvgcc Lith. pi. durys 'door'. In West Germanic and Norse u became o through the influence of an a or o of the following syllable. Part, passive OHG. botan (Mod. High Germ, ge-boten) O.Sax. bodan Ags. boden O.Icel. bodenn = Goth, budans, cp. Skr. part, budhand-s. OHG. tor = Goth, daur (prim. Germ. *dura-m) beside turi 'door. Cp. §§ 247. 299. R e m . Some scholars look upon this umlaut also as prim. Germ, (cp. § 35 rem.).
§ 5 2 . B a l t i c - S l a v o n i c . Lith. kruvina-s O.Bulg. kruvlnti 'bloody': Lat. cruentus. Lith. pi. musai (stem musa-) 'mould
44
Indg. u in Baltic-Slavonic.
§ 52.
on sour milk', O.Bulg. muchti 'moss': Lat. mus-cu-s, OHG. mos (prim. Germ, stem *musa-) 'moss'. Lith. su-nii-s O.Bulg. sy-nu 'son: Skr. su-nu-s 'son'. In some L i t h . dialects u is pronounced very open, so that it can be regarded as a closed o, e. g. sulcau suho (1. and 3. pers. sing. pret. of suku 'I turn') appear as sSkaii, sdko. S l a v o n i c u must, already in prim. Slav., have been a reduced sound, approaching the quality of o. u, in Modern Slavonic developments, became a full vowel in closed syllables (such syllables often arose through the falling out of an i or u in the following syllable), Russ. o, Serv. a, e. g. Russ. son Serv. san = O.Bulg. surtu from *supnu (§ 545) 'sleep': Gr. vn-vo-g; Russ. doc = O.Bulg. dusti 'daughter': Lith. dukte. On the other hand final u disappeared, e. g. Russ. syn O.Bulg. synu, loc. pi. Russ. rybacli Serv. ribdh = O.Bulg. rybachu (from rybafem. 'fish'), cp. Skr. loc. pi. asvcl-su (from diva- 'mare'). Medial u also, as a rule, disappeared in open syllables, e. g. Russ. Serv. sna = O.Bulg. suna (gen. sing, of sunu), Russ. spat = O.Bulg. supati 'to sleep', legit a = O.Bulg. ttguka nom. fem. of ttgu-ku 'easy' (further formation of an old it-stem by means of the suffix -ko-). All these modifications can be traced as far back as the period of old Church Slavonic. On the one hand they follow from such double spellings as Jcrovt kruvi 'blood', domochu domuchu (loc. pi. from the stem domu- 'house' = Lat. domu-~)\ these forms can only be understood by assuming that the final i and u were no longer syllabic; on the other hand from such double spellings as reksu rekusu dat. sing, of the stem rekuse-, that is, *rek-us-ioaccording to §§ 84 and 147 (part. pf. act. of rekq 'I say'), leguko beside liguko neut. to liguku. With these modifications of u, compare the parallel treatment of i § 36. Prim. Slav, iu became n (cp. ii from iy § 60 and ie from § 84). O.Bulg. zivq 'I chew' from *ziuvq: OHG. chiuwu (§§ 143. 147). igo 'yoke' from *ngo (§ 145) and this fr. *iugo:
io
S k r . yugd-m,
Lat.
jugu-m.
§ 52—56.
P r i m . Indg. u.
45
Indg. » in Ar. Arm. Greek.
R e m . In the Zographos gospel i f o r it occurs occasionally before palatal
vowels, e. g. bideti = budeti 'to wake', duvit m., but dive fern, neutr. 'two'.
I n d g . u. § 53.
I n d g . *dhu-mo-s 'ebullition, smoke': Skr.
Gr. ddfio-g (courage, passion'), Lat. fumn-s, O.Bulg. dymu.
dhumd-s,
Lith. pi.
dumai,
Indg. *bhii-, weak form of the the rt. bheu- 'be,
become' (§ 3 1 2 ) : Skr. bhii-td- 'become, been' dbhu-tam 'ye two were' bhu-ti-s 'good condition, well-being', Gr. (pv-raXtrj 'a planted place' 2. pers. du. scpv-rov, Lith. inf. bu-ti O.Bulg. by-ti 'to be'.
Indg. *mus- 'mouse': Skr. mus-, Gr. /.ivg, Lat. was
OHG. mus and O.Bulg. mysi fr. *mychi
mur-is;
(§ 588, 2 ) , that is,
originally *mus-i-s, show the change of the monosyllabic stem into the «-declension. Indg. -u as ending of the nom. acc. du. of ew-stems: Skr. sunu O.Bulg. syny from Indg. stem *su-neu- *su-nuzmogu
fr. *imogu
son', Lith.
(§ 664, 3 ) to smogu-s 'man', O.Ir. mug
fr.
*mogu (§ 657, 2 ) from stem mogu- 'servant'. § 64.
Aryan.
Skr. du-ra- Avest. O.Pers. dii-ra- 'far' from
rt. deu- 'withdraw* (Gr. Horn, ¿eveodcu 'fall short of).
Skr.
iu-ra- 'hero', Avest. su-ra- 'strong': Gr. a-xvgoi; 'weak, invalid*. Skr. wit Avest. nu 'now': Gr. vv-v O.Bulg. ny-ni 'now*. § 55.
Armenian.
I do not know any sure examples for
the regular representation of Indg. u.
The following comparison
of words may, however, be taken into consideration, ku-oy,
'dung' to Skr. gutha-
ku, gen.
'excrement', Avest. gupa-
'dirt'.
mukn, gen. mkan, 'mouse, muscle', to Skr. musika- musakcL- 'rat, mouse,' O.Bulg. mystka 'mouse', Lat. musculu-s (cp. Hiibschmann Zeitschr. d. deutsch. morgenl. Ges. X X X V 170, Osthoff Morph. Unt. I V 217 f.).
PI. srun-Ji ( - « i s the sign of the plural) 'shins,
shanks', to Lat. crur-a (Hiibschmann, as above, page 177). Compare also Hiibschmann Armen. Stud. I 62. § 58.
Greek,
oypvg bcpgvv 'eye-brow': Skr. bhru-S O.Icel.
bru-n Ags. bru 'eye-brow'. fiov-Xv-xo- 'the time for unyoking oxen': Lat. so-lu-tu-s, O.Icel. ludr 'bruised, crushed, deafened,
46
Indg. u in Italic, O.Ir. Germanic.
tired out' (cp. Ivto yovvata). sthu-la- 'coarse, thick, massive'.
otv-Io-q 'pillar': Skr.
§ 56-58.
sthu-rd-
The change from u to il went parallel with that of u to u (§ 48). The dialects, which retained w, retained also u: Boeot. EvOov¡.IO-Q = Att. Ev-8vf.to-g, Lac. /uovaidSsi (which according to Hesychius means XaXsT) — Att. *luvdi£ei. It is doubtful whether v became l through the dissimilating influence of a following v (cp. § 48). cpT-tv (neut.) 'twig, sprout' has been brought together with Skr. gerund bhu-tv-a and Lat. fu-tu-o. It may, however, be connected with Lat. fio and O.Ir. biu 'fio, sum', the original form of which was *bhu-ii6 (§§ 170. 175). It is still more uncertain whether iov-g has come from *c8v-g (Osthoff Morph. Unt. I Y 188 ff.). § 5 7 . I t a l i c . O.Lat. pf. fu-l fH-imus (Inscript. fuueit), Osc. F u u t r e i 'deae creatrici': Skr. bhu-ta- etc. (§ 53). Lat. su-s sii-bus: Skr. su-kara- 'pig, boar', Gr. v-g v-v, OHG. sm (Mod.HG. sau) from prim. West-Germ. *su-2 (§ 583). Lat. cupa 'vat, recess for the dead': Skr. kupa-s 'pit, hole'. It is quite possible that in Umbr. the i in p i r {nvg) beside abl. p u r e Osc. p u r a s i a i (loc. sing, of the stem purasid- 'igniaria'), in s i m (Gr. vv, Lat. suem) and sif (acc. pi.) beside s u r u m sorsurn 'suillum', and in frif acc. pi. 'fruges', represents an older u which may have been pronounced ii. (cp. O.Bulg. y fr. w beside u fr. m, §§ 52. 60.). It is however impossible to form a correct judgment owing to the quantities not being marked, and the scantiness of the material. Cp. Osthoff Morph. Unt. I Y 16, Zur Gesch. d. Perf. 432. § 58. O l d I r i s h , run (fem.)'secret': OHG. runa O.Icel. run f. 'secret, rune', dun (neut.) 'castle', Gall, -dunum in names of towns (Lupo-dunum and others): O.Sax. Ags. masc. tun OHG. masc. zun O.Icel. neut. tun 'hedge, enclosure'. Pron. tu 'thou': Avest. tu Gr. TV-V-TJ Lat. tu O.Icel. pu O.Bulg. ty. For such forms as acc. sing, ruin beside nom. sing, run see § 640. For u in unaccented syllables cp. §§ 613. 634. 657.
47 § 69. G e r m a n i c . In Goth., where u and u in the native alphabet were not kept apart, u can scarcely ever be established without the help of the other languages. Goth, f u l s OHG. Ags. f u l O.Icel. f u l l 'foul': Lith. pi. pulei 'matter', Gr. nvOio 'I make to rot, putrefy', Lat. pus puris 'matter, spittle' pu-ted 'I smell bad, stink'. Goth, rums OHG. O.Sax. m. rum, O.Icel. neut. rum 'room, open place', from the same root as Lat. rus ruris 'open field, country', Avest. ravah'free'. OHG. utar Ags. uder 'udder': Skr. udhar 'udder'. OHG. hlut Ags. O.Sax. Mud 'loud' from rt. fdeu- 'hear': cp. Gr. imperative y.Xv-rt, O.Bulg. sly-sati 'to hear', sly-tije neut. 'call, name'. § 80. B a l t i c - S l a v o n i c . Lith. tula-s adj. 'so many a', originally a substantive denoting 'a mass', O.Bulg. tylu 'nape', originally 'intumescence': Skr. tida- neut. 'panicle, tuft, cotton', Gr. tvXrj 'a swelling or lump, callosity', from rt. teu- 'tumere'. Lith. sura-s 'saline', O.Bulg. syru 'raw': OHG. sUr O.Icel. surr 'sour, bitter'. Lith. jus 'ye': Avest. yus yuzem 'ye'. O.Bulg. svekry 'mother-in-law': Skr. ivasru-s 'mother-in-law'. The change of u to y belonged to the prim. Slav, period. Prim Slav, iy became ii (cp. ii from iu § 52). O.Bulg. siti fr. *siiti (§ 147), and this fr. *siyti : Lith. siuti 'to sew', Skr. syuti-s fem. 'sewing, sack'. Compare also kraji instr. pi. of k r a j i 'edge' but raby instr. pi. of rabu 'servant'. I n d g . e. § 01. I n d g . pres. stem *bhero- *bhere- 'bear': Skr. bhdrUmi, Arm. berem, Gr. tpsgco, Lat. ferd, O.Ir. berirn, Goth, baira, O.Bulg. berq. Indg. *es-ti 'is': Skr. dsti, Gr. S6TI, Lat. est, O.Ir. is (proclitic), Goth, ist, Lith. esti (now archaic), O.Bulg. jestt. Indg. e in the root syllable of neuters in -es-, e. g. Indg. *regos 'vapour, darkness': Skr. rajas, Arm. erek, Gr. sQepog, Goth. riqiz-aIndg. *nebhos, 'cloud, atmosphere': Skr. ndbhas, Gr. vstpog, Lith. debes-i-s (the origin of the d- instead of n- is not clear), O.Bulg. nebo, comp. also with /-suffix Gr. vt(piXtj Lat. nebula O.Ir. nel (prim. kelt. stem *neblo-, § 526) OHG. nebul 'cloud, fog'. Indg. *qe 'and': Skr. ca, Gr. re, Lat. que. Augment
48
of the past tenses of the Indicative (originally an independent particle) Indg. e: 3. pers. sing. Skr. a-dpSat Gr. E-Jpccxe (rt. derk'see'), Arm. e-UJc Gr. s-hns (rt. leiq- 'leave'), Goth, i-ddja Skr. a-yat 'he went' (for Goth, ddj see § 142). •e Indg. ending of voc. sing, of o-stems: Skr. vfka Gr. Xvxs Lat. lupe 'wolf, O.Ir. maicc maic from prim. kelt. *makue 'son', Lith. vilkb O.Bulg. vluce 'wolf. The 'thematic vowel' Indg. -e(by the side of -o-): e. g. 2. pers. pi. Skr. bhar-a-tha, Gr. (ptQ-e-Te, Lat. veh-i-tis, Goth, bair-i-p, O.Bulg. ber-e-te. Nominal suffix Indg. -es-: gen. sing. Skr. ndbh-as-as Gr. vt O.Pers. 3. sing. pret. mid. a-pata-ta (read -ta): Gr. nits-rai, Lat. peti-t. Skr. pdri Av. pairi O.Pers. pariy (read pari) 'round, about'; Gr. nsQi. Indg. tautosyllabic ei, eu became ai, au in prim. Aryan, thence Skr. e, o, Av. ae, oi, ao eu, O.Pers. ai, au (cp. §§ 126. 160). Skr. e-ti Av. aeiti (= *ae-ti, § 638) O.Pers. ai-tiy (read aiti) 'he goes': Gr. slot. Skr. bodhate Av. baodaite : Gr. nevdsrui,. Indg. *bheudhetai from rt. bheiidh- 'guard, watch, observe'. O.Pers. daustar- 'friend', Skr. jostdr- 'one who desires' from rt. geus- 'choose' : cp. Gr. ysvariipto-v 'a thing to taste with, a cup'. R e m . AY. final ae became e, see §§ 78, 94. Here would have to be mentioned loc. sing, aspe beside aspoi and aspaf-ca {-ca — Gr. Tf, Lat. -que) — Skr. dive (from Indg. stem *chiio- 'horse'), if -ai, the prim. Ar. ending of o-stems, arose from Indg. -ei (cp. oixei). This, however, can also be traced back to Indg. -oi. Cp. § 116 and the accidence.
For Av. e, i, see § 94. Brugmann,
Elements.
o from prim, Iran., prim. Ar. a =
Indg. e 4
50
Indg. e in Armenian and Greek.
§ 63—64.
§ 63. A r m e n i a n , cer'oldman': Gr. ytpiov 'old man', Skr. járant'fragile, old man', eln (gen. eXin) 'stag': Gr. tXayog (orig. form *el%-bho-s) 'stag', Cymric elain 'cerva', Lith. éln-i-s 0.Bulg. jelen-i 'stag', cpf. *elen-. evfn] 'seven': Gr. hmá, Lat. septem, O.Ir. secht n- (§§ 211, 243), Goth, sibun, Lith. septyni, O.Bulg. sedmi, Skr. saptá. ev and, also': Gr. 'tm 'on, at', Skr. dpi 'also', cpf. *épi. e became i before nasals, hin (gen. hnoy) old': Gr. tvo-g 'old', Lat. sen-ior, O.Ir. sen 'old', Goth, sin-ista 'the oldest (man)', Lith. sena-s Skr. sdna-s 'old', Indg. *seno-s. cin 'birth' : Gr. yévog. hing 'five' : Gr. nivzi. Occasionally a is found instead of e, e. g. tasn 'ten': Gr. Séxa, Lat. decern. The circumstances under which a occurs for e have not been ascertained. Indg. tautosyllabic ei appears as e in final syllables, in other syllables as i. dizem 'I accumulate', orig. form *dhéighd (cp. Gr. rsTxog 'wall', Av. pairi-daeza'fence'), aor. l.sing. dizi, 3. sing. e-dez. lizem, lizanem 'I lick': Gr. Xsi'/w. jiun 'snow': Gr. xa/uwv. Indg. tautosyllabic eu appears -as oi in final, as u in nonfinal syllables, lois, gen. lusoy, 'light': Gr. ksvxó-g 'white', Goth. liuh-ap 'light', unain empty' from *oin-ain: Gr. ev-vi-g 'bereft'. eu probably in the first place became ou; cp. § 79 and nor (gen. noroy) new' as compared with Gr. vsfo-g, vs(^f)aQ¿-g. § 64. G r e e k , ennui 'he is together with, follows': Lat. sequitur, O.Ir. sechur 'sequor', Lith. sekü 'I follow', Skr. sácate 'he is together with, follows', Indg. 3. sing. pres. mid. *séq-e-tai. 1. sing, fa) acc. dasQa, Lith„ dever-l-s. gero, ingero, sceleris with r from s. (cp. § 569).
53 2. Finally. lupe : hiy.t. age agite : ays aysrs. quTnque 3. As a rule before double consonants: scelestu-s, (ob-sideo),
au-spex
:
nivrs.
ob-sessu-s
(au-spic-is).
Change of e to i: 1. According to prim. Lat. accentuation in unaccented syllables (§ 680) followed by a single consonant (except r). •colligo pro-tinus
: lego
Xiyw.
: tenus.
ob-sicleo agite
: sedeo
agito
£ Jo?,
: tiysrs
oppidum
dyerco.
: Pedum
agiminl:
nsSov. ayo/xevoi.
R e m . 2. The i instead of e in plied (Grr. nt/xia), O.Lat. spicio (rt. spek-) miniscor (rt. men-), and others, is to be explained from the fact that the form, w h i c h regularly arose in composition, made itself independent (com-plico, con-spicid, re-miniscor). R e m . 3. juven-em with -en- instead of -in- was, perhaps, influenced by juventus juvencu-s etc. ap-peto, ex-peto for *ap-pito, *ex-pito re-formed from analogy with the simple form peto, or else formed at a time w h e n the activity of the law had died out according to which colligo etc. a r o s e ; cp. exquaero beside exquiro, exaequo beside iniquo-s, conclausu-s beside conclusu-s etc.
2. Independent of accent, in closed syllables before nasals. quinque1)
: nivrs.
tinguo
: rsyya
'I wet',
in
intus
: tv
evrog.
To
which may be added dTgnu-s beside decet, as well as lignu-m, tlgnu-m, in case they really are related to lego 'I gather, collect' and rir/vrj; for gn was pronounced mi (§§ 500. 506). This change also affected en and em from n and qi (§ 237), e. g. mgintl: Dor. f iy.au, sim-plex : a-nat. There are, however, in both cases many exceptions, the explanation of which is difficult, e. g. offendimentu-m (fend— Indg. *bhendhand -mento= Indg. *-mnto-~), supine ventum ( = Skr. gantu-m, Indg. *gem-tu-m), part, in-ventu-s ( = Gr. [iaro-g, Skr. gatd-s, Indg. *gm-t6-s). Cp. further decern ( = Skr. ddsa, Indg. *deHm) beside undecim.
For Lat. -o- from -ue- see § 172, 3. Tautosyllabic ei seems to have still been a diphthong in Lat. at the period of the oldest monuments, cp. the spellings deivos deiva, deico, feido and others. The diphthong, however, 1) Concerning tignu-m s e e § 619.
the
long
vowel
in quinqué,
dignu-s,
lignu-m,
and
54
§ 65-66.
at an early date became an open i, and the sign ei now became also employed for the representation of open »-vowels (or close e-vowels) of other origin (see §§ 41. 73). In U m b r i a n and O s c a n Indg. e was mostly represented by e. Besides the examples already mentioned, the following may be compared: Umbr. petur-pursus 'quadrupedibus', Osc. petora 'TSTTU(JU, Umbr. co-vertu 'convertito', destram-e 'in dextram', Osc. dat. G e n e t a i 'Genitae' (to Lat. genitor). Besidethese occurs also i, Osc. i, e. g. Umbr. t i s i t 'decet', Osc. 1st 'est' (beside e s t u d 'esto'), n i s t r u s 'propiores' (beside nest mum 'proximum'). Owing to the scarcity of the material it is impossible to decide whether the change to i (open) took place under certain conditions, unknown to us, or whether e had everywhere the same close pronunciation in Umbrian-Oscan, and that we have here to do with a mere irregularity of the graphic representation of the sound. Tautosyllabic ei was represented in Oscan by e i , ei and (rarely) ii. d e i k u m deicum inf. 'dicere'. d e i v a i dat. 'divae', deivatud imper. 'iurato', d i i v i i a i dat. 'divinae' (a derivative like Umbr. deveia 'divina'). t e r e i loc. 'in terra', cp. Gr. o'ixsi. The spelling ii points to a passage of the diphthong into open i. In Umbrian ei certainly became closed e. e t u eetu 'ito'. deveia 'divina'. desire loc. 'in dextro', cp. Osc. loc. t e r e i . § 66. O l d I r i s h , e was often retained as such in chiefaccented syllables, con-dercar 'conspicitur': Gr. ¿¿(jxouca 'I see', from rt.
derfc-.
celim
'I hide':
O H G . hilu
' I hide', cpf.
*keld.
Gen. meda 'of mead', Skr. madhu 'sweetness, honey', Gr. fi£8v 'wine', OHG. metu 'mead', O.Bulg. medu 'honey', Indg. *medhu. ech, Gall, epo- (in Epo-redia name of a town): Lat. equo-s Goth. alhva'horse',
(in aihva-tundi
reTTapsg. news,
'thorn
b u s h ' ) S k r . dsva-s
Indg.
tech teg 'house' f r o m *tegos : G r . Teyoc- cethir
Cp. further
Indg.
*efcuo-s
'four': Gr.
Gall, ne^ns-dovla 'quinquefolium': Gr.
*pStsqe.
In such cases as: gen. eich 'equi' from prim. Kelt. *ekul (-» still in Ogam inscriptions, as maqui 'filii', and in Gall. Segomari etc.): Lat. equi; 3. sing. pres. as-beir 'he brings forward,.
55 says' from *berit, prim. Indg. *bheret; deich n- 'decern' from *deken, Indg. *dehm (§§ 242. 248) the i after the e only indicates the palatal timbre of the following consonant, which had been produced through the palatal vowel of the next following syllable. In a corresponding manner the «i-(o-)timbre of the following consonants is represented by u and o in the dat. sing. neurt, eoch from *nertò, *ekijó (nom. nert, ech, stem nerto'power, might', echo- 'horse'; with nerto- cp. Gall. Nerto-marus, Coh-nertus etc.). See § 634. Under certain conditions chiefaccented e became i before a following palatal vowel or u. tige fr. *tegesos, gen. sg. to tech 'house', as-bir 'thou sayest' fr. *biri(s), prim. f. *bhere-s. mid 'mead' fr. *medu, to gen. sg. meda, s. above, as-biur 'I say' fr. *biru, prim. f. *bherd. e in other than chief-accented syllables, in case it did not entirely vanish, (§§ 634. 657), became an irrational vowel, the quality of which adjusted itself mostly to the timbre of the following consonants. i. as-bérid 'you say' = Gr. cfs^ivs. is 'est' always proclitic = Gr. sffrt ; the accented form would be *eis. a. as- 'ex' in as-bérid, as-rénaim 'reddo' and others beside és- (O.Gail, ex-, Cymr. eh-: Lat. ex, Gr. f f ) in és-rechtaid 'exlex' és-orgun 'a striking, killing' etc. u. (ni) épur (fr. *éd-bur) beside as-biur 'dico' fr. *beru, prim. Kelt. *berd = (pipa, cóm-mus 'proportion, weight, worth' to méss 'measure, judgment' fr. *messu-s (*med \-tu-s, to Goth. mitan 'to measure'). In forms like só-nirti adj. pi. 'strong, brave', só-nirte 'strength, power'compar. só-nortu 'stronger' beside nért er had first become S. § 634. I n the later orthography só-nairti só-nairte beside só-nirt só-nirte a merely indicates the timbre of the preceding consonant. S. § 640. Indg. and prim. Keltic tautosyllabic e{ became è, whence ia, if a, o, or u followed, ad-féded 'narrabat', ad-fiadat narrant': Gr. fiióouai, rt. ueid-. F u t . 2. pers. pi. for-tèsid 1. pers. pi. for-tiasam, to 1. sg. pres. for-tiagaim 'I help': Gr. ÓTTI/w HTSI'^CO, Goth, steiga. Voc. de 'O God' from *de(u)e (Lat. deive), nom.
56
Indg. e in O.Ir. and Germ.
§ 66—67.
dia from *de(u)o-s (Lat. deivos). On the 2. sing, for-teis cp. § 640. e — Indg. ei is to be held for prim. Kelt. Cp. e. g. Gall, reda 'waggon Epo-redia, O.Ir. reid 'passable, free, even, smooth' Cymr. rhwydd 'prosper, expeditus' (st. *redi-), O.Ir. riadaim 'I drive, ride', rt. reidh-: OHG. ritan 'to move oneself away, drive, ride' and be-reiti 'ready, equipped' (the latter with rt. form roidh-). Indg. tauto- and heterosyllabic eu. The Irish and British branch concurrently point to oy,. This, before consonants, became in O.Ir. o, whence probably in the first instance ua, when a, o followed; in British u. O.Ir. tuath Cymr. Bret, tut Corn, tus 'people' from *tota: Goth, piuda, cpf. *teuta. O.Ir. loche (gen. lochet) 'lightning', luach 'white' in luach-te 'whitehot': Goth. liuhap 'light', Gr. Isvxo-g 'white'. O.Ir. dat. pi. srua-mannaib, nom. sing, sruaim 'stream': Gr. Qtv-/ua, Indg. stem *sreu-men-. In like manner we have 6 from antesonantal ou in O.Ir. noi n'nine' (beside Cymr. naw Corn, naw Bret, nav nao\ fr. *nouen, Indg. (cp. §§ 65. 174. 233); whereas u in O.Ir. clu 'renown' = Gr. xXbf-og and in nue 'new' beside O.Bret, nowid, to Skr. navya-s (cp. §§ 139. 154). R e m . I f the change from o to ua in O.Ir. was occasioned by an or o in the following syllable, then such forms as nom. sruaim fr. *srontin, dat. tuaith from *toti cannot have been formed according to law, but are merely due to a levelling with the forms of other cases.
a
eu is still found in Gall, in forms like Teuto-matus, Leucetius (Zeuss-Ebel G. C. 34 f.); beside these already also Toutus Toovnovg etc. (as quoted above). In like manner also Nivella, Nivisium, the i of which is also probably a direct continuation of e, but also already Novesium and Novio-dunum 'New-castle' (as quoted above 56). The question is probably here a difference of dialect within the Gallic branch itself. § 67. G e r m a n i c . Indg. e became i in prim. Germ, in the following cases: 1. before nasal + consonant. Goth. O.Sax. Ags. bindan OHG. bintan O.Icel. binda 'to bind', 1. sing. Indg. *bhendho : Skr. bdndhana-m 'a binding' fut. bandhisyd-ti 'he will bind', Lat. of-fendimen-tu-m 'chin-cloth'.
57 R e m . 1. That this i made its appearance already in prim. Germ., is confirmed by the fact that the verb Goth, rinna OHG. rinnu 'I run', whose orig. form *rl-ny,-o contains Indg. i, and which belongs to Skr. ri-nd-ti 'lets flow, lets run', ri-ril-ti 'begins to flow' (Osthoff Morph. Unt. IV. 45), passed over into the series binda : band in its tense-formation in every Germ, dialect (pret. Goth. OHG. etc. rann). This could only have taken place, if the i in bindan etc. had already arisen. 2. In hetero- and tautosyllabic ei. a. N o m . pi. Goth, preis *prii{i)z
*preiez
from *vgs(i)sg, gester cpf.
gestir
: O.Bulg.
Indg. *trei-es.
(§ 6 8 ) ,
Skr. avay-as
cp.
O.Sax. settian
'three' f r o m
Skr. trayas,
Goth, gasteis
'guests' from *^astii(i)z,
*qhostei-es,
iatjan
O H G . drl O.Icel. prlr trije
O H G . gesti
*^asteiez: (nom.
Gr.
sing,
rpng
O.Icel.
O.Bulg.
gostije,
avi-s).
Goth.
O.Icel. set)a infiri. 'to cause to sit, to set',
1. sing. Goth, satja
from *satiid,
'sit': cp. Skr. saddyol-mi
*sodSio,
causat. f r o m rt. sed-
'I m a k e to sit, set', Gr. tfofiko 'I m a k e
to flee, scare' f r o m *(pofist(o. b. Goth, steigan Germ.
OHG. stigan
1. sing. *stigd
O.Icel. stiga
from *stiv£0 : Gr. arn'/o),
'to climb', prim. *steigho.
Indg.
R e m . 2. The great antiquity of this i is confirmed by the fact, that the present: Goth, peiha OHG. dihu 'I extend', which in prim. Germ, had arisen from * pita/6, older *pen/o ( = Lith. tenku 'I reach out with something') (§ 214), gave occasion for the formation of the pret. form: Goth. pdih OHG. deh etc. (after the analogy of preterites from 6uch roots as steigh-). This was only possible, if the i in *siigo (goth. steiga) was already there. 3. W h e n the following syllable i(i)
or
I.
Prom
O.Sax. ¿si, A g s is. bharati)
*beredi,
Germ. *li%iana-n likkan
3. sing. *esti
contained a pre-Germanic son)
*isti : Goth.
F r o m 3. sing. *bhereti
thence *beridi:
( O . B u l g . bereti,
Goth, bairip,
'to lie* : O.Icel. liggja
: Gr. Xs^og 'bed' rt. legh-.
(.OHG. irdlri)
(—
O H G . birit.
O.Sax. liggian
Prim. Germ. Hrpmaz
OHG. Skr. Prim. OHG.
'earthen
beside *erpo 'earth' ( O H G . erda) : Gr. ep-a^e 'on
to the earth'. R e m . 3. "We have also here a proof of the great antiquity of the «. The verb *biSiana-n 'to beg' (Goth, bidjan O.Sax. biddian O.Icel. bidja), which belongs to Gr. nci6w 'I prevail upon, persuade' rt. bheidh-, would not have had a in the pret. of all Germ, dialects (3. sg. *bad(i) Goth. bap etc.), had not prim. Germ. *li%iana-n, *sitiana-n (rt. sed-) been pronounced with i.
58 4. Finally, e. g. 2. sing, imper. *fari (Goth, far) from *fare cp. Gr. ays. Such an -i, like Indg. i (see 3) had influence upon a preceding e; hence 2. pers. pi. *fariM (Goth. farip OHG. dial, ferit): Gr. aysrs; acc. *mi-ki (Goth. O.Icel. mik OHG. mih 'me'): Gr. ifts-yt, etc. Cp. Kogel in Paul u. Br. Beitr. V I I I 135 f. 5. In isolated cases, where no reason for the change has yet been discovered, especially in unaccented syllables: e. g. Nominal suffix prim. Germ, -iz- — Skr. -as- Gr. -so- Lat. -es(Goth. ag-iz-a- 'fear'). Cp. Paul's article in Paul - Braune's Beitr. Y I 85, Osthoff Zur Gesch. d. Perf. 419. The W e s t G e r m a n i c dialects in general stopped at this point. An e, preserved in uninterrupted continuity from the Indg. prim, language, is retained e. g. in OHG. beran 'to bear' = prim. Germ. *beratia-n (Gr. qsofiv) and reht 'right* = prim. Germ. *rexta-z (Gr. oye/.ro-g). R e m . 4. The change of e to i through the influence of a following u is questionable. E . g. in OHG. 1. pers. sg. biru, nimu, to beran, neman. situ m. 'custom', probably to Gr. eBo; 'custom*, sibun 'seven': Lat. septem. hiru$ 'stag* fr. prim. Germ. *xerut- (with anaptyctic u, s. § 6 2 8 ) : Gr. xrpn; 'horn'. See Paul in Paul-Braune's Beitr. Y I 78 f., Brate in Bezzenb. Beitr. X I 185, Braune Ahd. Gramm. p. 19. The i also of the root syllable in biris, birit etc. might only have been produced in "West Germanic through the following i.
The Norse dialects have also such an old e, e. g. O.Icel. bera
=
OHG.
beran.
On the other hand every e, which had remained in prim. Germ., became i in Goth. Then this i, through the influence of a special Gothic law, which also concerned i = Indg. i (§ 35), became e (at) again before r and h. niman 'to take', as opposed to OHG. neman O.Icel. nema — Gr. vepsiv. ga-vigan 'to stir, move', as opposed to OHG. wegan O.Icel. vega 'to move oneself' = Lat. vehere. fill (neut.) 'hide' (in pruts-fill 'leprosy'), but OHG. fel Ags. fell O.Icel fjall 'hide' from *fella*felna(§ 214) = Lat. pelli-s, Gr. 7IB\Aa; but balran from *biran — OHG. beran, raihts from *rihts = OHG. reht. The forms alrpeins = OHG. irdin 'earthen' and ga-raihtjan = O.Sax.
§ 67-68.
59
rihtian OHG. rihtan 'to judge' may serve as evidence to show that bairan and raihts had not an e which had been preserved in uninterrupted continuity through the influence of r and h. For it is scarcely possible to separate the at in these forms from that in airpa 'earth' and raihts 'right'. Indg. tautosyllabic eu appears in Goth, as iu. kiusa 'I choose, test, select': Gr. ysvoftai, cpf. *geus-o. ana-biuda 'I bid, command': Gr. ntidofiut, cpf. *bheudho. piuda 'people': Osc. tovtad 'civitate', O.Ir. tuath 'people', cpf. *teiitcl-. In High German it became, before a o vowels, in the oldest period eo, then io, in all other cases iu. deota diota 'people', but diutisc 'popularis' (Goth, piudisko adv. '¿Ovr/.w^'). reo$an rioym 'to weep' = Goth. *riutan: Skr. rodana-m 'weeping' (noun), rt. rend-. It is probable that the pronunciation eu still existed generally in prim. Germ, and that iu only arose at that time when there was an i in the following syllable, e. g. *piudiska-z (OHG. diutisc) beside *peiidd (OHG. deota). § 6 8 . B a l t i c - S l a v o n i c . Lith. vezii O.Bulg. vezq 'I drive': Gr. Pamphyl. f t / j t ) Lat. veho Goth, ga-viga Skr. vdhami, Indg. *ueghd 'I move, drive'. Lith. velkh O.Bulg. vlekq from *velkq (§ 281) 'I draw, drag': Gr. sXxio (cp. § 164 rem.), cpf. *uelqd. Lith. keturl O.Bulg. cetyrije 'four': Gr. ztrTuosg Umbr. petur- O.Ir. cethir Skr. catvdr-as. Nominal suffix -ter-, gen. pi. Lith. mo-ter-u O.Bulg. ma-ter-u 'matrum': cp. Gr. /.i^-Tsg-eg. Voc. sing, vilke O.Bulg. vluce 'wolf': Gr. Xvy.e, Lat. lupe. P e r sonal suffix of the 2. pers. pi. Lith. es-te esa-te O.Bulg. jes-te 'estis': cp. Gr. ¿are. In Lith. e is mostly pronounced very open and in some southern and eastern dialects it regularly became a, when initial, e. g. asu = esu 'I am' (§ 666, 2). Indg. hetero- and tautosyllabic eu apparently became ou, whence Bait, au (cp. § 84), already in the period of the BalticSlavonic prim, community. a. Heterosyllabic: Lith. av, O.Bulg. ov. Lith. tava-s 'tuus', sava-s 'suus': Gr. Tf(/)o-s, s(f)o-g. O.Bulg. novu new': Gr. vtfo-r. O.Bulg. plovq 'I flow, sail': Gr. nlt{f)w. O.Bulg. nom.
60
Indg;. e in B a i t - S l a y .
pi. synove "sons' from *suneu-es: cp. Gr. Jf(/)-ic, Goth, sunjus sons' from *sunius prim. Germ. *suneu-ez (§ 179), Skr. sundv-as. — Lith. devyni Lett, dewini 'nine', Pruss. newlnts 'the ninth', 0.Bulg. deveti nine' are to be explained as being formed from analogy with Lith. deszimt O.Bulg. desrfi 'ten'; we may observe also that the initial d- instead of n- cannot well be explained otherwise than as being borrowed from the word for ten. R e m . 1. It is doubtful whether the d- of the Lith. and Lett, word for 'nine' was due to Slavonic influence and neu- was the right form in prim. Baltic, which was retained in Pruss., or whether already in prim. Baltic-Slavonic de%t- took the place of nou- and the n- of the Pruss. form was borrowed from Germ. (cp. Bopp. Vergleich. Gramm. I 3 76, F i c k "Worterb. I I 3 588. 744, Bruckner Litu-Slav. Studien I 80, Osthoff Morph. Unt, I 93).
b. Tautosyllabic: Lith. aw, O.Bulg. u (fr. ou § 185). Inf. Lith. pldu-ti 'to rinse' O.Bulg. plv-ti (1. sing. pres. plovq), supine Lith. pldu-tn O.Bulg. plu-tu: cp. Gr. fnXtv-na, Skr. inf. plotu-m (3. sing. pres. mid. plav-a-te 'he swims'). Lith. raudmi (an archaic pres. form, now raudoju) 'I lament, complain': Skr. rodi-mi 'I lament, weep', Indg. *reud-mi or *reudd-mi. Lith. raugiii 'I belch':
cp. Gr. tQivyo(.iai.
new'.
O.Bulg. inf.
srotu-m
'to hear',
y.Xtfog
Skr.
irdvas
sluti
L i t h . nauje-s
'new':
'to be called', supine
to O.Bulg. slovo
Goth,
sliitu:
(stem sloves-)
niuji-s
Skr. inf. 'word':
Gr.
(neut.) 'renown'.
Heterosyllabic Indg. prim. Baltic-Slav, ei became ii in prim. Slav, and this i was, in the period of Old Church Slavonic, lengthened to i (cp. § 36 page 37). O.Bulg. vijq vijq: Lith. vejii 'I turn, wind', Skr. v&ycimi 'I weave', Indg. *ueid. Nom. pi. go stye, gostije 'guests' (nom. sing, gosti) from *gostei-e(s): cp. Skr. avay-as 'oves'. trije trije 'three': Skr. tray-as, Indg. *trei-es. Indg. tautosyllabic ei appears in Lith. as ei and e, in Slavonic as i. The course of development of the latter was *ei, *ii (*t?), 1, cp. the change of heterosyllabic ei. Lith. 1. sing, eimi 'I go', 3. sing, eiti (Pruss. eit): Gr. tua shi; supine eitu O.Bulg. itu: Skr. inf. etum from Indg. *ei-tu-m, rt. ei-. Lith. szlei-va-s szlei-vi-s 'having crooked legs' (Lett, slei-ja 'stripe', slei-ju 'I lean against') : Gr. y.Xsi-Tv-g, rt. Tclei-. Lith. veida-s 'face' O.Bulg. vidu
§ 68-69.
61
Indg. e in Balt.-Slav. Prim. Indg. e.
'looks' (noun), Lith. veizdeiu 'I see' (for the £ cp. § 547 rem.), O.Bulg. vizdq 'I look' from *uidiq (§ 147): Gr. fsidofiui, fetäog. Lith. geidziu, 'I long for', Pruss 3. pi. geide 'they wait', O.Bulg. ~idq 'I wait'. Lith. deivi/-s'ido\' deive (fem.) 'ghost' beside d£va-s(hett. diws)'god', Pruss. O.Bulg. deiwa-s 'god': Lat. deivos dlvus. Lith. ze-ma (Lett.J'hna) zi-nia 'winter': Gr. yj-T-f.a>., /si-^auy. Lith. de-na (Lett, din a) 'day', Pruss. dei-na-n acc. sing, 'day' from a stem *dei-n(*dei-en-) beside *di-n- (O.Bulg. gen. di-n-e 'day's'). Lith. snSga (Lett. snig) 'it snows': Gr. vut^u, rt. Sneigh-, Lith. lekü 'I leave' (Lett. Uhu 'I lay'): Gr. Xeinw, rt. leiq-. O.Bulg. pisq 'I write' from "pis-iq (§ 147): Gr. nsixio 'I scratch' rt. peiTc-. R e m . 2. The circumstances, under which ei at one time remained (ei) in Lith., and at another time became e, have not been ascertained (cp. Mahlow Die langen Yocale A, E, O, page 143 f. and Osthofif Morph. Unt. IV 112). We are inclined to think that only circumflexed ei, e. g. in eiti 'to go' (cp. § 691), regularly became c when the following consonant had not a palatal timbre, caused by an e- or i- vowel in the next following syllable (cp. the double forms e and ia in Irish § 66); hence devas beside
deivys deivi, eiml beside lekü. The supine eitu instead of *etu would have been formed after the analogy of eiti\ einu etna "I go' 'he goes' (instead of *ento *ena) with ei-, because they were not formed until after this sound-law had ceased to operate. Such verbs as leiiu leszti 'to lick' = O.Bui, liza (from Hiz-ia) beside such as geidziü geisti (Leskien Der Ablaut in Wurzelsilben im Litauischen p. 144 f.) present difficulties. For it is certainly hardly probable that the few forms as sup. lesztu etc. could have extended their e to the whole verb. Cp. the change a i : e § 84 rem. 2.
en, em before explosives and spirants became e in Slavonic e. g. svetü 'holy', zqbq 'dilacero, frigeo'.
S. § 219.
I n d g . e. § 69.
I n d g . dhe- 'set, lay': Skr. fut. dhü-syämi
(neut.) 'position', Gr. fut. 6/j-oco, Goth, ga-deps 'deed, position', Lith. fut. de-siu inf. de-ti. im-ple-tu-s,
inf. de-ti, O.Bulg. aor.
Indg. pie- 'fill': Skr. part, prä-td-,
full' from *ple-io-s, Gr. nlrj-o^r
(st.
dhdnaga-dedi-) dö-chü
Arm. Ii (gen. Hoy)
'full' aor. ^nlrj-oa,
Lat.
ple-nu-s
O.Ir. lin 'numerus, pars', O.Icel. fleire fleiri 'more'
(from st. *fle-iz-en*, cp. § 614). Optative suffix Indg. -iie- -ie-: 2. pers. sing. Skr. s-iyd-s
s-ya-s,
62
Prim. Indg. e.
Indg. e in Ar. Armen.
§ 69—71.
Gr. s'irjg from *ea-trj-g, Lat. s-ie-s from rt. es- 'be'. -te(r) nom. sing, ending of -fer-stems: Skr. mcL-ta Arm. mair from *ma-ter Gr. Dor. /.id-ttjo 'mother', Lith. mo-te and mo-te 'wife, married woman', 0 . Bulg. mati from *mate 'mother', -e Indg. ending of the instr. sing, of o-stems: Skr. uccd 'above', Gr. Lacon. nrf-noxa 'usquam' or 'unquam', Goth, pe to pa- 'the', live to hva- 'who' 1 ). Perfect stem Indg. ed- from rt. ed- 'eat': Skr. part, ad-iGr. iS-rjd-wg, Lat. ed-imus, Goth, et-um, Lith. part, ed-fs, 0.Bulg. part, jad-u. Imperfect stem Indg. es- from es- 'be': Skr. 1. sing, as-am 3. sing, as, Gr. 1. sing. Horn. r\a 3. sing. Dor. Indg. 1. sing. 3. sing. *es-t. Cp. § 114. The tautosyllabic diphthongs ei and eu probably did not frequently occur in the prim, language. E . g. Imperfect stem Indg. ei- from ei- 'go': 2. sing. Skr. ais, Indg. *ei-s. *pleisto'most': Gr. nXsioTo-g O.Icel. fiestr (§§ 116, 611, 614). Nom. sing. Indg. *di£u-s 'sky' : skr. dycius, Gr. Zfvg from Ztjug (§ 611).
-van,
§ 70. A r y a n . 3. sing. pret. Skr. d-dha-t Avest. da-p O.Pers. a-dcL from rt. dhe- 'put'. Skr. va-ti Avest. vaiti 'it blows' : Gr. a(f)ij-ai 'it blows', OHG. wcLan 'to blow', Lith. ve-je-s 'wind', O.Bulg. vi-je-ft 'it blows', rt. ue-. The change from e to d took place in the prim. Aryan period, after the vowel (like Indg. e) had palatalised the preceding 2-sounds, e. g. Skr. -jani-s 'wife, spouse' : Goth, qens (st. qeni-) 'wife', Indg. *qeni- (§ 445). For Avest. e, o, A from prim. Iran. prim. Aryan a — Indg. c see § 102. § 71. A r m e n i a n , mi 'that not': Skr. ma, Gr. Dor. (Gen. mti) 'sense' : Gr. u?jaXo-g. In like manner ursu-s fr. *or(c)so-s, mulctu-s fr. *molcto-s, whose or, ol arose from Indg. f , § 295. Final -o, like -a (§ 97), became e. Imper. sequere: Gr. s7is(a)o. Nom. olle (ille seems to be a new form after iste, ipse) fr. *ol-se (§ 571), iste a new form for *is-se (with t transferred from 8uch forms as istum isto): the second component of these compounds ia identical with Skr. sa, Gr. o (Indg. ^so), cp. the accidence. Alao in other caaea o in unaccented syllables seems occasionally to have become e, i, that is, an irrational vowel, which appears after i and in cloaed syllables aa e, and in other cases aa i. ilico fr. *in sloco. hospes hospitis fr. *hos(ti)-pit- (§ 633) beaide potis sum (com-pos -potis). anni-culus to anno-, auri-fex to auro-, socie-tils to socio-, novi-tcis (Gr. vso-tys) to novo-, legi-mus == Gr. leyn-utv. I t , of course, still remains to be determined in what relatione this change stands to the u in filius etc. and for what reasons here the one, there the other course of variation was introduced; cp. § 680. The question might also be raised as to whether Indg. e does not occur in anni-culus, auri-fex, socie-tols, novi-tds and in legi-mus: anniculus etc. like oiX£-r%', legi-mus, new formation after legi-tis = Gr.
ktys-Ti.
R e m . 1. The genitive suffix -es -is (salut-es nalut-is) has not arisen from -os -us (senatu-os, homin-us). -es and -os were both old inherited forms (see p. 67).
Thurneyaen Kuhn'a Ztschr. X X V I I I p. 154 ff. assumes the transition of ov to av, e. g. caveo: Gr. xo(/)ia> 'I mark'; cavo-s: Gr. xo'oi excavations'; lavo: Gr. Aos 'he washed'. Cp. however, Hiibschmann Daa Idg. Vocalay8tem p. 190. The fluctuation of the orthography between o and u in the Umbrian-Samnitic dialects pointa to a close pronunciation of the o. Since the Umbr. u was the sign both for o and u (§ 23), the tables written in the national alphabet contribute nothing towards determining the quality of the o-sound; the tables written
74
Indg. o in Italic.
in L a t . h a v e both o a n d u.
I n O s c a n u u, o u.
porca
' p o r c a s ' : L a t . porcu-s,
Grr. nogtw-c,
farah
(§ 2 7 7 ) , L i t h . parsza-s,
Umbr. uvef neut.
ovi
'flock of sheep', L i t h .
p r i m . st. *kor-en-
avi-s,
G r . nóm-g.
ovi-ca,
Goth,
arèpi
cpf.
*oiii-s
Gr. xooolvy 'crow',
U m b r . p u r - e 'quod', Ose. p u d G r . 770-, Goth, hva-,
N o m i n a l suffix -0- in U m b r . p u p l u - m h u r t u - m ' h o r t u m , t e m p l u m ' , dolo-m The
G r . ot-g,
O.Bulg.
farli
sucking-pig'.
Ose. p u t i a d p u t i i a d 'possit' p u t i a n s
' q u o d ' : L a t . F a l i s c a n quo-d,
' p i u m ' : L a t . dolo-m
OHG.
'pig,
'cornicem' : L a t . comix
*kor-n-.
' p o s s i n t ' : L a t . potis, pod
O . I r . ore,
cpf. *porìco-s
acc. 'oves': L a t . ovi-s,
'sheep'. U m b r . curnac-o
Umbr. p u r k a
dolu-m,
poplom
dolu-m
'dolum', Yolsc.
Falisc. vootu-m
d i p h t h o n g s oi a n d
Indg.
*qo-.
'populum',
Ose. piho-m
'votum'.
oil.
T a u t o s y l l a b i c oi in chief a c c e n t e d s t e m s y l l a b l e s b e c a m e oe, u in L a t .
oino-s
moini-cipio-m mama-s
oeno-s tìno-s : G r . oìvó-g,
moenia
munia:
Goth,
'exchange', G r . Sicel. fiol-ro-g
oe, 1 arose in final syllables. ploirume
(C. I. L . I H2), plurimi:
•oi, *toi f r . s t e m to- 'the, this'. -is
(equis)
f r . *-ois
(§ 612).
syllabic nom. pi. hi qui,
cpf. *oi-no-s
ga-mains
'one'.
'common',
Lith.
'requital'.
N o m . pi. poploe
(Saliar h y m n ) ,
G r . rot, lv-/.ot, G o t h .pai, Cp. i n s t r . pi. *-ois,
Indg.
-oes
(oloes)
I n like m a n n e r also 1 in t h e m o n o -
instr. pi. his quis.
qui a n d quis
h a v e o r i g i n a t e d f r o m a n u n a c c e n t e d position (si qui,
may
dli-qul
etc.)
a n d hi a n d his m a y be d u e p a r t l y to assimilation to qui,
quis,
and
illis.
partly
to assimilation
to t h e p o l y s y l l a b i c
C p . also t h e n o m . sg. qui f r . O . L a t . quoi Dvenos-inscrp., i. e. *quo
+
cp. B e r s u
particle i
f o r m s illi,
(qoi =
quoi
on
the
D i e G u t t u r a l e n , B e r l i n 1 8 8 5 , p. 40)
(l).
Rem. 2. oi is also said to have become t in chief accented syllables, when preceded by v or l: vìdit oì?*, vi nit-m oivoc, vicu-s olxo-.-, Ivpiit Xoinr, liber loebertatem and others. This sound-change cannot, however, be accepted as definitely proved, since almost all examples under this category can also be easily explained otherwise. See Osthoff Morph. Unt. IV p. 129, Zur Gesch. des Perf. p. 191, Danielsson in Pauli's Altit. Stud. 4. 156 f. Rem. 3. "Why feres, which I (against Thurneysen Bezzenberger's Beitr. Y i n 269) consider equal to Gr. yrpois Goth, bairdis, with e and not lì It is possible that è arose regularly in the 1. sg. (O.Lat. ferem)
75
Indg. o in Italic.
and 3. pi., and then from these two forms was extended to the other persons. I leave it undecided as to whether -em, -ent were the continuation of prim. Lat. *-oi-in, *oi-nt (cp. Gr. 1. sg. Att. TQitpoiv, 3. pi. Delph. Trag-ezoiY, Goth. 3. pi. bairdin-a), or that of Indg. *-oim, *-oi-iit, which, according to Lat. sound-laws (§ 237), must, first of all, have become *-o(i)em, *-o(i)ent. It is also conceivable that capies and nudies are regularly formed, but that *feris arose with the same contrast as that in which socie-tas and novi-tas stand to each other. In that case *ferls has been transformed to feres after the analogy of capiex; but then merfii, mediis (utaoi i/tami) must also be regarded as modifications of *medie, *medies after nnvl, novis. In Umbrian u = oi whereas the O.scan-Sabell. monuments still present the diphthong, as to which it is certainly questionable, whether the written diphthong still existed everywhere as such in the language itself. 'unita'.
U m b r . u n u 'unitin,
Osc. u i n [ i t u
Osc. m u i n i k a d abl. 'communi': O.Lat. comoinis.
Umbr.
k u r a i a 'curet', Pelignian coisatena 'curaverunt': L a t . coiro coero euro (the forms coraveront
C. I. L . I 73 and couraverunt
1419 are very singular; in the latter prim. Ital. stem. fern. *koisa-. Nuvlanuis
'Nolanis'
zicolois
u was certainly
Cp. also the 'diebus',
1. c.
meant),
Osc. instr.
Pelignian
suois
pi.
'suis',
U m b r . v e s k l e s - i s 'vasculis': L a t . oloes Hits; Umbr., like Lat., indicates here a special manner of treating oi in final syllables. Tautosyllabic Indg. ou, fell together with Indg. eu (§ 65) and is still only traceable in Osc. as d i p h t h o n g : Gen. sg.
castrovs
f r . stem castru-
'tribus';
rofu
'rubos': Goth, rdups,
gen. sg. fructus; giutan
'fundus' (§ 77).
cpf. *rouaho-s.
pf. fudit:
'to pour out').
ro/o-, gen. domos =
In Umbr. o: gen. trifor
noundinum
domus
(Augustus wrote domos);
(beside poublicom
nundinum).
'poured out'
(pres.
B y side of it also o: robu-s — U m b r .
the inscrip. forms with (long) o, losna § 570), poplicod
u arose in L a t . , e. g.
cp. Goth, gdut
(=
publicus),
luna
fr.
nondinum
cp. also *louc-sna, (beside
I t remains doubtful how this duality u
and o in L a t . is to be estimated.
I t was hardly a mere graphic
duality. Rem. 4. In a number of Ital. words containing it cannot be decided, whether this diphthong, represents prim. Indg. ey, (§ 65) or ojt, e. g. in those containing loyk- as Osc. gen. L l i v k a n a t e i s 'Lucanatis' L u v k i s L u v i k i s 'Lucius', Lat. Loucina Luucius etc., fr. rt. leyk- 'illumine' (Gr. Aevxdi Xovooor).
76
Ind^. n in O.Irish.
§ 82. O l d I r i s h . In chief accented syllables o. ocht eight* : Gr. oxrai. roth 'wheel* : Lat. rota 'wheel', OHG. rad 'wheel', Lith. rata-s 'wheel', ore 'pig' : Gr. nogxo-g. gorim 'I heat, warm, burn' : O.Bulg. goriti 'to burn', Lith. gara-s 'steam', fr. rt. gher-. Prep, eo n- 'with' : Lat. cum, Umbr. Osc. com. I t passes into u with the palatalisation of the following cons o n a n t : suide 'seat' f r . *sodio-
— L a t . soliu-m
(I — d,
§ 369),
rt. sed- 'sit' (§ 640). Cp. also mug 'slave' fr. *mogu-s (Goth. magu-s 'boy, servant'), ru- beside ro- (Indg. *pro 'before, in front', p. 67), cum- beside com- con- (co n-, see above). In secondary accented syllables o, so far as it did not entirely disappear (§§ 634. 657), became an irrational vowel, whose sound quality mostly regulated itself according to the timbre of the following consonant. a.
1. pi. as-biram
'dicimus',
3. pi. as-berat
: Gr. q>saofisv,
(Doric) (ptQovri, tffiQov, Goth, bairam, bairand; -berat may be middle = Gr. a dialect had also e for e and i before m n v, especially in final syllables, emavantVigorous': late Ay. amavantSkr. amavant-. Acc. kehrp-em: late Ay. kehrp-em 'figure, body', Skr. kfp-am. yem beside yirn: Skr. ya-m. xsqnmen-e, dat. fr. xsqn-ntan- 'grief, affliction*, cp. Skr. vdrtman-e. e-visti- 'want of acquaintance' beside late Av. e-vista(see aboye).
2. a became e after y before i, i, e, y, nh (fr. Ar. si, § 125), c or J in the following syllable. The orthography is also here inconsistent, since a appears also beside e. 3. sg. act. haray$iti, 3. pi. act. kclrayeinti, 3. sg. mid. kdrayete of k&rayemi, Causat. fr. kar- 'make', ye.snya- 'adorable' beside yama- 'offering, price': Skr. yajnd- Gr. ayvri-g. Gap. gen. masc. yehycL 'cuius' = Skr. yasya, late Av. yenhd (fem.) 'cuius': Skr. yasycis. ipyejah- 'need, danger': Skr. tyajas-. Cp. e fr. ct § 102. 3. a became o before u, il or v in the following syllable, especially after labial sounds, and before r + consonant. The spelling with o is also in this case not consistent; since beside it a also occurs, pouru-s (fr. *poru-s, § 638) 'multus' beside nom. pi. parav-d: Skr. puru-s, purdv-as (§ 290). vohu 'bonum': Skr. vdsu. mosu 'soon, quick': Skr. maksu. Gap. ver'zyotu 3. sg. imper. of ver'zyemi 'I work', as-bourva- 'having much food': cp. Skr. Yed. su-bharva- 'much consuming'. coreJ> 3. sg. aor. 'made.: cp. Skr. akar fr. *a-kart, prim. Ar. *a-cart. Cp. o fr. a § 102. § 95.
Armenian.
aseXn (gen. aslan)
'needle': Gr. ay.go-q
90
§ 95-96.
'point', Lat. acu-s 'needle', Lith. asz-t-rii-s O.Bulg. os-t-ru 'sharp'. argel 'hinderance': Gr. apxf'w 'I ward or keep off', Lat. arced. taigr,
gen. taiger
(g f r . u § 1 6 2 ) : Skr. devdr-,
(§ 96), L a t . levir (§§ 65. 368), Ags. tacor
Gr. ¿atjo
O H G . zeihhur
(with
remarkable guttural), Lith. dever-l-s O.Bulg. dever-t, Indg. prim. f. *daiuer- *daiu%- 'husband's brother, brother-in-law' (§ 155 rem. . § 98.
G r e e k iSaxpv: L a t . dacruma
Goth, tagr 'tear',
lacruma
O.Bret,
dacr
xaxy.o-g (Hesych.) 'little finger', xaxn-g 'bad'
c o m p a r . xaximv: A v . kasn- ' s m a l l ' c o m p a r . kasya.
tuay.oo-c;
'long':
AY. mascth- 'greatness' O.Pers. mapista- 'uppermost', Lat. macer, OHG. magar 'lean'. ayxro I press tight', uy/i 'near': Skr. dhas'tightness, oppression', Av. qzah- 'tightness, anxiety', Lat. ango angor, Goth, aggv-u-s 'close, tight', Lith. ankszta-s 'close, tight', O.Bulg. qzu-ku 'tight, close', rt. angh- 'straiten. Tautosyllabic 'I dry, make dry fr. *aavaw (§ 564), Ags. sedr 'withered, dry' OHG. sorcn 'to parch' O.Icel. saurr mud', fr. rt. sdus-. Lith. augmü (gen. augmeñs) growth, ex-
96
Prim. Indg. a.
§ 100—102.
crescence': Lat. augmen-tu-m. O.Bulg. u- 'off, away' (u-myti 'to wash off', u-dati 'to give away, ixdovvai OvyaTtga), Prues. au- 'off, away' (au-mu-sna-n acc. 'ablution'): Lat. au-, au-ferre. I n d g . a. § 101. I n d g . *bhaghu-s 'elbow': Skr. bahu-s 'arm', Gr. Dor. naxv-g 'fore-arm, elbow', OHG. buog O.Icel. bogr 'elbow' Indg. *md-ter- 'mother': Skr. matdr-, Arm. mair, Gr. Dor. /xCtrr/o, Lat. mater, O.Ir. mathir, OHG. muoter O.Icel. moder modir, Lith. mote ('wife'), O.Bulg. mati. Indg. fern, suffixes -d-, -tcL-, -id- etc.: Skr. aSvd 'mare', sd 'the', Gr. xa'jga 'land', Dor. a 'the', Lat. dat. abl. pi. dextrd-bus, O.Ir. dat. instr. pi. mndib fr. st. mncL- fr. *bna- (nom. sg. ben 'woman, s. § 520) = Skr. gna- 'woman', Goth. dat. pi. gibo-m fr. st. gibo-, nom. sg. giba 'gift', so 'the' f., Lith. instr. pi. ranko-mis fr. st. ranko-, nom. sg. ranka 'hand', O.Bulg. instr. pi. rqka-mi fr. stem rqka-, nom. sg. rqka 'hand'. Indg. conjunctive suffixes -a-, -id-, etc., stem *bhera- from bluer- 'bear': Skr. 2. pi. bharatha, Lat. ferO-tis, O.Ir. 1. pi. do-bera-m, O.Bulg. 1. sg. berq fr. *bherd-m (cp. Brugmann Morph. Unt. I p. 145, I I I p. 30 f., Kuhn's Ztschr. X X V I I p. 419) 1 ). Indg. pres. suffix form -nd-: Skr. mi-nd-mi fr. rt. me\- 'lessen', Gr. Horn, tfd/.i-v/]-/ni. -tdm Indg. personal ending of the 3. du. act.: Skr. abhara-tdm, Gr. Dor. ¿cfiSQe-rav, Indg. *Sbhere-tdm fr. rt. bher- 'bear'. The tautosyllabic diphthongs di and du were not, as it seems, frequent in prim. Indg. -di ending of the dat. sg. of d-stems: Skr. su(v)-apatydi (su(v)-apatya- 'having a beautiful posterity'), Gr. xulpa, O.Lat. Mdtutd 'Matutae', Goth, gibai 'to a gift'. Skr. ndus 'ship', Gr. vafg fr. *vaus (§ 611). § 102. A r y a n . Skr. sthana- n., Av. O.Pers. stdna- n. 'standing, standing-place, place of abode' : Gr. Svotijvo-t; fr. *i f1'- /c-Xtno-g. The prim. Indg. duality i and ii was preserved e. g. in ¡.itiiooc, /Litooq 'medius' (fr. *u(0-i;o-g = Skr. mddhya-s) compared with nuroiog 'patrius' (— Skr. pitriya-s), in ;oj 'I do, work' (fr. *foey-i.a), cp. § 120) compared with id-ao 'I sweat'. On the other hand the Lesb. forms tursygo:, aXXoreooog = Att. /.ihoiog 'moderate', uXXotQiog 'alienus' etc. show a sound-change which took place within Greek itself: ¡.itrgmg became *metfios, *meterios, thence /ufTtgpog (§ 626). It seems to me more probable that the f in Lesb. xa'of«, Zowvoog, Cypr. y.ooLa and others, beside did, xuodia, zliwvvaog (s. Meister Griech. Dial. I p. 127 if.) is due to a Gr. change of i to i than to regard it as an old inherited double form as is the case with /usoog and nargiog.
120
Indg. i in Greek and. Italic.
§ 131-133.
The Cypr. orthography Il«(pif.a, ¡(.ax^oav, Pamphyl. net'x'q. Hence specially in O.Bulg. through anticipation of the spirant *mey'd'y'a, *mej(t/q, further *mez'd'z'd, *mes't's'q — *mezd'za,
*meit'&q
—
meld'a,
meU'q.
The O.Bulg. forms sii-mostrja, u-mristvljq, biizdrje were also prepared for already in prim. Slav, in so far as i had softened t h e w h o l e c o m b i n a t i o n s tr, 1v, dr. But it must here be observed that the s and z depending on the anticipation of the palatalism are not always written in the monuments, e. g. su-motr{j)enije beside sii-mostr(j)enije n. 'consideration* (Miklosich Yergl. Gramm. 1 2 220). We do not decide as to how this fluctuating orthography is to be judged. That the O.Bulg. combinations id and si were spoken soft (dz\ &') follows from such spellings as nadezde (e sign for ja) beside nadezda 'hope', sqstju beside sqstu 'ovr' dat. sg. of the part. sy. W met with in most monuments. Cp. rem. 4. F r o m prim. Slav, lei, gi ( m o s t l y =
Indg. qi, qi, ghi, § 4 6 1 )
arose in prim. Slav, c (i. e. ts), dz, the latter w a s w e a k e n e d to z
in O . B u l g .
fr. luzq
*lugn
pri-tuca
: OHG.
'mentior
fr.
lukki *lugiq,
'comparison' fr.
*-tukia.
lust
'mendacious'
'mendacious' (prim. Germ. st. inf.
lugati.
A n a l o g o u s l y ski b e c a m e sc i. e. sts, h e n c e b y 1) i, I are the softened z, s. elsewhere.
*lu%ia-);
assimilation
Cp. Ar. i and $ §§ 20. 21. 396 and
134
Indg. i in Balt.-Slavonic.
§ 147.
of the sibilants èts, further historic st (more correctly ¿f, cp. rem. 1). istq istesi etc. fr. *îskiq, inf. iskati 'to seek*. Bern. 2. Cp. vusteti 'to begin' fr. *vùscetï = *vûz + cetl. Further on account of the O.Bulg. change of sts to st, the loo. sg. clovècïstè beside
clovècïscè from nom. clovècïskù 'human* and inf. istèlili beside is-ciliti 'to heal* (st fr. sts) are instructive.
Beside this treatment of prim. Slav, ki, gi there is also a change of them to c (i. e. ts) and dz, the latter was weakened to ^ in O.Bulg. in the beginning of the literary period (Miklosich Yergl. Gramm. I 2 251 if.). Nom. sg. in -ïcï m., -ice n., -tea f., e. g. junïcï 'young bullock' fr. *iûnïkiï, srïdïce 'heart' fr. *srïdïkie, ovïca 'ovis' fr. *ovikia. Neut. lice (gen. lied) 'face' fr. *likie. Fem. stiza 'way* fr. *stïgia. The treatment here in question took place at a later period of the prim. Slav, language than the one above, viz.: at the same time when the instr. sg. Hakëmï (O.Bulg. talcu 'talis') became tacëmï (§§ 84. 462). Therefore a *iûrilkn was only formed to *iûntku and an *ovïkjfl to *ovïka after the law, whereby ki became c, had already ceased to operate. The voc. jumce still belongs to *iûnïku (cp. vluce : vluku), so too Jumcî 'taurinus', ovicï 'ovinus' are also directly to be referred to Hûnikû, *ovika1). Prim. Slav, si, zi == Indg. hi, gi, g hi (§ 412) became already in prim. Slav, s, z. O.Bulg. pres. pisq pisesi etc. (inf. pïsati 'to write') fr. *plsiq *pïsiesï, rt. peiîc- 'to cut, cut straight'. zujq and *Mvq 'I chew' fr. *ziû-iq and *znvq *ziuvq (§ 52), the latter = OHG. chiuwu from rt. giaxu-, lièq liêesi etc. (inf. lizati 'to lick') = Lith. lëêiù 'I lick', prim. f. *leigh-iô. i with a preceding Indg. s became s in prim. Slav, siti 'to sew' fr. *siïtï, *siytî = Lith. siûti 'to sew' (§ 60). Adj. nasi 'noster', to gen. nasu 'nostri' (fr. *ncls-su). Part. neut. bysqste-je 'TÔ fiéXXov, to an obsolete fut. indie. *by-sq : Lith. bû-siu 'I shall be'. Part. pf. gen. sg. masc. nesusa (indie, pres. nesq 'I carry'): Lith. nêsz-us-io, cp. Goth. nom. pl. bêr-us-jôs 'parents' ('those who have brought forth). 1) No direct historic connexion, consequently, exists between jumci and Lith. jaunlkis, gen. jaunîkio 'bride-groom* (properly 'juvenculus') in spite of their similar suffix formation.
§ 147-149.
Indg. i in Balt.-Slavonic.
135
R e m . 3. In those cases where Indg. s had passed into ch in prim. Slav., s is not to be traced back directly to si, but first of all to chi (cp. snusinit 'like a daughter-in-law' adj. fr. snuchinu from snucha 'nurus' : Skr. snum, Indg. nusci). Cp. § 588, 2. 3.
These affections of prim. Slav, s ( = Indg. fc, s) and 2 ( = Indg. f), gh), caused by /, took place also when an I or n stood between them. O.Bulg. mysljq : inf. mysliti 'to think', blaznjq : inf. blazniti 'to lead astray'. Here i first palatalised n and thus extended its influence to s, z, but disappeared in the softened I, n, according to the remarks made above under ri, li, ni. Cp. ostrjq (inf. ostriti) fr. *ostrja above, rem. 1. mysljq : pisq, nesusa — ostrjq : tlusta. R e m . 4. That O.Bulg. c, z — orig a-m -ro-g, Lith. ketvirta-s). Cp. also prim. Ar. *atharun-
('fire priest') for *atharun-
(Av. dat. sg.
aparun-e) beside *atharua- —•- *atliarun- (Skr. dat. abl. pi. dtharva-bhyas), Skr. instr. dyim-a ('life') beside Gr. ai(f)bv ai(f)wv etc.
(The author Morph. Unt. I I 189 ff).
Rem. I, therefore, conjecture that Skr. instr. sg. devr-d (fr. *daiuer'husband's brother') does not represent an Indg. *daiy,r-, but a special Sanskrit new formation *dainer-, *daiyf- and *daiur- may have existed side by side of each other in Prim. Indg. The spondaic form rWfiuv, llias SI 769, can equally well be read cWytov (fr. *dat,vqiov) as SuiFqmv, conjectured by Ebel.
§ 156. F i n a l l y ,
u only occurred as an absolute final in the
second component of diphthongs.
E . g. voc. Skr. suno
Lith.
sunau O.Bulg. synu '0 son, Indg. *suneu or *sunou-, loc. Skr. sunWti O.Bulg. synu *sunou
(cp. § 85).
'in filio', Umhr. m a n u v - e 'in manu', Indg. Its treatment in the individual
languages
was almost entirely the same as that of anteconsonantal diphthongs.
u, when conditionally final, also stood after consonants,
perhaps *medhu esti == Skr. madhv asti 'mel est'. Cp. § 645, 2. Aryan.
§ 157.
Initially.
'we* : Goth. veis. vis-% 'vicus'.
Skr. vayarn Av. va$m O.Pers. vayam
Skr. wis- Av. vis- O.Pers. vip- ' c l a n : O.Bulg.
Skr. Av. O.Pers. va 'or': L a t . -ve.
§ 157—159.
Indg. y, in Aryan.
141
Initial u was lost before u and u in Skr. uru-5 'broad' fr. *uur-u-s, prim. f. *u^r-u-s (§ 290). urna 'wool' fr. *uurnd, Indg. *ul-na (§§ 151. 306). ur- was transposed in Av. with w-prothesis. "rvctta- n. 'decree, command': Gr. fpyro-g 'decreed'. Part. pres. mid. nrv&zemna'powerful' (z —