Studia Philologica Iranica: Gherardo Gnoli Memorial Volume 9788866871156


121 36 19MB

Italian, German, English, French Pages 467 [152] Year 2017

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Studia Philologica Iranica: Gherardo Gnoli Memorial Volume
 9788866871156

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

ISMEO

.UNIVERSITA DEGLI STUDI DI NAPOLI "L'ORIENTALE"

ASSOCIAZIONE INTERNAZIONALE DI STUDI SUL MEDITERRANEO E L'ORIENTE

DIPARTIMENTO ASIA, AFRICA E MEDITERRANEO

ISMEO-ASSOCIAZIONE INTERNAZIONALE DI STUDI SUL MEDITERRANEO E L'ORIENTE

SERIE ORIENTALE ROMA FONDATA NEL 1950 DA GIUSEPPE T UCCI DIRETTA DAL 1979 DA GHERARDO GNOLI

Scientific Board: Timothy H. Barrett, East Asian History, School ofOr. and African Studies, London Alessandro Bausi, Äthiopistik, Asien-Afrika-Institut, Universität Hamburg Peter Komicki, East Asian Studies, Cambridge University Daniel Potts, Ancient Near Eastem Archaeology and History, Inst. for the Study ofthe Ancient World, New York University

STUDIA PHILOLOGICA IRANICA GHERARDO GNOLI MEMORIAL V OLUME Edited by Enrico Morano, Elio Provasi and Adriano V. Rossi

Editor: Adriano V. Rossi

NUOVA SERIE Vol. 5

ROMA ISMEO 2017

ROMA SCIENZE E LETTERE 2017

This volume was published with a grant J,-om the: - PRIN 2009 Project (2009JHSEE7) "Sedi del potere, comunicazione politica e societa nell 'Iran achemenide: ricerche e studi archeologici ed epigrafici ", directed

at L 'Orientale University by Adriano V. Rossi;

- MIUR Project "Studi e ricerche sulle culture dell 'Asia e dell 'Africa: tradizione e continuita, rivitalizzazione e divulgazione ".

TUTTI I DIRITTI RISERVATI

ISBN 9788866871156

© 2017 Scienze e Lettere S.r.l. Via Piave, 7 - 00187 Roma Tel. 0039/06/4817656 - Fax 0039/06/48912574 e-mail: [email protected] www.scienzeelettere.com © ISMEO Associazione Internazionale di Studi sul Mediterraneo e l'Oriente, Roma www.ismeo.eu © Universita degli Studi di Napoli L'Orientale, Napoli www.unior.it Layout by Beniamino Melasecchi

Gherardo Gnoli in a photo from the early 201 Os.

CONTENTS

Preface by E. Morano, E. Provasi and A.V. Rossi ..................................

1x

M. Alram, Ein Schatzfimd des Hunnen-Königs Mihirakula .. . . . . .. .. . .. . . . .. . G . Asatrian, Middle Iranian Lexical Archaisms in Armenian Dialects H:R. Baghbidi, Three Etymological Notes .. . . . ..... . . . . ..... . . . .. . . .... . . . . . .. ........ . / C.G. Cereti, A Short Note on MHDA 38 .. . . . ..... . . . ...... .. . . . ........ .. . ........ .... . . . J. Cheung, On the Origin of the Terms "Afghan" & "Pashtun" (Again) C.A. Ciancaglini, Phonology, Etymology and Transcription Issues of Middle Persian Final Sequences and " _h..i .J � jl � (ed. SOTUDEH 1962, p. 72).

7 Probably the mausoleum of Jäbir al-An�ärT (11 th century CE), the son of the Herat poet Abdalläh al-An�ärT, cf. ÜIBB III, p. 591, fn. 209 . 8 �.J 0W�I � Jt:s,i r"l.:. �I ü-o Ä..ij\..l., � �.ß 0'11 4-:i.J ;;�I �..i.., uL � -.::..ul.S .J �l.S .)1 \...i_)L....i � 0W�I 4 � �.J .... .J 0� ofi � _».i$JI �.J J,i)=JI tlb.i �JiSI .J �_,§ �� .J yl.....:u .J J� .J � � � .J utl...ß .)1 lib..J � .J "4J.J�I .J\..,iS ü-o U"4c. �I � �W�I J,ic.L.wil &Jill �.JIJ �� �Y.J J,i.;JI � �.J �lil � liJI.?. ü,i.,,,. liS .J 0W�I � � (ed. DEFREMERY, SANGUINETTI III, pp. 89 f.). 9 A centmy later, his grandson Bahlul Khän succeeded in establishina the Lodi dynasty in 0 Northem India. 10 4J l;.A 0UJ....JI .l..t:icl! l+:ilc. .J� � � J� ü__,.iSL....i �.J 0W�I .i....._,il � .J y� .i......Jt:s,i � .UI rW a..i)4./ 0W�I Ü-0 °.J�.J (>" .)c. 1� 01 .i..!� .)1 '-,iiS.J (ed. DEFREMERY, SANGUINETTI III, p. 362.).

35

Johnny Cheung

On the Origin ofthe Terms "Afghan" & "Pashtun" (Again)

In barely 300 years Afghan settlements have greatly expanded, both north (into Kabul Province 11 ) and south (into India) of their homeland around the Sulaymän Mountains. This may well be ascribed to their legendary, independently minded, military prowess tl1at is so abundantly clear in later centuries. Several alleged ancient attestations of"Afghan" from the early Persian period, as cited by HABIBI 1968 (and mentioned in most Afghan websites), are obscure. A person with the name ABfAN (in Greek script, transcribing Pahlavi 'dwk'n) is mentioned in a rock inscription dated to the mle of the Sassanian king Sabühr I. From the inscription, it appears that ABfAN is the father of a Sassanian dignitary, Gundafarr, at the court of Shäbuhr I. SPRENGUNG (1940, p. 411) was the first scholar who connected this name to"Afghan," without any further contextual cor­ roboration. GIGNOUX (1986, no. 157) has not committed himself to any interpre­ tation, hence: "Etymologie inconnue." The circumstances of this attestation are unknown, as the inscription does not allude in any way to this person's ethnicity or origin, nor is the transcription fully reliable. Early poetical references may also be cited, although they are quite rare, some ofthem are rather dubious. Tue passages in which the name"Afghan" is supposedly found in FirdausI's epic, the Shähnämah (ca. 1000 CE), may be later interpolations: the Shähnämah glossary compiled by WoLFF (1935) does not refer at all to a name "Afghan" (the homonymous afgän 'wailing, moaning' is attested though). Among FirdausI's (near-)contemporaneous fellow-poets from the same Ghaz­ navid court, the term"Afghan" as a group or tribe has been alluded to in the oeuvre of notably FarmxI and 'UnsürI. The name is attested in a single bayt from 'UnsürI, which was quoted in AsadI's 11 th century Persian Dictionary (ed. IQBÄL, pp. 162 f.):

Again, the general traits of the Afghan people as described earlier, are quite well encapsulated in these two verses. The name "Afghan" itself does not appear to be a genuine, inherited Pashto fonnation. But would it be possible to provide an explanation or analysis of this formation in the first place? lt has a very Iranian "appearance,"13 anyway. lt does not resemble to any of the geographical references (ethnicity, place name etc.) in the earlier Hellenistic geographical manuals (such as Ptolemy, Strabo) that describe this region and the surroundings. This suggests that the name could have been an ad hoc term that was coined locally, i.e. by another Iranian-speaking people, pos­ sibly as a"nick-name" or provisional appellative, to their neighbours. In this case, arguably the best candidate to have "coined" the appellative would have been the Bactrians, the most powerful and culturally most advanced people of the region. In a recent discovery, a similar name has indeed been found in four small Bactrian fragmentary documents, aßayavo (4x), aßyavo ( l x), cf. SIMS-WILLIAMS, BD II, pp. 19, 90, 122, 144, 181a, which would make it the oldest attestation of"Afghan," viz. late 4th century CE. (ibid., p. 19). One of the recurring formations in those Bactrian documents are suffixed with -yavo (next to -avo, -Kavo, SIMS-WILLIAMS, BD II, p. 193), which denotes patro­ nymics or family names. Mechanically, the preform could go back to a formation such as *apä-klgäna-, reflecting the (residential) origin of this (family-, clan-) name. As aßayavo is more prevalent than aßyavo the middle vowel may have been originally a long *ä, which, in unstressed position, becomes shortened a and only in (much) later documents we may expect syncope. 14 An originally short, unstres­ sed, vowel would have disappeared in such a (trisyllabic) position, at an early stage in Bactrian, cf. *abi > aßo 'on, at, to' (with quasi-word-divider o). The most plau­ sible formation is *apäka-, abstracted from the old instr. case *apäkä 'in the di­ stance, far' of *apänc- 'being behind, off', whence aßayavo < *apäka- + äna-suffix (of belonging, ethnicity). This form is attested in both Avestan and (Vedic) Sanskrit, cf. Av. apaxtara- 'backward, northem' (with tara-suffix), Skt. (Rig-Vedic), instr./adverb apäkd 'far, distantly,' apänc- 'going, situated backwards,

34

The king of the world went on from Ghazna to rush upon the Afghans and the pagans of Kahbar. 12

The Ghaznavid court poet FarruxI cites afgänfyän (the plural form "Afghans") in a single verse from a long qa{idah that praises the exploits of sultan Mal1müd (ed. DABIRSIYÄQI, p. 42): ü

.) y.,.

.. .. -ß- (sometimes represented by ...J in manuscripts,21 but mostly inferred from the altemation in the spelling, .w /f/, _, /w/ and y /b/), e.g. zafan I zabän 'lan­ guage, tongue' (standardized: u�J), nibistan / niwistan (standardized: ��). ii. postvocalic -g- > -g- (t), e.g. ägös / ägös 'embrace' (standardized: ut_,c.i), sagäl I sagäl 'jackall' (standardized: J�; ultimately < Skt. srgäla-). 620 CE. Even the alphabetic nature of the Graeco-Bactrian script was accurately described by Xuanzang: "Their language and manners are slightly different from those of other countries. In their alphabet there are twenty-five letters [i.e. the 24 letters of the Greek alphabet plus the in­ vented p for /s/], by which various words are formed to express all things. Their writing is hor­ izontal from left to right, andtheir records have been gradually increased untilthey exceed those of Suli [= Sogdians] in number." (translated by LI RONGXI 1996, p. 32). 19 § ittfiIE�Hi+ 3i EI ft1=1Gtu* �giTt�ijs, o X fili :itf1ßijHijrJlt�]] o X W.it 1t{IJ§JJ:-f @;/ o (ed. online: http://www.suttaworld.org/gbk-txt/sutra/lon/other50/2053/2053-05.htm). 20 As I cannot claim to be an expert in Middle Chinese phonology, I have relied on the re­ search of two eminent linguistic (palaeo-)sinologists, :viz. BAXTER 1992 and PULLEYBLANK 1991. Their transcriptions have been quoted here side by side. 21 The use ofthree *e (= Ps. e), but . still early enough to join tht: i-epenthesis afterwards. Finally, the Sakaraukai correspondence of Khotanese pruha- 'dew' could be the source of Pashto parxa. Khot. -h- may well have been adapted to fricative -x­ at this stage of Pashto.

44

On the Origin of the Terms "Afghan··· & "Pashtun" (Again)

Johnny Cheung

A potential source of old borrowings into Pashto could have been Greek. lt seems that its impact on the Pashto language is minimal though. Despite the adop­ tion of the Greek alphabet, and cultural influences, even Bactrian is relatively de­ void of Greek loanwords, apart from certain commercially pertinent terms.3 ° For Pashto it seems that the following two forms have been borrowed, viz. mecan f. 'hand-mill, quem' (< ultim. Gr. µ11xav11), and macn6yza (mac6yna, macl6yza) 'sling' (< ultim. Greek µayyavtKa), on which see MACKENZIE 1989, p. 461, and MoRGENSTIERNE 1929, p. 200. lt suggests that in those (mountainous) areas where these Iranian-speaking peoples of the steppes settled, no Greek settlement was pre­ sent. Perhaps, through trade with the Bactrians (rather than with genuine Graeco­ phone residents), these advanced apparatus may have entered Pashto, although the Bactrian correspondences have not (yet) been attested. also have been Prior to the invasion the speakers of prehistoric Pashto would in Central ages langu in touch with the two (politically or culturally) important wing borro of ces instan Asia, viz. Khwarezmian and Sogdian. We may have two from Khwarezmian: lated and com­ - raxa f. 'envy, jealousy, hatred.' A preform *araxa- has been postu with Av. ara­ 1exion "Com pared with Khwar. 'rx 'id.' MacKenzie remarks that a ska-, Pers. rask, etc. is unlikely" (NEVP, p. 71 ). Khot. pruha- 'dew' - parxa f. 'dew, hoar frost; fine rain' < Khwar. *pwrx' or be all regular conti­ t canno (*parswä-)? Although clearly related, the formations reflects a different bly nuations of a Proto-Iranian fonnation. Av. araska- proba to which the re­ f.), 206 pp. ablaut fonnation *(H)rH-as- (cf. MAYRHOFER, EWA I, wed in several borro been lational suff. *-ka has been added. The Avestan form has Av. ara.fiia1J,t­ and ' 'envy Iranian languages. The Khwarezmian form, Pers. rask l *s. Un­ palata with stem t 'envious' on the other hand may go back to the varian ywx f. cf. s, * l palata a to like Pashto -x-, the Khwar. fricative -x- can go back r. 'rx Khwa from ord loanw 'ear' (< *gausä-). Therefore, Ps. raxa is probably a can -xr. Khwa As . parxa to (< * rsya- ), cf. Skt. 'ir,$ya f. 'envy.' This may also apply rez­ Khwa from wed borro go back to a palatal *s, this formation could have been an context (yet) ? mian, despite the fact that it has not been found in a Khwarezmi -h- for fricative with ese, Altematively, parxa is a borrowing from Saka/Khotan x- in early Pashto (see above). an as well: Pashto appears to contain one certain borrowing from Sogdi The connection 99). p. P, - xax, xax 'hard, stiff' < Sogd. (Man.) fr- 'stiff' (NEV 409. with Sogd. has already been made by BAILEY, DKS, p. 30 Such as, dating (e.g. cSriµmp1yavo name of a month, after the Greek goddess Demeter, x}:>ovo '(calendar, regnal) year'), cunency (cStvapo 'dinar' < Gr. 8rivap10v, 88paxµo 'drachma' < Gr. 8pax�nl) or goods (mporiyyo '(cloth) made of1inen or cotton' < Gr. ßucrcroi;), cf. SIMS-WIL­ LIAMS 2002, p. 228. The Bactrian title xoacSrio, xo88rio is obviously a calque on Greek aurnKpatwp, but not necessarily a direct translation.

45

F�nally, we may even consider one specific borrowing from Middle Persian/ Parthian: � hag m. 'pig'. This form is probably from Middle Persian or Parthian hwg [hüg] '1d.' (cf. New Pers. xük). A genuine Pashto form would have been twag or tway. Th� fom1 could well have been passed 011 during the (early) New Persian period, for 111stance as a mere "regional," dialectal word, or even via Bactrian as a tradable commodity. The number of (clearly) "Middle-Iranian" borrowings in Pashto appears to be small, but at this stage pre?ist�ri� (P�oto-)Pashto may well have adopted many _ more loanwords from 1ts lmgmstlc s1sters. They would have participated to all t�ose l�ter s�und devel�pments t? at make them almost indistinguishable from gen­ umely mhented format10ns (or, 111 the case of Middle Persian/Parthian from later New Persian borrowings). They do, however, attest to the earliest his;ory of con� tacts �etw�en (Proto-)Pashto speakers and their Middle Iranian cousins, prior to the m1grat10n of (Proto-)Pashto speakers into the Hindu Kush region . Of course, . one has to cons�der �hat the documentation of many Middle Iranian languages dis­ covered so far 1s still relatively limited in scope, for instances, references to do­ mestic affairs are few.

Conclusions The relation between "Afghan" and "Pashtun" can thus be differentiated as follows: the former term is derived from a geographical entity, while the latter has strongly ethnic-cultural overtones. E:_ver since t?e post-Classical (post-Hellenistic) era, the people around the Su­ layman mountam range are better known as the "Afghans." lt is first attested in seve�al Bactrian frag�enta:Y documents �late 4 century CE) spelled as aßayavo an� 111 a late �kt. treat1se of the astronomer Varäha Mihira (611i century CE) the Av­ . aga�a 1s me�t10ned: Moreover, it is referred to as ß6J1lf {J! *?a-bak-gi:anh/*?a-phak­ ken 111 the b1ogr�phies of the �hinese Buddhist scholar Xuanzang (629 CE). In the Islam1c era, the des1gnation "Afghan" is first mentioned in the 11th cen­ tmy CE, n?tably by the anonymous compiler of the ]fudüd al-'Älam and by the Khwarezmian polymath al-BirünI. lt is also referred to during the rule of the Turkic dynasty of the Ghaznavids, when several, famous Ghaznavid bureaucrats such as 'UtbI, BarJJaqI and GardezI were employed at the court. The Afghans ar� histori­ c.ally,,aep1cte� as fi.e:cely independent from any central authority (being "rebel­ h?us ), showmg m1htary prowess and making a living as either mercenaries or as h1gh:-7ay robb�rs/thieves. Another important observation is their deep, symbiotic . relat1?nsh1p w1th the �pagan) Hindus or Indo-Aryan groups, who are frequently mentloned together w1th th� Afghans. This general characterisation still holds to a certain degree, until today. The name "Afghan" looks very Iranian, but it cannot be explained within Pashto. The name may have been coined by their culturally, more advanced neighth

Johnny Cheung

46

äka- äna ­ yavo we may posit a formation *ap b ours, the Bactrians. For Bactrian aßa - 'being äka p a nc-l* ä ia),' derived from * ap 'People from a distant land (behind Bactr gän, rea ß a * form veloped from an older b ehind,off.' Pers. afgän w ould have de flecting Bactrian aßayavo.

ng the linguistic steppe peoples, includi The arrival of several North Iranian ere istan must have taken place, somewh ancestors of the Pashto speakers,in Afghan E IERN NST RGE Mo f. c , s tu odo the rnle of Di in the 2 d century BCE, if not during the P ashtuns with sible theory to connect 1979, p. 29: "lt is, a priori, a very plau c., into Sakastäna t e t, a r e H a enetrated, vi the Sakas, who in the 2nd century B.C. p 173). ,292 ff.; MORGENSTIERNE 1940, p. (Seistän)" (echoing TARN 1938, pp. 284 its actors) was e so that this event ( and The date is close enough to Strabo's tim bo had most likely ugh and recorded accordingly. Stra still remembered well eno iled near or on the "spot" ess reports that were comp access to Greek (Seleu cid) witn ports are therefore om the steppe . Those re of the invasion by the Iranian hordes fr ese predecessor, s the reports by his Chin more detailed about the invaders, wherea earsay, sed on an outsider 's view and h the historian Sima Qian, seem more ba most whom the Chinese would have been biased towards those "Barbarians" with familiar, viz. the Yuezhi/Tocharoi. u ced in this o mu st have be en int rod In the aftermath of the invasions, P asht , the non­ inly a rt ce m, o fr s r e the Greek-Seleucid settl area, after the expulsion of Greek loanwords s. The relative lack of urban centres in Bactria and its surrounding peakers. ese prehistoric Pashto s confüms the la te arrival of th as the linguis ­ rabo may be considered The invading Pasianoi mentioned by St ct that P ashto fa e also corroborated by th tic ancestors of modern Pashto. This is O ssetic and ," s r e t r "steppe sis oglosses with it s forme shares several exclusive is mmercially co d an Sogdian (two culturally Khotanese, even with Khwarezmian and Khotanese d n a c ti Eurasian steppes). O sse important Iranian languages used in the Asioi and e th t a th xy" of the closely related languages can be considered as a "pro , spoke rces u o s l ca entioned in the Classi Sakaraukai, the other invading tribes m respectively. ve conquered e, the Tocharoi m ay ha Finally, the most powerful invading trib in the region. n, a ctri a its language, Irano-B Graeco-Bactria and ultimately imposed g settlers eakin p s kee -Bactrians and Gr Genuine social interaction between the Irano Greek few ly e tiv a l e r e reflected in th must have been fairly limited, which is also most, e th At n). o gi e r n languages in the borrowings in Bactrian (and other Irania h ac­ c whi d, e y o mpl e king scrib es may have b een s ome rem aining, Gr eek-spea eek Gr e th d n a n a tri ac B e Gr eek script for writing counts fo r the adopti on of th the s a h c u s , s m e it l a i rc f c ertain comme names in the names of the date . Nam es o as k ee Gr m o fr ed w o borr r) and goods, have also been curren cy ( drachma and dina well.

On the Origin of the Terms "Afghan " & "Pashtun " (Again)

47

ABBREVlATIONS

. , . . . .Av. = Avestan; instr. = instrumental (case)· Ba t. = B actnan,.Bart. = Bartang1;f. = femmme; Kak. = Kakari Pashto (dialect)·' Kh,war. � Kh wareznuan; Khot. = Khotanese·, . . . n ,. Pers. = Persian; · haean,. 0 ss. = Ossetic· . .p ers1� Lat. = Latin·, Man. = Mamc , OP = Old . . . · Skt. = Sansknt; Sogd. = Sogdian; Plr. = Proto-Iranian; Rosh. = Roshani·, Sanq._ = Sanqoh, . Shugh. = Shu hni·, Wan. = Wanets1 Pashto (dialect); Waz. = Waziri Pashto (dialect); Yazgh. = Yazghulamf

11

REFERENCES M.G. ASLANov· Afgansko ru� k' lovar '[AD . ghan-Russian dicti?nary]. Moskva 1966. H.W. BAILEY: "Irano-Indi�a „ . Bulletin of the School of Orzental and African Studies 13/1 (1949), pp. 121-139. -::oa�a.': In: Kho:an�se Texts, Vol. V II. Cambridge 1985 pp. 110-142 As1�m �n? Pasiam." In : Bulletin of the Asia Institute 1'(1993), PP . 9·-10· DKS. Dz_ctwna? of Khotan Saka. Cambridge 1979. .. . . The Age oif the �teppe Warriors. C. BAUMER. The Hzstory of Central Asza. London 2012. . W. BAXTER: A Handbook oifOld Clunese Phonology. Berlm 1992. . . . .. C.I. BECKWITH: Empires of the Silk Road a Histo y of Central Eurasza jrom the Bronze , r ' . Age to the Present. Princeton 2009 e 1g by t�e.S�a, an Hwu� Li. London 1911 [repr. of 1888]. �.:���;� :a:h!:�7::t:.;�at Samhzta, -itith � Englzsh Translation Exhaustive Notes and Lz'terary Comments. 2 Paiis. Delhi 1981. BrnD�LPH: Afghan Poetry of the Seventeenth Century. London 1890. . AROE. The Pathans, 550 B.C.-A.D. 1957. London 1958 . J. CH�UNG: "Selected Pashto Problems II. Historical Phonol�gy 1. On Vocahsm and Etyma I. In: Iran and Caucasus 15 (2011), pp. 169-205. . · · --EDJV· l0!1·�al n _::. _ · Etvm 0/ the Iranian Verb. Leiden -Boston 2007. . . zctionary M . DABIRSIAQI: Dzwan-z Hakim-i Farr - s- t- 1- [ he Divan of master FairuxI SistänI]. T ehr�n 1335 [= 1957 CE] [repr. 134�� {�;; CE] T . - - u- tz· 'l-Hmd DAlRA:ru L-MA'ARIFI. , L-ÜSMA.NlA . . Kitabfi n hqzqor Al-B'irünf 's India (Ara. ; � bic Text)' an Account of the R�ligion z o zt rature, Geography, Chronology, . Astronomy, Customs, Laws and As;.olo ) ;fj, ; ; 1030 A.D. Hyderabad 1958. J. DARl\tffiSTETER: Chants populaires des Afffha,;s. ;a�/1[8u;. C. DEFREMERY, B.R. SANGUINET1T Voya es d'Ibn Batoutah, . texte arabe, accompagne d'une traduction. 4 Vois. Paris 1853-185{ A. DEGENER: Khotanische Suffixe. Stuttgart 1989. DKS: see BAILEY, DKS. . B. DORN: History ofthe Atg' hans, Tr, anslatwn of Makhzan-e Afghani. London 1829. EDIV: see CHEUNG·, EDT/!,.' EVP: see MüRGENSTIERNE, EV:P. r . . H.A.R. GIBB III: The Travels of Ibn Battüta A D 13 5-1354, Translatzon wzth Revisions

1t 1f:.

==

g�

_?

�/i

if,

D , 2 and Notes from the Arabic ·Text ed .b . C e and B.R. Sanguinetti, by H.A.R. Gibb. 5 Vols. Cambridgel958 [rep;, i�ndon 2�o�f?

48

Johnny Cheung

PH. GIGNOUX: Iranische Personennamenbuch. Band II: Mitteliranische Personennamen. Faszikel 2: Noms Propres sassanides en moyen-perse epigraphique. Wien 1986. A.H. HABIBI: "Afghan andAfghanistan." In: Afghanistan 22/2 (1348 [= 1968 CE]). Online: http://www.alamahabibi.com/English Articles\Afghan_and_Afghanistan.htm, 1968. H. HUMBACH, S. ZIEGLER: Ptolemy Geography, Book 6, Middle East, Central and North Asia, China. Part 1: Text and English/German Translation. Wiesbaden 1998. H. Hu:MBACH, K. FAISS: Herodotus's Scythians and Ptolemy's Central Asia; Wiesbaden 2012. ' A. IQBÄL, ed.: Kitäb-i Lugatu 'l-Furs [Dictionary of Persian]. Abü Mam;ür 'Ali b. Al;unad AsadITüsI. Tehrän 1319 s. /1940-1 CE. G. LAZARD: La langue des plus anciens monuments de la prose persane. Paris 1963. LI RoNGXI: A Biography of the TripitakaMaster of the Great Ci'enMonastery of the Great Tang Dynasty. Trans!. from the Chinese of Sramarza Huili andShiYancong. Berkeley 1995. - The Great Tang Dynasty Record of the Western Regions, Translated by the Tripitaka­ Master Xuanzang under Imperial Order, Composed by Srama,:za Bianji. Berkeley 1996. J. MARQUART: "Untersuchungen zur Geschichte von Eran." In: Philologus (Suppl.) X 1 (1905), pp. 1-258. M. MAYRHOFER, EWA: Etymologisches Wörterbuch des Altindoarischen, 3 Bände. Heidelberg 1992-2001. V. MINORSKY: ]fudüd al-'Älam: "The regions of the world," a Persian Geography, 372 A.H.-982 A.D. London 1937. G. MORGENSTIERNE: " Notes on Prof. Charpentier'sArticle." In: Acta Orientalia (Batava Danica Norvegica) 7 (1929), pp. 180-200. - 1940: "'Pashto'. 'Pathan' and the treatment of r + sibilant in Pashto." In: Acta Orientalia (Batava-Danica Norvegica) 18 (1940), pp. 138-144. -- "The Linguistic Stratification of Afghanistan." In: AfghanStudies 2 (1979), pp. 23-33. - EVP: An Etymological Vocabulary of Pashto. Oslo 1927. - NEVP: A New Etymological Vocabula,y of Pashto. Compiled and Edited by J Elfenbein, D. N .M. MacKenzie and NicholasSims-Williams. Wiesbaden 2003. W.R. PATON: Polybius, the Histories. Cambridge (MA)-London 1922-1927 [2011, revised by F. w. WALBANK and CH. HABICHT]. E. PULLEYBLANK: Lexicon of Reconstructed Pronunciation in Early Middle Chinese, Late Middle Chinese, and EarlyMandarin. Vancouver 1991. A. QAVIM: Tärfx-iYamfnl[The History ofYamfn].Tehrän 1334 [= 1955 CE]. S. RADT: Strabons Geographika, mit Übersetzung und.Kommentar. Band 3: Buch IX-XIII: Text und Übersetzung. Göttingen 2004. E. SACHAU: Alberuni's Jndia. London 1910. 0. SEEL: M. Iuniani Iustini epitoma Historiarum Philippicarum Pompei Tragi, Accedunt prologi in Pompeium Trogum. Stuttgart 1972 [repr. of 1886]. A. DE SELINCOURT: The Histories, Herodotus. Hannondsworth 1972 [several repr.]. S. SHERWIN-WHITE, A. KUHRT: From Samarkhand toSardis: A New Approach to the Seleucid Empire. Berkeley-Los Angeles 1993. N. SIMS-WILLIAMS: "Ancient Afghanistan and Its Invaders: Linguistic Evidence from the Bactrian Documents andinscriptions." In: Indo-Iranian Languages and Peoples. Oxford 2002, pp. 225-242. - BD II: Bactrian Documents from Northern Afghanistan. Vol II: Letters and Buddhist Texts. London 2007. N. SIMS-WILLIAMS, F. DE Bw1s: "The Bactrian Calendar: New Material and New Sugge-

On the Origin ofthe Terms "Afghan'' & "Pashtun" (Again)

49

stions." In: Languages of Iran: Past and Present. IranianStudies in memoria m David Nei!Mac.Kenzie, ed. D. WEBER. Wiesbaden 2005 (Iranica 8), pp. 185-196 . M. SüTUDEH: ]fudüd al-'älam min al-Ma.friq zla al-Magrib, Compiled in 982-3 A.D. = 372 A.H. Tihrän 1340 [= 1962 CE]. M. SPRENGLING 1940: " Shahpuhr I, the Great on the Kaabah of Zoroaster (KZ)." In: The American Journal ofSemitic Languages and Literatures 57, 4 (1940), pp. 341-429. W.W. TARN: The Greeks in Bactria and India. Cambridge 1938 [repr. 1951, 1980]. W.M. THACKSTON: The Baburnama, Memoirs of Babur, Prince and Empero r. New York 1996. A.V. TRIPÄTHl: Brhatsaf!lhita by Varahamihiracärya, with the Commentary of Bhattotpala. 2 Parts. Varanasi 1968. R. TURNER: A Comparative Dictionary of the Indo-Aryan Languages. 4 Vols. London-New York 1966-1985. W. V OGELSANG: The Afghans. Oxford 2002. K.T. WITCZAK: "The Hittite Name for 'Garlic' ." In: Acta Orientalia Hungarica 59, 3 (2006), pp. 341-345. F. WOLFF: Glossar zu FirdosisSchahname. Berlin 1935 [repr. Hildesh eim 1965]. S. YERASIMOS: Ibn Battüta Voyages, traduction de l 'arabe de C. Defrem e1y et B.R. San­ guinetti (1858). Introduction et notes deStephaneY erasimos. Paris 1982 [contains the . repnnt of the French translation of DEFREMERY, SANGUINETTI 1853-1858]. Zambasta: .Khotanese Texts (edited by H W. Bailey). Vol. VI: Prolexis to the Book ofZam­ basta. Cambridge 1967. E. ZÜRCHER: "The Yüeh-chih and Kani�ka in the Chinese Sources." In: Papers an the Date of Kani$ka, ed. A.L. BASHAM. Leiden 1968, pp. 346-390.

CLAUDIA A. CIANCAGLINI Roma

Phonology, Etymology and Transcription lssues of Middle Persian Final Sequences and in two or three different ways. Schematically:

, and their corres­ ponding transcriptions. 2. as -rag, more rarely as -lag, on the basis ofthe ManMP and/or NP equivalents ofthe word under consideration. • The last time I met Prof. GHERARDO GNOLI, at IsIAO, we talked about my book on the Iranian loanwords in Syriac and he emphasized the relevance of this parallel tradition as a test for the consistency ofMAcKENzIE's transcription ofMiddle Persian. After a year since Gherardo passed away, with sadness and gratitude I dedicate this paper to his memory, which in a way continues our last conversation.

52

Phonology, Etymology and Transcription Issues ofMP Final Sequences dg> and 'complaint, lamentation' < gil- 'to weep, com­ plain' (CPD 36), NP gila 'complaint, reproach.' In all the cases transcribed with final -rag or -lag the MP words involved are derivative fonnations, suffixed with the outcome of the Qlr. adjectival suffix *-ka-, added to a MP stem ending in -r or -!. MP -ag is the most common outcome of Olr. *-ka-; -a- in -ag is part of the Qlr. stem wrongly analysed as belonging to the suffix; in other words, MP -ag < Early MP -ak < Olr. *-a-ka-. 4 Qnly once in CPD is the sequence 'chicken, hen' (CPD 50), NP kark 'id.,' which is an exceptional instance of absence of the voicing of -k after a voiced consonant: this fonn is probably the outcome of an onomatopoetic word, Ir. *krka- 'hen' (cf. YAv. kahrka0 in kahrkäsa­ 'jackal,' lit. 'eater of chickens'). 5

1 Cf. SKJJERV0 2009, p. 201; SlJNDERMANN 1989, pp. 144-145; WEBER 1997, p. 613 with bibliography. I shall treat this complicated topic in a forthcoming volume on the Old and Middle Iranian velar suffixes. 2 The MP phoneme /1/ is rare; PIE */1/ mostly merged with */r/ in IIr., though *III was pre­ served in some diatopic and diastratic variants of Oir. and Oind. In OP III only appears in loan­ words. The MP instances of11/ are either inherited or developments of0Ir. consonantal clusters. See SUNDERMANN 1989, p. 145; WEBER 1997, pp. 615 f.; SKJJERV0 2009, p. 201. For a linguistic interpretation of the diastratic and diatopic distribution of the IIr. isogloss represented by the merger of PIE */1/ and *lrl see LAZZERONI 1968, pp. 105 ff.; 1969. 3 Cf. SKJJERV0 2009, p. 20L 4 There are many other MP outcomes ofthis 0Ir. suffix, more or less widespread and pro­ ductive: see CIANCAGLINI, in press, and below. As regards 0Ir. *-ka- in the Old Iranian languages, see CIANCAGLINI 2012. 5 See MAYRHOFER, EWA 1, p. 388.

53

3. : -rg, -rag As opposed to

As stated above, my claim is that the problem is the transcription ·-rag of the sequence . MAcKENZIE, in his Dictionary, uses this transcription for a group ofderivatives containing the MP outcome ofthe 0Ir. suffix *-ka-, as he does for those considered in the previous section (e.g. cärag 'means, remedy'), which however are spelled with the final sequence as well as -sometimes-· represent the same phono­ logical sequence -rag. The instances ofthese derivatives, having the spelling and transcribed as -rag in MAcKENzrn's Dictionary, are the following:

MAcKENZIE, but also the derivatives (at least those recorded in his Manual), where MAcKENZIE adopts the transcription -rag. lt seems clear, indeed, that NYBERG dis­ tinguishes between and or , and the transcription -rg when the final -g is not the MP suffix continuing 0Ir. *-ka-, but belongs to the stem or represents a different suffix. In particular, MAcKENZIE's transcription -rag ofthe spelling in the group of words like abarag, wastarag etc. seems also conditioned by the lack of a NP continuant that might have helped in deciding the transcription, so we may suppose that he considered these MP words as showing the most frequent outcome of 0Ir. *-a-ka-, namely MP -ag. In fact, in the cases in which the etymon and the formation are known, most of the examples transcribed with -rg by MAcKENZIE do not represent outcomes ofOlr. *-ka-: this is surely the case for gurg 'wolf,' marg 'death,' warg 'leaf' (where -g is part of the stem) and mastarg 'skull' (where -g traces back to a different suffix). But this explanation does not hold for wuzurg 'big, great,' sturg 'fierce, con­ tentious; gross, coarse,' and xwarg 'ember,' which diachronically are derivatives containing another outcome of 0Ir. *-ka-, namely -g, not the more common -ag. Among the MP derivatives the basis ofwhich ends with a continuant /r/ or /1/, we may also think that MACKENZIE in his transcription distinguishes between:

54

aharag 'robber' (CPD 10) < appar 'robbery, theft' < appurdan, appar- 'to steal;' NYBERG (1974, p. 28) has only appar and appurdan; parrag 'in, between, among,' wistarag 'bedding, cover' etc. as MP synchronic derivatives in -ag formed respectively from MP abar 'higher, superior,' andar 'in, among, towards, conceming,' wistar 'bed­ ding, cover' etc. This cannot be excluded in principle, but again it is not proven that the pro­ nunciation is affected. On the contrary, I believe that, when taking into account the lack of any lin­ guistic proof supporting a different interpretation, the evidence provided by the different spellings in the archaizing Pahlavi script, which does not record the MP voicing of Oir. * /p t k/ in intemal and postvocalic position. On the other hand, this same archaizing writing system does record the results of the voicing contact assimila­ tion of an older voiceless stop preceded by /n/ and /r/: this is an independent con­ ditioned change that took place before the voicing of Olr. * /p t k/ in internal and postvocalic position.

HÜBSCHMANN (1895, p. 239) suggests a more subtle diachronic distinction: the older voicing assimilation, already in OP, only involved the stops after /n/, whereas the voicing after /r/ arose later, in the early Sasanian period, or at the end of the Parthian. H()BSCHMANN grounds his claim on the parallel traditions: as regards the voicing after /n/ he quotes examples like Gk. l:apciyym (Hdt. 3, 93), the name of the inhabitants of Drangiana, OP Zranka-; for the assimilation after Ir/ he mentions instances like Gk. apyanEt11� (J r 'a coin, measure' (Frahangxvi: 7; xxix: 12), NP däng 'the fourth part of a dram; a sixth of anything,' from OP *dänaka- 'name of a small Persian coin. ' 20 Here the presence of the final -g in MP and its retention in NP .is due to the early loss of the short -a- preceding the suffix. 21 The fact that the voicing, though recorded in the Pahlavi script, was not yet accomplished in OP is proven by the parallel traditions, which testify a MP previous phase *dän(a)k: see Gk. 8av, and that this voiced allotrope -g, spelled , merged with the out-

56

4. Different Voicing Phenomena: a Historical Survey BACK17 claims that diachronically the sonorisation of the Old Iranian voiceless plosives in intervocalic position is to be distinguished from the voicing after /n/ and /r/, because this latter voicing assimilation took place already in Late OP, whereas the f01mer phenornenon is a MP innovation. The sonorisation after /n/ and /r/ would be a different and independent phenomenon, because it affects only a specific segmental feature in the assimilation process, whereas the sonorisation in intemal, intervocalic position and in final, postvocalic position - according to BACK - is included in a Lautverschiebung which involves many different reduction processes (for example the change of the 0Ir. voiced plosives into glides, and the subsequent changes affecting these glides in particular contexts). These processes are favoured, if not caused, by phonic contexts of a maximum degree of voicing and characterized by different segmental assimilations. 17 BACK 1981, p. 179 and fn. 5; p. 184. See also WEBER 1997, p. 613; SUNDERMANN 1989, pp. 144 f.; SIWERV0 2009, p. 201.

See also Syr. LW 'rgp( 'id.:' cf. CIANCAGLINI 2008, p. 115, with bibliography. 19 See, for instance, MP tang 'nan-ow, tight' (CPD 82, NYBERG 1974, p. 191), NP tang 'id.,' cognate with 0Ind. tafi.c- 'to contract,' from PIE *tenk- 'to contract, coagulate': cf. MAYRHOFER, EWA I, p. 615. 2 ° Cf. HINZ 1975, p. 82. 21 From the same root see, by contrast, the wordMP dänag 'seed, grain' > NP däna 'id.' 22 Cf. CIANCAGLlNI 2008, p. 152, with bibliography. 18

59

Claudia A. Ciancaglini

Phonology; Etymology and Ttanscription Issues ofMP Final Sequences dg> and *-rk > -rg. This is the case of wuzurg 'big, great' < OP vazra-ka-, and perhaps of MP vastarg, if the basis is *vastra-, as argued by BARTHOLOMAE (see above, § 4), and the formation has been *vastra-ka- > *vastr-ka- > *vastar-ka (with secondary anaptyctic -a-) > wastarg. This second way ofvoicing assimilation is presumably later than the first one, and this may account for the fact that the parallel traditions testify that it was not yet accomplished in the early MP period: see, for instance, the abovementioned Syr. bstrl 'a part ofthe wedding-dress' +- MP wastarag 'clothing, garment' etc.; and also Tahn. Aram., Syr. dnq' /dänqä/ 'the sixth part ofthe dirham' +- MP dang < OP *danaka- 'name ofa small Persian coin.'

Beside the compelling evidence provided by the different scripts (MP /rg/, MP corresponds to NP -ra (e.g. MP carag 'means, remedy,' NP cara 'remedy, eure'), whereas MP -rg < *-rka-, spelled , is continued in NP by -rg (e.g. MP wuzurg 'big, great,' NP buzurg 'id.'), or more rarely by -r@, as for instance in NP gugar < MP -i,vidarg (MACKENZIE: widarag) 'path, passage.' Although the majority ofcases of transcribed as -rag by MAcKENZIE lacks a NP continuant, the instance ofNP gugar is significant: its loss ofthe final plosive is better explained ifwe posit a MP final sequence -arg. In other words, a change MP -arg> NP -ar is phonologically natural, whereas a hypothetic change MP -arag > NP -ar (instead of-ara) needs an explanation.24 Excluded from this picture are at least two instances in which is preceded by MP -rr- < *-rn- and that have to be read with final -rag, i.e. parrag 'wing,' NP parra 'id.' (see § 4), which in fact has the double spelling

transcriptions -rf! -ra:;!, -lag, (-rk)

24 NP bistar 'bedding, cover,' if derived from MP wistarg 'id.' (MAcKENZIE: wistarag) might

23 NYBERG (197 4, p. 186) also argues that this word is probably related to YAv. xsviwra­ ' swift' and Skr. k�ipra-.

be another case of -0, but it cannot be excluded that it continues MP wistar 'id.' 25 On the etymon of MP .nvarrah and the initial altemance xw- -f- see GNOLI 1999, with bibl.; 1996; SKJJERV0 1983.

Claudia A. Ciancaglini

60

Historically, MP -rg, spelled , may have different origins; it may derive from: a) 0Ir. *-rka-, where *-r- is part of the stem and *-ka- is the suffix, as in MP widarag 'path, passage' < 0Ir. *vitar-ka- ; b) 0Ir. *-rka-, where *-ka- is part of the stem, as in MP gurg 'wolf' < PIE *wlk117 c) 0Ir. *-rga-, where *-ga- is a different suffix, as in MP mastarg 'skull' < Ur. *mast-r-s/-; d) Oir. *ra-ka-, where *-ka- is the suffix, and with the early loss ofthe presuffixal vowel -a-, as in MP wuzurg 'big, great' < OP vazra-ka-. -;

On the other hand, MP -rag/-lag (and exceptionally -rk), spelled 'chicken, hen' < Ir. *krka- 'hen,' which is an onomatopoetic word. lt is also noteworthy, in my opinion, that the Book Pahlavi script is not always as archaizing as generally thought, given that in our case it records the contact as­ similation ofthe voicing of */-r+k-/ though this phonic change should be dated to the Early MP period. The Pahlavi script, however, does not record the voicing in intervocalic position, which developed later in MP, as is confirmed, apart from the Pahlavi script itself, by the parallel traditions, especially Middle Aramaic and Armenian. A last consideration, quite marginal to the topic of this paper, concerns the possibility that MP -g (in wistarg, wuzurg etc.) may have been perceived as an autonomous suffix, different from the most common MP-ag

61

ABBREVIATIONS CPD= D.N. ivIAcKENZIE: A Concise Pahlavi Dictionary. London 1971 Frahang = Frahang i Pahlavzk, ed. H.S. NYBERG, B. UTAS. Wiesbaden 1988 Ilr. = Indo-Iranian Ir. = Iranian LW= loanword ManMP= Manichaean Middle Persian Md. = Mandaic Mir. = Middle Iranian MP= Middle Persian NP = New Persian Oind. = Old Indian Olr. = Old Iranian OP= Old Persian Paz. = Pazand PIE= Proto-Indo-European Skr. = Sanskrit Talm. Aram. = Talmudic Aramaic

BIBLIOGRAPHY M. BACK: "Die mittelpersische Lautverschiebung: ein Stilwandel." In: Sprache 27 (1981), pp. 178-186. H.W. BAILEY: Dictionary of Khotan Saka. Cambridge 1979. CH. BARTHOLOMAE: Altiranisches Wörterbuch. Strassburg 1904. C.A. CIANCAGLINI: Iranian Loanwords in Syriac. Wiesbaden 2008. ---- "Outcomes ofthe Indo-Iranian suffix *-ka- in Old Persian and Avestan." In: DARJOSH Studies II. Persepolis and his Settlements. Territorial System and Ideology in the Achae­ menid State, ed. A.V. RossI, G.P. BASELLO. Napoli 2012, pp. 77-86. - "Allomorphie Variability in the Middle Persian Continuants of the Old Iranian Suffix *-ka-." In: Studies on Iran and The Caucasus. In Honour of Garnik Asatrian, ed. U. BLÄSING, V. ARA.KELOVA, M. WEINREICH. Leiden 2015, pp. 291-308. GH. GNOLI: "Über das iranische htj,arnah-: lautliche, morphologische und etymologische Probleme. Zum Stand der Forschung." In: Altorientalische Forschungen 23 (1996), pp. 171-180. - "Farr(ah)." In: Encyclopaedia Iranica 9 (1999), pp. 312-319. W. HINZ: Altiranisches Sprachgut der Nebenüberlieferungen. Wiesbaden 1975. P. HORN: Grundriss der neupersischen Etymologie. Strassburg 1893. H. HüBSCHMANN: Persische Studien. Strassburg 1895. R. LAZZERONI: "Per una definizione dell'unita indoiranica." In: SSL 8 (1968), pp. 131-159, repr. in In.: Scritti scelti. Alessandria 1997, pp. 103-126. - "Considerazioni su -as > -o.in sanscrito e in avestico." In: SSL 9 (1969), pp. 185-197. D.N. M simply means "pipe-conduit." 8 In 1962 a further contribution to the debate was made by ILYA GER­ SHEVITCH, whd attributed to Av. näuuiia- both the meaning of "navigable" and that of"running in channels."9 The channel-the01y has been embraced by many scholars, * This paper is part of a research planned in the framework of various Projects on Achaemenid linguistics and epigraphy funded by the Italian Ministry for Education (MIUR) since 2005 (PRIN Projects: 2005105580, 2007ZKPPSM, 2009JHSEE7). 1 In fact, the form näuuaiia- is more frequently found than the (expected) näuuiia-. On the -aiia- vs. -iia- suffix alternation see PANAINO 2004, p. 274 fn. 2 with references . 2 See HERZFELD 1930, pp. 62-64; 1938, p. 324; 1947, pp. 563 ff. 3 See, e.g., WEST 1892, p. 119 ("flowing in a channel (naev-tiik)"). See below, fn. 43 for a different behaviour in WEST 1880. 4 See DARMESTETER 1887. 5 See JUNKER 1912, p. 86 6 HERZFELD 1938, p. 324 fn. 1. 7 HERZFELD 1947, p. 564. 8 Ibid., p. 565. 9 ÜERSHEVITCH 1962, p. 173.

100

101

Ela Filippone

On the Meaning ofAvestan näuuiia- and Pahlavi *näydäg

among which ERNST HERZFELD, HAROLD W. BAILEY, 10 GHERARDO GNÖLI, ANTONIO PANAINO, DAVOUD MONCHI-ZADEH11 and others. However, it has not been universally accepted. Many have expressed their thinking on the matter with more or less elab­ orated comments or simply making their own lexical choices in text translations. More recently two articles have been published on this specific issue. ANTONIO PANAINO (2004) has refashioned the arguments supporting the channel-theory, 12 PAUL WIDMER (2007) has suggested a new interpretation for Av. näuuiia;.., also challenging its current and never contested etymology. 13 The result ofthis long debate is that nowadays several different interpretations are found ofAv. näuuiia- and/or Pahl. is not confined to the semantic aspect; in fact, we are still groping for its pronunciation. In M ACKENZIE 1986, p. 58, it is recorded as *näydäg "deep, unfordable, navigable," with the asterisk marking, as usual in M AcKENZIE's Pahlavi dictionary, doubtful transcription or meaning (in this case, seemingly, doubtful transcription).22 Päzand texts generally transmit it as näwaöä. The more usual Pahl. spelling seems to be , ). And since the same spellings may in theory be used to record other Pahl. words,23 we cannot even be sure that in a given passage one is entitled to transcribe as *näydäg graphic sequences like or or in one ofthe other 'alter­ native' spellings, whatever its actual pronunciation, was not a tenn rooted in the lexical repertoire ofthe authors ofthe Pahl. religious texts (translations or exegesis ofthe Avesta). Possibly, it was expressly created to render Av. näuuiia- in transla­ tion and has always remained mentally associated to it (in Av. quotations or remi­ niscences), though in the course of time it might have suffered deterioration and loss of its phonetic substance, with possible subsequent reinterpretations. All this will make it harder to define the exact reading and meaning of, the BAILEYunpubl. I, pp. 76-77 "flowing in channel." MoNCHI-ZADEH 1975, p. 112 (the name ofthe river 'Röt Nävatäk' meaning "Kanal-Fluß"). 12 See already PANAINO 1990, pp. 115-116. 13 See below, § 5. 14 First tentatively suggested by HENNING 1948, p. 309. For HENNING's earlier views see below fn. 147. 15 C f. BENVENISTE 1933, p. 236. 16 Cf. KREYENBROEK 1985, pp. 99-100 n. 4.6 See also below § 3.2. 17 SKJ1ERV0 2005a, p. 315, 2011b, p. 326. 18 WIDMER 2007. 19 All of them have inherited their broad semantic range from Lat. canalis "(1) a channel or conduit for supplying water [ ...]; also, a trough [ ...]; (2) an open drain or gutter [ ...] along the roof of a building; (3) the channel or bed of a river [...]" (GLARE 1982 p. 263). In English, the term canal is used unambiguously with reference to artificial products. 10

11

zo WIDMER 2007, pp. 216-222.

WIDMER 2007, pp. 221-222, fn. 10-11. Cf. also FARAVASI 1967, p. 319 navtak "navigable, navigable river [qäbel-e kestirani, rud­ e qäbel-e kefürani];" ABR.AMJAN 1965, p. 125 navtak"navigable." According to ZAEHNER 1955, p. 214, "N'YWT'K should perhaps be read as näyutäkwith a light u as inpaf'xsay (P'TWXS'Y, normal spelling in P ahlavi) with the normal development v > y." 23 One example for all, for ewtag "alone, single," see below § 2.7. 24 Words for 'water' are commonly used for 'river' in some Iranian (and not only Iranian) languages; see also FrLIPPONE 2016, p. 31 fn. 27. According to MALLORY, ADAMS 2006, p. 126, the attested meaning of words for 'water' derived from IE *h2ep- "suggests an original 'living water,' i.e. 'water on the move. "' 21

22

102

On the Meaning ofAvestan näuuiia- and Pahlavi *näydäg

Ela Filippone

2.1. In the Mihr Ya§t (Yt 10), the mythical land inhabited by the .Aryan people, the first and best creation ofAhuramazda, is described as a land where tall, shel­ tering mountains provide ample pasture for cattle, where deep lakes stand with surging waves and where (1) äpö nävaya para8wis xsaoömJha fJwaxsante äiskatam pourutamca mourum häröyum gaomca suxöam x''äirizamca (Yt 10.14)

"näuuiia- waters/rivers25

rush, broad, with a swell towards Iskata and Parutia, Margu, Haraiva and Gava, Sogdiana and Chorasmia."

As anticipated above, ÜERSHEVITCH' interpretation ofapö nauuiiß, as "naviga­ ble waters" was strongly challenged by Gherardo GNOLI. When GERSHEVITCH (1962, p. 79) admitted that Av. näuuiia-, besides meaning "navigable," could also be used with reference to "waters running in channels,"26 he did not mention the Mihr Yast passage under discussion, and therefore we do not know in principle which was his last stance on this matter.27 However, his statement (1962, p. 80) that "Av. apö navaya is used in the sense of 'channel waters' " permits US to inter­ pret his thinking with a certain confidence, including the apö navaya in the Mihr Yast among the attestations ofthe alleged "channel waters." According to PANAINO (2004, pp. 281-282) in Yt 10.14 "le acque näuu(a)iia­ non sono [ . . .] evocate tanto per la loro navigabilita (in senso negativo), bensi per la loro positiva funzione, indispensabile per il nutrimento del bestiame e per l'ir­ rigazione delle piante e dei campi, in un contesto in cui e la maestosita dello Airyana Vaejah [ ...] ehe deve essere pienamente celebrata." That in this passage the function ofwater, vital element for the cosmic system, is highly exalted, is un­ deniable; the same thing happens in an impressive number of other Avestan or Zoroastrian passages. To my opinion, however, there is nothing in these lines which may be understood as an actual reference to structures built by humans (i.e., water canalization) for the "inigazione delle piante e dei campi." The only image which comes out from them is that of a mythical, mountainous environment, with the magnificently impressive energy of the primordial natural force. This majestic scenery resembles that depicted in the Farvardin Yast (Yt 13.910). Speaking to Zaratustra, Ahuramazda reveals to be assisted by the Fravasis in maintaining the great and wide earth, bearing all the bodily world, with its high mountains rich in water, rich in pasture land, upon whom GERSHEVITCH 1959, p.81 "navigable rivers" (see also p. 174, n. 144); GNOLI 1966, p. 71, 1967, p. 15; PANAINO 2004, p. 282 "acque canalizzate;" DARMESTETER 1883, p. 123 "wide-flow­ ing rivers;" MALANDRA 1983, p. 60 "wide irrigation waters;" SKJ.2ERV0 2011a, p. 51 "flooding waters;" MENDOZAFORREST 2011, p. 14 "rivers in spate." 26 ÜERSHEVITCH 1962, p. 173. 27 See alsoGNOLI, 1967, pp.83-84 fn. 2. 25

103

(2) paiti Braotö.stäcö äpö tacil:zti näuuaiia (Yt 13 .10) "nauuiia- waters/rivers28 flow flowing in streams"

and many sorts ofplants grow up. We have again a description ofa mythical land, the product ofa divine creation. Challenging the 'channel-theory,' ZAEHNER29 points to a particular passage from a P ahl. hymn to the Creator (Dadar foayenzdarzh "Hommage to the Creator") contained in the Zand Xorda Avesta,30 where the creation on the earth of the fol­ lowing entities are attributed to the Creator: (3) röd f *näydäg, köf f buland f purr-wästar f äbömand ud göhrömand ud därf wuzurg-täkfräx-*säg3 1 ud zofr-resag ätaxs-iz fsuxr fsözäg mardöm fwas-ewe­ nag ud göspand fpanj-ewenag fwas-sardag "*näydäg-rivers,32 high mountains füll of pasture, füll of water and miner als, and trees with big branches, wide boughs and deep roots and also red and buming fire, men of many kinds and small cattle of five kinds and many species."

Note that for the word transcribed as *naydäg by the editor both the mss used ) by DHABHAR transmit the spelling < ywt' k>.33 34 ZAEHNER's remark that the "creator can scarcely be credited with the creation of 'canals"' has been rejected as an "osservazione [ ...] gratuita" by GNOLI 1966, p. 71, fn. 5 and as a "questione [... ] mal posta" by PANAINO (2004, p. 283), who finds it to be not "impressionante ehe anche le acque incanalate nel loro letto o in percorsi artificiali siano state considerate nel corso del tempo come una benedi­ zione del cielo," since the Zoroastrian tradition has always attributed to Ahura­ mazda all that is considered good on the world. This is an unquestionable fact. Furthermore, any human activities aimed at improving the overall conditions of the earth are felt as highly meritorious. However, there is something unclear in PANAINO's statement: does he consider "le acque incanalate nel loro letto" and those "incanalate in percorsi artificiali" as equivalent concepts for which the same expression ca� be used? What are in fact "le acque incanalate nel loro letto" ifnot 'natural' river's waters? Rivers do flow in (natural and open) channels! And what would mean "cham1elled rivers" if one just intends rivers flowing in their bed? In fact, the arguments produced by scholars in support of their opinion on the m�aning ofAv. nauuiia- and P ahl. *näydag (including the arguments produced in th1s paper) may not rarely be felt as shaky because oftheir circularity and indemon­ strability. Nevertheless, I find it difficult not to agree with ZAEHNER, and not to 28 MA LANDRA 1971, p. 113 "channel waters" (see also p. 164 "translated in the light ofGER­ SHE VITCH' remarks on Olr. *näu- 'channel; boat' "); KELLENS 1975, p. 36 näuuaiia- "navigable;" SK.TrERV0 201 lb, p. 65 "waters ... in spate." 29 ZAEHNER 1955, p.21411. A. 30 DHABHAR 1927, pp. 252-253; transcribed in ZAEHNER 1955, p. 217. 31 On the readingfräx-*säg and translation "bough" see DHABHAR 1963, p. 441 n. 19. 32 DHABHAR 1963, p. 464 with fn. 4; ZAEHNER 1955, p. 198,214 n. A "navigable rivers." 33 Cf. DHABHAR 1927, p.404 (23.6). 34 ZAEHNER 1955, p. 214 11. A.

Ela Filippone

On the Meaning ofAvestan näuuiia- and Pahlavi *näydäg

consider at least 'odd' to open a list of divine creations on the earth with the men­ tion of artificial channels. The Pahl. Riväyat accompanying the Diidestän T Denfg contains one of the several descriptions of Kangdiz, the legendary, paradise-like stronghold of the Kayanians, located somewhere in Bast Iran. Kangdiz is here portrayed as a pleasant place surrounded by seven walls with silver palaces and golden pinnacles;

Nothing in this passage authorizes one to think that the described phenomena were not the result of natural geological processes, though presented in a narrative context ofmythic cosmology. lt is well known that sources miginate rivers, rivers may originate other rivers; they can either have a single stream ofwater or several streams com1ected with each other. But what is it that distinguishes *näydäg rivers from other rivers? The emergence of the first two rivers, the subsequent eighteen rivers and all the others are also referred to in another passage of the Bundahisn. There,44 the eighteen rivers are not qualified as *näydiig, but as miidagwar "principal" (iin ha§t­ dah röd miidagwar). In the Wizfdagfhii f Zädspram, the eighteen rivers are said to be 'great' (ha§tdah röd T wuzurg). 45 To know which are these rivers, suffice it to read their names as they are listed in the Bundahisn: according to the body of knowledge and beliefs of the Zoroastrian tradition, they are among the greatest and most distinguished rivers in the world. 46 Another possible occurrence of *näydäg in the Bundahisn, this time in collo­ cation with iib, is in a passage dealing with a deed (or more probably a misdeed) attributed to the Turanian Fräsfäb, the arch-enemy of the Iranians in the Kayanid period, to which traditionally an 'excellent ability in constructing channels' is at­ tributed: 47 in the Kayänseh Sea he spurd 48 a thousand sources ofwater, the sources ofthe Zarrenömand (also said Hedömand) river and also (6) u-s xän f Wätaenf röd ud fos äb f *näydäg49 andar ham zreh spurd ud mardöm nisastag be kard (Ir. Bd.11 a.32)50 "he spurd the source of the Wataeni river and six51 *niiydäg waters/rivers52 in the same sea and made people settle (there)."

104

(4) u-s cahärdah köf andar u-s haft röd f *näydäg3 5 andar u-s hafl murw [...] andar (Pahl. Riv./DD 49.7)36 "and (there are) 14 mountains in it, and seven *näydäg rivers37 in it, and seven meadows [ ...] in it." The description of this mythical place, with the combination of monumental buildings, urban structures and mountainous environmental elements within the city walls, points to an ideal microcosm and reproduces in the main the outline of the Aryan prototypical landscape. Another account of Kangdiz (but with no reference to mountains) is found in the Ayadgiir f famiispfg. Notably, there the seven river flowing inside the blessed reahn are qualified as "deep" (in Pärsi: haft röd biilii). 38 gical system is repeatedly 2.2. The mythical formation of the cosmic hydrolo 39 the first two rivers, Arang d Ohrmaz of recalled in the Bundahifo. By the agency the and Weh, started flowing from the North (Alborz mountain), one going toward rivers other of waters the d receive earth, the d east and one toward the west, encircle in one and melt again into the Fräxwkard Sea, their original source. In particular, passage one reads that hastdah röd f *näy­ (5) ka än da röd be tazfd büd pas az ham bun-xänän f awe§än 41 tazfd hend (Ir. Bd. 4 fräz än däg 0 be tazfd pas abarfg äbfhä az än *näydäg 42 6.b.21) "when those two rivers had flowed, then from the same sources as those, eight­ een *näydäg rivers43 flowed out, then other waters/rivers flowed out from those *näydäg ones." . WILLIAMS 1990 I, pp. 190-191, II p. 89. S 1990 II, p. 89 "*navigable rivers." 37 BoYCE 1984, p. 64 "navigable (?) rivers;" WILLIAM See also WILLIAMS 1990 I, p. 320 (näydtig (or näwdäg?) "navigable"). 38 AGOSTINI 2013, p. 56 [(7.5)]. Bd. 11.1; TD1,DH vs. TD2, K.20, 39 Note the mss divergences on this point (Ir. 146). p. 2005, PAKZAD by out M51 ), as pointed 40 in TD1,TD2,DH; < n'wt'k> inK.20,K.20b,M51; cf. PAKZAD 2005,p. 99 fn. 181. 99 fn. 183. 41 114 in both mss, the editors KoTWAL and KREYENBROEK have proposed to read *näydäg'fh, a reading which is however very uncertain and open to question: (23)

ka mähfgän hend äb pad *näydägfh ne ämär (Ner./Pahl. 49.11) "Ifthere are fish the water/river is not to be considered *näydäg."

KOTWAL, KREYENBROEK 2003, p. 221"navigable;" see also p. 290 (Glossary). The Av. deictic adverb ißa has here a cataphoric function. Translating "similarly" KOTWAL, KREYENBROEK (2003, p. 221) attribute to it an anaphoric function. 11° C f. KOTWAL, KREYENBROEK 2003, p. 220 fn. 1401, 1402, p. 222 fn. 1419. 111 Ibid., p. 221 "deep." See also p. 309 (Glossary) "deep, unfordable, navigable." to edon an 112 Translating "the same" KoTWAL, KREYENBROEK (2003, p. 221) attribute anaphoric function. See fn. 110 above. 113 Ibid., p. 221 "but." In my opinion, the Av. discourse marker äaJ has not been used here with an adversative value. 114 So ibid., p. 222 fn. 1425. I wonder, however, in which way this sequence would differ from the sequence . 108

109

115

Even more debatable is the reading *niiydiig proposed by KoTWAL and KREYENBRO�K em�nding SPäzand) razq gqda (in both the mss) in the following �assage dealmg _with the Ab-zöhr ritual, which I quote according to the transcrip­ tion and translat10n ofthe two editors, who also note that 'deep water' would pre­ sumably refer to 'water from a stream:' (24) ab 'i*näydäg säyed hän ke bun fxän *gandag (Ner./Pahl. 30.15) "One may use deep water in case the (water) of the lower part of the house stinks."(KOTWAL, KREYENBROEK 2003, pp. 136-137).

My only remark here is that bun 'f xän would be better interpreted as "source" and that also * gandag is a very uncertain reading. 115 WIDMER (2007, p. 221) points out that in Ner. 48-49 "[d]rei verschieden Arten Wasser �erden� ... ] behand�lt:fratat.carat- _'laufend, fliessend,' armae§ta- 'ruhig s!ehend und, mit letzterem m engerer Verbmdung, näuuiia-." In fact, the alleged !mk ofthe näuuiia- water with the stagnant water is absolutely questionable,116 as 1s �roved by the fact that the only detail mentioned in the text (i.e., quantity ofof­ fenngs accepted) equates the näuuiia- water with the rnnning water. Running and stagnant waters represent the most fundamental division ofwater ecosystems and are referred to in opposition in many other Zoroastrian texts as we11; 117 from Ner. 48-49 ':e may ded:uce that the referent here qualified by näuuiia- has something to do with the runmng water. WIDMER rightly observes that different types ofwater have been listed in other Mazdean texts, and never the näuuiia-1 *näydiig waters have found their place in these lists. In the Bundahisn, 118 e.g., seventeen kinds (sardag) ofwater are listed: the first is dew (sitting upon vegetation), the second, the one_ flowing _fr?m the mountains, i.e. the rivers (än 'f garän tazisn ast 'f röd'ihä), the third, the rammg water, the fourth, the water of wells, the stagnant (armest) and other unspecified waters, the fifth, human and animal sperm, etc. (going on �p t? seventeen with different human and animal secretions, bodily liquids and hqmds �f the plants). 119 Similarly, in the Yast dedicated to AhuränI (hidden in Y 68) different types of water and forms in which it appears are mentioned (Y 68.6); 120 again, no niiuuiia- water is mentioned. lt is possible then that even in the case ofthe Nerangestiin, Av. äp- and Pahl. äb have to be interpreted as "river, watercourse," and not as "water." Mss , cf. KOTWAL, KREYENBROEK 2003, p. 136 n. 647. I suppo e that the interpretation given by WIDMER (and KoTWAL, KREYENBROEK as well) � _ ofAv. iBa and aal, (see above fn. 110 and 112) is responsible fro such an assumption. 117 Cf. Vd. 6.30, 39 (MOAZAMI 2014, pp. 174, 178); Reväyat-e Äzur Farnbay 38 (with rawäg • mstead oftazäg; cf. REzAI BAGHBIDI 2005, p. 30), etc. 118 Ir. Bd. llb (PAKZAD 2005, pp. 158-159); Ind. Bd. 18 (BEHZÄDI 1989, p. 48). 119 Interestingly, most ofthe·elements ofthis list are introduced in the Pahl. translation ofY 38.� (MAL�NDRA, Ic�APORlA 2013.' pp. �5, 225-226) as an explanation for the seven epithets wh1ch prov1de the vanous names w1th wh1ch the waters are venerated (HINTZE 2007, pp. 227 ff.). 1 0 2 Misprinted Y 82.6 in WmMER 2007, p. 221. 115

116

Ela Filipp one

On the lvfeaning ofAvestan näuuiia- and Pahlavi *näydäg

2.7. Röd f *Näydäg (the *Näydäg River) is attested in the Bundcihisn as a h�­ dronym (Ir. Bd. 29.5)121 and a toponym (Ir. Bd. 29.9).122 The usual and problemat1c graphic variance (n' yw0 vs. ' yw0 ) permits alternative readings. GNOLI is prudent in his pronouncement ("fiume Nävtäk, (Näyötäk or Evtäk)," 1967, p. 14) also be­ cause a river called Ewtäg is attested in other Pahl. texts; GRENET, e.g., complains of the systematic emendation of Ewtäg into *Nawtäg "comme on le fait depuis Darmesteter" (2002, p. 207). In a glossary ofPärsI words copied by Manucihr son of Burzö Sanjänä in 1023/1655, and edited by SACHAU (1871), a form \j}.:i is quoted and explained as "name of a river [näm-e rud ];" 123 a few lines below 124 the same form (this time vowelled with zir under väv, i.e., nävidä) is explained as "name of a river (which) is very deep and always füll of water [ näm-e rud ast besiär camiy ast va hamise por bäfod]." As for the toponym Nautaka, the name of an East Iranian region transmitted in Greek by the historians ofAlexandre and often related by scholars to Pahl. *näy­ däg, 125 GRENET (2002, pp. 209-212) has suggested interpreting it as "the nine wa­ tercourses" (< *nawa-täka), a name well fitting the Sogdian area from the Zara:fsän Mountains to the Iron Gates where it is still possible to trace nine rivers. 126

but the Aramaic word underlying the Aramaeogram has not been identified (cf. HOFTIJZER, JONGERLING 1995, p. 1199, s.v. stmh). The sequence in one ofthe mss). 127 Before *näydäg and STMT', one finds one after the other the graphic sequences (in prepositional phrases headed by PWN/pad) m Vd. 5.51; cf. JAMASP 1907, I p. 192. See also ABRAMJAN 1965, p. 127 and FARAVASI 1967, p. 319. But, as already noted by KAPADIA 1953, p. 34 the meaning "ploughshare" "does not seem to b� appro­ priate here." KAPADIA read instead 'vachak' and interpreted it as "o!fspring, young-one;" lt should be noted however that this meaning does not seem to be appropnate here as well. A füll under­ standing of the phrases PWN n'pk and PWN n'pyk has yet to be reached (cf. ANKLESARIA 1949, pp. 125-126 and MOAZAMI 2014, pp. 156-157). . , 19s ELFENBEIN 1990, p. 11 O (but not found in ELFENBEIN s Anthology and m any other Bal. lexicographical sources). l96 Cf. BELLEW 1901, p. 163. 197 SCARCIA 1965, p. XCIV. 19s See also now "valley" in the dialect ofAvaräzän inADIB-E Tus11964, p. 374. 199 Cf. K.lYÄt�I-E KULIVAND 2010, p. 1125. 200 See alsoAsATRIAl�, LIVSHITS 1994, p. 95 ("nav/wal 'gorge, vallon, vallee'" (> madar, « frere » baradar et « fille » doxtar. Le XIXe siecle est le temps de la philologie. Dans le cas de l'iranologie, l'at­ tention s'est portee vers l'antiquite et sur deux terrains. D'un cöte, EüGENE BuR­ NOUF, philologue frarn;ais, dans son etude critique comparative de l'Avesta, posa les bases des etudes avestiques et de la linguistfque iranienne. Ces travaux furent continues,en Allemagne et ailleurs,par toute une pleiade de savants tels que BAR­ THOLOMAE, SALEMANN et d'autres. C'est alors que parurent de grands ouvrages de fond comme le Grundriss der iranischen Philologie et l'edition ÜELDNER de l'Avesta. De l'autre cöte, des specialistes s'attaquerent a l'etude des inscriptions achemenides redigees en trois langues. Les versions en vieux-perse copiees par RAWLINSON,officier britannique, et progressivement dechiffrees par un effort col­ lectif,ouvrirent l'immense domaine de l'assyriologie. Le vieux-perse apparnt clai-

PIERRE LECOQ

Paris

Le -a final en vieux perse Comme on le sait, l'ecriture cuneiforme vieux-perse est d'une simplicite re­ marquable, comparee a l'ecriture cuneiforme assyro-babylonienne. C'est un veri­ table alphabet, plus perfectionne que l'alphabet phenicien et ses derives, puisque les voyelles y sont notees. Mais cette simplicite de prime abord est perturbee par un grand nombre de conventions orthographiques, qui restent jusqu 'a ce jour inex­ pliquees.1 L'un des problemes qui intrigue le plus, est l'emploi du signe en fin de mot. On peut exposer ce probleme par les exemples suivants vp. hainä adä utä abara

« armee» « il donna» « et» « il portait»

Le premier exemple montre que le signe note le -ä final, que l'on peut appeler primaire, puisque la longueur de cette voyelle est assuree par le temoignage du sanskritsenä et de l'avestique haenä-. 11 est vrai que dans cette demiere langue, le -ä final s'est abrege au nominatif singulier: haena, 2 mais on admet qu'il s'agit d'un phenomene propre a l'avestique recent. Dans le deuxieme exemple, le meme signe note egalement un -ä final, mais celui-ci doit etre considere comme secondaire, puisqu'il resulte de la chute d'une consonne finale en vp. : skr. adhät� av. (a)däJ. La graphie du troisieme mot pose un probleme. En indo-iranien, on a affaire a un -a breffinal primaire, comme l'attestent le skr. uta et l'av. uta. La transcription habituelle du vp., utä, avec un -ä, escamote en quelque sorte le probleme, d'autant plus que le bref final secondaire, issu de la chute d'une consonne montre que, dans ce cas, le breffinal n'etait pas note: abara conespond bien au skr. abharat et a l'av. (a)baraf. Des le debut des etudes, l'idee que le signe en finale notait un -a long re­ sultant d'un allongement du bref final, semble l'emporter. Deja RAWLINSON 3 considerait que le final primaire s'etait allonge partout en vp. et il transcrivait, par exemple, uta (a = ä). SPIEGEL4 suit la rneme doctrine et ecrit uta, sans trop se

-a

-a

-a

-a

1 Sauf si l' on considere que l' mihographe represente fidelement la prononciation du vieux perse comme l'a soutenu K. H6FFMAJ.'W (1976). 2 Cf. HOFFMANN, FORSSMAN 1996, p. 54 et p. 121. 3 RAWLINSON 1949, p. 80. 4 SPIEGEL 1881, p. 156.

219

Pierre Lecoq

Le -ajinal en vieux perse

demander pourquoi la voyelle se serait allongee. II est suivi par BARTHOLOMAE, dans le Grundriss, 5 qui ne s'attarde pas non plus sur ce probleme. ANTOINE MEILLET est l'un des premiers a s'etre interesse plus longuement a cette anomalie. 6 II ecrit : « En fin de mot, aucune distinction entre -ii et -ä ; le signe de a est nute dans tous les cas, ce qui, suivant l'usage, conduit toujours a transcrire -ä en fin de mot. On n'est fixe sur la quantite d'un a final que devant un mot enclitique ; la distinction de -a et de -ä ressort alors du fait que ä seul est note ». 7 II semble bien que pour !'eminent linguiste, il ne s'agit pas ici d'un fait phonetique, l' allongement du -a final, mais plutöt d'un simple fait de graphie. KENT, dans son edition des textes vieux-perses, 8 note bien que le a final est ecrit avec le signe qui marque la longueur [c'est-a-dire le signe ], mais que ce signe peut egalement noter le ä Iong. 9 Dans la partie de son ouvrage consacree a la phonetique historique, il repete cette constatation en ajoutant: comme si cette voyelle s' etait allongee. 1 ° Cet auteur suggere donc une explication, mais sans s'aventurer plus loin. Les ouvrages les plus recents sur le vieux perse sont trop restreints pour aborder ce sujet en detail. Dans sa description de l'ecriture cuneiforme vieux-perse, TESTEN 11• n'evoque meme pas le probleme du signe en finale. Toutefois, R. ScHMITT 12 ecrit que selon toute probabilite, le en fin de mot doit representer un veritable allongement du -a bref ancien. Dans les annees soixante, ui1 article de W. HINZ declencha une vive mais fertile polemique. 13 II y reprenait une idee deja ancienne, selon laquelle, Darius, dans le chapitre 70 de l'inscription de Bisutun, se declarait l 'inventeur de l'ecriture cu­ neiforme vieux-perse. Les nombreux articles parus dans les annees suivantes ont au moins eu le merite de reprendre l'examen interne de cette ecriture et, surtout, d'evaluer ses rapports avec la realite phonetique du vieux perse. Parmi toutes les explications proposees pour resoudre notre probleme, on si­ gnalera l'une des plus ingenieuses, fournie par J. KURYLOWICZ. 14 Le savant polonais explique que la distinction entre -ii bref et -ä long s'est neutralisee, non seulement en finale absolue, mais aussi en syllabe finale, avant la chute d'une consonne (cas des voyelles finales secondaires). Dans le cas des finales vocaliques anciennes, ce phenomene est possible, mais dans celui des finales vocaliques entravees, il est

peu probable, meme si l'on fait intervenir un facteur prosodique comme l'accen­ tuation. Sans entrer ici dans une refutation detaillee, on a l'impression que les scribes devaient avoir des notions de phonetique historique, ce qui est naturelle­ ment impossible. Quoi qu'il en soit, la tentative de KURYLOWICZ a le merite de chercher une explication dans une perspective diachronique. Nous y reviendrons. On a aussi fait remarquer cent fois que l'hypothese d'une creation ex nihilo de cette ecriture par Darius, se heurte a des problemes orthographiques insolubles. On abandonne maintenant l'interpretation du chapitre 70 de Bisutui1 que HINZ avait proposee. La demiere traduction du texte elamite, par F. VALLAT, 15 permet d'affirmer une fois pour toutes que Darius, dans ce texte, parle de traduction et non pas d'ecriture. Ainsi s'impose de plus en plus l'idee que cette ecriture a une histoire, qui pourrait remonter a l'epoque mede et meme au-dela, comme on l'a deja souvent pense. 16 C'est precisement dans une perspective historique que nous voudrions propo­ ser une solution a notre probleme. Rappelons qu'il nous faut expliquer la diffärence de graphie, en finale, entre Ie -a bref primaire, uta ecrit utä, tandis que le -a bref secondaire n' est pas note, comme dans abara. Tout s'explique si l'on considere que l'ecriture a ete creee avant la chute des consonnes finales. On a d'abord ecrit *abarat, puis plus tard, simplement abara. Cela signifie qu'a une epoque ancienne on ecrivait les -a brefs et les -ä longs avec le seul signe . Ceci ne saurait surprendre puisqu'on se trouve dans la meme configuration qu'a !'initiale, ou ii- et ä- sont ecrits de la meme maniere, avec le seul signe . Dans ui1 certain nombre de mots se terminant par ui1 -a, cette voyelle est notee par le signe comme s'il s'agissait d'une longue (type utä). Nous avons vu que l'explication la plus courante veut que cette voyelle se soit allongee, mais cette ex­ plication d'ordre phonetique ne saurait etre retenue : la fixation de l'accent sur la penultieme ou sur l'antepenultieme, responsable de la chute des consonnes finales, ne peut en aucun cas justifier un allongement des voyelles dans la meme position. Or, dans ces memes mots, a la suite d'ui1e enclise, le -a final n'est pas note et vaut donc a bref, conformement a l'etymologie, comme par exemple utä : utamäm, utamaiy (14 exemples), utasim, etc. Si le -a final etait devenu long, il se serait pro­ page en toute position. 17 Le suffixe -dä (idä « ici » pour *ida, cf. av. ü5a, skr. iha), dans le mot avadä « la; alors » est presque toujours ecrit avec ii bref devant un enclitique : avadasa, avadasim, avadasis (une seule fois : avadäsim : DB §47 74). 18

218

BARTHOLO:MAE 1895-1904, p. 160 et utä, p. 142. MEILLET 1915; 2e ed. revue par E. BENVENISTE 1931, pp. 91-93. 7 Jbid., p. 91. 8 KENT 1953. 9 Jbid., p. 17 (§ 36). lO Jbid., p. 26 (§ 61). 1 1 TESTEN 1996, pp. 134-137. 2 BRANDENSTEIN et 1 SCHMITT 2004 (cf. pp. 720 et 726). C'est deja la these de W. M. MAYRHOFER (1964, pp. 28 et 30 (§ 20) ). 13 HINZ 1968, pp. 95-98. Voir aussi: Id. 1952, pp. 28-38; etc. 14 KURYLOWICZ 1964, pp. 103-107 ; cette explication sera reprise dans Id. 1972, pp. 47 s. 5

6

15

VALLAT 2010, pp. 492-500. Ainsi, R. SCHMITT (cf. SCHMITT 2004, p. 722) admet maintenant que l'ecriture a pu exister a l'epoque de Cyrus. 17 II est vrai que 1'on trouve egalement utäsaiy, etc., mais en moins grand nombre. Tous ces exemples de graphies contradictoires montrent simplement l'embarras des scribes. 18 Dans le cas de avaßä ou 1'etymologie suppose un -ä long, i1 est naturel de retrouver celui­ ci devant enclitique. La graphie avaea-d[im] de XPI §4 19 doit etre une erreur de scribe. 16

220

Pierre Lecoq

Le -afinal en vieux perse

Les cas Oll la conjonction -cä (av. -ca, skr. -ca) est ecrite -c sont trop peu nom­ breux (deux exemples) pour etre significatifs, mais on remarquera, a l'interieur d'une locution: aväkaramcamaiy (DNb §7 27-28) etyacamaiy (DNb §3 13 = XPI §3 15). De meme, la locution avah yarädiy est ainsi orthographiee et jamais *avah yä-. Mais d'autre part, il existe un certain nombre de cas Oll un breffinal primaire n'est pas note (type abara), conformement a sa valeur phonetique. Ainsi les noms de mois sont-ils regulierement ecrits -hya, au genitif, alors que partout ailleurs on trouve -h yä. 11 s'agit d'une orthographe Oll le nom du mois est traite comme m1 proclitique, voire meme comme un premier element de composi­ tion flechi, malgre l'utilisation du trait de separation des mots, dont l'usage peut etre tardif. 11 n'y a pas d'exception:

-a

a9iyädiyahya (DB §18 89, §38 18) ädukanaisahya (DB §31 69) anämakahya (DB §19 96, §25 26, §29 56, §45 63) garmapadahya (DB §11 42, §36 7-8, §42 46) ßuraväharahya (DB §26 36, §27 41, §30 61, §41 39) viyaxanahya (DB §11 37, §35 98, §46 68) 19

Cette interpretation est renforcee par une autre graphie dans le nom a(h)urahya mazdah (XPc § 3, 10). 20 Notre corpus est trop limite pour etre corrobore par d'autres exemples. Toutefois, ce genre de compose a probablement ete plus frequent qu'il n'y parait, comme l'atteste le moyen perse kärezär « champ de bataille », litt. « de l'armee », en un seul mot (persan kär(e)zär), oll le premier terme remonte a vp. *kärah ya (nos textes n'ont que kärah yä, sans exception, conformement a l'orthographe usuelle). 21 On trouve probablement le meme type de fom1ation dans le pehlevi sabexön (prs. sabixun) « expedition (litt. sang) de nuit » ; le mp. manicheen 'rdywhyst (prs. ordibeh est) « 2e mois de l'annee », < *artahya ; d'anciens composes du type prs. kadizäd « natifde la maison», kadzvar « maitre de maison», etc. ; des n0111s pro­ pres comme Garsevaz < * krsahya, Gunde säpür puis analyse en Gund-e säpür, etc. 11est plus difficile d'expliquer la meme orthographe dans quatre noms de per­ sonnes, d'autant plus que deux de ces noms ont parfois la graphie attendue:

Les noms de Haldita, Nabutaina et (H)uvaxstra sont suivis de puc;a «fils de» o� de, tau(h )mäY_ä « de la famille de», deux mots qui, par leur sens, conviennent b1en a la format10n d'un compose a premier terme flechi. Mais en face d'eux, il y a un grand nombre d'exemples de genitifs determinant les mots « fils de» et « famille de » . et qui ont la graphie requise : -h yä. Pourquoi ces trois noms ont-ils cette fonn� �nattendue ? En outre, le nom Vahyazdäta apparait dans m1e construc­ . tlon �n ezdfe.: « l'an_nee, (celle) de V.», oll il ne peut etre question de compose. l:n� �tourdene du s�nbe est exclue: il y a peu, mais tout de meme trop d'exemples repetes de la graph1e -hya. 11 reste que le scribe a pu, dans des passages proches les 1:111s des �u:res, employer une graphie qui rendait mieux la phonetique reelle. C�c1 �ourrait etre encore un argmnent contre l'allongement spontane d'un -a bref pnmaire. Si notre hypothese est exacte, les trois signes vocaliques se trouvent donc sur le meme pied : ils notent a la fois, respectivement, les breves et les lon�u�s. ?n a souvent fait remarquer que dans le cas des voyelles de timbre i et u, la d1stmct10n de longueur est peu pertinente. 22 Dans le cas de /a/ et /ä/ la distinction de longueur presente un rendement bien plus eleve et la distincti�n de timbre 23 �es scribes ont prefäre de ne pas noter 1� . �omm� e� p�r�an, y est deja ��pitale. a bref a 1,mten �r �u mot � etalt un p1s-aller qui montre a quel point le respect ; , _ , d un systeme d ecnture hente exen;ait une contrainte telle que tout changement . etalt exclu. Plutöt l'imprecision que l'im10vation.

7

halditahya (pU Kh. dyü su�­ _ gests. As for the meaning, one may add that the OKh. phrases �e JU ye and ma JU ye mean 'no one' in negative and prohibitive clauses re_spectively (�MMERICK, forthcoming, § § 11.5 and 25.22), although na can be substituted forma m the later language, so that, despite the different syntax, LKh. Suv 3.83 majve and 3.86 na _ jve 'no one' are close matches of Sanskrit mä ... kasyacid 'for no o�e' an� ma ... kvaci(d) 'nowhere' respectively. Thus, the Su� passages �d�r cons1deration can , be simply translated 'May no one know suffenngs (or) pams (3.83) and May no one know the name ofsuffering' (3.86). MM]

veloped the aspiration to awakening (bodhicitta) upon hearing the Buddha's preaching. Accordingly, SKTJERV0's side a is the recto and side b is the verso. MM]

r smal� Sgh 43 [6] batakujve tt!mapäre,�däpharu ye vfväfupajäysd� �1:�wev . � ttarrai va o bznasta ku .3 97 the seed that one sows 1s, one gams much reward; (anyhim gives one no thirsty, or hungry gets he hämäte nai jve hecjä 'when

2.67d: In the translation, substitute 'went' (tsute) (also RevTr) for 'came.' 2.72cd: Translate, with LEUMANN, 'Not (ni) for the sake ofhonour, not (ni) for profit, let us display (our) magic powers (Skt.rddhi), Buddha' (also RevTr) instead of 'For the sake of our [ni] honour, let us display rddhis as a blessing to them [ni], Buddha.' 2.89d: In the translation, substitute 'quickly' (yslttaru bär;f,u) (also RevTr) for 'a short time.' 2.96b: In the translation, substitute 'Watch out!' (vä äysda yanfru) [cf. LEU­ MANN: 'Her kund machet euch [her schauet]'] for 'May you protect me!' 2.128a: In the translation, substitute 'restrained me' (muho ... pathfye) (also RevTr) [with Lemnann] for 'restrained us.' 2.233c: In the translation, substitute 'They encircled (karä tsutändä) the Bud­ dha' (balysä karä tsutändä) (also RevTr) for 'They even [karä] came to the Bud­ dha' following a suggestion made by ALMUTH DEGENER (cf. Dict. 55 s.vv. karvfnaa­ and karvlra, [EMMERICK] in Studies 2.32-34 s.v. karä, DEGENER 1989a, p. 134 § 21.3.6 and Suv, vol. 2, p. 253 s.v. kara- and karvfra). 4.34d: In the translation, substitute 'cannot all enter it' (ni bissä hä tranda hämäre) with present potential construction for 'have not all been able to enter it.' The potential construction consists of a past participle + the auxiliary verbs yan­ 'to make' (+ the superfluous past participle yur;f,u 'made') for transitives or häm­ 'to become' for intransitives. lt expresses either potentiality or anteriority ("con­ summation of an action," see EMMERICK 1987, pp. 279-288). Cf. below on 5.76b and 19.92ab. 4.42b: Read gya(if ne [as LEUMANN] instead ofgyar;f,lne (19 April 1984). [The suggested word division results in a better translation: 8

280

thing).'

[Cf. also Sgh 73 .2 nejve mä#äna gyasta balysa �a hauta stä 'no one, gracious Lord Buddha, has that ability.' MM] Anantamukhanirhäradhärar,f 16.6 IOLK.hot 24/3 Catalogue 121 (= Kha. i.119 KT3.128) v5 khojve harridarye hecjä ... 'As one gives to another ... '

[Note that the tiny fragment Or. 12637 /59.1 Catalogue 150 (= Hardinge 075.6 KT 5.277 (599) = Hardinge 075.10 KT 5.278 (603)) is in all likelihood f�om a variant manuscript ofthe Anantamukhanirhäradhärarzf (§ § 16.6-7; for the T1betan text see ed. INAGAKI 1987, p. 264 and trans. INAGAKI 1999, pp. 176-177). Sr({�jifl-�: 'im�j(��fg� = Exhibition ofCentral Asian Artifacts from the Ötani Expeditions Preserved in the Library ofRyukoku U niversity, 2-10 Sep­ tember 1983, the Library ofthe Omiya Campus (Shichijo-Omiya, Kyoto). Kyoto 1985.

289

Philadelphia 1955 (Transactions ofthe American Philological Society n.s. 45.5). K.AGAWA TAKAO �J 11 �!i: Muryöjukyö no shohon taishö kenkyü �:i:�;fil(J) �*�flijivf � [A Comparative Study ofthe Texts ofthe Larger Sukhävätfvyüha-sütra]. Kyoto 1984. 0. KLfMA: Review ofZ. In: AO 37 (1969), pp. 625-626. KT l-7 = H.W. BAILEY: Khotanese Texts. Cambridge: vol. 1, 1945; vol. 2, 1954; vol. 3, 1956 (vols. 1-3: 2nd ed. in one vol. 1969; repr. 1980); vol. 4, 1961 (repr. 1979); vol. 5, 1963 (repr. 1980); vol. 6: Prolexis to the Book ofZambasta, 1967; vol. 7, 1985. S. KoNow: "Zwei Handschriftenblätter in der alten arischen Literatursprache aus Chine­ sisch-Turkistan." In: SPAW (1912), 1127-1139. - "Primer of Khotanese Saka: Grammatical Sketch, Chrestomathy, Vocabulary, Bibliography." In: NTS 15 (1949), pp. 5-136. A. KUNST, Review ofZ. In: BSOAS 33 (1970), pp. 630-633. KV = M. MAGGI: The Khotanese Karmavibhanga, Roma 1995 (Serie orientale Roma 74). E. LEUMANN: Buddhistische Literatur nordarisch und deutsch. 1: Nebenstücke. Leipzig 1920 (Abhandlungen für die Kunde des Morgenlandes 15.2) [repr. Nendeln 1966]. - Das nordarische (sakische) Lehrgedicht des Buddhismus: Text und Übersetzung, aus dem Nachlaß hrsg. von M. LEUMAl\'N. Leipzig 1933-1936 (Abhandlungen für die Kunde des Morgenlandes 20). - "Zu zwei khotansakischen Fragmenten." In: ZDMG 113 (1963), pp. 80-86, 1 pl. M. MAGGI: Review of SDTV3. In: JIJ 41.3 (1998), pp. 282-288. - "11 Grande veicolo in Asia Centrale: il buddhismo di Khotan e i1 Libro di Zambasta"; "11 libro di Zambasta: capitoli 1-2, 5, 13-4, 24." In: La Rivelazione de! Buddha. 2: II Grande veicolo, a c. e con un saggio di R. GNOLI, introd. ai testi tradotti di C. CicuzzA e F. SFERRA con contributi di M. MAGGI e C. PECCHIA. Milano 2004a (1 Meridiani. Clas­ sici dello spirito), pp. clii-clxv, ccxv-ccxviii, 1193-1285. - "The Manuscript T III S 16: Its Importance for the History of Khotanese Literature." In: Turfan Revisited: The First Century of Research into the Arts and Cultures of the Silk Raad, ed. D. DURKIN-MEISTERERNST et al. Berlin 2004b (Monographien zur indi­ schen Archäologie, Kunst und Philologie 17), pp. l 84a-190b, 457. - "Annotations on the Book ofZambasta, II: Khotanese ma"fJ1,kyä -." In: BAI l9 (2005 [2009] = Iranian aildZoroastrian Studies in Honor of Prods Oktor Slg'cervß), pp. 103-109. - "Nagarjuna's Quotation in the Khotanese Book of Zambasta." In: ARIRIAB 10 (2006 [2007]), pp. 533-535. - "Annotations on the Book ofZambasta, I." In: Literarische Stoffe und ihre Gestaltung

in mitteliranischerZeit: Kolloquium anlässlich des 70. Geburtstages von Werner Sun­ dermann, hrsg. v. D. DURKIN-MEISTERERNST, C. RECK, D. WEBER. Wiesbaden 2009

(Beiträge zur Iranistik 31), pp. 153-171, pi. i. M. MAGGI, G. MARTINI: "Annotations on the Book ofZarnbasta, III: Chapter 18 no more." In: Scripta: an International Journal of Codicology and Palaeography 7 (2014), pp. 139-158. G. MARTINI: "Tracing the Sources ofthe Book ofZambasta: The Case ofthe Yak$a Painter Simile and the Kasyapaparivarta." In: JIAAA 3 (2008), pp. 91-97. - "A Large Question in a Sm.all Place: The Transmission ofthe Ratnaküta (Kasyapapari­ varta) in Khotan." In: ARIRIAB 14 (2010a), pp. 135-183. - "Studies on the Book ofZambasta," diss., Universita degli studi di Napoli L'Oiientale, 2010b.

290

291

Mauro Maggi

-- "Mahämaitrf in a Mahäyäna Süt:ra in Khotanese: Continuity and Innovation in Buddhist Meditation." In: C HBJ24 (2011), pp. 121-194.

- "Bodhisattva Texts, Ideologies and Rituals in Khotan in the Fifth and Sixth Centuries." In: Multilingualism and History of Knowledge, ed. J.E. BRAARVIG et al. 1: Buddhism among the Iranian Peoples of Central Asia, ed. M. DE CmAR A, M. MAGGI, G. MARTINI. Wien 2013 (Iranische Onomastik 11), pp. 13-69. MW = M. MONIER-WILLIAMS: A Sanskrit-English Dictionary. Oxford 1899. PED = T.W. R.HYS DAVIDS, W. ST EDE: The Pali Text Society's Pali-English Dictionary. Ox­ ford 1921-1925. RevTr = R.E. EMMERICK: Unpublished working revised translation of most of the Book of

Zambasta. L. SCHMITHAUSEN. In: CAJ 15 (1971), pp. 75-77. SDTV3 = R.E. EMMERICK, M.I. VOROB'EVA-DESJATOVSKAJA: Saka Documents. Text Volume III: The St. Petersburg Collections, with contributions by H. KUMAMOTO et al., London 1995 (Corpus Inscriptionum Iranicarum. 2: Inscriptions of the Seleucid and P arthian P eriods and of Eastem Iran and Central Asia. 5: Saka. Texts 3). Sgh = G. CANEVASCINI: The Khotanese Sanghätasütra: A Critical Edition. Wiesbaden 1993 (Beiträge zur Iranistik 14). SGS = R.E. EMMERICK: Saka Grammatical Studies. London 1968 (London Oriental Series 23).

N. SIMS-WILLIAMS: "Chotano-Sogdica, II: Aspects of the Development ofNominal Mor­ phology in Khotanese and Sogdian." In: Proceedings of the First European Conference

of Iranian Studies Held in Turin, September 7th-llth, 1987 by the Societas Iranologica Europaea. 1: Old and Middle Iranian Studies, ed. G. GNOLI, A. P ANAINO. Rome 1990 (Serie Orientale Roma 67.1), pp. 275-296. Skazanie = V.S. VüROB'EV-DESJATOVSKIJ, M.I. VüROB 1EVA-DESJATOVSKAJA: Skazanie o Bha­ dre: norye listy sakskoj rukopisi "E ": faksimile teksta, translaipcija, perevdd, predislovie, vstupitel'naja stat'ja, glossarij i prilozenie. Moskva 1965 (Pamjatniki pis'mennosti

Vostoka 1). P.O. SKJLERV0: "Khotanese v- < Old Iranian *dw-." BSOAS 48.1 (1985), pp. 60-73. Studies 1-3 = R.E. EMMERICK, P. 0. SKJLERV0: Studies in the Vocabulary of Khotanese. Wien: 1, 1982 (Veröffentlichungen der iranischen Kommission 12); 2, 1987 (Veröffentlichun­ gen der iranischen Kommission 17); 3, ed. R.E. EMMERICK, contributed by G. CANEVASCINI et al., 1997 (Veröffentlichungen der Kommission für Irani�tik 27). Suv = P.O. SKJLERV0: This Most Excellent Shine of Gold, King of Kings of Sutras/ The Kho­ tanese Suvan:zabhäsottamasutra. [Cambridge, Mass.,] 2004 (Sources of Oriental Lan­ guages and Literatures 60-61), 2 vols. SWTF= H. BECHERT (ed.): Sanskrit-Wörterbuch der buddhistichen Texten aus den Turfan­ Funden und der kanonischen Literatur der Sarvästiväda-Schule, begonnen von E. WALDSCHMID T. Göttingen 1973-. VkN = P.O. SKrnRv0: "Khotanese Fragments of the Vimalakirtinirdesasütra." In: Ka­ lyä(lamiträräga(lam: Essays in Honour of Nils Simonsson, ed. E. KAI--IRS. Oslo 1986 (Serie B: Skrifter 70), pp. 229-260, 6 pls. A. WAYMAN: Review of Z. In: JAST29 (1969-1970), pp. 151-152. U. WOGII--I ARA: Bodhisattvabhümi: A Statement of Whole Course of the Bodhisattva (Being Fifteenth Section of Yogäcärabhumi). Tokyo 1930-1936. Z = R.E. EMMERICK: The Book ofZambasta, A Khotanese Poem an Buddhism. London 1968 (London Oriental Series 21).

JOURNAL ABBREVIATIONS

AM AO AR!RIAB BAI BSO AS CAJ CHBJ JAST 111 JIAAA JRAS MSS NTS SPAW Stil TP ZDMG

Asia Major. Archiv orientalni. Annual Report of The International Research Institute for Advanced Buddhology at Soka University. Bulletin of the Asia Institute. Bulletin of the School of Oriental and African Studies. Central Asiatic Journal. Chung-Hwa Buddhist Journal. Journal of Asian Studies. Indo-Iranian Journal. Journal of Inner Asian Art and Archaeology. Journal of the Royal Asiatic Society. Münchener Studien zur Sprachwissenschaft. Norsk tidsskriftfor sprogvidenskap. Sitzungsberichte der Preußischen Akademie der Wissenschaften. Studien zur Indologie und Iranistik. T'oung pao. Zeitschr(fi der Deutschen morgenländischen Gesellschaft.

298

Enrico Morano

Eruc PIRART Liege

BIBLIOGRAPHY F.C. ANDREAS, W.B. HENNING: "Mitteliranische Manichaica aus Chinesisch-Turkestan, II." In: Sitzungsberichte der Preußischen Akademie der Wissenschaften (1933), pp. 292363. A. BENKATO: The Manichaean-Sogdian Parahle Book: a New Edition and Literary Analy­ sis, PhD dissertation, SOAS 2015. M. BOYCE: A Catalogue of the lranian Manuscripts in Manichaean Script in the German Turfan Collection. Berlin 1960. J. CHEUNG: Etymological Dictionary of the Iranian Verb. [Leiden Indo-European Etymo­ logical Dictionary Series, ed. A. LuBOTSKY, 2.]. Leiden-Boston 2007. I. GERSHEVITCH: A Grammar of Manichean Sogdian. Oxford 1954. W.B. HENNING: "Sogdian Tales." In: BSOAS 11 (1945), pp. 465-487. [repr. HENNING 1977, pp. 169-192]. W.B. HEmHNG: Selected Papers II. Teheran-Liege 1977 (Acta Iranica 15). F. JOHNSON: Hitopadesa. The Sanskrit Text with a Grammatical Analysis Alphabetically Arranged by F Johnson. Hartford-London 18642• E. MORANO: "A Working Catalogue of the Berlin Sogdian Fragments in Manichaean Script." In: lranian Languages and Texts from Iran and Turan. Ronald E. Emmerick Memorial Volume, ed. M. MACUCH et al. Wiesbaden 2007 [2008], pp. 239-70. ---- "Sogdian Tales in Manichaean Script." In: Literarische Stoffe und ihre Gestaltung in

mitteliranischer Zeit. Kolloquium anlässlich des 70. Geburtstages von Werner Sunder­ mann, ed. D. DURKIN-MEISTERERNST, CH. RECK, D. WEBER. Wiesbaden 2009 (Beiträge

zur Iranistik 31), pp. 173-200 + Taf. II-VI. E. PROVASI: "Parabole e testi dottrinali." In: ll Manicheismo. Volume III ll mito e la dottrina. Testi manichei dell 'Asia centrale e della Cina, ed. G. GNOLI. Fondazione Valla, Milano 2008, pp. 111-225, 374-478. CH. RECK: "The Middle Iranian Manuscripts from the Berlin Turfan Collection: Diversity, Origin and Reuse. In: Eurasian Studies XII (2014), Lecteurs et copistes dans !es tradi­ tions manuscrites iraniennes, indiennes et centrasiatiques. Texts reunis par N. BALBIR, M. SZUPPE, pp. 541-553 (pl. XXI-XXIII). N. SIMS-WILLIAl'vfS, D. DURKIN-MEISTERERNST: Dictionmy of Manichaean Textt..,Volume III, 2: Texts from Central Asia and China (Texts in Sogdian and Bactrian), flUmhout 2012. L. STERNBACH: The Hitopadesa and its Sources. New Haven 1960 . W. SUNDERMANN, Mitte/iranische manichäische Texte kirchengeschichtlichen Inhalts Berlin 1981 (Berliner Turfantexte 11).

Les Soleils de l' Avesta I. Introduction Prolixe a propos de celui des Scythes, Herodote ne dit rien du pantheon qui etait venere des Medes et n'est guere disert sur celui des Perses. De surcroit, comme les inte,pretationes graecae occultent les theonymes iraniens, nous ne pouvons ve­ rifier ce que l'historien grec veut dire au juste en nous informant que les Perses sa­ crifiaient aussi au Soleil, a la Lune, a la Terre, au Feu, a l'Eau et aux Vents. Quel est donc ce dieu Soleil ? Ce1ies, il est fort probable que ce füt le dieu honore avec le Xvarsed Yast meme si comme nous verrons, d'autres possibilites existent. C'est l'appellatif vedique su var « Soleil » qui correspond au nom avestique habituel du dieu honore du Xvarsed Yast, Huyar0, car, pour la designation gD: dieu Soleil, la langue vedique recourt plutöt a d'autres mots, principalement Sür 1ya et _ Vivasvant. J'analyse ci-dessous cette divergence, mais sans revenir sur ce qm a deja ete expose concemant le Soleil dans Les Adorables de Zoroastre. 1 Dans notre recherche, nous devrons bien evidemment tenir compte de la com­ paraison indo-iranienne. En effet, plusieurs dieux zoroastriens portent un meme nom que des dieux vediques. 2 En plus de divinites visibles telles q�e les Rivieres, les Monts, la Terre ou l'Aurore, citons Aspin = Asvin, A.sa = fi.ta, Aramaiti = AraStar! @Rar mati' ,,:;tflör ; . b ' Mi8ra = Mitra, Vaiiu = Vayu. Sans compter IYJJ o JJ oxstar = Tvastr qu'Ahura pmie le meme nom qu'Asura ou que les Varyhu / Vohu sont homonymes des Vasu. Et nous devons ajouter que de nombreux heros p01ient un meme nom de part et d'autre. Des exemples intermediaires peuvent exister: le correspondant vedique Aparh· Napa't de Bar0 za(r1t) Apqm Napa't n'est jamais appele Brhant. Et nous rechercherons ce qui diffärencie le Soleil avestique de Mi8ra, car d'au­ cuns, des 1'Antiquite, donnent ce dernier pour un dieu solaire. Le nom pehlevi de ce dieu, dans la prononciation tardive mihr, est d'ailleurs passe en sanscrit sous la forme mihira- pour designer le Soleil. II. HuyarLes Cantates ( Ga8a) mentionnent le Soleil, les Etoiles et la Lune, notamment aux vers Yasna 44.3cd 1 PIRART 2010b, pp. 85 sqq.

Le cas de figure de demons iraniens homonymes de dieux vediques a souv�nt ete mis en exergue, mais il est relativement peu frequent. Je n'en conna�s que cinq illustrat10ns : hzdra = indra, Sauruua = Sarva, Nil7JhaiBiia = Ndsatya, Tauruui = Turvi, Daeuua = Deva. 2

300

Brie Pirart kasna xuv a,:zg (4)+starqmca3 dä1 aduuanam m ka ya ma uxsiieitf nar,Jsaitf 8ßa1

Qui donc a etabli le chemin du Soleil et des etoiles ? Et qui est celui par lequel la Lune croit, puis decroit ?4 mais aucune personnification n'apparait: le genre grammatical du nomdu Soleil, huiwr3 au nominatif-accusatif (< proto-indo-europeen * seH2 uH if) et xuv 'ättg au ge­ nitif (< proto-indo-europeen *seH2 uH 1 ens), 5 est le neutre. En revanche, dans l'Avesta recent, meme si le genre neutre est maintenu pour hu1wr0 nominatif-ac­ cusatif, huü geniti±: l'epithete auruuat.aspa- « possesseur de chevaux d'attaque » nous invite a considerer une certaine personnification. Aucune diffärence morpho­ logique ne permet donc de distinguer l 'appellatif du theonyme : ce dernier tout aussi bien arbore le genre neutre alors meme que, dans les cas d'Ätar0 , de MiOra ou de Var0 8rayna, le neutre a ete modifie en masculin. Par contre, le Veda, en introduisant un suffixe 0 a- ou 0 tya-, distingue soigneusement le theonyme, Sdra (< proto-indo-europeen * seH2uH 1 lo-) ou Sdriya (< proto-indo-europeen *seH2uH 1 liH1 o-), de genre masculin, de l'appellatif neutre s uvar. Avec son suffixe 0 iya-, le nom du Soleil vedique Sdr tya co:incide etymologi­ quement jusque dans le detail avec celui qu'il porte en grec, homerique 'HEAto� < proto-grec *häyelios 6 < proto-indo-europeen *seH2uH 1 liH 1 o-. Notons que le nom grec, masculin, n'est concurrence par aucun neutre. En latin, la situation est radicalement differente puisque le mot neutre, sans recevoir aucun suffixe, a pu­ rement et simplement acquis le genre masculin : Sol < proto-indo-europeen * seH2 uH 1 f. Pour sa part, le germanique possede le mot de genre neutre, gotique sauil, mais concurrence par un derive feminin, gotique sunno, Ie soleil qu'il fait ou 3 Le genitifpluriel de star - (contre KELLENS 1974, p. 369; KELLENS, PIRART 1988-1991, I ö p. 149 ; II p. 316) doit obeir aux memes exigences morphologiques que celui de vedique nf- « homme, seigneur» :: naräm (sur ce demier, voir PIRART 1989). 4 Traduction KELLENS, PIRART 1988-1991, I p. 149. l 5 Je connais les exemples suivants de la possibilite de proto-indo-europeen *eHu > vedique ü : vedique sifr iya- (aussi sdra-) = proto-grec *häweliyos (homerique 11t1cioO < *Yama-xsäwan "Yama the king") on the versa of king Huviska's coin. 48 Besides Kusans, a parallel tradition in Arsacid coinage,49 also showing a strong phase of hellenization with Semitic nuances,50 44 The Elamite name Siyatiparna < * siyäti-farnah- can give early evidence of such a link, pertaining to an Ancient Iranian phraseology that I have discussed in PIRAs 1994-1995 [ 1996]. See the auspicious greeting for Narseh's enthroning (NPi § 89) too: "and be hfilmy by Your Own g1Qry and realm" (ud pad xwes xwarrah ud sahr säd bavverl) (HUMBACH, SKJJERV0 1983, p. 69), with emphasis on political tenets and ideologies. 45 See the exhaustive treatment of the question in SKJJERV0's entry "Jamsid i: Myth of Jamsid" (http://iranica.com/articles/j amsid-i). 46 The absence of such royal imagery in the strict sense from the Avesta has to be underlined: even if in later interpretations Yima could have been considered as a prototype of the heroic and demiurgic power, shared by the Persian king too, as I tried to demonstrate in the case of Xerxes' crossing of the Hellespont (PIRAS 2011, pp. 129-130). 47 KELLENS's introduction (p. 14) to the volume füma/Yima. Variations indo-iranienne sur la geste mythique, ed. S. AZARNOUCHE, C. REDARD, Paris 2012. 48 For the weight of the Iranian and Zoroastrian component in Kusan ideology and Rabatak inscription, see GNOLI 2009. For the Kusan dynastic cult and the xv aranah- see VERARDI 1983, pp. 269-272. 49 For Parthian numismatics, see the in-depth survey of VARDANYAN (2001). With regard to our subject, see the motif of a bird with a coronet/garland in its beak (especially Phraates IV's coins, but also over a wide span of time from 50 B.C. to 130 A.D.). After Pacorus II (78-105 A.D.) the wreath remains the ori.ly sign of investiture (VARDAJ'\/YAN 2001, p. 101). 50 In Parthian times (inscription of 3rd cent. B.C.), the interpretatio aramaica of the :nvarrah as the god Gad in a falcon shape, is another point of Iranian-Semitic interactions noticed by ÜNOLI 1995a.

320

X varnnah- and the Garlands

Andrea Piras

may have provided iconographical models ofthe investiture scenario by combining different symbols (star, sun, lunar crescent, comucopia), insignia (tiara, diadem, wreath}, gods (Nike, Artemis, Tyche, Nanaia) and crowning ceremonies (a goddess offering a ribboned wreath; an eagle with a garland). The Sogdian Yima, during the eastem propagation ofManichaeism, featured some traits ofthe Ancient lranian legend (the splendour) about the primaeval Yima, together with new aspects pointing to the royal connotations, such as the garland/diadem or the public acclamation51 (§irnäm) of Yima's great kingship (mazek axsäwan). This Yima was a Manichaen divinity corresponding to the Rex Honoris, 52 whose role of protection of heavens and earth was typified by royal traits, perfectly fitting in the Avestan glory of Yima, as an Iranization of funda­ mental concepts that Mani leamed and interpreted. The myth, name and royal fea­ tures of Yima provided then a noteworthy case of adoption and adaptation for a missionary strategy addressed to an Iranian audience (Persian, Parthian, Bactrian, Sogdian). In conclusion, the importance of Manichaean literature in its multilin­ guistic variety, offering remarkable evidence of an Iranian tradition, receives fur­ ther proof of its prominence. Not only to reconstruct the history of Iranian texts but Central Asian ones as a whole, in accordance with the methods ofcomparative stylistics of Buddhist languages, 53 to recognize the shift of meanings and mutual influences from one cultural area to another

51 Tue "good name" (sirnäm) addressed to the king Yima brings to mind the Pahlavi notion of fame / good reputation (husrawfh) of Sasanian kings, an epithet thoroughly investigated by GNOLI, who also pointed out the connection between husrawfh andx varanah- (GNOLI 1995, p. 134). 52 Cf. SKJIERV0 1995, p. 207, and 221, for the distinction between secondary Iranization, described above, and additional Iranization,which occuned as the texts were adapted, as they were distributed, to the Bast Iranian communities of Parthian, Sogdian and Bactrian languages. Cf. SKJJERV0 2009, p. 270 as well. 53 I do reference to the methodology of Pinault (PINAULT 2003,p. 53) with regard to Uigur, Tocharian and Sogdian. Further languages such as Sanskrit, Prakrit, Chinese, Khotanese and Tibetan,should equally be worth considering.

321

ABBREVIATIONS APAW

AoF AOASH BSOAS

EW JIAAA SOL SIAL Stlr TIES

Abhandlungen der Preussischen Akademie der Wissenschaften Altorientalische Forschungen Acta Orientalia Academiae Scientiarum Hungaricae Bulletin of the School of Oriental and African Studies East and West Journal of Inner Asian Art and Archaeology Studi orientali e linguistici Studies in Inner Asian Languages S tudia Iranica Tocharian and Indo-European Studies

R.EFERENCES S. AsHTON HARVEY: Scenting Salvation. Ancient Christianity and the Olfactory Imagination. Berkeley-Los Angeles 2006 (The Transformation of the Classical Heritage 42). H.W. BAILEY: Dictionary of Khotan Saka. Cambridge 1979. M. BoYCE: The Manichaean Hymn-Cycles in Parthian. London 1954 (London Oriental Series 4). -A Reader in Manichaean Middle Persian and Parthian. Teheran-Liege 1975 (Acta Iranica 9). P. BRYDER: The Chinese Transformation of Manichaeism. Löberöd 1985. L. CLARK: "The Manichaean Turkic Pothi-Book." In: AoF9 (1982),pp. 145-218. - Uygur Manichaean Texts. Texts,Translations,Commentary. Volume II: Liturgical Texts. Tumhout 2013 (Corpus Fontium Manichaeorum, Series Turcica II). K. ERDMANN: "Die Entwicklung der säsänidischen Krone." In: Ars Jslamica 15-16 (1968), pp. 87-123. I. GERSHEVITCH: The Avestan Hymn to Mithra. University of Cambridge Oriental P ublica­ tions, 4. Cambridge 1959. B. GHARIB: "New Light on Two Words in the Sogdian Version of the 'Light Paradise or the Realm of Light'." In: Studia Manichaica, ed. R.E. EMMERICK, W. SuNDERMANN, P. ZIEME. Berlin 2000,pp. 258-269. M. GHOSE: "Nana: The 'Original' Goddess on the Lion." In: JIAAA 1 (2006) ,pp. 97-112. G. GNOLI: "On Kushan and Avestan Yima." In: Archaeologia lranica et Orientalis. Mi­ scellanea in honorem Louis Vanden Berghe, vol. II, ed. L. DE MEYER, E. HAERINCK. Gent 1989,pp. 919-927. - "Cosroe dall 'Anima inunortale o della doppia felicita." In: Un ricordo ehe non si spegne.

Scritti di docenti e collaboratori dell'ütituto Universitario Orientale di Napoli in memoria di Alessandro Bausani. Napoli 1995 (Series Minor 50),pp. 119-146. - "Un emprunt iranien en arameen ignore jusqu'a present." In: Au carrefour des religions. Melanges offerts a Philippe Gignoux. Bures-sur-Yvette 1995a (Res Orientales 7), pp. 87-93. -- "Farn als Hermes in einer soghdischen Erzählung." In: Turfan, Khotan und Dun­ huang. Vortrage der Tagm)g "Annemarie von Gabain und die Turfanforschung," ver­

anstaltet von der Berlin-Brandenburgischen Akademie der Wissenschaften in Berlin (9.-12.12.1994),ed. R.E. EMMERICK,W. SUNDERMANN,I. WARNKE,P. ZIEME. BBAW, Berichte und Abhandlungen, Sonderband 1, pp. 95-100. Berlin 1996.

322

Andrea Piras

X varnnah- and the Garlands

- "Note ku�ä:t).ä. A proposito di una recente interpretazione di Pharro." In: La Persia e l'Asia Centrale. Da Alessandro al X secolo. Roma 1996a (Atti dei Convegni Lincei 127), pp. 685-702. - "Farr(ah)." In: Encyclopcediairanica 9, ed. E. YARSHATER. New York 1999, pp. 312-319. - "Nuove note sullo huamah-." In: Griente e Occidente. Convegno in ricordo di Maria Bussagli, ed. C. SILVI ANTONTh1, B.M. ALFIERI, A. SANTORO. Pisa-Roma 2002, pp. 104-108. - "Some Notes upon the Religious Significance ofthe Rabatak Inscription." In: Exegisti monumenta. Festschr(ft in Honour of Nicholas Sims-Williams, ed. W. SUNDERMANN, A. HINTZE, F. DE BLOIS. Wiesbaden 2009 (Iranica 17), pp. 141-159. F. GRENET: "Notes sur le pantheon des Kouchans." In: Stlr 12 (1984), pp. 253-261. - "Yima en Bactriane et en Sogdiane." In: Yama/Yima. Variations indoiranienne sur la geste mythique, ed. S. AZARNOUCHE, C. RtDARD. Paris 2012, pp. 83-94. W.B. HENNING: Ein manichäisches Bet- und Beichtbuch. Berlin 1936 (APAW 10). - "The Book ofthe Giants." In: BSOAS 11, 1 (1943), pp. 52-74. - "Sogdian Tales." In: BSOAS 11, 3 (1945), pp. 465-487. --- "A Sogdian Fragment of the Manichaean Cosmogony." In: BSOAS 12, 2 (1948), pp. 306-318. D. HITCH: "The Kuchean Hymn in Manichaean Script." In: TIES 6 (1993), pp. 95-132. H. Hu:MBACH, P.O. SKJt'ERV0: The Sassanian Inscription of Paikuli 3. 1. Wiesbaden 1983. H. HUMBACH, P.R. ICHAPORIA: Zamyäd Ya§t. Wiesbaden 1998. H. HUMBACH: "Yima/Jamsed." In: Varia Iranica. Orientalia Romana 7, ed. C.G. CERETI, B. MELASECCHI, F. VAJIFDAR. Roma 2004 (Serie Orientale Roma 97), pp. 45-58. A. HINTZE: Zamyad Yast. Wiesbaden 1994 (Beiträge zur Iranistik 15). J. KELLENS: "Yima, magicien entre les dieux et les hommes." In: Orientalia J. Duchesne­ Guillemin Emerito Oblata. Leiden 1984 (Acta Iranica 23), pp. 267-281. - "De la naissance des montagnes a la fin du temps: le Yast 19 ° ." In: Annuaire du College de France 98 (1997-1998), pp. 737-765. D.N. MAcKENZIE: "Two Sogdian HWYDGM'N Fragments." In: Papers in Honour of Pro­ fessor Mary Boyce II. Teheran-Liege 1985 (Acta Iranica 25), pp. 421-428. M. MAYRHOFER: Etymologisches Wörterbuch des Altindoarischen II. Wiesbaden 1996. G.-J. PINAULT: "Contacts linguistiques en Asie Centrale a la turniere des textes tokhariens." In: Indien und Zentralasien. Sprach- und Kulturkontakt, ed. S. BRETFELD, J. WILKENS. Wiesbaden 2003 (Veröffentlichungen der Societas Uralo-Altaica 61), pp. 45-83. - "Bilingual Hymn to Mani. Analysis ofthe Tocharian B Parts." In: SIAL 23 (2008), pp. 93-120. A. PIRAS: "A proposito di antico-persiano siyäti-." In: SOL 5 (1994-1995 [1996]), pp. 91-97. - "The 'Quiet ofWorship'. An Avestan-Manichaean Middle Persian Parallel." In: EW 49 (1999), pp. 281-284. -- "Sulla gioia e sul diadema. Interazioni culturali fra zoroastrismo, manicheismo e Vicino Oriente." In: Varia Iranica. Orientalia Romana 7, ed. C.G. CERETI, B. MELASECCHI, F. VAnFDAR. Roma 2004 (Serie Orientale Roma 97), pp. 183-211. - "Serse e la flagellazione dell'Ellesponto. Ideologia avestica e conquista territoriale achemenide." In: Studi iranici ravennati I, ed. A. PANAINO, A. PIRAS. Milano-Udine 2011 (Indo-Iranica et Orientalia, Series Lazur 1), pp. 111-136. - "Note di epigrafia iranica. L'iscrizione persepolitana di Serse (XPf 30-3 7) e alcuni con­ fronti testuali." In: OCNUS 20 (2012), pp. 271-278. E. PROVASI: "Sogdianfarn." In: Religious Themes and Texts of pre-Islamic Iran and Central

61h December 2002, ed. C.G. CERETI, M. MAGGI, E. PROVASI. Wiesbaden 2003 (Beiträge zur Iranistik 24), pp. 305-322. _ "Testi medio-iranici III. Parabole." In: ll Manicheismo, volume III. Il mito e la dottrina. Testi manichei dell'Asia Centrale e della Cina, ed. G. GNOLI with A. PIRAS. Milano 2008, pp. 111-157 (translation); pp. 374-411 (commentary). c. RECK: "Die Beschreibung der Daenä in einem soghdischen manichäischen Text." In: Religious Themes and Texts of pre-Islamic Iran and Central Asia. Studies in honour of Professor Gherardo Gnoli on the occasion of his 65th birthday on 61h December 2002, ed. C.G. CERETI, M: MAGGI, E. PROVASI. Wiesbaden 2003 (Beiträge zur Iranistik 24), pp. 323-338. J.R. RussELL: Zoroastrianism in Armenia. Cambridge Mass.-London 1987 (Harvard lranian Series 5). P.0 SKJJERV0: "Thematic and Linguistic Parallels in the Achaemenian and Sassanian In­ scriptions." In: Papers in Honor of Professor Mmy Boyce. Teheran-Liege 1985 (Acta Iranica 25), pp. 593-603. - "lranian Epic and the Manichaean Book of Giants. Irano-Manichaica III." In: AOASH 48 (1995), pp. 187-223. - "Reflexes ofIrani an Oral Traditions in Manichaean Literature." In: Literarische Stoffe und ihre Gestaltung in mitteliranischer Zeit, ed. D. DuRKIN-MEISTERERNST, C. RECK, D. WEBER. Wiesbaden 2009 (Beiträge zur Iranistik 31), pp. 269-286. S. STARK: "Some Remarks on the Headgear of the Royal Türks." In: JIAAA 4 (2009), pp. 119-133. L. VANDEN BERGHE: "Les scenes d'investiture sur les reliefs rnpestres de l'Irän ancien: evo­ lution et signification." In: Orientalia Iosephii Tucci Memoriae Dicata 3, ed. G. GNOLI, L. LANCIOTTI. Roma 1988 (Serie Orientale Roma 56), pp. 1511-1531. R. VARDANYAN: "Tendenze culh1rali e ideologiche nell'impero partico riflesse nella mon­ etazione." In: Parthica 3 (2001), pp. 25-132. G. VERARDI: "The Ku�ä:t).ä Emperors as Cakravartins. Dynastie Art and Cults in India and Central Asia: History of a Theory, Clarifications and Refutations." In: EW 33 (1983), pp. 225-294.

Asia. Studies in honour of Professor Gherardo Gnoli on the occasion of his 651h birthday

011

323

Euo PROVASI Bologna

Some Notes on Sogdian Phonology: Prothetic Aleph and Labialised Velars* sahr-i I'aznfn na hamän ast ki man dzdam pär cijittäda stki imsäl digargun suda kär raft u mä-rä hama becära wu darmanda bi mänd man na dänam ki ci darmän kunam fn-rä wu ci kär (FmruxI SistanI)

The writing systems of semitic origin which where commonly used for Sog­ dian, with their lack of markings for short vowels (especially short ä) often make the task of ascertaining the number of syllable of a given word difficult. Even in Sogdian texts written in Syriac script the vocalic points are used very sparingly. Here ptäl.J_a (dot above and below, for short /a/) and zqäpä (two dots above for long /ä/) 1 are used only sporadically, and, on one hand, rwäl.J_a (dot above wäw for /o/) and alä�ä (dot below wäw for /u/), and, on the other hand, assäqä (two 26 (< *hwähaya-), 27 e.g.: (24) /pxway/ "to cut off'' (< *apa­ hwiihaya-, Yaghn. piixöy): 28 (a) S pxw'y JJ [Padm /42/] vs. (b) S pxw y'nt w [P2 /513/], M pxww y [BBB a /7/], C [p}wx y [E5.28 /r/11/]; (25/1) /sx äy/ "to lift" (< *uz-hwähaya-, Yaghn. suxöy-) (cf. GMS §§ 417, 627): (a) S 'sxw y [P3 /262,267/; So 10129 /I/v/12/],'sxw y y [Dhy /24/J),M 'sxw y [M 7420 /16/] vs. (b) S sxw y [P3 /51/; P3 /261/; Huy 36.1 /I/v/8/J, sxw y y [Padm /43/; VJ /1465/], sxw yt [PBuch /116/], sxw 'ymk'm [TaleH /14/] vs. (c) C swx y 'z [E27.12 /v/12/],swx YY: [E5.125-I26 /I/r/16/],swx y-ymsq [E5.125-126 /I/r/20/], sw[x] yt (/sxwäyät/) 3sg. opt. [E5.127 /r/10/],and also in the pret. stem ofthe pre­ ceding verb, (25/2) /sxwast-/, p .p. /sxwaste/, past inf. /sxw asti/): (a) S 'sxwst 'kt [So 20220 /II/r/1/],'sxwsty [So 18196 /r/9/], vs. (b) S sxws[ty} [So 12601 /r/9/], M sxwst [M 815 /v/2/J,29 sxwst[y] [M 1134+3611 /v/8/],sxwstyy [M 1204 /r/6] vs. (c) C swxsty [E5.72 /v/13/], swxsty. [E6.6 /r/4]; in the two preceding examples, both the prothetic aleph in S script and the alternation between (S M) and (C) point to initial clusters respectively of/p/ or /s/ followed by a labialised /xw/;30 (26) /ptxwäy/ "to kill"(< *pati-xwiihaya-) (Yaghn. ti7xöy-) (cf. GMS § 226): (a) S ptxw '.v- (passim), Mptxw ,yy [M 127 /v/9/],ptxw y [M 760 /v/5/],pfxw'·J/t [BBB e /5/J, pfxw yn [M 648 /A/6/], C ptxw y [E28.23 /r/16/], ptxw yt [E28.23 /v/21/], ptxw Jit' [E5.125-126 /I/v/12/], ptxw ynt [E23 /403/],ptxw yntq' [E5.75 [E27.69 /r/19/]. The /I/v/3/] vs. (b) C ptwx y [E5.75 /II/r/4/,E27.7 /v/18/],ptwx tihöy, 31 which is Yaghn. cf. but aleph, prothetic with written never is last verb probably a reduction of/ptxwäy/ with simplification ofthe initial 3-consonant group (rather then /patxwäy/ with unexplained loss of the first syllable). Before secondary /ä/ < *aha in initial position we have: (27) /xwär/ "sister" (Av. xva-,:;har-, Yaghn. x( 1jör) (cf. GMS § 227): S xw'rh [P2 /149/; Or 8212/1793 /2/; KG.3.3 /451/]; xw'r'ysth pl. [P2 /927/],M xw'r [M 760 /v/2/],C xw'r [E5.91 /r/16,22/; E27.68 /v/19/; E28.27b /r/4/],and from *h and *w in secondary contact after the fall of a short vowel: (28/1) /xwäkar/ (< *wxiikar < *wahakara-) "mer-

chant" (cf. GMS §§ 392,422): S xw' 'kkr [Vim /157/],xw'krt pl. [So 14000 /8/], M xw'qr [TaleD /9, 15, 29/], with the extended *-aka- form (28/2) S xw'kkr'k [TaleK /27/],xw'kry [Mug A9 /r/2/],xw'krytt pl. [So 14000 /19/] and the derived abstract (28/3) C [x]w'qry' [E5.125-126 /I/r/3/]. Let us now examine a few examples of before an original short vowel. In some ofthe following cases there are clues (such as prothetic aleph,graphical metathesis or other) to the possibility of a labialised /xw/. In initial position,before an original short vowel *a, e.g.: (29a) /xwar:/ "to eat" (< *hwara-) (Yaghn. x war-): S M C xwr- (passim), and the noun (29b) /xwart/ "food"(Av. x"arati-): S xwrt, M xwrt (passim), C xw 'rt [E6.3 /r/19]; (30) /xwaßn­ '; "sleep,dream'' (< Av. xvafha-, Yaghn. xuvn) (cf. GMS § 312): S xwßny [P2 (3x); SCE /57/; P6 /4/; P7 /66/; Vaj2 /4/; TaleI /39/],xwßnw [VJ /3,4/],xivßn '[P6 /112/; Rustam /26/],xwßnyh [P2 /409/; P6 /109/],M xwßnyy [BBB /695/; M 1088 /r/5/], xwßny [M 5606 /r/ii/5/], C xwbny' [E27.48 /v/9/]; (31) /x"'fäw/ "lord" (< *hwatiiwä): 32 S xwt'w (passim), with obl. xwt'wy [xwt'w JJ in Mpn3 /v/9/], voc. xwt'w ',pl. dir. xwt 'wt, pl. obl. xwt'wty, and the abstract xwt'wy' "lordship" [L71 /7/] (but in one case S xt'wty pl. obl. [Bez C /12/],Mxwt'w-xwt'w (passim), with pl. dir. xwt'wt [TaleB /46/] (but in one case M xt'w [M 6330 /r/1/J), C xwt'w (passim), with pl. dir. xwt'wt [E26.5 /v/9,15/] and pl. obl. xwt'wty [E26.27 /r/28/]. In some Chr. texts this word is written with voiced : C xwdw [E8 /r/5/: E24.1 /r/2, 7/; E24.7 /v/10/; E24.9-10 /I/v/2/; E24.9-10 /I/v/10/], xwdw' (voc.) [E24.6 /v/5,10/],xwd 'w' [E26.28 /v/19/]. MARTIN SCHWARTZ compares the voicing of/t/ in this word (as well as in xwdy "self ') to the one appearing after /r/ in mrdxmy. 33 Although this word never appears to be written with prothetic aleph, both the ab­ sence of w after x in M xt 'wand S xt'wty [Bez C /12/]34 and the voicing of lt/ in Chr. xwd'w [xwdäw] point to the possible absence of a vowel between /x(w)/ and /t/,hence a form /xwtäw/. Cf. also,in Arabic script, xdynh (for Sogd. :xwt(}yn(h)) "queen" (HENNING apud GMS § 133 n. 1).35 In initial p�sition,before an original short vowel *a umlauted to /i/: (32) /xwin­ / "to be called" (pres. mid.) (< *hwanya-): 36 S xwynty 3sg. [So 16186 /4, 9, 14/; Lll + So 20248 + So 20164 /22/], xwyn'nt 3pl. [So 10305+10660+18273 /v/3/], M xwyndyy 3sg. [M 549 /II/v/25/]. Before an original short vowel *a, in medial position: (33) /saxw n 'l "word, speech"(< *sahwan-) (cf. GMS §§ 38 n.1,222,417): (a) S sxwnw [VJ (passim);

yw

The uncompounded stem is found in S xw' 'y [So 18400 /r/18/] /xwäy/ "breaking (of the law)" (pres. inf.). 27 On which cf. HENNING 1937, pp. 58-59. 28 Cf. HENNING 1937, p. 59 on 505. 29 P.p. without ending [= Angad Rösnä.n Ia.12b (at the end ofthe verse)]. 30 The Christian forms could be altematively be interpreted as representing a later develop­ ment /pxw/ > /pux/ and /sxw/ > /sux/ as inYaghnobi, but on the other handYaghn. forms such as suxoy- and similar ones are probably best interpreted phonematically as containing a labialised /xw/ and a non-phonemic anaptyctic vowel (thus, /sxwoy/ [s�xwny]). In fact, in similar cases in 11 ANDREEV's transcriptions from Yaghnobi, the ultrashort u altemates with zero, cf. e.g. p xoy-, w pxoy- (ANDREEV, PESCEREVA 1957, p. 310), t"xoy-, txoy- (ibid., p. 339). A phoneme /x / is present in the phonological system ofYaghnobi (cf. XROMOV 1972, pp. 13, 14), but at the phonetic level it is (facultatively?) delabialized (or perceived as such by investigators) before the back low 11 11 half-rounded vowel /o/ [n] (< *ä); cf. x11at "self' vs. the altemations x ar vs. x orta - xorta "to 11 eat" and x or -xor "sister" (ANDREEV, PESCEREVA 1957, pp. 364-365). 31 See the preceding note. 26

,,

1

333

Elio Provasi

332

Cf. GMS §§ 38, 131, 133 n.1, 222, 269. STSC2, p. 151: " ... the consonantism of xwd 'w { C 49 (T II B 40 No. 1) /r/5/}, which con­ trasts with the more usual Chr. spelling xwt 'w, Man. xJ 'w, S. ywt 'w, probably indicates a genuine Sogdian dialectal variation. The text group STii 1-3 has d for Olr. lt/ after /xu/ not only in xwdw but also in xwdy 'self' (and after r in mrdxmy and msydrd). Note also, in Arabic script, xoynh 'lady' B. ywt 'ynh, Chr. xwtyn (BSTii), on which see Henning apud GMS § 133 fn." 34 On the delabialisation of /xw/ cf. below. 35 Cf. also the proper name in Hebrew characters lr1-vdynq [Xudenakk] in the newly discov­ ered Judaeo-Persian letter from Dandan Öyliq [DU II /3/, see ZHÄJ-TG, SHi 2008, pp. 83, 84, 89]. 36 Cf. Khot. hvafi- (EMMERICK 1968, p. 156). 32 33

Elio Provasi

Same Notes an Sogdian Phonology: Prothetic Aleph and Labialised Velars

Mpn3 /r/5/; PBuch /132/; KrIV.879.4263 /36/], sxwn [ALett II /r/30/,ALett III /18, 22/J,37 sxwnt' pl. [PBuch /134/],M sxwn- (only in sxwn tnb 'r [M 14 /v/7, 25] < WMir.) vs. (b) S sxnw [TaleJ /4/] vs. (c) S swxny [P2 /1036/], C swxnw [E29.4 /v/1,17/]. Once written in Sogd. script sxnw [TaleJ /4/],without waw either before or after . If this is not a scribal error (which I think improbable), it could be indicative ofthe delabialisation of /xw/ in contact with a following consonant (cf. xwt \v above). Initial /xw/ can also develop from initial *ux, where the labialisation is trans­ w ferred to the consonant and the short vowel is deleted: (34) /x star'-/ "camel" (< *uxstra-, Av. ustra-) (cf. GMS §§ 36, 162,257,428,483,497): (a) S 'xwstry [VJ (4x); P2 /771/; P3 /183/; SCE /176, 371/], 'xwstr'y [SCE /288/], 'xwstrw [ALett I /r/11/; VJ /163,d56/], 'xwstr' [P3 /184/] and S 'xwst'ry [VJ /41 6,c13, 38 1165/],'xwst 'r 'yh [VJ /86/],*'xw t',yh [VJ /618/] vs. (b) M xwsfryy [M 568 prothetic /r/5/],>.-wstryyb [M 127 /v/10/]. The altemation offorms with and without w/,rather than /x labialised a as group initial ofthe tion interpreta the aleph supports a progressive metathesis across /xi (contra GMS § 428). Labialisation of /xw/ in the original group *xsw appears in the numeral (35/1) 39 /xwsuJ "six" (< *xswasam, cf. Yaghn. uxs): (a) S 'xwsw [Dhu /14/; Mug A4 /r/1/; Mug BI /19/; Mug A7 /3/] vs. (b) S wxwsw [VJ (23x); P2 (6x); SCE /119,121/; P6 /131/; P7 /53/; P9 /30, 35/; Vim /15/; Dhu (3x); Dhy /117/; KG.3.3 /489/; So 18400 /v/6,13,16,25*/; L93 /13/; P. Chinois 3134 (7x); Mug Novl Ir/ (3x); Mug AS /3/J vs. (c) C xwsw [ES.127 /v/3/J,40 as well as in the two forms ofthe ordinal "sixth:" (35/2) /xwsmI/: (a) S 'xwsmy [T II Tm /6/ (cit. GMS § 1331 n. 23)] vs. (b) S wxsmy [So 18311 /4/] vs. (c) S wxwswmy [P2 /1094/; P8 /166/; Xian /13,18/] and (35/3) /xwsrnlk/: (a) S wxwswmykw [So 18248 /II/v/22/] vs. (b) S wxsmy-k [Huy 36.9 /v/10/],wxfoiykw [PBuch /92/; So 10202 /8/] vs. (c) Cxwsmyq [E27.39 /v/4/], xwsmyqy [E27.68 /v/11,13/]. In the word for "parasang" the labialisation of in final position which ap­ pears in two cases is probably ofsecondary origin. The spellings point to forms �s different as /fsäi)X and sometimes altemate in the same text,e.g. in the Dhüta-sütra. 41 The M formfns 'x shows an occasional metathesis /f(a)sm.).xw/ > /fal)sax(w)/. 42 Cf. GMS § 1092 and GERSHEVITCH 1946a, p. 146. 37

38

335

43 Cf. GMS §§ 11, 236 , 391; HENNING 1937, p. 86 on 73 0. According to the latter, it is per­ haps a loan from Middle Persian (cf. ManMPers i5x, ZorMPers. axw). 44 I.e. /caxwmanI/, a contraction or allegro-form of the hendiadys. 45 t 'nxw(h) "äme, coeur" according to BENVENISTE (1946 , p. 94 on 18d), has been shown by HENNING to be read 'zxw(h) "anxiety" and connected with Av. qzah- (HENNING apud GMS § 403 1). This is then possibly a compound /al)zaxw/ from < qzah- + *ahwä-. Tue lack of after the initial aleph probably points to the substitution of the diphthong with a nasa1ised vowel (/azaxw/). 46 I.e. [ßzäx"kw], where the final group is disjoined by an anaptyctic vowel and the labiali­ sation is transfen-ed to the last consonant (/xwkf > /xkw/). 47 I.e. [ß"zäxCwlk], with a labialised anaptyctic vowel after /ß/ and the labialisation of /x/ left unmarked. 48 Both the adjective /ßzäxwk/ and the abstract /ßfäxwtya/ derive from an unattested non­ suffixed form */ßfäxw/. �9 This form, written withput interestingly shows a delabialisation of /xw/ (cf. the spellmgs of /xwtäw/ and /saxwn-/ above). 50 Tue spelling of the abstract�ending as -ty '(without final -k) in Mßj 'xwty 'band C bi 'xwry ' shows that the preceding stem is heavy (cf. SIMS-WILLIAM:S 1981b, p. 12). 51 The spelling indicates a reduced pronunciation of unstressed short /a/: [sät;:lxw].

337

Elio Provasi

Same Notes on Sogdian Phonology: Prothetic Aleph and Labialised Velars

[M 178 /I/r/8/], Cs >twxy' [E27.51 /r/3/, /v/6/], s )hvxy: 'obl. [E27.57 /r/8/]; (41/1) /sänaxw/ "excellent:"52 S s 'nwx [Or 8212/89 /2/], Ms'nwxt (pl.) [M 617 /II/r/9/], Cs'nwx [E27 (2x)J, with the abstract (41/2) /sänaxwya/ "excellence:" Cs 'nwxy' [E26.5 /r/4/; E27.C /v/3/]. The fo1ms for "blood" are problematical and cannot be derived from a single prototype, but all the graphic alternations point to a labialised velar: (42) /yixwn-/ - /yixwrn-/ ,..., /waxwrn-/ ,..., /xwrn-/ "blood" (< *wahunf- X *·wixrun-, Av. vohunfla-) (Yaghn. ivaxm - waxn) (cf. GMS §§ 208, 209, 350, 360,391,417): (a) M only in the compound ywxn '- psyyk [M 549 /II/v/4/] "blood-spilling," C ywxny nom. [E26.31 /v/14/], ywxnw acc. [E27 (2x)J, ywxny. gen. [E27.57 /r/19/] vs. (b) S yxwnw [P22 /15/], Myxwny [BBB f/59/] vs. (c) Myxwrnw [M 5912 /v/1/] vs. (d) S wxrnw [P2 /569/J, wxrnwh [P2 /476/], wxrnyh [P2 /454/]) vs. (e) S xwrny [P2 (2x); P22 /16/], xwrn y [P2 (2x)J, xwrnw [VJ /278, 783/], xwrn' [P2 (2x)J; M xwrnyy [BBB /763/; ManLett 1 /21, 29/; M 141+6795 /v/12/], xwrny[y] [M 834 /I/r/11/],xwrn[yy] [ManLett 2 /13/J, xwrny[yb] [M 6450 /II/r/i/6/]; C xwrny [E6.7 /v/10/; E23 /318/]. This is always a light stem. Probably the different written forms represent (as SCHWARTZ suggested) different spoken forms, deriving from a cross­ ing of*wahunf- (Av. vohunf!a-) with *wixrun- (Av. xrüniia- etc.)). 53 At least some ofthe spellings (such as the variation between forms (a) and (b) could indicate a labialised velar /xw/. The final -m ofYaghn. waxm is probably a reflex ofa transfer ofthis labialisation to the final nasal. The labialised velar fricative /xw/ probably originated also from a merger with the second element ofa preceding diphthong au, as in the adjective (43/1) /raxwsn'­ / "luminous" (Av. raoxsna-, Yaghn. raxsin "dawn") (cf. GMS §§ 36,40,127, 428): (a) S rxwsn [SCE /76/],54 rxwsny [P3 /147/; SCE /495/; P6 /170/; P8 /23/; P9 /137/; Padm /14/,28; Dhu (14x); Dhy (15x); Mpnl /1/; KG (5x); So 18248 /I/v/22/; Huy 36.10 /v/11/], rxwsn y [So 14195+ 141986 /r/13/; So 14197+ 14198+ 14199 /v/13/], r.nvsn' [KG.3.4 /567/; (-') Huy 36.11 /r/3/],rxwsnw [SCE /78/; Dhu /245/]; rxws'nt (pl.) [P25i /2/] vs. (b) S rwxsny [P14 /33/; Mpn2 /v/7/ (rw[x}sny); So 10650(17) /5, 15/; So 14411 /II/r/18/; So 14411 /II/v/18/; So 14638 /I/v/17/; TaleK /2/; L93 /16/; KrIV.879.4263 /8/], rwxsnt pl. [P25ii /3/], rwxsntyh pl. obl. [PBuch /121/]; rwxsnty-' [PBuch /98/J, M rwxsny [PBuch c /20/], rwxsnyy [BBB f /80/; M 14 /v/6/; M 635 /11/r/1/; M 672b /r/7/; M 715c /I/r/3/; M 1469 /r/3/], rwxsmvw [M 178 /l/r/11/], rwxsn' [BBB b (4x); M 178 /I/r/19/; M 5266 /r/10(11)/], M rwxsnd' pl. [BBB f/46/],rwxsnd'b [M 178 /I/r/5/], rwxsndt' [M 5271./v/1/], rwxsnt' [M 6330 lr/4/], rwxsnt'b [M 684 /v/6/], C rwxsny [E23 /165/; E26.3 /v/4/; E26.6 /a/3/].55

The same alternation ofspellings with and can be observed in the com­ parative (43/2) (a) S rxwsntr [Huy 36.16 /r/7/]-vs. (b) S rwxsntry [Talel /34/], M rwxsndryy [M 264a /v/9/]56 and in the derived abstract (43/3) /raxwsnyäk/: (a) S rxwsny'k [P6 /139/; P9 /57/; Dhy /148/], -h [P9 /64/; Intox /5/; Dlm /102/; Dhy /219/; KG.3.2 /376/; Huy 36.3 /v/5/; L80 /7/],-y [Huy 36.7 /v/15/] vs. (b) S mxsny 'k [P23 /12/; So 14411 /11/v/9/], nvxsny ''k [L93 /6/], -h [Vim /12/; Ch/So 101001 /v/18/; PBuch /28/], - 'kw [So 14411 /11/v/3,8/],- 'ky [PBuch /97/J, M rwxsny' 'q [M 14 /r/25/,/v/19/],rwxsny'k [M 178 /I/v/11/; M 129 /v/9/; M 583 /11/r/h/; BBB /485/], -y(v) [M 129 /v/12/; M 674 /v/7/], C rwxsny 'q [ES (passim); E27 (passim); E23 /217/], rwx§ny'qy obl. [E5.48 /r/16/], as well as in the word for "paradise:" (43/4) (a) S rxwsn'yröm'nwh [P8 /34/], rxwsn 'yrJmnwh [VJ /821, 829, 1041, 1234/], rxwsn'yrJ[mnh] [Huy 36.1 /v/13/], rxwsn'yrömnyh [So 20165 /15/] vs. (b) S rwxsn 'Jwbmnwh [TaleG /13/; Or 8212/82 /8/], rwx§n 'yrbmnyh [Or 8212/84 /5/], M rwxsn'yrömn [M 178 (3x); KG.25.1 /2338*/],rwxsn 'yrbmn �v [M 178 /I/r/11/]. The fact that this is a light stem indicates tlmt the assimilation *aux > /axw/ preceded and prevented the normal monophthongisation *au> !öl (which occurred, e.g., in words like ßwb(bh) /ßö8/ "perfume" (Av baobi-, Yaghn. vüd), yws lyösl "ear" (Av gaosa-, Yaghn. yüs), rwt /röt/ "river" (OPers rautah-, Yaghn. rayt). On the other hand,/taxm-/ "seed" (a light stem) is always written without w; 57 cf. (44) /taxrn/ "seed" (. u 5IUl JUNnRö reuMJllj=_x��, WATANABE KA10Y0K . -1932 Tökyö 1924 terbuch. Sprachmaterial der vorislamischen UigWb = K. RöHRBORN: Uigu risches Wör Wiesbade n 1977-1998. türkischenTexte aus Zentralasien , Lief. 1-6. of the Mahäyäna Mahäparinirväf/,asütr a in D.A. UTZ: "An Unpublished Sogdian Version . diss., Harvard University. Cambridge, Ph.D the German Turfan Collection," unpubl. Mass., 1976. e Dogmatik aus chinesischen und iranischen E. WALDSCHMID T, W. LENTZ: "Manichäisch Texte." In: SPAW 1933, pp. 480-607. im Mitteliranischen. Göttingen 1970 (DisserD. WEBER: Die Stellung der sog. Inchoativa tation Göttingen). Volume 2. Winona n ologies of Asia and Africa, - "Pahlavi Phonology." In: A.S. KAYE: Pho Lake 1997, pp. 601-636. . A.L. XROMOV: Jagnobski j jazyk. Moskva 1972 & Journal of Asian and Af rican Studies 1 In: es." nitiv Infi Y. YosHIDA: "On the Sogdian (1979), pp. 181-195. hibun no sogudo-go han ni tsuite" -/J 7 /\Jv YOSHIDA YUTAKA aEB�: "Karabarugasun asun ["On the Sogdian Version of the Kara Balg -jj' J.. :/fi�xO) �) 7" �'M!MEU.:.-:::) �) -C n Ajia kenkyü [Bulletin of the Societyfor West­ lnscription"]. In: "E§ffi 7 s.P Tfiff:JE Seina 2. niversity] 28 (1988), pp. 24-5 ern and Southern Asiatic Studies, Kyoto U ungakubu shozö no b aku daig to Kyo jo mon o - "Mujö o toku manikyö sogudo-g � - :gffl5*�x�f:fß?fr}ijOJ shashin shüyö kara" �'rt �äi < i (WERBA 2006, p. 289 fn. 137); progressive weakening of word-final syllables (KORN 2013, pp. 84-85); *s > h (TREMBLAY 2005a, p. 3; LIPP 2009/I, pp. 318-122; FRANCFORT, TREMBLAY 2010, p. 90; ÜETTINGER 2015, p. 57); st-st (LIPP2,009/1, pp. 318-122); stop,..., fricative altemation (KIM 2010); for /8äi/ (lTTZES 2014, p. 363); /E)r/ > /hr/ (SCHMITT 2006, p. 100 fn. 120, 142; TAVERNIER 2011b); *Br> hr (TAVERNIER 2009, p. 212); vowel elision (TAVERNIER 201 la, p. 239-240); /xw/ > !xi (TAVERNIER 2013a, p. 277). The correspondences 'OP phoneme -AE grapheme' have been studied by TAVERNIER 201 la. (For phonetic/phonemic adaptations ofOP words to the phone ­ mics of the single languages spoken in the Empire cf. also the relevant treatments contained in the works quoted at 4.2 below).

3.3 Morphosyntax.

ä- (JANDA 2009, pp. 151-152); accusative subject (D ANESI 2014, pp. 224 ff., 250 ff.); a-band- (SCHWEIGER 2005a, p. 582); ta-band- (SCHMITT 2009b, pp. 425427); akunaj (SKJJERV0 2011a, p. 328b); alignment in OP (HAIG 2008, pp. 23-88); aniya- (indef.) (FILIPPONE 2016, pp. 41-42); bav- + pred. (LINCOLN 2015, pp. 136137); belong-constmctions (BEN VENUTO 2012); caxriyä (KIM 2010, p. 173 fn. 15); collocation of attributive genitive (SCHMITT 2014b, p. 204); dative-genitive syn­ cretism (BEN VENUTO-P OMPEO 2012); dativus (in)commodi (?) (SCHMITT 2016a, p. 110); demonstratives (FILIPPONE 2005); t-di ("postpon. emph. Partikel," WERBA 2006, p. 282 fn. 103); enunciative structure (SCHMITT 2014b, pp. 205-207); ex­ pression of goal (PoMPEO 2008); functions of the genitive (PoMPEO, BEN VENUTO

3.4 Word Formation. -ana- (JANDA2009, pp. 151-152); composition (in an Indo-Iranian perspective) SADOVSKI (2007 and 2013, with lit.); outcomes ofI A word-formative *-ka- in Av and OP (CIANCAGLINI2012a, CIANCAGLINI 2012b; I TTZiis 2014, pp. 364-365); -u­ adjective vs. -ra-adjective (SCHMITT 2012a, pp. 356-358); vrddhi derivation (I TTZES 2014, pp. 362-367); *xsay-aßa- (SCHMITT 2014b, p. 205); see also § 4.2 below. 4.1 Lexicon. SCHMITT 2014 is to be systematically consulted for any OP lexeme, and therefore not quoted in what follows. Specific lexical fields/series have been recently treated, such as month names ( DE BLOIS 2006), colours (Ross1 2006a), borrowingsOP > AA (SHAKED 2005, NAVEH, SHAKED 2012).One should also refer to the etymological dictionaries with markedOld Iranian interests, in primis EWA; on the - courageous - enterprise regarding the first everIranian etymological dic­ tionary, whose third (R..AsTORGUEVA, EDEL'MAN 2008), fourth (EnEL'MAN 2011) and fifth volumes (EDEL'MAN 2015) fall within the period under examination cf. the (mostly shareable) remarks by TREMBLAY 2005a, p. 13. CHEUNG 2007 is a useful repertoire ofIr. verbal families including more than 100 OP verbs and nouns (but cf. SCHMITT2008b, 312: "Die Darstellung der altpersischen Verben ist absolut un-

364

365

Adriano V. Rossi

Ten Years ofAchaemenid Philology

befriedigend, denn die Altpersisch-Forschung des letzten halben Jahrhunderts ist kaum zur Kenntnis genommen worden;" only original etymological proposals by the author are entered below). The following list contains specific comments on single words and/or bases (with the exclusion of proper names, cf. 4.2 below): *ardi-.fja- (BRUST 2008, pp. 106-107); abi-jävaya- (SKJJERV0 2005b, p. 313); *ä-c,:na- (NAVEH, SHAKED 2012, pp. 57, 182); *ädranga- (NAVEH, SHAKED 2012, pp. 55,115); *ädu- (HENKELMAN 2008a, p. 406 fn. 929); *ä-dv(a)isa(h)- (NAVEH, SHAKED 2012, pp. 98-99); *äfrasta- (NAVEH, SHAKED 2012, pp. 55, 116); ägriya­ /ä-gariya- (SIMS-WILLIAMS 2010, p. 55; HERRENSCHMIDT 2015, p. 126 fn. 1); aiva- (SCHMITT 2012c, p. 172 fn. 1); *ä-k,:sta- (NAVEH, SHAKED 2012, pp. 57, 217); *äla- (?) (SCHMITT 2006a, pp. 215-217); ämäta- (CHEUNG 2007, p. 255); amiixam (SCHMITT 2006a, p. 208); *ang-iira- (BRUST 2008, p. 36); antar (NAVEH, SHAKED 2012, p. 73); *anya-ravi(a)- (NAVEH,SI-:lAKED 2012, p. 105); *apa-kana­ (ARFA'EE 2008b, p. 264); apadäna- (STOLPER 2007, pp. 233-234; JANDA 2009; SCHMITT 2010d, p. 157; SCHMITT 2012b, p. 409; FERRER-LOSILLA 2014, pp. 72-73 fn. 187); *apärga- (TAVERNIER 2007c, (l)); apariyäya (SKJJERV0 2005a, p. 83 n. 28); arika- (SKJJERV0 2005a, p. 57, 82 n. 16; LINCOLN 2012, pp. 24-25; AMIRI­ PARYAN 2015, p. 2); ärstika (ITTZES 2014, pp. 365-366); *ardaba- (STOLPER 2007, pp. 233-234; TAVERNIER 2007a, p. 449); artäcii (AHMADI 2015, pp. 4, 14; PIRART 2015, pp. 51-52; LIPP 2009/I, p. 193 fn. 102); artävan- (BUYANER 2007, pp. 361 ff.; AHMADI 2015, p. 18; PIRART 2015, pp. 50-51); aruvasta- (SKJJERV0 2005a, p. 77); arya- (KELLENS 2005; PERRY 2010; HENKELMAN 201la, pp. 611-612; PANAINO 2015, pp. 93 ff.); aryaci$U- (SOUDAVAR 2006, pp. 170-177; SOUDAVAR 2009, pp. 443 ff.); *aspa-manga- (NAVEH, SHAKED 2012, p. 55, 223); *asp,:na- (NAVEH, SHAKED 2012, pp. 55,116); *aOanga- (FRANCFORT,TREMBLAY 2010,p. 164); *äOu­ (KORN 2009, p. 201); ava-hard- (SKJJERV0 201la, p. 327a); ava-näda- (?) (STOLPER-TAVERNIER 2007, pp. 14-15); ava-rada- (LIPP 2009/II, pp. 171-172); *ava-stäka-? (NAVEH,SHAKED 2012,pp. 55,229); *ä-v,:du- (NAVEH, SHAKED 2012, p. 182); ayadana- (JANDA 2009, pp. 151-152; ROSSI 2016); axsata- (LIPP 2009/II, p. 291; GARNIER 2013, p. 62); äxsn(a)u- (SCHWARTZ 2013, p. 69); azdä (GOTÖ 2005,pp. 200-204); *azgand-'? (NAVEH, SHAKED 2012, pp. 55,108); baga- (PIRART 2015, pp. 48-49); *bagaica- (NAVEH, SI-:lAKED 2012, pp. 58, 142, 196); *bagina (NAVEH, SHAKED 2012, pp. 56, 184; CANEPA 2013, p. 338); *bagya- (NAVEH, SHAKED 2012,pp. 36,56,184); biiji- (TUPLIN 2008, pp. 326ff,384); *bakadausiya­ /bagadauriya- (HENKELMAN 2008a, pp. 381-382); HENKELMAN 2011b, p. 104); band- (SCHWEIGER 2005a, pp. 581-582; SCHMITT 2008b, p. 312); *bänuvant-7 (NAVEH, SHAKED 2012, p. 238); baristama- (ARFA'EE 2008b, p. 266); *biit(i)ja­ ka-lbiita- (BRUST 2008, pp. 141-142); biixtri- (DE BLOIS 2013; CHEUNG 2015, pp. 53-54); 0 bigna- (CHEUND 2007, p. 2); *brazma-däna- (TAVERNIER 2007a, p. 438); brazmani- (PIRART 2015, pp. 51-52; LIPP 2009/I, p. 193 fn. 102); *ctna- (NAVEH, SHAKED 2012, p. 121); *caraka- (HENKELMAN 2010, p. 753; NAVEH,SHAKED 2012, pp. 56,181); Ci$a- (SOUDAVAR 2006, pp. 170-177); *daba- (NAVEH,SHAKED 2012, p. 72); Daha- (ScI-IWARTZ 2012a, p. 277); dahyu- (SOUDAVAR 2012, p. 71 fn. 124);

*damya- (NAVEH, SHAKED 2012, pp. 33, 57, 181-182, 210, 261); *dämya-data­ (NAVEH,SHAKED 2012, pp. 182,210); *diimya-data-kiina- (NAVEH, SHAKED 2012, PP· 55, 210); dastakarta- (JANDA 2006, pp. 84-86); *daOapati- (SIMS-WILLIAMS 2007, p. 226); *dau,;a- (HENKELMAN 2008a, pp. 212-213, 231; SKJJERV0 2013, PP· 559-560); *dau,;iya- (HENKELMAN 2008a, pp. 212-213, 231; NAVEH, SHAKED 2012, p. 184); *dausa-xwara- (NAVEH, SHAKED 2012, p. 82); didä- (ROSSI 2009a; FERRER-LOSILLA 2014, pp. 52-53 fn. 122); dipi[ ... ] (TUPLIN 2005a, pp. 224-225; ROSSI i.pr.2); drauga- (SCHMITT 2016a, p. 673); duvarOi- (MAYRHOFER 2006, pp. 22-23); duvitäparanam (AHMADI 2014); duzyära- (PIRART 2015, pp. 54-55); *fra-biira-? (NAVEH, SHAKED 2012, pp. 57, 218); *fra-käna- (NAVEH, SHAKED 2012,p. 97);/ra-mii- (SCHMITT 2016a,p. 69 fn. l);framätar- (HENKELMAN 2008a, p. 235 fn. 517; HENKELMAN 2011a, pp. 619-620); *fra-stä-van- (SMITH, MARTIN 2009,p. 35);/rasa- (SKJ.tERV0 2005a,p. 82 n. 14; SKJJERV0 2011a, p. 327a); *frasa­ (LINCOLN 2012, p. 211 fn. 41; STOLPER 2015, pp. 14-21); *frasam kar- (SCI-lMITT 2010b, pp. 393-394); *frataci- (TAVERNIER 2007a, p. 421); fratama- (TUPLIN 2005b; NAVEH,SHAKED 2012, p. 105); *frataraka- (NAVEH,SHAKED 2012, p. 74); fraO- (FILIPPONE 2007); *fravarti- (NAVEH, SHAKED 2012, pp. 36, 56, 196-197); *friiyah- (NAVEH,SHAKED 2012, pp. 56,136); gaub- (SKJJERV0 2005a,p. 82 n. 20); *gafrista- (NAVEH, SHAKED 2012, p. 143); *ganzabara- (NAVEH, SHAKED 2012, pp. 55,171); *gari-datika-? (NAVEH,SHAKED 2012,pp. 56,217); gäOu- (TAVERNIER 201la,pp. 209-210); *gau-varza-? (NAVEH,SHAKED 2012,pp. 57,212); *gäwarOa­ (ScHWARTZ 2012b, p. 99); *grastapati- (SCHWARTZ 2015); *grda-taka-? (NAVEH, SHAKED 2012, p. 217); *griva- (?) (STOLPER-TAVERNIER 2007, p. 13); *hadahra­ pati- (SCHMITT 2007c, p. 356); *haftaxva-piitä-l*haftaxva-pati- (SMITH, MARTIN 2009, pp. 49-50 and fn. 64); hainä- (PIRART 2014, p. 242 fn. 409); *hamagäOu-, *hamagiiOavii (PIRAS 2012b, p. 276); *hamärOa- (SCHWARTZ 2012b, p. 99); *hamhiiga-(MlJTZAFI 2014, p. 60); hami$ya- (DE BLOIS 2006, pp. 49-51,BASELLO 2012b, pp. 95-97); hanatä- (?) (LINCOLN 2012, pp. 250-251); *handarza- (NAVEH, SHAKED 2012, p. 82); hasiya- (LINCOLN 2012, pp. 30 ff., 225 fn. l); *ha-yät/0u­ (WERBA 2006, pp. 289-290); *haxay- (ARFA'EE 2008b, p. 264); *hmara-(NAVEH, SHAKED 2012, p. 74); (NAVEH, SHAKED 2012, pp. 70-71); *(a)xsaina- (NAVEH, SHAKED 2012, pp. 155-156); *(h)ulä- (TAVERNIER 2007±); *kam-, *kam-bujiya- (REMMER 2012, pp. 321-323); kamna- (LINCOLN 2012, pp. 425-428); * kan- "to long for" (TAVERNIER 2005b); kapautaka- (FRANCFOR1� TREMBLAY 2010, pp. 161-162); kära- (TUPLIN 2005a, pp. 230-231); *kära-paOt­ (MAYRHOFER 2005, p. 122 fn. 91; STEER 2011, pp. 246-247); *kiira-tanu-ka­ (NAVEH, SHAKED 2012, pp. 29, 84); *kas-ciy (LIPP 2009/I, p. 118 fn. 62); *kiisakaina-(NAVEH, SHAKED 2012, pp. 56,182); kayiida (WERBA 2006, pp. 289296); magu- (PANAINO 2016a); *maiOmäniya-(NAVEH, SHAKED 2012,pp. 55,184); *mari- (BRUST 2008, pp. 449-451); marika- (SIMS-WILLIAMS 2007,p. 231; BRUST 2008,pp. 445-447; HAJIANI,�MJOO,RAHIMIAN 2010); maOista- (LINCOLN 2012, pp. 429-432); mayüxa- (FRANCFORT, T REMBLAY 2010, pp. 162-163; BASELLO 2012a, pp. 36 ff.); miOa- (kär-) (SIMS-WILLIAMS 2007, p. 233; CI-IEUNG 2007, p. 258); mu(n)O- (CHEUNG 2007, p. 275); *nahvaka-/*nävaka- (HENKELMAN,

367

Adriano V Rossi

Ten Years ofAchaemenid Philology

STOLPER 2009, p. 301 fn. 88; ROLLINGER, HENKELMAN 2009, p. 333; TAVERNIER 2013a); *nava-taca-(NAVEH, SHAKED 2012,p. 72); näviya-(SIMS-WILLIAMS 2007, p. 240; WIDMER 2007, pp. 223 ff.; SKJ/ER V0 2005b, p. 315; SKJ/ER V0 201 la, p. 326b, fn. 3-4; FILIPPONE 2016, pp. 30-34); *ni-mita- (NAVEH, SHAKED 2012, p. 20?); [ni]pista-marya- (?) (SMITH, MARTIN 2009, pp. 57-58 and fn. 94); *nistävan(a)- (NAVEH, SHAKED 2012, p. 75); niyäka- (ANDRES-TOLEDO 2009, p. 398);paisyä (CIPRIANO 2007,p. 31 no. 16); *panfra-lpanfvara- (HENKELMAN 2010, p. 735); paratara- (TUPLIN 2011, pp. 174-175 n. 52); *pari-bar-1 (NAVEH, SHAKED 2012,p. 218);paridaida-+(WERBA2006,pp. 279-283; HENKELMAN 2008a, pp. 427-452; BRUST 2008,pp. 509-517; MAYRHOFER 2010, p. 5; PANAINO 2012a· LINCOLN 2012,pp. 59-85; ALVAREZ-PEDROSA2015 ; PANAINO 2016b);pärsa-(Sou� DAVAR 2012, pp. 64, 70-71; WIDMER 2012b); *pasca-,pascäbara-,pasyäbara­ (NAVEH,SHAKED 2012, pp. 55,197); *-pati-s(=AE 0battis) (HENKELMAN,STOLPER 2009,p. 303);patiyäva(n)h-(CHEUNG 2007,pp. 405-406); *patiyävahyai-(Gorn­ MAN 20!5, p. 177, fn. 905); *patikaravant- (NAVEI-1, SHAKED 2012, p. 56, 223); *patizbäna-(NAVEH, SHAKED 2012, p. 73); paBi- (STEER 2011); 1ttff(l"{U(; (Ross1 2013,p. 367-368); *piBra-käna-(NAVEH,SHAKED 2012,pp. 28,55,209); *piBvä­ (NAVEH, SHAKED 2012,pp. 55, 143, 209); rädi (PATRI 2005); rästa- (LIPP 2009/I, pp. 197-198, LIPP 2009/II, pp. 171-172); ,:täcä brazmaniya (SK.TJER V0 2011a, pp. 327b-328b; LIPP 2009/I, p. 193 fn. 102); *ratu- 1 (NAVEH, SHAKED 2012, p. 125); Saka- (HERRENSCHMIDT 2915, p. 129 fn. 2); *säma-gauna- (NAVEH, SHAKED 2012, pp. 56, 219); *sarakära-(NAVEH,SHAKED 2012,pp. 29, 56,195); *särtapawan-(SCHWARTZ 2009,p. 179); *sita-(NAVEH, SHAKED 2012,pp. 56,181); *si�a- (FRANCFORT,,TREMBLAY 2010, p. 160); skal;!Bi- (MAYRHOFER 2006, pp. 2223 and fn. 24; ITTZES 2014,p. 367); stünä- (SCHMITT 2016a, p. 46 fn. 3); *suxta­ (ka)- (NAVEH,SHAKED 2012,p. 56,223); s(väta-(SCHMITT 2006a,p. 187 fn. 164); siya1i- (LINCOLN 2012, pp. 258-268); tauma- (TUPLIN 2005a, pp. 229-230); *tauxma-kani-(NAVEH,SHAKED 2012,p. 115); taxs-(LIPP 2009/II,pp. 230-231); TVKra (BRUST 2008,pp. 658-660); Bä- 'to announce'(SKJ/ER V02005a,pp. 56-57; LINCOLN 2012, pp. 34-37); *Baigra-ka- (WERBA 2005/2006, p. 416; TR EMBLAY 2009�,p. 331 fn. 20); *Bigra-(MAYRHOFER 2006,p. 22; LUBOTSKY 2012,p. 103; ITTZES 2014, pp. 364-365); *()uc- (WERBA 2006, pp. 265-286); *upa-bariya­ (NAVEH,SHAKED 2012,pp. 55, 184); *upa-daya-(NAVEH,SHAKED 2012,p. 207); *upadita- (NAVEI -1, SHAKED 2012, pp. 26, 28, 56, 260); *upa-hmata- (NAVEI-I, SHAKED 2012, p. 98); *upa-ita/i- (NAVEH, SHAKED 2012, pp. 55, 82); upastä­ (BRUST 2008,p. 489); *upayäta-(NAVEH,SHAKED 2012,p. 262); *usp,:na-(NAVEH, SHAKED 2012,pp. 55,116); *?us()ra-(FRANCFORT,TREMBLAY 2010,pp. 169-170); *u,ftra-päna-(NAVEH, SHAKED 2012,pp. 52,56,70); uvamarsyu-(STOLPER2015); Uvza- (ÜETTINGER 2015, p. 57); *uzbara- (BIELMEIER 2008); *uz-gäma- (SIMS­ WILLIAMS 2007, p. 188; NAVEH, SHAKED 2012, pp. 55,206,210); uzmayä pati (PüMPEO 2012c, pp. 195-197); *väcaka- (NAVEH, SHAKED 2012, p. 135); va�a­ bara- (PIRART 2010, pp. 110-111); van- (CHEUNG 2007, pp. 205-206); *varda­ (HENKELMAN, STOLPER 2009, pp. 286-287); vasaj (LINCOLN 2012, pp. 214-216; LINCOLN 2015,p. 137); vasnä auramazdäha (SKJJERV02011a,p. 326b; SCHMITT

2016a, p. 41); *vätahya- (NAVEH, SHAKED 2012, pp. 36, 56, 184-185); vasna­ (SCHWARTZ 2013,p. 70); vaz,:ka- (SCHWARTZ 2013,p. 70; PIRART 2007,p. 98; PI­ RART 2010,pp. 112-113; LINCOLN 2012,pp. 232-236); *vi-dayld'i- (WERBA 2006, pp. 292-296); vi-mrad- (CHEUNG 2007, pp. 179-180); vi-nasta- (LIPP 2009/I, p. 198-199); viBa-(TUPLIN 2005a, PP· 229-230); viyax(a)na- (LUBOTSKY 2012, pp. 97-99); *v,:kfizana-(LUBOTSKY 2012,p. 104); xsa�a-(SKJ/ERV02005a, pp. 58-59; ÜNOLI 2005; ÜNOLI 2007); *xsä()ra-käna-, *xsä()rf-käni- (NAVEH,SHAKED 2012, pp. 56,209); *xsaOrapati- (SCHWARTZ 2005 [2009]); xsaya-(LIPP 2009/II,pp. 306309); xsäya()ya- (WERBA 2010; PIRART 2014, p. 241 fn. 395); *xsvip- (NAVEH, SHAKED 2012, pp. 57-165); yaci (BENE VENUTO 2012); *yäda- (BU YANER 2014, p. 16); yakä- (MORANO 2012);*yäma-(SHAKED 2013,p. 251); *yäsista- (NAVEH, SHAKED 2012, p. 105);yätumä(WERBA2006,pp. 289-296);yäumaini-(TREMBLAY 2005b,pp. 178-179; MAYRHOFER 2010, p. 6; SKJlERV0 2011a,p. 327a-b; LINCOLN 2012,pp. 263-268);yauna-(ROLLINGER,HENKELMAN 2009); *yava-däh- ? (NAVEH, SHAKED 2012,pp. 55,209); *rJJänali- (NAVEH, SI-IAKED 2012,pp. 55-159).

366

4.2 Onomastics(for space reasons only a general overview is offered in what fol­ lows). Most Opera minora related to the incessant onomastic activity by R. SCHMITT are gathered and indexed in SCHMITT 2000. Afterwards,updatings may be obtained through the bibliographies in SCHMITT 2006a,SCHMITT 2007,SCHMITT 2009e, SCHMITT 2011b,SCHMITT,V ITTMANN 2013, SCHMITT 2016b; ZADOK's re­ searches in the field are mostly retrievable in ZADOK 2009. An updated index of the OP documentation attested in non-Iranian sources ('Nebenüberlieferung,' cf. SCHMITT 2009c,p. 147 fn. 2) is now available in TAVERNIER 2007a, improving in many details HINZ' books, but accepting much material iranologically unscruti­ nized (cf. ScHMITT's works quoted above for detailed criticism of many TAV­ ERNIER's proposals,and TREMBLAY 2009b,p. 6). On GERSHE VITCH's activity in the field(in close collaboration with R.T. HALLOCK) cf. RossI 2006b, pp. 76 ff. Olr. (and sometimes OP) names and epitheta in an Indo-Iranian perspective are studied in SADO VSKI 2007 and SADOVSKI 2013(where further bibliography by the author on the subject is mentioned). On Greek reinterpretations of OP nouns cf. SCHMITT 2007b. OP names in the Biblical tradition are studied by HuTTER(2006,2009,2015). One should also refer to the extant articles in Encyclopaedia iranica and Reallexikon der Assyriologie, and the indexes of the published IPNB volumes. 5 Editions, Re-interpretations/New Translations, Comments. CMa(SouDAVAR 2012,pp. 62-64); DB/OP(VALLAT 2010,pp. 492-500,based on the Elamite text); DB/OP(HENKELMAN 2008apassim); DB/OP §§ 1, 8-14,32, 54-55, 58, 60-64 (SIU/ER V02011b,pp. 228-231); DB/OP I, 9-11 (AHMADI 2014); DB/OP §14(RossI 2016); DB/OP I,31-32,50-51(SKJJERV02009b,pp. 280-281); DB/OPI,41(SCHMITT 2016a,p. 60); DB/OP §18 (SKJIERV02011a,pp. 326b-327a); DB/OP I,48-50 (FILIPPONE 2015, pp. 30-34); DB/OP I, 85-87 (FILIPPONE 2016);

368

Adriano V. Rossi

Ten Years ofAchaemenid Philology

DB/OP IV, 68f. (SCHMITT 2016a, p. 86); DB/OP §70 (BAE 2008; BAE 2012; Wu 2014, p. 237 fn. 55; RossI i.p.1); DB/OP §§ 72, 75 (JACOBS 2014); DBa/OP § 4 (BA­ VANT 2014a, p. 189); DE/OP (SKJJERV0 2011b, p. 44); DNa/OP 48-51 (FILIPPONE 2015, pp. 34-37); DNa/OP 58-60 (LIPP 2009/II, pp. 171-172); DNa/OP §6 (Filip­ pone 2012b, p. 137 fn. 38); DNa/OP §§ 4-6 (Skjrerv0 2011b, p. 231); DNb/OP §§ 1, 3, 8 (SKJJERV0 2011b, p. 176), §§ 2, 4-11 (SKJJERV0 2011b, pp. 191-192); DNb 13f. (SCHWEIGER 2005b, p. 135); DNb/OP 13-15 (BENVENUTO 2012); DNb 26 (SCHWEIGER 2005b, p. 136); DPd (SKJJERV0 2011b, p. 232); DPd 13-24 (PIRART 2015, pp. 54-55); DPe 23 (GARNIER 2013, p. 61); DPe § 3 (LINCOLN 2012, p. 407 and fn. 6); DSe/OP (VALLAT 2010, p. 303); DSe/OP (1.9) (FILIPPONE 2012b,pp. 145146); DSe/OP § 5 (LINCOLN 2012, p. 251 fn. 41); DSf/OP (VALLAT 2010, pp. 304305; SKJJERV0 201 lb, p. 232 [11. 55-58]); DSf/OP 12-15 (AHMADI 2014, p. 50); DSf /OP 23-25 (FILIPPONE 2015, pp. 39-42); DSf/OP 50 (SCHMITT 2016a, p. 75 fn. l); DSk/OP (VALLAT 2010, p. 305); DSk/OP §§1-2 (SKJJERV0 2011b, p. 232); DSab/OP (VALLAT 2010, p. 312-313); DSad/OP (SCHMITT 2009b, p. 421); XPf/OP 30-37 (PIRAS 2012b); XPh/OP (HENKELMAN 2008a passim); XPh/OP §4 (FILIPPONE 2010); XPh/OP §§4-6, 8 (SKJJERV0 2011b, p. 236), § 7 (SKTJERV0 2011b, p. 181); XPh/OP 35-41 (PIRART 2015, p. 54); XPh/OP 43-46 (FILIPPONE 2015,pp. 35-37); XPl 5-9 (SCHWEIGER 2005b, pp. 134-135); XPl 14-17 (BENVENUTO 2012); XPl 26f. (SCHWEIGER 2005b, pp. 135-136); XPl 52 (SCHWEIGER 2005b,p. 137); XPm/OP (SCHWEIGER 2005b, p. 136); XSa/OP (VALLAT 2010, p. 314); 3 XSd/OP (VALLAT 2010, p. 315); XV/OP 20-25 (SCHMITT 2014b, p. 199 and fn. 11; FI LIPPONE 2015, pp. 37-39); A2Hb (FILIPPONE 2012b, pp. 132-133 fn. 28), A2Sa/OP (WERBA 2006; VALLAT 2010, p. 316); A2Sa/OP §§1-3 (SKJJERV0 2011b, pp. 232-233); A2Sc (VALLAT 2010, p. 316); A2Sd/OP (VALLAT 2010, p. 317). Tablettes: Fort. 1208-101 (S TOL PER, TAVERNIER 2007); Susa emailed brick (Schweiger 2005a, Schmitt 2009b, pp. 425-427).

than 80 glosses on AE tablets (progressively increasing in number, cf. AzzoNI 2008, p. 253, ibid. also for remarks on the functions ofAA); interesting information on this matter comes from the AA corpus of late Achaemenid documents from Bactria published by NAVEH, S HAKED (2012). AA in the Achaemenid period (SCHNIEDEWIND 2007, pp. 143-147; BLACK 2008, PP· 58 ff.; PANAIN O 2012b, pp. 39-44); AA Oll the AE tablets (S TOL PER, TAVERNIER 2007, pp. 19-20); AE vs. AA relationships (RAZMJOU 2008, p. 55; HENKELMAN 2008a, pp. 89-93). Aramaic pattems for OP script (WERBA 2006, p. 284 fn. 113); AA Vorlage (WERBA 2006, p. 290 fn. 137); AA intermediator between the centre and Egypt (SCHMITT 2008a, pp. 501-502). See also A 2 above on the Achaemenid linguistic situation.

6 Relationships between AA, OP and AE. FoLMER (1995, pp. 9-15) wams against applying such denominations such as Reichsaramäisch to AA; the more fresh AA documents are published, the clearer appears that AA was used as a lingua franca (only) by the scribes at the orders ofthe trade communities and landlords, tlms ex­ plaining the reduced amount of AA documents found at the periphery of the Achaemenid state (cf. lastly NAVE H, S HAKED 2012; also JACOBS 2012, pp. 100102; a very well formulated statement is GzELLA 2012, p. 11: "under Achaemenid supremacy [ ... ] did one of the existing varieties ofAramaic (presumably a Baby­ lonian dialect) provide the common language of a highly centralized scribal cul­ ture"). The last essay on the interference between OP and AA is GELLER 2005, who leans toward a strong linguistic pressure OP � AA; cf. also CIANCA GLINI (2011, pp. 15-16). The Orientallnstitute of Chicago has launched a Project for the publi­ cation of ca. 600 AA Persepolis epigraphs on tablets, labels etc., as well as more 3 XSb, XSe (VALLAT 2010, p. 315) are only AB and in principle should be excluded from the present report.

369

1 Relationships between Language and Writing. Barring few cases (Rossr 2005, pp. 937-938, BASELL O 2012c, pp. 5-46), scholars interested to OP script do not have discussed in the last few years the intemal structure ofthe signary (some in­ sights in FEAR 2015, pp. 78-82), but rather the problem ofthe invention ofthe OP cuneiform (artificial invention politically motivated vs gradual adapting ofexistent pattems) and the connected interpretation ofDB/OP §70 - DB/AE L. These texts have been considered for long time as a kind of 'foundation document' ofthe OP script, while it is clear from the context that they refer - in a not yet understood way- to the whole monument (text + images + localization) (Rossr 2000; i.p. 1). Interest for the alloglottographic theory (writing in Elamite and reading in OP, cf. GERSHEVITCH 1979, never seriously discussed: cf. RtJBIO 2007, p. 67: "lt is some­ what puzzling that, despite GERS HEVI TCH' scholarly stature, this theory of his has been almost completely ignored by Achaemenid scholars and, for the most part, has received neither explicit criticisms nor endorsements ofany kind") has greatly decreased. Rossr (2006b, pp. 83-84 n. _86) outlines how this theory was received; since then few comments have been added, mostly sceptical (S TOL PER, TAVERNIER 2007, pp. 7-9; TAVERNIER 2008, pp. 75-76; YAKUBOVICH 2008, pp. 205-207; HENKELMAN 2011a, pp. 614-622; SCHMITT2011a, p. 333; VALLAT 2011b; SCHMITT 2015b, p. 256; RossI i.p.1), rarely in favour (Rumo 2007, pp. 39-40). Aramaic pattems for OP script (WERBA 2006, p. 284 fn. 113); 'confusion' of signs (WERBA 2006, pp. 278-279); crossing ofcunei (SCHMITT 2014b, p. 202-203); invention of OP script (BAE 2012, pp. 146-149; SCHMITT 2014b, pp. 212-213); "multiliteracy" (S TOL PER 2015, pp. 12-13); OP in Egyptian cartouches (Wu 2014, p. 229 fn. 42); < 0 v 0> for /öl (WERBA 2006, p. 285); writing Oll wood (BRIAN T 2014); word-divider (SCHMITT 2014b, p. 202; BASELL O 2012c, pp. 105-123; BASELL O i.p.); writing of diphtongs (WERBA 2006, pp. 278-279); for /e/ (WERBA 2006, pp. 284-285). The history ofthe decipherment is treated by HUYSE 2006a. 8 The 'Median 'Language. Acceptable evidence, dialectological collocation and sociolinguistic role of an alleged (NW) Iranian language conventionally referred

370

Adriano V. Rossi

Ten Years ofAchaemenid Philology

to as 'Median' raise a specific question, whose literature has beeri reviewed by Rossr (2009b, i.p.1, i.p.2). Strong upholders of a dialectal focalization which is highly subjectivist if referred to phases of scanty documentation such as Achaemenid Iran are scholars ( e.g. HoFFMAN, MAYRHOFER, to a lesser extent SCHMITT) who believe in the possibility to isolate punctual shibboleths valid both for Olr. and modern Ir.; the geographic collocation of these could allow the pro­ jection in antiquity ofdialectal features possibly found in anthroponyms and ethno­ toponyms of 'Median' entities defined as such in old sources (Assyrian, Babylonian, Greek etc.) According to the present writer (cf. already Rossr 1981 pa ssim), this ambiguous formulation originates from the classification of the 0Ir. languages such as it was conceived at the times of the Grundriss der ira nischen Philologie, when a collocation ofthe Av. dialect in West Iran was considered possible, very few Mid­ dle Iranian languages were really described, modern Iranian dialectology was still in its infancy, and most Neo-Assyrian and Neo-Babylonian documentation was still unexplored. The importance of the 'Median' polities in I millennium BCE Near Bast is strongly reduced in the modern approach to 'Median' archaeology, to minor arts, to the 'Median' digressions in Greek and Latin sources. Consequently, the interest for an alleged 'Median' writing, instrument ofa 'Median' bureaucracy in the serv­ ice of an alleged 'Median' state, decreased greatly after Padua 2001 Conference (LANFRANCHI et al. 2003), even if TAVERNIER 2007a shows the persistence of the usage (introduced by HINZ) to define most 11011-0P elements in AE as 'Median.' RossI (i.p.2) suggests to figure out in the 'Median' space a coexistence of a plurality of ethnic/linguistic entities perceived by foreign descriptors under the common label of 'Median.' This does not imply to negate the existence ofan Jran­ ian linguistic phase linked to the names attested in the Neo-Assyrian and Neo­ Babylonian collateral tradition pointing to the geographic areas in which Ma da local dynasts were active; the criticism only concerns the possibility to join in a coherent whole oflinguistic features all the divergencies from what appears as the linguistic base ofOP. On the diffusion ofthe (Irano-)'Median' varieties cf. ZADOK 2013, pp. 416-418. On PIRART's (2004, p. 36) proposal that an 'aspa-dialect' (i.e. 'Median,' as op­ posed to a southern ' a sa- dialect') may be pointed out as the dialectal base ofAv, cf. TREMBLAY's 2008a, pp. 11-12: "Especially Pirart's favorite isogloss asa :: a spa is no sensitive marker, since dialects where PIE *ky, PIIran. *cy become *sp (Northwestem Iranian, Scythian, all Eastem Iranian except Sakan) constitute no unity." Cf. also PIRART (2014, p. 240 fn. 394: "C'est par convention que je parle de 'mede' pour designer la langue iranienne ancienne qui est a la source de l'un des deux dialectes arbitrairement meles dans la versione dite 'vieux-perse' des in­ scriptions cuneiformes achemenides"), and T REMBLAY (2009b, p. 2). 'Median' features in (Proto-)Achaemenid Elamite onomastics (MANCINI 2008a, p. 34 fn. 41; TAVERNIER 2011a, pp. 236-238; TAVERNIER 2013a, p. 271); 'Median' features in Achaemenid titulary and formulae (SCHMITT 2014b, p. 204-

205; PIRART 2014); 'Med.' z - OP d (W ERBA 2006, pp. 280-281); 'Median' as a conventional term (PANAINO 2016b, p. 40 fn. 7)..

371

9 Falsifications. SCHMITT 2006b, 2007a; VAHDATI 2007; ZIEGLER 2008, pp. 88-93; SCHMITT 2010a, 2012b, p. 408, 2015a; SCHMITT, STOLPER 2016. SCHMITT discusses various 'suspect' inscriptions (some of which also treated in SCHMITT 2009) in his Pseudo- a ltpersische Inschriften (2007, pp. 25-34); the 'falsification' question is partly linked to the datation of the first documents car­ rying OP writing (cf. §7 above) and the concept of the linguistic evolution ofOP which one agrees to (cf. §2 above). B. Achaemenid Elamite 0 No recent Forschungsberichte for AE are available, barring Rossr 2008. The bibliography in HINZ, KoCH 1987, pp. 1317-1368 is complete till about 1986 as far as the AE lexicon is concerned, but not for other fields ofElamite linguistics. All periodical surveys quoted at A O above contain Elamitistic material. The digital series NABU andARTA - available at www.college-de-france.fr- have contributed a lot to the recent discussions ofAE themes. 1 Epigraphy, Corpora . In the last decade no major discoveries of AE epigraphic material were an­ nounced. Minor fragments and tablets emerged on the market or reappered from museal reordering have however greatly contributed to the enrichment of our knowledge ofAE. Etat des questions et perspectives de recherches on Persepolis Elamite docu­ ments kept at The Oriental Institute, Chicago have been discussed at College de France (3-4 Nov.ember 2006; proceedings in BRIANT, HENKELMAN, STOLPER 2008 (composition and dimensions of the archive are discussed by JONES 2008, its via­ bility by TUPLIN 2008, pp. 324-325, multilingualism by TAVERNIER 2008, pp. 62 ff). As far as the Persepolis Archive is concemed, the "Persepolis Fortification Archive Project" is a major research and publication project of the University of Chicago's Oriental Institute. This important ancient archive - found in 1933 by Oriental Institute archaeol­ ogists working at Persepolis - is currently being studied and published by experts in Chicago and elsewhere (Stolper, Henkelman and others). The "Persepolis For­ tification Archive Online" (PFAOnline) is a vehicle for documenting and dissem­ inating the contents of the archive in an electronic form suitable for advanced philological research. The only recent edition of the whole AE royal inscriptions corpus, with Spanish translation by ENRIQUE QUINTANA CIFUENTES is available at http ://www.um.es./ipoa/cuneiforme/elamita.

Adriano V. Rossi

Ten .rears ofAchaemenid Philology

HENKELMAN 2008a includes many Persepolis tablets interpreted and com­ mented with original remarks oflinguistic interest. Two PhD thesis on the AE texts ofDB have been defended recently by CLAU­ DIA ROMAGNUOLO (Naples L'Orientale University, 2014) and Salmän ALIYÄRI BABOLGHÄNI (Tehran University, 2015); the last has been subsequently published as ALIYÄRI 2015.

tions of OP /8av0/ (TAVERNIER 2006b, p. 191); AE < 0 s 0 > for "OP" 8 (SCHMITT 2006a, pp. 139-140); -s (WERBA 2006, p. 281 fn. 102); - alternance for OP /sa/ (WERBA 2005/2006, pp. 414-415); */ 0sa/ > */ 0ca/ written (SCHMITT 2010c, p. 291); with [tJl > ? (SCHMITT 2010c, pp. 290-291); (TAVERNIER 2007d, pp. 269-271); /u/ > /i/ (TAVERNIER 2007d, pp. 279-289); OP vs. non-OP ('Median') variants in AE, AB, AA (SCHMITT, VITTMANN 2013, p. 97); vowel syncopation (TAVERNIER 2006b, p. 191; HENKELMAN 2011b, p. 154 fn. 145); Late AE word-divider (SCHMITT 2010c, pp. 289-290; BASELLO 2012c, PP· 121-123; BASELLO i.p.); zhN for /cV/ (LIPP 2009/1, pp. 121-122).

372

2 Linguistic Description. 2.0 Some grammatical outlines for Elamite have been recently made ava1lable: STOLPER 2008 (20041); GRILLOT-SUSINI 2008 (second edition, partly revised, ofthe 1987 edition, with main focus on Middle Elamite). QurNTANA has published two introductions to the language (2010, 2013), with short anthologies and glossaries (also available at http://www.um.es./ipoa/cuneiforme/elamita). Very briefdescrip­ tions of Elamite are KREBERNIK (2005) and TAVERNIER (2011c). BAVANT (2014a, pp. 235-336; also 337 ff.) is an interesting general overview from a specific point of view. WINDFUHR 2006 has a main focus on the OP interrelation with AE (pp. 386-390). On AE as an administrative language cf. STOLPER, TAVERNIER 2007, pp. 16-18; BASELLO 2011; HENKELMAN 2013. 2.1 Writing and Phonology. AE adaptation of OP stems in -ä- and - 'f- (KORN 2009, p. 207 fn. 72); alloglottography (Rumo 2007, pp. 39-40, 67-69; DANESHMAND 2015, p. 325; KuDRINSKI, YAKUBO VICH 2016, pp. 55-58; see also A § 7 above); -as (SCHMITT 2009c, p. 148; SCHMITT 201 la, pp. 331-333; SCHMITT 2015b, pp. 256-257); (TAVERNIER 2007d, p. 265 fn. 3); broken writings (TAVERNIER 201 la, p. 229); tCu-aC (HENKELMAN 2008a, p. 560); determinatives in AE (V�LL�T 2005); (BASELLO 2012c, pp. 112-121; BASELLO 2013, p. 255); �emmat1?n (T�v­ ERNIER 2005a); graphic variation in AE (BASELLO 2006); graph1c semphficat10n ( instead of , instead of , SCHMITT 2010c, p. 290); h-signs (TAVERNIER 2011a, pp. 218-219); h - k (PIRART 2015, p. 54); (SCHMITT 2006a, p. 133 fn. 16; TAVERNIER 2007d, pp. 265-269); historical evolution ofAE (BLACK 2008, pp. 48 ff., 54 ff.); in transcriptions ofOP (KLOTZ 2008, p. 111); AE ideograms (STOLPER 2013, pp. 402-403); /inne/ (TAV­ ERNIER 2007d p. 273 fn. 50); interchanges b -p, d - t (TAVERNIER 2005a); /mat/ (TAVE�IER 2011a, p. 219); LIL, HE.GAL (TAVERNIER 2007e); lateral affricate /-phoneme (HENKELMAN 2008a, p. 407); (tSIGiSA7) (STOLPER 2013, p. 403); (SCHMITT 2010b, p. 399); for mi, OP -mai (SCHMITT MEs 0 2010c, p. 291); 0me/ 0ma for OP / vai/ (HENKELMAN 2008a, pp. 434-435); < > 0 0 (HENKELMAN 2008a, pp. 448 fn. 1040, 526); n#k [IJk] (SCHMITT 2010b, p. 392); (TAVERNIE-R 2007d, pp. 272-274); /ne/ (TAVERNIER 2007d, p. 278); AE for OP /f/ (FRANCFORT, TREMBLAY 2010, p. 75); phonetic com­ plements (TAVERNIER 2007c); phonological recons:ruction of (A)E (SMITH 2?07); � cune1fo�m Reiner-Test (TAVERNIER 2011a, p. 234); relationsh1ps between AE/0 0> m transcnp p. scripts (Rossr 2005); - bddt Herausgebern (S. 59) erwogene Alternativlesung und -de utung al edergegeben wi " s Oxu he t Vaxsu-abda-data" das mit Created miraculously by mit griech. >brdt von g n �ird. Was immerhin überra;�ht, ist die Übereinstimmu o_h l durch w �.) s u 2 1, 5, eia Aßpa86:cäi;, dem Namen, den Xenophon (Kyroupai� siche r get ch m r abe er d t, g _ Rückübertragung einem fiktiven König von Susa be1le . er t deutet ist. 84 Doch diese Beziehung hilft auch nicht wei ießl 7. Namen mit dem Fl uß- und Gottesnamen *Vaxsu- finden sich sc�l ich nft en der nsch I auf n e h c s auch am Unterlauf des Flusses im C h w a r e s m i s g nicht in d r �e al r : ish b � ag sl g . älteren Zeit (1.-3. Jh. n.Chr.). Hier ist die Forschun . es­ r �war e c ub ck bh e Ub � e d � enfassen sehr günstig gewesen, denn einen zusamm d ich en e t konn s mau h r e Darub mische Personennamen bot erstmals L1vs1c 2011. so r e ­ P hen Iranisc „ es d ikel z s Fa en ersten Entwurf des Te xtes für den einschlägig ­ rt e r w e nkens da burg) er kt-Pet n (Sa nennamenbuches" benutzen den PAVEL LuRJE ' Nachdruck w eise vorbereitet und aur den schon hier ausdrücklich und mit . . hingewiesen sei. . für den t e t e bi 2) 1, . r (N hr. C . v s Jh. 1. es Ein Ostrakon aus Kalaly-gyr 2 wohl d Ox os von ,, kau-däta*Vaxs < / -8atak Namen eines Steuerbeamten wxswdtk- /Waxs Träg en t n :nn ge eile Z en d 1:_ e g ie r� e �,r gegeben",ss ganz parallel zu dem in der darüb . en � e geg u sn Ra von ,, / u-8atak sn /Ra kdes ebenfalls theophoren Namens rsnwdt . ech gn ch au en d r, o v_ en N�m m e d zu dung Hier liegt also eine triviale ka-Weiterbil . n gel ie �p er w1d 6.) en b o . s ( dt 'O�u86:c11� (s. oben 2.) und aram. wl;Jw und So wie der in diesem Namen vorl iegende Anlaut wxsw- aram. whswd ryk 'xws-m � i ( ,xwsgd. so zu g o al n � a . sogd. wxwsw- ents pricht, so findet sich e m di rm, o F e es di /; -8at-en /Waxs -yn >xwsdt ,xws-prn) auch chwaresm. )xws- in hr.) zuta­ dem Palast von Toprak-kala auf einem Inschriftenfragment (2.13. Jh .. n .C �at V. A. b, e bh en t l a rh e on � auf s l ina g i -Or er d e getrat, das nur als Abdruck des L hnung eic z ch �a r e d o g n du il 86 Abb hne o d n u n gabe LIVs1c ohne irgendwelche An a n u Erklar e agen schl ge vor . v . s !1 LURJE, �,' c� der beka nntgemacht. Die von PAVEL 1st, u?e�­ die Form im Prinzi p eine V ariante von wx�wdt�- ,;vo� Oxos geg:ben okonstl ­ zeugt un eingeschränkt; genau genommen, hegt m xwsdt-yn nur em Hyp anischen Sprachen �ut bekannt kon auf -yn /-en / vor, wie sie a us mehr eren mittelir her au�ge­ sind. Die schon bei L1vs1c, a.a. O. angedeutete und von LURJE, s . v. n� er gleich­ führte Alternativlesung und -deutung 'xwsrt-yn, die auf mehr oder wemg * -dät ­ bedeutendes *-räta- als Zweitglied zielt, darf wohl auf sich beruhen, da _ � d dies für als Kompositionszweitglied durch Parallelen gut gestützt wird, währen *-räta- nicht der Fall ist. ent, 2. /3. Aus ebensolcher Quelle (Palast von To prak-kala, Inschriftenfragm 87 unter Jh. n.Chr.) stammt auch der Name wxswbrk-y /Waxs-ßarak/, den LIVSIC Vgl. hierzu SCHMITT 2011, S. 63 Nr. 2. S. 155 Anm. 28; LIVSIC Vgl. LIVSIC 2004, s. 189-193, v.a. s. 190ab; ferner LNSHITS 2003, 2011, s. 163. 86 LIVSIC 1984, S. 264. 87 LIVSIC 1984, S. 265; vgl. LIVSIC 2004, S. 190b; LIVSIC 2011, S. 163. 84 8 5

423

Vergleichtmg von spätchwaresm. ßrk, sogd. ßr 1, ßryy „Fruc ht" als „Frucht des Oxos" versteht. Andere Namen, die auf *-baraka- zurückgehen, scheint es nicht zu geben; auf mo.rp�ologisch mehrdeutiges mittelpers. brk( )n) /Barag/, das wohl eher als Hypokon stlkon zu interpretieren ist, 88 läßt sich nicht bauen. Ratsam ist es auch, von einer Vergleichtmg des Namens *MiBra-bara- (belegt in demot. Mjtrbr) �bzusehen , zu de1:1 es offenbar keine Parallelen gibt und dessen genaue (onomas­ tlsche) Analyse mcht ohne Schwierigkeiten ist.89 Unsic her �leiben zwei weitere Namen, die ebenfalls im Palast von Toprak ­ kala bezeugt smd: Auf dem zuerst von S. P. TOLSTOV edierten Dokument 1 9° mit ?1'eis�altige� Namenliste �wohl wiederum 2. /3. Jh. n.Chr.) erkannte L1vs1c später m Zeile 18 1m Rahmen em er wesentlich verbesserten Lesung des Gesamttextes wxwsmry, d. h. ,,Wax(u)smarI" 91 - wxws- zeigt eine Schreibvariante des Namens im Sogdischen belegt i t -, von dem aus er eine Brücke schlug zu sogd'. d�e auch � v ()wxsm1yk usw., das er damals J edoch noch nicht zu deuten wußte. Der dadurch implizierte Ansa tz v on * Vaxsu-marya- ,,Diener des Oxos" unterscheid et sich aber von allen anderen vergleichbaren Bildtmgen dadurch, daß er nicht auf erweitertem *-mwyaka- b�siet!, wie es durch baktr. oax}Jo µap11yo und sogd. ()wxsmryk vo­ :.ausg��et�t w1rd. Uber di�se � chwierigkeit hilft auch nicht die (in meinen Augen uberflussige) Annahme emer mner-ostiranischen Entlehnung hinweg. In dem Register über Getreide lieferungen des Jahres 252 der chwaresmischen Ara K-1 steht in Zeile 14 wxwsps(.)[]: Während L1vs1c diesen Namensrest ohne weiteren Kommentar als „Vax vaspas []" las,92 enthält er nach LURJE, s. v. im Anlaut den Namen des Oxos, ist also mit einem unklar blei benden Rest etwa als * Vaxs­ pas0 zu verstehen. 93 8. Zusammenfassend lassen sich also folgende Namenkomposita mit dem Erstglied *Vaxsu- als sicher anerkennen : * Vax�u-bandaka- in sogd. wx(w)s(w)ßntk (s. 4.) und aram. wb§wbndk (s. 6.); * Vaxsu-brta- in baktr. oax}Joßop8o (s. 5.); *Vaxsu-däta- in griech. 'O�u8frr11� (s. 2.), aram. wl;,swdt (s. 6.) und chwaresm. wxswdtk- usw. (s. 7.); *Vaxsu-farnah- in sogd. wxwswprn usw. (s. 4.); *Vax_su-�iaryaka- �n sogd. {'.)wxsmryk usw. (s. 4.), baktr. oax}Joµap11yo usw. (s. 5.), vielleicht a uch m chwaresm. wxwsmry (vgl. aber o ben 7.); *Vaxsu-vahista- in aram. wbswhst (s. 6.); *Vaxsu-vanant- in baktr. oax}Jooavtvöo (s. 5.) und * Vaxsu-vazdah- in griech. 'O�ußasrn; (s . 2.). „

88

Hierzu vgl. GIGNOUX 1986, S. 55 Nr. 186. Vgl. SCHMITT, VITTMANN 2013, S. 66 Nr. 34. :� ToLSTOV 1962, S. 218 (Abb. a): 219 (Nr. 8) las „WffWSMN7K7". Vgl. L1vs1c 1984, S. 268 f sowie S. 283 Anm. 82, wo als alternative Deutung "gut rechnend/denkend" erwogen ist. · 9 2 LIVSIC 1984, S. 256. 93 „ Die ursprünglichen Lesungen von TOLSTOV 1958: 210 bzw. TOLSTOV 1962: 220 sind langst obsolet. 89

425

Rüdiger Schmitt

Der Flußgott Oxos in der iranischen Anthroponymie

. Vaxsuößär (s. 4.) u nd baktr. Ei nzelsprac hlic he Sonderfälle liege n bei sogd e kom�t man auc h kaum.daran�. oa?(Pmaµpo (s. 5.) vor; und nach Lage der .Di�g ionstyp1:1s wie diese ?aktnsche Form auc h die z�e1 herum, dem gleic hen Komposit oa�ou (s. 2.) und aram. mtnvbsw Reflexe von * Mißra-vax§u-, nämlich gnec h. Mt8p . (s. 6.), zuzurec hnen . bdruck (auf To�) unbe�annter 9. Anhangsweise sei hingewiesen auf den Siegela is an) ?1it zwe1sprac h1��r InHerkunft (wohl aus Gandhära, etwa Nordwest-Pak � ,,[sri }vakhadavasa „des · l .,, OXON (?]" und alti n94doar. (i n Brähm1) schn'ft gnec1 „the name seems to daß , nd ung si Vakhadeva" da dessen Herausgeber der Mei n both „ in �a ng�ages" �S. 13 a\ refer to the �od of the river Oxus" (S. 80a), und zwar m1 kh zu iran. * xs brah h oc n h. X Diese Ansic ht ist jedoch verfehlt, da weder griec von * Vaxsu- stimmt.

G.D. DAVARY, H. HUMBACH: Eine weitere aramäoiranische Inschrift der Periode des Asoka aus Afghanistan. Akademie der Wissenschaften und der Literatur. Abhandlungen der Geistes- und Sozialwissenschaftlichen Klasse, 1974/1. Mainz/Wiesbaden 1974. G. FussMAN, D. KöNJG: Die Felsbildstation Shatial. Mainz 1997 (Materialien zur Archäo­ logie der Nordgebiete Pakistans 2). P. GIGNOUX: Noms propres sassanides en moyen-perse epigraphique. Wien 1986 (Iranisches Personennamenbuch II/2). G. GNoLI: Zoroaster in Hist01y. New York 2000 (Biennial Yarshater Lecture Sedes 2). F. GRENET: ,,L'onomastique iranienne a Ai Khanoum". In: BCH 107 (1983), S. 373-381 (,,Appendice" zu RAPIN 1983). H. HUMBACH: ,,Die sogdischen Inschriftenfunde vom oberen Indus (Pakistan)". In: AVA-B 2 (1980), S. 201-228. -,,The Sogdian Inscriptions of Thor-Shatial". In: JCA 8 (1985), S. 51-57. -Review of SIMS-WILLIAMS 2000. In: BSOAS 65 (2002), S. 415-418. F. Jusn: Iranisches Namenbuch. Marburg 1895 (Nachdruck Hildesheim 1963). V.A. LIVSHITS: ,,Three Silver Bowls from the Isakovka Burial-Ground No. 1 with Khwa­ rezmian and Parthian Inscriptions". In: ACSS 9/1-2 (2003), S. 147-172. V.A. L1vs1c: Juridiceskie dokumenty i pis 'ma. Ctenie, perevod i kommentarii. Moskva 1962 (Sogdijskie dokumenty s gory Mug 2). - ,,Dokumenty". In: Toprak-kala. Dvorec, red. J.A. RAPOPORT, E.E. NERAZIK. Moskva 1984, S. 251-286. - ,,Nadpisi i dokumenty". In: Kalaly-gyr 2: Kul'tovyj centr v Drevnem Horezme IV-II vv. do n.e., red. B.I. VAJNBERG. Moskva 2004, S. 188-212. - Sogdijskaja epigrafika Srednej Azii i Semire6:;a. Sankt-Peterburg 2008. - ,,Licnye imena v horezmijskom jazyke". In: VDI 4 (2011), S. 156-166. P.B. LUR'E: ,,Sogdijskaja nadpis'". InMaterialy Pendzikentskoj arheologiceskoj ekspedicii. XII: Otcet o polevyh rabotah ekspedicii v 2008--2009 gg., red. P.B. LuR'E, I.K. MAL­ KIEL', A.V. OMEL'CENKO. Sankt-Peterburg 2010, S. 40-42. P. LURTE: Personal Names in Sogdian Texts. Wien 2010 (Iranisches Personennamenbuch II/8 = Iranische Onomastik 8). J. MARKWART: Wehrot und Arang. Untersuchungen zur mythischen und geschichtlichen Landeskunde von Ostiran, hg. H.H. SCHAEDER. Leiden 1938. J. MARQUART: Untersuchungen zur Geschichte von Eran. II. Leipzig 1905 (Philologus. Supplementband X/1). J. NAVEH, S. SHAKED: Aramaic Documents from Ancient Bactria (Fourth Century BCE.) From The Khalili Collections. London 2012. C. RAPIN: ,,Les inscriptions economiques de la tresorerie hellenistique d' Ai Khanoum (Af­ ghanistan)". In: BCH 107 (1983), S. 315-372. W. RAu: Staat und Gesellschaft im alten Indien, nach den Brähma�i a-Texten dargestellt. Wiesbaden 1957. G. ROUGEMONT: Inscriptions grecques d'Iran et d'Asie centrale. London 2012 (Corpus Ins­ criptionum Iranicarum II/1). C.E. SACHAU: The Chronology of Ancient Nations. An English Version of the Arabic Text of the Atluir-ul-Baldya of Albfrun� or „ Vestiges of the Past". London 1879. R. SCHMITT: ,,Die theophoren Eigennamen mit altiranisch *Mi()ra-". In: Etudes Mithriaques. Actes du 2e Congres International Teheran, du J er au 8 septembre 1975. Leiden usw 1978, S. 395-455 (Acta Iranica 17). - Iranische Personennamen in der griechischen Literatur vor Alexander d. G1: Wien 2011

424

ABKÜRZUNGEN ACSS AVA-B BCH BSOAS CRAI JCA NC VDI

ria Ancient Civilizations from Scythia to Sibe - Beiträge ie äolog Arch ende leich Verg und ne emei Allg ue eniq Hell Bulletin de Conespondance an Studies Bulletin of the School of Odental and Aföc ptions et Belles-Lettres Inscd des emie Acad ces. sean des us Comptes rend Journal of Central Asia The Numismatic Chronicle Vestnik Drevnej Istorii

BIBLIOGRAPHIE

Griechischen bis zur Kaiserzeit. Halle F. BECHTEL: Die historischen Personennamen des . . . (Saale) 1917 (Nachdruck Hildesheim 1964). . z am Rhem 2005 (Matenahen zur ArM. BEMMANN: Die Felsbildstation Dadam Das. Mam . chäologie der Nordgebiete Pakistans 5). Ai K.hanoum (Afghamstan)" · In: CRAI P. BERNARD: ,,Campagne de fouilles 1976-1977 a . (1978), S. 421-463. A . , e dokumenty. Ueme,perevod i· kommenM.N. BoGOLJUBOV, O.I. SMIRNOVA: Hozjajstvenny Mug 3). . . tarii. Moskva 1963 (Sogdijskie dokumenty s gory me III: Zoroastnamsm under Ma­ Volu . nism astria M. BoYCE p GRENET· A History ofZoro (Handbuch der Orientalistik I/8/1/2/2). cedonia,; and Ro�i an Rufe. Leiden usw 1991 t Yue-ti. Part III". In: NC III 12 (1892), A. CuNNINGHAM: ,,Coins of the Kushans, or Grea S. 40-82 und 98-159. 94 Vgl. UR RAHMAN, FALK 2011, s. 13a und 80a Nr.

06.05.04.

426

Rüdiger Schmitt

(Iranisches P ersonenna�1enb ch V/5A = Iranische Onomastik 9). � l1e N,ame. 11 i n ägyptischer Neben überlieferu ng. Wien ScHlvIITT' G. VITTMANN. Iramsc . 1 10) · nbuch VIII = Iramsche Onoma st'k ame n 2013 (Iranische s P ersonen . erneur. Documen ts arameen s du S. SHA KED: (2004) Le satrape de Bactrzane et son gouv . . · . ne p ans 2004 (P ersika 4) IV s. avan t n otre e, re p10 . venan. t de. Bactrza . • . n 7 n ian 1n scri'J:' tion s of the Upper In dus. 1. London N. SIMS-WILLIAMS: Sogdian an d oth�r 1ra 1989 (C orpus lnscriptio_num lra�1c�rnm ��:, ll)U 1n - Sogdian an d other Iraman b_1scrzpt10 ns o; ie pr er In dus. II. London 1992 (C or pus Inscriptionum Ira nicarnm 11/3/2). - Bactrian Docum�nts_.from Nort�e.m A.ß. h . t n 1. Legal an d Economic Documents. Lon­ don 2000 (�tud1es �n the �ahh �ollg�;:� ; � C orpus Inscriptionum Iranicarum, 11/6). - ,,The So gdian Anc1ent Le��er II, . In. Philolo ica et Linguistica. Historia, Pluralit�s, _ Hum. b ac�' hg. M.G. SCHMIDT, W. BlSANG. Tner R·

U niversitas. Festschrift fur Helmut

2001, S. 267-280.

II· Letters an d Buddhist Texts. London st - Bactrian Docun:en tsfrom Jl!?rthern,Afj_ : ghm�. ;�� pu� Inscriptionum Iranicarum II/3). 2007 (Studies m t he Khahh C ollectrnn 3 . . 1\r es. Wien 2010 (Iramsches P ersonennamenb ue h 11/7 = Iranische -Bactnan Personaz lVam Onoma stik 7). - Bactrian Documen ts from Northern Avg. h.an .� ta�I· Legal an d Econ omic Documents � ;lili Collection 3 = Corpus Inscrip : (revised edition). London 2012 (Stud ies m e ­ tionum Iranicarum II/6). . J TAVERNIER: Iran ica in the Achaememd . Perzod (,ca . 'i50_330 B.C.). Lexicon of Old Iranian . _ �I Proper Names and Loamvords, Attested m N,on ranian Texts: Leuven usw. 2007 (Orientali a L o vaniensia Analect a 1 58). . S.P. ToLSTOV: ,,Raboty Horezmsk�J �?eologo-etno rafice skoj ekspedicii A� SSSR v 1949, rte raboty Horezmskoj Ekspedicii 19491953 gg.". In: Arheologicesk1e z etnograifizcesl 1953, red. S.P. ToLSTOV, T.A. ZDANKO. Moskva 1958, S. 7-258 -Po drevnim del'tam Oksa i Jaksart . os a � M k; �i:�� from Gan dhära. Wiesb ad en 2011 A. UR R AHMAN, H. FALK: Seals, Sea 1:n gs �i r � M (Studies in t he Ama� ur R ahman C l el c wn 1 = onographien zur indischen Archäo­ _ lo gi e, Ku nst und Phtlolo gi e 21). C · WERBA: Die arischen Person en namen u n d z'h�e Träger bei den Alexan derhistorikern z Dissertati. on . w·1en 1982 · (Studien zur iran ischen An th roponomast'k)

°

MARTIN SCHWARTZ Berkeley

An Achaemenid Position, and Gathic Composition: OPers. *grasta-(pati-), OAv. grähma-, and PIE -Vghres A title fPALTAIIATIB (grastapatis) occurs in t1i a 4 century B.C.E. Greek inscrip­ tion from Mylasa in southwest Anatolia, most recently treated in DESCAT 2011, 1 who, restoring the text, translated it as follows: "Aux divin ites bienfaisants d'Hekatomnos et d' Aba, en etant grastapatis de Mausole, a cons acre (les brasiers?) Artimes fils de Targelios." Nothing more is known about Artim es, son of Targelios. lt is clear, T as DESCA notes, that Artimes' title is a Greek rendition of an Old Persian noun *grastapati-, nominative *grastapatis, in which *-pati-, as in many other compounds, means "chief, master, someone in charge of something (in this instance, *grasta-)."2 DESCAr took *grasta- as referring to provision of food, seeing in *grasta- the equivalent ofVedic grasta- "something devo ured" (root gras), which has a variety ofGreek cognates which refer to devouring or eating (by animals or people). Postu­ lating an equivalent Iranian root *grah, from which HUMBACH and MAYRHOFER had taken the obscure Old Avestan noungrahmaas "voraciousness," DESCATtentatively proposed for *grastapati- "chef des choses devorees, chef de ce qui est mange." Thereby DESCAT envisaged two possibilities for the function of the grastapatis. Firstly, DESCAT called attention to the interest ofthe Hekatomnids in the raising ofherds and the profits therefrom, which is well attested for the instance ofKon­ dalos (Pseudo-Aristotle, Econ. II 2, 14a), who was responsible for nourishing the animals belonging to the satrap. This activity is related to the central role ofanimal husbandry in the Achaemenid system; cf. the tax mentioned in Pseudo-Aristotle (Econ. II 1, 4, with BRIANT 1996, pp. 453456). In this environment grastapati­ would refer to being in charge of the anim als' eating. A second possibility men­ tioned by DESCAT is that the grastapatis' pres idence over eating would consist of the provision of meals for the satrap. Tue fact however is that there is no evidence whatever for an Iranian root grah "to eat, to devour;" words for eating and feeding are always :from the root hwar in Iranian. Before further considering the prepotmd *gra sta-, one must note the survival of *grastapati- as West Middle Iranian gras t(e)bed, attested as Pahlavi and Parthian , both in two passages of Shapur I's trilingual inscription at

1 I thank Professor PIERRE BRIANT for putting this article at my disposal, and Professor RAYMOND DESCAT for a lively online discu ssion ment as well as Doctor HRAcH Mt\RTIROSYAN, of the data, and both scholars for the encourage­ who provided invaluable remarks on Annenolo and Iranistic matters. gical 2 For another example of an Old Iranian compound in *-pat i- attested in late Achaemenid southwest Anatolia, see SCHWARTZ 2005 .

428

Martin Schwartz

lleled by Greek TOY EIII ANNQNHl, Ka 'be - ye Zardosht, in each instance para ofprovisions ofsupplies (from the prothe latter amounting to "the one in charge vinces to the capital)." 6, pp. 99-100), who compared the These words were discussed by BAILEY (195 al, such as horse, mule, and espe­ im an prepound with the Iranian loanword grast " contexts of Shapur 's trilingual it r e rg ." From the la cially ass, for carrying loads with the title grast(e)bed bore a unique is clear that the two successive persons antecedent t hat Arm. grast has as its imperial economic position . Given now nally have e l atter term should o rigi *grasta-, the prepou nd of *grastap ati-, th e A chaemenid ch came to designate a l at meant "chief of caravan (animal)s," whi imperial office.3 natolia) and later a Sasanian A st e hw t sou t a d ste tte a ( t s po l satrapa rivation of *grasta- from an alleged O ne must re j ect BAILEY's (loc. cit.) de whence supposedly OAv. garazdi- (corIndo-Iranian base *ghar-dh "to tak e up," . gradior). rectly "pathway, approach, access," to Lat *grastando- lra nian roots , the form ofl y rtor Given the presently known repe h-ta - > gra B-ta-) a * gr]Jt a past passive participle (vi could be , in the ory, ta k en as er, this root is manifest v e How " . t kno a e i t "to from the Indo-Iranian root granth , . g. Sogdian yrans, Persian girih "a knot " in Iranian only as the noun *granßya- > e les. sadd k ac p of o notable role in the constmction and knotting generally has n ided by the Greek word for pack Etymological guidance fo r grasta- is prov "press, constrict." A par­ 4 t rJaK, pres . rJa.rrw (*rJaKiw) saddle, rJa.yµa, from the roo ack has been *grasta- "anima l upon which a saddle p allel etymology would have ion" from a ct ri t s n o ected to pressure or c imposed" = past passive participle "subj k"). as o t " s a ras ta- from the rootfr root grah "to press, constrict" ( lik e Av.f gr5hm a-, an t es Av is confirmed by Old Such a root grah, I shall try to show, s Indo­ riou a v of rds with cognates in wo which I tak e as meaning "oppression," ndoo-I t Pro a from e 5 . grahm a-, I deriv European languages, for which, with OAv ict." European root *g hre s "to press, to constr es the three t h t he pas sage whic h preced wi s ord The meaning "oppression" acc .12-14: attestations of g rahm a - at Y32 ugh the robbery of inherited Y 32.1l. ..the wrongsome ones, who thro milords and miladies with their property, show themselves off as 6 from Best Mind. grandeurs, as they tear the righteous away charge ofprovision 3 lt is notable that on the eastem trade roads caravan chiefs, who were infor this function un­ titles the and ions, posit s for the animals, occupied 1ucrative and prest; igiou 2009 [2013]. ARTZ SCHW and , -WILLIAMS 1996 derwent semantic expansion; cf. DIEN 1962 SIMS . 2004 DIEN cf. yra; Palm The situation was similar for the res foyyartä at syrn f. Muqad4 I have derived from this word the Khwarezmian word for "pack saddelle," saddle"), a pack 's "cam sutur lan-i ä p . Pers rabl, . pälän dimat al-Adab 47 .5-6, 1242, 316.5 ( = Arab. Pers for ng girdi al later ofbi basic sense point ofcultural-historical interest. Note the also " nd. arou (and Sogd. pyrön) < 0Ir *pari-däna- "placing4, with dismissal ofconnection with Vedic grasate 5 Cf. already SCHWARTZ 2006, p. 461, fn. Middle Iranian forms for "wealth." "devours" and of [HENNING's] comparison with grähma- as Pahlavi parag, Parsi Sanskrit 6 This context would explain the glossation of laik ä, and Persian risvat "bribe."

An Achaemenid Position, and Gathic Composition

429

These evildoers are the kauui-s the local warlord rulers, 7 charact�nzed as dom' . s ineering tyrants (Y32.12 below) h

!!�if;��:i::!:�:��:i:� ti!

;,�:: ;�;�:e!,es:::���S:!:�i:�::Ifif u

Y32.12a yäis grähmä a.sät karapä varatä xsa9ramcaä fsanqm dmjam Through such things, via *�ppress10n, t11e karapan-priest chooses . · . domm (at)'10n and the tyrants wrongness instead of Rightness.

Y32.l3� yä x_Sa?rä grähmö hfsasat acistahiiä damäne-mana11hö b m1haus maraxtärö ahiiä h ough such d��in(at)�on, *oppression tiess the destroyers of this existence � ; he H ouse of vv orst Mmd. ..

a-· n z-k-auuaiiascft xratus nf. dada t Y32 .14a ahiiä *grähn1a-9 a.- 1z-· oiaoz hrough the *oppression of this [existence], the kauui�s indeed put � t eir mtelhgence mto ensnarement. .. . ,, lt shall now be argued that gräl a * "op re,ss1o · n �tams an association with h f ; :: th e idea ofconstriction posited abo or 1 t�e"��!:tricr: �s!! �s sp;ifically associated e m�y proceed from t.he fact t hat at Y32.13 a'and Y32 14 a ' hzsasat andgh-�9a_-: oz oz respect1ve1y nn · med�· ate1y follow derivatives ofthe root häy "to . . tie (up)" ( h' h I h a1;;e sema�tlc �eld as "constrict, constrain") and occur , : :ft � �ä�1;� immediat e a - responsible for th� ty�ng of the oppres:Ors the�selv e;. ��!;:;;�atJt:s ;;�c;ay are among a senes m the . second half ofY32, wi' th cu1mm · at10n · at Y32 ·16c aenav, h-e draguu - ... avha iiä ato . "may I tie up (ayhaiiä < /ä häyaY_w' ;;/'\ the wrongsome ones for their harmfulness ..." . In Y32 the 10CUS c Oll tymg10 proceeds from Y . 46 re�alled back wards; specifically . Y46.19-Y46.6 coµtn'butes a lex1cal and themat1c bas1s for Y32 · 7- Y32 ·16, 11 as sha11 ·. be charted below. In th'i s compos1t10nal recasting, y46 • 7 ha e-l7na hn..a- "of ensnare-

:1ra:

!:�

�f =��::::

-v

) )) • 7 SCHWARTZ 2013. To the Sogdian evide shou be d��d_S. kwy, kw y, kw .v discussed rn SIMS-WILLIAMS 1992, pp. 54 and 75 Not;��e spe!fmg :w y m the personal name wnkw, 'y . and Man. S. sm 'wtry kw' J' "sea giant ,: which sho�s th�� the_Sogd. word is /kawäy/, from the Olr. strong-stem acc. *kaw äyam· Tl1e.'oAv fiorms kauuauascit and k äuuaiiascä must have defectively placed ä. s For the meaning and etymology of hfsasat see ScHWARTZ 2006, p. 461, fn. 5 and the cross-textual composition evidence in ID · 2010'p.- '219 and Chart 1 on 221, and most recently ID. 2013, p. 72. 9 Mss. grähmö by influence ofY32.13a' grähmö 10 For the reI �v�nt pumtlve · · and other ties, see ScHWARTZ .2006. PP ·. 460-466 . · u For the prmc1ple, see In. 2002 ' d1scuss10n 3-8 and the charts, w1th thereto, in ID. 2003, 2009, and 2010. The relationshi eenl Y46. 19 remembered backwards and Y32.7 seq. was first treated, in elaborate detail, i� ID · I99 , PP· 143 and 161.

'Jf� ,

430

An Achaemenid Position, and Gathic Composition

Martin Schwartz

fr aryh aiiä "may I tie u p" �m the as positional correspondent Y32.16 plmes of tra the r ente to d e 'h said to be doom same ·"\J a-y. At Y46 · 6, the wrongdoer is .. - - T he nex t stanza, Y46. 7 , asks . o- da-mq· n liae9ahzz a g a-t · - · - hztu Wrongness: d ru10 hush:�� to hann ho would lay hold ofZarat Mazdä's protection from the evildoer w .�. May t�e Y46 ws, o foll or ated in the stanza wh 1ch him. T he idea is then e lab y m t bu ·, · · me � lt ,�ome with against me .. . not rea.ch threat of him who plots harm i fe .. · J:Iere �he d oo g m o fr � . is p erson, and �eep h nn matchin g enmi ty against h rec1 proc1ty. 1ve umt p f o s ent strainü:g go_ wi tl� state� themes of tying down and re 14, �h� re d an .13 Y32 md hes beh T his ideology o f punitive rec1proci ty 1:�tnct10n 0 c nt at l the t th � to indicate � g rahma- + '1häy may consequently be said oppresa m grah f o on not1 asta- as root ('1grah), which we have noted fro� *g r . . sion " ties up the agents of oppress1on. etween ed fonnal relat10nsh1p� b o i t n e m re fo a e h t ws � ho s t The following char of the g rtm ha he T . ve abo d ': se ment10ne Y46.19-16 and Y32.7-16, including tho me fr ve at o ll o c a s e :"'ork vid � � � related words also pro correspondences offonnally tems h i 1c h w hm t wi lly, a i ant d purely sem � for recognizing words which are relate a- and st ra g een w t be t10n ec fo r the ety mol ogical conn there is fur ther evidence g rahma-. d ä, mos� kno,;'ing" tu(uam) ( ... ) mazdä vaedistö "Tho�, Maz .procat1ve Y46.19 & Y32.7 ly rec1 fy sat1s "to vfci9a- "decision;" --Jx§n(a)u Y46.18 & Y32.8 " hear "to a)u ,/sr( " . ,, sa]Jgh- "to declare; Y46.l7 & Y32.9 mJhö "the power ofGood Mmd man aus vw1h + ifa 9 Y32. & 16 Y46. dä9ä,:zg "lawful" + .Jdä "put, establish" Y46.15 & Y32.10 maz- "great" Y46.14 & Y32.ll '\.1/sr(a)u "to hear" + marata- "morta1s" 12 Y46.13 & Y32. . . aeibiiö ... mazda "to them Mazda speaks" Y46.12 & Y32.12 ;" ahu- + --Jmarklc .''destroy ex1sions inat dom "vi·a s� , · z , (; n ra " xsa17 Y46.11 & Y32.13 est (peJ. )" tence;", damäna- "house;" karapan- "pri ]" e tenc exis [ ahiiä "ofthis ,, " y46.1O & Y32.l3 I gau- '"cow;" mrau-1mrau- say Y46.9 & Y32.14 jiiätaus/jiiätoöis "oflife" Y46.8 & Y32.15 ä9(r)is "threat" y 46.8 & Y32.16 aenayhe "for hanning" . ,, .16 Y32 & Y46.7 vahista- best ,/häy "to tie;" draguua,:zt- ''wrongsome; Y46.6 & Y32.16 ns, Y46 .11 th the most lexical �orr elati� Now in both corresponding stanzas wi vi a thev dom­ karapan-s, "the de�troy ers of_ex 1stence, and Y32.'13, it is said t hat the l ofx�a9ra­ ell; m both the mstmmenta · nation(s) " are doomed to the "House" ofH y oke/s�bJug�te at Y 46.11 by yüjan "they �'domination" is immediately follow ed u bJugat�on oe oppres�ion ." Not only d !.! [mankind]" and at Y32.13 by grahmö*" yup n to y okmg ppression, but the hteral r�ference.of here semantically p arallel * o ab� ve, tow ard e, or constrict10n, posit �d would parallel the imposition of p ressur g a pack saddle" - and 1� g rah7:ia-. the meaning of '1g rah in *g rasta- "havin ference to encumben ng ammals, h_as Furthermore, '1yu j "y oke," in its basic re f the three preimmediately continues the thought o a parallel at Y32.15, which

ment" has

.v

"

,,

431

ceding stanzas, wh ich a ttes tgrahma-: "Through these things the karapan -dom and kauui-dom are lost, together with those whom they put into harness (aib'i ... dab:ztz); those latter, with the fonner two groups, will not be brought into the House o fBest Mind by Those W ho rule over li fe as They wish." Here, again collocated with the theme ofretributive impedance seen alongside with grahma-, aib'i ... dail:zt'i "they hamess" parallels yüjan "they y oke," both expressing the imposition of constric­ tion, which, again, with animals as object, characterizes grasta-, the meaning of whose root, grah, as "to put under pressure, constrict, constrain," is corroborated by the con textual and compositional details attending g rahma- < Oir. *grah-ma-. The final support for the meaning o fthe latter root comes from broader Indo­ European considerations, for which we may compare the Iranian exclusively with a series of Baltic forms, which, with the Iranian, point to a PIE verb root *g 11res. OAv. grähma-, with -m V- formant, comp ares wi th Lith uanian g resme, g rasme "threat, compulsion."12 Allied Baltic fo1ms are Lith. gresiit, gresti "to threaten (Lat ­ vian grasät, gras'it 'id. '), to make loathsome;" g rasits "disgusting;" grasa "threat, restriction;" gristü, g r isti "be bored." S haring meanings found for Lith . gresiit and grasits, Armenian g arsim "I loathe, am disgusted" (whence Arm. noun g ar.§ "loathing") is easily seen as a 1 cognate, S ogd. /üzba/, /wizbä/ "fright, wor ry,· anxiety," Khwar. )wzß )hyd "bored, was tedious to" (SCHWARTZ 1970, p. 299). In conclusion, a late Achaemenid Old Persian title *grastapati-, reflected in a Greek inscription o f Mylasa, has as a later correspondent, Middle Persian and Parthian grast(e)bed, whose Greek equivalent indicates the transpo rtation of provincial supplies. From all these data one may trace the development o f *grastapati- as an important economic office originating from "chief o f caravan animals," whose prepound *grasta- (> Armeniangrast), re ferring to the imposition ofpack saddles, is cognate wi th OAv. g rähma-, which, on the basis ofGathic con­ textual and compositional criteria, must mean "oppression, subjugation, constric12 PoKORNY 1959, p. 445, under *ghers-, *ghres- "Widerwille, Abscheu, Ekel" gives Av. _ grahma- as "Sünder, Frevler" but, although mentioning the other relevant Baltic forms omits mention ofLithuanian gresme, grasme. PIE *g hres- is not represented by Rlx ET AL. 1998.

432

An Achaemenid Position, and Gathic Composition

Martin Schwartz

tion," from a Proto-Iranian root grah- < Proto-Indo-European *ghres (also attested in Baltic, A rmenian, and Tocharian) "to press down, constrict, oppress."

ADDENDUM

A novel explanation of OAv. grähma- (or, theoretically, *grähman-) has re­ cently been offered by HUMBACH (2010, pp . 50-53), who, comparing the Vedic root gras "to devour, to graze," Greek ypacrTv; "fodder, forage" and Latin gräm en "leaf of grass" < *grasm en, takes the OAv. word as parallel to English slang grass, Arabic !Jasf§ ("grass") for "cannabis," and suggests that the OAv. refers to some unspecified drug plant "and the gang of those who are addicted to it." Given that 110 such word for "grass" is attested in (Indo-)Iranian, and the close analysis of Y32.12-14 does not support such an interpretation, but points to a very different meaning and ( an inner-Iranian) etymology. 13

433

.REFERENCES D.Q. ADAMS: A Dictionary of Tocharian B. R evised a nd Greatly Enlarged. Amsterdam 2013. H.W. BAILEY: "Armeno-Indoiranica." In: TPS (1956), pp. 88-126. C. BARTHOLOMAE: Altiranisches Wörterbuch. Strassburg 1904. P. BRIANT: De Cyrus a Alexandre . Pans 1966. R. DESCAT: "Autour de la tombe d'Hekatomnos. Nouvelle lecture d'une inscription de Mylasa." In: Z eitschrift der Papyrologie und Epigraphik 178 (2011), pp. 195-246. A.E. DIEN: "The Sapao Problem Revisited." In: JAOS, 82.3 (1962), pp. 335-346. - "Palmyra as a Caravan City." 2004. H. HUMBACH, K.. FAISS: Zarathushtra and His Antagonists. Wiesbaden 2010. J. POKORNY: Ind ogermanisches Etymologisches Wörterbuch. Bern and München 1959. H. RIX, M. KÜMMEL, T. ZEHNDER, R. LIPP, B. SCHIRMER: Le xikon der Ind ogermanischen Verben . Wiesbaden 1998. RoNG XINJ1ANG: "New Light on the Sogdian Colonies along the Silk Road." 2009. M. SCJ-IWARTZ: "On the Khwarezmian Version ofthe Muqaddimat al-Adab as edited by Johannes Benzing." In: ZDMG 120 (1970) [1971], pp. 288-304. - "Sound, Sense, and 'Seeing' in Zoroaster: the Outer Limits ofOrality." In: K.R. Cama Orien tal Institute Proce edings (51h-81h January 1989). Bombay 1991, pp. 127-163. - "How Zarathushtra Generated the Gathic Corpus: lnner-textual and Intertextual Composition." In: BAI 16 (2002) [2006], pp. 1-8. - "Women in the Old Avesta: Social Position and Texiual Composition." In: BAI 17 (2003) [2007], pp. 53-64. - "Apollo and Khshathrapati, the Median Nergal, at Xanthos." In: BAI, 19 (2005) [2009], pp. 45-50. - "The Gathas and Other Old Avestan Poetry." In: La langue poetique ind o-europeenne.

Actes du colloque de travail de la Societe des Etudes ind o- europeenne, Paris 22-24 oc­ tobre 2003, ed. G.-J. PINAULT, D. PETIT, Leuven-Paris 2006, pp. 459-498.

- "Pouruchista's Gathic Wedding and the Teleological Composition of the Gathas." In: Exegisti Monumenta: Festschrift in hon our ofNicholas Sims-Williams, ed. W. SUNDER­ MANN, A. HINTZE, F. DE BLOIS. Wiesbaden 2010, pp. 429-448. - "Sartha- and Other Caravan Words." In: BA/23 (2009) [2013], pp. 177-178. - "Lexical Cruces ofYasna 29 and the Cross-textual Composition ofthe Gathas." In: An-

cient and Middle Iranian Studies: Proceedings ofthe 611i European Conference ofIran­ ian Studies, ed. M. MACUCH, D. WEBER, D. DURKIN-MEISTERERNST. Wiesbaden 2010. - "Avestan kauui-s and the Vocabulary of Indo-Iranian Institutions." In: Gifts to a Magus, Ind o-Iranian Studies Hon oring Firoze Kotwal, ed. J. CHOKSY, J. DUBEANSKY.

New York 2013, pp. 63-76. - F orthcoming, BAI 23. N. SIMS-WILLIAMS: Sogdian and Other In scription s of the Upper Indus. London 1992. - "The Sogdian Merchants in India and China." In: Cina e Iran da Alessandro Magn o alla dinastia Tang, ed. A. CADOJ\i"'NA, L. LANCIOTTL Firenze 1996, pp. 45-47. 13 A variant version of this article has appeared in the online journal ARTA, 2015. http: //wvvw.achemenet.com/document/ARTA_2015.005-Schwartz.pdf

Shau1Shaked

448

. .

NICHOLAS SIMS-WILLIAMS London

n 1988-1991.

vie z�-a�es!·iques. Wiesbade versiy Press. London 1971. J· KELLENS' E. P1RAR.T: .c: rd Uni · Ji za vi. Dzctwnwy. 0 xio - Boulder, - e-tan. . pät-z Em D N MAcKENZIE: A d VI) , by Ätur a r, k e n (D es a g i an S san a e S ofth W dom s.· S�AKED: The is . Les textes Conczse pa

and Iranian Mytholog1cal Coloradol979. Sem1"t'1.c-I anian Syncretism volution and Permanence on Notes King. t Firs Man, ; a sformation Re - "First ssa y on Transformations." In: Gilgul. � e � to R2� ;wi Werblo�vsky, ed. S. SHAKED, D. SHUL­ edzcat in the His tory ofReligions. D . n Relz-. n 1 87, P· 3 8-256 . 1de �e A. MS ROU G ST a tte rns in Ira nza P e nt rr MAN, G. . cu R e In: _P �ure" ? Ltter Islamtc (30th - "Same Iranian Thernes m m. Procee ; s .ofthe Round Table Held in Bamberg s ent des Suji cem to van A l' pour n gions from Mazdaism ciatio 1), ed. PH. ��N OUX. Asso September-4th Getobe,: 199 . a, · 143-158· a nd Ch.ristos . Studzes h ia ess M Etudes Iraniennes. Pans 1992 PPand · "In ogy. 1 . Eschato · ogony HAKED, - "The Myth of Zurvan: Cosm Presented D ·d Flusse r ed. I. GRUENWALD, S. s · nity a i t _ s in the Origins ofCh�·i o. �19-a;l . Sasa , nz·an Ira n. School of Onental . .on in G .G . STROUMSA. Tubmgen 1992. b,. PP·01J Relzgz es 1 e z rz V a n. a tio form n s Tra . - Dual.zsn'• i·n . '1_' s. London 1994 . tion ' Interpretation, Dis-� nsla Tr· a andAfrican Stud1e · )· d an (Z sta Ave ary on the · ademta - "The. Traditional Comment l 'As ia Centra1e da Ales ,;;a ndro al X se colo. Acc e s i a P e r a · L . In t o rt10n?" 6: aedia Ira nica , nazionale dei Lincei. Rome 199 d Practices." In: Encyclop an deas I man Ira by ced --- "Had1"th. v· As Influ en 453-457. : Ti.ex·t on Religious Customs. Hamburg 1930 . ed. E. YARSH ATER. New York 2002hl, aPP vz A Pa ra hams. J.C. TAVADIA: Säyast-ne-säyast. Concepts a nd ße z·ze.1· 1s . Tlra ns la ted by Israel Ab . s : Tli e zr ge a S e Th H: BAC E.E. UR Jernsalem 1979. . . utt art 1965. G. WIDENGREN: Die Religz. one n Irt;r:� �ae!ma . Oxford 1955. as z R.C. ZAEHNER: Zurvan. A oro

,

The Name ofthe Kushan Goddess Oµµa A goddess Oµµo has long been known from a rare coin of Huvishka which shows a god and goddess labelled as 0111:>o and Oµµo respectively. The former is represented in one ofthe usual ways, as a four-armed Siva, while Oµµo is a "robed female holding flower." 1 GöBL's early reading2 011µ0 rather than Oµµo and his identification of the figure as the deified king Vima Kadphises has been generally rejected, and was given up by GöBL himselfin his later work. 3 Consequently, most cormnentators both earlier and later have identified Oµµo as the Indian Umä , the consort of Siva. 4 In recent years, the discovery in the Rabatak inscription of the older form of the name Oµµa, with clear feminine ending -a, has made it clear that Oµµa/Oµµo is indeed a goddess and has led to renewed speculation concerning her name and identity. In my first edition of the inscription I wrote: 5 "Oµµa ... is evidently the goddess named as oµµo on a coin ofHuvishka, where she is depicted as a consort of011po = Siva. A parallel issue depicts 011po with Nana, thus indirectly associating Nana and Umma as in lines 9-10 ofthe present text. Oµµo is generally taken as a transcription of Skt. Umä, a name of the wife of Siva. This fits the context but does not account for the double-µµ-, which is most easily explained as due to as­ similation from *-ß-µ- (cf. aµo < aßo µo 'to the' in SK4, M, line 11). In that case the divine name oµµa/oµµo might be compared with Av. upama- 'highest' and ex­ plained as the name or epithet ofa female deity, who was perhaps conflated with Umä on account of the phonetic similarity of the two names." This analysis re­ ceived valuable support both from GERARD FussMAN and from GHERARDO GNOLI. 6 A partially simi:lar proposal was put forward by MUKHERJEE, who suggested that Oµµo derives from Indian Ambä, another divinity belonging to the circle of Siva; 7 but, as FussMAN correctly pointed out, all the other divinities listed in this part of the Rabatak inscription have Iranian rather than Indian names. 8 Moreover: "Ambä, ROSENFIELD 1967, p. 94. First apud HUMBACH 1960, pp. 57-59. 3 e.g. GöBL 1984: 44: "Ommo (oder Nono)-Oeso." 4 e.g. BANERJEA 1956, pp. 126-127; ROSENFIELD 1967, pp. 94-95; ÜNOLI 2009, pp. 143-145. 5 SIMS-WILLIAMS, CRIBB 1996, p. 84. 6 Fussrv1At� 1998, p. 587; GNOLI 2009, p. 143. 7 MVKHERJEE 1998, p. 41 and elsewhere. 8 FussMAN, loc. cit. The "exception which proves the rule" is the interlinear text between lines 9 and 10 which specifies the name "in Indian language" of one of the Iranian deities, prob­ ably Sroshard (see SIMS-WILLIAMS 2008, p. 64b). 1

2

450

The Name ofthe Kushan Goddess Opµa

Nicholas Sims-Williams

ndue serait nite mineure et la f�rme atte quoique fort populaire plus tard, est une divi HE�JEE MUK by ted development posi *amma." One may add that the phonological 1ves surv -µßce n e , where the sequ seems not to be attested elsewhere in Bactrian 9 the language. unchanged throughout the known history of O GNOLI �i� me the_ g�eat honour of RARD GHE ague Our deeply lamented colle on the Rehg10us S1gmficance of the contributing to my Festschrift "Some Notes 10 d, am?ng�t o�her matters: the name Rabatak Inscription ," in which h e discusse contnbution m Ghe:ardo s memory, and identity of the goddess Oµµa. In this short is question and presentmg some further I would like to reciprocate by returning to th e goddess. . .. thoughts on the etymology of the name of th he wntmg with double -µ� t that fact e h t is ed h blis esta T he first point to be m are ts n a n so n co ble Dou t. significan cannot be ignored but must be regarded as le conso�an�s common_ly attested doub ly n o e h T g. n i writ n general rare in Bactria and -88-, wh�c� 1s systematically used are -yy-, which represents [tJg] as in Greek, ar pronunciatlon of 8 after a palatal in later Bactrian texts to indicate a particul mi -fourth ce�tury onwards, t�e word vowel (11 or 1) or before p.11 In texts from th� � ß), 1s often wntten aßßo. In th1s case, for "water," [äß] (Manich aean Bactrian ythe word for '\:ate( fro� the even the reason is almost certainly to distinguish 1s demon ­ nichaean Ba�tnan ß), as commoner word aßo [aß] "to, for, etc." (Ma bother to wnte th� double -ßß- ,;"hen strated by the fact that the scribes often do not ede� ?Y the �1de t- : iaßo (the) the noun forms part of a compound or is prec e preposit1011 aßo 1s never preceded by water" is of course unambiguous, since th ts a double -c;;c;;- as the outcome o_f a 13 the article.12 One document of 610 C.E. attes tu-c'i- or �p�rtz,_-ciya -. The Indian complex duster in the word noc;;c;;o "debt" < *par o, no doubt 1mitatmg the double con­ loanword novo "merit" is once written novv 14 :i a-, Pali pufifia-. In tl�e Kush�n sonant of a Middle Indian form such as Pkt. pwp po "glory, majesty" (pa�s1m) an� m period a double -pp- < *-rn- is attested in cpap etc., a�d a_ double -cm- ;n combma­ 1 c;;opptyt "time" (Rabatak), later cpapo, .soptyo, d endmg m [s] or [1s]. tions of the enditic partide -cn/-cro with a wor attests the ar?it�ary doubling �f lt will easily be seen that none of these cases a mere transcnptlon of Skt. Uma. a consonant as would be the case if Oµµa was r differentiation, as in �e case of It is unlikel; th at Oµµa/Oµµo is so writtefffo or *oµo is not attested. lt 1s th erefore aßßo � aßo, since a contrasting word *oµa

almost inevitable that the double -µµ- results from assimilation of an earlier con ­ sonant duster, as in the cases of -pp-, -ss-, and perhaps -öö-. 16 T he assimilated con­ sonant could be a labial stop or fricative, as previously proposed, but it seems worthwhile considering whether there is any alternative. In a recent book on Bactrian personal names., I discussed the Bactrian name I>oµoyoßav8ayo "Slave of Shumug," identifying the divine name "S humug" with a word .for "moon" wh ich is widespread in th e modern languages of north em Afgh�n�stan : Munji yum�ga, Yidgha imoyo, Sanglechi wulmjk, Ish kashmi lömik, Wakh1 ;amak, all of wluch may perhaps be derived from a feminine adjective *uxsma-kä- "growing, crescent." 17 The same divinity, I suggested, is also referred to in the Sogdian personal name 'x,fwmßntk "Slave of Akhshum" and the month ­ name 'xswmyc/xswmyc, h ere in a shorter form derived from *uxsmä-, without the *-k-suffix. 18 The que�tion arises whether the name of the goddess Oµµa/Oµµo may represent exa�tly th1s fo�m and thus be identical with that of the Sogdian deity Akhshum. T h1s hypothes1s has the advantage that Akhshum is actually attested as the name of a divinity, one of sufficient importance to be the dedicatee of a month whereas the alternative *upamä- "Supreme (deity)" is a mere paper construction: At first sight it may appear that the different ph onological development of *uxsma -kä- to I>oµoyo- is enough to rule out the new hypothesis. However, it is clear that many Old Iranian sounds develop in Bactrian in differen t ways in dif­ ferent words, presumably as a result of dialect mixture. 19 In the case of Old Iranian *s,�, we have ,�airly clear examples of (i) the preservation of the original duster, as 111 vtpµapo cmnplaint," v1pµap-, v1pµ11P- "to complain" < *ni-smära (-ya)-; (ii) a development to [hm], as in oauµap-/oauµop8o "to muster" < *wi-smära -/*wi­ smrta -; and (iii) the loss of *§ without trace, as in the second person plural pronoun 1:roµaxo, i:oµaxo, mµaxo < *usmäxam (with added prefix for differentiation from the first person plural pronoun aµaxo, iaµaxo < *ahmäxam). 20 Tlms the assimilation of *sm, perhaps via *hm, to *m(m), is well within the bounds of possibility. Whether the same result may be expected in the case of the heavier duster *x§m is difficult to say, as there seems to be no other likely example apart from the very words under consideration . In Yidgha-Munji, according to MORGENSTIERNE, "in the heavy groups 16 _In_Io .. 1999, p. 197, it is argued that the double -88-, at least originally, represented [dd] by ass1m1lat10n from [dy], this being the immediate outcome of a palatalized *d. � ID. 2�10, pp. 157-158 (no. 558). Various Pashto dialect words for "moon, moonlight" _ (spogmay, ogmay, wagmay etc., also tarogmay "moonless night" - all ofthem feminine nouns), cf. MoRGENSTIERNE 1927, p. 69; 2003, p. 75, also seem likely to belong here though the origin of the initial sp- of spogmay is unclear. 18 This suggestion is approved by LURJE 2010, p. 121 (no. 212). 19 A study on this subject is expected from Frarn;;ois DE BLOIS. 20 For all these forms and their etymologies, see the glossary entries in SIMS-WILLIAMS 2007. For less ,ertain,�xamples see also t�e entries under auµapo "account," 1pouµivo "enemy," rcaµavo[ 7 wool. The so far unexplamed second element of the personal name paoavo-miµo (B�RATTE et al. 201_3, p. 23b) is possibly the Bactrian outcome of *casman- "eye." cf. the vo­ callsm of late Sogdtan cym- (beside csm- and cm-). 7

crmµßo, etc. 9 Cf. SIMS-WILLIAMS 2007, glossary, s.vv. aµßayo, aµßapo, ya�Lßo, Kaµßampo, 10 GNOLI 2009. . rarely occurs m any other context. 11 See SIMS-WILLIAMS 2007, pp. 38-39. Double -88. 12 See ibid., p. 182a, s.vv. aßßo, aßßo8axµo, aßßoA.avo . 13 See ibid., p. 257b. . s been assumed that a d�uble -vvtnne 14 See LEE, SIMS-WILLIAMS 2003, p. 171. lt has some r ethnic name Alkhan. In my v1ew, how­ is also attested in certain spellings of the personalntand/o single v, see SIMS-WILI�LA.MS 2010, P· 33. ever the "w-shaped" character is merely a varia of1: ap11µcrcro "also the third," aßtcrcrt "also 20-2 lines 'i 5 See SIMS-WILLIAMS 2008, p. 57, many."

451

452

Nich o las Sims-Williams

, and the resulting groups Sm, sn were XSm, xsn the X was dropped at an early date "moon," he 21 case o f Sanglechi wulmfk . 22 further reduced to m, n. " Similarly, in the The voic­ " was reduced to sm at an early date assumes that "the group xsm e" < *cas­ "ey ply early loss ofx, cf. cAim ing of *§ in Wakhi �mak also seems to im of *xsn, at th ears not to run parallel to man-.23 In Wakhi the treatment of *xsm app xsna-, *rau < " ch i s l ost, c f. ruxn "white where it is the *§ rather than the *x wh i t of men p elo v in B actrian , where the de raxnfg"fire ."24 This may also be the case de­ s it and an, gratitude" < *xsnaußr25 *xsn to xn is attested by axvcopo"satisfactio , y case an n I axvcopo"recompense"). rivatives (axvoptyo"satisfied , grateful", na8 l ity ibi oss p the ufficient reason to reject the contrasting treatment of *xsn is not a s me of the goddess Oµµa/Oµµo . na e -µµ- in th o f a develop ment o f *xsm to sim pler to ew, it would of course be From a purely phonological point of vi N either kä-. maa-kä- rather than *uxsmä- and *uxs start fr om *usmä- and *usm ogdian S om fr rt a s discu s s ed ab ove , ap Oµµa/Oµµo nor any of the potential cognate refore the ht mig truction with *x. One 26 or 'xswm-, actually contain x or require a recons ra*ust < " camel as second ary, as in xwstr-" x n ia d g o S e th g n i d r a reg k f n o thi 27 and reconstructing the root as *(H)us in Old Avestan xsma- "y ou" < *usma-, rease," Indo­ r, the root * (H)waxsl(H)uxs"to grow, inc 28 rather than *(H)uxf Howeve whereas the , n "waxing" of the moo European * h i!!,eks, is typically used of the ongly asso­ r st s i , dawn ," Indo-European *h2l}es, root (H)1vvahl(H)us "to brighten rm for the te a ed 29 and is hardly likely to have provid ciated wi th the moming sun here from ssed cu s i d words oach may be to derive the moon. A more promising appr f Pahlavi o s i as b e by BENVENISTE on th the root *sam"to shine," which was posited "west ," y nsm n a i d i- "sunset ," Sog fräsm "daybreak," Avestan hü fräsmö. däit ki äsma a z a Z and r 'dysmy "moon ," Parthian nys 'm "darkness ," Pahlavi Psalte accept ­ y l ri ssa e nec t u ve added -witho 31 "id ."30 To this collection of words others ha d.," "i ä üsm i an g n ri e -Harzani ösma "moon," K 33 ing the premise of a root *Sam 32 L�ym-, sm-"to blush." BENVENISTE' s Avestanfräsmi-"*glowing," and Sogdian )sama-ka-" as the basis ofMunjiyu­ reconstruction of"une forme commune *a(x is hard o f north- eastem Afghanistan magd"moon" and its cognates in the dialects MORGENSTIERNE 1938, pp. 66-67. Ibid., p. 312. 23 Ibid., p. 471. etc. might be early loanwords. 24 Cf, ibid., p. 466, with the caveat that ruxn prob different treatment is seen in aßrü>vropo, A vv. s. , ssary o 25 g1 , See SIMS-WILLIAMS 2007 o anword (ibid., p. 184b). l a r o iant r va ctal diale a e b ably meaning "without mercy", which may 26 GERSHE\!1TCH 1954, §257. 27 HOFFMANN 1975, p. 190. 288-289. 28 CHEUNG 2007, pp. 428-429; Rrx 2001, pp. 29 CHEUNG 2007, p. 202; RJX 2001, pp. 292-293. 30 BENVENISTE 1936, pp. 230-231. 31 HENNING 1955, p. 174. 32 GERSHEVITCH 1959, p. 233. CHEUNG 2007, p. 371. 33 SIMS-WILLIAMS 1985, p. 180; differently

The Name ofthe Kushan Goddess Op1-w. to reconcile

453

with the proposed root *. 1 reference to "formations que F preverbe diffärencie" one 111igh��:du:::a� �ENVENISTE regarded the init. ial · *a- as a preverb (perhaps a sl1ortened .e:1om1 of *ä-?) ·., but S_tICh an mterpretat 10 n seems arbitrary. One m.ight ., do b-etter to reconstru ct a form wi th preverb *us-"up' " \ ka-. as an antony111 t_ o tl1e s ogdrn giving *u/sl-fma-., *u 1�s;-sma' n and Parthian forms 1'/ ' with preverb *ni- "down' " h. l ref,er to the _d1sappearance or absence of light. _ So far as I can see this is th:o�� 1 �u�I< w��ch woul� allow one :o co111bine all ::: :: the "1110011" w,o rds dis cu s s ed a unJI yumaga, P�s�to ogmay, Pahlavi 'dy.fozy, Zazaki ä.foza etc . - under a sin e root.--�s such, ,1 t i s ?o le ss worthy of , consideration than the derivation fro111 t( :�;::,� to wax, wh 1 ch only see111 s to be applicable to the Bast Iranian"111oon" Unfortunately, the only known portrait of Oµµa/0 �µo, th�t on the coin ofHu'.· vi shka where she is paired with O o = S1va , co�tams .no i conographic detail which wo uld allow one to affinn or1 y a connect1011 w1th the 11100 n. Given the well-established lunar connections f � na, one coul� arg�e that the parallelis111 � � between .the issues inscribed Oµ.µo 0r11„o and those mscnbed Nava - 0�l'ho, t. o­ �ether wlth the clear association between Oµµa and. Nava m · the Rabatak mscnpt 10 n , i s at least consi stent with h nnect 1 on ; but the ,�rgu111ent � s not ' :��111:ntt�at Oµµa a�d Nava were cons1dered compelling. To 111y 111ind , GNoLI s s to be equivalent" goes beyond wha� can really be ascertamed fro111 our li111 ited 111a· · terial .34 Si111ilarly' Oµµa's l1'kely .ep1tl1et omapp -r O"g 1onous" m the Rabatak inscription line 9 is certainl p r ough for a 111oon - dei ty but not sufficiently ;��:::i:�� distinctive �o, be reg arXe� ! seu l le

21

22

34

GNOLI 2009, p. 144.

454

Nicholas Sims-Williams

REFERENCES

rphy ' 2nd ed. C alcutta 1956. . . T J. N . BANERJEA'. The Development ofHmdu 1conogra F BARATTE O · BOPEARACHCHI, R · GYSELEN _, ' N . SIMS-WILLIAMS: "Un plat romam mscn't - . en bactrien et pärs ig.,, I n: ObJets et documents inscrit en parsig. Bure s-sur-Yvett e 2012 [2013] pp.9-28 (R es ori entales 21). . . 228 (1936) 'pp.193E. BENVE�STE: "Notes parthes et sog di ennes." In: JournalAsiatlque, 235. . C l . l D' t· narv of the Jranian Verb. L ei den 2007. . . �. F�:;;'?:��n�i;�;io:�tRabat�k e t l'origi ne de l'ere saka." In: Journal Asiatlque, 286 (1998), pp. 571-651. . I. GERSHEVITCH: A Grammar ofManich�an Sogdian. Oxford 1954. -TheAvestanHymn toMithra. Ca��ndge � 95_9: G. GNOLI: "Som e No tes o n t he R ehg1ous S1gmüca nc e of �h e R a_bat. ak lnscriptio n." In : r Exegisti monumenta. Festschrift in honour of Nicholas S1ms Wi llia':1s, d W· SUNDER MANN A HINTZE F.DE BLOIS.Wi esbaden 2009, pp. 141-15.� (� ramca ;7) . . R . G LB Sy t Chronolo ie derMünzprägung des .Kusänreiches. �1enna � 984.. W.B.oHE.NNING. . , .: �n:;;;e Ancient L!nguage of Azerbaijan." In : Transactions oj the Phzlological _ Society (1954) [1955], pp. 155-177. K. HOFFMANN: Aufsätze zur Indoiranistik, 1. Wi esbaden \975. H. HUM BACH: Die Kaniska-Inschrtft von Surkh-.Kotal. 'Yt�sbaden 196�. J. LEE, N. SIMS-WILLIAMS: "The A ntiquiti es an d I nscnp t10n ofTang-1 S a:fiedak· „ In· Silk Road Art and Archaeology, 9 (2003), PP· 159-184· . P.B. LURJE: Personal Names in Sogdian Texts. Vienna 2010 (lrarusches personennamenbuch' WS)· · G.MORGENSTIERNE An Etymological Vocabulary of Pashto. O slo 1927. - Jndo-Iranian Frontier Languages, II. Oslo 1938.

. . , . - - ,J

���=����

6::�:i::;::�:::���:1:J;��tuseum

B.�. Bulletin. Calcutta (1995) [1998] . H · RIX' ed...LTV. · · Lexicon der indogermanischen Verben, 2nd e d. wi esbaden 2001. l e1ey-Los A ng el es 1967 JM ROSENFIELD: The Dynastie Arts of the .Kushans . B erc . SIMS-WILLIAMS: The Christian Sogdian manuscript C2. Berlm 1985 (Ber1'mer· Turf:an-

N.

PRODS ÜKTOR SKJ.tERV0 Harvard

·

t

-- ��1 ;�;f;ian D eed of Manumission." In: Silk Raad Art and Archaeology, 5 (1997/98) [199�], pp. 191-21/ · ArorthernA-F. Jl.ghanistan II: Letters andBuddhist Tex ts. London --Bactrian D ocumen,sfirom 1v1

- ,�i:

Bactrian Inscrip tion ofRabatak: a N ew R ea di ng." I n: Bulletin of theAsia Ins titute 18 (2004) [2008], pp. 53-68. - Bactrian Personal Names. Vi enna 2010 �Iranisc�e s .P ers nn.a nb eh 11/7) N. SIMS-WILLIAMS, J. CRIB :B "A New Bactn an Inscnp tton o;�ams;;a e Great ,; In: Silk Raad Art and Archaeology, 4 (1996), PP· 75-142.

t:

Khotanese Land Purchase Deeds One ofmy fondest memories ofthe great scholar and incredibly generous col­ league and friend who was GHERARDO GNOLI is an evening we sat in his library at his home in Rome discussing the existence of Zarathustra. We never got to share the same view on this topic, but not much ofthe Johnny Walker Black Label we did share on this occasion was left at the end of the discussion. Of the five documents presented here, two (Or.6393/2, Or.6397/1) are in the British Library, Lond.on, and three (SI P 96.8, 103.17, and SI P 103.2+100.3) in the Oriental'Institute, St. Petersburg. Or.6397/1 and Or.6397/1 are online at the idp database (http://idp.bl.uk/) in good, readable reproductions. The four pieces in the Oriental Institute were published in R.E. EMMERICK and M.I. VoROB' YEVA-DESYATOVSKAYA's edition (see below); in EMMERICK's Nachlass, there was also a complete set of high-resolution .psd files of the entire published collection, which have made it possible to improve on some of the readings. The documents all concem "sale" of lands. Three of them have dates in the reign ofVisa' Väharµ, who reigned 767-802 (his 36th year, latest dated document), one in his 20th year = 786. The persons involved are, apparently, all from the same period and the same location, the Six Towns of Khotan, and appear together or with the same persons in numerous other documents. "General" (späta) SI