The Origin of Latin Haud and Greek ou; And the Extensions of the Originally Unextended 9781463221959

Horton-Smith offers an explanation for the negatives haud and ou in Latin and Greek respectively, suggesting a root word

181 107 2MB

English Pages 27 [31] Year 2009

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

The Origin of Latin Haud and Greek ou; And the Extensions of the Originally Unextended
 9781463221959

Citation preview

T h e Origin of Latin H a u d and Greek ou; And the Extensions of the Originally Unextended

A n a l e c t a Gorgiana

ÒSI Series Editor George Anton Kiraz

Analecta Gorgiana is a collection of long essays and

short

monographs which are consistently cited by modern scholars but previously difficult to find because of their original appearance in obscure publications. Carefully selected by a team of scholars based on their relevance to modern scholarship, these essays can now be fully utili2ed by scholars and proudly owned by libraries.

The Origin of Latin Haud and Greek ou; And the Extensions of the Originally Unextended

Lionel Horton-Smith

l gorgias press 2009

Gorgias Press LLC, 180 Centennial Ave., Piscataway, NJ, 08854, USA www.gorgiaspress.com Copyright © 2009 by Gorgias Press LLC Originally published in All rights reserved under International and Pan-American Copyright Conventions. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise without the prior written permission of Gorgias Press LLC. 2009

1

ISBN 978-1-60724-611-4

ISSN 1935-6854

Extract from The ^American Journal of Philology 18 (1897)

Printed in the LTnited States of America

LA TIN HA UD AND

GREEK OY.

69

II. Idg. *gho- *ghi-: Lat. ho- in hie (from *ko-i-ce) from the Idg. demonstrative pronom. stem *gko-' this.' 1 Skr. hi (and -hiy~\ Nom. acc. sing neut. of the Idg. demonstr. Avest. zl (and zl)1 ! pronom. stem *ghi- 'this' (whence in A r m . zi2 f Skr. Avest. and Arm. the meaning G k . -xi in ov-\l 3 J 'for'). G k . -x in ov-x, abridged from - x i in oi-^i.4 III. Idg. *qo- *qi-: G k . xe in oS-re from the uninflected Idg. *qe ' h o w ' (indefinite ' s o m e h o w ' and ' a s also' = ' a n d ' ) of the Idg. interrogative and indefinite pronominal stem *qo-, and identical with Skr. and Avest. ca, Lat. -que? G k . TI in oi!-n, nom. acc. sing. neut. of the Idg. interrog. and indef. pronom. stem *qi-, whence come also Skr. nd-ki-s (for *nd-ci-s) ' n o one,' Lat. quid, etc. 5 A r m . c in o-c: either (i) from Idg. *qe (: G k . re in ou-Te), meaning 'some-how,' so that o-c — 'no-how, in no wise.' 5 or (ii) from Idg. *qi (: G k . n in otf-xi) 'somewhat,' so that o-c = 'not at all.' 5 LIONEL

HORTON-SMITH.

53 Queen's Gardens, L a n c a s t e r G a t e , L o n d o n , W . , and S t . J o h n ' s C o l l e g e , C a m b r i d g e ; E n g l a n d . 2 Supra, §9, pp. 65, 66. Supra, §8 (a) (1), s. v. ov-x'i, p. 62. §8 (a) (1), s. v. ob-xi, pp. 61, 62, and (d), p. 64. 4 Supra, §8 (a) (2) ad fin., p. 64. 5 Supra, § 1 0 ad fin., p. 68. 1

3 Supra,

I V . — T H E O R I G I N O F L A T I N HA UD A N D G R E E K OY; A N D THE EXTENSIONS OF T H E ORIGINALLY U N E X T E N D E D FORM.1 A . — T H E O R I G I N OF L A T .

haud

GK.

oi

'NOT.'

§I. Introduction. Lat. haud and Gk. oi 'not' have long been the subject of discussion, but it will hardly be contended that the question of their origin has yet received a satisfactory answer. Under these circumstances I would venture to offer a new explanation in the following pages. §2. The three forms hau, haud, haut. The evidence of (a) Inscriptions, (J>) MSS and Libri, (c) The Ancient Grammarians. The form *au established as the earlier Latin form. One of the chief difficulties lies in the Latin form. Hence we shall do well to examine this word first, to see what its earlier form was in Latin. The three forms hau, haud, haut. W e have apparently three forms to deal with—namely, hau, haud, haut. Ritsehl, Prol. ad Plaut. Trin., pp. xcix-cii (1848), writes : " Corruptelis autem etiam haut scriptura non raro proditur, velut cum pro eo aut positum est, e. g. Trinummus, vv. 362, 721. Sed novum est quod duabus haut et haud formis tertia hau accessit, suscepta a me ex Ambrosiano vv. 233 (hau liquet), 462 (hau bonumsf), in eodemque codice aliis in fabulis tam saepe exstans, ut de calami lapsu cogitari nequeat." (a) Inscriptional Evidence. The usual form of the word on inscriptions is haud, e. g. C. I. L. I 1306 quoniam haud licitum, but we find hau in one inscription, 1 T h e present is the paper to which an advance-reference has already been made in the Essay on the ' E s t a b l i s h m e n t and Extension of the L a w of T h u r n e y s e n and Havet,' Part II (Amer. Journ. of Philology, vol. X V I I , part 2, July, 1896), §6, p. 180, n. 3, and §8, p. 189, n. 1, and p. 193.

44

AMERICAN

JOURNAL

OF

PHILOLOGY.

C. I. L. I 1007 ( = Orelli 4848 = Gruterus, p. 769) heic est sepulcrum hau pulcrum pulcrai feminaehau does not occur again in C. I. L. I, nor does it occur at all in the indices to C. I. L. I I - V (inch), V I I - X (incl.), XII, XIV. The form haut is found in C. I. L. II 562 'haut licitum' (an inscription 'aevi Antoniani' probably), and XII f T 499 1 haut dispar' ('ex titulis Christianorum'), but is not found in the indices to C. I. L. I,2 I I I - V (incl.), V I I - X (incl.), XIV. (b) Evidence of MSS and Libri. Otto Ribbeck, Scaenicae Romanorum Poesis Fragmenta, vol. I (Tragicorum Fragm.), 1871, gives the following readings :— hauquaquam, Att. 618. haud in Ennius 330 3 ; Pacuvius 325, 426 s ; inc. inc. fab. 30.3 haut in Livius 35; Enn. 340; Attius 108, 115, 193, 330, 360, 466; Fabul. Praetext. 31. Otto Ribbeck, op. cit., vol. II (Comicorum Fragm.), 1873:— hau, Liv. 3 (haut codices); Naev. 16 (Jiau om. libri) ; Titin. 181 (The MS has room for three letters only) ; Afran. 58 (aut libri) ; Pub. Syri Sententiae 461 ( = 694) (reading very doubtful). haud, Fabul. Palliat. inc. inc. 47*; Afranius 12*; Sententiae Turicenses 693 ( = 850). haut, Naevius 60, 112; Caecilius 181; Turpilius 9, 10; Fab. Pall. inc. inc. 74; Titinius 127, 166; Afran. 7, 51. L. Miilier, Q. Enni Carminum Reliquiae (1885), reads as follows:— haud, Annales 278, 389, 578; Fabulae 127, 424 ( = Ribbeck, vol. I, Enn. 340 haut, v. supra). haudquaquam, Ann. 293. haut, Ann. 499. In connexion with Ribbeck's 'velut cum pro eo (s. c. haut) aut I C . I. L . I 1007 ( = also F . D. Allen, Remnants of Early Latin, No. 138, where it is included among ' Epitaphs dating from about the Gracchan period on') and I 1306 are among the " Inscriptiones a bello Hannibalico ad C. Caesaris mortem" (see Mommsen in C. I. L . I, pp. 5, 43), and they are, I believe, the only instances of the negative in question to be found in C. I. L . I . 2 haut does not occur in C. I. L . I 1306, as might perhaps be at first sight inferred from Stolz, Lat. Gr. 2 , §69, p. 316. This particular inscription (quoted above in the text) shews ' haud licitumwith which contrast haut licitum in C. I. L . I I 562 (also quoted in the text above). 3 Wrongly placed under hatit in Ribbeck's index to vol. I. 4 Wrongly placed under haut in Ribbeck's index to vol. I I .

LATIN

HAUD

AND

GREEK

OY.

45

positum est' we may note that in the following of the abovementioned passages the reading aut is supported either by Libri or by at least one M S : Ribbeck, op. cit., vol. I, Enn. 340; Pac. 426; Att. 108, 115, 193. 33°Ribbeck, op. cit., vol. II, Naev. 60, 112; Turpil. 9, 10; Pall, inc. inc. 74; Titin. 166; Afran. 7, 12, 51, 58. Miiller, op. cit., Enn. Ann. 499; and Fab. 424 ( = Ribbeck, op. cit., I, Enn. 340). Ritschl, in his edition of Plautus, Trin. 1848, reads as follows (I add critical notes in brackets) : hau in lines 233 (hau A , ut videtur. haud reliqui), 462 {hau A, haud reliqui). 1 haut in lines 60 (haut A , haud reliqui, et sic constanter nisi ubi contrarium testabimur) ; 62 (haut A . R., haud reliqui); 90 (aut H.) ; 115 (haut B, haud reliqui) ; 362 (aut A, haud reliqui) ; 445 (haud B C D E , hau Camerarius, au Palmerius Spicil., p. 859) ; 584 (haud dare Pius, haddare B, addere C D E Z ) ; 625 (haut B, haud reliqui) ; 721 (haud Dousa iv. 24, Scaliger. aut libri) ; 835 ; "57( f ) The opinions of the Ancient Grammarians. Marius Victorinus (fior, about 360 A. D.), according to the reading of Ritschl, Proleg. ad Plaut. Trinum., page c, writes : "Hau adverbium est negandi et significai idem quod apud Graecos ov : sed ab antiquis cum adspiratione, ut alia quoque verba, dictum est et adiecta d littera, quam plerisque verbis adiiciebant. d tamen litteram conservât, si sequens verbum incipiat a vocali ut haud aliter muros et haud equidem. at cum verbum a consonanti incipit, d perdit ut hau dubiam et hau multa et hauplacitura refer.""1 A n alternative reading given by Ritschl (1. c.) runs as follows : 1 T o these statistics w e may add that Georges, L e x . der L a t . Wortf., g i v e s hau in Plaut. B a c c h . 506; M e n . 927; Most. 434, 720, 9 1 9 ; Pers. I I , 23, 500; Poen., Prol. 94; Pseud. 215. N i p p e r d e y , Ritter read hau in T a c i t u s , A n n . (e. g.) I I 36 and V I 43 (49).—P. S. R e f e r e n c e m a y profitably be made also to F r i e d r . N e u e , F o r m e n l . d. L a t . Spr. I I s 664 sqq. 2 T h e reading " a t cum verbum a consonanti incipit, d perdit ut haut dudum et haut multum et haut placitura refert, et inducit t" cannot possibly stand. K e i l ' s reading is, in the main, identical with that of Ritschl, and runs thus : " at cum sequens verbum a consonanti incipit, d perdit, ut hau dudum et hau multum et ' hau placitura refer' [et inducit /]."

46

AMERICAN

JOURNAL

OF

PHILOLOGY.

"Haud adverbium est negandi et significat idem quod apud Graecos oi, et fuit au: sed ab antiquis etc." 1 Flavius Caper (flor. before the end of the 4th century A. D.), according to the reading of Keil, Gramm. Lat., vol. V I I , p. 96, 1. 4, writes:— "'hau dolo' per d recte scribitur, etenim d inter duas vocales esse debet, quod si consonans sequitur, d addi non debet, ut ' hauscio.' "2 Charisius (flor. some time between the middle of the 4th and end of the 5th centuries A . D.), Institut. Gramm., bk. I, §xv ad fin., after a brief discussion of the particle sed, continues:— "Haud similiter d littera terminatur: cm enim Graeca vox 3 d littera termina[ri apud antiquos] coepit quibus mos erat d litt[eram omnibus] paene vocibus vocali littera finitis adiungere, ut quo ted hoc noctis \dicam prö\ficisci foras. Sed et per t scribi sonus vocis admittit." Charisius thus gives authority also for the form haut. Results of the foregoing investigation. (i) The three forms hau, haud, haut are established by the united evidence of Inscriptions, M S S and Grammarians. (ii) The earlier form of the Latin word under discussion seems beyond all reasonable doubt to have been *au. Thus we have before us Lat. *au 'not' : Gk. ov 'not'; and the problem is to find the connexion between them. The explanation of the h and d : t of hau haud: haut, which have been shewn above to be non-original extensions of the earlier Latin form *au, and the examination of the extensions of Gk. oi, viz. OV-K ov-X OV-KL ov-XT, will be deferred to the latter part of the present essay (§8, pp. 61 sqq.), where the various Latin 1 T h i s , w i t h t h e o m i s s i o n of R i t s c h l ' s ' e t f u i t au' is t h e r e a d i n g g i v e n in K e i l ' s G r a m m a t i c i L a t i n i , vol. V I , p . r5,11. 21, 22. 2 K e i l gives t h e f o l l o w i n g critical n o t e : " haud olo per d recte scribitur M ( = C o d e x M o n t e p e s s u l a n u s 3 0 6 ) ; haud dolo sic recte scribitur C ( = C o d e x B e r n e n s i s 3 3 8 ) ; haud dolo sic alioqui recte scribitur B ( = C o d e x B e r n e n s i s 3 3 0 ) ; r e c t i u s e r a t hau dolo per unurn d recte scribitur, nisi p o t i u s haud aliter s c r i p t u m e r a t , q u o d ex V e r g i l i i versu A e n . V i l l i 65 M a r i u s V i c t o r i n u s p. 15, 24 a d s c r i p s i t . " L i n d s a y , T h e L a t i n L a n g u a g e , c h . I I , §136, p. 122, s u g g e s t s a n e w r e a d i n g : haud uolo. 3 R i t s e h l (1. c., p . ci) p r e f e r s t o r e a d " hau e n i m , g r a e c a vox oil, d l i t t e r a e t c . " K e i l , G r a m m . L a t . , v o l . I , p . 112, 1. 8, a n d L i n d s a y , T h e L a t i n L a n g u a g e , c h . X , §18, r e a d " ov e n i m G r a e c a vox a?littera," etc.

LA TIN

HA UD AND

GREEK

OT.

47

and Greek extensions of the forms *au : oi, together with some kindred forms in the same or in some other Idg. languages, will be dealt with in detail. § 3 . Older

explanations

of haud

: oi

examined.

Before I venture to put forward my own views on the vocalism of Gk. oi and Lat. (h)au{d), it may be well to examine one or two of the older etymologies or explanations offered. Corssen, Ausspr. Vocalism. und Beton. d. Lat. Spr. 2 , vol. I (1868), p. 205, regards the au of (h)au{d) x as the 'Pronominal Particle' au, which in au-fero au-fugio has the meaning 'away, apart,' and occurs also, according to Corssen, 1. c., in Lat. au-tem, Osc. av-ti, Umbr. o-te, Lat. au-t (see id. ib., p. 157). This au, he says further, corresponds to Skr. dva, which properly means 'down, downwards,' but which, he adds (on the authority of Benfey, Chrestom. Gloss., p. 32 f.), in composition often contains the pure negative meaning ' -less, un-, not.' On the ground, too, that Pott (Etymol. Forsch., part II, 1836, pp. 64, 134) identified Gk. oi-K with Skr. avd-k, Corssen (1. c.) identifies Greek o'v with the same Skr. dva,1 which, according to him, has the form au in Lat.

h-au-d.

Thus Corssen identifies the vocalism of Lat. ( h ) a u ( d ) , au(fero), au(fugio), uu{teni), au{i), Umbr. o(te), Osc. av(Ji'), Skr. dva, Gk. ov. Leaving Lat. h-au-d and G k . oi for the moment out of the question, it may be remarked that only on one condition—viz. that Skr. dva and the au- of Lat. au-fero au-fugio represent an Idg. au (a view which I believe to be right, v. infra, p. 50, n. 3,) —can we regard as correct Corssen's identification of these with Lat. au(tem), aa(f), Umbr. o{ie), Osc. av{ti~), which must represent Idg. *au (see Brugmann, Grundriss, I, §§96, 97; Lindsay, T h e Lat. Lang., ch. X , §§4, p. 599, 5, p. 601). If, therefore, Skr. dva, Lat. au{fero) au(fugio) au(tem) au(f), Umbr. o(te), Osc. av(ti) are all to be identified together as representing Idg. *au, they must all be separated from G k . oi ' not' (which cannot represent Idg. *au"), and probably, therefore, also from Lat. *au 'not.' Thus Corssen's identification of Gk. oi with Skr. dva [based on 1

F o r Corssen's view on the h and d of h-au-d see beiow, §7 a and b, pp. 55,

592 B o p p (as I conclude from Brugmann, Gr. Gr. 2 , p. 236) was the first to identify G k . ov with Skr. dva.

48

AMERICAN

JO URNAL

OF PHIL

OLOGY.

Pott's incorrect identification of G k . oh< with Skr. avak (on the latter syllable of which v. infra, §8, p. 63)] cannot possibly stand. Osthoff, in Hiibschmann's Das Idg. Vocalsystem (1885), pp. 190, 191, regards " L a t . (JC)aud\ Gk. oh (from *ov8)" as shewing different ablaut-grades of the same root. For the ablaut " G k . ov : Lat. au" Osthoff, 1. c., compares Gk. otfara (Att. S r a from *oara) : Lat. auris, from Idg. y*aus- 'to draw, gather, take, obtain' [Gk. ava Lat. haurio (from *aus-id, Osthoff, Zur Geschichte des Perfects, p. 486) O.Norse ansa],1 the 'ear' thus meaning 'the grasping organ'—middle grade of G k . our in Lesb. irap-aia 'cheek' and weak grade in Avest. usi-.2 Osthoff's derivation of G k . oh from * o i S seems unlikely. It is much more probable that oh was the earliest Greek form, and that the forms such as ohd-ds shewing 8 came in later (cf. Brugmann, Gr. II, §31 ad fin.). With respect to Osthoff's view that Lat. haud Gk. oh belong to the a-series, I would not deny that 0 appears beside a in the strong grade of this series,3 but I would raise the objection that there is (so to speak) no Indo-Germanic ' p e g ' on which to hang Lat. haud Gk. oh, thus referred to the a-series. Victor Henry, in M6m. d. 1. Soc. d. Ling., vol. V I , part 5 (1889), pp. 378 sqq., seeks {unsuccessfully, I think) to justify Bopp's and Corssen's above-mentioned identification of Skr. dva with Gk. oh. He observes that at first sight the disparity of meaning between dva and oh is difficult, but remarks (on p. 378) that it is possible to see in Skr. the commencement of the proceeding, by which 1

It s h o u l d b e m e n t i o n e d that H a v e t , in M e m . de l a S o c . d. L i n g , d e Paris,

v o l . V I , part I (1885), p. 18, a n d K i n g a n d C o o k s o n , S o u n d s a n d I n f l e x i o n s in G r e e k a n d L a t i n , chh. V , p. 86, I X , p. 187, regard I d g . *ous- as the origin of L a t . auris G k . OVQ. B u t these scholars s e e m c e r t a i n l y to b e m i s t a k e n in their view.

C f . Osthoff, Perf., pp. 486 sqq.; H u b s c h m a n n , D a s I d g . V o c a l s . , p. 159 ;

L i n d s a y , T h e L a t i n L a n g u a g e , ch. I V , § 3 1 , a n d the p r e s e n t w r i t e r in his E s s a y on the ' E s t a b l i s h m e n t a n d E x t e n s i o n of the L a w of T h u r n e y s e n

and

H a v e t , ' Part I I ( A m e r . Journ. of P h i l o l o g y , vol. X V I I , part 2, J u l y , 1896), § 8 , p. 194, n. 3. 2

T i l l r e c e n t l y ' u n d e r s t a n d i n g ' w a s the only k n o w n m e a n i n g of A v e s t . usi-

(cf. A r m e n . us ' u n d e r s t a n d i n g , ' r e g a r d e d as a b o r r o w e d w o r d by H u b s c h m a n n , A r m . S t u d . I , p. 47), but the m e a n i n g ' e a r ' has b e e n e s t a b l i s h e d b y the n e w f r a g m e n t s of the ' N i r a n g i s t a n , ' v i d . D a r m e s t e t e r , L e Z e n d - A v e s t a , f r a g m e n t vi, verse 26, p. 95 kvaeibya usibya ' with the t w o ears.' 3Cf.

the a b o v e - m e n t i o n e d E s s a y on the ' E s t a b l i s h m e n t and E x t e n s i o n of

the L a w of T h u r n e y s e n and H a v e t , ' P a r t I ( A . J. P., vol. X V I , part 4, D e c . 1895), § 3 (P- 447, note 1).

LATIN dva,

HAUD

AND

GREEK

49

OT.

" q u i a parfois e n sanscrit le sens inversif," has b e e n

of transformation

into a n e g a t i v e

particle.

capable

But to pass on

to

w h a t is p e r h a p s m o r e i m p o r t a n t , h i s a t t e m p t e d e x p l a n a t i o n o f t h e form

o f t h e w o r d s : — h e s u g g e s t s t h a t dva m a y h a v e c o m e f r o m a n

I d g . *oy.o- [ a d d i n g t h a t L a t i n is n o t a g a i n s t t h i s , s e e i n g t h a t s h o r t e n e d f o r m au- (au-fero,

e y s e n a n d H a v e t , m a y r e p r e s e n t I d g . *ou j u s t a s w e l l a s I d g . T h e O l d I r i s h p r e p o s i t i o n o iia ( s i c 1 ) , w h i c h c o m e s f r o m I d g . would

correspond

t o S k r . dva,

f r o m I d g . *du(o).2

if t h i s l a t t e r b e r i g h t l y

A c c o r d i n g to B r u g m a n n ' s law Henry

expected

and

*dud

to

become

Skr.

*avd,

in

the

particles,

e. g . S k r .

apa

: Gk.

*du,

derived

would

not

Skr.

a d d i n g , h o w e v e r , t h a t t h i s r u l e is n o t a b s o l u t e , b e i n g notably

*a%].

t h a t I d g . d in

a n o p e n s y l l a b l e b e c o m e s a in S k r . , V i c t o r Idg.

the

e t c . ) , in v i r t u e o f t h e L a w o f T b u r n -

dno;

he

have dva,

violated further

e x p l a i n s dva b y s u p p o s i n g t h e c o - e x i s t e n c e o f t w o o r i g i n a l f o r m s , a full o n e *dud ( o r * o u a or *oue), a n d a s h o r t e n e d o n e *du ( w i t h w h i c h h e c o m p a r e s L a t . au-), *ava),

t h e f o r m e r y i e l d i n g S k r . *ava

(or

t h e l a t t e r y i e l d i n g S k r . *dv, a n d c o n t a m i n a t i o n o f t h e t w o

producing

ava.

S o f a r a s I o n i c A t t i c a l o n e is c o n c e r n e d , V i c t o r

' ¿ a (as accentuated by Victor Henry), which should strictly represent iia (for the 'accent,' which is written in Gaelic, is really no accent at all, but only a mark of quantity), may be criticised as not being quite an exact way of representing the true sound of the word. Some diphthongs in Gaelic may be ( i ) short, (2) long with respect to the first vowel, (3) long with respect to the second, but there are also (4) a few others, which are always long\ to this latter class ua belongs. Diphthongs of this latter class are never ' accentuated'; thus iia, as written by V. Henry, is both right and wrong—wrong with respect to the notation, right with respect to the quantity (ua being a long diphthong). The form o belongs to Scots (as well as Irish) Gaelic, e. g. 0 urnuigh ' from prayer,' St. Luke xxii. 45 ; o Ghalile ' from Galilee,' id. xxiii. 5. (This 0 of Scots Gaelic is of course long, as in Irish Gaelic, but Scots Gaelic very rarely makes use of the 'accent.') T h e form ua, however, is, so far as I know, peculiar to Irish, and even there is, I believe, retained only in the prepositional pronouns which are formed with this preposition, e. g. tmim (= zta -(- me) 'from me,' uait ( = tta-\-tii) 'from thee' [all the simple prepositions in Irish being thus compounded with the personal pronouns, an odd feature common to Gaelic and the totally unrelated languages, Hebrew (e. g. Immanu-el 'with-us God') and, I am told, Hungarian], Of ua I can find no trace in Scots Gaelic. Curiously enough, in O'Donovan's Irish Dictionary, the only form given of the preposition in question is ua, but the form d ( = 0) is now almost invariably written, even in printing old texts which shew ua. 2

See infra, §4, p. 52, n. 1.

50

AMERICAN

JOURNAL

OF

PHILOLOGY.

H e n r y o b s e r v e s that ov m i g h t c o m e from the full form *oFo,1 but that as this w o u l d not suit L e s b . , Boeot. and D o r i c , in which the said *6fo w o u l d have b e c o m e *«, he prefers to derive ov from the s h o r t e n e d form *oF ( c o m p a r i n g Trap = napd, Hv = ava, etc.), w h i c h satisfies all the phonetic needs, the F before consonants f o r m i n g a d i p h t h o n g with the p r e c e d i n g vowel, so that, e. g., *oF (¡¡api ( w h i c h , a c c o r d i n g to V . H e n r y , = S k r . dva bhami) b e c a m e ov r)/u. O n V i c t o r H e n r y ' s t h e o r y the following r e m a r k s m a y be m a d e . H e is s u r e l y w r o n g in s u p p o s i n g that B r u g m a n n ' s law, viz. that " I d g . 6 in o p e n syllables b e c a m e a in the Prim. A r y a n p e r i o d " ( B r u g m a n n , G r . I, §78), is intended to include final syllables, for B r u g m a n n himself, in G r . I l l , §409, r e g a r d s S k r . sa (: G k . o) sd-s as the S k r . representative of I d g . *5 seems beyond all reasonable doubt to be correct. T h e common I d g . ground-form of this G k . -xl Skr. hi, beside which latter we find also Skr. -hi,1 is *gk% as given by Osthoff, 1. c., i. e. *ghi. T h i s is proved to be correct by the Avestic zi, beside which we find also Avest. zi 'for.' 2 Skr, hi and G k . -xl, taken by themselves, might quite regularly be derived from either Idg. * g h i (cf. Brugmann, Gr. I, §§386, 405) or Idg. *%hi (cf. Brugmann, Gr. I, §§425, 445, 454), but if (as seems undoubtedly right) they are to be identified with Avest. zi, whose z can only represent Idg. *gh (cf. Brugmann, Gr. I, §405, and also §452 on Idg. 3h as represented in Iranian), it follows that *ghl must be set up as the original Idg. form (cf. also infra, p. 65, note 2, on O l d Pers. *di ' for'). T h e meaning of this Idg. *ghi (Gk. -xi Skr. hi Avest. zi) m a y be fixed with tolerable certainty by a comparison of the Latin demonstrative pronominal stem *ho-, in Lat. hie (from earlier Lat. *ho-i-ce, the -ce of which comes from the Idg. demonstr. pronom. stem *ko-), etc., which can come quite regularly from ' B e s i d e Skr. nahi (from nd-\-hi) we find naht before ml in R i g v e d a 167, 9 ; 314, 4 ; 623, 1 3 ; also Prätis. 442, 483 [vid. Grassmann, Wörterb. zum Rigv. (1873), s. v. nahi]. Is the i of this Skr. -hi to be compared to the I of Avestic zi, or is it due merely to metrical reasons ? W e find also nahi nü in R V . i. 80, 15 ; vi. 27, 3 ; naht nü in R V . viii. 21, 7 (v. Osthoff in Morph. Untersuch., vol. I V , p. 340). 2 T h e i of Avestic zi is curious; beside it we find once zi' for,' Y t . xiv. 12 (v. Justi, H a n d b . d. Zendspr., 125 b), also -zi in ya-zi ' if, w h e t h e r ' from *yad-zi seen in yazi-ca ' a n d if,' Y t . xxiv. 47 (" lies yezica" ? says Justi, op. cit., s . v . yazica),ye-zi ' if, w h e t h e r ' for regular ya-zi [v. Osthoff in Morph. Unters., vol. I V (1881), p. 240].

LATIN

HAUD

AND

GREEK

OT.

63

I d g . *gho-} T h e relation between I d g . *gho- and I d g . *ghi would be the same as that between I d g . *ko- and I d g . *ki(Brugmann, Gr. I I I , §409, p. 329), and that between I d g . *qoand I d g . *qi- (Brugmann, Gr. I I I , §411, p. 333). Thus I d g . *gho- (: Lat. stem ho-) *gki- ( G k . -^i Skr. hi Avest. zt) would be a demonstrative pronominal stem, identical in meaning with I d g . *ko- *ki- 'this.' T h e development of the meaning ' f o r ' in Skr. and A v e s t . is easily traced: I d g . *ghi would originally mean ' t h i s ' (the meaning perhaps still discernible in the -^l of Gk. ov-xt), and might have been used in answers to a question, as we often say 'Just this,' 'Just so,' ' W h y , this,' ' W h y , just this,' before proceeding with our answer to the question. F r o m the use in such phrases, it might well have developed gradually into a pure conjunction 'for.' It may be observed in this connexion that the Skr. V e d i c negative nahi, besides its usual meaning ' f o r not,' shews sometimes the simple meaning ' n o t ' or 'indeed not' ( v . Grassmann, W ö r t e r b . zum R V . , s. h. v., and Delbrück, Vedische Chrestomathie, p. 84, s. h. v.), in which cases it comes very close to G k . ov-xt. OV-Kt. — T h e -kI of ov-kI, quite distinct from the -^i of ov-xt (just discussed), is the neuter singular of the demonstrative pronominal stem *ki- 'this' (cf. Osthoif, 1. c.; Brugmann, Gr. Gr. 2 , §95. P- 131. and Gr. I I I , §182, p. 49, E. E., §331, p. 330, E. E . ) , so that oi-Kt meant originally 'not this,' and is thus identical in meaning with oü-xi"not this' (v. supra). (2) OVX, OVK. N o n e o f the theories yet advanced in explanation o f Gk. ovx, ovk seems at all satisfactory. W e may at once dispose of Pott's above-mentioned (§3, p. 47) identification of Gk. ovk with Skr. Aväk. Skr. Aväk is of course, strictly, the nom. acc. sing. neut. of Skr. Aväc ' turned downwards,' which is compounded of Skr. Ava 'off, d o w n ' and -ac- 'bent in a certain direction, turned' (from I d g . *-nq-, seen in Gk. nod an-6-s, L a t . p r o p - i n q u - o - s ) ; cf. Whitney, Skr. Gr., §§407,409; Brugmann, Gr. I, §228, p. 195. Osthoif, 1. c., explains ovk : oix in the following w a y : — h e holds that in cases o f apostrophe o f the -1 before a following aspirated 'Brugmann, Gr. III, §409, Rem. 1, pp. 330, 331, was doubtful how to derive the Latin stem ho-.

64

AMERICAN

vowel,

JOURNAL

OF

PHILOLOGY.

and OIXL fell together in ofy'; so that e. g. OIX on, OIX can belong to both OVKI and OIXL; before EN he thinks there probably existed both OV-K- and OI-X- (ov-«-m and OIL~X~€TL), comparing the E t y m . M a g n . 368, 30 ¿IFFXLXE YAP XE'YEIRDAI OIX EPXOP-AI. Then at a later date, according to his theory, *OIX-En and *OIX' EPXOPAI g a v e way, the use of OIX, beside OIK from OVKI, becoming confined to the position before spiritus asper on the analogy of other cases, such as ACF>' beside ¿71-', KU8' beside NAR, DUD 1 beside so that thenceforth OIX appeared only as a phonetic (or graphic) modification o f o i * . Victor Henry, in M6m. d e la S o c . de Ling., vol. V I , part 5, 1889, pp. 379 sq., without taking mid into consideration at all, oilK a s both derivable primarily from oixl. regards oix His explanation is that " before an initial vowel the I is elided, e. g . ofy and if the following consonant was an aspirate, then the X had to lose its aspiration, whence owe e^o), OI< RJXDOV ; whereupon," and according to his theory, " t h e relation of OIK e^x others similar, g a v e rise to the belief that the X of this last combination was due to the rough breathing of e|\E." But surely there is at hand a much simpler explanation than either of these two latter. If, as is most probable, oi-xl and ov-kt both had originally the same meaning 'not this' (v. supra, pp. 62, 63), then both would be used indiscriminately. This being so, it was but natural that in cases of the elision of the I before a following aspirated word the form oixCO would be chosen, while on the other hand before a following non-aspirated word the form used would be OI