Ancient Survivals in Ossetic

Ancient Survivals in Ossetic Author(s): Ilya Gershevitch Source: Bulletin of the School of Oriental and African Studies, University of London, Vol. 14, No. 3, Studies Presented to Vladimir Minorsky by His Colleagues and Friends (1952), pp. 483495 Published by: Cambridge University Press on behalf of School of Oriental and African Studies Stable URL: http://www.jstor.org/stable/609110 Accessed: 26/11/2010 06:05 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=cup. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

School of Oriental and African Studies and Cambridge University Press are collaborating with JSTOR to digitize, preserve and extend access to Bulletin of the School of Oriental and African Studies, University of London.

http://www.jstor.org

Ancient Survivalsin Ossetic By

ILYA GERSHEVITCI

Laegcal aevzagizona ual legi'y - uoyyau du daer, mraaxuzrgzenaeg. Praetextus HEN Soslan went to the land of the giant Uomar, son of Tara, he pretended to be his own hireling: Tari furt Uomari zaenxaemae ranzexstUeraey ye'xuzersti r§euoni, Pam. ii 60. The Dictionary translates Dig. raeuonaeas ' cause, pretext, appearance ', and quotes our passage to illustrate the third meaning: 'in the appearance of his hireling.' There is, however, another passage, in which rteuonaeclearly means 'disguise'. Uaskergi had placed watchmen at the gorge of Zadaelesk,to prevent Ceenxigol,the spirit of salt, from leaving the heights. The villain Sirdon advises Ceenxigol: mnengadcay rewaendtey ku naerayervzezayuaedtadaeramardzzenencae' if you do not escape by means of my bitch's raeuaendtae (plur. of raeuonae)they will kill you'; thereSirdoni Caenxigol upon gadcay cari bacudzey ma uotemaey qalaurt3ebael "Sirdoni rayevhudcey, gadca aey!" zzehgaey' Cenxigol entered the hide of Sirdon's bitch and thus passed by the watchmen, who said: " it is Sirdon's i 102). Miller translated maen gadcay bitch "' (Miller, OcemuucKuue 3mioObQ as 'by the ruse of my bitch'; evidently the meaning is 'by rneuzendtaey is also used, according disguising yourself as my bitch '. The plural rneuaendtse to B., my Digor friend, when the word means 'pretext'. for ' disguise ', the two passages quoted Although B. would not use rweuonxe show that speakers of an earlier generation were aware of this meaning. The semantic connexion between 'disguise' and 'pretext' is known in other languages. Hiibschmann, Pers. St. 32, quoted parallels from Latin and German in support of a derivation of NPers. bahane 'pretext' from OIr. *uahdna-, belonging to uah- 'to dress'. Dig. r§euon§e,from *fra-uahdna-, confirms the correctness of this derivation. Abayev, PyccKo-ocemu7cKuui cJzoeapb, s.v. npeO.ozt, gives also Dig. Iron §efson,for 'pretext'. The former, which rzeuagBe,and Dig. §efsoyn§e, the and means to B. was ' cause', according Dictionary compared by Miller, with ' Skt. We announcer'. Ossetisch,35,2 pravdka may prefer to derive it from *fra-uaha-ka-. While in NPers. bahdne, Oss. rzeuonaeand raeuaggethe disguising aspect of 'dressing' is stressed, the relation between dress and custom or disposition, 1 Read ma instead of ne. 2 Where the derivation of reuaeg ' light, quick ' from *fravaka- must, at any rate, be preferred to the one proposed by O. Szemerenyi, ZDMG 1951, 209 sq.

484

I. GERSHEVITCH-

which finds expression in such words as Ital. costume,Lat. habitus,suggests that also Oss. uag (Dig. uagae) 'custom, normal condition, character', Plur. uagtae 'manners ', belongs to the base uah- (*uahaka-). In tefson, for which the Dictionary also gives ' cause, reason' the meaning 'pretext' is perhaps to be understood as 'adduced fault'; the example in the Dictionary cey aefsonmmnna radtay dae baex 'on what grounds did you not give me your horse' means to B. 'because of which (= whose) fault or kaenunce ' they say guilt, etc.'; moreover I have noted from B. uoy wefsoynae'y it is his fault (lit. make it his fault) '. If one compares Lat. caussa 'cause, pretext, guilt', and its relation to causor and accuso, it seems possible that in tefson, from *abi-sihana-, an Indo-Iranian legal term has survived. Skt. abhi-'amsamna 'accusation' has been discussed by Hopkins, JAOS 45 (1925), 49 (a reference I owe to Professor Bailey). In the Avesta (Yt 14, 34) ZaraOustra asks for Ahura Mazdah'sadvice on how to behave if he is aiwi.sast5 aiwi.smardat by many opponents. Bartholomae's translation was ' verwiinschtin Wortenund Gedanken', but ' accused or assailed verbally and mentally ' agrees better with what we know of abi-sth-, even though its converse, pati-syh-, has a magical acceptation in Ahura Mazdah's reply (cf., on the other hand, paiti.sa.nhdm 'contradicendo' in Vend. 4, 54, and the juridical connotation of Pahl. passoxv (< *pati-sahvan-) ' sureties', v. Bailey, BSOS vi 83).

payah-, payuIn Oc. 3m. ii 86 Miller doubtfully compared Oss. ftezdon(ae)'udder' with Av. fstdna-, NPers. pistdn. The comparison no longer appeared in Ossetisch, since the voicing of -st- could not be justified. There is, however, an easy alternative etymology at hand. -don(ae)may be the second term of compounds which continues -ddna- ' container'. Comparisonwith Bal. goddn, Ps yulanja 'udder' < *gau-ddna(c5)-,cf. Morgenstierne, NTS xii 263, EVP 41, further suggests that fwtz- represents Av. payah- 'milk'. Hence *pay(a)z-ddna- (on -az-, cf. Bailey, BSOAS xii 331) > *frzzdon (on the assimilation of y to a following consonant, cf. Bailey, Asica 12 sq.) > fxzdon; cf. Wx pez,n, Morgenstierne,IIFL ii 536; payah- also survives in Ps pa?, etc., v. EVP 55. Iron Dig. fiyyau, Iron fiyau, fayyau, fiyyau ' shepherd ' probably represents a metathesis of OIr. pdyu- ' protector '. The Ossetic restriction to the protector of sheep may have been inherited from IE times, cf. Greek 7rciv ' flock of sheep'. The same metathesis can be recognized in Oss. yau(ce) 'energy, vital force', if, accepting Morgenstierne's etymology, NTS xii 269, we derive it directly from dyu-, instead of assuming secondary lengthening of a. One may compare 1 enseuag, enuag, for which the Dictionary only gives 'slovenly, ill-bred, depraved' (= ' without manners '), also means ' unusual ': Fidselti dogi nse bserzondxusenxti Ku iscseykodta senseuagsestong 'at the time of the fathers there arose on our mountains an unusual famine', Maliti Geusergi,Irasf, Ordzonikidze, 1935, p. 33.

ANCIENT SURVIVALS IN OSSETIC

485

the Sogdian metathesis which Henning established in fsy'ws 'gentleman'
486

I. GERSHEVITCH-

For imis- one would normally reconstruct OIr. *mis- or *mis- + -ui-. A base *mis- (OPers. *miO-) used with the same preverb, has been postulated by Henning, ZII ix 179, for MPers. wmyh- 'to shape, fashion, create'. This verb is only found in three passages, all of which refer to the creation of man by Az, the equivalent of Hyle in Iranian Manicheism. The circumstances of this act of creation must be borne in mind. It consists in reproducing from memory the image of a seducing divinity whom the Archontes, fettered in the sky, had once seen; by devouring the abortions of the Archontes, Az had inherited the vision of this image. Numerous passages, in which the place of Hyle is usually taken by a demon, insist that man was modelled after this image: ich will euch ein Bild herstellennach dem Bild des Erhabenen (Keph. 1383) = sic quippe illius magni, qui gloriosus apparuit, imaginemfingam (August. De nat. boni 46); the eKcovof the seducer is imposed on the 7rhdTA4ua of Adam and Eve (Keph. 13717),etc., cf. Mary Boyce, BSOAS xiii 910 sq., Baur, Man. Rel. 132, 138 sqq., Cumont, Recherchesi 42 sq. In the MPers. Cosmogony (Mir. Man. i) wmyh- is twice followed by dys(20229sq. 20328sq.) 'to build' (v. Henning, BBB p. 56 bottom), once by kyrdn ' to make' (19530). Since neither verb adequately correspondstofingo, lrAaaaco (Keph. 13717),or wcoypace'o (Keph. 13518,13814),the artistic connotation of the latter will belong to wmyh-. In addition the three passages apparently contain no reference to the imitative character of this creation, which in the corresponding descriptions in other languages is explained in detail; therefore the meaning of wmyh-may have been such as to supply this reference. If Oss. mnsin is connected it would appear that wmyh- meant 'to compose ,or fashion according to a mental pattern'. The suggested connexion excludes Morgenstierne'setymology of Oss. misin as belonging to Skt. vimrr-, NTS xii 268. fraxstiMiller, Ossetisch31, translated lxxsta (Dig. lixste) kJenunas ' to pray, cross oneself'; he assumed -taeto be the Plural ending, and the second meaning to be the original one; he accordingly connected the word with Skt. laksa ' sign '.1 Munkacsi's Iron informants used ltexste kxnun for 'to request, ask ', and ' interpreted lexstae as Plur. of lehz polished ', hence, with ksenun, 'to use polished words ',2 Bliiten der oss. Volksdichtungi 38 sq. Since the meaning ' to cross oneself' appeared to be secondary, Munkacsi rejected Miller's etymology. The Dictionary only has lexstyse ' imploring ', and lexstg kxnun ' to implore, 1 Internal Iron se corresponds to Dig. i (after I or before r) also in lehz 'polished ', daerzaeg ' rough ', sevdserzin' to irritate (skin) ', ent'serin 'to chase away, banish ', and xaerlnk'a'pocket knife ', v. Abayev, Oc. s3. u 96oJm. i 369. Of these words the first has been compared by Miller with Skt. Slaksna, Ossetisch32, the second (to which presumably the third belongs, v. Dictionary) with Av. darazista-, Oc. Sm. iii 151. 2 B. views this popular etymology with disfavour.

ANCIENT SURVIVALS IN OSSETIC

487

pray, entreat'. When the beautiful Gidannaebegs her father not to force her to marry the JEldar's son, her voice is lixsta 'imploring' (radzoruyyin lixsts haeslassy,Maliti Geusergi,37). In the Ossetic tradition on the Tukka disease it is said (Miller,Oc. 3m. i 100, 17): Xucauaencid balixsts kandzinan caemaeynm sas xuarzenxaeua, uoytuxxaey'let us pray to God that His benevolence be with us '. In such expressions, B. tells me, kaenunis also used with the Plural lixstitas. One therefore wonders if laexsta need be a Plural. In Yt 1, 30, we read: ahe nars asaonofravasm yazamaide yo asmo.xvanvd nyma, aSdt anyaesgcmasaongmfraxsti (MSS.fraxsti, fraxsta, fraxasta, fraxasti, fraxsasta, fraxusta, fraliusta) yazdi fravardta 'We worship the Fravasi of the truthful man A.; then I worship fraxsti (the Fravasis) of the other truthful Bartholomae uneasily translated fraxsti by 'abundantly', and (men)'. Lommel, Ydscht's, 18, n. 3, explained: as abundantly as they are worshipped in Yt 13. The translation 'with prayer' or ' entreatingly' (Instr.) does not any the less justice to the context, and lasxste, which could be derived from any of the Av. variant readings, may be adduced in its support. The reason for preferringfraxsti to the variants in Yt 1, 30, is the occurrence offraxsti.dd- as an epithet of MiOrain Yt 10, 65 (var. frastio). On the strength of the presumed meaning of fraxsti- this epithet is translated as 'giving abundance'; it could equally well mean 'granting the entreaty '.1 Bartholomae separated fromfraxsti- the first name by which Ahura Mazdah calls himself in Yt 1, 7: fraxstya- (Var. fraxstia, fraxstaya, fraxrastaya, fraxstoye, fraxstauye, fraxastuye, fraxa.stuye). This he analysed as Fut. Pass. Partc. offras- ' to ask' ' der zu befragenist' (?). The suggested meaning is acceptable, the form less so; for Fut. Pass. Participles in -tya- are not otherwise recognized in Iranian. Without incurring this difficulty a similar meaning, ' implorandus', is obtained if we relate fraxstya- to fraxgti- as yasnya- is related to yasna-. The origin of these words remains uncertain. For Avestan, Bartholomae's connexion with fras- could be maintained and extended to fraxsti-; the -tinoun with intrusive x might have acquired a meaning distinct from that of the regularfrasti-. However, while intrusive x is also known in Sogdian (GMS § 257), it has so far been assumed to occur in Ossetic only in Iron yaxs, Dig. sexss ' whip ', which Miller, IF xxi 328, doubtfully derived from Av. yaxsti-. If this etymology should prove to be correct, it would open the door to a derivation of laxsta fromfraxsti- even if -ta were the Plural suffix. One could then compare Dig. *lixsas: Plur. lixstitaswith Dig. uose ' woman': Plur. uostitx (connected by Miller, l.c. 332, with Skt. yosit 2); lixstaswould then be secondary, either as a Plural built on *lixsB, or as a Singular abstracted from lixstitx. 1 Cf. the meaning of Dig. iskurdiadse(abstract of korun ' to request, ask ') ' ability, or right, to receive from God what one asks for'. ' 2 Hence Dig. uozun, Iron uzin to rock, swing' < Av. yaoz-, and Dig. uodun, Iron udtn 'to exert oneself' < Av. yaod-, cf. Parth. ywdy-.

488

I. GERSHEVITCH-

haxaThe meaning of Av. haxa- was established as 'sole of the foot' on the strength of the Pahlavi translation ' under the foot ', and the context in the two parallel passages Vend. 8, 69 sq., and 9, 24 sq. In these Nasu, the demon of putrefaction, is chased down the body along successive stages, of which the last six are as follows: knee, calf,l ankle, forepart of the foot (frabda-, cf. Turner, BSOAS xii 641), haxa-, and 'under the toes'. To be able to sprinkle the haxa- with water it is sufficient to raise the heels, standing on one's toes. The Avestan context alone arouses the suspicion that haxa- is not the sole as 'under surface of the foot', but that part of the sole which is situated between the heel and the beginning of the toes. This suspicion can now be confirmed by reference to Oss. xx ' hollow of the foot between heel and toes' (adj. zxdzn), which evidently continues Av. haxa-.2 hiz-, xizIn Asica 32 sq. H. W. Bailey derived Oss. xezun ' to climb, pass over' from a base *haiz- attested in Sogd. -xyzyy ' creeping', MPers. (etc.) dxez- ' to rise ', Khot. pah7ys- 'to flee', va° 'descend', bio 'decrease', (tti)hTys-'rise', and Av. pdiri.haeza- ' go around'. With Av. pdiri.haeza- W. B. Henning connects MPers. phryz- ' versari' (= Parth. prxyz-) and 'to abstain', assuming that with the latter meaning the preverb was para-, v. BSOAS xi (= Tales) 62 n. 2, 724 on 1114,3xii 56 on 25. Previously Horn had referred to Oss. xezun and to Av. °haeza-(for which he quoted the variant °xaeza-)in connexion with NPers. xez-/xdst- (Npers. Et. 102 sq.), and Morgenstierne in connexion with Ps xez-/xat- (EVP 98). Various attempts have been made to reconcile the vocalism of NPers. xez- and xdst-. Andreas thought of an occasional Iranian replacement of e by d, v. Lentz, ZII iv 270, 292. Bartholomae, WZKM 25, 257 sq., suggested two alternatives, of which he preferred the second: (1) IE 1 Av. ascu-. The expected OIr. *asku- survives in Oss. sgauta (Pam. iii 413), skutas (Narti Kaddztas, Dzaeudilhaeu, 1946, 3521) 'haunches (as food) ', Plural of sgu (Dict.), isgu (Abayev, Pycc;o-oc. cJ.oeapb,s.v. 6eOpo). 2 A compound with sx may be recognized in Dig. anguldzasxta ' dance on tip-toes', which, acc. to Abayev, Oc. s3. u 06OAbK.,i 441, corresponds to Iron k'ax-k'uxtll kaft. As k'ax-k'ux is 'footfinger = toe ', so anguldzax may be ' ax-finger ', with the order of compound terms inversed as in yaugaf 'fish-millet = caviare', xasffindz ' nose-matter = snivel', and others discussed by Abayev, op. cit., 234. senguldzaxta kasnun 'to dance on tip-toes' is used by Maliti Geusergi, Irsef, p. 48; B. knows anguldzasxtaskafun. One is reminded of the expression gsndzaxta cashdin 'to be seized with convulsion, cramp', cf. Dict., 389, 1639. It occurs, e.g. in Pam. iii 521(Iron), where the corresponding Digor version (Pam. ii 126)has tasppasztacashdun. The latter is explained p. 180 n. 68 as 'to beat, shake one's extremities '. gsndzaxts, which, acc. to the Diet., is also Dig., may thus contain sx ; gasndz-is not clear; cf. Dig. gasndzu,Iron gascci ' tooth ' (nursery word) ? Since haxa- may belong to Av. haxti (Dual) ' thighs' (as Professor E. Fraenkel has convinced me by referring to the relation of Lith. kulnis ' heel' to kiulse ' hip', cf. Walde-Hofmann, Lat. Et. Wb.3, 144), its closest cognate will be Gr. taXt o0av's (Hes.), v. Meillet, MSL 23, 259. 3 Chr. pryyz- ' umstellen ' may contain a base *yiz-, which is attested in Dig. hizun, Iron qtxzn ' to threaten, wish ill', cf. Pam. ii 4332.

489

ANCIENT SURVIVALS IN OSSETIC

*khag-/*khakto-; from regular Ir. *xista- a Present *xaiz- could have arisen analogically, while xdstan would have preserved the root vowel a under the stress; (2) IE *khaig- might have given an Iranian Past *xdista- from which *xdsta-. Henning contemplated, but rejected the assumption that two Ir. bases, xaz- and xiz-, had influenced each other, BBB p. 105. The previous discussions leave out of consideration the variety of meanings of Oss. xezun, which does not seem to be irrelevant to the involved problems raised by this group of supposedly related words. It is desirable first to present the evidence. (1) Av. pdiri.haezarha, 2 Sg. Impt. of unclear meaning, occurs in Vend. 21, 4. 8. 12. 16. Here is a table of the variants :4

8

12

16

K1

pdiri.haizawha

pdiri.xaizavuha

pdiri.haezavuha

L4

pdirizaauha

pairi.xaizavuha

pdiri.haezawha

Jpl

pairi.haFzawha

Mf2 L2

paiti.haizawaha

,,pairmhaizawha pairi.xsaezavuha

Brl

,

Dhl

,,

K10

,,

M2

,,

pdiri.xsaezavuha

pairi.xsaezavuha

pairi.xsaezavuha

pairi.xsaezaauha

pairi.xsaezaaha

pairi.xsaezawha

pairi.xsaezavuha

pairi.xaezavuha

, ,,

02

pairi.s'aezauuha,,

B2

pairi.xsazar uha

L1

pairz.haezawha

,,

pairi.xsazavuha

pairi.xsaezavuha

pairi.xsaizawha

pairi.xsaezawha

(2) Parthian. (a) prxyz-: prx'st- 'to stand around, be about, versari', sometimes 'to stand around for the purpose of looking after someone = serve, nurse, protect', often merely 'to be' (Henning, Tales 62 n. 2). (b) 'xyz-: 'x'st- ' to rise'. (c) Once 'x'z- 'to rise', supposedly scribal error, cf. Ghilain, Essai, 61 sq. (d) mwrd'xyz, mwrd'hyz', adj., 'arousing the dead.' (3) MPersian. (a) xyz- 'to creep', v. Henning, BBB p. 61 on 512. (b) x'yst (with 'br) 'to rise' (v. Henning, ZII ix 178, where also the following six). (c) 'xyz-: 'xyst- ' to rise', Caus. 'xyzn-. (d) 'wxyz-: 'wxyst- 'to descend'.

490

I. GERSHEVITCH-

(e) prxyz- (fra-) 'to advance' (?). (f) whyz-, Pahl. Iss. why6-, as Professor Henning tells me, 'to move,' as from one country to another. (g) mwrd'xyz,mwrd'hyz,as (2) (d). (h) ryst'hyz 'resurrection'. (i) phryz- : phryst-, meaning as (2) (a) ; Arm. pahest,Arab.fthrist' deposit, Aufbewahrung', v. Henning, BBB p. 113b. (k) phryz- 'to abstain, beware of'. (4) Sogdian. (a) (a) Chr. 'yz- 'to rise (from the dead) '; (/) Man. "x'zyy4 M 900, 5 (no context). (b) Chr. 'xs't' arose ' i 53, 5,' got up ' ii 4, 12. Giw 232. (c) Causative Chr. 'xy'- 'to place, appoint' i 40, 1, 'raise up' i 30, 15; 'yyz- 'to resurrect' Giw 84. (d) Man. Chr. 'wyz-, Man. 'wxz- 'to descend', Past Chr. 'wxs't-,Caus. B. w'y'yz (Jmpf.). (e) Man. 'nyz-, mnxz- (Impf.), Past 'nx'gt-, B. 'ny'St- (anxas't-)'to rise' Caus. B. 'ny'yz-. Man. txyz ' setting (of sun and moon) ', v. GMS § 662. (f) (g) prxyz- ' versari', BBB f 91. (h) xyz- 'to creep' in z'yxyzyy. (i) Man. pcxyz-, Chr.pcyyz-, B. pcy'yz- : B. pcyst- ' to kneel ',cf. Henning, BBB p. 105. The Man. spelling, which confirms the z conjectured by Henning, occurs in the following passage (M 130 ii V): [p]cyxyz 00 'rrtyy w'nw (2) [p]tS'kw't8'rto kt w'f (3) [m]zyx yw'n kyy 'tyy 'zw (4) [']qmw8'n oo 'rtyy s't (5) trw yrp'_8'ryy oo 'tymyy (6) en mzyx knelt and addressed him thus: as many great sins as I have .. .'he committed, you have known (them) all; from my great...' (5) Khotanese. (a) hTys-'to rise', or ttih;iys-('to descend'? cf. (4) (f)). (b) pahTys-: pahissta-'to flee' (or simply 'go away', as Professor Bailey tells me). (c) vahTys-: vahista- 'to descend'. (d) bihiys- to decrease (6) Khwdrazmian xyz- 'to get up', v. Freiman, CoaemcnoeBocmooa6eOenue vi 83. (7) Ossetic Dig. xezun : xist-, Iron ~x~zin: xist-. xezun ' to climb a tree, (a) ' to climb ' with Loc. ext. : bekasame, xonxmnaw a mountain'; bunma x. 'to climb down' (B.). (b) 'to climb over, across; pass over' with Adess. or Loc. int. kaua?balx. 'to climb over, across a fence' xanxtebbl x. 'to climb over mountains' (B.), Iron sau dendiilzi sarti x. 'to pass over the Black Sea' (Dict.).

ANCIENT

SURVIVALS

IN OSSETIC

491

(c) 'to wait' (this meaning is found only in Dig., v. Abayev, Oc. S3. u 96OJibP. i 485): ba-mtr-xeza 'wait for me'; d' ercudtdin xezun 'I am waiting for your arrival'; d'ercudtmaedin xezdzarnern'I will wait until you arrive '; duute saxadti ftexxiston fal n'erbacudtey'I waited for two hours, but he did not arrive' (B.); Kettagi hasumefteccasuy,Uoy Ket ye'mburdiku xezuj 'the Ket-man returns to the village-the Ketians are waiting for him in the assembly ', Maliti Geuaergi,66. (d) ' to protect ': ked nte Xucau xezidteyecifidbiliztay ' so that God may protect (or deliver) us from this evil' (Miller, Oc. 3mn. i 100, 7); Iron xi xizmn'sich hiiten ' (Dict.); dtexe si xezae 'beware of him'; zelgtrdonaeydaexexeze ' keep off whirlpools' (B.); cf. also the pr. name Uadsexez(te),interpreted as 'protecting from wind', Pam. ii. p. 183 n. 121. (e) 'to pasture,' trans. and intrans.: yeytefons xezuy 'he pastures the cattle '; fons buduri xezuy (fexxistaey) ' the cattle is (was) grazing on the plain'; fons xezgaekrnuy ' the cattle is grazing ' (B.); v. also Dict. NPersian. (8) (a) xez-: xdst- 'to rise'. (b) Kabul Pers. Past xest-. (c) rastaxiz ' resurrection'. (d) xezidan 'to rise, leap; totter, slide, go softly; creep on all fours' (Dict.). xazldan 'to creep'. (e) parhez-: parhext- 'to abstain ', v. Henning, ZII ix 215. (f) (9) Ps xez-: xatdl (3 Sg. Past xot) ' to rise'. (10) Sangl. xaz-: xot- 'to rise '. (11) Par. xist kan- 'to rise, jump'. (12) Wx gzz-: gdzd-(< *han-xaiz-, v. Morgenstierne,IIFL ii 450) ' to rise '. The divergent meanings of Oss. xezun are unlikely to have evolved from that of a single base. On first consideration at least three different verbs seem to have come to coincide in Ossetic. One, meaning 'to pasture ', has no equivalent among the other Iranian forms quoted above. The second verb will be represented by the meaning 'to climb, pass over'. This agrees with the notion of' rising ' and other movements (varying according to the preverb) which inheres in some of the forms quoted from other Iranian languages. However, the meaning' to creep ' of MPers. and Sogd. xyz-, NPers. xe;zTdan, xazTdan,which Henning closely associates with Sogd. pcxyz- ' to kneel ', is not found in Ossetic, and may belong to a base distinct from the ' rising ' one-; cf. below on the Sogdian evidence. The third verb is seen in the meaning ' to wait '. This need not be separated from the notion of 'protecting ', as comparison with Germ. warten shows, cf. OSax. wardon 'protect, watch, look after, be on one's guard', Engl. ward, and the LWs French garder, Ital. guardare (Kluge, Et. Wb. d. deutsch. Spr.7,

492

I. GERSHEVITCH-

postulated an original meaning ' auf jem. oder nach jem. schauen').1 It seems to me that the words quoted under (2) (a), (3) (i), (k), (4) (g), and (8) (f), belong to the Oss. 'waiting and protecting' base, rather than to the verbs meaning to rise (etc.) '. Bearing in mind the distinction suggested by the meanings of Oss. xezun, we may proceed to examine the variation in the root-vowel of the forms collected, assuming for the moment that the initial consonant was OIr. h-. Dig. xez-/xist- and Khot. h;ys-/°hdst- point to OIr. *haiza-/*hista-. Kabul Pers. xest-, Par. xTst may have been influenced by the Pres. xe/Iiz-; MPers. x(')yst- may be explained in the same way, unless it stands for xist< *hista-. On the other hand, the natural derivation of the Sogdian forms quoted under (4) (a)-(e), is from *haz-/*hast- (Chr. °x/yyz-, B. °y'yz- < *hdzaya-). It is therefore likely that in Middle Iranian one of the two past stems, xistor xast-, was built analogically on a secondary present stem. The present stem with e/i cannot be secondary 2; hence, if we are dealing with a single OIr. base, its form must have been *hiz-, of which MIr. xdz- appears to be a secondary development. The normal reduplication vowel of i bases is i. There is, however, one solitary example, attested three times in the Avesta, of an i base reduplicating with a, viz. daisy- from dz- 'to look'. Since obviously something exceptional has happened to our base, let us assume that it sufferedan exceptional reduplication of the form *hahiza-. Comparisonwith the preverbs pati-, pari-, and abi-, which normally lose their i before single consonant in Middle Iranian, suggests the possibility of *hahiza- having become *hahza- and further either xaz-, or, according as the loss of the second h entailed or not compensatory lengthening, xaz-.3 Acceptance of this theory commits us to recognizing h- as the initial OIr. consonant of the base which meant 'to rise (etc.) '. Beside Sogd. °xaz-: °xast- we find prxyz-, txyz, °xyz-, and pcxyz-, all apparently from *haiza-, and B. pcyst- from either *hist- or *hast-. The possibility that at least in °xyz- and pcxyz-/B. pcyst- Sogdian has preserved a distinction between two originally different bases,4 can be contemplated if pceyt- represents *hista- and NPers. xazzdanan older *xizzdan (cf. zam ' winter < zim, Horn, GIP i2 20 sq.). The verbs meaning ' versari, protect, wait' have been separated above from the ' rising (etc.) ' and ' creeping' ones. If we reconstruct for their sake a base *xiz-, we shall avoid recognizing in Parth. and Sogd. prxyz-, MPers. phryz-, and NPers. parhez-, the same irregular treatment of OIr. h after i as we find in Khot. pahiys- and bihTys-from *hiz- 'to rise'. 1 Cf. also NPers. (etc.) pay- ' to protect ' and ' to wait ', Horn, Npers. Et., 63; here, however, the difference in meaning may reflect a difference in origin, v. Hiibschmann, Pers. St. 37, Henning, ZII ix 203, 214. 2 OIr. *haz(a)ya- would have had palatalized z in Khot. and Sogd. 3 For the absence of palatalization in Sogdian, cf. zrywn ' vegetable' < zairi.gaona-. 4 Viz. *hiz- ' to rise' (Sogd. °xaz-/°xact-) and *2hiz- or *xiz- ' to creep' (Sogd. xez-/°xiSt-).

ANCIENT

SURVIVALS

493

IN OSSETIC

As to the Avestan word, we are not bound to consider all variants except 02 as incorrect forms or spellings of what ought to have appeared as either pairi.saeza- or *pairis.haeza-; without prejudice to the reality of pdiri.haeza-, pairi.xaeza- appears to be an irreproachable formation from a base'*xiz-. In Avestan, pari- may have been used with both *hiz- 'to rise (etc.)' and *xiz' to stand around, versari, wait, look after '; whatever was at first intended to stand in the Vend. passage (in which the difference between 'arise and go around' and 'arise and stand around, be about' does not amount to much), the copyists may have confused and contaminated pairi.xaeza- with *pdiris(h)aeza- and the simple *haeza-, thus making it impossible for us to recover the original reading. If we now return to Oss. xezun 'to pasture ', it appears from a comparison with NPers. caridan 'to pasture' (etc.) beside Av. cara- 'versari ', Skt. car' to move around ' and ' to pasture ', that this meaning is compatible with that of the base *xiz- as postulated for Oss. xezun ' to wait' and 'to protect '.1 In conclusion we may attempt to sort out the available evidence as follows: (i) OIr. hiz- 'to rise (etc.)'. (1) Pres. haiza-. Av. °haeza-, later hiz-, xe/iz-, 2b, d; 3c, d, e, f, g, h; 4f (this possibly under ii, 1); 5 ; 6; 7a, b; 8a, c, d (?); 9; 12. (2) Pres. xaz- (y(a)z-) < *hahiza-, 4a a (perhaps also ,), d, e; 10. (3) Pres. xdz- < *hahiza-, 2c; 4a ,8; 10. All doubtful; 10 may have suffered secondary lengthening. (4) Past h/xist-, 5; 7a, b; possibly 3b, c, d. (5) Past xast- (built on xaz-), 4b, d, e; possibly 9. (6) Past xast- (built on xdz-, or lengthened from xast- on the analogy of other Past stems, cf. Av. rasta-, tasta-), 2b; 8a; possibly 9 and 10. Past xe/lst- (built on xe/z-), 8b; 11; perhaps 3b, c, d. (7) OIr. 2hizor 2xiz- 'to creep '. (ii) (1) Pres. xez- < *h/xaiza-, 3a; 4f (but cf. i, 1), h, i; 8d. (2) Pres. xaz- < *h/xiza-, 8e. (3) Past xist-, 4i. (iii) OIr. 'xiz- 'to wait, versari, look after, protect, pasture'. (1) Pres. Av. °xaeza-, later °xez-, 2a; 3i, k; 4g ; 7c, d, e; 8f (cf. Horn, GIP i2 95).

(2) Past °xist-, 3i; 7c, d, e. (3) Past °xast- (borrowed from i 6), 2a. *ui-kurIron guirin,Dig. igurun 'to be born' can hardly go back to anything but *ui-kur-. It is tempting, .therefore, to connect it with Sogd. wk(')wr, which occurs in two passages. In the colophon of P 8 the text is said to have been ' 1 This view seems ' preferable to deriving xezun to pasture from OIr. *xdzaya-, denominative of *xaza-, cf. Khot. khays- ' food ' (< *xdd-s-, v. Bailey, BSOAS x 598), Parth. x'z- ' to devour ', NPers. xdaydan ' to chew' (Ghilain, Essai, 59). VOL. XIV.

PART 3.

34

494

I. GERSHEVITCH-

written 'for the benefit (v. Henning, BSOAS xi 736 sq.) of the whole family (knt kwtr, v. Benveniste, JA 1951, 113 sq.), the wk'wr ywtyyw§tyt, the dead and the living, the near and the far ones, etc.' Since wk'wr ywtyywstyt is surrounded by asyndetic pairs of collective nouns, there is a presumption that 'wk'wr and friends' rather than ' wk'wr friends' was intended. In this case wk'wr, too, may be a collective, meaning 'relatives by birth, kinsfolk, gens'. The abstract wkwry', which Henning, l.c., quoted from the other passage, could then be understood as 'kinship ', or, more vaguely, 'familiarity'. daDig. radtun (Iron radttn), wedtun,dasdtun(Iron dasdttn),tastun,all meaning 'to give', appear to have developed from OIr. dada-, cf. Miller, Ossetisch,66. radt- may be fromfra-d(a)da-,1 and sdt- from a-d(a)da- with shortening of the preverb (cf. Miller, op. cit. 81 n.); dadt- is perhaps a contamination of an earlier *dasd-with radt- and adt-. The Past Iron radton2 presumably has -dt- < *-dat< -data-. The base da- also seems to have survived in Dig. fedtan (Iron fidtcn) and fedton (Ironfidton) ' payment, compensation '. The Dictionary understands the former as 'what is to be paid ', in agreement with the gerundive function of other derivatives in -en from Present stems (cf. Miller, op. cit., 93 sq., Abayev, Pycciio-ocemuHcKuui cJosapb, 593). Since -on, too, forms nouns from Present stems, it is legitimate to seek in fedt- the survival of an alternative Present to fed- in Dig. fedun (Iron fidtn) 'to pay '. The two stems are best explained as representing OIr. -d(a)da- and -d(a)- respectively. In fe- we may have the result of pati-d- (cf. Asia Major (New Series) ii 138),3 so that the original meaning will have been ' to give in return '. The semantic development and the forms recall *para-dd-, which is widely used in Eastern Iranian for 'to sell '; there, too, the Present goes back sometimes to -d(a)-, sometimes to -dada- (in Sangl. to -daya-), v. Morgenstierne,IIFL ii 237; for Sogdian cf. GMS § 122; Professor Bailey has shown me the Khot. Present stem pirdth- in his Khotanese Buddhist Texts 4160-61,

44178,

Past para (< pardta-) 4172.

The Past stem offedun, fist- (Ironfist-) cannot be separated from Yd. p9rZst-, Mu. p9r7st-,and parust- 'to sell'. Morgenstierne, I.c., derives the latter from *pardsta-, which he considers to have been built on the Present *parad-, as if this were a simple verbal stem. Alternatively one may recognize in Oss. and Yd.-Mu. -st- an ancient survival corresponding to Skt. -tta-, v. Wackernagel, 1 Or from *rand- ? Cf. Wx rand- 'to give', Morgenstierne, IIFL ii 537, and Oss. bsedtun > band-. 2 The Dig. Past laevasrdton,which belongs to fra-bar-, should be added to the forms quoted, JRAS 1946, 181, n. 3. 3 Or does fe- merely represent pati- in special circumstances ? fe- may correspond to an irregularMPers. pdai- infedis (Ironfidis) ' blame ', if MPers. p'dys'gyh' slander ' can be compared; the latter was derived by Henning from *pati-dis-, v. ZII ix 229; either meaning may have developed from ' stigmatizing '. In festeg ' pedestrian ' < *pastika- (v. Morgenstierne, NTS xii 267) no original long vowel seems to be involved.

ANCIENT

SURVIVALS

IN OSSETIC

495

Ai. Gr. iil 98, cf. also Av. paityJsti-, paitydstar-. The vowel of fist- must in either case be secondary, adapted to the i of other Past stems that have e in the Present, cf. Miller, op. cit., 60 sq. INDEX

Skt. Av. Av. Av. 0. 0. 0. 0. 0. 0. NP. Av. Av. Av. 0. 0. 0. 0. 0. 0. 0. 0. Arab. 0. OP. Av. Av. Av. 0.

abhisama)zsana484 aiwi.sast6 484 a&Ju- 488 n. 1 &yu- 484 ademon- 485 adtun 494 afson 484 wnwuag 484, n. a3nguldzaxta 488 n. 2 xx 488 bah&ne 483 d&- 494 daenman- 485 daicy- 492 dadtun 494 485 faldeman fa3rat 485 n. 1 fmzdon 484 fe/Tdis 494 n. 3 494 fe/idtxn fe/idton 494 fe/idun 494

fthrist 493 iii 2

ftyyau 484 frabar- 494 n. 2 frax§ti- 487 fraxgti.dd- 487 fraxJtya- 487 gawndzexte 488 n. 2

0. gurin 493 Sogd. Oyz-493 i 2 Av.

OhaJza- 493

Av. Av. 0. Khot. 0. 0. Gr. Khot. 0. 0. Khot. Bal. 0. PI. Khot. Av. NP. Av. Av. MP. Sogd. Sogd.

haxa- 488 haxti 488 n. 2 hizun 488'n. 3 OhIys-493 11 igurun 493 imisun 485 Xaxt483 n. 2 khaysa- 493 n. 1 lalixsta 487 1Fevardton494 n. 2 maysiridi 485 n. 3 maz&r485 n. 3 mis8n 485 mzarai 485 n. 3 pada 485 n. 1 paity&sti/ar- 495 parhiz- 493 iii 1 payah- 484 p&yu- 484 p'dys'gy4 494 n. 3 pryy'- 488 n. 3 pcxyz- 490

Y-M.

parfst- 494

Khot. pir&th- 494 Tokh. porat 485 n. 1 Gr.

i7Tj'i 484

0. qlzin 488 n. 3 0. radtun 494 rand- 494 n. 1 Wx. 0. rwuagm 483 0. rwuwg 483 n. 2 0. rwuonx 483 0. sgA 488 n. 1 NP. tabar 485 n. 1 Arm. tapar 485 n. 1 0. tatun 494 Russ. topor 485 n. 1 0. uag 484 0. ualdemon- 485 0. u(o)din 487 n. 2 0. u(o)zin 487 n. 2 Av. 3vah- 483 Sogd. wk(')wr 494 MP. wmylh- 486 0xaJza- 493 Av. NP. x&yidan 493 n. 1 Parth. x'z- 493 n. 1 Pers. (etc.) oxjz- 493 0. xizt'n 491 Sogd. xz- 492 Av. yaod- 487 n. 2 Av. yaoz- 487 n. 2 0. yau 484 zmarai 485 n. 3 PI.