Problem of Archaism and Innovation in the Eastern Iranian Languages (dissertation)

Univerzita Karlova v Praze Filozofická fakulta Ústav obecné lingvistiky Obecná jazykověda – Indoevropská srovnávací jazykověda Ľubomír N o v á k Problem of Archaism and Innovation in the Eastern Iranian Languages K problému archaismu a inovace ve východoíránských jazycích Teze vedoucí práce – Doc. PhDr. Petr Vavroušek, CSc. 2013 The Eastern Iranian languages form an independent group within the Iranian branch of the Indo-European languages. The presented thesis aims to present an outline of development of the Eastern Iranian languages – as languages develop, they usually start to differ from its relatives by development of various innovations and/or by preservation of archaisms. The spread of innovations and preservation of archaisms may vary in individual languages or dialects and study of sets of common innovations and/or archaisms may characterize grouping of languages of a given branch. To see the Eastern Iranian archaisms and innovations I have decided to focus on three fields of study – 1) an outline of the Eastern Iranian languages, 2) historical grammar of SogdianнandнYaghn bīнandн3ц lexical study. The first part will be dedicated to the description of the attested Eastern Iranian languages and dialects – each language (or a subgroup) will be briefly described with focus on common data about the individual language(s), with an overview of main phonetic changes and grammar outline. For the overviews I will mark only some archaic and innovative features of the individual languages as for each language can be written separate book on its historical grammar and phonology. I would also like to (re)examine commonly accepted grouping of the Eastern Iranian languages into the Northern and Southern branches as it seems to me that this grouping needs a new revision. TheнsecondнpartнwillнpresentнcomparationнofнdevelopmentнofнSogdianнandнYaghn bīн – i.e. two languages that are considered closely related by many scholars (e.g. BOGOLYUBOV 1956; KLIMCHITSKIY 1935), but none of them has ever presented thorough study of their differences – Yaghn bīнwasнinнcommonнjustнconsideredнasнaнdialectнquiteнdifferentнfromнliteraryнSogdianынByн comparation of phonology and morphology of both languages I would like to show main differences between them and if possible I would like to try to define interrelationship of Sogdianн andн Yaghn bīын Theн comparativeн studyн ofн Yaghn bīн andн Sogdianн hasн beenн takenн intentionally – as both languages are comparable from diachronic point of view, their comparisonн mayн answerн moreн questionsн thanн justн theirн “dialectal”н relationshipын Historicalн developmentн ofн Sogdianн andн Yaghn bīн willн beн comparedн withн theн otherн Easternн Iranianн languagesнwithнfocusнonнtheнPāmīrнgroupы Theн thirdн partн willн presentн aн studyн ofн Sogdianн andн Yaghn bīн lexiconын Iн have compared Yaghn bīн andн Sogdianн lexiconн accordingн toн theн “Swadeshн List”н ofн я07н words combined with “StandardнWordнListнItems”нpresentedнinнSociolinguistic Survey of Northern Pakistan. The lexical items of both languages will be supplemented by their etymology. The choice of the Swadesh List was not motivated by attempts of glottochronological study of both languages – I just wanted to exploit an accepted list of basic vocabulary. Both lists try to present unbiased choice of basic vocabulary so in this issue I have also to study eventual loans (mainly in case of Yaghn bīцы As can be seen from outlines of all three parts, my study of the Eastern Iranian archaisms and innovations aims to present new classification of the Eastern Iranian branch with focus on position of Sogdian and Yaghn bīнwithin this language branch. ·1· *** The Iranian languages form a group of genetically related languages and dialects that developed from the Indo-Iranian branch of the Indo-European languages. By use of methods of historical and comparative linguistics we can explain the origin of the Iranian languages as a split of the Indo-Iranian branch of *Proto-Indo-European language. The original *Proto-Indo-Iranian languageнbrokeнupнintoнtheнfourнmainнbranches:нIranianōнNūristānīнхorнKāfirцōнDardicнandнIndoAryan. Particular prehistoric dialects of Indo-Iranian share with *Proto-Indo-European (and also with many other Indo-European languages) many common features – so called archaisms as well with series of innovations that set them apart from the proto-language. Some of the innovations can be observed in more branches of the Indo-European languages, but are not phenomena proper to the original system of reconstructed *Proto-Indo-European. The Iranian languages are divided into two main branches – Western and Eastern. Their division is based on agreed conventional brake up of two Old Iranian dialects according to their geographical location to the East and West respectively from the deserts of Central Iran (ÈDEL MAN 1986, 3). Present geographical spread of the Eastern and Western Iranian languages and their speakers has changed due to historical migrations of the Iranian peoples (e.g. Western IranianнBal chīнisнnowadaysнlocatedнinнEasternнIranнandнWesternнPakistanнorнtheнEasternнIranianн Ossetic is to be found on the Caucasus), the contemporary location of the Iranian languages is not relevant for their classification. The Iranian languages can be thus considered as an offspring of the Indo-European proto-language with which they are connected by genetic relationship and a preservation of some (*Proto-)Indo-European archaisms, on the other hand they differ from *Proto-Indo-European by several innovations which define this language family from historical point of view. We are informed about the history of the Old Iranian languages by means of indirect sources. Herodotus for example mentions several Scythian words, in one case he even presents an etymology (HERODOTUS IV, 110; HINGE 2006). He also mentioned that the Sauromatians speak the language of Scythia, but they do not speak it well because the Amazons did not learn properly the Scythian language – Herodotus mentioned that the Amazons married some Scythians and by this the Sauromatian nation came into being (HERODOTUS IV, 117). Herodotus also writes about an older poem, Arimaspea, written by Aristeas of Proconnesus (HERODOTUS IV, 13). It is said that Aristeas described the habits and the language of Scythian Issedonians (Issedones) and Arimaspians (Arimaspi) who dwelled in regions to the North-East of the Pontic or Black Sea (ALEMANY I VILAMAJÓ ю999цынUnfortunatelyōнAristeas нArimaspea has not came down up to these days, it is only mentioned in the Histories of Herodotus and also in Π ὕ by Longinus and in Chiliades (or Book of Histories) by John (Ioannes) Tzetzes (TZETZES, Chil. VII, 686-692). In the Anabasis of Arrian there are mentioned several local tribal and personal names of Central Asia, but we miss any reference to the languages of the region, the only relevant information is that the river Ἰ ά η хSīrн Daryāцн wasн calledн Ὀ ά η in a language of barbarians of Sogdiana (ARRIAN IIIōн30ыю3цынInнStrabo sнGeography is mentioned, that ·2· the northern part of Ἀ (i.e. approximately area of modern Afghanistan, Eastern Iran, Tajikistan, Turkmenistan, and North-western Pakistan) is inhabited by Bactrian and Sogdian peoples who do speak similar languages (STRABOōнGeographyōнXXVōня:8цынTheнcityнofнKūrkatнinн present northern Tajikistan is known from the antiquity – it is spelled either as Κῡ ύ or as Κῡ έ ; we can discover more about the local Iranian dialect by the analysis of both Greek names: Κῡ ύ is probably a calque of Iranian appellative дKūruš-kąϑā- cityнofнCyrus ынWhatн is even more interesting is the form Κῡ έ χα α, it can be an attempt to render the local name дKūruš-kąϑā- (cf. Tjk. and Pers. Kūrkát); the Greek name is probably contaminated by another Greek word ἐ ά η theнfarthest нхprobablyнbyнan influence by the name of the city of Alexandria the Farthest – Ἀ ά Ἐ ά η, present Khujand, in the Soviet period known as Leninabad). City of Ῥ ά η mentionedн byн Ctesiasн ofн Cnidusн canн beн connectedн withн cityн ofн R shānн хR shānī ⁾ , Tajik цнinнTajikнBadakhshānнхABAEV 1949, 178). The Iranian language family is conventionally divided into two basic groups – Eastern and Western Iranian. Differences between these two groups begun to appear probably in the Old Iranian period and became more distinctive in the Middle Iranian period. Each of these groups later split into two subgroups – South and North subgroup. Among the North Eastern Iranian are classified Scythian dialects and *Sauromatian (in the Old Iranian period), Sarmatian, Alanic, SogdianнхMiddleнIranianнperiodцнandнOsseticнandнYaghn bīнхNewнIranianнperiodцынSouthнEasternн IranianнlanguagesнareнrepresentedнbyнdialectsнofнtheнSakaнхmainlyнKh taneseнandнTumshuqeseцōн 1 Bactrianн хMiddleн Iranianн periodцōн theн Pāmīrн languagesн хShughnī-R shānīнgroup ōн Yazghulāmīōн Wanjīōн Wakhīōн Ishkāshmī-Sanglēchī, Munjī-Yidghā andн probablyн Sarghulāmīцōн Pasht н andн Waṇetsīн хNewн Iranianн periodцын Questionableн isн classificationн ofн theн Avestanн languageн – it is probablyн oneн ofн theн Southн Easternн Iranianōн Khwārezmianн isн variously classified as North or Southн Easternн Iranian;н theн mostн complicatedн isн classificationн ofн Parāchīн andн rmuṛīн – some scholars claim them as North Western Iranian but some other hive off new – Southeast branch within Eastern Iranian. The South Western Iranian languages and dialects differ from other Iranian languages by significant isogloss Ir. *ʦ, *ʣ, *ʦu > *ϑ, *d (< * ??), *s; such isogloss, however, does not separate North Western Iranian languages from Eastern Iranian, cf. development of Ir. *ʦ, *ʣ, *ʦu > *s, 2 *z, *sp . Differences between the (North) Western Iranian and Eastern Iranian have to be looked up within other features. However, distinctive features cannot be found only on phonological level. There were not many phonological differences between the Eastern and Western Iranian in the Middle Iranian period, one of the essential features was development of word-initial voiced stops *b-, *d-, *gand development of clusters *ft and *xt. To establish a border between the Eastern and Western 1 2 Id est ShughnīōнShākhdaraīōнBajūīōнKhūfīōнR shānīōнBartangīōнRāshārvīнandнSarīq līы But in Wakhī *ʦu > š and in Kh tanese *ʦu > śś [ʆ]. ·3· Iranian, lexical (e.g. in many works presented example *gari- ×ндk‘u”‘- mountain нandн*kap - ×н ʦ ‘- fish ;н cfын SIMS-WILLIAMS 1989a, 168-169) and grammatical differences should be also taken into account. There can be mentioned some other words from lexicon that can be considered typical for the Eastern Iranian area: √ yr √’₍r /√ īr/, Khwārын yrʾ yr-, *abi-ar- to find, to obtain н>н Yazgh. vir-, Yagh. vīr-; двhгá d‘- blind н>нKh tынhanaʾnt ʾnd /aṁd/, Munj. ₍ā dəy, Pasht. ṛū dōн rmын hō d (but cf. Parth. hand); д‘u‘-sú⁾t‘-(ka-) cleanōнpurified н>нKh tынTumshuqынvasutaʾwsw ty, ʾws(ʾ)w tʾk ʾwsw tyy ь su d/, Oss. (without prefix) s dæg ‖ su dægōн Khwārын (with other prefix) (ʾ)fs d; дdráu‘- hair н>нKh tы drauẓw-y й ə⁽íй, Yagh. d‘ráu ‖ d ráu, Oss. ærdu ‖нærdo, Shugh. cī⁽ōнR shынcō⁽, Yazgh. c rmынdrī ×нPersын ō < д ‘ud‘-; дkąt‘- house н>нSogdын ktʾy, ktʾk qt, qty(y), ktyy qṯy /kə ( ) /kad(a)g/, Yagh. kat, Shugh. čīd, R shынKhūfынčod, Bart. čȫdōнRāshrvынčǖdōнSarīqынč“d, Yazgh. k d, Munj. -ka- (but Sogd. ḱay, Yidgh. kʶε , Ave. kata- (+ Parth. Pahl. kdg) ×н Persын ⁾ā á < ʾnʾk(h) xʾnʾ xun, Ishk. xon, Sangl. ⁾ā ); ʾkwt-y kwt-y, qwt-y /ᵊkʷətíьōнBactrын /kud/, Yagh. kut, дkút‘-, дkutīOss. kʵ ʒ ‖нkuy, Shugh-R shынkudōнSarīqынk d, Yazgh. k°od, Ishk. k d ×нPersынsag < *ʦuą-ka-, Med. σπάκαн хbutн Kh tын śv“, Wakh. š‘č, Pasht. spay (f. spə ), Waṇ. sp‘ в”и spīгōн rmын ᵊspuk, Parāchынᵊspȫ); д á ϑamy m(ʾ)y my , my(y) myϑ, mỿϑ, myd ьmēϑьōнKhwārынmyϑ ьmēϑ/, Yagh. mēs ‖ mēt, Shugh. ēϑōнR shынKhūfынBartынRāshrvын īϑōнSarīqынmaϑ, Yazgh. miϑ, Ishk may, Sangl. ē , Munj. Yidgh. ī⁾ ×н Persын rō₎ < дráuč‘- (but Pasht. wraʒ, rwaʒ, Waṇ. wrezōн rmынwriez, wrioz); √pc ʾ(ʾ)z ь√pə pc ʾz-; *pati-gaʣ- to accept н>нKh tынp‘jā₍ssʾn ьsānьōнOssынson ×нPersынduš á < *ʦ - enemy н>нKh tынTumshuqынsā ‘дduš-mana- (cf. SIMS-WILLIAMS 1989a, 169; SIMS-WILLIAMS 1996b [online]). Division of Eastern Iranian languages into Northern and Southern branch (and eventually South-easternнbranchнifнweнwillнconsiderн rmuṛīнandнParāchīнasнmembersнofнtheнEasternнIranianн languages) is often used by many scholars, only few of them explain the criteria of such classification, so it seems that this division was more based on (modern) geographical distribution of the Eastern Iranian languages. Some criteria for dividing the Eastern Iranian languages were presented by Vera Sergeevna RASTORGUEVA (1966, 198), but only a few of them fit the classification of the Northern and Southern branches of East Iranian. Comparation of selected sound changes and other features can characterize some isoglosses in the Eastern Iranian languages, but some changes are common for many of these languages regardless to their ranking to the Northern or Southern branch. Instead of classification of the Northern and Southern branch, there can be better postulated a dialect continuum than two different branches; the only (?) branch that seems to show more distinctive features is the South-eastern ·4· branchнwhichнcontinuesнinнtheн rmuṛī-ParāchīнsubgroupынAsнdistinctiveнfeaturesнofнtheнSouthн Eastern Iranian branch can be considered 1) preservation of archaic formation of plural (i.e. absence of innovation of plural form by adding an abstract suffix *- ); 2) sonorization of intervocalic *-š-; 3) change of Ir. *rd, *rt; 4) change of Ir. *rʣ, *rʦ and 5) emergence of innovated form of the second person plural personal pronoun from combination of forms of the second person singular and first person plural. All the above mentioned changes have not emerged in all South Eastern Iranian area: feature 1) haveн notн tookн placeн inн Yazghulāmīн хandн except some non-productiveн formsн inн Ishkāshmīц;н intervocalic *-š- has not been sonorized in BactrianнandнprobablyнalsoнinнSarghulāmī;нchangesнunderнtheнpointн 3) have not taken place in BactrianнandнWakhī;нinнMunjīōнYidghāнandнWakhīнхandнprobablyнalsoнinнBactrianцнhasнnotнtakenн place change point 4); innovated forms of plural the second person plural (point 5)) are present inнallнSouthнEasternнIranianнlanguagesōнbutнinнParāchīōнSakaнdialectsнandнinнKhwārezmian they come from a different source than from the above mentioned. The Modern Eastern Iranian languages can be divided into five branches: I Northern (Sogdo-Scythian) group; II North-eastern (Saka) group, III Central (Pā īr) group, IV Southern (Paṭhā ) group and V South-eastern (H dūkush) group. Group I can be defined by innovated plural ending *- - хcomparableн toн Yazghulāmīцōн preservationн ofн intervocalicн *-š- (shared with BactrianнandнWakhīнbutнexcludingнOsseticцынGroupsнIII, IV, V have undergone common change of form of the second person personal pronoun, in languages of these groups there are innovated forms of plural, they may be influenced by Indo-Aryan or Dardic pronouns. Innovated forms of the second person plural often comes from combination of personal pronoun of the second person singular with form of the first person plural *ta/u-*ah -(k/xam-), or *ta/u- ṣ ma-нcopiedнfromнIndo-AryanнхcfынMaiyāнtus; Ṣiṇāнtsa/o;нLahndāн tusцнdifferentнformнisнjustнinнParāchīынGroupsн II and IV share sonorization of word-initial *fr-, *ϑr-, *xr-. Some isoglosses can be demonstrated on following examples: дčáš ‘ - eye I cвšг -y cm-y(y), cš -y c(y)m-y, cš -y ьčɨхšцmíь;нOssынcæst, casm ‖нcans window-opening II Kh tынts“ʼ ‘ III Ishk. com, Sangl. cāṃōн Zēbы cō , Munj. čō , Yidgh. č‘ , Shugh. Baj. cē , , Wakh. čə(ẓ)m R shынKhūfын cā , Bart. cē ōнRāshrvынcī ōнSarīqынcem, Yazgh. V rmынc ī, čī , c m ? Khwārынcm-, c - /camma/, Ave. č‘š ‘ *ϑr ‘- three ʾ ry ʾ ryw (ʾ) ry ʾ ry(y) š₍ /ᵊšai/, Yagh. s‘rá₍ ‖нt rá₍, Oss. ærtæ I II Kh tынdrai, Tumshuq. dre III Yidgh. ⁾ ray, ⁾uro₍, Munj. ⁾ r‘₍, Shugh. arayōн Bajын Bartын R shын ‘rā₍ōн Sarīqын ·5· aroy, Ishk. r ₍, Sangl. rō₍, Yazgh. c ₍, Wakh. trūв₍г {Bactr. α ηιο /hərēyь} IV Pasht. drē, Waṇ. dre V rmынṣȫ, ṛīōнParāchынšī, šu ? Khwārынš₍ ьšēьōнAveынϑrā ō д uš you I (ʾгš ʾ w, ʾš ʾ h ʾš ʾx(w), š ʾx š ʾx /ᶤ Oss. s ‘⁾ ‖ sumax II Kh tынuhu, , u äōнLKh tынama , Ishk. t ⁾, Sangl. təməx, Munj. ō”, Yidgh. III Wakh. R shын Khūfын tama, Bartын Rāshrvын t‘ āšōн Sarīqын t‘ ‘š {Bactr. , ьt māxōнtumāxōнtamāxь} , , Waṇ. tās IV Pasht. V Parāchын⁽āōн rmынtōs, t₍ūs ? Ave. ₍ū əmōнKhwārынh y šᵘ , , , Shugh. , , дg‘uš‘- ear I Sogd. ⁽š / šьōнYaghын ūš, Oss. qus ‖н os, Scyth. ⃝ ο II Kh tынgguv ‘-, ggū III Wakh. iṣ, Ishk. ḷ , Sangl. ōḷ, Shugh. ōн R shын ō⁽ōн Sarīqын awl, Yazgh. əvon, Munj. ū₍, Yidgh. IV Pasht. ⁽‘ǵ, waẓ V rmынgōī, gō₍ōнParāchынgū ? Khwārын wx / ōx/, Ave. g‘oš‘- The issue of reclassification of the Eastern Iranian languages was only outlined in this thesis, the question still waits for its thorough examination. Valentina Stepanovna Sokolova studiedнgeneticнrelationsнofнYazghulāmīнandнtheнShughnī-R shānīнgroupнхSOKOLOVA 1967) and laterн relationsн ofн theн Shughnī-Yazghulāmīн groupн withн Munjīн хSOKOLOVA 1973). Studies of geneticнrelationsнofнMunjīнandнYidghāнwithнBactrianнandнalsoнinterrelationsнofнBactrianнwithнtheн Paṭhānн languagesн canн answerн theн questionн ofн positionн ofн Bactrian within the Eastern Iranian groupынInнaнsimilarнwayнcanнbeнstudiedнrelationshipнofнWakhīнandнtheнSakaнlanguagesн – Wakhīн appears to share several isoglosses with the Saka languages, but the language shows probable adstrate or substrate phenomena thatнlinkнitнcloserнtoнtheнlanguagesнofнPāmīr. Classification of theн languageн ofн Khwārezmн remainsн toн beн ratherн complicatedн – Khwārezmianн sharesн severalн isoglosses with Alano-OsseticнlanguagesнandнwithнtheнlanguagesнofнPāmīrнonнoneнhandōнonнtheн other hand there are some similarities with North-Westernн Iranianн Sangesārīн хcfын Aʿ AM – WINDFUHR 1972), there are also some isoglosses shared with Sogdian (cf. SIMS-WILLIAMS ю989aōню70ц;нsummaryнofнpossibleнconnectionsнofнKhwārezmianнwithнAvestanнhaveнbeenнpresentedн by David Neil MACKENZIE (1988) and by Vladimir Aronovich LIVSHITS (1962, 140). ·6· Majority of the Eastern Iranian languages can be studied mainly from synchronous point of view – these languages and dialects are attested as individual stages of the Eastern Iranian branch but with some exceptions we do not know their older development stages. There is exception within the North Eastern Iranian branch – in this case bothн Yaghn bīн andн Osseticн canн beн compared with their closely related ancestors. The development of Ossetic can be continuously observed from the Old Iranian period – there are many similar features in the Scytho-Sarmatian dialects and in Alanic that can be compared with Ossetic and we can even suppose that Ossetic is a modern descendent of one of Alanic (or Sarmatian or even Sauromatian) dialects. Similar situationн appliesнforн Sogdianн andн Yaghn bīн – these two languages are very similar from many points of viewōнYaghn bīнhasнbeenнevenнlabelledн N“o Sogd ‘ by some authors (BOGOLYUBOV 1956; KLIMCHITSKIY 1935; SKJÆRVØ 1989, 375-376), nowadays many scholars are inclined to believeн thatн Yaghn bīн mayн comeн fromн someн non-attested non-literary dialect of Sogdian (BIELMEIER 1989, 480; SIMS-WILLIAMS ю989bōню73цōнAl bertнLeonidovichнKhromovнexpressedнanн opinionн thatн Yaghn bīн couldн haveн originateн inн aн non-attestedн Sogdianн dialectн ofн Ustr shanaн (KHROMOV 1987, 645), unfortunately there is no relevant data to confirm this hypothesis. On the basis of the above mentioned data we can declare that a thorough diachronic and synchronic study of the Eastern Iranian languages is possible in its Northern branch – but in the case of Ossetic comparable material lies mainly in lexicon, development of grammar and syntax is blurred (cf. ABAEV 1949). It is of course possible to outline historical development of other (New) Eastern Iranian languages, but in these cases it is necessary to deal only with methods of historical and comparative linguistics because there are not attested direct ancestors of these languages. Based on the above mentioned facts the main theme of this thesis is the comparison of SogdianнandнYaghn bīн– information on Sogdian are available in a large text corpora from which weнcanнlearnнaboutнSogdianнgrammarōнlexiconнandнsyntax;нYaghn bīнasнaнlivingнlanguageнisнsoнfarн undrawn repository of knowledge – toн linguistsн Yaghn bīн isн knownн aн littleн bitн moreн thanн hundred years, within that period of time some texts, grammarsн andн lexiconsн haveн beenн publishedōн atн theн presentн timeн aн researchн onн theн Yaghn bīн languageн andн ethnographyн isн underн patronageн ofн theн Academyн ofн Sciencesн ofн theн Republicн ofн Tajikistanōн whereн underн theн Rudak н Institute of Language and Literature falls the D“p‘rt “ t o” Y‘gh ō’ī Stud “s. In case of Yaghn bīнхandнtheнotherнModernнEasternнIranianнlanguagesнexceptнOsseticцнitнisнalsoнnecessaryн to follow development of Modern Persian, mainly its variety in Tajikistan. AнcomparisonнofнtheнSogdianнandнYaghn bīнmaterial can solve the issue of the relationship of both of these languages. It can be supposed that both languages developed from one common North Eastern Iranian proto-language or proto-dialect, such proto-language will be labelled *Proto-Sogdic (i.e. a Centralн Asiaticн varietyн ofн“Scythian”цн hereын Laterн чProto-Sogdic split into two (or even more) main dialects – *Proto-Sogdian and *Proto-Y‘gh ō’ī. Both *Proto-Sogdian and *Proto-Yaghn bīнareнreconstructedнasнpredecessorsнofнtheнattestedнlanguagesн– Sogdian and Yaghn bīōн besidesн thoseн twoн languagesн thereн mayн haveн beenн Sogdianн dialectsн ofн Bukhārāōн ·7· Ustr shanaн andн Zhetisuн – чBukhāranн Sogdianн isн attestedн byн severalн shortн textsōн чZhetisuн Sogdian is attested on several inscriptions and from historical sources while чUstr shanianн remainsн toн beн aн hypotheticalн Earlyн Mediaevalн ancestorн ofн Yaghn bīōн чUstr shanianн isн alsoн thoughtн toн beн anн ancestorн ofн hypotheticн чZarafshānīн languageьdialectн whichн remainedн asн substrateнinнTajikнdialectsнofнMastch hōнFalgharнandнF nы The *Proto-Sogdic language split into two reconstructible dialects – *Proto-Sogdian and *Proto-Yaghn bīын Forн descriptionн ofн theн historicalн phonologyн ofн Sogdianн itн isн necessaryн toн outline several stages of development of the Sogdian language, see following scheme: th th th th 4 -5 cent. 7 -9 cent. th half of the 11 (?) cent. (middle ages) up to cca. 1900 from cca. 1900 *Proto-Sogdic *Proto-Sogdian / *Proto-Yaghn bī *Old Sogdian Preclassical Sogdian Early Classical Sogdian Classical Sogdian ва ьukhār‘ d ‘l“ctг Postclassical Sogdian (& Zhetisu dialect) (death of Sogdian) чZarafshānī EarlyнModernнYaghn bī ContemporaryнYaghn bī language of Sogdian translation of Aṣəm vohū the Ancient Letters Christian document C 2 majority of texts ьrāh ī docu “ tsж Chr st ‘ docu “ t C 5 preserved only in central Tajik dialects pr“s“rv‘t o o” ‘jhūl ō ‘ d ϑ Yaghn bīн appearsн inн someн aspectsн moreн archaicн inн comparisonн toн Sogdianн – it preserves archaicнpositionнofнstressōнaugmentнхthoughнtheнaugmentнhasнbeenнinnovatedнinнYaghn bīцōнand also better preserves inherited vowels (i.e. there is no reduction of unstressed vowels to Schwa as there were no phenomena related to the Sogdian Rhythmic Lawцн andн Yaghn bīн dialectsн showн that origins of both dialects can be of an old date. Archaic is also formation of ergative constructionн inн Yaghn bīн andн anotherн archaismн sharedн withн Avestanōн Kh taneseн andн Khwārezmianн isн preservationн ofн archaicн preterite ending of the third person plural *-ār. On contrary, Sogdian shows archaic features mainly in morphology – the operation of the Sogdian Rhythmic Law preserved archaic inflectional system for light stem words, and also verbal morphology – Sogdian preserves moreнinheritedнverbalнformsнthenнdoesнYaghn bīы Both languages share some innovations – main similarity is development of nominal inflectionн inн Yaghn bīн andн inн caseн ofн theн heavy stems in Sogdian – development of direct and obliqueнcasesнisнcomparableōнmoreoverōнYaghn bīнlostнvocativeнcaseынAnotherнsharedнinnovationн (typical also for other North Eastern Iranian languages) is formation of plural with the abstract suffix д-tвu -. Sogdian innovated ergative construction as it replaced copula by the verb дdārto hold нforнtransitiveнverbsнхcfынsimilarнdevelopmentнinнKhwārezmianцōнanotherнinnovationsнcanн be seen in new suffixed forms of verbal inflection. The most important innovation in Sogdian was a stress-shift х“Stress III”ц and subsequent operation of the Rhythmic Law – in this case originally phonetic change strongly influenced morphology and phonology of the language. Yaghn bīнinnovationsнshowнspreadнofнprefixedнaugmentнbyнanalogyнto all verbal forms regardless ·8· of their original prefixes and also reanalysis of verbal endings – original durative ending - št serves to form simple present and future tenses or as durative marker for the imperfect. Development of stress in the *Proto-Sogdic language is essential to understand phonology of Sogdianн andн Yaghn bīн andн alsoн toн discoverн differencesн betweenн bothн languagesын Itн isн notн necessary to focus on position of stress in *Proto-Iranian because there was a stress shift in *Proto-Sogdic from which both languages developed. The reconstruction of *Proto-Iranian stress is complex – it can be supposed that the *Proto-Iranian stress was mobile and its position was similar to Vedic. For the reconstruction of Old Iranian stress is essential to study stress in Pasht н хGRYUNBERG – ÈDEL MAN 1987, 38-39). Position of stress changed also in the other Easternн Iranianн languagesōн mainlyн inн theн Pāmīrн languagesн whereн stressн shiftsн causedн eitherн syncopation of unstressed vowels or changes of stressed vowels under operation of ā- or i-Umlaut;нnowadaysнallнPāmīrнlanguagesнofнBadakhshānнhaveнstressнonнtheнlastнsyllableы Itн seemsн thatн predecessorsн ofн bothн Yaghn bīн andн Sogdianн underwentн theн sameн orн veryн similar stress shifts, the results of operations of stress slightly differ in both languages. Some Sogdian words point to original *Proto-(Eastern-)Iranian stress, the place of this stress (Stress I) can be reconstructed after operation of izyrn /zeṙn/ < *ʣár‘ ‘- gold н (SIMS-WILLIAMS 1989b, 181). Stress later shifted to another position (Stress II): the stress fell on penultimate or antepenultimate syllable. Words with penultimate stress were either disyllabic words or words with a penultima containing long syllable i.e. syllable containing either long vowel (long either naturally or rhythmically) or a diphthong (diphthong could have been formed also by a nasal or д r) in a closed syllable; in other positions the stress shifts on antepenultima. Positionнofн stressнinн Yaghn bīн comesнfromн the results of operation of the Stress II, this stress can be observed in Sogdian in results of operation of i-Umlaut of several words. Such stress shift is also probably related with change of its strength – manyнunstressedнvowelsнхinнYaghn bīнoftenн all syllables) were reduced or even syncopated, mainly short vowels directly preceding or following a stressed syllable. Other stress shift (Stress III) took place only in Sogdian, and this change is related operation of the Sogdian Rhythmic Law; but no such shift hasн takenн placeн inн Yaghn bīын Theн Rhythmic Law, which was originally only a phonological feature caused many other changes in Sogdian morphology. The Rhythmic Law divides Sogdian words into two groups – in so-called light and heavy stems (cf. SIMS-WILLIAMS ю984;нGMSн§484-530). As the heavy stems we can classify words with stressed root syllable, in fact stress falls on the first possible rhythmically long syllable (i.e. either on a long vowel or on a diphthong – in this case diphthongs are considered groups , u, Vṙ, Vṁ in closed syllable), the heavy stems end with a consonant in majority of words. In the light stems stress shifted to the ending – the light stem words do not have rhythmically long root syllables and the stress shifted towards the end of the word, and thus *Proto-Sogdian endings have been preserved. Emergence of the Rhythmic Law also influenced reduction of vowels in unstressed syllables, mainly when they followed stress – in the heavy stems the original endings ·9· disappeared but they remained in the light stem forms. Subsequently the last stress shift (Stress IV) appears – this stress shifts to the ultimate syllable (Nicolas Sims-Williams suggests thisнdevelopmentнafterнanнanalysisнofнSogdianнdocumentsнinнtheнBrāhmīнscriptōнsomeнevidenceнofн this feature can be found in several vocalized documents in the Syriac script; SIMS-WILLIAMS 1996a, 312-313) As indicated above, mere shifts in stress position presented a significant feature which resultedнinнfurtherнsoundнchangesнinнSogdianнandнinнYaghn bīынBothнlanguagesнprobablyнsharedн similar changes of stress for quite a long period of time during their common development. Yaghn bīнretainedнoriginalнstressнonнхanteцpenultimaнхiыeынStress II) Sogdian, however, was more progressive and there developed another innovation in stress (Stress III), this shift was motivated by rhythmical weight of a syllable – the operation of Stress III and the Sogdian Rhythmic Law is oneнofнtheнmostнimportantнdistinctiveнfeaturesнdistinguishingнYaghn bīнfromнSogdianы We can distinguish three development stages of stress changes: Stress I, Stress II and Stress III – the first two stages can be observed in both languages (there are sources for position of the Stress I mainly in Sogdian, but they can be suggested in Yaghn bīцōн Stress III is just Sogdian development – in the scientific literature the Stress III is labelled as the Sogdian Rhythmic Law. In the presented thesis I will use the term h₍th c L‘⁽ just for the outcome of the operation of the Stress III in all its complexity, mainly as a feature influencing Sogdian grammar; the label Stress III means only phonological shift of stress. In Late Sogdian Stress IV followed. A good example of all stress shifts can be seen in the following example: Stress I *aʣá I нхPasht. zə; Waṇ. ze; Munj. za; Yidgh. zo, zə; cf. Ave. azəm, Ved. ‘há ; Ide. *h “g ) > Stress II дá₎‘ (Proto-Sogdic дá₎u; Yagh. д ‘₎; Wakh. wuz; Ishk. az(i); Sangl. azə; azi; Yazgh. az; Shugh. (w)uz;нR shынaz;нKhūfынRāshrvынSarīqынwaz; Bart. ā₎) > Stress III ʾzw, /əzúьн> Stress IV Sogd. zw /zu/ (?). Thereн areн alsoн severalн phoneticн differencesн inн developmentн ofн Sogdianн andн Yaghn bīн – these features can be considered dialectal and probably they originally led to the assumption that Yaghn bīн may be a dialect of Sogdian. According to the analysis of stress shifts in languages derived from *Proto-Sogdic it can be suggested, that phonological development was also influenced by stress, namely in *(Proto-)Sogdian, where original short unstressed vowels changed to Schwa (ə or its allophone ɨцōнbutнremainedнunchangedнinнYaghn bīы Inн morphologyн theн differencesн betweenн Yaghn bīн andн Sogdianн ariseōн mainlyн dueн toн theн operation of the Rhythmic Law, but there are also other phenomena that have not been influencedн byн stressын Fundamentalн isн developmentн ofн augmentн inн Sogdianн andн Yaghn bīн – in Sogdian augment has been lost for all non-prefixed verbs, but it has been preserved as so-called internal augment for prefixed verbs (i.e. reflects of augment can be seen after a verbal prefix, in thisнcaseнprefixнusuallyнchangesнitsнphoneticнformнwhenнfollowedнbyнaugmentцōнbutнinнYaghn bīн augment remained as a distinctive feature of imperfect and was reanalysed by analogy for all verbs as a prefix even for those containing historical verbal prefixes. Other essential ·10· morphologicalн featuresн areн twoн archaismsн preservedн onlyн inн Yaghn bīн – preservation (and reanalysis) of peripheral preterite ending -ōr < *-ār < Ide. *-(o)ro / -(o)ror and preservation of imperfect ending of the first person plural -ō < *-ā ‘ inн Westernн Yaghn bīн хinн Easternн Yaghn bīнandнinнSogdianнtheнimperfectнendingнofнtheнfirstнpersonнpluralнhasнbeenнreplacedнbyн original optative ending д-‘ ‘ -ī , Sogd. -ē цын Theн factн thatн Yaghn bīн dialectsн developed two different imperfect endings of the first person plural may indicate an early split of *Proto-Yaghn bīн andн чProto-Sogdian, and subsequent innovation of imperfect endings in (*Proto-)Sogdian and *Proto-EasternнYaghn bīы During the development of the Sogdian language, Sogdian nominal morphology gradually simplified inflectional cases and light stem nouns changed their case endings and analogically switched to agglutinative inflection as is attested for heavy stems – the light stems formed minority of nominal roots and as there was double system of nominal inflection in Sogdian the language tended to avoid such dichotomy. As the light stem inflection switched by analogy towards the heavy stem inflection, there remained system of three cases – direct, oblique and vocative, i.e. case system similar to *Proto-Yaghn bīын Thisн reducedн inflectionalн systemн isн attested in late Sogdian Christian document C 5 (cf. SIMS-WILLIAMS 1982). Also verbal endings tended to be unified for both light and heavy stems. Similarity inн“agglutinative”нsystemнofнlateн SogdianнinflectionalнsystemнwithнYaghn bīнisнstrikingōнbutнonlyнformallyнхorнsayнonнsynchronicн level), but diachronically the development in both languages differ. The late Sogdian (or “C 5-Sogdian”цнsystemнofнnominalнinflection cannot be considered as a source for development ofнYaghn bīнinflectionalнsystemнasнthereнareнstillнdifferentнpatternsнofнstressнdevelopmentнinнbothн languages – diachronicallyн Yaghn bīн stillн preservesн stressн onн itsн positionн asн itн wasн inн *Proto-Sogdic (i.e. Stress II), but (*Proto-)Sogdian certainly developed later stress shift – Stress III that influenced also morphology of the language (i.e. so-called Rhythmic Law), and probably later on another stress shift appeared in (late) Sogdian – Stress IV. The shift towards the Stress IV can be probably connected with the above mentioned simplification of nominal inflectional cases as attested in the document C 5 – the tendency to equalize the three-case system of the heavy stems and the six-case system of the light stems led towards a heavy stem-like agglutinative system. There was probable opposite tendency in stress – it tended to shift towards the end of a word, such tendency can be seen in analysis of Sogdian versification by Elio PROVASI (2009, 351-353) whereas the final state of the Stress IV shift can be seen in the Sogdian documentsнwrittenнinнtheнBrāhmīнscriptнхSIMS-WILLIAMS 1996a, 312-313). Lexiconн ofн bothн Sogdianн andн Yaghn bīн differsын Thisн factн canн beн causedн byн twoн factsн – 1) Sogdian is attested in various documents, but majority of texts are religious texts so the vocabularyн oftenн doesн notн describeн “basic”н vocabularyн connectedн withн everyday life of peasants andнotherнcommonнpeopleнinнSogdianaōнbutнsuchнvocabularyнisнwellнattestedнinнYaghn bīнasнtheн Yaghn bīsн areн semi-nomadicн pastoralistsн andн theirн languageн preservesн manyн “indigenous”н terminology connected with animal husbandry and life in the mountains; and 2) there is ·11· approximatelyнaнthousandнyearsнlongнgapнbetweenнSogdianнandнхModernцнYaghn bīōнduringнthisн periodнtheн“worldнofнtheнSogdians”нchangedнconsiderablyнandнthisнdevelopmentнmayнbeнobservedн inнdevelopmentнofнYaghn bīнlexiconы After the fall of Sogdiana and gradual disuse of the Sogdian language (Arabic and) Persian became the lingua franca of Central Asia and Persian strongly influenced not only (Pre-Modern) Yaghn bīōн butн alsoн manyн otherн languagesн suchн asн theн Pāmīrн languagesōн Pasht ōн Indo-Aryan UrdūōнtheнNūristānīнandнtheнDardicнlanguagesнorнTurkicнUzbekōнKyrgyzнetcынModernнYaghn bīн preserves approximately 27% of indigenous vocabulary, other parts of lexicon are borrowings, calquesōнorнYaghn bī-PersianнхYaghn bī-Arabic etc.) compounds. Sogdian lexicon contains also number of borrowings, mainly from SanskritōнOldнTurkicнandнAramaicнхbutнexcludingн“Sogdian”н words written with Aramaic ideograms). Both languages also show similar patterns of word-formationōнevenнYaghn bīнcalquesнfromн Tajik show some Sogdic patterns of word-formationын Inн Yaghn bīн thereн stillн remainн manyн suffixes attested in Sogdian, unfortunately many of such suffixes are unproductive in the contemporaryн languageн хcfын GMSн §935-1166; LIVSHITS – KHROMOV 1981, 434-449; KHROMOV 1987, 665-670). Someн Yaghn bīн wordsн haveн noн Sogdianн responsesōн Sof ya Petrovna Vinogradova quotes several of them: úrd‘ eye ōн ayk daughter ōнrax mouth 3, ōs- to take нхVINOGRADOVA 2000, 310), there are many other words without Sogdian etymology, but some of those words have etymologyн inн theн Pāmīrн languagesōн eыgын Yaghын ayk daughterōн girl н mayн beн connectedн withн Yazgh. ‘č‘ǵ, Shugh. ācōн R shын acōн Sarīqын oc; Yagh. Saponaria Griffithiana Boiss. plant н ~ Khūfын wu m; Yagh. p‘rá Cousina umbrosa Buge plant н ~ Khūfын piram, Yagh. ‖н home-made paper-likeн thinн cottonн cloth н ~ Shugh. ⁾ ⁽ī , Bart. ⁾ ⁽ī čōн Khūfын ⁾ ⁽ī č, ⁾u⁽‘ ōнR shын⁾ ⁽ū č; Yagh. xᵘ crowōнmagpie н~ ShughынKhūfынk ⁾ pc and many other. The Yaghn bī–Pāmīrīнvocabularyнmayнbeнconnectedнwithнlocalнecologyнandнsemi-nomadic lifestyle or itн mayн evenн beн associatedн withн theн Pāmīr-Hindūkushн Sprachbund (cf. PAYNE 1989, 422-423). Unfortunately there are no attested counterparts in Sogdian. SomeнotherнYaghn bīнwordsнhaveнbeenнrecordedнinнpastнyearsōнbutнtheyнareнnotнusedнinнtheн modern language: man apple ōн red ōн yellow ōн šōu black ōн white ōн vᵘrʉk eyebrow ōн , ‘l’‘lá muchōн many н andн manyн otherн хcfын BOGOLYUBOV 1966, 359; KLIMCHITSKIY 1940; NOVÁK [in print]цōнsomeнotherнчEarlyнModernнYaghn bīнwordsнthatнwereн also similar in Sogdian were replaced by their Tajik similar-sounding counterparts: дvō (Sogd. ā цн garden н×нTjkын>нYaghын’ō , д ō⁾ (Sogd. ā⁾цн moonōнmonth н×нTjkын>нYaghын ōh month н (cf. BOGOLYUBOV 1966, 359). Fromн theн aboveн mentionedн pointsн itн thusн canн beн suggestedōн thatн Sogdianн andн Yaghn bīн areн closely related languages, but there is no evidence that showsнthatнYaghn bīнdevelopedнdirectlyн 3 Yagh. rax has attested Sogdian form r ʾk /rə . ·12· fromнSogdianынIfнweнassumeнthatнYaghn bīнdevelopedнfromнaнSogdianнdialectнweнhaveнtoнdefineн such dialect, but evidence of Sogdian dialects is quite deficient. It is certain that both Sogdian andнYaghn bīнdevelopedнfrom the same proto-language, but this proto-language equally differs from both languages in focus – I labelled the proto-language as *Proto-Sogdic which I find appropriateн forн explanationн ofн developmentн ofн bothн Sogdianн andн Yaghn bīн ratherн thanн *Proto-Sogdian as there has to be suggested a an intermediate development stage between *Proto-Sogdic and (literary) Sogdian. Bothн Yaghn bīн andн Sogdianн showн manyн differencesōн someн ofн themн areн causedн byн approximately thousand years of discontinuity of development of both language as Sogdian has been replaced by Persian in the 10th and 11th both languages were gradually influenced by Persian, strong influence of Persian is visible mainly inнYaghn bīынAsнbothнlanguagesнdiffer according to their attested forms, it can be said that from diachronic point of view they are two similar dialects/languages, both comparable in historical development as Sogdic dialects within the North Eastern Iranian language group. Abbreviations: Ave. Bactr. Baj. Bart. Ide. Ir. Ishk. Kh tы Khūfы Khwārы Munj. rmы Oss. Pahl. Parāchы Parth. Pasht. Pers. Rāshrvы R shы R shānī Sangl. Sanglēchī Sarīqы Sarīq lī Scyth. Scythian Shugh. Shughnī Shugh-R shы Shughnī-R shānīнgroup Sogd. Sogdian Buddhist Sogdian Christian Sogdian Manichaean Sogdian Sogdian in Sogdian script – Mount Mugh documents Sogdian in Sogdian script Tjk. Tajik Tumshuq. Tumshuqese Ved. Vedic, Old Indic Wakh. Wakhī Yagh. Yaghn bī Yazgh. Yazghulāmī Yidgh. Yidghā Avestan Bactrian (in Greco-Bactrian alphabet) Bactrian in Manichaean script Bajūwī Bartangī (Proto-)Indo-European (Proto-)Iranian Ishkāshmī Kh tanese Khūfī Khwārezmian Munjī rmuṛī Ossetic MiddleнPersianōнPahlavī Parāchī Parthian Pasht (Classical) Persian Rāshārvī ·13· Bibliography: ABAEV 1949: В И ч :н н н н ын н– ōню949ы ALEMANY I VILAMAJÓ 1999: ыugustí AlemanyнiнVilamajó:нElsн«Cantsнarimaspeus»нd ArísteasнdeнProconnèsнiнlaнcaigudaнdelsнZhouн occidentals. Faventia 21/2, 1999, p. 45-55. Aʿ AM н– WINDFUHR 1972: Cheragh Ali Azami – Gernot Windfuhr: A Dictionary of Sangesari, With a Grammatical Outline. Tehrānнх: Franklin Book Company), 1972. .۱۵۳۱ ،)‫ موس ه انت ا ات فرانکلین‬:( ‫ تهران‬.‫ با مقدمەای ا دستو ان بان‬،‫ واژەنامٔ سنگ ری‬:‫ ویندفوهر‬.‫چرا علی ا ظمی – گرنت ل‬ BIELMEIER 1989: Roland Bielmeier:н Yaghn bīын In:н üd g“r Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 480-488. BOGOLYUBOV 1956: ч :н н х цн ын н н ын н н н ё н н н н ын ню956ы BOGOLYUBOV 1966: ч :н н ынIn:нВи Ви н(ed.): н н ын н :н н ын нх:н ), 1966, p. 342-361. ÈDEL MAN 1986: Д И ч Э :н н н н ын ын нх:н ), 1986. GMS: Ilya Gershevitch: A Grammar of Manichaean Sogdian. Oxford, 1954. GRYUNBERG – ÈDEL MAN 1987: А ч н– Д И ч Э :н н ын In:н В н хedыц:н н н ын н н IIын – н ын нх:н ), 1987, p. 6-154. HINGE 2006: George Hinge: Herodot zur skythischen Sprache. In: Glotta 81 2005[2006], p. 86-115. KHROMOV 1987: А ч :н н ынIn:нВ нхedыц:н н н ын н нIIын– н ын нх:н ), 1987, p. 644-701. KLIMCHITSKIY 1935: и Ии :н н ын In:н н н н н н·нVIын ōню935ыню5-25. KLIMCHITSKIY 1940: и Ии :н н н н н н ё ын In:н н н н– н н ōн ынIXн– 1938 – н– н– ынAkademijaji Fanho ·14· SSSR: Asarhoji ʙ‘₎‘j Toç k sto ж ç ld IX – Tarix – zaʙon – adaʙijotын н– н (: И А ), 1940. 104-117. LIVSHITS 1962: В А ч :н н ынIn:н и и н– и Аи н– и и н – и Аи н хedsыц:н н н н н н Iын н (: И А ), 1962, 138-140. LIVSHITS – KHROMOV 1981: В А ч н– А ч :н н ын In:н В н хedыц:н н н ын н н Iын нх:н ), 1981, p. 347-514. MACKENZIE 1988: David Neil MacKenzie: Khwarezmian and Avestan. In: East and West, Vol. 38, No. 1/4 (: Instituto It‘l ‘ o p“r lʼы”r c‘ “ lʼOr “ t“), 1988, p. 81-92. NOVÁK [in print]: Ľu’o ír Novák:н Yaghnobi:н anн Exampleн ofнaн Languageн inн Contact. In:н Chatreššarн я0ююын Praha (: F lo₎o” cká ”‘kult‘ U v“r₎ t₍ K‘rlov₍ v Pr‘₎“), 2011, p. XX-YY. PAYNE 1989: John Payne:н Pāmirн Languagesын In:н üd g“r Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 418-444. PROVASI 2009: Elio Provasi: Versification in Sogdian. In: Werner Sundermann – Almut Hintze – Fr‘ ço s de Blois (eds.): Exegisti monumenta. Festschrift in honour of Nicholas Sims-Williams. Wiesbaden (: Harrassowitz), 2009, p. 347-368. RASTORGUEVA 1966: В :н н ын В . In: Ви Ви н (ed.): н н ын н :н н ын нх:н ), 1966, p. 194-211. SIMS-WILLIAMS 1982: Nicolas Sims-Williams: The double system of nominal inflection in Sogdian. Transactions of the Philological Society 1982, Oxford, 1982, s. 67-76. SIMS-WILLIAMS 1984: Nicolas Sims-Williams:н Theн Sogdianн “Rhytmicн Law”ын In:н Wojciech Skalmowski, Alois van Tongerloo (eds.): Middle Iranian Studies: Proceedings of the International Symposium organized by the Katholieke Universiteit Leuven from the 17th to the 20th of May 1982. Leuven (: Katholieke Universiteit te Leuven), 1984, p. 203-215. SIMS-WILLIAMS 1989a: Nicolas Sims-Williams: Eastern Middle Iranian. In: üd g“r Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 165-172. SIMS-WILLIAMS 1989b: Nicolas Sims-Williams: Sogdian. In: üd g“r Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 174-192. SIMS-WILLIAMS 1996a: Nicolas Sims-Williams:н Theн Sogdianн manuscriptsн inн Brāhmīн scriptн asн evidenceн forн Sogdianн ·15· phonology. In: Ronald Eric Emmerick et ali. хedsыц:нTurfanōнKhotanнundнDunhuang:нVorträgeнderн Tagungн “Annemarieн vonн Gabainн undн dieн Turfanforschung”н veranstaltetн vonн derн Berlin-Brandenburgischen Akademie der Wissenschaften in Berlin (9.-12. 12, 1994). Berlin, 1996, p. 307-315. SIMS-WILLIAMS 1996b [online]: Nicolas Sims-Williams: Eastern Iranian Languages. In: sā Yārshāṭerн(ed.): EncyclopædiaнIranicaн [online edition]. Costa Mesa, California. URL: <http://www.iranicaonline.org/articles/eastern-iranian-languages> [quot. 05. 12. 2012, 12:34] SKJÆRVØ 1989: Prods Oktor Skjærvø:н Modernн Easternн Iranianын In:н üd g“r Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 370-383. SOKOLOVA 1967: В :н н н н н н н н ын нх:н ), 1967. SOKOLOVA 1973: В :н н н н н н н н ын ōню973ы VINOGRADOVA 2000: я :н н ын In:н н ын н н IIIын – н ын нх:нИ Д И ), 2000, p. 290-310. ·16· 

Univerzita Karlova v Praze Filozofická‎fakulta DISERTAČNÍ PRÁCE 2013 Ľubomír‎Novák Univerzita Karlova v Praze Filozofická‎fakulta Ústav‎obecné‎lingvistiky Obecná‎jazykověda – Indoevropská‎srovnávací‎jazykověda Ľubomír N o v á‎k Problem of Archaism and Innovation in the Eastern Iranian Languages K problému archaismu a inovace ve východoíránských jazycích Disertační‎práce vedoucí‎práce‎– Doc. PhDr.‎Petr‎Vavroušek, CSc. 2013 Prohlašuji,‎ že‎ jsem‎ disertační‎ práci‎ napsal‎ samostatně‎ s využitím pouze‎ uvedených‎ a‎ řádně‎ citovaných‎ pramenů‎ a‎ literatury‎ a‎ že‎ práce‎ nebyla‎ využita‎ v rámci‎ jiného‎ vysokoškolského‎ studia‎či‎k‎získání‎jiného‎nebo‎stejného‎titulu. p›nʾm y ʾm ʾnk (Mugh Letter 1.I) I would like to thank to my family and to my friends for support. My thanks also belong to κʥʲщлἱuϑʧʣлπϑʦüʲʣʲováлʤoʲлpʲooʤʲʣadʧnʥлandлto κʥʲщлηanлBʧčovʳ₎ýчлἵʦщαщ for proofreading and for valuable comments on phonology and other issues. My special thanks belong to Reiner Lipp, Ph.D. who read the text and corrected many of my mistakes. My gratitude belongs to αoϑщлἵʦαʲщлἵʣtʲлσavʲoušʣ₎члCπϑ. who inspired me to write this thesis. My gratitude also belongs to the YagʦnōϐХлpʣopὕʣлуʣʳpʣϑʧaὕὕyлtoлἵʲoʤʣʳʳoʲлπayʤʧddХnлκХʲzōzōdaл and his family) – thank to them I had the opportunity to learn their extraordinary language. Ľuϐomíʲлλová₎ Prague, 2nd April 2013 PROBLEM OF ARCHAISM AND I N N O V A T I O N I N T H E EASTERN IRANIAN LANGUAGES K PROBLÉMU ARCHAISMU A INOVACE VE VÝCHODOÍRÁNSKÝCH JAZYCÍCH Table of contents Table of contents ..................................................................................................................... i Abbreviations ...........................................................................................................................v Languages ............................................................................................................................v Maps and figures .................................................................................................................. viii Abstract ................................................................................................................................. xii Abstrakt ................................................................................................................................ xii I. Introduction ............................................................................................................................. 1 I.1. An outline of history and classification of the Eastern Iranian languages .......................... 2 I.1.1. Overview of the Eastern Iranian languages ...................................................................3 I.1.1.1. *Proto-Indo-Iranian and *Proto-Iranian periods .................................................. 4 I.1.1.2. Old Iranian period ................................................................................................ 6 I.1.1.2.1. Avestan .......................................................................................................... 6 I.1.1.2.2. Scythian and Sauromatian dialects, Cimmerian ............................................. 9 I.1.1.3. Middle Iranian period ...........................................................................................11 I.1.1.3.1. Sogdian ...........................................................................................................11 уʣʸϑuʲʳʧonльфлπoʥdʧanлdʧaὕʣϑtʳлoʤлBu₎ʦāʲāчлςʳtʲōʳʦanaлand Zhetisu ....................... 15 I.1.1.3.2. Sarmatian, Alanic and Jassic.......................................................................... 16 ζщьщьщющющлἰʦʷāʲʣzmʧan ................................................................................................. 17 I.1.1.3.4. Bactrian........................................................................................................ 20 ζщьщьщющ5щлἰʦōtanʣʳʣлand Tumshuqese, Saka dialects ................................................... 21 I.1.1.4. New Iranian period .............................................................................................. 23 I.1.1.4.a. North Eastern Iranian ...................................................................................... 24 ζщьщьщящьщлḲaʥʦnōϐХ ...................................................................................................... 24 уʣʸϑuʲʳʧonлэфлḲaʥʦnōϐХлdʧaὕʣϑtʳ ............................................................................. 26 (excursion 3) Sogdo-ḲaʥʦnōϐХлʳuϐʳtʲatʣлʧnлtʦʣ ḳaʲaʤʳʦān-Tajik dialects ...............27 I.1.1.4.2. Ossetic ......................................................................................................... 29 I.л.л.ᵙ.’. Th“ PāmД› l‘ngu‘g“s ......................................................................................... 32 ζщьщьщящющлτan₍Х ............................................................................................................ 35 i I.1.1.4.4. YazghuὕāmХ ...................................................................................................36 ζщьщьщящ5щлρʦʣлπʦuʥʦnХ-οōʳʦānХлʥʲoup........................................................................ 39 ζщьщьщящ6щлπaʲʥʦuὕāmХ .................................................................................................. 44 ζщьщьщящ7щлκun₍Хлand Ḳʧdʥʦā ......................................................................................... 45 I.1.ьщящřщлζʳʦ₎āʳʦmХчлπanʥὕēϑʦХлand ḳēϐā₎Х ..................................................................47 ζщьщьщящŚщлτa₎ʦХ...........................................................................................................50 I.1.1.4.c. South and Southeast Eastern Iranian................................................................... 53 ζщьщьщящьыщлἵaʳʦtōлandлWazХʲХ ....................................................................................... 53 I.1.1.4.11. τa ʣtʳХ ........................................................................................................ 55 ζщьщьщящьэщлἵaʲāϑʦХ ........................................................................................................ 56 ζщьщьщящьющлŌʲmuṛХ ........................................................................................................ 58 I.1.1.5. Other Eastern Iranian languages .......................................................................... 59 I.1.2. Classification of the (Eastern) Iranian languages ....................................................... 60 II. Archaism and innovation in Sogdian and ḲaʥʦnōϐХ ............................................................. 66 II.1. Historical phonology .......................................................................................................68 (excursion 4) Sogdian orthographical systems ........................................................68 II.1.1. Stress ........................................................................................................................ 75 II.1.1.1. Stress I ................................................................................................................. 77 II.1.1.2. Stress II ............................................................................................................... 78 II.1.1.3. Stress III and the Sogdian Rhythmic Law .............................................................80 II.1.1.4. Stress IV .............................................................................................................. 81 II.1.2. Vowels and diphthongs ............................................................................................ 82 II.1.2.1. *a, ŋą ................................................................................................................. 90 II.1.2.2. ŋā, ŋā ................................................................................................................ 92 II.1.2.3. *i ........................................................................................................................ 93 II.1.2.4. *Д ...................................................................................................................... 94 II.1.2.5. *u ...................................................................................................................... 94 II.1.2.6. * ...................................................................................................................... 95 II.1.2.7. *› ....................................................................................................................... 95 II.1.2.8. * ..................................................................................................................... 96 II.1.2.9. * u .....................................................................................................................97 II.1.3. Consonants ...............................................................................................................97 II.1.3.1. *p ...................................................................................................................... 103 II.1.3.2. *t ...................................................................................................................... 103 II.1.3.3. *k ..................................................................................................................... 104 II.1.3.4. *č...................................................................................................................... 108 II.1.3.5. *b...................................................................................................................... 108 II.1.3.6. *d ..................................................................................................................... 108 (excursion 5) Lambda Sogdica?.............................................................................. 108 II.1.3.7. *g....................................................................................................................... 113 ii II.1.3.8. *ǰ ....................................................................................................................... 113 II.1.3.9. *f ....................................................................................................................... 113 II.1.3.10. *ϑ .................................................................................................................... 114 II.1.3.11. *x .....................................................................................................................116 II.1.3.12. *x , *hu ............................................................................................................116 II.1.3.13. * ...................................................................................................................... 117 II.1.3.14. *ž ..................................................................................................................... 117 II.1.3.15. *m .................................................................................................................... 117 II.1.3.16. *n .................................................................................................................... 118 II.1.3.17. *r ..................................................................................................................... 118 II.1.3.18. *l (?) ................................................................................................................. 119 II.1.3.19. *s...................................................................................................................... 119 II.1.3.20. *h .................................................................................................................... 119 II.1.3.21. *z .................................................................................................................... 120 II.1.3.22. *ʦ .................................................................................................................... 121 II.1.3.23. *ʣ .................................................................................................................. 122 II.1.3.24. * .................................................................................................................... 123 II.1.3.25. *u ................................................................................................................... 123 II.1.3.26. *ʜ................................................................................................................... 124 II.1.4. Syncope and reduction ........................................................................................... 124 II.1.5. Prothesis and epenthesis ......................................................................................... 125 II.1.6. Assimilation and dissimilation ............................................................................... 127 II.1.7. Metathesis .............................................................................................................. 128 II.1.8. Analogy .................................................................................................................. 129 II.1.9. Syllabic structure .................................................................................................... 130 II.2. Historical grammar ....................................................................................................... 132 II.2.1. Nominal inflection ................................................................................................. 132 II.2.2. Pronominal inflection ............................................................................................. 141 II.2.3. Numeral inflection ................................................................................................. 146 II.2.4. Verbal inflection .................................................................................................... 147 (excursion 6) Ergative ........................................................................................... 154 II.2.5. Adpositions ............................................................................................................. 158 II.2.6. Conjunctions ......................................................................................................... 159 III. Lexicon .............................................................................................................................. 160 III.1. Pronouns ..................................................................................................................... 164 III.2. Numerals ..................................................................................................................... 168 III.3. Adjectives (i)................................................................................................................ 170 III.4. People ......................................................................................................................... 172 III.5. Animals ....................................................................................................................... 176 III.6. Plants .......................................................................................................................... 179 iii III.7. Body parts ................................................................................................................... 182 III.8. Verbs ............................................................................................................................ 188 III.9. Celestial objects ........................................................................................................... 196 III.10. Nature (i) ................................................................................................................... 197 III.11. Weather ...................................................................................................................... 199 III.12. Fire ........................................................................................................................... 200 III.13. Settlement .................................................................................................................. 201 III.14. Tools ......................................................................................................................... 203 III.15. Nature (ii) ................................................................................................................. 204 III.16. Colours ..................................................................................................................... 204 III.17. Time .......................................................................................................................... 205 III.18. Adjectives (ii) .............................................................................................................207 III.19. Adpositions ................................................................................................................. 211 III.20. Conjunctions ............................................................................................................. 212 III.21. Name ......................................................................................................................... 213 IV. Conclusion ......................................................................................................................... 215 V. Bibliography ....................................................................................................................... 222 iv Abbreviations abl. acc. arch. C coll. colloq. dat. dial. dir. du. encl. gen. GMS imper. impf. inf. instr. itr. lit. loc. nom. ablative accusative archaic any consonant collective colloquial dative dialect, dialectal direct case dual enclitic genitive GERSHEVITCH 1954 imperative imperfect infinitive instrumental intransitive verb literary locative nominative obl. occ. opt. pers. pf. pl. poet. postp. prep. pres. pret. pron. sbjn. sg. tr. V voc. oblique case occasionally optative person perfect plural poetical postposition preposition present preterite pronoun subjunctive singular transitive verb any vowel vocative mark of an incomplete word, e.g. ⃝ th“ s₍m’ol s us“d to om t a part of compound word Afghan ἵʣʲʳʧanчлαaʲХ Arabic Avestan Azərbayjanian Bactrian (in Greco-Bactrian alphabet) Bactrian in Manichaean script Ba₍ūʷХ BaὕōchХ BaʲtanʥХ BeṅʥāὕХ Breton ἰἱρ лdʧaὕʣϑtʳлуʧщʣщлἰʣʲnʣʷʣʥъ Cornouaillais – Leoneg/ŭéon‘›d – Tregerieg/T›égo››o s) Gwenedeg/Vannetais dialect BukhāʲanлAʲaϐʧϑ Byzantine Greek C., Class. Chaghat. Chin. Cimm. Corn. Cr.Goth. Cr.Tatar. CSl. Dard. Elam. Eng. γāʲʳщ Fr. Gael. Georg. Ger. Gmc. Goth. Classical Chaghatāy Chinese Cimmerian Cornish Crimean Gothic Crimean Tatar Church Slavic Dardic Elamite English Contemporary Persian of Iran French Scottish Gaelic Georgian German Proto-Germanic Gothic ⃝ Languages AfghP. Ar. Ave. Azərb. Bactr. Baj. Baὕōch. Bart. Beṅʥāὕ. Bret. BukhAr. ByzGre. v Gre. εazāʲщ Hind. Hitt. Hung. IAr. Ide. IIr. Ir. Irl. Ishk. Jass. ἰāϐщ ἰāmvʧʲщ Khōtщ Khūʤщ Khʷāʲщ Kurd. Lat. Latv. LHChin. Lith. LKhōtщ κāzandщ MChin. MGre. Mid. Mod. Munj. λūʲщ O. OChin. OCS. OHG. OIcel. OIrl. OKhōtщ OPers. Ōʲmщ Old Persian ŌʲmuṛХ Baʲa₎Х-Barak dialect ἰānХʥuʲāmлdʧaὕʣϑt OScand. Old Scandinavian Oss. Ossetic Digoron dialect Iron dialect Ott. Ottoman Turkish OUygh. Old Uyghur Pahl. κʧddὕʣлἵʣʲʳʧanчлἵaʦὕavХ Middle Persian in Manichaean script ἵaʲāch. ἵaʲāchХ Parth. Parthian Pasht. Pashtō Pers. (Classical) Persian Prkt. Prakrit Pruss. Prussian ἶāʲaqaὕpщ ἶāʲaqaὕpāq Qashq. QashqāyХ οāshrv. οāshāʲvХ Roman. Romanian οōsh. οōshānХ Rus. Russian Sangl. πanʥὕēchХ Sargh. SarghuὕāmХ Sarghul. SarghuὕāmХ πaʲХqщ πaʲХqōὕХ Sarm. Sauromatian, Sarmatian Scyth. Scythian Shākhd. ShākhdaʲaХ Shugh. ShughnХ Shugh.-οōsh. ShughnХ-οōshānХлὕanʥuaʥʣ group Skt. Sanskrit Sogd. Sogdian Sogdian in Sogdian script – Anϑʧʣntлἱʣttʣʲʳ Buddhist Sogdian BʲāʦmХлπoʥdʧan Christian Sogdian Sogdian in Sogdian script – (Attic) Greek Doric Ionic Hellenistic Koine εazāʲaуʥХф εʧndХ Hittite Hungarian Indo-Aryan (Proto-)Indo-European Indo-Iranian (Proto-)Iranian Modern Irish Ish₎āshmХлуοanХф Jassic γāʲʳХ-ἰāϐuὕХ ἰāmvʧʲʧ Khōtanʣʳʣ KhūʤХ Khʷāʲʣzmʧan Kurdish Latin Latvian Late Han Chinese Lithuanian Late Khōtanʣʳʣ κāzandʣʲānХ Middle Chinese Modern Greek Middle Modern κun₍Х Central dialect Northern (Lower) dialect Southern (Upper) dialect λūʲʧʳtānХ Old Old Chinese Old Church Slavic Old High German Old Icelandic Old Irish Old Khōtanʣʳʣ vi ρāὕysh. TBukh. Tehr. TFalgh. Thrac. Tjk. TMast. Tokh. Chilhujra documents Sogdian in Sogdian script – ʧʳōʲa₎лdoϑumʣntʳ Manichaean Sogdian Sogdian in Sogdian script – Mount Mugh documents Sogdian in Sogdian script Sogdian in Sogdian script – Zhetisu documents ρāὕyshХ Tajik dialect of Bukhāʲā Teheran dialect of Modern Persian Tajik dialect of Falghar Thracian Tajik Tajik dialect of Mastchōʦ Tokharian A лdʧaὕʣϑt B лdʧaὕʣϑtчлἰūchean Tr. ρüщ Tumsh. Turkm. TVanj. TVarz. TYagh. Urd. Ustr. Ved. Wakh. Wanj. Yagh. Yazgh. Yidgh. ḳēϐщ vii Turkish Turkic Tumshuqese Turkmen Tajik dialect of Vanj Tajik dʧaὕʣϑtлoʤлσaʲzōϐ Tajik dialect of Yaghnōϐ ςʲdū ςʳtʲōshanian Vedic, Old Indic WakhХ τan₍Х YaghnōϐХ Central (Transitional) dialect Eastern dialect Western dialect YazghuὕāmХ Yidghā ḳēϐā₎Х Maps and figures Map 1 Historical territories of Central Asia (DE LA VAζππζÈοβл2005, 14, Map 1) viii Map 2 Distribution of Modern Eastern Iranian Languages (1 Pashtō, 2 Wa etsХ, 3 κun₍Х-Yidghā, 4 Ish₎āshmХ-πanʥὕēchХ, 5 WakhХ, 6 ShughnХ, 7 οōshānХчлBaʲtanʥХчл οāshāʲvХ, řл πaʲХqōὕХ, 9 YazghuὕāmХ, 10 YaghnōϐХ, 11 Ossetic, 12 ἵaʲāchХ, ьюлŌʲmuṛХ), URL: <http://www.ljplus.ru/img4/b/a/bahmanjon/Iranian_lang_rus.jpg> [quot. 02. 01. 2013 21:52], edited by Mgr. Veronika Ḥ k“ ová. Map 3 Location of the YaghnōϐлvaὕὕʣyлʧnлTajikistan (yellow) and an approximate distribution of ZarafshānлTajik dialects (grey). ix Map 4 Iranian ὕanʥuaʥʣʳл oʤл ἵāmХʲл andл ad₍aϑʣntл ʲʣʥʧonʳщл ь κun₍Хчл э Yidghā;л юa Ish₎āshmХ, 3b πanʥὕēchХчл юc ḳēϐā₎Х;л 4 WakhХ;л 5 YazghuὕāmХ;л 6 τan₍Х;л 7a ShughnХчл 7b ShākhdaʲaХчл ř Ba₍ūХчл Ś οōshānХчл ьы KhūʤХчл ьь BaʲtanʥХчл ьэ οāšāʲvХчл 13 πaʲХqōὕХл уϐyл Yuriy Borisovich KORYAKOV, Russian Academy of Sciences 2001), URL: <http://lingvarium.org/maps/asia/pamir-lgs.gif> [quot. 02. 01. 2013], “d t“d ’₍ Ḥg›. V“›on k‘ Ḥ k“ ová. x Figure 1 Sogdian alphabet of the Ancient Letters. Figure 2 Sogdian cursive script. Figure 3 Manichaean Sogdian alphabet. Figure 4 Syriac Sogdian alphabet. xi Abstract The presented dissertation aims to bring new information concerning the classification of the Eastern Iranian languages. Instead of commonly accepted two branches of Eastern Iranian (Northern and Southern) it seems that there can be classified at least five branches of Eastern Iranian languages, moreover, Avestan can form its own branch, which possibly may include also Khʷāʲʣzmʧanщлρʦʣлmaʧnлʧʳʳuʣлoʤлtʦʣлpʲʣʳʣntʣdлtʦʣʳʧʳлʷaʳлtoлʳʦoʷлaʲϑʦaʧʳmʳлandлʧnnovatʧonʳлoʤл the language group in focus. Such task is an issue for numerous studies so the main attention was paid to historical development of Sogdian and YaghnōϐХл – two closely related Eastern Iranian languages. Linguistic proximity of Sogdian and YaghnōϐХлʦaʳлϐʣʣnлoϐʳʣʲvʣdлʳʦoʲtὕyлaʤtʣʲлdʧʳϑovʣʲʧʣʳлoʤл the first Sogdian documents in Chinese Turkestan on the beginning of the 20th century, for a long time it has been supposed that YaghnōϐХлʧʳлaлmodʣʲnлdʣʳϑʣndʣntлoʤлπoʥdʧanщлByлanaὕyʳʧʳлoʤл phonology, grammar and vocabulary of both languages I tried to find clues that may answer this question. From diachronic view there is no much difference between Sogdian and YaghnōϐХчл indivʧduaὕл ϑʦanʥʣʳл mayл ϐʣл ʧntʣʲpʲʣtʣdл aʳл dʧaὕʣϑtaὕ чл ϐutл tʦʣʲʣл ʧʳл onʣл pʦʣnomʣnonл tʦatл influenced different development of both languages – operation of the so-called Rhythmic Law in Sogdian, but not in YaghnōϐХщл γoʲл tʦʧʳл ʲʣaʳonл ζл ʦavʣл reconstructed an older common ancestor of both languages – *Proto-Sogdic, i.e. proto-language before the operation of the Rhythmic Law. Abstrakt ἵřʣd₎ὕádanáлdʧʳʣʲtaϑʣ ʳʧл₎ὕadʣлzaлϑíὕлpřʧnéʳtлnovéлʧnʤoʲmaϑʣлoʦὕʣdnЕл₎ὕaʳʧfi₎aϑʣлvýϑʦodoíʲánʳ₎ýϑʦл ₍azy₎ůщлκíʳtoлoϐʣϑnЕлa₎ϑʣptovanýϑʦлdvouлvýϑʦodoíʲánʳ₎ýϑʦлvЕtvíлуʳʣvʣʲníлaл₍ʧžníфл sʣлzdáчлžʣлby ϐyὕoлvʦodnЕ₍šíлtytoл₍azy₎yлʲozdЕὕʧtлmʧnʧmáὕnЕлnaлpЕtлʳ₎upʧnщлκožnouлšʣʳtouлʳ₎upʧnuлpa₎лmůžʣл tvořʧtлavʣʳtštʧnaчлʳpoὕuлʳ níлpřípadnЕлʧлϑʦóʲʣzmštʧnaщлεὕavnímлtématʣmлpřʣd₎ὕádanéлpʲáϑʣлvša₎л ϐyὕлzámЕʲлʳὕʣdovatлaʲϑʦaʧʳmyлaлʧnovaϑʣлvʣлvýϑʦodoíʲánʳ₎ýϑʦл₍azyϑíϑʦщлαů₎ὕadnéлzpʲaϑováníлtétoл pʲoϐὕʣmatʧ₎yл ϐyл ʳʧл zaʳὕoužʧὕoл řaduл odϐoʲnýϑʦл ʳtudʧíчл pʲotoл ϐyὕoл danéл témaл zúžʣnoл zʣ₍ménaл naл ʳὕʣdováníл ʦʧʳtoʲʧϑ₎éʦoл vývo₍ʣл ʳoʥdštʧnyл aл ₍aʥʦnóϐštʧnyл – dvouл ϐὕízϑʣл příϐuznýϑʦл výϑʦodoíʲánʳ₎ýϑʦл₍azy₎ůщ σzá₍ʣmnáл ϐὕíz₎oʳtл ʳoʥdʳ₎éʦoл aл ₍aʥʦnóϐʳ₎éʦoл ₍azy₎aл ϐyὕaл zpozoʲovánaл ₎ʲátϑʣл poл oϐ₍ʣvʣníл pʲvníϑʦл ʳoʥdʳ₎ýϑʦл tʣʸtůл z Čínʳ₎éʦoл ρuʲ₎ʣʳtánuл začát₎ʣmл эыщл ʳtoὕʣtíщл ηaʥʦnóϐštʧnaл ϐyὕaл do₎onϑʣл poл dὕouʦouл doϐuл považovánaл zaл modʣʲníʦoл po₎ʲačovatʣὕʣл ʳoʥdštʧnyщл οozϐoʲʣmл ʤonoὕoʥʧʣчлʥʲamatʧ₎yлʧлὕʣʸʧ₎aлoϐouл₍azy₎ůл₍ʳʣmлʳʣлpo₎uʳʧὕлna₍ítлodpovЕďлnaлotáz₎uлvzá₍ʣmnéʦoл vztaʦuлtЕϑʦtoл₍azy₎ůщлḳ diaϑʦʲonníʦoлpoʦὕʣduлmůžʣmʣлpovažovatлʲozdíὕyлmʣzʧлoϐЕmaл₍azy₎yл₍ʣnл ₍a₎oл nářʣční odὕʧšnoʳtʧчл ₍ʣл zdʣл vša₎л ₍ʣdʣnл ₍ʣvчл ₎tʣʲýл způʳoϐʧὕл ʲozdíὕnýл vývo₍л v oϐouл ₍azyϑíϑʦл – půʳoϐʣníлtzvщл›₍tm ckého ₎ákon‘ v ʳoʥdštʧnЕчл₎ʣл₎tʣʲémuлvša₎лnʣdošὕoлv ₍aʥʦnóϐštʧnЕщлḳ tohoto důvoduл₍ʳʣmл ʲʣ₎onʳtʲuovaὕ лʳtaʲšíʦoлʳpoὕʣčnéʦoлpřʣdϑʦůdϑʣлoϐouл₍azy₎ůл– хpʲotoʳoʥdʧčtʧnuчлt₍щл prajazyk z doϐyлpřʣdлpůʳoϐʣnímл›₍tm ckého ₎ákon‘. xii I. Introduction The Eastern Iranian languages form an independent group within the Iranian branch of the Indo-European languages. The presented thesis aims to present an outline of development of the Eastern Iranian languages – as languages develop, they usually start to differ from its relatives by development of various innovations and/or by preservation of archaisms. The spread of innovations and preservation of archaisms may vary in individual languages or dialects and study of sets of common innovations and/or archaisms may characterize grouping of languages of a given branch. To see the Eastern Iranian archaisms and innovations I have decided to focus on three fields of study – 1) an outline of the Eastern Iranian languages, 2) historical grammar of Sogdian and YaghnōϐХлandлюф lexical study. The first part will be dedicated to the description of attested Eastern Iranian languages and dialects – each language (or a subgroup) will be briefly described with focus on common data about the individual language(s), with an overview of main phonetic changes and grammar outline. For the overviews I will mark only some archaic and innovative features of the individual languages as for each language can be written separate book on its historical grammar and phonology. I would also like to (re)examine commonly accepted grouping of the Eastern Iranian languages into the Northern and Southern branches as it seems to me that this grouping needs a new revision. The second part will present comparation of development of Sogdian and YaghnōϐХл– i.e. two languages that are considered closely related by many scholars (e.g. BOGOLYUBOV 1956; KLIMCHITSKIY 1935; SἰηÆοσØ 1989a, 375-376), but none of them has ever presented thorough study of their differences – YaghnōϐХлʷaʳлʧnлϑommonл₍uʳtлϑonʳʧdʣʲʣdлaʳлaлdʧaὕʣϑtлquʧtʣлdʧfferent from literary Sogdian. By comparation of phonology and morphology of both languages I would like to show main differences between them and if possible I would like to try to define interrelationship of Sogdian and YaghnōϐХщл ρʦʣл ϑompaʲative study of YaghnōϐХл andл πoʥdʧanл has been taken intentionally – as both languages are comparable from diachronic point of view, their comparisonлmayлanʳʷʣʲлmoʲʣлquʣʳtʧonʳлtʦanл₍uʳtлtʦʣʧʲл dʧaὕʣϑtaὕ лʲʣὕatʧonʳʦʧpщлεʧʳtoʲʧϑaὕл development of Sogdian and YaghnōϐХл ʷʧὕὕл ϐʣл ϑompared with the other Eastern Iranian languages wʧtʦлʤoϑuʳлonлtʦʣлἵāmХʲлʥʲoupщлρʦʣлἵāmХʲлὕanʥuaʥʣʳлʷʧὕὕлϐʣлuʳʣdлaʳлaлϑompaʲative material for two reasons – 1) ʧtл ʳʣʣmʳл tʦatл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл andл ḲaghnōϐХл ʳʦaʲʣл ʳomʣл historically non-documented areal contacts and 2) for I have collected many material on the ἵāmХʲлὕanʥuaʥʣʳлʳoлζлϑanлϐʣttʣʲлuʳʣлtʦʧʳлmatʣʲʧaὕлin my study. I have not compared development of Sogdian and YaghnōϐХлmuϑʦлʷʧtʦлʲʣὕatʣdлἴʳʳʣtʧϑлϐʣϑauʳʣлoʤлa probable ʣaʲὕyлʳpὕʧtлoʤл ἵontʧϑл πϑytʦʧan л andл Cʣntʲaὕл Aʳʧanл πϑytʦʧan л dʧaὕʣϑtʳл oʤл λoʲtʦл βaʳtʣʲnл Iranian branch and also because of long-standing intensive contact of Ossetic with the Caucasian languages, which caused different development of this branch of Scythian. Materials on other Eastern Iranian languages such as Pashtōл oʲл Saka dialects were also available to me, but I focused mainly on ʳtudyл oʤл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл – there can be supposed a common development also in the Middle Iranian period. Example can be seen in many common features shared in Bactrian (as ·1· Bactrian can be considered as a relative of *Proto-ἵāmХʲлὕanʥuaʥʣʳфлandлπoʥdʧan on one hand andлʳomʣлʤʣatuʲʣʳлʳʦaʲʣdлϐyлBaϑtʲʧanлʷʧtʦлtʦʣлἵāmХʲлʥʲoupлуmaʧnὕyлʷʧtʦлḲʧdghāлandлκun₍Хф. The third part will present a study of Sogdian and YaghnōϐХл ὕʣʸʧϑonщл ζл ʦavʣл oʲʧʥʧnaὕὕyл intended to compare YaghnōϐХл andл πoʥdʧanл ὕʣʸʧϑonл aϑϑoʲdʧnʥл toл tʦʣл Swadesh List of 207 words. Later I found πtandaʲd Word List ζtʣmʳ presented in the five-volume Sociolinguistic Survey of Northern Pakistan (see http://www.sil.org/sociolx/pubs/ssnp.asp) by the National Institute of Pakistani Studies, Quaid-i-Azam University and Summer Institute of Linguistics, so I decided to combine both lexical lists to present a more thorough study of basic vocabulary of both languages. In the lexical parts lexical items of both languages will be supplemented by their etymology. The choice of the Swadesh List was not motivated by attempts of glottochronological study of both languages – I just wanted to exploit an accepted list of basic vocabulary, this motivation also led to supplement the Swadesh list by the SIL πtandaʲd Word List ζtʣmʳ . Both lists try to present unbiased choice of basic vocabulary so in this issue I have also to study eventual loans (mainly in case of YaghnōϐХфщ As can be seen from outlines of all three parts, my study of the Eastern Iranian archaisms and innovations aims to present new classification of the Eastern Iranian branch with focus on position of Sogdian and YaghnōϐХлʷʧtʦʧnлtʦʧʳлὕanʥuaʥʣлϐʲanϑʦщ I.1. An outline of history and classification of the Eastern Iranian languages The Iranian languages form a group of genetically related languages and dialects that developed from the Indo-Iranian branch of the Indo-European languages. By use of methods of historical and comparative linguistics we can explain the origin of the Iranian languages as a split of the Indo-Iranian branch of *Proto-Indo-European language. The original *Proto-Indo-Iranian language broke up into the four main branches: IranʧanчлλūʲʧʳtānХлуoʲлἰāfir), Dardic and IndoAryan. Particular prehistoric dialects of Indo-Iranian share with *Proto-Indo-European (and also with many other Indo-European languages) many common features – so called archaisms as well with series of innovations that set them apart from the proto-language. Some of the innovations can be observed in more branches of the Indo-European languages, but are not phenomena proper to the original system of reconstructed *Proto-Indo-European. The Iranian languages are divided into two main branches – Western and Eastern. Their division is based on agreed conventional brake up of two Old Iranian dialects according to their geographical location to the East and West respectively from the deserts of Central Iran (ÈDEL MAN 1986, 3; about the classification of the Iranian languages see Chapter I.1.2. of presented work). Present geographical spread of the Eastern and Western Iranian languages and their speakers has changed due to historical migrations of the Iranian peoples (e.g. Western IranʧanлBaὕōchХлʧs nowadays located in Eastern Iran and Western Pakistan or the Eastern Iranian Ossetic is to be found on the Caucasus), the contemporary location of the Iranian languages is not relevant for their classification. The Iranian languages can be thus considered as an offspring ·2· of the Indo-European proto-language with which they are connected by genetic relationship and a preservation of some (*Proto-)Indo-European archaisms, on the other hand they differ from *Proto-Indo-European by several innovations which define this language family from historical point of view. We are informed about the history of the Old Iranian languages by means of indirect sources. Herodotus for example mentions several Scythian words, in one case he even presents an etymology (HERODOTUS IV, 110; HINGE 2006). He also mentioned that the Sauromatians speak the language of Scythia, but they do not speak it well because the Amazons did not learn properly the Scythian language – Herodotus mentioned that the Amazons married some Scythians and by this the Sauromatian nation came into being (HERODOTUS IV, 117). Herodotus also writes about an older poem, Arimaspea, written by Aristeas of Proconnesus (HERODOTUS IV, 13). It is said that Aristeas described the habits and the language of Scythian Issedonians (Issedones) and Arimaspians (Arimaspi) who dwelled in regions to the North-East of the Pontic or Black Sea (ALEMANY I VζἱAκAηÓ 1999). Unfortunately, Aʲʧʳtʣaʳ лArimaspea has not came down up to these days, it is only mentioned in the Histories of Herodotus and also in by Longinus and in Chiliades (or Book of Histories) by John (Ioannes) Tzetzes (TZETZES, Chil. VII, 686-692). In the Anabasis of Arrian there are mentioned several local tribal and personal names of Central Asia, but we miss any reference to the languages of the region, the only relevant information is that the river Ἰ (SХr Daryā) was called Ὀ in a language of barbarians of Sogdiana (Arrian III, 30.13). ζnлπtʲaϐo ʳ Geography is mentioned, that the northern part of Ἀ (i.e. approximately area of modern Afghanistan, Eastern Iran, Tajikistan, Turkmenistan, and North-western Pakistan) is inhabited by Bactrian and Sogdian peoples who do speak similar languages (STRABO, Geography, XXV, 2:8). The city of Kūrkat in present northern Tajikistan is known from the antiquity – it is spelled either as ῡ ύ or as ῡ ; we can discover more about the local Iranian dialect by the analysis of both Greek names: ῡ ύ is probably a calque of Iranian appellative ŋŬ ›u -kąϑā- city of Cyrus (i.e. Gre. ῡ ύ <ἡ ύ ). What is even more interesting is the form ῡ έ χα α, it can be an attempt to render the local name ŋŬ ›u -kąϑā- (cf. Tjk. and Pers. Ŭ ›kát) 1 ; the Greek name is probably contaminated by another Greek word the farthest (probably by an influence by the name of the city of Alexandria the Farthest – Ἀ Ἐ члpresent Khujand, in the Soviet period known as Leninabad, Tjk. ). City of Ῥ mentioned by Ctesias of Cnidus can be connected ʷʧtʦл ϑʧtyл oʤл οōshānл уοōsh. ⁾ n, Tjk. n) in Tajik BadakhshānлуABAEV 1949, 178). I.1.1. Overview of the Eastern Iranian languages Within following pages I present a short overview of the Eastern Iranian languages and dialects. The description of individual languages is not meant to be absolute; it contains just basic 1 But see Greek popular etymology « , ὂ ύ ·3· » (STRABO, Geography, XI, 11:4). information about the history of each language supplemented with an outline of its grammar and main traits of its development. The aim is to present the most important innovations and archaism of each language in focus. The innovations and archaisms will be presented also in (historical) phonology and also in (historical) morphology. The examples of archaisms and/or innovations will be presented in general; the documentation of changes on examples will be (with a few exceptions) waived. 2 I.1.1.1. *Proto-Indo-Iranian and *Proto-Iranian periods The Iranian languages separated from the older Indo-Iranian branch of the Indo-European languages. The formation of Indo-Iranian proto-language can be characterised by a series of changes that which caused that this branch started to differ from its parent proto-language – the *Proto-Indo-European language. Characteristic phonetic differences include following chain of changes: 1) merger of Ide. *k, *k > *k; *g, *g > *g; ŋgʰ, *g ʰ > ŋgʰ; 2) aspiration of *p+h , *t+h , *k+h > ŋpʰ, ŋtʰ, ŋkʰ; 3) palatalization of *k, *kʰ, *g, ŋgʰ > *c, *cʰ, * , * ʰ before , , ŋ ; 4) Bʲuʥmann ʳ law: ŋŏ > ŋ in an open non-final syllable; 5) merger of , , > . In addition to this chain of changes we can mention a number of others: rhotacism *l > *r; effect of the RUKI rule: *s > ŋ > ŋ following *r, , ŋu, *k( ), *g( )(ʰ), *k, ŋg(ʰ), , ŋ ; satəmization ŋk, ŋg, ŋgʰ > ŋć, , but ŋk, ŋg, ŋgʰ (or ŋć, , ) next to a stop > * , * , * ʰ and later development, previously thought as the thorn problem : ŋtk, ŋdg, ŋdʰgʰ > *tć, *d , *dʰ ʰ > *t , *d , ŋd(ʰ) ʰ; merger of the laryngeals *h , *h , *h > *ʜ and subsequent vocalization *ʜ > *ə > *i in certain positions; vocalization of *n, *m > *ą > *a; *nḧʜ, *mʜ > *n, *m > ŋā ш ŋā and so on. Probably already in the Indo-Iranian period we can also expect the creation of opposition *a ×лŋā хψaл×лȴ”л~ ȵ”ϊлуcf. SIMS-WILLIAMS 1981a, 357, 35731), this change is evident in the New Iranian languages (mainly in the New Eastern Iranian languages we can see change ŋā > (*) 3). consonants labials dentals palatals velars labiovelars pharyngeals glottals stops p b ϐʰ t d dʰ ₎ ʥ ʥʰ k g ʥʰ k g gʰ affricates fricatives ʳ уz) sonorants mu nrl ʧ vowels iХ uū eē h h h oō aā Table 1 Sound system of *Proto-Indo-European. 2 I would like to thank to Reiner Lipp, Ph.D. for his valuable comments on the development of Ide. sound system in *Proto-Indo-Iranian and *Proto-Iranian. 3 This change is spread over a wide area of Central Asia, such as we find it in YaghnōϐХчлPashtōчлShughnХ-οōshānХл ʥʲoupчлκun₍ХлandлḲʧdghāчлζsh₎āshʧmХчлπaʲghuὕāmХлуŠфчлϐutлaὕʳoлʧn the South-West Iranʧanлρā₍Х₎лandлεazāʲaуʥХфчлʧnл Turkic Uzbek or in Central Asian Arabic dialects. ·4· consonants labials stops affricates fricatives p pʰ b ϐʰ sonorants mu vowels iХ uū dentals alveolars t tʰ d dʰ sz ć ćʰ ʰ postalveolars š уžф retroflexes c ϑʰ ʰ palatals velars nr ʧ k ₎ʰ g ʥʰ aā ʜ glottals Table 2 Sound system of *Proto-Indo-Iranian. consonants stops labials pb affricates fricatives sonorants f m (m) u sz šž n (n) r (ʲ) vowels iХ uū dentals alveolars td postalveolars li čǰ ʧ palatals velars labiovelars glottals kg x x h aā (ʜ) Table 3 Sound system of *Common Iranian. The Iranian languages later underwent further changes, that differentiate them from the > *b, *d, *g, * , Indo-Aryan branch: the loss of aspiration of voiced stops ŋ’ʰ, ŋdʰ, ŋgʰ, * ʰ, , , ; *c, * (ʰ) > ŋč, ŋǰ;лtʦʣлϑʦanʥʣлoʤлtʦʣл ʳatəm лandл tʦoʲn лϑonʳonantʳ ŋ(t)ć, > *ʦ, *ʣ; fricativization of ŋpʰ, ŋtʰ, ŋkʰ > *f, *ϑ, *x and also fricativization in front of another consonant *pC, *tC, *kC > *fC, *ϑC, *xC 4; change of ŋtć, > ŋ , ŋž; shift of *s > *h (but not *s in front of a stop) and subsequently ŋhu > *x ; change *T-T > ŋTˢT > *ST; and probably also ŋ›ʜ > ŋ› > *ar (i.e. diphthong (?) хψaʲу”фϊфлandлὕoʳʳлoʤл*ʜ. The vocalic system recognises four short (*a, *i, *u, ŋ›5) and three long (ŋā, ŋД, ŋ ) vowels and three short (ŋ‘ , ŋ‘u, *ar) and two long (ŋā , ŋāu) diphthongs – however, it is possible that diphthongs (and triphthongs) could also consist It is probable, that the fricativization of ŋpʰ, ŋtʰ, ŋkʰ > *f, *ϑ, *x took place in a *post-Proto-Iranian stage of *Common Iranian – there is no such change in WakhХчлBaὕōchХлandлʧnлtʦʣлπa₎aлdʧaὕʣϑtʳщлκaʲtʧnлἰümmʣὕлʳuʥʥʣʳtʳчл that *Proto-Iranian possessed voiceless aspirated stops, so WakhХчл BaὕōchХл andл πa₎aл pʲʣʳʣntл anл aʲϑʦaʧϑл ʳtatʣл (KÜκκβἱ 20. 11. 2012, lecture On historical phonology, typology and reconstruction , Lectures at Charles University, Prague 19-20 November 2012). 5 The syllabic ŋ› is in fact not a vowel but a syllabic core – as it often behaves as vowels it will be for simplification considered as a vowel in this theses. 4 ·5· of consonants *r, *m and *n, after a stressed (?) vowel in front of a stop or fricative; i.e. ŋV›, ŋV›, ŋVm, ŋVn and so on.6 According to the development of the Eastern Iranian languages in the Middle and New Iranian periods it can be assumed that a number of dialectal differences has its source already in the Old Iranian period. Based on a non-existent comparative material we cannot establish a deeper division of these dialects yet, but it seems that by the end of the Old Iranian period the two main Eastern Iranian groups (Northern and Southern) begin to appear. I.1.1.2. Old Iranian period There is only one Eastern Iranian language directly attested from the Old Iranian period – Avestan, but we know also some other languages like Scythian and Sauromatian dialects attested in glosses, mainly onomastic. Classification of Avestan within the Eastern Iranian branch has not been successfully solved yet (cf. ÈDEL MAN 1986, 6-7 with bibliography) I will not attempt to solve the problem of Avestan classification in this thesis and Avestan will be considered as the oldest preserved member of the Eastern Iranian branch. Grammatical system of the *Proto-Indo-Iranian and *Proto-Iranian languages is not much different from the proto-language state. It has preserved a rich inflectional system of nouns, pronouns and verbs, also there are many archaisms compared with the other Indo-European languages, notably the preservation of the verbal injunctive. *Proto-Indo-Aryan grammar is reconstructed mainly on the basis of Vedic Sanskrit, similarly the reconstruction of *ProtoIranian is based mainly on Avestan – proto-languages of both branches are then confronted with the *Proto-Indo-European state. I.1.1.2.1. Avestan Avestan (in older sources also Old Bactrian) is one of the Eastern-Iranian languages. It is closely related to Old Persian, and also comparable with the Indo-Aryan Vedic language, although differences with Vedic go to greater extensions then compared to Old Persian 7 . Unlike Old > k ; ŋć, , I will briefly describe the development in Vedic Sanskrit: *c, ŋcʰ, * , * ʰ > c, cʰ, j, jʰ; * > ; ŋtć, > ś, j, h (but ŋć before a stop > ); ŋtˢt, *dᶻt, ŋdʰᶻt > ŋtˢt, *dᶻd, *dᶻdʰ уBaʲtʦoὕomaʣ ʳлἱaʷфлşл*tt, *dd, ŋddʰ; ŋ₎d(ʰ) > dd(ʰ); *s > word finally of before a pause; ŋsć(ʰ) > ŋ(c)cʰ; *n, *m > ṃ in front of y, r, l, v, ś, , s, h; emergence of retroflex sounds ṭ, ṭʰ, ḍ, ḍʰ, ṇ; nasal assimilation: *n > ṅ in front of k(ʰ), g(ʰ); *n > ñ in front of c(ʰ), j(ʰ); *n > ṇ in front of ṭ(ʰ), ḍ(ʰ); *-n > -m in front of l. Dialect origin are probably sounds l, (ḷ), l. The vowels continue entirely consistent with Indo-Aryan, just the diphthongs change: ŋ‘ , ŋ‘u > e, o; ŋā , ŋāu > ai, au. 7 For better documentation of similarities of Avestan and Vedic we have to look at a short Avestan text converted into Vedic: Ave. Təm amauuaṇtəm yazatəmж | s ›əm dām hu səuu təm, | Miϑrəm ₍‘₎ā ₎‘oϑ›ā’ (Yasht 10.6a-c); Ved. ṭʰ‘mж Ḥ t›ám ₍‘j‘ hót›ā’ʰ₍‘ (Indo-Iranian *tám ŋám‘u‘nt‘m ŋ ‘ ‘támж ŋć ›‘m ŋdʰ m‘su ŋćáu tʰ‘mж ŋḤ t›ám ŋ ‘ ā ŋ ʰáut›ā’ʰ ‘s), in English «Th s po⁽“›”ul d“ t₍ st›ong ‘mong th“ liv ng th“ st›ong“st Ḥ th›‘ж I honou› ⁽ th l ’‘t ons» (JACKSON, 1892, xxxi-xxxii). Similarly other Avestan texts can be converted into Sanskrit or vice versa without losing any the basic metrical principles of both languages (VAσοἴςŠβἰ 2007, 23-24). 6 ·6· Persian, Avestan has no modern successor. This fact is not overshadowed by the relative recency of the surviving Avestan manuscripts, because Avestan is in fact much older than Old Persian. In contrast with the other Iranian languages, we do not know which Iranian tribe or ethnos used the language or in which territory it was spoken. We even do not know the time-span when Avestan was used and we also do not know the original name of the language itself or in primary either in secondary sources. These questions can be answered only generally: Avestan was a language of an unspecified Iranian tribe (or tribes) that lived in the east part of the territory inhabited by the Iranian-speakers. We can suppose that Avestan was spoken in what is called Aⁱ› anəm V‘ēǰ 8 in Avestan (VХdēvdāt 1.1-2) and probably Avestan was the mother-tongue of Zarathushtra. Dating is controversial, we can assume roughly the period of 1200-700 . The name of the language is also questionable; we do not know the original name 9 ; Avestan is based on the name of the Holy Book of Zoroastrism – The Avesta. But this name is not original, it dates back to the Middle-Iranʧanл pʣʲʧodл andл ϑomʣʳл ʤʲomл κʧddὕʣл ἵʣʲʳʧanл уἵaʦὕavХф ʾp(₍)stʾk /abeʳtā₎ ~ a eʳtāʥ/ praises < Ir. ŋup‘-st u -k - (KELLENS 1987); Pers. Avest (A e ). Another plausible etymology is that the (Middle) Persian form comes from Ir. upa- -ka- ʤoundatʧonчл ϐaʳʣлуtʣʸtф луοʣʧnʣʲлLIPP, pers. comm.). consonants stops labials p b ttd dentals affricates fricatives sonorants m v/uu čǰ fу ф уф s z š ž alveolars postalveolars alveopalatals palatals velars labiovelars glottals k g ṧ уz) ʸ (ẏ) xу ф x h vowels iХ uū r hr š eē oō ń ʦ y/ii ʦ h əə aā a 10 Table 4 Sound system of Avestan (values in parentheses represent allophones) . The Avestan language as it is known today had undergone a complex development, part of which we cannot document according to known sources. One of the most important facts we have to realise is that the preserved form of the language had been already dead at the time It is not known where exactly was the territory of Aⁱ› ‘nəm V‘ēǰ , but it may be comparable with area of Ἀ ια ή (area of Afghanistan, Eastern and South-eastern Iran, Tajikistan, Turkmenistan and North-western Pakistan) mentioned by Strabo. 9 The language was probably called * A›₍‘n i.e. Iranian (Ave. ‘ⁱ› ‘- < Ir. *a› ‘-) by its speakers; similarly Old Persian has been called A›₍‘n (OPers. ariya-; OPers. ‘› ₍ānām; Ave. ‘ⁱ› ‘nąm, Aryan, Iranian (gen. pl.) > Pers. чл γāʲʳщл , Iran; Ave. ‘ⁱ› ‘nəm (adj.)) in the times of Darius ζщл aϑϑoʲdʧnʥл toл tʦʣл BХʳotūn (Behistun) Inscription (DB IV 98). 10 For a detailed description of Avestan sound-system see MORGENSTIERNE 1973. 8 ·7· when it was put down in writing (for the fiʲʳtл tʧmʣл ʧnл tʦʣл πāʳānʧanл pʣʲʧodчл ээяIndividual parts of the Avesta were originally passed on orally; the oldest preserved Avestan manuscripts come from the end of the 13th Linguists divide Avestan into two dialects – Older Avʣʳtanл уoʲл δāthāл Avʣʳtanчл δāthic) and Younger Avestan. Those two ὕanʥuaʥʣʳ лdoлnotлpʲʧmaʲʧὕyлʲʣpʲʣʳʣntлtʷoлϑʦʲonoὕoʥʧϑaὕὕyлdʧfferent stages of one language but they are two dialects of the same language – ἴὕdл уδāthic) Avestan being spoken in an older period and Young Avestan from the younger (KELLENS 1987; see also FRYE 1972). Avestan was passed on orally for a long time, perhaps for more than one thousand and five hundred years. The oldest preserved manuscript (K 7a) dates from , but there was th pʲoϐaϐὕyлoὕdʣʲлtʲadʧtʧonчлtʦʣлπāʳānʧanлaʲϑʦʣtypʣлʤʲomлtʦʣл5 (KELLENS 1987) were written in a new script created by an extension of the Pahlavi cursive script (this script was derived from the Aramaic alphabet). The Avestan script was been occasionally used to write in Middle Persian, such documents are called Pā₎‘nd (or Pāzend). Avestan alphabet consists of fifteen graphemes for vowels and forty graphemes for consonants, some phonemes can be written using multiple graphemes. Avestan script, regardless of graphical doublets, contained more graphemes than phonemes, the orthographical difference between the original phonetic system and writing was caused by a long oral tradition but also by inclusion of sub-phonemic material (e.g. the Schwa etc.). Since Avestan has already been registered as dead language, and there was no firmly codified spelling of the language, many words are often written in different ways in the same text – some notations therefore express different varieties of pronunciation that may have arisen in a later period. Avestan differs from *Proto-Iranian mainly in the following phonetic innovations: *-rt-, *-rϑ- > *[ɬ] (cf. MACKENZIE 1988, 90), *ʦ, *ʣ > s, z; *ʦu > *(t)sϕ > sp; ŋč > ṧii; palatalization or labialization of *h > ⁾, x 11, but *h, ⁾, x between vowels often > h, h, h; *nh > ṇg (or h); allophonic realisations x ~ / g ~ ; ϑ ~ / d ~ ; v/uu ~ / b ~ ; emergence of t *[d]12; nasalization of vowels (mainly *a > ą); in front of a nasal often ə; ŋ› > ər(ə), in front of voiceless consonant əhr; ŋ› > *ar > ar(ə); ŋ‘ , ŋā , * u > ‘ē, ā ~ i ~ ‘ē ~ əi, āu ~ əu ~ ao; i- and u-Umlaut; shortening of ŋā > ă in front of ŋ. , ŋ.u. Inлδāthā Avestan also lengthening of wordfinal vowels (perhaps a feature of recitation?). In Young Avestan there is often documented change of intervocalic *b, *d, *g > , , typical for the Middle Iranian period. Avestan grammar preserves much from *Proto-Indo-Iranian, majority of grammatical categories is similar to Old Persian and/or Vedic. Avestan preserved eight cases in three numbers (singular, dual, plural), declension is based on stem system, with vocalic stems (terminating in -a, -ā, -i, -Д, -u, - ) and consonantal stems (terminating ⁾ and x (also transliterated as h, h ) were in complementary distribution with hii and huu. There was threefold opposition of dental stops in Avestan: t : t : dлψtлśлdл: d] (i.e. t: +tense -voice, t: -tense -voice, d: -tense +voice; Reiner LIPP, pers. comm.) was probably an allophone of d word finally after a vowel, *r and *g (-Vt, -rət, -gət) and word initially before *k and *b (tk-, tb-). 11 12 ·8· in -n, -nt, -s, -z, -t, -d, -r, -r/-n, -m, -p, -k, -g, -h/- ). There was no difference in declension of the nouns and the adjectives. Avestan verbal categories are almost the same as they are in Vedic – the verb distinguishes three persons in three numbers, four tenses (present, imperfect, aorist, perfect and injunctive), five moods (indicative, conjunctive, optative, and imperative) and four voices (active, middle, stative and passive). Individual verbal forms are formed by connecting primary or secondary endings to a stem, and/or by adding augment or by reduplication of the stem. Each form can differ a lot from another because they may be influenced by position of stress. Avestan is in many respects more archaic than Old Persian and it provides better evidence for the state of *Proto-Iranian, on the other hand, the reconstruction of *Proto-Iranian is in many aspects based on Avestan. 13 I.1.1.2.2. Scythian and Sauromatian dialects, Cimmerian We have the information on the languages or dialects of the Scythians, Sauromatians (Sarmatians) and Cimmerians from Greek and to a lesser extent, from Latin, Old Persian and Assyrian sources. Language material is relatively modest, several dozen personal and ethnic names and a few glosses are known. When analysing the Scytho-Sauromatian data we can reconstruct some three hundred Scythian and/or Sauromatian roots (ABAEV 1949, 151-190), but their phonology is problematic. Since neither Greek nor Latin graphic system was suitable for accurate representation of Iranian languages phonology. In addition to personal names known from Scythian cities in the Northern Pontic region and some glosses in secondary sources we found also know one Scythian inscription written in Hittite hieroglyphs from the 7th at Saqqez (Kurd. Seqizфлʧnлἴʳtān-ʣлἰoʲdʣʳtānчлIran (HARMATTA 2002b). It is also believed that an undeciphered inscription in an archaic Kharo ṭʦХлʳϑʲʧpt (?) found on a silver bowl from Yesik (Issyk14) kurgan, Kazakhstan dated to the end of the 6th/beginnings of the 5th century is also Scythian (MENGHIN – PARZINGER – NAGLER 2007, 167, Abb. 131; AKISHEV 1978, 53-61). There are probably some Scythian inscriptions written in Aramaic script in the Northern Pontic region (HARMATTA 2002a). Herodotus noted that the Sauromatians spoke the language of Scythia but they did not speak the language well. There was a legend that the Sauromatian nation was formed after the Amazons married Scythian men, the Amazons initially did not learn the Scythian language properly and thus the Sauromatian language differed from the Scythian, «Φ ῆ л л ῆ чл л ὐ ῆ ἀ ἀ чл ὐ ῶл ὐ л Ἀ .» (HERODOTUS IV, 117). Sound system of Scythian is reconstructed approximately as for *Proto-Eastern Iranian – vowels and diphthongs probably continue *Proto-Iranian system without a change *a, ŋā, *i, ŋД, 13 For the purposes of this thesis, I decided to divide dialects of the Sauromatians / Sarmatians according to the historical sources into two chronological phases – I call the Old Iranian dialects as Sauromatian, by Sarmatian I mean the follower of Sauromatian in the Middle Iranian period. 14 λotлtoлϐʣлϑonʤuʳʣdлʷʧtʦлtʦʣлὕa₎ʣлÏʳïq-₎öὕлʧnлKyrgyzstan, Russian ь; Kazakh Ïstï‹-köl. ·9· *u, ŋ , ŋ and ŋ u; consonants also quite conservatively continue *Proto-Eastern Iranian stage, but we can observe several innovations: change *d > * > l; change *-rn- > l(l) (a dialectal feature?). Sauromatian has a similar evolution as Scythian but there are some different innovations: palatalization *r and subsequent shift to l before , ŋ ; disappearance of *f before *r; transition *p > f; and probably no change *d > l. Both dialects share the change *ʦu, *ʣu > sp, zb (later in Sarmatian *sp > *sf > *fs, *zb > *zv > *vz), loss of word-initial *h- (but not before , , ŋu), palatalization *t > c before , change *ϑ > t (?), often loss of word-initial ŋu- and metathesis *Cr > *rC (ABAEV 1979; HARMATTA 1970; VITCHAK 1992). It is difficult to assess whether merger of quantity of high vowels: *i, ŋД > ĭ and *u, ŋ > (difference in quantity of *a and ŋā remained preserved) and monophthongization ŋ , ŋ u > ē, started already in the Old Iranian period. consonants stops labials pb td dentals affricates c čǰ alveolars postalveolars fricatives f mw l sz šž nr palatals velars labiovelars glottals kg sonorants x x{ h vowels iХ uū ŋ‘ > ē < ŋ‘u yl у ф aā Table 5 Sound system of the Scytho-Sauromatian dialects. Documented data also provide poor evidence for Scytho-Sauromatian morphology. According to the Greek transcription we would assume that the Scythian nominal system maintained thematic vowels in the nominative, but they slowly started to disappear in the first centuries . Noun plural was formed by adding the ending *-tā- derived from abstract suffix ŋ-t(u -/*-ϑu -. The analysis of Scythian and Sauromatian personal names shows a number of word formation suffixes, many personal names were formed as tatpuru a composites (ABAEV 1979). In the Saqqez inscription we can recognise two forms of the preterite tense (HARMATTA 2002b). Scytho-Sauromatian dialects were developing through approximately 1000 years. Based on preserved material we cannot determine the exact chronology of individual changes. In the materials dated into the Christian era we can see changes that are typical for languages of the Middle Iranian period. It is questionable whether we can consider Cimmerian an Old (Eastern) Iranian language. From the rather scarce data we can assume that the Cimmerian language was a relative of Scythian, to which point shared innovations e.g. the same development *d > l. Apart from that, no much else can be said about the language. ·10· It is highly probable that already in the Old Iranian period there were some Scythian dialects which gave rise to ancestor(s) of Sogdian and YaghnōϐХщл Unfortunately we have no data that could confirm this theory. On the other hand there are three Central Asian Scythian (Saka) pʣʲʳonaὕл namʣʳл ʲʣϑoʲdʣdл ʧnл tʦʣл BХʳotūnл ʧnʳϑʲʧptʧonл уἴἵʣʲʳщл Skuⁿ⁾‘-, king of the S‘kā t g›‘⁾‘udā defeated by Darius I.; DB V 27) and in the Histories of Herodotus ( ό ι , queen of 15 the Massagetae and , son of Tomyris, Massagetian general; HERODOTUS I 207, 221). Those names do not give us much information about the language/dialect, but we may observe a similarity in [ ] and Sogd. to⁾mí; Yagh. taxm ʣʥʥ л< Ir. ŋtáu⁾m‘n- offspring, ʤamʧὕy (but see OPers. t‘umā-16). I.1.1.3. Middle Iranian period Languages of the Middle Iranian period can be characterised by four main innovations that took place throughout the Eastern Iranian language area: 1) monophthongization of diphthongs ŋ , ŋ u > *ē, * ; 2) change *b, *d, *g, ŋǰ > * , * , * , ŋž 17; 3) transition of *xt, *ft > * d, * d and *ʦ, *ʣ > *s, *z and *ʦu, *ʣu > *sp, *z (but in Saka and WakhХлşлŋś(ś), ŋź(ź)). Another common feature is a reduction or syncopation of unstressed vowels and gradual tendency to simplify nominal inflection. Verbal inflection also undergoes gradual changes, especially in past tenses where we observe a tendency to replace the original system by preterite formed from past participle. Together with the development of preterite the importance of ergative construction emerges. There are five rather well attested Eastern Middle Iranian languages – Sogdian, Khʷāʲezmian, Bactrian, Khōtanese and Tumshuqese; to a lesser extent we have information on the other languages and dialects such as Sarmatian and Alanic, Sogdian dialects of Bukhārā, Zhʣtʧʳuлandлςʳtʲōshanaчлoʲлʳʣvʣʲaὕлπa₎aлуŚa₎aфлdʧaὕʣϑtʳ from Chinese Turkestan (Uyghuristan, Xinjiang), Eastern Iran, Afghanistan, Pakistan and India. I.1.1.3.1. Sogdian Sogdian (Sughdian, Soghdian; ʳʷ y‚ʷл z ‚k /sə { y uл /) occupies a special position among the Eastern Iranian languages – its uniqueness can be viewed at two levels. From historical point of view it was probably the most successful Eastern Iranian language – it served Cf. Scythian чл₎ʧnʥлoʤлtʦʣлπϑytʦʧanʳлandл₎ʧnʥлoʤлtʦʣлAʥatʦyʲʳʧanʳлуtʷoл₎ʧnʥʳлʷʧtʦлtʦʧʳлnamʣлaʲʣл known; HERODOTUS I 78, IV 76). 16 It is questionable whether *Massagetian and Old Persian shared the change *xm > *m or whether should we read ρomyʲʧʳ лnamʣл ŋTóh/xmyris; see also Θ (< ŋTáh/xmyris ???) by POLYAENUS (Stratagems of War, 7.11.8), by other authors also ώ чл ύ ᾱ (JUSTI 1895, 328, 330), cf. also ς₍₍ayʧnХ and κāὕvāл Saka •tʰuma /tʰūmaъчл offspring (HARMATTA 1989, 305). Classical Persian form of the name is Tahm; Tomyris may be connected to ἵʣʲʳщлρaʦmХnaчлdauʥʦtʣʲлoʤлπamanʥānчл₎ʧnʥлoʤлρūʲānчлmotʦʣʲлoʤлπuʦʲāϐлʧnлFiʲdauʳХ ʳлShāʦnāmaлуʧϐʧdщчлюэřчлюьŚф. 17 ρʦʧʳлϑʦanʥʣлtoo₎лpὕaϑʣлʧnлaὕὕлpoʳʧtʧonʳлʣʸϑʣptлἵaʲāchХлandлŌʲmuṛХлʷʦʣʲʣлtʦʣлϑʦanʥʣлdʧdлnotлtoo₎лpὕaϑʣлʷoʲdinitially. In Ossetic there is no change *b- > * - word-initially; in the Saka dialects there is no change of wordinitial *g-. In some languages there is a change *d > (ŋ >) *l, see Excursion 5. 15 ·11· as a lingua franca on Central Asian route of the Silk Road (cf. DE LA VAζππζÈοβл 2005), it was not just a language of trade, many documents concerning three different religions – Buddhism, Manichaeism and Christianity were also translated into Sogdian. From present point of view we can consider Sogdian as a language that is preserved by a large amount of texts and it is also the language for which we know a closely related offspring – YaghnōϐХ (see I.1.1.4.1). Despite its outstanding status Sogdian practically did not survive Arab invasion to Central Asia, its influence slowly declined from the second half of the 8th century, during the 10th and 11th centuries it was gradually replaced by Persian, and Sogdian language enclaves survived only on the peripheries of Sogdiana. Geographically the Sogdian documents are attested from quite vast areas of Central Asia and its surroundings – majority of texts comes either from Sogdiana itself or from Sogdian colonies in Eastern Turkestan and Western China, other texts come from Mongolia, Zhetisu in Kazakhstan, Merv in Turkmenistan, or from Ladākh and arā oram in Pakistan; some ancient Turkic monumental inscriptions were also written in the Sogdian language. The language of Sogdian literal monuments appears to be relatively homogeneous despite the fact that the period between the oldest and the youngest documents is approximately five centuries long. Linguistic homogeneity can be observed mainly due to texts written in the so-called Sogdian script – orthography in this script was based on archaic form of Sogdian and emerged in 4th or 5th century and was preserved until the 8th century (or even up to the 11th century). Orthographies in Manichaean script and Syriac Esṭrangēlā script document classical stage of the language, but Sogdian of the 6th to 9th centuries did not differ much from its oldest attested form18. Archaic form of the language is known from so called Ancient Letters found in Chinese Dunhunag, other archaic features can be observed in Christian manuscript C 2; on the other hand, the Christian Sogdian texts contain many late-Sogdian features, such as the reduction of nominal inflection as it is documented in Christian manuscript C 5. Although the Sogdian documents are preserved in three different alphabets – Sogdian, Manichaean and Syriac 19 (and even fragmentary in North Turkestan variety of the BʲāʦmХ script), we cannot speak about three different dialects. Sound system of Sogdian is known only fragmentally – the language was written in consonantal alphabets of Semitic origin so there were no special graphemes for vowels 20, for An exception is a Sogdian translation of the Zoroastrian prayer A əm voh found in manuscript Or. 8212/84 (Ch. 00289) – this short text presents really archaic stage of the language (GERSHEVITCH 1976). 19 To mark the script of Sogdian documents I will use Ancient Letters gh Sogdian script from fortress of ʧʳōʲa₎ Zh 18 πoʥdʧanлtʣʸtʳлʷʲʧttʣnлʧnлtʦʣлBʲāʦmХлʳϑʲʧptщ 20 Vowels were written by so-called matres lectionis – ā by the letter āl‘p <‚şчл and by the letter waw <w> and , , ə and by the letter yuḏ <y>, diphthongs and were written by digraph waw-yuḏ <wy> or eventually as waw alone. By orthographical conventions in each script the letters āl‘p, waw and yuḏ could have been doubled, or some vowels could have been written by combination āl‘p-waw or āl‘p-yuḏ even word-internally. For word-final - also ·12· consonants often no distinction was made between voiced and voiceless sounds21. Despite the difficulties with interpretation of the Sogdian graphic systems, we can reconstruct Sogdian sound system. With the help of methods of historical and comparative linguistics, for more accurate reconstruction of phonology we can also utilize Sogdian fragments written in the BʲāʦmХ script. Sogdian vocalism was strongly influenced by position of stress – we can observe two main stress shifts: the first took place in an early stage of the language and its results can be seen not only in Sogdian but also in YaghnōϐХчл tʦʣл ʳʣϑondл ʳʦʧʤtл ʧʳл tʦʣл ʳo-called Sogdian Rhythmic Law – position of stress within a word depended on quantity of stem vowels. Both stress-shifts caused reduction, shortening or syncopation of unstressed vowels or even syllables. Basic development of Iranian vowels in Sogdian can be described as following: *a > a, ə (under the influence of i-Umlaut > e; due to a contact with a labial sound > u; under the influence of u-Umlaut > o~u); ŋā > ā, a (under the influence of i-Umlaut > ē, due to a contact with a labial sound > ); *u > u, ə, (under the influence of i-Umlaut > > ); ŋ > , u, ŋ‘u > (in front of ŋ⁾ n, *xm > o); ŋāu under the influence of i-Umlaut > ē; ŋ ‘u under the influence of i-Umlaut > ; ŋu > ; ŋu‘ , (or palatalized ŋu‘, ŋ‘u) > (later > ); *i > i, ə, ; ŋД > Д, i; ŋ‘ > ē; ŋ› > əʳ, iʳ, uʳ. It is also necessary to add some diphthongs to the vocalic system of Sogdian, apart from rising diphthongs and there were probably falling diphthongs ā and āu. Also nasals and r in front of a consonant in closed syllable (i.e. diphthongs like /Vṁ, Vṙ/, phonetically probably [Vəъσ , V ]) were of diphthongal nature. Consonantal system does not differ much from the form reconstructed for common Middle Iranian, significant changes include *ϑr, * r > , ž; *ʦr, *ʦ > ; *ʣr, * › ʣ, *ʣ > ž; *mp, *nt, *nk, ŋnč > ṁb, ṁd, ṁg, ṁǰ; *xt, *ft > d, d and in some cases palatalization of *k, *t > č when in contact with , ŋ (or ʳ < ŋ›). Iranian *x (< ŋhu) usually keeps its labial characteristics when word-initial before , in other positions it changes to non-labial x; in rare cases, however, there is a change *x ‘- > ⁾u-. Unclear is the development of Iranian *d, the sound is written in the Sogdian script by an Aramaic letter lāmaḏ <l> (in Sogdian it is transcribed as < şфчлin the Manichaean script letter < şлis based on the shape for lāmaḏ, but in the Syriac script the letter is written by dālaṯ <d>. It is possible that *d changed to a dental approximantлψ ], which continued in some dialects as and in some others as l (see excursion 5). The above mentioned Sogdian Rhythmic Law did not have an impact just on phonology – although it was originally a phonological rule, it strongly effected also morphology: Sogdian words split into two groups, so-called light and heavy stems according to the position of stress. the letter hē <h> could have been used in the Sogdian (and occasionally in the Syriac) script. The Syriac script also utilised diacritic marks for vowels: a <ᴤчл×> or < чл×>, ā <ᴟчл×>, <ỿчл×>, <ỵ>, <ẇ>, <ẉ> ( × m“‘ns ‘n₍ l“tt“›), but those diacritics were used rarely in Christian documents. 21 In the Sogdian alphabet there were only separate graphemes for and x, but forms of these letters usually merged together. The only script that had graphic symbols for both voiced and voiceless sounds (except and ϑ) was the Manichaean alphabet. In all three alphabets there was a clear distinction just between z and s and partially between ž and . ·13· Light stem endings were retained because they bore stress, unstressed endings of the heavy stems were lost or transformed. Substantives had three genders: masculine, feminine and neuter; neuter, however, survives only in a few relict forms. Nouns also maintain three numbers, the original dual was transformed into numerative (SIMS-WILLIAMS 1979). In the light stem inflection Iranian stem system continues in a transformed way (i.e. a-, ā-, i-, Д-, -, -, ya-, ₍ā-, and r-stems and also so-called contracted aka- and ākā-stems), heavy stem inflection consists just of three cases – nominative (direct case), vocative and oblique case; the light stems had six cases – nominative, vocative, accusative, genitive-dative, locative and instrumental-ablative. Plural was formed by adding the ending -t(a), animates have can have endings in or - . Adjectives are declined as nouns, but they gradually turn to uninflected forms. Personal pronouns had forms just for first and second person singular and plural, they were declined in two cases (direct and oblique), enclitic forms distinguished within oblique accusative, genitivedative and instrumental-ablative. Demonstratives distinguished triple deixis and were used also for the third person of personal pronouns. The definite article evolved from forms of the demonstratives of III. deixis. consonants stops bilabials p уϐ) affricates fricatives mw dentals t уd) postalveolars retroflexes (ṭ) (ḍ) (c) ( ) č уǰф sz šž velars uvulars labial velars ₎ уʥ) (q) у₎{ф уʥ{) (l) nr iХ uū eē ə oō () šлž (ṇ) x y у ф уü) palatals vowels ɨ f labiodentals alveolars sonorants a ā ʸ{ у {ф labial uvulars labiovelars ʸ° labiouvulars glottals (h) Table 6 Sound system of Sogdian (consonants in italics mark sounds appearing only in loan-words). Verbal system is based on present and perfect stems. Imperfect tense was originally formed by addition of augment to a present stem, in Sogdian augment was preserved only as so-called internal-augment between verbal prefix and stem, augment of non-prefixed verbs disappeared due to operation of stress. Perfect stem is derived from participles in *-ta-(ka-). Perfect distinguishes transitivity and intransitivity. Transitive verbs form perfect from the perfect stem and auxiliary verb ʾr, to have; perfect stems of the heavy stems have no ending, light stems end in -ú < *-am (< accusative singular of masculine). Intransitive verbs form perfect from the ·14· perfect stem and copula (but in forms of the third person singular there is no copula at all), forms of the light stems end in -í < *-ah (< nominative of masculine), the heavy stems have no ending. (excursion 1) Sogdian dialects of Bukhārā, Ustr shana and Zhetisu Sogdian seems to be a homogeneous language. It is quite difficult to observe several dialect differences – features that distinguish the languages of individual documents can be interpreted as developmental stages rather as dialects. We can observe some dialectal features in the preserved Sogdian texts; e.g. durative suffix ʾ tn (cf. Yagh. - t) appears in some Buddhist texts (e.g. V“ss‘nt‘›‘ jāt‘k‘) but in the majority of Sogdian texts there is the suffix ʾskwn and its forms. The phenomenon of the Sogdian dialects was solved by Walter Bruno HENNING (1958, 105-108) who notes that many differences between the language of the Christian documents in the Syriac script and the documents recorded in Manichaean and Sogdian alphabets can be in the case of Christian Sogdian interpreted rather as colloquial forms of later stages of the Sogdian language (HENNING 1958, 105). There is mention of a Sogdian dialect of Bukhāʲā in scientific literature. There are several inscriptions in the Old Bukhāʲanл уoʲл Sogdian-Bukhāʲanфл dʧaὕʣϑtл уϑʤщл LIVSHITS – KAUFMAN – α YAKONOV 1954; LIVSHITS – LUKONIN 1964‎‎), the authors unfortunately do not mention the differences between Literal Sogdian and Bukhāʲan-Sogdian. Based on my own analysis of several Bukhāʲanлinscriptions I suppose that in Bukhāʲanлthe Rhythmic Law was not applied and thus the Bukhāʲanлdʧaὕʣϑtлʷaʳлʳʧmʧὕaʲлtoлaлdʧaὕʣϑtлoʤлςʳtʲōshana. The *ςʳtʲōshanian dialect has been premised by Aὕ bert Leonidovich KHROMOV (1987, 645) and after him also by some other Tajik scholars (e.g. BUZURGMEHR 2005, 117). *Ustrōshanian is not attested in known sources, the premise of its existence is based on a hypothesis that from this dialect the YaghnōϐХлὕanʥuaʥʣл could have developed (KHROMOV, ibid.). Sources for knowledge of *Ustrōshanian may be taken from the fortress of Chilhujra in the South-Western part of the Ferghānaл vaὕὕʣy. The texts from Chilhujra have been published by Vladimir Aronovich LIVSHITS (2003). By my opinion these texts do not differ from other Sogdian texts. According to recent discoveries in Tajikistan ʷʣлϑanлʳuppoʳʣлaὕʳoлaлvaʲʧʣtyлoʤлςʳʦʲōshanian of the Mastchōʦлʲʣʥʧonл – documents found at the fortress of ʧʳōrak yet need a detailed analysis to be done (cf. LUR€E 2011; 2012). Apart from the above mentioned dialects we can also assume a Sogdian dialect of the Zhetisu (Semirech e) region. We have several Sogdian documents from Zhetisu from the 6th century, the use of a local Sogdian vernacular can be supposed till after the half of the 11th century (LIVSHITS 2008, 350-352). Zhetisu Sogdian is attested by two sources – the first are several rock inscriptions and ostraca, the other notes concerning (Zhetisu?) Sogdian in the Old Turkic lexicon Kitābu dēvānu lughāti ʼt-t rk by Ma mūd bin usayn bin Mu ammad AL-KĀSHGHAοФ. There are also some clues that show similar development of Zhetisu Sogdian and YaghnōϐХчлe.g. Zhetisu Sogdian word pwn /pun(n)/ corresponds to Yagh. pun(n) ×л pwrn-y ъpuwníъ full < ŋp n‘-; also the change *ϑ > t is similar to development of *ϑ in the ·15· Western dialect of YaghnōϐХщлZhetisu definite article is recorded as ʾyny / instead of Literal Sogdian ᵊ . We have no more precise clues then the above mentioned, therefore a precise reconstruction of the dialect of Zhetisu is still questionable. It is known from the historical sources that local Sogdian population adopted Turkic clothing and customs, but they had preserved their own language for quite a long time – e.g. Sogdian influence on lexicon and phonology of local Turkic dialects has been recorded (cf. LIVSHITS 2008, 350-351). I.1.1.3.2. Sarmatian, Alanic and Jassic Sarmatian and Alanic represent a dialect continuum based on Sauromatian dialects, it can be considered as language(s) of the Sarmatian, Alans, Roxolani, Jazyges, Aorsi, Siraces and Asi. (HARMATTA 1970), The beginnings of these languages can be dated from the 3rd century their development continues on Caucasus up today as the Ossetic language, or more precisely, it presents two dialects – Iron, the literal and standard form, and the quite archaic Digoron. Under Mongolian pressure together with the Cumans ( ypchā s) the Alanic Jassians migrated into Hungary. Both Sarmatian and Alanic material is scarce, we have mainly onomastic material and some borrowings in languages such as Hungarian or Chŭvash. Besides Sarmatian and Alanic glosses there is also a short Alanic inscription on a grave-stone from the 10th century from Zelenchuk ʧnлἰuϐan лdʧʳtʲʧϑt in Russia and two Alanic phrases were recorded in the 13th century by a Byzantine poet John (Ioannes) Tzetzes in his poem Theogonia. With regard to the scarce material it is difficult to draw the line between Sarmatian and Alanic, the label for the languages has been taken from the ethnic names of its speakers as they are known from historical sources. consonants stops labials pb affricates fricatives sonorants fv mw sz ž nr vowels dentals alveolars td postalveolars c čǰ palatals velars labiovelars glottals kg x ʸ° (h) ē ə ō yl у ф aā Table 7 Sound system of Sarmatian and Alanic dialects. Development of Sarmatian continues directly from Old Iranian Sauromatian, phonetic changes observed in Sarmatian show completion of the development outlined for Sauromatian above (I.1.1.1.2.); Sarmatian and Alanic vowels are reconstructed as a, ā, ē, > i, , > u. In front of word-initial consonantal clusters there appears *ə. Consonant system can be described as follows: *f (< *p), *t, *ϑ, *k, ŋč, *c > v, d, t, g, ǰ, ʒ; development of intervocalic clusters *ϑr, * r, *fr, *xr, * r > rt, › /rd, rv, rx, › as well as *sf, *z (< *ʦu, *ʣu) > fs, vz (HARMATTA 1970, 58- ·16· 97). A question is whether the change * > *d took place already in Sarmato-Alanic period or whether it was an Ossetic development. On morphology we have just fragmentary information. From the attested material we ascertained genitive singular ending -i, and nominative plural ending -ta. Original genitive plural ending *-ān‘m > *-ān lost its original function and became a suffix of some adjectives derived from nouns. Endings of the nominative singular disappear except a-stem feminines where *-ā- > *- (> Oss. -æ) remained (but > -ø). In phrases recorded by John Tzetzes there can be recognised some Alanic words, their grammatical forms have not been thoroughly analysed yet. Jassic is attested in one manuscript from the year 1422 which contains a brief Jassic word-list with their Latin and/or Hungarian translation. Forty three words are attested, while in the first part of the document there is a Jassic phrase and then a brief glossary follows, some other Jassic lexemes can be found in toponymy and onomastic of Hungarian district of Jászberény. The language extinction can be dated before the year 1693. Jassic is formally very similar to the Digoron dialect of Ossetic, the main feature that distinguishes Jassic from Ossetic is the preservation of * before nasals, in Ossetic there is an innovation o < ŋā /_{m, n}. The exact phonetic form of Jassic cannot be reconstructed on attested material – Jassic words are written in a similar way as medieval Hungarian, on one example we can suppose an ejective sound k <kh>, we can also suppose change ŋ , ŋž, ŋč, ŋǰ > s, z, c, ʒ known also from Ossetic 22 (see NÉκβρε 1959). I.1.1.3.3. Khwārezmian Khʷāʲezmian (Kh ›‘zmian) was a language of ancient Chorasmia, i.e. region of Khʷāʲʣzmл located in the KhХvaлoaʳʧʳл (present Qoraqalpogiston Autonomous Republic in Uzbekistan) on lower reaches of Āmūлαaryāлnʣaʲлtoлʧtʳлʣʳtuaʲy to the Aral Sea. Historically there are two stages of the Khʷāʲʣzmʧanл ὕanʥuaʥʣл – Middle 23 and Late Khʷāʲʣzmʧan. Middle Khʷāʲʣzmʧan is attested from two short inscriptions on ceramic vessels from the 3rd or 2nd century from Qoy-Qirilgan-Qal€a, other texts are known from inscriptions on coins, from silver-bowls from the Ural-area, documents written on wood and skin from Toproq-Qal€a and Yakka-Porson, from ossuary at Toq-Qal€a and from an ostracon from Xumbuz-Tepa. The Middle Proximity of Ossetic and Jassic can be illustrated also on ethnic names of both peoples – the name Jassian (forms of plural: Lat. Jazones / Jassones, Jazyges / Jaziges, Gre. Ἰ члἸ члεunʥщлjás₎ok, Russ. , Roman. áșĭ, Ger. Jassen) and Ossetian (from Russian , the Russian name comes from Georgian s“ti) have the same origin, see also Greek names of Scytho-Sarmatian tribes Ἀ ῖ чл Ἄ щл In contemporary Ossetic As ‖л As(s)i labels Caucasian Balkars and Balkaria, in Abkhaz the region of Northern Caucasus is called Aś (ABAEV 1958, 479-480; NÉκβρε 1959, 5-13). The Ossetians call themselves either Ir ‖лГ›æ or D gu› ‖лDigor according to their language and ethnicity. 23 Helmut Humbach proposes for the oldest attested from of Khʷāʲʣzmʧanл aл ὕaϐʣὕл Middle Kh⁽ā›“₎m ‘n (HUMBACH 1989, 193), the term Old Kh⁽ā›“₎m ‘n remains untapped, it probably serves as a label for the oldest, unattested form of the language from the Achaemenid period. 22 ·17· Khʷāʲʣzmʧanлtexts were written in a local variety of the Aramaic alphabet. Late Khʷāʲezmian is a language of documents written in adapted Perso-Arabic script. Main sources of the Late Khʷāʲezmian language are the following works: interlinear translation of encyclopaedia Muqaddimaẗ al-adab by ηārullāʦ Abū-l-Qāʳʧm Ma mūdл ϐʧn ƒUmar az-ZamakhsharХ from the year 1135, glosses in a legal document Qunyaẗ al-munyaẗ li-tatmДm al-ghunyaẗ by NajmiddХn Abū Ra₍ā Mukhtār bin Ma mūd az-ZāhidХ al-GhazmХnХ from the 13th century (Qunyaẗ al-munyaẗ contains also Khʷāʲezmian quotations from YatДmaẗ ad-dahr fД fatāwâ ahl al- asr by Mu ammad bin Ma mūd ƒAlā‚uddХn ƒAbdurra Хm at-TarjumānХ al-MakkХ al-KhuwārazmХ), glosses from Qunyaẗ al-munyaẗ and Y‘tím‘ẗ ad-dahr were collected in Risālaẗ al-alfā al-khuwārazmiyyaẗ allatД fД qunyaẗ al-mabs t by JamāliddХn al-ƒImādХ al-Jur₍ānХ around the year 1350. Calendar, astronomical and medical terms together with names of kings of Khʷāʲʣzmлaʲʣлattʣʳtʣdлʤʲomл the works of Abū-r-Ray ān Mu ammad bin A mad al-BērūnХ Kitāb al-āthār al-bāqiyaẗ an al-qur n ‘l-khāliyaẗ and Kitāb as-saydana fД-t-tibb from the beginnings of the 11th century (HUMBACH 1989, 193-194, ZARSHβλĀπ 1357, 57-59). Khʷāʲezmian became extinct sometime in the 14th century when it was replaced by Oghuz- ypchā variety of Turkic. In the so-called Kh⁽ārezm-T ›kД language there were numerous influences of Khʷāʲʣzmʧanлʳuϐʳtʲatʣчлʳomʣлoʤл the Khʷāʲʣzmʧan words can be heard in Uzbek dialects of Xorazm (Khʷāʲʣzmфл even today (LIVSHITS 1962, 140). Classification of Khʷāʲʣzmʧanл ʧʳл unϑὕʣaʲ – Dzhoy Iosifovich Èdʣὕ man assignes it to Northern group of the East Iranian languages (ÈDEL MAN 2000a, 95; ÈDEL MAN 2008, 6), but in her older work she claimed Khʷāʲʣzmʧanл toл ϐʣл tʦʣл πoutʦ Eastern Iranian language (ÈDEL MAN 1986, 6). Khʷāʲʣzmʧanлʳʦaʲʣʳлʳomʣлʤʣatuʲʣʳлʷʧtʦл Alano-Ossetic dialects, ʳomʣлotʦʣʲл ʤʣatuʲʣʳлὕʧn₎лʧtл ʷʧtʦл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳ; many similarities with Sogdian are also interesting. Cherāgh-ƒAlХ AƒẓamХ and Gernot Windfuhr see some similarities between Khʷāʲezmian and North Western Iranian SangesārХ (Aƒ AκФ – WINDFUHR 1972). consonants stops bilabials pb affricates fricatives sonorants mw vowels iХ uū f labiodentals dentals alveolars td postalveolars c čǰ sz šž palatals velars labiovelars glottals kg x x{ h eē nrl y у ф ə ō aā 24 Table 8 Sound system of Khʷāʲezmian . ζnлtʦʣлἰʦʷāʲʣzmʧanлadaptatʧonлoʤлtʦʣлἵʣʲʳo-Arabic script there are also letters used only in Arabic (i.e. , s, t, z (d), , , qфчлϐutлtʦʣʧʲлpʲonunϑʧatʧonлʧnл ἰʦʷāʲʣzmʧanлʧnлnotл₎noʷnчл tʦʣyл ʷʣʲʣлpʲoϐaϐὕyл pʲonounϑʣdл ʧnл aлʳʧmʧὕaʲл way as in Classical Persian. 24 ·18· Since Khʷāʲʣzmʧanл is recorded by alphabets of Semitic origin, we have no clear idea of Khʷāʲʣzmʧanл voʷʣὕʳчл voϑaὕʧϑл ʳyʳtʣmл ʧʳл ʲʣϑonʳtʲuϑtʣdл aʳл ʤoὕὕoʷʳśл a, ā, i, Д, e, ē, , u, , ə25. In development of Khʷāʲʣzmʧanлvoϑaὕʧϑлʳyʳtʣmлʧtлʧʳлʧmpoʲtantлtoлundʣʲʳtandлopʣʲatʧonлoʤлʳtʲʣʳʳл– short unstressed vowels (including ŋ›) were reduced, long unstressed vowels were probably shortened. Vowels that were not affected by operation of stress generally did not differ much from the Middle Iranian stage. The only exception was *a, that often changed to i. Besides oral vowels there were also nasalized vowels that emerged after deletion of nasals in front of a consonant or in word-final position, nasalization was often not marked in writing. The stress was mobile, it remained on word-stem. Due to the stress shift vowels within a word changed, some changes were also influenced by sandhi. In Khʷāʲʣzmʧanлʳomʣлϑonʳonantʳлʷʣʲʣлpaὕataὕʧzʣdл in front of and ŋ (or even ŋ‘ ): *ḱ, ŋt > c; ŋǵ, ŋd > ʒ/z; *ϑ, ŋś, *ṧ > s; some other consonants were depalatalized ŋč, ŋǰ, ŋž > c, ʒ/z, z. After palatalization and depalatalization has been completed, voiceless consonants were probably sonorized when post-vocalic or after a nasal: *-p-, *-t-, *-k-, *-c- > *-b-, *-d-, *-g-, *-ʒ-26. Other differences from the Middle Iranian consonant system are: *ϑu > f; * u > ; ŋ” > f; *fr > f, fr (word-initially also r-, -, h-); ŋ m > m; * r- > ϑ-; *-ϑr- > r; *ϑr- > - (in other cases hr-, Vr-, rc-); *ϑn, *rn > n; ŋt› > c, č; *rs, ŋ› , *sr, ŋ t› > ; *rz > ž; ŋ > x, f, h, s, y27; ŋ⁾ , ŋ⁾ u > x; ŋ⁾u > ⁾° (in front of ), x. (ÈDEL MAN 2008, 13-26) Khʷāʲezmian nouns and adjectives distinguished two genders (masculine and feminine) and two numbers (singular, plural; for nouns as a relict also dual). Nouns were inflected in three cases in singular: direct (nominative-accusative), oblique (labelled also as ablative, locative or instrumental) and genitive (possessive), in plural there are just two cases: direct and oblique. Personal pronouns of the first and second persons singular have four cases (nominative, accusative-dative, ablative-locative and genitive), in plural there are again just two cases (direct/nominative and oblique/genitive), and for personal pronouns of the third person demonstratives were used. Demonstratives have triple deixis, they do distinguish gender but inflectional system was greatly simplified. Khʷāʲezmian has a definite article (one form for masculine and plural, the other just for feminine singular). The definite article originates in forms of the demonstratives of III. deixis. Verbal system preserves quite a large range of moods: indicative, imperative, conjunctive, irrealis, optative and injunctive, there are also grammatically expressed categories of transitivity and intransitivity and aspect. The verb has three stems – Long vowels were written with matres lectionis: alif <‚ş – ā, ⁽ā⁽ <w> – , , ₍ā₍ <y> – ē, Д; short vowels were occasionally marked by Arabic vocalic signs (h‘›‘kāt), kasra was used for i and also for e and ə. To mark the position of stress Arabic sign tashdДd уtʲanʳϑʲʧϐʣdлaʳл<ὴşлoʲл<¯şфлϑouὕdлʦavʣлϐʣʣnлuʳʣdщлуÈDEL MAN 2008, 12) 26 Sounds g and ʒ are not marked by special letters, about their voiced pronunciation is considered analogous to the evolution of *-p- and *-t-. 27 Development of Iranian ŋ is diverse in Khʷāʲʣzmʧanл– in vicinity of ŋ‘u, it changes to x, however, after labial consonants ŋ > f (e.g. ŋg‘u ‘-, ear, ŋm -, mouse, > wx ъ ōʸъл ×л mwf ъmūʤъф;л ʷʦʣnл paὕataὕʧzʣdл oʲл ʧnл ʤʲontл oʤл suffixed *s it changes to s; word-internally (after a palatal ??) ŋ > y (e.g. *fra-p ‘-, to thrash, > p₍-); in other cases ŋ > h. 25 ·19· present, imperfect and preterite. Present tense comes from Iranian present stems, imperfect stem is formed from the present stem with addition of reflexes of augment; perfect is based on Iranian participles in *-ta-(ka-) and auxiliary verb ʾ›ȳ- < ŋdā›-, to have. Characteristic feature of Khʷāʲezmian verbs is use of postverbs – enclitic particles determining direct or indirect object of a clause. Postverbs were derived either from enclitic pronouns or from particles or prepositions. (ÈDEL MAN 2008, 26-54) I.1.1.3.4. Bactrian Bactrian (also called Eteotokharian, Tokharian, Kushān ‘n or Kushāno-Bactrian), language of Bactria, is attested from several dozen inscriptions written in a local adaptation of the Greek alphabet and also from several texts written in the Manichaean script from a period from the 2nd th to the 9 centuries mainly from Northern Afghanistan and Southern Tajikistan, to a lesser extend from ἶaὕƒa-yi Afrāsiyāϐлnear Samarkand, from the ρuʲʤān oasis in Eastern Turkestan or form the Hunza Valley in Pakistan. Some scholars believe that Bactrian can be closely related to κun₍Хл andл Ḳʧdghā. By comparing words attested in the Greco-Bactrian alphabet with those written in the Manichaean script we can quite well reconstruct the phonology of Bactrian – the advantage of Greco-Bactrian alphabet is especially the ability to record vowels, which writing systems derived from the Aramaic alphabet do not allow well enough. consonants labials dentals stops affricates fricatives sonorants vowels ( ) ( ) alveolars postalveolars ϸ() у ) palatals velars labiovelars glottals Table 9 Sound system of Bactrian (given in letters of the Greco-Bactrian alphabet). Phonological development of Bactrian can be characterised as follows: * > l; *ϑr > hr; *p, *t, *k > , d (~ ), g (~ ); ŋč, ŋǰ > ʦ, ʣ (> s, z); in Manichaean Bactrian *ϑ > h. In later stages of the language articulation of h is lenited or even lost. Comparison of texts in the Manichaean and Greco-Bactrian alphabets proves maintaining differences in quantity of vowels. In morphology there was ascertain a reduction of Old Iranian inflectional system into two cases – direct and oblique, dual was lost and neuter merges with masculine. Attested is a definite article that distinguishes gender, reflexive article Д (m) / ya (f) performs a function similar to Persian izāfaẗ. Verbal morphology is based on a system of two stems: present and past; inflection is based on stem endings in ŋ-‘ ‘-, which is comparable with the Western Middle ·20· Iranian languages. Past tense is formed by ergative construction (STEBLIN-KAMENSKIY 1981; SIMS-WILLIAMS 1989c; LIVSHITS 2000). I.1.1.3.5. Kh tanese and Tumshuqese, Saka dialects Khōtanese (Kh t‘n S‘k‘; OKhōtщлhvatanau; LKhōtщлhva au, hvaṃ) and Tumshuqese (Tumshuq Saka, ё₍āźd“s“, ё₍āźd ‘n28, in older works also Maralbashi Saka;л ʥyāźdʧyā- ?) are two closely related languages of the Saka (Śa₎a) of Eastern (Chinese) Turkestan. Both languages were written in Turkestan varieties of the BʲāʦmХ script, but each language had its own orthographical conventions – Khōtanʣʳʣлuʳʣdлmaʧnὕyлdʧʥʲapʦʳлtoлʲʣpʲʣʳʣntлʳoundʳ not present in BʲāʦmХ but Tumshuqese used new ak aras (so called Fremdzeichen). Tumshuqese was a language of the Gyāźdi region/kingdom, it is attested in fifteen texts from the 7th and 8th th th centuries (or even from the 4 and 5 centuries; cf. EMMERICK 2009, 379; EMMERICK 1989, 204) found on archaeological sites Tumshuq, Maral-bashi (Barchuq) and Bäzä₎ὕʧ₎л уκuʲtuqф. Tumshuqese is more archaic relative of Khōtanese – a language attested form Buddhist texts from the 7th to the 10th century from territory of ancient kingdom of Khōtan (OKhōtщлђv‘tän‘-, LKhōtщлHvaṃ(na-), Chin. Yutien, Hetian), from the ρuʲʤān Oasis and from Chinese Dunhuang. In Khōtanese there can be observed two stages of language development: Old Khōtanese (language of the kingdom of Khōtan) and Late Khōtanese (language of the ρuʲʤān oasis)29. Phonological development of Khōtanʣʳʣлʧʳлquʧtʣлϑompὕʧϑatʣdлʳoлζлʷʧὕὕлmʣntʧonл₍uʳtлʧtʳлϐaʳʧϑл features. The vocalic system has been largely rebuilt, there is a reduction of vowels on one side and compensatory lengthening on the other side, primary or secondary diphthongs were monophthongized, many vowels were also palatalized, labialized or contracted. The development of Iranian ŋ› is also complex. Old KhōtanʣʳʣлʦadлtʣnлdʧffʣʲʣntлvoʷʣὕʳśлъʧчлХчл , члaчлāчлoчлuчлūчлə/, these are reduced to four vowels and one diphthong in the later stage of the language: / / < OKhōtщлъ чл , /; /a/ < OKhōtщлъaъ;лъɔ/ < OKh ə/ < OKhōtщлъə/ or an unstressed vowel, and diphthong / ɔ/ < OKhōtщлъūъщлαʣvʣὕopmʣntлoʤлϑonʳonantʳлʧʳл₍uʳtлaʳлϑompὕʣʸśлʷoʲdinitial consonants remained unchanged (except *f-, *ϑ-, *x- şлъpʰчлtʰчл₎ʰъ;л*fr-, *ϑr-, *xr- > /br, dr, gr/ and ŋ - > OKhōt ъǵъл şл ἱKhōt ъǰъфчл voʧϑʣὕʣʳʳл ϑonʳonantʳл уʣʸϑʣptл *s) were sonorized in word-internal and word-final positions and later they have undergone other changes such as syncopation, palatalization or they may have formed a diphthong (which was later usually 28 Rong Xinjiang proposes instead of naming Tumshuqese (made by modern place-name Tumshuq in Eastern Turkestan, where the documents in the language had been first found) a more appropriate name derived from the ʦʧʳtoʲʧϑaὕлʲʣʥʧonлoʤлδyāźdʧлуρumsh. ё₍āźd -, Chin. Jushide, Tibetan Gus-tig) – ё₍āźd“s“ or ё₍āźd ‘n (RONG 2005). In this work I am going to keep the label Tumshuqese as it is customary in other scientific works. 29 Leonard Georgievich Gertsenberg characterizes interrelationship of Old and Late Khōtanʣʳʣл aʳл ʲʣὕatʧonʳʦʧpл oʤл Latin and Modern Italian (GERTSENBERG 1981, 234). He sees the archaicity of Old Khōtanʣʳʣлpoʳʳʧϐὕyлʧnлanлoὕdʣʲл scribal tradition in Khōtanл andл ἱatʣл Khōtanʣʳʣл ʧʳл ʣʸpὕaʧnʣdл aʳл aл vaʲʧʣtyл oʤл ϑoὕὕoquʧaὕл ὕanʥuaʥʣл oʤл tʦʣл ἰʦōtanʣʳʣл people in ρuʲʤānл уʧϐʧdщфщл οonaὕdл βʲʧϑл βmmʣʲʧϑ₎л ϑὕaʧmʳчл tʦatл aϑϑoʲdʧnʥл toл paὕaʣoʥʲapʦʧϑл anaὕyʳʧʳл tʦʣл oὕdʣʳtл th th Khōtanʣʳʣлtʣʸtʳлϑanлϐʣлdatʣdлaὕʲʣadyлtoлtʦʣл5 and 6 EMMERICK 2009, 378), it is possible that the orthography of Old Khōtanʣʳʣлdʣvʣὕopʣdлʧnлthat period. ·21· monophthongized). Palatals ŋč, ŋǰ are depalatalized to /l, i/ when preceding back vowels (but ŋč > /lʰ/); *ʦu, *ʣu ϑʦanʥʣʳлʧntoлъśчлźъ30 etc. In the development of consonants there is also a significant difference between the Old and Late Khōtanʣʳʣщлρʦʣʲʣлaʲʣлaὕʳoлʣvʧdʣntлζndo-Aryan influences on Khōtanʣʳʣ consonantism – emergence of retroflex sounds 31 and a transition of non-sibilant voiceless fricatives into aspirate stops *f, *ϑ, *x > pʰ, tʰ, kʰ32 (see EMMERICK 1989, 209-216 for details). Syncopation of consonants could have caused changes in tonal colours of surrounding vowels, such feature could be expected especially in cases of *-r- and *- - уşлъž/ > ø oʲлъ‚ъфчлʧnʳtʣadлoʤлtʦoʳʣлʳoundʳлtʦʣʲʣлʧʳлaлʦoo₎ < > ʷʲʧttʣnлϐʣnʣatʦлaлὕʣttʣʲлʧnлtʦʣлBʲāʦmХлʳϑʲʧptл – the hook is usually transliterated as an apostrophe at the end of a syllable or as subscribed hook (i.e. ‘ʼ or aфщл ἶuʣʳtʧonл ʧʳл ʷʦatл ʳoundл doʣʳл tʦʧʳл ʦoo₎ л ʲʣpʲʣʳʣntśл ἱʣonaʲdл δʣoʲʥʧʣvʧch Gertsenberg supposes that it marks some tonal quality (GERTSENBERG 2000, 49) or even a ʥὕottaὕлʳtopлъ‚ълуGERTSENBERG 1981, 237), Ronald Eric Emmerick does not specify its phonetic value (EMMERICK 1989, 209) or claims it to be a marker of a breathed syllable (EMMERICK 2009, 381). consonants stops affricates p pʰ (b) bilabials dentals tt tʰ d alveolars postalveolars retroflexes ṭ ṭʰ palatals k ₎ʰ gg velars labiovelars glottals € tc ts js c ϑʰл₍ k ky gy fricatives b t s ys śśлś sonorants mv n r rr l iХ uū ä o/au /ə/ ( ) ñy h: g hv h vowels ṅ/ṃg e/ai ā a 33 Table 10 Sound system of Old KhōtanʣʳʣлуvaὕuʣʳлʧnлtʦʣлtaϐὕʣлaʲʣлϐaʳʣdлonлtʲanʳὕʧtʣʲatʧonлoʤлtʦʣлBʲāʦmХ script) . A series of changes occurred also in morphology. In nominal inflection the Old Iranian stem system was heavily transformed into a new system of almost two dozen inflectional classes. Genitive case merged with dative, and instrumental merged with ablative. Neuter usually merged with masculine but in some cases neuter was preserved as newly-build n-stems. Dual was lost, with some exceptions. Number of cases has been further reduced in Late Khōtanese, prepositions or postpositions were used to a greater extent to express cases. πʧmʧὕaʲл ϑʦanʥʣл ʧʳл attʣʳtʣdл aὕʳoл ʧnл ἵāmХʲл τakhХ: Ir. ŋáʦu‘-, horse > Khōtщл ‘śśä [aʆу”ф ], Wakh. ₍‘ ×л Avʣщл aspa- (but OPers. asa-), Ved. áśv‘-. 31 Due to contact with the Indo-Aryan languages the retroflex consonants can be met also in other Iranian languages, e.g. in PashtōчлτakhХчлζsh₎āshmХ-πanʥὕēchХ, YidghāлoʲлBaὕōchХ. 32 πʧmʧὕaʲлʤʣatuʲʣлϑanлϐʣлʳʣʣnлaὕʳoлʧnлἵaʲāchХчлŌʲmuṛХлandлλoʲtʦ-West IranʧanлBaὕōchХ. 33 In Late Khōtanʣʳʣлъśчлźълaʲʣлuʳuaὕὕyлʷʲʧttʣnлaʳл<śşчл<ś ъśşлу×лἴKhōtщл<śśşчл<śşфлandлъšчлž/ as < >, < ъʳşлу×лἴKhōtщл < >, < >). For OKhōtл<ṭʰṭʰ> and <k > /ṭ ʰ/ stands just <k > in Late Khōtanʣʳʣщ 30 ·22· consonants stops affricates p pʰ b bilabials dentals t tʰ d dʰ alveolars postalveolars ts dz c ϑʰл₍ retroflexes w ḏ sz śź ẓ ky gy palatals velars fricatives k g glottals kh ǥ h sonorants mv vowels ä i u nrṟl e ñy ṅ /ə/ a o ā Table 11 Sound system of Tumshuqese (values in the table are based on transliteration of the BʲāʦmХ script). Verb distinguished all inherited moods as well as active and middle voice. Also verbal endings continue from *Proto-Iranian, in this case the forms of the endings may differ due to Khōtanʣʳʣлʳoundлϑʦanʥʣʳщлζnnovative is transformation of tenses – Khōtanʣʳʣлdʧʳtʧnʥuʧʳʦʣʳл₍uʳtл opposition of present and perfect. Perfect is based on opposition of transitive and intransitive verbs – each of these categories has its own set of endings (EMMERICK 1989; GERTSENBERG 1981). Knowledge of Tumshuqese is poor in comparison to Khōtanese. Tumshuqese is generally much more archaic, both in phonology and in morphology; there is e.g. no sonorization of word-internal voiceless vowels or no palatalization of vowels (EMMERICK 1989, 204-205). As was observed by Janos Harmatta, beside Khōtanese and Tumshuqese there are also some other Saka dialects, so-called dialects of Southern Saka – SХstān Saka, Gandhāra Saka, Mathurā Saka and ς₍₍ayʧnХ and κāὕvāл πa₎a. The dialects of Southern Saka are attested mainly on onomastic material in some Prakrit texts written in the BʲāʦmХ and Kʦaʲō ṭʦХ scripts, occasionally there are some glosses in the Greek alphabet (HARMATTA 1989), another Saka dialects of the Eastern Turkestan attested by several glosses are Murtuq Saka (a variety of Tumshuqese?), Krōraina Saka, ἰāshghar Saka (Kancha₎Х, ἰan₍a₎Х) and Indian Saka (GERTSENBERG 1981, 234). Question is whether unattested languages of S‘kā t g›‘⁾‘udā and S‘kā h‘um‘v‘›gā known from Old Persian sources were the proper languages of the Saka, or whether they were spoken by Central Asian Scythians. I.1.1.4. New Iranian period In the New Iranian period is attested majority of the known Eastern Iranian languages. Three languages – τan₍Хчл ḳēϐā₎Хл andл SarghuὕāmХ – died in on the beginning of the last century. There are now 20 living Eastern Iranian languages spoken by approximately 32˙809˙000 people (excluding Pashtōлsome 809˙000 people). Only Ossetic and Pashtōлʦavʣлoʲtʦoʥʲapʦyлoʤлʧtʳлoʷnчл the other languages have no written tradition. Modern Eastern Iranian languages differ considerably one from the other. All the languages have simplified nominal declination to maximally three cases system. Verbal inflection was in many languages much simplified, majority of past tense verbal forms is based on ergative ·23· construction. Typical Iranian subject-object-verb word order continues in all Eastern Iranian languages. I.1.1.4.a. North Eastern Iranian I.1.1.4.1. Yaghn bī 35 YaghnōϐХ (Yaghnā’Д, incorrectly also Neo-Sogdian34; ) is a language originally spoken in a high-mountain valley on the upper reaches of the river Yaghnōϐ in Aynî district in North-Western Tajikistan. In the 18th century some of the YaghnōϐХʳ settled southern slopes of the iʳāʲлrange ʧnлnoʲtʦʣʲnлpaʲtʳлoʤлtʦʣлσaʲzōϐл district South of Yaghnōϐ and several villages in Ghōnchî district in the Ferghāna Valley; later in the half of the 20th century some YaghnōϐХʳ settled southern parts of Varzōb and Northern iʳōr regions (BUZURGMEHR 2005). In the years 1970 and 1971 all the population of the Yaghnōϐ valley was forced to move to the Zafarōbōd district in the Hungry Steppe (MДrz č l; LOY 2005), some of the YaghnōϐХʳл ʲʣtuʲnʣdл ϐaϑ₎л toл tʦʣʧʲл ʦomʣὕandл ʧnл the ʣaʲὕyл ьŚŚы ʳ, today there are approximately 500 people living in the Yaghnōϐ Valley36 (MФοḳŌḳŌαA 2008, 6). There are some ьэ˙5ыылpeople who consider themselves YaghnōϐХ, of which approximately 8000 speak YaghnōϐХл αʣʳʧʥnatʧonл λʣo-πoʥdʧan л ʷaʳл ʲaʲʣὕyл uʳʣdл ʧnл oὕdʣʲ scientific literature (cf. BOGOLYUBOV 1956). Nowadays YaghnōϐХлʧʳлaὕʳoлϑaὕὕʣdлsu , Sogdian by some of its speakers. This is a quite recent phenomenon caused by the emerging national self-awareness of the YaghnōϐХʳщ 35 by some of its speakers. The name of the language is The language is called also ya n ⁽ or even ya derived from the name of the Yaghnōϐл ʲiver and its valley (Tjk. Ya , Yagh. Yá n ’, Yá n u). The original name of the river and its valley has two possible etymologies: 1) it either comes from Yagh. ya d ~ yaxt wide (Sogd. y (ʾ)›t-y, yr t y rṯ-y /yəуwф dí/) and n u valley, dale > *ya d-n u > Yá (d)n u > Tjk. Ya n (but also Yá (d)n u); 2) or ʧtлϑomʣʳлʤʲomлρā₍Х₎л cold, icy or yaxn cold place (cf. Sogd. yxn(w) /yəʸnúл~ xn(u)/ ice ) and ’ water (Yagh. p) > - ’> > Ya луϑʦanʥʣлъʸnълşлъ nълϑanлϐyлʣʸpὕaʧnʣdлaʳлvoʧϑʣлaʳʳʧmʧὕatʧonчл ϐutл ʳuϑʦл aл ϑʦanʥʣл ʧʳл attʣʳtʣdл nʣʧtʦʣʲл ʧnл ρā₍Х₎л noʲл ʧnл ḲaghnōϐХ;л ʧtл mayл ϐʣл ʣʸpὕaʧnʣdл aʳл dʣvʣὕopmʣntл ϑauʳʣdл ϐyл ρā₍Х₎-YaghnōϐХлϑontaϑtл??) – this etymology can be supported by YaghnōϐХлtoponymyщлζnлtʦʣлἶūὕлσaὕὕʣyлtʦʣʲʣлʧʳлaл brook called Ḗ⁾ ḥ u (or Ḗ⁾n uфл ζϑʣ αaὕʣ лʧnлḲaghnōϐХлуḲagh. ē⁾, Д⁾, ice, Tjk. yax < Ir. ŋ‘ ⁾‘-фщлρʦʣлῦkhi λōʷл brook is located in the southern part of the Yaghnōϐл σaὕὕʣyл andл ʧtл flows into the Shōʷkhōnл ʲiver (i.e. main tributary of the Yaghnōϐлʲiver in the YaghnōϐлσaὕὕʣyлʧtʳʣὕʤфщлAὕonʥлtʦʣлʲiver Shōʷkhōnлʲunʳлonʣлoʤлtʦʣлуʦʧʳtoʲʧϑaὕὕyфл most important paths connecting the valley with the σaʲzōϐлʲʣʥʧonчлʳoлmayϐʣлtʦʣлḲaghnōϐлvaὕὕʣyлreceived its name through ρā₍Х₎лʲʣanaὕyʳʧʳлoʤлḲagh. Ḗ⁾( ) ḥ u: tʦʣлρā₍Х₎ʳлanaὕyʳʣdлtʦʣлḲaghnōϐХлʦydʲonymлaʳлŋ ⁾(. )n- u and it was later calqued as Tjk. *Yax( )n- ’ ψρā₍Х₎лdoʣʳлnotлdʧʳtʧnʥuʧʳʦлvoʷʣὕлquantʧtyлoʤл йД and uй inherited from Persian] > . Ya Both theories i.e. Yaghnōϐл aʳл τʧdʣ αaὕʣ л oʲл ζϑʣ αaὕʣ can be considered correct, or maybe the name of the Yaghnōϐл σaὕὕʣyъʲiver emerged from a combination of both names, since it is considered that the name as it is ₎noʷnл todayл ʦaʳл ϐʣʣnл adoptʣdл ϐyл tʦʣл ρā₍Х₎ʳщл ἵʦonʣtʧϑaὕὕyл *ya d-n u is more accurate than Ḗ⁾( ) ḥ u or - ’. 36 Before the forced migration there were approximately 2500 people (KHROMOV 1972, 4), 1794 of them were ḲaʥʦnōϐХ-speakers in 1952 (ibid.: 6). 34 ·24· (MФοḳŌḳŌαA pers. comm.). YaghnōϐХл ʳpὕʧtʳл ʧntoл tʦʲʣʣл dialects – Western, Transitional (or Central) and Eastern37. The language does not have any literary tradition. First books written in YaghnōϐХ (dictionaries, text-books etc.) ϐʣʥanлtoлappʣaʲлʧnлtʦʣльŚŚы ʳ, today the task to create YaghnōϐХлorthography is in progress. A Tajik form of the Cyrillic alphabet serves as the basis for written YaghnōϐХ. YaghnōϐХл ʳoundл ʳyʳtʣmл ʧʳл ʲʣὕatively archaic – vowels have not been affected much by Umlaut, consonants continue from the Middle Iranian stage, with only little changes. The development of vowels is closely related with stress, it seems that *Proto-YaghnōϐХл ʳtʲʣʳʳл corresponds to position of stress in archaic Sogdian before operation of the Sogdian Rhythmic Law. Under the influence of stress many Iranian vowels were changed in unstressed positions: ŋД and ŋ were shortened to i, u; also short vowels (or even all syllables) were lost when preceding a stressed syllable. Compared to Sogdian in YaghnōϐХлtʦʣʲʣлtoo₎лpὕaϑʣлaлchain shift of ā, , > , , ʉ/ , (Middle) Iranian ŋā changes to Sogdian ē under i-Umlaut, in YaghnōϐХлtʦʣʲʣлʧʳл ‖л ‘ . Consonants do not differ much from Sogdian, major difference may be ŋ , ŋ > v, d; transition of , x, ⁾° from velars to uvulars; quite recent is a development of *ϑ > s ‖лt38. Unlike Sogdian there is no change *ϑr, * r > , ž, in YaghnōϐХчлtʦʣʲʣлʧʳл ʲʣʥuὕaʲ лdʣvʣὕopmʣntлto s(ⱽ)› ‖лt(ⱽ)›, d(ⱽ)›; YaghnōϐХл mp, nt, nk, nč respond to Sogdian ṁb, ṁd, ṁg, ṁǰ; and perhaps (*Proto-)Sogdian * d, * d, * ₎d > YaghnōϐХлxt, ft/vd ‖лft, st ‖лzd. (KHROMOV 1987, 653-661) consonants stops bilabials pb affricates td čǰ alveopalatals fv sz šž uvulars labiouvulars pharyngeals glottals kg q vowels Х ū ⁱi nrl u ē ( ) ō y у ф palatals velars sonorants mw labiodentals alveolars fricatives x ʸ ( ) (ᴥ) h a уāф Table 12 Sound system of YaghnōϐХ. 37 From now on I will distinguish different forms in the Eastern and Western dialect by double vertical line: i.e. {Eastern dialect} ‖лʻWestern dialect}. The Transitional dialect stands between the Western and the Eastern one – some of its features correspond with the Western dialect, some other with the Eastern (for more information on the Yagh KHROMOV 1972, 97-105; NἴσÁἰ 2010, 243-246). At the present time the majority of speakers use the Western dialect, its speakers settled also areas in the Ghōnchîлandлςppʣʲлσaʲzōϐлdʧʳtʲʧϑtʳщ 38 Before the year 1913 there was still ϑ in YaghnōϐХлуJUNKER 1930, 126, 128-129). See chapter II.1.3.10. ·25· YaghnōϐХ nouns have two numbers and two cases (direct and oblique), the distinction of gender has been lost39. Plural is formed with the ending -t (in words ending in -a the final vowel was prolonged before the plural ending: -a+t > *-āt > - t) < *-tā-; oblique case ending originates in Iranian a-stem genitive singular: ŋ-h ‘ > -i (after vowels - , if a word ends in -a, this -a is palatalized: -a+i > - ‖л -‘ ). Adjectives are indeclinable; they have neither case nor gender. Personal pronouns have forms for first two persons, for the third person demonstrative pronouns are used. Personal pronoun of the second person singular and demonstratives of both numbers are declined in two cases40; demonstratives distinguish double deixis. Verbs have two stems – present and imperfect, there is a similar pattern also for participles – i.e. present and past participles. The present stem comes from Old Iranian present stems; the imperfect stem is formed from the present stem with addition of augment a-. Personal endings of the present tense correspond to Old Iranian primary endings (but the ending of the third person plural was replaced by original perfect ending), imperfect endings come from Iranian optative and imperfect endings. By adding a suffix - t to personal endings was originally formed durative of verbs, later this old durative was reanalyzed: in present the durative ending serves as new present, the old present than changed its function as a dependent verb; durative of imperfect was reanalyzed as preterite. Perfect tense is derived from the Iranian past participle. Perfect is connected with split ergativity: perfect of intransitional verbs is formed from the past participle and copula, transitional verbs have subject in oblique followed by copula of the third person singular. Forms of progressive (durative) present and perfect are formed from the infinitive, these forms are also influenced by the ergative (formed analogically as in the perfect tense). (KHROMOV 1987, 662-694) (excursion 2) Yaghn bī dialects There are recognised two common YaghnōϐХлdʧaὕʣϑtʳл– Eastern and Western YaghnōϐХщлAὕ ϐʣʲtл Leonidovich Khromov recognises also third, Transitional, dialect which shares some features of Eastern YaghnōϐХл andл some other of the Western variety. I will not describe the differences between the dialects as this issue has been described well in Khʲomov ʳл ḲaghnōϐХл δʲammaʲл (KhROMOV 1972, 97-105), an outline of YaghnōϐХлdʧaὕʣϑtʳ with a short dialectal word-list is also presented in the grammatical appendix of the YaghnōϐХ-Czech dictionary (NἴσÁἰ 2010, 243246). In many works that mention YaghnōϐХл dʧaὕʣϑtʳл tʦʣʲʣл aʲʣл observed basic differences of development of historical *ϑ (and *ϑr-) and i-Umlauted ŋā, i.e. development such as ŋmá ϑa- > mēs ‖л mēt day ;л *ϑ ‘- > s‘›á₍ ‖л tⁱ›á₍ tʦʲʣʣ л andл ŋu ‘- > w ‖л ⁽‘ ʥʲaʳʳ щ Less often πomʣл ʤʣmʧnʧnʣл ʤoʲmʳл ʷʣʲʣл ʧntʲoduϑʣdл vʧaл ρā₍Х₎л ʤʲomл Aʲaϐʧϑл oʲл ʤʲomл οuʳʳʧanл уϑʤщл ϑoὕὕoquʧaὕ л maall má, teacheress; u₎’éčk‘, Uzbek woman). 40 Robert Gauthiot provides direct case of the first person singular az. Such form is not mentioned in other works on YaghnōϐХчлtʦʣʲʣлʧʳл₍uʳtлʳʧnʥὕʣлʤoʲmлman for both cases (originally man < *mana is oblique (< genitive) of * az < ŋá₎u < ŋá₎‘m < *aʣám; cf. GAUTHIOT – BENVENISTE 1929, 108-109). 39 ·26· differences in verbal endings are given, e.g. for present indicative of the third person singular -č ‖л -t t. All the above mentioned examples are distinct in contemporary YaghnōϐХл dialects, but they are not as important from diachronic point of view (see e.g. BIELMEIER 1989, 487; VINOGRADOVA 2000b, 309-310; JUNKER 1930, 123-131; BOGOLYUBOV 1966, 359 etc.). What is more interesting than the above mentioned isoglosses s ‖лt, ‖л‘ and -č ‖л-t t is imperfect and simple preterite ending of the first person plural -Дm( t) ‖л - m( t) – Eastern YaghnōϐХл -Дm is derived from optative ŋ-‘ ma (KHROMOV 1987, 681) 41 , but the Western YaghnōϐХл ʣndʧnʥл - m continues from imperfect *-ām‘. This feature was unfortunately left unnoticed by majority of scholars. The two different sets of YaghnōϐХл ʧmpʣʲʤʣϑtъʳʧmpὕʣл preterite endings of the first person plural show deeper history of the language, even deeper than the other commonly presented dialectal differences. In this case Eastern YaghnōϐХлʳʦaʲʣʳл innovation with Sogdian while Western YaghnōϐХл уʷʦʧϑʦл ʳʦouὕdл ϐʣл ʥʣoʥʲapʦʧϑaὕὕyл ϑὕoʳʣʲл toл literary Sogdian) preserves archaic Iranian imperfect. This observation may be another clue that proves that YaghnōϐХлʷaʳлnotлdʧaὕʣϑtлoʤлπoʥdʧanлϐutлπoʥdʧanлandлḲaghnōϐХлʳpὕʧtлmuϑʦлʣaʲὕʧʣʲщ (excursion 3) Sogdo-Yaghn bī substrate in the Zarafshān-Tajik dialects It is not exactly known when the territory of present Tajikistan underwent language shift in favour of Persian; it can be supposed that Persian gained its prestigious position during reign of tʦʣлπāmānʧdлdynaʳty (819-999). Sogdian was then gradually displaced by Persian, but its dialects survived several centuries in mountainous regions on upper reaches of the Zarafshōnл ʲiver. Nowadays Tajik is spoken in these regions, respectively its Central (of Zarafshānф dialects (RASTORGUEVA 1964). ZarafshānлTajik can be split into three (sub)dialect groups – dialects of historical regions of Mastchōh (cf. KHROMOV 1962), Falghar (cf. KHROMOV 1967; KERIMOVA 1963) and Fōn (RASTORGUEVA 1964, 8; the last two mentioned regions form together with the Yaghnōϐ Valley present Aynî district, the first mentioned region forms present district Kuhistōni MastchōʦфщлSubstrate words from a Sogdian dialect survived in these dialects. Sogdian substrate in Zarafshānлdʧaὕʣϑtʳлϑanлϐʣлoϐʳʣʲvʣdлʧnлphonology, lexicon and in toponymy. In phonology the Zarafshānл dʧaὕʣϑtʳл ʳʦaʲʣл ʳʧmʧὕaʲл ʤʣatuʲʣʳл ʷʧtʦл YaghnōϐХчл maʧnὕyл ʧnл aл change of vowels initiated by labialization of ŋā and subsequent chain-shift of ŋ and (Figure 5). In the Zarafshānл dʧaὕʣϑtʳл aʳл ʧnл λoʲtʦʣʲnл Tajik merged ŋД, *i > i and ŋ , *u > u probably before the chain-shift, but this feature is not observed in YaghnōϐХл уdʣvʣὕopmʣntл ʧnл YaghnōϐХл ʧʳл aл ₎ʧndл oʤл ϑompʲomʧʳʣл ϐʣtʷʣʣnл tʦʣл ʳϑʦʣmʣʳл (a) and (b) at Figure 5, the development *u, ŋ , , *i > [ члuу”фчлy”члɪ]) differs. Substrate consonantism generally does not differ from Tajik, Zarafshānлdʧaὕʣϑtʳлmoʳtὕyлʲʣtaʧnлϑὕuʳtʣʲʳлmb, nd, ng, nǰ, in YaghnōϐХлtʦʣʲʣлʧʳл mp, nt, nk, nč instead. 41 And is directly related to Sogdian ending -ēm. ·27· Classical Persian Zarafshān dialects of Tajik (a) (b) Figure 5 Chain-shift of back vowels in the Zarafshānл dʧaὕʣϑtʳл oʤл Tajik: (a) dialects on the right bank of Lower 42 Mastchōʦ, several dialects of Upper Mastchōʦл and majority of Falghar dialects (including Tajik dialect of the Yaghnōϐ Valley), (b) majority of Upper Mastchōʦ dialects, dialects on the left bank of Lower and several Upper Falghar dialects; dashed arrow represents conditioned change (IαŌ 2009, 68). The Sogdian substrate can be recognised in lexicon – problem of Sogdian loan-words in Persian was solved by Walter Bruno HENNING (1939). The list of Sogdian and YaghnōϐХлʷoʲdʳл in the Zarafshānл dialects and in Tajik ʷaʳл ʳtudʧʣdл ϐyл Aὕ ϐʣʲtл ἱʣonʧdovʧch KHROMOV (1962; 1388). In the Zarafshānлdialects there are 74 words of Eastern Iranian origin – nine of them are of Sogdian origin without attested responses in YaghnōϐХ; 16 words are attested both in Sogdian and YaghnōϐХ; 28 are attested only in YaghnōϐХ and other 21 words are of Eastern Iranian origin, but their Sogdian and/or YaghnōϐХлʳouʲϑʣлcannot be found. Another important source for the study of Sogdo-YaghnōϐХ substrate is toponymy, from Sogdian sources there are known some place-names of North-Western Tajikistan that are used even today e.g. Anz b (Sogd. ʾnzʾ h), Isk dár (Sogd. ʾskʾtr), Farmētán (Sogd. ⃝ prnmy n ), Fal ár (Sogd. pr rh), Madm (Sogd. m mh), Dar (Sogd. r h, Yagh. ⃝ Dar ), Rarz (Sogd. rzrh), Falm t (Sogd. ʾ tmʾwt , Yagh. , TFalgh. F‘lm ǘt), ⃝ Xu ēkát (Sogd. (ʾ) sykn h, ʾ sykt , TFalgh. X ēkát), Mard kát (Sogd m›t kt-; TMast. Ḥ‘›d kát, Mardu kát; today generally called Mastchōh), Zar vátk (Sogd zrʾw kh), Varz(-i ⃝ Mⁱn r) (Sogd rz-; present Aynî), V dí” ⁽ʾt₍ ); other toponyms are known also ⃝ 43 rm ), Varz b (Sogd. rz- + ʾʾp(h)) etc. (cf. from neighbouring areas: Γarm (Sogd. KHROMOV 1966; BOGOLYUBOV – SMIRNOVA 1963, 101-108; SMIRNOVA 1963; BUSHKOV – NOVIKOV 1992; LUR€E 2004; NἴσÁἰ [in print], NἴσÁἰ 2009). 42 In majority of the Upper Falghar dialects (with an exception in dialect of Rarz) and in some Lower Mastchōʦл dialect of right bank of the river Zarafshōnлtʦʣʲʣлψʏ] later changed to [ɪ] (KHROMOV 1962; KHROMOV 1967b). In the presented thesis the Zarafshānлρā₍Х₎лvoʷʣὕʳлψ ] and [ʏ] will be transcribed as u, . 43 It is either city of Gharm in RashtлdʧʳtʲʧϑtлʧnлἶaʲōtʣʥХnчлoʲлʧtлϑouὕdлϐʣлvʧὕὕaʥʣлoʤлGhaʲmēnлʧnлḲaghnōϐлуBUSHKOV – NOVIKOV 1992). ·28· On the basis of the substrate in the Zarafshānл dʧaὕʣϑtʳл it can be assumed that the local dialect originated from the same basis as Sogdian and YaghnōϐХ – this hypothetic language (dialect) can be called *ZarafshānХ. It is possible that ZarafshānХлϑould originate in a dialect (?) attested in documents from fortress of ʧʳōʲa₎лʧn by Mardūshkat in Mastchōʦ (cf. LUR€E 2011; 2012, 455-456). I.1.1.4.2. Ossetic Ossetic (Ossetian) needs to be understood as two varieties of one language – Iron уʧʲonлǦvzaʥ, ironauл‖лХʲonлǦvzaʥчлХʲonau; in older works also Tagaur – Northern Iron and Twal – Southern Iron) and Digoron (Digor; d ʥuʲon Ǧvzaʥч d ʥuʲonau ‖л dʧʥoʲonл Ǧvzaʥ, digoronau) 44. Iron is official language in North Ossetia-Alania and South Ossetia (formerly autonomous region of Georgia), Digoron is spoken in western parts of North and South Ossetia. Iron is considered as a literal form of Ossetic, total number of speakers of Ossetic vernaculars is estimated to 5яэ˙000 people (ISAEV 1987, 539). Both dialects are historically close one to the other, but due to sound changes that started in Iron approximately two hundred years ago both languages are intelligible with difficulties (THORDARSON 1989, 457); to these two dialects also a transitional dialect of τǦὕὕaʥ₎omл can be added (ISAEV 1966, 101-111). The oldest book written in Ossetic was a translation of catechism by Gay Takaov in the year 1798, the language was written in old (i.e. Church Slavic) variety of the Cyrillic alphabet, in the past Ossetic was written in various modifications of Cyrillic, Georgian alphabets Khutsuri and Mkhedruli or in modified Latin alphabet (THORDARSON 1989, 457-459); Digoron speaking Muslims also used the Arabic script. Modern Ossetic nowadays uses the Cyrillic alphabet extended by a letter æ and nine digraphs (in Digoron there is also digraph iy ʤoʲлъХълandлalso a letter h may be used). consonants bilabials stops affricates p p b t t d postalveolars ϑ čϑǰ f v c s z labiovelars uvulars labiouvulars k ₎ g ₎{ ₎{ ʥ{ q q{ vowels i u nrl o e y у ф palatals velars sonorants mw labiodentals alveolars fricatives Ǧ a x ʸ{ { Table 13 Sound system of Iron Ossetic. In this work Ossetic words will be marked in three ways – words that are the same in form will marked just as ἴʳʳщ чл ʷʦʣnл tʦʣʲʣл aʲʣл dʧfferent forms in the Iron and Digoron dialects, those forms will be separated by double vertical line: {Iron} ‖л{Digoron}. If a word exists only in one Ossetic dialect, it will be marked by a small capital letter: 44 ·29· Ossetic is a direct descendent of Alanic, which originates in Scytho-Sarmatian dialects. Though the origins of Ossetic can be traced to the 7 th linguistic data concerning its ancestor(s) – the problem lays mainly in an insufficient graphical system in which the old Scytho-Sarmatian languages were recorded and also in a fragmentariness of data which do not provide us with much information concerning morphology and syntax. consonants stops bilabials ppb affricates ttd sonorants mw labiodentals alveolars fricatives ϑϑ fv sz vowels Х i nrl e alveopalatals y у ф palatals velars ₎₎g o Ǧ a labiovelars uvulars u q x labiouvulars glottals (h) Table 14 Sound system of Digoron Ossetic. Vocalic system of Ossetic, mainly of its Digoron dialect, is rather archaic – reduction of unstressed vowels in Alano-Ossetic dialects did not occur to such extent as it is known in other Eastern Iranian languages. The development of vowels was as follows: *a > æ, a, o; ŋā > a, o; > ( ) ‖лu; ŋ› > æ›, ar; palatalized ŋ‘( ) > i ‖лД, e; *u‘ ŋ (a) > i ‖лe; ŋ u(a) > u ‖лo; > ‖ i; after a velar or uvular > o, æ ‖лwa, ⁽æ; and e from contraction: -æ + æ- or -æ + i- ‖ ye-. In AlanoOssetic, the quantity of high vowels was lost: *i, ŋД and *u, ŋ developed to i and u in Digoron, in Iron they all merged into . Qualitative changes can be observed for low vowels *a and ŋā, in this case quantitative difference was replaced by difference in quality: *a = æ /ȳ/, ŋā = a /ȴъчл*a in front of two tautosyllabic consonants merges with /ȴъл andл tʦʧʳл nʣʷ л ъȴъл ὕatʣʲл ϑʦanʥʣdл toл o when followed by a nasal. Consonant system continues from Alanic without major changes, but it has been enriched by contact with Caucasian languages, so in Ossetic there are also glottalized consonants p, t, k, c, and in Iron also č. Ossetic innovation when compared to Alanic is the switch ŋ , ŋž, ŋč, ŋč, ŋǰ > s, z, c, c, ʒ45. Velars and uvulars were labialized in front of old o and u (Iron u, ): k, k, g, q, x, > k , k , g , ‹ , ⁾ , . Iron differs from Digoron in two Development of s, z, c, c, ʒ continues also recently, in (Northern) Iron they are realized as [ʃ, , s, l€, z]; in Digoron they remain as [s, z, l, l€, i] when followed by back vowels (i.e. æ, a, o, u), before front vowels e and they are palatalized: [ʃ, , m, m€, j]. Different development can be observed in some southern dialects of Iron: sibilants and ŋč develop the same way as in northern Iron, palatal affricates probably retained their pronunciation th until half of the 19 century, nowadays pronunciation of ŋč, ŋǰ remained when geminated or when following n, in all other positions they changed to palatal sibilants: ŋč, ŋǰ şлъšчлžъл×л * čč, * ǰǰ, * nč, * nǰ (THORDARSON 1989, 457). 45 ·30· fundamental changes: change of word-initial ( )- > ‹( )- and affrication of palatal velars before front vowels e and : ḱ, ḱ, ǵ > č, č, ǰ (in southern Iron dialects > c, c, ʒ, in Digoron they remain ḱ, ḱ, ǵ). It should be noted that labialization and palatalization preceded change * i, * u > 46 (ISAEV 1987, 552-580; THORDARSON 1989, 459-466). Bilabial approximants wлψ л~ ] and y [j ~лʧ] are non-phonemic and often form falling or rising diphthongs. ive, genitive, dative, allative, ablative, inessive, superessive (/adessive), elative (/equative) and comitative (the last mentioned case is not present in Digoron), it has two numbers (singular and plural) and does not distinguish gender. Ossetic is by the number of cases comparable to Old Iranian, nevertheless Ossetic cases do not respond to the Old Iranian cases functionally; only endings of four cases – nominative, genitive, ablative and inessive (< locative) are considered to be inherited from Old Iranian. All the other case endings newly emerged from prepositions, adverbs or due to contact with languages of Caucasus (BELYAEV 2010). There is also an opinion that Ossetic originally possessed only two inherited Old Iranian cases: nominative and genitive (> oblique) and the other cases are an innovation due to contact with Caucasian languages (KIM 2003; 2007). Ossetic verbal morphology is quite conservative, it preserves most of Old Iranian verbal moods, an innovation is shift of past tenses into single past tense – preterite, also the forms of future tense are new. Conservativism can be observed clearly also in personal endings which are in many cases inherited (THORDARSON 1989, 473-477; ISAEV 1987, 664-632). There are distinguished transitional and intransitional verbs, transitivity is expressed morphologically in preterite – to a past stem (formed originally from *-ta- past participles) are added personal endings, for transitional verbs formed from copula, for intransitional verbs formed from verb to have (ISAEV 1987, 619). It is evident that the preterite endings confirm ergative construction which have been lost in modern Ossetic, but it has just preserved its trace in two sets of the preterite personal endings47. For Ossetic is characteristic the use of preverbs – calque from the 46 Velars were probably palatalized quite recently, some 150 or 200 years ago. In the first book printed in Ossetic there are no marks of palatalization in orthography (but see notation of palatalized and non-palatalized velars in the Romance languages), either the change i, u > has not taken place although the book was written in the Iron th dialect (KOZYREVA 1974, 64). The issue of Ossetic phonology at the end of the 18 century is complicated – Tamara Zaurbekovna Kozyreva in her analysis of Ossetic Catechism does not deal with phonology and notes that the analysis needs a separate study (ibid.: 14). Palatalization of velars had to be completed before the year 1844, ʷʦʣnлʦadлAndʣaʳлηoʦanлπ₍ögren published the first grammar of Ossetic (SηÖδοβλ 1844). The solution perhaps may be found in translations of religious texts to Southern Ossetic (written in the Khutsuri alphabet), which were th published in the early 19 century by Ivane Yalghuzidze (THORDARSON 1989, 458), unfortunately I have not seen those sources. The clue for the issue of velar palatalization can be found in different results of palatalization in the Southern and Northern Iron dialects, or possibly in the development of the transitional Digoron-Iron dialect of τǦὕὕaʥ₎omл– aϑϑoʲdʧnʥлtoлσʳʣvoὕodлγëdoʲovʧch Miller the velars were seldom palatalized before the year 1880, but before the year 1957 palatalization was fully implemented (ISAEV 1966, 106-107). 47 The comparation of ergative with Ossetic inflectional system could be interesting – tʦʣʲʣлaʲʣлmanyл nʣʷ лϑaʳʣʳл formed due to contact with Caucasian languages but it has not preserved or borrowed ergative as a separate case, by ·31· Caucasian languages, but morphologically formed from Iranian sources; preverbs have two functions – locative and modal (THORDARSON 1989, 475; ISAEV 1987, 612-616). I.1.1.4.b. The Pāmīr languages ρʦʣлἵāmХʲлὕanʥuaʥʣʳл(or BadakhshānД languages) form a significant group within the Southern branch of the Eastern Iranian languages48. ρʦʣлἵāmХʲлὕanʥuaʥʣʳлϑanлϐʣлdivided into two groups: Northern ἵāmХʲХл уoʲл ShughnХ-YazghuὕāmХ фл ʥʲoupл andл πoutʦʣʲnл ἵāmХʲХл ʥʲoup. To the Southern group belong WakhХлandлζsh₎āshmХ-πanʥὕēchХчлall the otʦʣʲлἵāmХʲлὕanʥuaʥʣʳлϐʣὕonʥл to the Northern group49. Formerly it was supposed that the languages come from a * ProtoἵāmХʲХ proto-language (cf. PAKHALINA 1983), nowadays it seems that sources for these languages vary, maybe the languages of the ShughnХ-YazghuὕāmХл ʥʲoupл mayл ʦavʣл aл ϑommonл ancestor (cf. ÈDEL MAN – DODYKHUDOEVA 2009, 773; PAYNE 1989, 420-423; SOKOLOVA 1967; SOKOLOVA 1973). We do not have much information about the уpʲʣфʦʧʳtoʲyл oʤл tʦʣл ἵāmХʲ-εʧndū₎ush area before the Middle Agesчл ϐutл ʧtл ʳʣʣmʳл tʦatл ἵāmХʲ was settled by Iranian speaking people in several waves. We do not know from where the Iranian-speakinʥлἵāmХʲʧanʳ came, there may be a clue only for WakhХл ʷʦʧϑʦл ʳʦaʲʣʳл ʳomʣл isoglosses with the Saka dialects. κaʲtʧnл ἰümmʣὕл suggests that (Old) WakhХлʷaʳлoriginally a Western Saka dialect (KÜκκβἱ 2008, 1) – nowadays WakhХлϑʣʲtaʧnὕyл belongs toлtʦʣлἵāmХʲлʥʲoupчл a study of the WakhХлmatʣʲʧaὕлʳʦoʷʳлtʦatлtʦʣʲʣл may be two (or even more) language layers50. It can be supposed that a Saka-WakhХ language contrast, it completely dropped it, despite the fact that ergative is present in languages such as Georgian or Svan (BELYAEV 2010, 309-310). 48 The most widely accepted classification of the Eastern Iranian languages divides those languages into two branches – Northern and Southern. I have not found any exact criteria by which both branches are defined. It can be assumed that the inner development especially in the Southern branch could have been much more difficult. It seems that the Eastern Iranian languages should be reclassified. They can be newly divided into five branches: I Northern (or Scythian group; to this group belong Sogdian, Scytho-Sarmatian dialects, Ossetic and Yaghn ’Д), II North-eastern (or Saka; Saka dialects, maybe also WakhД), III CʣntʲaὕлуoʲлἵāmХʲ;лYazghulāmДж ShughnД- shānД g›oupж ḤunjД-Yidghā, WakhД, IshkāshmД-S‘nglēchД), VI Southern (or Paṭʦān;лPasht ‘nd W‘ṇetsД; maybe ḤunjД-Yidghā and SarghulāmД can belong to this group) and V South-eastern ( ›muṛД ‘nd P‘›āchД). Questionable is a position of Bactrian (member of the Paṭʦānлʥʲoupлoʲлκun₍Х-YidghāлἵāmХʲлʳuϐʥʲoupл ??) and Khʷāʲʣzmʧanлʷʧtʦʧnлtʦʣл aϐovʣл mentioned groups. The proposed classification is based mainly on contemporary (often geographically conditioned) proximity of the languages. Such classification needs to be based on more thorough study of isoglosses within all members of the Eastern Iranian group, some criteria will be shown later in this thesis. 49 ρʦʣлpoʳʧtʧonлoʤлκun₍ХлandлḲʧdghāлʷʧtʦʧnлtʦʣлἵāmХʲлʥʲoupлmayлϐʣлquʣʳtʧonaϐὕʣчлtʦʣʲʣлaʲʣлʳomʣлautʦoʲʧties who do not recognise them ἵāmХʲлὕanguages and link them with Pashtōлandлτa etsХщлκoʲʣлϑompὕʧϑatʣdлʧʳлthe position of SarghuὕāmХщлζлʷʧὕὕлtʲʣatлtʦʣmлaὕὕлaʳлmʣmϐʣʲʳлoʤлtʦʣлἵāmХʲлʥʲoupлʧnлtʦʧʳлʷoʲ₎щ 50 They can be observed mainly in different development of intervocalic voiceless consonants – in some cases they remain voiceless, but in some other instances they were voiced. There are even some examples of roots with forms with both voiced and voiceless responses in WakhХщ ·32· was ἵāmХʲʧzʣd , i.e. overlaid ϐyл aл ἵāmХʲл ʳupʣʲʳtʲatʣ 51 . It is quite difficult to determine the development of the ἵāmХʲл ὕanʥuaʥʣʳ. As I have mentioned above, there is no reason to reconstruct a *Proto-ἵāmХʲХлὕanʥuaʥʣчлʷʦʣnлaлpʲoto-ὕanʥuaʥʣлoʤлtʦʣлἵāmХʲлaʲʣaлʧʳлnʣʣdʣdлtʦen it should be reconstructed just for the ShughnХ-YazghuὕāmХлὕanʥuaʥʣʳщлAὕʳoлκun₍Х-Yidghāлуandл SarghuὕāmХл ??) probably belonged to this group, but they probably split earlier 52 . The Ish₎āshmХ-πanʥὕēchХлὕanʥuaʥʣʳ are quite close toлtʦʣлλoʲtʦʣʲnлἵāmХʲлὕanʥuaʥʣʳчлϐutлtʦʣyлdʧffer in some aspects, some authors even suppose that Ish₎āshmХ-πanʥὕēchХл dʧffer more from YazghuὕāmХлandлShughnХ-οōshānХлtʦanлdoʣʳлκun₍ХлandлḲʧdghāлуÈDEL MAN – DODYKHUDOEVA 2009, 773, 775-777; PAYNE 1989, 420-423; SOKOLOVA 1973). Genetic affiὕʧatʧonл oʤл tʦʣл ἵāmХʲл languages is thus problematic. To explain similarities between the individual languages of the area we can postulate ἵāmХʲл linguistic area (Sprachbund), ʧщʣщл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл of Badakhshānл уʣʸϑὕudʧnʥл κun₍Хл andл Yidghā). Tʦʣл ἵāmХʲл ὕʧnʥuʧʳtʧϑл aʲʣaл tʦen belongs to a wider linguistic area of tʦʣл ἵāmХʲεʧndū₎ush region that includes all the ἵāmХʲл ὕanʥuaʥʣʳл уʷʧtʦл κun₍Хл andл Ḳʧdghāфл andл tʦʣл αaʲdʧϑлandлλūʲʧʳtānХлὕanʥuaʥʣʳщлρʦʣʲʣлϑanлϐʣлʣvʣnл a wider linguistic area – Central Asian or εʧmāὕayanлπpʲaϑʦϐundлtʦatлʧnϑὕudʣʳлtʦʣлὕanʥuaʥʣʳлoʤлtʦʣлἵāmХʲ-εʧndū₎ush Sprachbund, other Iranian languages (i.e. Pashtō, Wa etsХчлἵaʲāchХчлŌʲmuṛХчлBaὕōchХ), some Indo-Aryan languages ( ōmā₎Х, Western PahāṛХчл ἵan₍āϐХл andл mayϐʣл ἱaʦndāл andл πʧndʦХфчл ʳomʣл πʧno-Tibetan languages (BaltХ, ἱadākhХчл τʣʳtл εʧmāὕayish languagesфчл αʲavʧdanл BʲaʦūХл and the language isolate BuʲūshaskХ (PAYNE 1989, 422-423). Some place names ʧnл ἵāmХʲл show probable non-Iranian origin, according to ρat yana 53 Nikolaevna Pakhalina the name of Ish₎āshim should be Indo-Aryan and Yazghuὕāmл andл Sarghuὕām probably contain a non-Indo-Iranian continuant of Ide. *dʰégʰ m ~ ŋdʰgʰém- earth 54 (PAKHALINA 1976b). 51 Ivan MikhaЧὕovʧčл STEBLIN-KAMENSKIY (1976) sees some pre-WakhХлtʲaϑʣʳлʧnл toponymy of Western part of the WakhХ-speaking territory: Khandūtл уτakh. ənd t < Ir. *x an-dāt‘-, given by the Sun; Tjk. Xand ) and λamatʥūtл уτakh. Nəmətg t < Ir. *namata-gāt-йgāϑ-, place of prayer/adoration; Tjk. Namatg , earlier also ḥ‘m ₎gáh which ʧʳлρā₍Х₎лϑaὕquʣлoʤлtʦʣлτakhХлnamʣ). «It s poss ’l“ th‘t th“ n‘m“s Kh‘nd t ‘nd ḥ‘m‘tg t o› g n‘t“ in some [unknown] Eastern Iranian dialect that was close or even identical with an ancestor of the contemporary WakhД language and they [i.e. the place-names] were formed in a period when Old Iranian form- and word-formation models ⁽“›“ st ll p›“s“›v“d.» (STEBLIN-KAMENSKIY 1976, 185) 52 ζtлʧʳлʣvʣnлpoʳʳʧϐὕʣлtʦatлanлanϑʣʳtoʲлoʤлκun₍ХлandлḲʧdghāлʷaʳлaл ἵāmХʲʧzʣd лdʧaὕʣϑtлʳʧmʧὕaʲлtoлBaϑtʲʧanщ 53 The name of Ish₎āshim (originally name of a territory, later on also name of the cites of Esh₎āshem and Nut): Ishk. (ь)ko ьm, Pers. I kā ím, Tjk. I k ím, AfghP. E kā ém; Wakh. (ə)k um; Shugh. k um) has probably derived from Indo-Aryan ŋś‘k -sam - (sic! PAKHALINA 1976b, 178; probably ŋś‘k - am -фл ἱandл oʤл tʦʣл πa₎a чл ϑʤщл Ved. ŋś‘k -k am- (PAKHALINA 1976b, 178-179). Probable etymon for (ь)ko ьm/I kā ím should be (Old) Indo-Aryan ŋś‘k -k am - with loss of *k in *k am - as a result of dissimilation: ŋś‘k -k am - > ŋś‘k am - > *Proto-Ishk. ŋ (ă)k (ă)m- > Ishk. (ь)ko ьm, (ь)ko mí; Pers. I kā ím, kā ( )m (WakhХл andл ShughnХл ʤoʲmʳл aʲʣл ὕoanʳл ʤʲomл Ish₎āshmХлoʲлἵʣʲʳʧanфщ 54 ρat yana Nikolaevna Pakhalina sees development of Ide. *dʰgʰém (IIr. ŋǵđʰ m-; Ir. *ʣam-; Ave. zam-; Pers. zam ; Ved. k am-; Gre. ώ and adv. , on the earth; Lat. humus; Hit. tēk‘n; TokhA. tkaṃ; TokhB. kaṃ; ·33· ρʦʣл ἵāmХʲл ὕanʥuaʥʣʳл ʳʦaʲʣл manyл ʳʧmʧὕaʲл ʤʣatuʲʣʳл ʧnл pʦonoὕoʥyл andл moʲpʦoὕoʥyщл σoϑaὕʧϑл similarities can be seen in operation of i- and ā-Umlaut. Almost in all of the ἵāmХʲлlanguages there were secondary palatalized tectals prior to front vowels (including * and often *r), also postalveolar fricatives were depalatalized in almost all of the languages. Palatal sibilants tended to change to retroflex sounds or even to velar fricatives. Intervocalic voicing of voiceless stops and sibilants appeared in all languages, except WakhХлʷʦʣʲʣлtʦʧʳлʤʣatuʲʣлappʣaʲʳлpaʲtὕyчлpʲoϐaϐὕyл due to influence of substrate or adstrate. In morphology we can see also many common features – gradual reduction of cases into two case system (but this development historically differs from one language (subgroup) to another) and its replacement with adpositional constructions, OCS. ₎“mlj‘; Lith. žẽmė) in several responses that exclude Iranian development: Yazghuὕāmл ϑanл ϐʣл tʲanʳὕatʣdл aʳл ἱandлoʤлtʦʣл*Asi pʣopὕʣ щлρʦʣлʲootл* ʦ - (??) can be compared with the name of the Ossetians or Jassians or the Ἀ , Ἀ ῖ (PTOLEMY, Geography V 9:16) or Ἄ (STRABO, Geography XI, 8:2) or with Ave. asu-, fast (ABAEV 1958, 79). The Ide. root *dʰgʰém changed to * g đ ʰ m- > *gd m- > *g/ ᵊ m. This *g/ ᵊ m was borrowed as * in Persian and besides Yaz it also appears in the name Sar уʧщʣщл хςppʣʲл ἱand ;л ϑʤщл aὕʳoл SarghuὕāmХл dʣvʣὕopmʣntл *d > > l – so the Persian form was probably borrowed from SarghuὕāmХл oʲл anotʦʣʲл related but otherwise unknown language). YazghuὕāmХлnamʣлoʤлtʦʣлḲazghuὕāmлσaὕὕʣyлʧʳлY ₎dom, its origin is the m (Tjk. Yazgul m, in Southern dialects Yazg lóm) < *Ἄ /Ἀ ῖ /Ἀ -* g đ ʰ m- > same as of Pers. Yaz * ás-(g)d m- > Y ₎dom. There lays also the origin of the name of the Yazghuὕāmл ʲiver, Yazgh. Z(ə) ǵ ‘m“nǰ < * as-(d)ǵ m na- - (also Yazgh. z(ə) ǵ ‘m ǵ, a person from Yazghuὕāmл <л * as-(d)ǵ m -ḱ -); probably there were lately two (or even three) continuants of * g đ ʰ m- in the Yazghuὕām-Sarghuὕāmл aʲʣaśл *(g)d m- and *(d) ǵ m-/*g(d) m-. (cf. PAKHALINA 1976b, 179-181) A variety of *(d)g m-/*g(d) m- < Ide. *dʰgʰém appʣaʲʳл aὕʳoл ʧnл ʳʣvʣʲaὕл уαaʲdʧϑŠфл toponymʳл ʧnл εʧndū₎ush: Ṣiṇe-gam ἱandл oʤл tʦʣл i ā чл Kala -gum ἱandл oʤл tʦʣл ἰaὕā a чл V“› -gum < ŋV“› k-gum ἱandл oʤл tʦʣл σʣʲchik (=B(u)ʲūshōфл pʣopὕʣ л уPAKHALINA 1976b, 179фчл ϐutл κaʲtʧnл ἰümmʣὕл ϑonnʣϑtʳл ga/um with Skr. g›ām‘-, troop > village (KÜκκβἱ, pers. comm.). In zero-grades Ide. *dʰgʰ m - appears as k m-, gm-, jm- in Vedic, see declination of Ved. k am- (f): sg. nom. k ās, gen.-abl. gmas / jmas / k mas, dat. k e, acc. k am, loc. jman / k ám , instr. jmā; du. nom. k mā; pl. nom. k māйД)s, acc. k ās, loc. k su (MAYRHOFER 1992, 424-425; MONIER-WILLIAMS 1964, 326); cf. Avestan zam-: sg. nom. ₎ , gen. zəm , acc. ₎ąm, loc. zəmē й ₎əm ; du. nom. ₎ ; pl. acc. zəmas, voc. zəm (BARTHOLOMAE 1961, 1662-1665). For the Iranian languages there is no attested zero-grade †gm- as in Vedic, according to Avestan there had to be Iranian zero-grade *ʣm-. For Dardic we can suppose zero-grade (or a reduced form) *g(V)m-. These examples do not explain origin of [Y ₎]dom, [Yaz] ul and [Z(ə)] ǵ ‘m[“nǰ]/[z(ə)] ǵ ‘m[ ǵ]. Is it a form of an otherwise unknown centum (?) Indo-European language (*Eteo-ἵāmХʲХл??) that was different from *Proto-Tokharian (: *tkam-). The *ʣm- and *g(V)m- roots can be compared with Greek χα αί у×лχϑώ ). The above mentioned examples are an extension to proposal given by ρat yana Nikolaevna PAKHALINA (1976b) ʧnлaлʳʦoʲtлʳtudyлonлἵāmХʲлtoponymyщлρʦʧʳлʧʳʳuʣлʧʳлʳtʧὕὕлopʣnʣdлʤoʲлʤuʲtʦʣʲлdʧʳϑuʳʳʧonчлϐutлʧtлʳʣʣmʳлtʦatлtʦʣлἵāmХʲл region was once linguistically richer than it is today. Question is whether my postulation of the centum *EteoἵāmХʲХл ʧʳл ϑoʲʲʣϑtл oʲл ʷʦʣtʦʣʲл tʦʣл dʣvʣὕopmʣntл oʤл tʦʣл ζdʣщл ʲootл *dʰgʰém in Y ₎dom, Z(ə) ǵ ‘m“nǰ/z(ə) ǵ ‘m ǵ, Yaz ul and Sar ul ϑanлϐʣлoϐʳʣʲvʣdлʧnлαaʲdʧϑлуoʲлmayϐʣлλūʲʧʳtānХфлὕanʥuaʥʣʳчлʧnлαaʲdʧϑлtʦʣлoutϑomʣлoʤлtʦʣл tʦoʲnлϑὕuʳtʣʲʳ лʳʦouὕdлϐʣлŋč, * (KÜκκβἱ, pers. comm.). The names Y ₎dom, Yaz ul , and Sar u can be also connected with IIr. *dʰām‘n- pὕaϑʣ члϐutлtʦʧʳлdoʣʳл not explain the initial parts of the presented toponyms. Yazghuὕāmл mayл ϐʣл ʣʸpὕaʧnʣdл aʳл *azga-dām‘n- ϐʲanϑʦpὕaϑʣ луKÜκκβἱ, pers. comm.), but Ir. *azga- (IIr. *ʜ‘₎gʰ‘-) is attested only in Western Iranian (Pahl. azg, Pers. ‘₎á ). ·34· development of ergative construction, which later tends to be lost. From demonstratives emerged definite article which became one of the most important part of speech since it determines gender (in those languages, where it is preserved), case and often subject, the demonstratives preserve triple deixis (except YazghuὕāmХчл ʷʦʣʲʣл tʦʣл ʳyʳtʣmл oʤ deixis has been innovated). ρʦʣлἵāmХʲлὕanʥuaʥʣʳлaʲʣлaὕʳoлvery similar in means of syntax. I.1.1.4.3. Wanjī τan₍Х (V‘njД or ld W‘njД; w/vanǰʧ, vanǰʧʷoʲ(í)чл vanǰivor) is an extinct language of the Vanj Valley in northern part of the Vanj district in Tajik Badakhshān. The first information on τan₍Хлaʳлthe language differing from Tajik comes from the year 1906 from a book Vostochnaya Bukhara by AndreЧ Evgen evich Snesarev (LASHἰAοBῦἰἴσ 2008, 61), first linguistic data were brought by Ivan Ivanovich Zarubin, who wrote that: «The inhabitants of valley of the Vanj river, pouring into the river Panj northwards of Yazghulām [and] where is now [spoken] one of the Mountain-Tajik dialects, do remember that their ancestors used to speak a different language. In the year 1915 there were living some elders who had us“d to h“‘› th“ W‘njД l‘ngu‘g“ from their grandfathers in childhood and could tell several words which were preserved in their memories. Despite their small number they [i.e. the words] allow to consider th“ lost l‘ngu‘g“ ‘s on“ o” th“ PāmД› [languages]» (ZARUBIN 1924, 79-80) – those several τan₍Хлʷoʲdʳлʲʣpʲʣʳʣntлaлὕʧʳtлoʤл33 words and phrases (ibid.: 80). Ten years later the Vanj Valley has been visited by Mikhail Stepanovich Andreev who confirmed that already in a half of the 19th century the language was spoken only in the furthermost villages of Upper Vanj. Andreev even met one of the informants of Ivan Ivanovich Zarubin – an old man of advanced age, who hardly recalled two-three dozen words of the forgotten language (ANDREEV 1945, 66). There are attested 64 τan₍Хл ʷoʲdʳ altogether (ZARUBIN 1924, 80; ROZENFEL D 1964, 141) and one derisive couplet recorded by Hannes SἰÖἱα (1936, 18-19; LASHἰAοBῦἰἴσ 2008, 622), some lexemes can be observed by an analysis of the Vanj toponymy and other words can be found in Tajik dialect of Vanj; together we can reconstruct some 500-600 τan₍Хлὕʣʸʣmʣʳ (LASHἰAοBῦἰἴσ 2008, 63). consonants stops bilabials pb affricates fricatives sonorants mw vowels iХ ü fv labiodentals ē dentals alveolars td postalveolars (c) ( ) čǰ sz šž palatals velars uvulars glottals uū kg q ʸ x h nrl y у ф eē ə aā Table 15 Sound system of τan₍Х. ·35· o οʣϑonʳtʲuϑtʧonл oʤл τan₍Хл pʦonoὕoʥyл ϑaʲʲʧʣʳл ʧtʳл oʷnл pʧtʤaὕὕʳл – the main problem is real phonological inventory which has been influenced by Tajik adstrate; for the reconstruction of τan₍ХлpʦonoὕoʥyлϑὕoʳʣὕyлʲʣὕatʣdлḲazghuὕāmХлandлShughnХ-οōshānХлὕanʥuaʥʣʳлaʲʣлʦʣὕpʤuὕщлρʦʣл development of vowels can be summarized as follows: *i, *u > i, ə; ŋД > i; ŋ > u; ŋ‘ > i, e, ai; ŋ‘u > au, aw~av; ŋ› > ø, ir; *a > a, u, e, ə; ŋā > o; ŋД > i; ŋ ‘ > e; ŋ›, *ar, ŋā before a nasal > ai; *a, ŋā under i-Umlaut > i, e; *a, ŋā in vicinity of a labial > o, u. For consonants is typical sonorization of voiceless stops when they follow sonors of voiced consonants and shifts ŋ > ⁾55; *ϑr- > r; *-ϑr- > c~č56. It seems that Middle Iranian sounds *ϑ, * 57 ʲʣmaʧnʣdлʧnлτan₍Хчлϐutлʧnл some cases there may be observed shift * > l (LASHἰAοBῦἰἴσ 2008, 64-89). Realization of v and w is disputable – whether they were two separate phonemes or free varieties of one sound as in Tajik58 (LASHἰAοBῦἰἴσ 2008, 75-77). τan₍Хл moʲpʦoὕoʥyл ϑanл ϐʣл ʲʣϑonʳtʲuϑtʣdл onὕyл paʲtʧaὕὕyл ʤʲomл tʦʣл attested material. Wan₍Хл probably distinguished masculine and feminine genders, some feminines were formed with i-Umlaut of the root vowel similarly as in other ShughnХ-YazghulāmХл ϐʲanϑʦл oʤл tʦʣл ἵāmХʲл languages. Plural of nouns was probably formed by adding an ending -ev. There is no information about the inflʣϑtʧonaὕл ʳyʳtʣmл oʤл τan₍Х. For adjectives there is attested the comparative ending -tar < Ir. *-tara-. Also information about verbal morphology is very poor. Several verbal stems are attested, for some of them we also know a past stem in *-ta-(ka-). Infinitive was formed by adding an ending -ak. Neither personal endings are attested, except imperative of the second person singular which was equal to the present stem. Marginally are attested also several demonstrative and relative pronouns and few postpositions (LASHἰAοBῦἰἴσ 2008, 95-103). A reconstruction of morphology is difficult, though there has ϐʣʣnлʲʣϑoʲdʣdлonʣлτan₍Хлϑoupὕʣd (bayt) – this couplet can be interpreted as Tajik ʷʧtʦлτan₍Хл lexicon (LASHἰAοBῦἰἴσ 2008, 622). I.1.1.4.4. Yazghulāmī YazghulāmХ (YazgulāmД; yůzdom(i) z(ə)vəg, z ǵ amíǵʧлz(ə)vəg, z ǵ amʧǵáyʧлz(ə)vəg)59 is a language spoken approximately by 3000 people in the Yazghulām valley in southern part of the Vanj In words recorded by Zarubin and Andreev ⁾ appʣaʲʳлʣʧtʦʣʲлaʳл<šşлoʲлaʳлuvuὕaʲл<ʸşчлʧnлtʦʣлρā₍Х₎лdʧaὕʣϑtлoʤлσan₍л there is also either or x ʤoʲлτan₍Хл * ⁾. 56 In records of Zarubin and Andreev instead of c tʦʣʲʣл ʧʳл <čşчл ʧnл σan₍Хл ρā₍Х₎л tʦʣʲʣл ʧʳл noл ъϑъл pʦonʣmʣчл ʧtл ʧʳл ϑonʳʧʳtʣntὕyлʲʣpὕaϑʣdлϐyлъčъщ 57 In words recorded by Zarubin and Andreev ϑ is spelled as <s> and is mostly spelled <d>, sporadically <z>. In the same way the continuants of *ϑ and * aʲʣлʲʣaὕʧzʣdлʧnлσan₍Хлdʧaὕʣϑtʳлoʤлρā₍Х₎щ 58 κodʣʲnл ρā₍Х₎л ʦaʳл ₍uʳtл onʣл ъvъл pʦonʣmʣл ʷʧtʦл poʳʧtʧonaὕл aὕὕopʦonʣл ъʷъл уPERRY 2005, 24-25), contrary Afghan αaʲХлʦaʳл₍uʳtлʳʧnʥὕʣлъʷълʳoundлуKISELEVA 1985, 27). 59 Persian name of the language sounds yaz ulāmДчлʧnлρā₍Х₎лtʦʣʲʣлaʲʣлtʷoлvaʲʧʣtʧes of the name: yaz ul mî (quite archaic) and ₍‘₎gul mî (the second variety can be influenced by Russian г or г ; but see ρā₍Х₎л dʧaὕʣϑtaὕл yazg lomí). YazghuὕāmХл dʣʲives its name from either the local name of the river Yazghuὕōmл – Z ǵ ‘ménǰ, or the name of the valley – Y ₎dóm. 55 ·36· district in Tajikistan (ÈDEL MAN 2000b, 274), from the year 1954 some YazghulāmХʳ live in KuЧbyshevsk district (nowadays Abdura mōn ηōmХл dʧstrict). There are no historical records about YazghulāmХщ The language does not have its own written form; the role of literary language is played by ρā₍Х₎Х Persian. YazghulāmХлʦaʳлtʷoлdʧaὕʣϑtʳл– Lower (Western) and Upper (Eastern), there is no clear border between these two dialects; internal differences are minimal, both dialects differ mainly in lexicon and pronunciation – especially in articulation of palatal tectals ḱ and ǵ (in the Upper dialect [c, ], in the Lower dialect > [n, k] or even [m, j]) etc. (ÈDEL MAN 1966, 9-11). consonants stops bilabials pb affricates fricatives sonorants mw vowels i ů fv labiodentals e dentals alveolars td postalveolars palatals velars labiovelars uvulars labiouvulars glottals u ǵ kg ₎° ʥ° q q° c čǰ sz šž ʸ ʸ° x ʸ° ° (h) nrl ə y у ф aā o Table 16 Sound system of YazghulāmХ. *Proto-Iranian vocalic system was completely remodelled in YazghuὕāmХчл vaʲʧouʳл transformations of vowels in stressed and unstressed positions occurred, and many changes were influenced also by ā- and i-Umlaut. Vowels a and ā do distinguish quantity, vowels e, i, o, , u are all short and ə is a super-short vowel. Peculiarity of YazghuὕāmХлʧʳлtʦʣлoppoʳʧtʧonлoʤлpaὕataὕчл velar and labial series of tectals – ḱ : k : k°, ǵ : g : g° and the opposition of labialized and plain (non-labialized) sounds continues also for velar fricatives (⁾ : ⁾°) and uvulars (q : ‹°, x : ⁾°, : °). Palatal tectals originate in plain velars that were palatalized by and ŋ› in so-called neutral position or under i-Umlaut. Labialization is a result of historical exposure to and ŋu ( has later underwent other sound changes, previous tectal was not labialized if has been changed by i- or ā-Umlaut). Tectals in front of front vowels (i, e) were also palatalized, on the other hand labialized sounds before back vowels (u, ) often lose their labial character. Original voiceless stops (together with ŋč) were sonorized between vowels. *Proto-Iranian ŋ , ŋž through stage ŋ , ŋž changed into ⁾, (but intervocalic *- - > *- - > w, ⁾); consonant groups *sp-, *st-, *sk- changed to > ŋ p-, ŋ t-, ŋ k-/ŋ ḱ- before and later came the change ŋ > ⁾ and in wordinitial clusters an epenthetic vowel was inserted between ⁾ and p/t/k/ḱ. Among other sound changes should be mention ŋ m > m; *dr-, *ϑr > c; ŋ⁾ > ⁾, ; or palatalization -d-, ŋ-t-, *-ḱ- > y. In consonant groups * › d and * › n the * › after a vowel formed a diphthong, such diphthong ·37· could have been monophthongized: *Vrd > *Vw / (when palatalized > *Vy /i ), *Vrn > *Vwn/ n; group * › t through intermediate *ḍ changed into g. (ÈDEL MAN 1987b, 353-381) YazghulāmХ nouns distinguish two genders – masculine and feminine, but the original gender system was transformed: the masculines include male names and persons and nouns denoting things and inanimate entities; female names and persons and animals (irrespective of their natural gender) are feminines. There can be traced some relics of old gender diversity, e.g. 1) plural ending -“žg appears with some feminines, 2) words ending in -“nǰ are old feminines; 3) in many words original feminine form can be observed due to reflexes of ā- and i-Umlaut and 4) the difference between original masculines and feminines can be seen in diverse reflexes of suffixes *(-a)-ka- ×л *(- )-kā-, *(- )-č -. Plural of nouns is formed by adding an ending -áϑ < *-ϑu‘-, plural of animate nouns can be also formed by adding an ending -én (with varieties -gén and -₍én for words ending in -a or -i respectively) derived from old genitive plural ending *-ānām. Another, yet non-productive plural endings are: -“žg for old feminine and rarely appears also an ending -án60. Old kinship terms in *-tar – ə d daughter and v(ə)›ád brother form plural by adding an ending -á› : ə dá›, v(ə)›‘dá›. YazghulāmХлʦaʳлtʷoлϑaʳʣʳл– direct and oblique, case is not expressed morphologically, it is expressed by a form of demonstrative pronoun; in singular there can appear attributive suffix -(y)i which is a reflex of Iranian genitive singular ŋ-h ‘. Adjectives are indeclinable, they do distinguish neither number, nor case nor gender, but gender categories are preserved in remnants – some adjectives have feminine forms that differ from masculine by operation of ā- or i-Umlaut of a root vowel. Personal pronouns distinguish direct and oblique cases in singular, in plural there is just one form for both cases; moreover, there is a possessive pronoun, which has separate forms for the first and second persons singular, in other cases it is formed with a suffix -i. Personal pronouns in the third person have two forms – one of them marks the third person in common and the other has an emphatic function – it points to a closer object. Oblique forms of the personal pronouns of the third person distinguish gender. Demonstrative pronouns originally had a system of triple deixis, this system changed to double deixis in course of the development of the language. From the original forms of demonstrative pronouns further developed forms of the third persons personal pronouns (for emphatic personal pronouns there fused the forms of I. and II. deixis – direct case is based on the I. deixis, oblique of masculine and feminine and of plural comes from the II. deixis; form of common third person pronoun originates in forms if the III. deixis); demonstrative pronouns yu(k) and du(k), which also serve as definite article, are based on the forms of the I. and II. deixis. 60 Plural ending -án is, similarly as above mentioned ending -(gй₍)én, a reflex of old genitive plural ending of a-stems. It seems, that the original -én was contaminated by Persian animate plural ending - ; the ending -án should be genuine YazghuὕāmХчлnoʷadayʳлʧtлappʣaʲʳл₍uʳtлʷʧtʦлtʦʣлʷoʲdлwex, man, pl. ⁽“⁾án, men (ÈDEL MAN 1987b, 382-383). ·38· YazghuὕāmХлverbal system is based on two stems – present and past. Present stems continue form Old Iranian verbal stems, but in forms of the third person the root vowel often undergoes i-Umlaut. Past (or preterite) stem originates in Iranian past participles in *- . To the present stem are added personal endings derived from Iranian primary endings, past tenses have endings derived from forms of copula – these endings are often added to the subject of clause. In past tenses ergative construction is applied, personal endings of the third person singular have different forms for transitive and intransitive verbs; intransitive verbs can even have no ending – it is often replaced by a subject in oblique case. (ÈDEL MAN 1987b, 381-401) I.1.1.4.5. The Shughnī-R shānī group The ShughnХ-RōshānХ language group is a family of eight mutually related languages and dialects which can be divided into four main dialect subgroups, individual languages/dialects are divided as follows: 1) ShughnХ (ShughānД, Shighn(ān)Д; xu nun(i) zivчл ʸu ni ziv), ShākhdaraХ (Shakhd‘›‘Д; ʸōʸdaʲāл zivчл ʸaʸdaʲāл ziv)61 and Ba₍ūХ (B‘j ⁽Д; ϐaǰū(w) ziv); 2) KhūʤХ уʸūʤл ziv) and RōshānХ (R shānД; ʲʧʸun ziv)62; 3) BartangХ (bārtāng ziv)63 and οāshārvХ (or Orosho›Д;лʲōšōʲvлziv); 4) SarХqōlХ (Tāsh‹ ›ghānД, wrongly (S‘›Д‹ lД) Tajik64; tuǰʧ₎лziv, πaʲХquli ziv)65. The languages of the ShughnХ-RōshānХл ʥʲoupл aʲʣл aὕtoʥʣtʦʣʲл ʳpo₎ʣnл ϐyл moʲʣл tʦanл ь77˙ыыыл people: ShughnХ is spoken by more than 10ы˙ыыы speakers in the Shughnōn and οōsht-Qalƒa districts of Tajikistan (ÈDEL MAN – YŪπςγBῦἰἴσ 2000a, 225) and some юы˙ыыылpeople in Afghan district of Sheghnān (BAKHρФBῦἰἴσл 1979, 3); RōshānХлʧʳлʳpo₎ʣnлϐyл18˙ыыы people on right bank of the river Panj in the Tajik Rushōn district (ÈDEL MAN – YŪπςγBῦἰἴσ 2000b, 242) and 2000-3000 speakers live on the opposite bank of the river Panj in the northern part of Afghan Sheghnān district (FAḲḳOV 1966, 5), KhūʤХ is spoken by more than 2300 people in the Khūʤ river valley in the Rushōn district (ÈDEL MAN – YŪπςγBῦἰἴσ 2000c, 254); BartangХл ʧʳл ʳpo₎ʣnл ϐyл appʲoʸʧmatʣὕyл 2500 speakers on the middle reaches of the river Bartang in the οushōn district (ÈDEL MAN – YŪπςγBῦἰἴσ 2000d, 259) andлοāshārvХ is used by some 2000 speakers on the upper reaches of the Bartang river in the οushōn district (ÈDEL MAN – YŪπςγBῦἰἴσ 2000e, 264); SarХqōlХ is a mother-tongue of more than 20˙ыыылspeakers in the Tāshqōrghān Tajik Autonomous County (T‘sh ku “›g‘n Tajike Zizhixian) in the Chinese Turkestan (ÈDEL MAN – YŪπςγBῦἰἴσ 2000f, τʧtʦлaὕʲʣadyлdʣadлBaʲʷāzлʳuϐdʧaὕʣϑtлуϐaʲʷōzʧлziv). τʧtʦлςppʣʲлу ēʲtānʥлzivфлandлἱoʷʣʲлуpōytaʸtлziv) subdialects. 63 With BaʳХdлуϐaʳХdлzivфчлBaʲdaʲaлуϐāʲdaʲāлzivф;лπʧpān₍луʳʧpōnǰлzivфлandлοaʷmēdлуʲaʷmēdлziv) subdialects. 64 Chinese authorities officially accept only one Iranian language in the Xinjiang-Uyghur autonomous region – the ρā₍Х₎л ὕanʥuaʥʣл уtajike-yuфчл ʦoʷʣvʣʲчл undʣʲл tʦʧʳл dʣʳʧʥnatʧonл ʤaὕὕл tʷoл ἵāmХʲл ὕanʥuaʥʣʳл – πaʲХqōὕХл уseleku'er-yu) and WakhХ (wahan-yuфщл λʣvʣʲtʦʣὕʣʳʳчл tʦʣʳʣл tʷoл ὕanʥuaʥʣʳл ʦavʣл notʦʧnʥл ʧnл ϑommonл ʷʧtʦл ρā₍Х₎л уʧщʣщл Cʣntʲaὕл Aʳʧanл variety of Persian), there are no Persian-ʳpʣa₎ʧnʥлρā₍Х₎ʳлin UyghuristanщлἱaϐʣὕὕʧnʥлoʤлtʦʣлπaʲХqōὕХʳлandлτakhХʳлaʳл ρā₍Х₎ʳлʧʳлϐaʳʣdлonлaлὕoϑaὕлὕaϐʣὕлoʤлtʦʣлπaʲХqōὕХʳлaʳлtuǰ k (< Pers. ) (cf. GAWARJON 1996, 257-266). In the past the term TājДk was used for Iranian-speaking population of Central Asia. 65 τʧtʦлρāshqōʲghānлуtošq ʲ onʧлzivчлvaʲšʧdʣлziv), WachaлуʷačaлzivфлandлBuʲunʥʳāὕлуϐ(ů)ʲůnʥʳoὕлziv, b( )r ngsol ziv) subdialects. 61 62 ·39· 269). The first historical record about the ShughnХ-RōshānХ languages can be found in the Travels of Marco Polo – he writes that the inhabitants of province of Balas(c)ian or Badas(c)ian (i.e. Badakhshān) have their own language (MARCO POLO, XLVII), ShughnХлʧʳлnotлmʣntʧonʣdл directly, but there are mentioned ruby-mines under the mountain Sighinan (i.e. Shughnān). The languages have no written tradition of their own, the only exception is ShughnХлʤoʲлʷʦʧϑʦл was created a Latin alphabet based on Tajik (and Pan-Turkic) variety of the Latin alphabet in tʦʣльŚюы ʳ66 (cf. ŞAκBζḳἴαĀρ 1931; ŞAMBIZODA 1937), but this alphabet has not been used for a long time. Currently there are some efforts to create a custom alphabet for each of the languages on basis of the Tajik Cyrillic alphabet (either by adding new diacritical marks or using digraph when letters and ь substitute diacritics) 67 , ʧnл tʦʣл ϑaʳʣл oʤл πaʲХqōὕХл tʦʣʲʣл ʦaʳл ϐʣʣnл created a local variety of the Latin alphabet based on Chinese Pinyin68 (cf. GAWARJON 1996). consonants stops bilabials pb affricates fricatives sonorants mw fv labiodentals td c čǰ sz šž palatals uvulars glottals ūu uu eē ȫ postalveolars velars iХ i dentals alveolars vowels kg q ʸ x (h) o ə nrl oō () y у ф Ǧ aaā Table 17 Sound system of the ShughnХ-RōshānХ languages (values in italic represent SarХqōlХлvoʷʣὕʳ). Individual languages and dialects of the ShughnХ-οōshānХл ʥʲoupл aʲʣл mutuaὕὕyл vʣʲyл ϑὕoʳʣл one to each other, substantial differences can be observed especially in vowels – ShughnХлdʧaὕʣϑtʳл andлοōshānХлʦavʣлtʣnлvoʷʣὕʳчлKhūʤХлʦaʳлʣὕʣvʣnлvoʷʣὕʳ69члBaʲtanʥХлandлοāshāʲvХл₍uʳtлnʧnʣлvoʷʣὕʳл andл ʧnл πaʲХqōὕХл tʦʣʲʣл aʲʣл onὕyл ʳʣvʣnл voʷʣὕʳл andл tʷoл dʧpʦtʦonʥʳ 70 . Valentina Stepanovna 66 ShughnХлLatin alphabet looked as follows (in parenthesis there are values of the letters corresponding to their scientific tʲanʳϑʲʧptʧonлuʳʣdлʧnлtʦʣлpʲʣʳʣntʣdлʷoʲ₎фśлaлāлϐ~ʙ (b) c (č) çл(ǰ) ꞓ (c) dлῥл( ) e (ē) ə ( ) ʤлʥлꞕл( ) ʦлʧлХл₍л(y) k l m n o ( ) ( ) pлqлƣл( ) ʲлʳлşл( ) tлꞛл(ϑ) uлūлvлʷлʸлꞗл(⁾) zлƶл(ž) ꞙ (ʒ). 67 γoʲл vaʲʧʣtʧʣʳл oʤл tʦʣл Cyʲʧὕὕʧϑл aὕpʦaϐʣtл ʤoʲл tʦʣл ἵāmХʲл ὕanguages of Tajikistan see ÈDEL MAN – DODYKHUDOEVA 2009a, 778 – Table 14a.1. 68 πaʲХqōὕХлpʧnyʧnлу Tuj k Ziv‘n P n₍ n ) looks as follows (values given in parenthesis show standard transcription of πaʲХqōὕХлaʳлʧtлʧʳлuʳʣdлʧnлpʲʣʳʣntʣdлʷoʲ₎фśлaлϐлϑл(⁾) d dz (ʒ) e f g gc ( ) gh ( ) h (x) hy (h) i j (ǰ) k kh (q) l m n o p q (č) r s ss (ϑ) t ts (c) uлüл( ) v w x ( ) y z zy (ž) zz ( ) (GAWARJON 1996, 1-2). 69 KhūʤХлæ and oлaʲʣлʲatʦʣʲлʲʧʳʧnʥлdʧpʦtʦonʥʳлψʧǦϊлandлψuɔ] respectively. 70 There were also long vowels ā, ē, ə, Д, , , ʧnлπaʲХqōὕХчлϐutлdʧfference in quantity has been lost (CIT). Instead of an opposition in vowel quantity, there is nowadays an opposition of stable (a, e, ə, o, u) vs. unstable (i, ) vowels. From the stable vowels e, o, u may be prolonged in speech. Schwa (ə) is considered an allophone of . (PAKHALINA 1966, 6) ·40· Sokolova reconstructs *Proto-ShughnХлvoϑaὕʧʳmлaʳлʤoὕὕoʷʳśл*a > ŋö, ŋā; ŋā > ŋȫ; > *a, *u; > *i; unstressed *i, *u > *ə; ŋ > ŋ“ ; ŋ u > ŋou (SOKOLOVA 1967, 63-78), in later development there took place other changes of vowels as effects of ā- and i-Umlaut, operation of stress and openness/closeness of syllable. The relationship of vowels in the ShughnХ-οōshānХлʥʲoupлϑanлϐʣл seen in scheme in Table 18. Consonantal system shares many common features: postalveolar affricates were depalatalized ŋč, ŋǰ > c, z-/ʒ; there happened second palatalization of velars *k, *g, *x > *ḱ, ŋ , ŋ⁾ > č, ž, in front of original front vowels (including ); *Proto-ShughnХлpoʳtvocalic voiceless sounds were sonorized *p, *t, *k, *ḱ, *c > b, d, g, ǰ/ž, ʒ/z; ŋ , ŋž changes through ŋ , ŋž into ⁾, , but post-vocalic *- - changes firstly to ŋ-ž- and it has later underwent different development in individual dialects: Shugh. члBa₍ūщл or w;лοōshщчлBaʲtщчлοāshrv. wчлπaʲХqщл or l (only occasionally w). Some other changes took place in consonantal groups: ŋ m > m; *ϑr > r (but word-initially ar-); *ʦr-, *ʣr- > * -, * - > ⁾-, -; *ʦ , ŋč > s; *gt, *kt > yd/wd; * › t > *ḍ > Shugh. dчл οōshщчл Baʲtщчл οāshʲvщчл πaʲХqщ g (rarely also * › t > r /V ); *rn > (w)n; * › ʦ, * › ʣ > Shugh. ⁾c, ʒ/ zчл οōshщчл Baʲtщчл οāshrv. ws, wzчл πaʲХqщл rs, rz. Old suffixes *-ka-, *-čД- usually changed to -ǰ and -ʒ (in the second case also with i-Umlaut of stem vowel). (SOKOLOVA 1967, 63-78; ÈDEL MAN 1987a, 238-284) Shugh. Х ē ō u ū ā a i u Khūʤ. οōsh. Bart.-οāshrv. SarХq. o Ǧ Х ē ō u ū ā a i u o ē Х ē ō u ū ā a i u ȫ ē Х ē ō e i y o u w ū ā a i u o a i Table 18 The relationship of vowels in the ShughnХ-RōshānХлὕanʥuaʥʣʳл(after: SOKOLOVA 1953b, 135; modified). There is distinguished masculine and feminine gender in the ShughnХ-οōshānХлὕanʥuaʥʣʳщл Gender differentiation is expressed in three ways: 1) morphologically – gender affiliation is maintained in reflexes of root vowels: masculines are words with reflexes of vowels in so-called neutral position and words ending in -ǰ < *-ḱa < *-ka-, feminines are words with reflexes of ā- and i-Umlaut and words ending in -ʒ < *- < *-čД-; 2) lexically – this way natural gender of animals and human beings is expressed as well as place-names, which belong to the masculine; 3) syntactically (or semantically) – syntactically gender is applied for majority of majority of nouns: feminines are entities perceived as individual unit, masculines can be the same words when perceived as collectives (morphologically in singular) – e.g. ‘ppl“ is feminine, if it is ·41· perceived as a single unit – (on“йth s) ‘ppl“ , but when it is perceived as ‘ppl“s ( n ‘ common s“ns“)ж m‘n₍ ‘ppl“s it is masculine 71 . In πaʲХqōὕХл tʦʣʲʣл ʲʣmaʧnʣdл ʳomʣл ʲʣflexes of gender in morphological and lexical level, in this case it is preservation of distinction of natural gender, syntactically the category of gender typical for the other ShughnХ-οōshānХл ὕanʥuaʥʣʳл ʷaʳл completely lost. Nouns distinguish two cases – direct and oblique, cases are often not expressed morphologically, in singular the direct and oblique cases are the same, formally they are equal to stem, in plural the situation is comparable – both cases are formed by adding a plural ending, onὕyлʧnлπaʲХqōὕХлtʦʣʲʣлaʲʣлtʷoлdʧfferent endings for direct and oblique case plural (under WakhХл influence?). Cases are expressed syntactically often with use of demonstratives. Plural can be formed by use of several endings. Plural of inanimate (and optionally animate) nouns is in Shughщчлοōsh., Bart. formed by adding an ending -ēn (following a vowel -₍ēn;лʧnлοāshrv. the ending -(₍)ēn appʣaʲʳлʲaʲʣὕyфчлandлʧnлοāshrv. -Д” (following a vowel -₍Д”флandлπaʲХqщл-ef (following a vowel -yef; used only in the oblique case), some animate nouns form plural from other endings: Shugh. -₍ n, -g n, -ǰ nчл οōshщчл Baʲtщчл οāshrv. -₍ n, -g n, -ǰ n; Shugh. - rʒ, - ›ǰ, οōsh. -ē›ʒ, - ›ǰ (Khūʤщл aὕʳoл -ā›ʒ), Bart. -ā›₎, -ē›₎, - ›₎чл οāshrv. -ā›ʒ, -ā›ǰ 72. Plural of some words is formed not just by adding the plural ending but also with an Umlaut of a root vowel. There is another set of collective plural endings: Shugh. -⁾ēlчл οōshщчл Baʲtщчл οāshrv. -⁾Дl (> πaʲХqщл-⁾ ₍l ʤoʲл pὕaʧn лpὕuʲaὕлʣndʧnʥлʧnлdʧʲʣϑtлϑaʳʣф;л Shughщчлοōshщчлοāshrv. -g‘lā, Bart. ; Shugh. -gu”tā, Bart. and relict Shugh. -Дčчл οōsh. -ēč; forms of collective plural can also take plural endings in -ēn. Adjectives do not differ in number or case, but some adjectives have different forms for masculine and feminine. Personal pronouns have forms just for the first and second persons, the third person is expressed by demonstrative pronouns. Both personal and demonstrative pronouns have two cases and two numbers (but the first and second persons plural have the same forms in the direct and oblique cases), the demonstratives distinguish gender in the oblique case (in ShughnХлtʦʣʲʣлaʲʣлmaʳϑuὕʧnʣлandлʤʣmʧnʧnʣлʤoʲmʳлaὕʳoлʧnлʤoʲmʳлoʤл demonstrative pronouns of III. deixis). Demonstratives distinguish triple deixis and they fill a syntactic function of definite article and they govern case of a noun besides the function of demonstrativʣʳлandлtʦʣлtʦʧʲdлpʣʲʳonлpʣʲʳonaὕлpʲonounʳщлπaʲХqōὕХлdʣmonʳtʲatives have preserved See the use of the word mā⁽nчлappὕʣлʧnлʤoὕὕoʷʧnʥлοōshānХлʣʸampὕʣʳśлdum {this: f. obl. sg. II. deixis} māwn {apple: f. sg.} mu-r {to me} dāk {givʣм}л ʥivʣлmʣлtʦʧʳлappὕʣ ;лdum {this: f. obl. sg. II. deixis} māwn {apple: f. sg.} ’ā⁾ ki {share!} ʳʦaʲʣлtʦʧʳлappὕʣ л×лday {this: m. obl. sg. II. deixis} māwn {apple(s): m. sg.} tar {to} ’ ₎ › {bazaar, market} ₍ s {carry!} ϑaʲʲyлtʦʣʳʣл appὕʣʳлtoлtʦʣлϐazaaʲ ;л day {this: m. obl. sg. II. deixis} māwn {apple(s): m. sg.} tar {to} ₎‘stāv {gate} ₍ s ʻϑaʲʲyм}л ϑaʲʲyлtʦʣʳʣлappὕʣʳлtoлtʦʣлʥatʣ луÈDEL MAN 1987a, 289; PAYNE 1989, 428). 72 Apart from the above mentioned plural endings there are many other endings, which are used only marginally: οōsh. -ʒēn, Shugh. -ʒin-ēn; Bart. -ʒ n, -ʒ‘n nчлοāshrv. -ʒ n;лBaʲtщчл οāshrv. -₎ ›;лBaʲtщчл οāshrv. - n; Bart. - ₍ā; Shughщчлοōsh., Bart. -ā› (this ending is added only to the word v › d ϐʲotʦʣʲ лśлv › dā›фщлζnлBa₍ūХлуandлpaʲtὕyлʧnл other dialects of ShughnХфл andл ʧnл BaʲtanʥХл tʦʣʲʣл ʧʳл aὕʳoл tʦʣл ʣndʧnʥл Ba₍щчл Shugh. -(ǰ)ēv (Shākhd. -(ǰ)ē”), Bart. -Д”, which is used in adverbial function indicating multiplicity of action, the same ending appears also in many placenamʣʳлуʧtлʧʳлtʦʣлʳamʣлʣndʧnʥлaʳлpὕuʲaὕлʣndʧnʥлʧnлοāshāʲvХлandлπaʲХqōὕХфщлуÈDEL MAN 1987a, 291-295) 71 ·42· forms of masculine and feminine, but the feminine forms are used rarely. (ÈDEL MAN 1987a, 284-316) ShughnХ-RōshānХл vʣʲϐaὕл ʳyʳtʣmл ʧʳл ϐaʳʣdл onл ʤouʲл ʳtʣmʳ: present, preterite, perfect and infinitive stems. The present stems continue from Old Iranian present stems, the preterite stem originates in Iranian past participles in *-ta- (m.), *-tā- (f., pl.) > -t/-d/*-ḍ (in feminine and plural forms there is ā-Umlaut of a root vowel), the perfect stem originates in extended perfect stem: *-ta-ka- (m.) // *-ta-čД- (f.) // *-ta-kā- (pl.) > -(C)č/-ǰ // -ʒ/-c // -(C)č/-ǰ (in feminine forms there is i-Umlaut of a root vowel, in plural ā-Umlaut takes place)73. Preterite and present stems distinguished gender and number, such distinction remained in majority of intransitive verbs forms, and transitive verbs are based on form of masculine, ʳamʣлaʳлπaʲХqōὕХлpreterite and perfect stems of intransitional verbs. Infinitive stem comes from Iranian verbal noun ending in *-ti, infinitive itself has two forms, short infinitive, which is equal to the infinitive stem and long infinitive – i.e. infinitive stem with the ending Shugh., οōsh., οāshrv. - ⁽, Bart. - (⁽), SarХq. - w. Personal endings of the present tense are consistent with Old Iranian primary endings, just the second person plural comes from optative ending ŋ-‘ t‘, forms of the third person singular often use i-Umlaut of root vowel with the ending -d/-t < *-ti. Past tense endings originate in forms of copula. The ShughnХ-RōshānХл ὕanʥuaʥʣʳл ʦadл oʲʧʥʧnaὕὕyл ʣʲʥative construction in the past tenses, ergative has been preserved in RōshānХ, KhūʤХ and BartangХ, however, in these languages the ergative construction tend to be substituted by absolutive construction as it is in ShughnХчлοāshāʲvХлandлπaʲХqōὕХ. Although the category of ergative has been lost in some languages (or it is slowly substituted by absolutive), the difference in transitive and intransitive verbs remains – in ShughnХ, οōshānХ, KhūʤХ, BartanʥХ and SarХqōὕХ the transitional verbs have an enclitic ending -i in forms of the third person singular (ʧnлοōshānХл and KhūʤХ use of the ending is optional, it is used mainly in phrases, in which there is not expressed subject; in SarХqōὕХл uʳʣл oʤл tʦʣл ʣndʧnʥл ʧʳл aὕʳoл optʧonaὕчл ϐutл ʧtл ϑanл ϐʣл uʳʣdл aὕʳoл ʤoʲл intransitive verbs; in RāshārvХлandлʧnлtʦʣлBaʳХdлdʧaὕʣϑtлoʤлBartanʥХлtʦʣʲʣлʧʳлnoлʣndʧnʥлatлaὕὕ), the intransitive verbs have no ending for the third person singular. BartangХ (and earlier also οāshārvХ) has special forms of enclitic ending for the third person plural. (ÈDEL MAN 1987a, 317337) 73 Reflexes of participles in *-ta- se do differ in individual dialects in front of preterite endings *- -/*-čД- e.g.: *tak-ta- -/-čД- (preterite stem of the verb to l“‘v“ ) > Shugh. t ₍ǰ // tДc // t ₍ǰ (m. // f. // pὕщфчлοōsh. tu₍ǰ // tayc // tā₍ǰ, Khūʤщлtu₍ǰ // tiyc/tДc // t ₍ǰ, Bart. t ₍ǰ // tayc // t ₍ǰчлπaʲХqщлt ₍ǰ (single form); ŋč u-ta- -/-čД- (preterite stem of the verb to go ) > Shugh. su ǰ // sic // sa ǰчлοōsh. suǰ // siʒ // s‘ǰ, Khūʤщлsuǰ // sic // s‘ǰ, Bart. suǰ // sic // s‘ǰчлπaʲХqщлse ǰ; transitive verbs have a single form based on masculine: ŋ’›-ta-ka- (preterite stem of the verb to ’› ng ) > Shugh. v ǰчлοōsh. (‘)v ǰ, Khūʤщлvugǰ, Bart. v ǰчлπaʲХqщлvə ǰ (ÈDEL MAN 1987a, 320). ·43· I.1.1.4.6. Sarghulāmī SarghulāmХ (or SaraghlāmД) 74 is a dead language from upper reaches of the Sarghulām (or Saraghlām) river in Afghan Badakhshān. The language became extinct at the beginning of the 20th century, the only reference about the language has been published by Ivan Ivanovich Zarubin, who in the year 1916 recorded several SarghulāmХл ʷoʲdʳл ʤʲomл aл κun₍Хл pʣʲʳonчл ʷʦoл claimed that he knew the SarghulāmХл ὕanʥuaʥʣ. From the list of SarghuὕāmХл ʷoʲdʳл ma₍oʲʧtyл ʷʣʲʣлἵʣʲʳʧanлoʲлκun₍Хлlexemes; Zarubin notes, that only three words could have been identified as SarghulāmХл ʷoʲdʳ75 – ⁽ol ké / ⁽ol kí water ; k ó cow , and ₎oДk boy , and he quotes these words with selected responses from other Eastern Iranian languages (ZARUBIN 1924, 79). Despite poorly documented linguistic material, we can get many valuable information about the language if we thoroughly analyse the attested words76. From the attested material we cannot judge much about SarghulāmХ – one can only guess that it ʧʳлonʣлoʤлtʦʣл λoʲtʦʣʲnлἵāmХʲлὕanʥuaʥʣʳчлoϐvʧouʳὕyлʲʣὕatʣdлtoлκun₍Х. However, we can observe two certain SarghuὕāmХлʧnnovatʧonʳ: change *d > l (thus a phenomenon that is known aὕʳoлʧnлBaϑtʲʧanчлκun₍ХлoʲлPashtō) and semantic shift of Ir. ŋuād -, irrigation channel > water77 (ZARUBIN 1924, 79; MORGENSTIERNE 1974, 99). Iranian suffix *- - should be attested je in words ⁽ol ké/⁽ol kí < ŋuād -kā- and ₎oДk < *ʣaha-ka- (or *ʣāt‘-ka- ??) child (cf. MORGENSTIERNE 2003, 103-104). Voiceless consonants were probably retained in intervocalic positions, in addition to example of suffix *- - similar feature can be seen in the word k ó78 < ŋk‘u - -/ - - cow 79 (cf. PAKHALINA 1987b, 484). Word-final long vowels were probably preserved; about the effʣϑtʳлςmὕautлaʳлʧtлʧʳл₎noʷnлʧnлotʦʣʲлἵāmХʲлὕanʥuaʥʣʳчлonлtʦʣлϐaʳʧʳлoʤлtʦʣл preserved material can be suggested only with reservations. By comparison with some other ἵāmХʲлὕanʥuaʥʣʳлʷʣлϑanлcome to a conclusion that (oblique?) plural ending was *or *- . ʧщʣщл tʦʣлʳpʣʣϑʦлoʤлmazāʲл(shrine) лaʤtʣʲлaлmazāʲлὕoϑatʣdлʧnлvʧὕὕaʥʣлoʤл The language was also known as laf Sarghuὕāmлуπaʲ GhuὕāmфлnʣaʲлtoлAfghanлγayzāϐād. 75 One can only say that it is a great pity that Zarubin did not specify also those words he did not consider SarghuὕāmХл – ʣvʣnлʤʲomлtʦʣл ʳtudyл oʤлϐoʲʲoʷʧnʥʳлʷʣл ϑouὕdл dʣduϑʣлmoʲʣл aϐoutлtʦʧʳл ὕanʥuaʥʣчл tʦʣл ʧʳʳuʣл oʤлκun₍Хл borrowing might be interesting – could they be a contamination caused by the first (?) language of the informant or were the SarghuὕāmХлandлκun₍Хлʷoʲdʳлʳoлʳʧmʧὕaʲчлtʦatлḳaʲuϐʧnлʧdʣntʧfiʣdлtʦʣmлaʳлκun₍Хлʷoʲdʳчлoʲлtʦʣʧʲлoʲʧʥʧnaὕл SarghuὕāmХлʤoʲmлʷaʳлʥaʲϐὕʣdлϐyлtʦʣʧʲлκun₍Хлʲʣʳponʳʣʳщ 76 Moreover there are several place-names in the Sarghuὕāmлσaὕὕʣyлtʦatлϑanлϐʣлʧdʣntʧfied as of SarghuὕāmХлoʲʧʥin: Ḥ‘l‘ngāu or Ḥ‘l‘ngā’, ŭučД⁽ and Gh‘›ālД⁽ (MORGENSTIERNE 1938, 439), I will not analyse them in detail in this work. 77 For a similar semantic shift see Wanj. wol ʷatʣʲ л ×л Ḳazgh. wa , Shugh. ⁽ , Wakh. wod/ ϐʲoo₎чл ʳtʲʣamчл уʧʲʲʧʥatʧonфл ϑʦannʣὕ чл Avʣщ v‘ⁱ i- ʧʲʲʧʥatʧonл ϑʦannʣὕ л <л ŋuād - (LASHἰAοBῦἰἴσ 2008, 83); Yazgh. ⁾“⁾, Vanj. ⁾ïk ʷatʣʲ л<лIr. ⁾āh‘- ʷʣὕὕчлʳpʲʧnʥ луMORGENSTIERNE 1974, 99) or Oss. don ʷatʣʲчлʲivʣʲ л<лIr. ŋdānu- ʲivʣʲ луABAEV 1958, 366-367). 78 ζnлκun₍Хчл ShughnХ-οōshānХлὕanʥuaʥʣʳч Ish₎āshmХ-πanʥὕēchХлoʲлʧnлPashtōлʧntʣʲvoϑaὕʧϑлх- - changed to *-ž- and later has undergone other phonetic changes. 79 Cf. Munj. / ϐuὕὕ члζshk. kьž k, Yagh. . 74 ·44· I.1.1.4.7. Munjī and Yidghā Mun₍Х (ḤunjānД or Minj(ān)Д; mən ʲōyчл mən Хwar, mŭn Хwar) is spoken by some 2000-2500 people in valley of the river Mun₍ān in Korān wa Mon₍ān district in Afghan Badakhshān 80 (GRYUNBERG 2000, 154; DECKER 1992, 54), Yidghā (Y dghā; yıd(ə) ā, lūṭ₎ūʦʷaʲ; Munj. yə ə ʲōy) is spoken by 5000-6000 speakers in the ἱūṭ₎ōʦ Valley in Pakistani Chitrāὕ (Yidgh. ) (DECKER 1992, 48). The Mun₍ān and ἱūṭ₎ōʦ Valleys are divided by the Hindū₎ush massif, the only path connecting both areas goes through the Dōrāh Pass ʧnл tʦʣл εʧndū₎ush, through which it is possible to pass further to the Sanglēch Valley. Both languages are closely related 81 , though both languages are hardly mutually intelligible today. Among the Yidghāʳл there is a legend, tʦatлtʦʣyлϑamʣлʤʲomлκun₍ānл – this fact can be also compared with the fact that majority of place-names in the ἱūṭ₎ōʦ σaὕὕʣyлʧʳлunὕʧ₎ʣлʧnлκun₍ānлaлnon-Iranian (mainly Dardic) origin and also that Yidghāлdoʣʳлnotлʳpὕʧtлʧntoлdʧaὕʣϑtʳчлϐutлκun₍Х has three dialects – Upper (Southern), Central and Lower (Northern). It is assumed that the Yidghāʳлϑamʣлtoлtʦʣ ἱūṭ₎ōʦ Valley sometime in the 11th-13th centuries (DECKER 1992). History of Mun₍ān is th unknown, the only historical record dates to the 7 century from the pen of Chinese traveller Xuan Zang, who within Tokhaʲʧʳtān mentions kingdom of Mungjin in Badakhshān (XUAN ZANG, I, 24, XII, 6; MORGENSTIERNE 1938, 7). Both languages do not have a written tradition of their own. Both languages are often classifiʣdлaʳлtʦʣлἵāmХʲл languages82, Valentina Stepanovna Sokolova classifies Mun₍Х and Yidghā aʳл mʣmϐʣʲʳл oʤл tʦʣл λoʲtʦʣʲnл ἵāmХʲл ʥʲoup (other members of this group are ShughnХ-RōshānХ languages, YazghulāmХ, Wan₍Х and probably SarghulāmХ; see SOKOLOVA 1973). Diffʣʲʣnϑʣʳл ϐʣtʷʣʣnл κun₍Хл dʧaὕʣϑtʳл andл Ḳʧdghāл ϑanл ϐʣл ʳʣʣnл maʧnὕyл ʧnл pʦonoὕoʥyчл correlation of vocalic system is summarized in Table 21щл κun₍Хл voʷʣὕʳл ă ( ) and ( ) merge with ə ʧnлϑoὕὕoquʧaὕлʳpʣʣϑʦщлκun₍Хлvoϑaὕʧϑлʳyʳtʣmлʷaʳлʣnʲʧϑʦʣdлϐyлἵʣʲʳʧanчлvoʷʣὕʳлa and u were introduced together with Persian loans (GRYUNBERG 1972, 400-401; GRYUNBERG 1987, 163-164), ϐutл tʦʣʳʣл ἵʣʲʳʧan л ʳoundʳл uʳuaὕὕyл mʣʲʥʣл ʷʧtʦл ʳʧmʧὕaʲл ʳoundʳл ʧnл κun₍Хл a ~ ă and u ~ (~ ə). Historical development of vocalism can be outlined as follows: *a > ă (~ə) ‖ o (in closed syllable > ) ‖ Д (in open syllable > ‖ a; in various positions > āйă ‖ ‖ a; under i-Umlaut > Д); ŋā > Д ; under i-Umlaut > ē); *i > ə ‖ i; ŋД, ŋ‘ > Д; *u > фл‖ ; ŋ > ; ŋ‘u > (under i-Umlaut > Д). Consonantal system of both languages has undergone many changes, which have comparable analogies within other Eastern Iranian languages. Development of voiced stops is the same as in Aϑϑoʲdʧnʥлtoлtʦʣлὕatʣʳtлʧnʤoʲmatʧonчлmoʳtлoʤлtʦʣлκun₍Хʳлὕʣʤtлκun₍ānлaʤtʣʲлtʦʣлʳtaʲtлoʤл Afghan Civil War (19891992) and they moved to different places in Pakistani Chʧtʲāὕчл manyл κun₍Хʳл mʧʥʦtл ϐʣл ₎ʧὕὕʣdчл andл manyл oʤл tʦʣʧʲл villages destroyed. Refugees themseὕvʣʳлʳayлtʦatлtʦʣyлʷouὕdлὕʧ₎ʣлtoлʲʣtuʲnлtoлκun₍ānлaʤtʣʲлtʦʣлʷaʲлʣndʳщлуDECKER 1992, 50) 81 For simplicity, in the following text the examples will be givʣnл ʧnл ϐotʦл ὕanʥuaʥʣʳл andл κun₍Хл ʤoʲmл ʷʧὕὕл ϐʣл separated from Yidghāл by double vertical lines ‖ ; to indicate the Lower, Central and Upper dialect I will use 80 82 ἴnлtʦʣлotʦʣʲлʦandчлκun₍ХлandлḲʧdghāлʳʦaʲʣлʳʣvʣʲaὕлʳʧmʧὕaʲлʤʣatuʲʣʳлʷʧtʦлτa etsХлandлPashtōщ ·45· the Middle Iranian period, in a later period there was a shift *d (> * ) > l; other characteristic changes are: *ϑ > ⁾; *-p-, *-t-, *-k-, *-č-, *- - > v/w ‖лv, ₍йø, gфл‖л , ž/y gфл‖лøй₍; *rt > r ṛфл‖лṛ; ŋ t, *rst, ŋ› t > ḱ ‖л č; *rn, ŋ› n, ŋ›žn > ńǵ ‖лṇ; ŋ m > m. One of the typical changes presents a loss of a nasal before (voiced) stop in Yidghāл andл ςppʣʲл κun₍Хщл Denominal abstract suffix *-(a-)ka- changed to -əy (-iy) ‖л -ëй-ə. Secondary palatalization of word-initial *k links both languages with YazghuὕāmХл andл tʦʣл ShughnХ-οōshānХл ὕanʥuaʥʣʳщл αuʣл toл tʦʣл ϑontaϑtл ʷith Dardic and Indo-Aryan languages cerebral sounds emerged in Yidghāщл уGRYUNBERG 1987, 171180; SἰηÆοσØ 1989c, 412-413; MORGENSTIERNE 1938, 36-109) consonants stops bilabials pb affricates td retroflexes alveopalatals palatals velars uvulars sonorants mw labiodentals alveolars fricatives c č čǰ ǵ kg q glottals fv sz šž šž ʸ x ʸ h vowels Х ū ŭ nrl ṛ u ē ə ō ńy у ф ă ā a ǥ Table 19 Sound system of κun₍Х. consonants stops bilabials pb affricates fricatives c čǰ č ϕ fv sz šž šž ʸ x labiodentals alveolars td postalveolars retroflexes palatals velars uvulars glottals ṭ ₎| ʥ| kg (q) sonorants mw vowels Х ū ŭ i n rлʲ l eē u ë ə ṛ o y у ф ä aā h Table 20 Sound system of Yidghāщ κun₍Хл andл Ḳʧdghāл nounʳл ʦavʣл tʷoл ʥʣndʣʲʳл уmaʳϑuὕʧnʣл andл ʤʣmʧnʧnʣфчл tʷoл numϐʣʲʳл уʳʧnʥuὕaʲчлpὕuʲaὕфлandлtʷoлϑaʳʣʳлуdʧʲʣϑtлandлoϐὕʧquʣфчлκun₍Хлʦaʳлadditionally predicative genitive and vocative. Adjectives have categories of gender and number but they do not distinguish case. Pronouns retain system of the direct and oblique cases together with the predicative genitive, demonstratives distinguish triple deixis. Verbal system is based on three stem system: present, pʲʣtʣʲʧtʣлandлpʣʲʤʣϑtщлκun₍Хлʤoʲmʳлoʤлpaʳtлtʣnʳʣʳлdʧʳtʧnʥuʧʳʦлtʲanʳʧtʧonaὕлandлʧntʲanʳʧtʧonaὕлvʣʲϐʳчл in Yidghāл tʦʣл ϑatʣʥoʲʧʣʳл oʤл уʧnфtʲanʳʧtivity have been lost. Moreover Yidghāл ʤoʲmʳл duʲative ·46· present and some verbal forms in Yidghāл oʲʧʥʧnatʣл ʤʲomл ʤoʲmʳл ϑaὕquʣdл ʤʲomл αaʲdʧϑл ἰʦoʷāʲл (ChʧtʲāὕХфщлуGRYUNBERG 1987, 180-229; SἰηÆοσØ 1989c, 413-415; MORGENSTIERNE 1938, 110-167) Munj. Х ē ū ō ă Yidgh. Х Х ū ē Х ū ō ā ǥ ē Х ū ō ă ə ŭ ə ə ŭ ā -əy ā -əy ā -əy Х 83 ēчлä i ū a 84 85 o i ə ŭ члä a -ë 86 Table 21 ρʦʣлʲʣὕatʧonʳʦʧpлoʤлvoʷʣὕʳлʧnлκun₍Хлdialects and in Yidghā (after: GRYUNBERG 1987, 169; modified). I.1.1.4.8. Ishkāshmī, Sanglēchī and Zēbākī Ish₎āshmХ (or Ishkāsh mД, ‘nД, RɪnД; š( )₎ošmʧлz vů₎члʲ nʧлz vů₎), πanʥὕēchХл(S‘nglДchД; sangl čʧл 87 88 of the Soutʦʣʲnл ἵāmХʲл lavz, sangl čʧл zəv k) and Zēϐā₎Х are three closely related languages group. They are spoken in south-eastern part of Tajik and north-western part of Afghan Badakhshān. IshkāshmХлis spoken by approximately 2000 speakers, majority of them lives in the village of Ran (Ishk. ьn), several Ish₎āshmХ speaking families live also in places such as Ishkōshim (Ishk. ḥьt or (ь)ko ьm), Sumjin, Mulvō₍ and Namatgūt (Wakh. Nəmətg t) on the Tajik bank of the river Panj and in vicinity of Afghan city of Eshkāshem (PAKHALINA 1987b, 474-475; PAKHALINA – QςοBŌλἴσ 2000, 197). In Afghan Badakhshān there live more than 1300 speakers of SanglēchХ (YŪπςγBῦἰἴσл – DODYKHUDOEVA 2008, 110) in the Sanglēch Valley < ŋā. In suffix -ēk‘ // - . 85 In ending of masculine nouns. 86 < *i. 87 All three languages are often referred to as Ish₎āshmХчл ʧʤл nʣϑʣʳʳaʲyл toл dʧʳtʧnʥuʧʳʦл ζsh₎āshmХл pʲopʣʲчл ʧщʣщл tʦʣл variety spoken on right bank of Panj the language is often called ›ьn ₎ьv k – ‘nД й ɪnД , or ‘n IshkāshmД . 88 According to information givʣnл ϐyл λazaʲл λazaʲzōdaл (an Ish₎āshmХл nativʣл ʳpʣa₎ʣʲчл mʣmϐʣʲл oʤл tʦʣл οuda₎îл Institute of Language and Literature of the Academy of Sciences of the Republic of Tajikistan) who has visited the πanʥὕēch valley in Afghanistan ʧnлэыы7члπanʥὕēchХ and Ish₎āshmХлaʲʣлmutuaὕὕyлʧntʣὕὕʧʥʧϐὕʣлὕanʥuaʥʣʳщ 83 84 ·47· southwards from the city of ḳēϐāk уḳēϐщл Iziv ); ḳēϐākХл dʧaὕʣϑt/language has been quite recently replaced by Afghan Persian and by Lower SanglēchХл dʧaὕʣϑt (YŪπςγBῦἰἴσл 2000, 186187). In addition to the above mentioned language varieties it is necessary to mention a mixed Ish₎āshmХ-SanglēchХ-WakhХл dʧaὕʣϑt of the village of Warg in Afghan Wakhān (MORGENSTIERNE 1938, 287, 291-292). Of all the three vernaculars only SanʥὕēchХл ʳpὕʧtʳл ʧntoл two dialects – ἱoʷʣʲл уλoʲtʦʣʲnфл andл ςppʣʲл уπoutʦʣʲnфл πanʥὕēchХчл ϐotʦл dʧaὕʣϑtл dʧffer slightly one form the other. There are not many information concerning Ish₎āshmХлʧnлʦʧʳtoʲʧϑaὕлʳouʲϑʣʳчл probably the first mention of the language can be from the Travels of Marco Polo, he mentions an indigenous language of province of Casem (or Scasem, Scasunen; MARCO POLO, XLVI) – it was either some today unknown language of region around the city of Keshem or it was a language of Ishā₎āshim89. All three languages do not have written tradition of their own, in recent years there are efforts in Tajikistan to create IshkāshmХлaὕpʦaϐʣtлϐaʳʣdлonлTajik variety of the Cyrillic alphabet. consonants stops bilabials pb affricates td postalveolars retroflexes ṭ c čǰ č fv sz šž šž uvulars glottals kg q vowels i u ů nrl e ḷ o y у ф palatals velars sonorants mw labiodentals alveolars fricatives a x (h) Table 22 Sound system of IshkāshmХ. Sound system of all the three dialects differs only a little bit, mainly in vowels. Vowel ʳyʳtʣmлoʤлπanʥὕēchХлappʣaʲʳлtoлϐʣлtʦʣлmoʳtлaʲϑʦaʧϑлoʤлtʦʣлtʦʲʣʣлvʣʲnaϑuὕaʲʳчлʧtлʲʣtaʧnʳлdʧʳtʧnϑtʧonл of five long and short vowels , 90, , , and a reduced vowel ə; on the opposite side stands Ish₎āshmХл voʷʣὕл ʳyʳtʣmчл ʷʦʧϑʦл doʣʳл notл dʧʳtʧnʥuʧʳʦл voʷʣὕл quaὕʧty91 – it has been replaced by opposition of stable versus unstable vowels, the stable vowels include a, e, i, o, , u; vowel ь92 is 89 The interpretation of the name Scasem or Scasunen is quite complicated, in the Travels there is attested also form Casem – this can be the Afghan city of Keshem or Esh₎āshem in Afghan BadakhshānлуϑʤщлYULE – CORDIER 1993, book 1, chapter 28, note 4; LENTZ 1933), it seems that probably two similar place names merged into interchangeable forms: Casem = Keshem ~ Scasem/Scasunen = Ish₎āshim/Esh₎āshem. 90 πanʥὕēchХл ē is often realised as rising diphthong /ʧē/; status of short e is unclear (YŪπςγBῦἰἴσл – DODYKHUDOEVA 2008, 116-117). 91 In descriptions of Ish₎āshmХл ϐyл δʣoʲʥл κoʲʥʣnʳtʧʣʲnʣл andл δʣoʲʥʣл Aϐʲaʦamл δʲʧʣʲʳonл ʷʣʲʣл ʲʣϑoʲdʣdл aὕʳoл ὕonʥл vowels (GRIERSON 1920; MORGENSTIERNE 1938, 283-427), in the description of Ish₎āshmХлpʦonoὕoʥyлϐyлσaὕʣntʧnaл Stepanovna Sokolova there is no mention about vowel length (SOKOLOVA 1953c). 92 Ish₎āshmХль is also spelled ɪ or I, mainly in non-Russian works. ·48· unʳtaϐὕʣщлḳēϐā₎ХлvoʷʣὕлʳyʳtʣmлʧʳлϑὕoʳʣʲлtoлtʦʣлπanʥὕēchХлonʣчлϐutлʧnлmanyлaʳpʣϑtʳлtʦʣʲʣлϑanлϐʣл ʳʣʣnл tʲanʳʧtʧonʳл ʤʲomл πanʥὕēchХл to Ish₎āshmХ;л unʤoʲtunatʣὕyл ḳēϐā₎Хл voϑaὕʧʳmл nʣʣdʳл aл moʲʣл dʣtaʧὕʣdлʳtudyчлʷʦʧϑʦлʧʳлʧmpoʳʳʧϐὕʣлduʣлtoлtʦʣлʤaϑtлtʦatлḳēϐā₎Хлʥavʣлʷayлtoлἵʣʲʳʧanлandлʲʣmaʧnʣdл aʳлaлʳuϐʳtʲatʣлʧnлἱoʷʣʲлπanʥὕēchХлуYŪπςγBῦἰἴσл 2000, 186). I tried to indicate the relationship of vowels of all the three vernaculars in Table 24. Due to a complex development of *ProtoIranian vowels in the Ish₎āshmХ-πanʥὕēchХл ὕanʥuaʥʣʳл tʦʣʧʲл ʣvoὕutʧonл ϑannotл ϐʣл ϑʦaʲaϑtʣʲʧʳʣdл briefly; the individual changes were influenced by ā- and i-Umlaut, openness or closeness of syllables and position of stress also played its part. The consonant system is in contrast to the vowels more or less the same in all the three dialects. There can be observed several similar features such as e.g. sonorization of voiceless stops in intervocalic position and their subsequent spirantization, partial shift ŋč, ŋǰ > c-/ʒ/z, ʒ/z; sonorization of intervocalic ŋ and its later change in ḷ or change ŋ m > m, ŋ t > t (in Ishk. also ṭ), *ϑr > r etc. In Ish₎āshmХл andл πanʥὕēchХл secondary palatalization of velar stops took place, its results vary: *ḱ, ŋǵ > č-, ǰ-йž-/y-, intervocalically y/i/ǰ/ž. Some other changes did not take place consistently in all languages: * (< *d, *-t-) changed into d word-initially in all the three vernaculars, in Ish₎āshmХлandлḳēϐā₎Хлуandл oʤtʣnл ʧnл ἱoʷʣʲл πanʥὕēchХфл tʦʧʳл ϑʦanʥʣл ϑontʧnuʣdл aὕʳoл ʷoʲd-ʧntʣʲnaὕὕyл ϐutл ʧnл πanʥὕēchХл - - is often preserved after vowels; *ϑ ϑʦanʥʣdлtoлπanʥὕēchХлandлḳēϐā₎Хлt but in Ish₎āshmХлtoлs; *rn > Ishk. r(n), Sangl. ṇ; *nd, *nt > Ishk. nd, Sangl. ṇd/nd; *ʦ t > Ishk. t, Sangl. t;л ʧnл πanʥὕēchХл there is ḷ (< *-rd-, *-rt-, *- -) preserved better than in Ish₎āshmХлуthere often ḷ > l); in Upper πanʥὕēchХл , often change to ⁾ and ž, ž to . (PAKHALINA 1987b, 476-496; YŪπςγBῦἰἴσл – DODYKHUDOEVA 2008, 117-174; MORGENSTIERNE 1938, 228-333; GRIERSON 1920) consonants stops bilabials pb affricates fricatives sonorants mw dentals td postalveolars retroflexes ṭ c čǰ s z š ž š ž palatals velars uvulars glottals iХ ū f v labiodentals alveolars vowels kg (q) уʸ) у ) x nrl ē (e) ē u ə oō ḷ y у ф aā (h) Table 23 Sound system of πanʥὕēchХ. The Ish₎āshmХ-πanʥὕēchХлὕanʥuaʥʣʳлdoлnotлdʧʳtʧnʥuʧʳʦлʥʣndʣʲлoʲлϑaʳʣчлoʲʧʥʧnaὕлʥʣndʣʲлʦaʳл been preserved only in several nouns; case is expressed syntactically by use of demonstratives. πanʥὕēchХл andл ḳēϐā₎Хл maʧntaʧnл хἵʲoto-Ish₎āshmХл pὕuʲaὕл ʣndʧnʥл dʣʲived from Old Iranian genitive plural ending *-ānām, in Ish₎āshmХл ʳuϑʦл ʣndʧnʥл ʧʳл uʳʣdл onὕyл ʷʧtʦл ʳʣvʣʲal animate nouns; Ish₎āshmХл ʤoʲmʳл pὕuʲaὕл ʷʧtʦл ʣndʧnʥл -o уʧnл πanʥὕēchХл - ), which is a loan from ·49· Persian -hā. Adjectives, same as the nouns, do not have categories of gender and case, moreover they do not distinguish number; in Ish₎āshmХл ʤoʲmʳл oʤл ϑompaʲative in *-tara- were lost. Personal pronouns distinguish direct and oblique case and a predicative genitive; the same categories are distinguished by demonstrative pronouns, which also distinguish triple deixis. Verbal system is based on two stems – present and past, the present stem continues from *Proto-Iranian present stems, the past (preterite) stem is derived from Iranian past participles in *- . Ish₎āshmХл ʤoʲmʳл paʳtл tʣnʳʣʳл ϐyл addʧnʥл ʣndʧnʥʳл dʣʲivʣdл ʤʲomл ϑopuὕa;л ʧnл πanʥὕēchХл paʳtл tenses of transitional verbs are formed by ergative construction, for intransitional verbs the situation is analogous to Ish₎āshmХщл уPAKHALINA 1987b, 496-536; YŪπςγBῦἰἴσл – DODYKHUDOEVA 2008, 175-227) Ishk. o a ů u e i Sangl. ē ā ō ū ə a ā ō o ū o u ə ē Х i Zēb. o ā ō ū ọ a ā ō o o ū u ə ē Х i Table 24 The relationship of vowels in IshkāshmХ, SanglēchХ and Zēϐā₎Х. I.1.1.4.9. Wakhī WakhХ (WakhānД; ʸ k z ₎члʸ kwor; in Pakistan also ʥuʦǰaὕʧ/guhyali – ё j‘lД ) is the second most uʳʣdлἵāmХʲлὕanʥuaʥʣлafter the ShughnХлὕanʥuaʥʣ. Its speakers live on territory of four states – Tajikistan, Afghanistan, Chinese Turkestan (Xinjiang) and Pakistan. The total number of WakhХлʳpʣa₎ʣʲʳлʧʳлʣʳtʧmatʣdлatляы˙ыыылpeople (REINHOLD 2006, 1), this number appears to be exaggerated. In Tajikistan there are 7000-ьы˙ыыы WakhХл ʳpʣa₎ʣʲʳл ὕiving in the Ishkōshim district (PAKHALINA 1987a, 408); in Afghanistan roughly 7000 speakers live in the Wakhān district; in Pakistan there are 7500-ьы˙ыыы WakhХʳл ʧnл tʦʣл Gō₍al (Upper Hunza), Ishkōman, ḲāsХn andл Ḳārkhūn Valleys; and approximately 6000-7000 WakhХʳл ὕive in SarХqōὕл ʧnл tʦʣл ·50· ρāshqōʲghānл Tajik Autonomous County in Chinese Uyghuristan (BACKSTROM 1992, 61-62). The WakhХʳ that live in the valleys of Northern Pakistan started to settle those areas sometime after the year 1880, the second wave of immigration continued between the years 1935 and 1940 (BACKSTROM 1992, 60). The WakhХ language is quite homogenous on all of its territories, it splits into three dialects – Lower (Western) and Upper (Eastern; including SarХqōὕ WakhХ) in Badakhshān andл δō₍al (Hunza) dialect of Pakistan (δō₍al WakhХл ʧʳл oʤtʣnл notл ϑonʳʧdʣʲʣdл aʳл individual dialect and it is often considered as a variety of Upper WakhХ), between the Upper and Lower dialects there is sometimes distinguished Central WakhХлdʧaὕʣϑtл(PAKHALINA 1987a, 408-409; PAYNE 1989, 419-420; BACKSTROM 1992, 65-69). The first historical record on WakhХл comes from Marco Polo; he notes that inhabitants of province of Vocan (i.e. Wakhān) have a speech of their own (MARCO POLO, L). WakhХл doʣʳл notл ʦavʣл ʧtʳл oʷnл ʷʲʧttʣnл tʲadʧtʧon, in Tajikistan there are efforts on to create WakhХлaὕpʦaϐʣtлϐaʳʣd on the Tajik Cyrillic alphabet, in Pakistan there is used a modified Latin alphabet based on scientific transcription of WakhХчл ʳomʣtʧmʣʳлtʦʣлςʲdūлaὕpʦaϐʣtлmayлϐʣлuʳʣd. consonants stops bilabials pb affricates fricatives sonorants mw vowels iХ uū fv labiodentals dentals alveolars td postalveolars retroflexes ṭ palatals velars uvulars glottals kg q c čǰ č sz šž šž n r (ʲ) l ( ) ʸ y (y) у ф (e) уēф əə oō (ṛ) ḷ уὕ) aā уa) x (h) Table 25 Sound system of WakhХ. Vowel system of WakhХлʧʳлʧnлϑommonлbased on six short (a, ə, i, o, u, ) and seven long (ā, ē, ə, Д, , , ) vowels93; historically *Proto-Iranian vocalic system has been influenced by series of changes, e.g. vowels in so called neutral position changed as follows: ŋā > , , ; *a > o, u, ; The vowel ē appears only in Lower WakhХчлʧnлtʦʣлotʦʣʲлdʧaὕʣϑtʳлtʦʣʲʣлʧʳл Д instead. Pakhalina claims that also ē can have its short counterpart (PAKHALINA 1987a, 410). Pronunciation of , varies, in the Central and Upper dialects as their pronunciation shifts from [ у”фϊлtʦʲouʥʦлψ у”фϊлtoлψɨу”фϊлуPAKHALINA 1987a, 410; BACKSTROM 1992, 410). WakhХл was variously transcribed ʉ, or ɷ in non-Russian works, ʉ is also used in the WakhХлἱatʧnлaὕpʦaϐʣtл in Pakistan. Some scholars believe that in WakhХл tʦʣʲʣл ʧʳл noл oppoʳʧtʧonл oʤл ὕonʥл andл ʳʦoʲtл voʷʣὕʳл уPAKHALINA 1987a, 410), with certaʧntyл ʧtл ϑanл ϐʣл ʳaʧdл tʦatл tʦʣл ὕʣnʥtʦл ʷaʳл notл ʲʣϑoʲdʣdл duʲʧnʥл tʦʣл ὕatʣʳtл ʲʣʳʣaʲϑʦʣʳл onл δō₍aὕл WakhХлуBACKSTROM 1992). In contemporary WakhānлτakhХлtʦʣʲʣлʧʳлʧnʳtʣadлoʤлoppoʳʧtʧonлʧnлquantʧtyлoppoʳʧtʧonл of stable ((e), ə, i, o) vs. unstable (a, u, ) vowels (ÈDEL MAN – DODYKHUDOEVA 2009a, 778). Persian ā уʧnлαaʲХл [ȵ”ϊчлʧnлρā₍Х₎лψɔ”ϊфлʧʳл oʤtʣnлʲʣaὕʧʳʣdл aʳл in WakhХчлʧnлtʦʣл δō₍aὕлdʧaὕʣϑtлʧtлʧʳл ʲʣaὕʧʳʣdл ὕʧ₎ʣл [ȵ”ϊлуʷʲʧttʣnл â in the Latin alphabet used for WakhХлʧnлPakistan). 93 ·51· ŋ > ē (/Д); ŋ u > ; , > , ə; but due to i-Umlaut the vowels shifted towards close front vowels, under ā-Umlaut there was a shift towards back open vowels (PAKHALINA 1987a, 412-419). WakhХл ϑonʳonantʧʳmл ʧnл quʧtʣл ϑonʳʣʲvative in some aspects, mainly due to the fact that the voiceless stops are usually retained (but in some cases they are sonorized or even spirantized), other archaic feature is preservation of Indo-Iranian clusters *tr and *kr (in *Proto-Iranian they shifted to *ϑr, *xr)94, partially there is preserved also Indo-Iranian group *k > k (in *ProtoIranian > ŋ⁾ ), or > щлπʧmʧὕaʲὕyлtoлtʦʣлotʦʣʲлἵāmХʲлὕanʥuaʥʣʳчлtʦʣʲʣлoϑϑuʲʳлaлʳʣϑondлpaὕataὕʧzatʧonл of velars. There is an interesting feature that links WakhХл ʷʧtʦл Khōtanʣʳʣśл *ʦu > (Khōtщл ъś-, -ź-ъл×лotʦʣʲлβaʳtʣʲnлIranian *spфщлρoʥʣtʦʣʲлʷʧtʦлʳomʣлotʦʣʲлἵāmХʲлὕanʥuaʥʣʳлτakhХлʳʦaʲʣʳл change ŋč, ŋǰ > c, ʒ. For many consonants there is often difficult to determine their development clearly, there are many alternations, e.g. ŋ > -, -, - -, - -, -ž-, -ž-, -⁾-; *g > g, , , ǰ, , ž, (z); *p > p, b, v, (f); ŋ t > st, t, t, ⁾t, -, t etc. Scholars who dealt with historical phonology of WakhХл (PAKHALINA 1983, 24-56; PAKHALINA 1987a, 420-438; MORGENSTIERNE 1938, 450-476), do not explain this curious feature, the exception is Ivan MikhaЧὕovʧch Steblin-Kamenskiy, who explains certain alternations as a result of sandhi and as an influence of areal contacts with neighbouring languages (STEBLIN-KAMENSKIY 1999, 17-40). Specifics of WakhХл dʣvʣὕopmʣntл can be interpreted as the influʣnϑʣл oʤл ϑontaϑtʳл ʷʧtʦʧnл tʦʣл ἵāmХʲ-εʧndū₎ush ethnolingvistic region (or in a wider view in Central Asian Sprachbund), John Payne offers a hypothesis that WakhХлʷaʳлtʦʣлoὕdʣʳtлуIranʧanфлὕanʥuaʥʣлʧnлtʦʣлἵāmХʲлʲʣʥʧon and later it was superstrated by the otʦʣʲлἵāmХʲлὕanʥuaʥʣʳлуPAYNE 1989, 421-423), Valentina Stepanovna Sokolova connects WakhХл closely with the Ish₎āshmХ-πanʥὕēchХлὕanʥuaʥʣʳлandлpʲopoʳʣʳлtʦatлtʦʣyлϑanлtoʥʣtʦʣʲлʤoʲmлtʦʣʧʲл oʷnл ʳuϐʥʲoupл oʤл tʦʣл ἵāmХʲл ὕanʥuaʥes (cf. SOKOLOVA 1973). In case of WakhХл tʦʣʲʣл ϑanл ϐʣл supposed early and intensive contact with Persian, many Persian loans had undergone intraWakhХлdʣvʣὕopmʣntлуSTEBLIN-KAMENSKIY 1999, 17-40), similarly was WakhХлʧnлquʧtʣлʧntʣnʳive contact with some Indo-Aryan language(s), there may have been some really old Indo-Aryan influences on WakhХлуPAKHALINA 1976a). WakhХл appʣaʲʳл toл ϐʣл aʲϑʦaʧϑл notл onὕyл ʤʲomл pʦonoὕoʥʧϑaὕл poʧntл oʤл vʧʣʷчл ϐutл aὕʳoл ʧnл morphology we can observe survivals of some archaic features that have not been preserved in other Eastern Iranian languages. Nouns do not distinguish gender, but according to operation of ā- and i-Umlaut there can be observed forms of feminine that certainly existed in older stages of WakhХлуʲʣὕʧϑtʳлoʤлnʣutʣʲлaʲʣлunϑὕʣaʲ;лPAKHALINA 1987a, 444-446). Unique archaism presents the preservation of traces of Old Iranian dual forms in WakhХśлʳomʣлnounʳлʷʦʧϑʦлdʣnotʣуdфлpaʧʲʣdл body parts and some other appellatives culturally perceived as pair (e.g. yoke or door) are in contemporary WakhХлϑonʳʧdʣʲʣdлaʳлʳʧnʥuὕaʲчлϐutлtʦʣʧʲлʤoʲmʳлaʲʣлϐaʳʣdлonлʦʧʳtoʲʧϑaὕлduaὕлʤoʲmʳл (plural of such words is then formed by standard addition of WakhХлpὕuʲaὕлʣndʧnʥʳфщлγoʲmaὕὕyл the survivals of dual do not differ from forms derived from historical singular, traces of dual can 94 The origin of groups tr, kr in WakhХлϑanлϐʣлϑonʳʧdʣʲʣdлaʳлanлʧnnovatʧonлʲatʦʣʲлaʳлaʲϑʦaʧʳmśлζIr. *pr, *tr, *kr > Ir. *fr, *ϑr, *xr (> (Saka) ŋpʰ›, ŋtʰ›, ŋkʰ›) > Wakh. *(f)r, *tr, *kr (cf. STEBLIN-KAMENSKIY 1999, 17-18). ·52· be observed in operation of a root vowel Umlaut (PAKHALINA 1987a, 447). Nouns are inflected in two cases – direct and oblique. Case is formally not distinguished in singular, in plural there are two endings: -Д (t) for the direct case, and -əv for the oblique; the ending of oblique plural ʦaʳлpaʲaὕὕʣὕʳлʧnлotʦʣʲлἵāmХʲлὕanʥuaʥʣʳ;лtʦʣлʣndʧnʥлoʤлtʦʣлdʧʲʣϑtлϑaʳʣлϑanлϐʣлʲʣὕatʣdлtoлπoʥdʧanл non-productive nominative plural ending - . In addition to the above mentioned there are also other plural endings in WakhХчл ʳomʣл oʤл tʦʣmл ϑomʣл ʤʲomл ἴὕdл Iranian genitive plural: - n/- n, -Дn (< *-ānām, *-Дnām); endings - ›g, - ›č (< *-ā-tra-ka-) have analogies in the ShughnХ-οōshānХл ʥʲoupл уʣщʥщл Shugh. - ›ǰ); and there are also some other endings: -āl, - l (< *-); -Д” (< ŋ‘ - ‘- < ŋ‘ -’ ‘-). The other plural endings except -Д t : -əv are non-productive and they appear only in forms of few nouns. According to the results of Umlaut can be assumed that some WakhХлnounʳлthat are currently considered as direct forms could have been derived from other cases than from nominative (PAKHALINA 1987a, 446-447). Adjectives distinguish neither gender (traces of original gender distinction in WakhХл ϑanл ϐʣл ʳʧmʧὕaʲὕyл aʳл ʤoʲл nounʳл observed in effects of Umlaut), nor number or case. Comparative is formed by adding the ending -tər < *-tara- but it can be formed analytically, there is not a separate form for superlative – it is expressed only analytically. Personal pronouns have forms just for the first and second persons singular and plural. Demonstratives distinguish triple deixis and they are used also for the third person of personal pronouns and as definite article. Pronouns are inflected in two cases – direct and oblique. WakhХлvʣʲϐaὕ system is primary based on opposition of present and preterite stem, from the preterite stem are derived some other forms of past tenses. Present is formed by adding personal endings to verbal stem, in the past tenses the endings are substituted by enclitic forms of copula. Preterite stem is normally formed by adding an ending derived from *Proto-Iranian preterites in *- -, occasionally, however, there are also forms derived from the suffix ŋ-n( -/*- - (PAKHALINA 1987a, 459-466). I.1.1.4.c. South and Southeast Eastern Iranian I.1.1.4.10. Pasht and Wazīrī Pashtō (or Pakht , Pusht , Paṭhān, Afghan; pəšt žəbə // paʸt žəba) is an Eastern Iranian language. Number of its speakers is the greatest among all Eastern Iranian languages – the language is spoken by more than 23 million people (ROBSON – TEGEY 2009, 721); speakers of Pashtōлὕive mainly in Southern Afghanistan and in North-western Pakistan, to a lesser extent there are some Pashtūn enclaves in Northern regions of Afghanistan and in Eastern Iran; Pashtōл is together with Afghan ἵʣʲʳʧanл уαaʲХфл ʲʣϑoʥnʧzʣdл aʳл official language of Afghanistan. Pashtōл distinguishes four main dialect groups: North-western and North-eastern (i.e. Hard or Eastern dialects – p‘⁾t ) and South-western and South-eastern (Soft / Western dialects – pə ), notʣʷoʲtʦyлʧʳлaὕʳoлτazХʲХлуdʧaὕʣϑtлoʤлtʦʣлτazХʲХлtʲʧϐʣчлʲʣmaʲ₎aϐὕʣлdʧaὕʣϑtлʷʧtʦʧnлπoutʦ-eastern Pashtō;лʷazХʲayлžəbba), which markedly differs from other Pashtō dialects. The Pashtūnʳлmay ·53· be connected with the tribe л mentioned by Ptolemy in area around Ἀ ᾱ (Ave. Harax ‘ⁱtД, OPers. Haraʰuv‘t ) and the river Ἐ ύ (present Hilmand; cf. SἰηÆοσØ 1989b, 384), from historical sources we know also some Pashtūnл tʲʧϐʣʳ, e.g. the AfrХdХл tʲʧϐʣл ϑanл ϐʣл connected with the Ἀ ύ mentioned by Herodotus. Pashtōлʧʳлʷʲʧttʣnлʧnлtʦʣлἵʣʲʳo-Arabic script supplemented by graphemes for Pashtōл ʳoundʳ. The oldest written monuments come from the 8th century (GRYUNBERG – ÈDEL MAN 1987, 7), literary tradition evolves from end th of the 16 century (ROBSON – TEGEY 2009, 721). consonants stops bilabials pb affricates palatals velars uvulars glottals vowels íХ úū (f) td postalveolars retroflexes sonorants mw labiodentals alveolars fricatives ṭ ǵ kg (q) c čǰ sz šž šž ʸ x (h) i nrl ē uō ə ṛ y у ф a ā Table 26 Sound system of Pashtō. Sound system of Pashtōлʦaʳлundʣʲʥonʣ a complex development; its characteristic feature is syncope of unstressed vowels, due to syncopation of unstressed vowels consonant clusters appear often syllable-initially. Development of Pashtō vowels can be summarized as follows: *a > a, ə, ā, (under i-Umlaut > ); ŋā > , ā, a (under i-Umlaut > (₍)ā); *i, ŋД > i; *u, ŋ > u, ə; * u > wa, ; ŋ u‘ > , ; * > i, ē; position of stress influenced quality and quantity of vowels in *ProtoPaṭʦān. Voiceless consonants were voiced after a vowel (also *-f- > * > w, but *-t- > *d (> * > l), ø); ŋč was depalatalized to c-, -ʒ(-); *d > * > /l/ and ŋ > -~⁾-, -ž(-)~-ǵ(-). From consonant groups containing *r or ŋ emerge cerebral sounds, e.g. *sr-, *str > ~⁾, *rd, *rt > ṛ, *rn, ŋ⁾ n > ṇ, *rs > t~⁾t. Clusters are often simplified, in some cases one of the consonants disappears or a consonant is changed into another one. Due to i-Umlaut the vowel can have prothetic y, i(-) > -), čē(-); a frequent phenomenon is which can cause secondary palatalization, e.g. also metathesis, assimilation or dissimilation and pre-nasalization of consonants (SἰηÆοσØ 1989b, 398-406; GRYUNBERG – ÈDEL MAN 1987, 21-38). In Pashtō there is mobile stress, words are divided into two stress patterns: barytones (words with a stressed root) and oxytones (words with a stressed ending or suffix). Study of operation of stress in Pashtō can help in reconstruction of stress in *Proto-(Indo-)Iranian – in some cases position of stress in Pashtōл appears to be more archaic than stress attested in Vedic (GRYUNBERG – ÈDEL MAN 1987, 38-39). Pashtōл nouns and adjectives distinguish two genders (masculine and feminine), two numbers, in plural is also distinguished animacy or inanimacy. Nomina are inflected in three cases: direct, oblique and vocative, some masculines can moreover form oblique II. Nouns are ·54· inflected in eleven paradigm classes (seven masculine and four feminine classes), the adjectives form four inflectional classes; there are also inflectional subgroups in each of the classes, inflectional forms often differ due to operation of stress. Verbal system has triple structure similar to other Eastern Iranian languages: present, preterite and prefect. In past tenses there is distinguished transitivity and intransitivity. Aspect of verbs is formed either by prefix wə- or by suppletive forms or stress shift (SἰηÆοσØ 1989b, 390-398; GRYUNBERG – ÈDEL MAN 1987, 44135). I.1.1.4.11. Wanetsī Wa etsХ ((ʳpХn) tar nōчлwa ʣϑХлz(i)bə/zəϐō, č algari) has been for a long time considered a dialect of Pashtō (it was often called corrupted Pashtō and is recognised as a kind of Pashtō by its own speakers, see ELFENBEIN 1984a, 54-55), nowadays it is mostly considered to be an independent language closely related to Pashtō (HALLBERG 1992, 45-47). Wa etsХ shares many features with Pashtōчлmaʧnὕyлʷʧtʦлἰā₎aṛХ dialect and Soft Pashtō in Quetta area, Pakistan. On the other hand Wa etsХл«diff“›s ”›om ‘ll oth“› P [Pashtō dialects] in phonology, morphology, and l“⁾ con so much ‘s to ’“ ‹u t“ un nt“ll g ’l“ to oth“› P [Pashtōϊлsp“‘k“›s ( n ‘ ⁽‘₍ th‘t “.g. W‘₎Д›Д s not) (ELFENBEIN 1984a, 55). The supposed number of speakers exceeds э5˙ыыылpeople living in Harnāy (Wa . A›n‘(h)Д) and πan₍āvХ ta ʳХὕs ʧnлπʧϐХлdʧʳtʲʧϑtлeastwards from Quetta, province of BaὕōchʧʳtānчлPakistan; many of the speakers live also in Quetta (HALLBERG 1992, 47-48). The language is spoken by Məkhyā Хл andл τa etsХ tribal subdivʧʳʧonʳл oʤл πpХn ρaʲХnл tribe 95 . The language itself does not possess any prestige in its socio-linguistic area, even among its own speakers it does not enjoy adequate prestige and is even disdained by the Pashtūnʳщлἵʲʣʳʣʲvatʧonл of the language in such socio-linguistic situation is connected with tribal matters as each tribe identifies itself through its own dialect. (ELFENBEIN 1984a, 55-56) The language has no written tradition, nor was thoroughly described in scientific literature. All Wa etsХʳлaʲʣлϐʧὕʧnʥuaὕлʷʧtʦл Pashtō and recently, as the importance oʤлςʲdū rises, many Wa etsХʳлʳpʣa₎лalso ςʲdūщ Phonologically Wa etsХлdoʣʳлnotлdʧffʣʲлmuϑʦлʤʲomлnʣʧʥʦϐouʲʧnʥлἰā₎aṛХлdʧaὕʣϑtлoʤл Pashtōщл Historical development of vowels is quite similar to that of Pashtōśл*a > , ə; ŋā > , ; *i, ŋД > ; *u, ŋ > , ə; * u > wa, ; * > , ; just i-Umlaut or epenthesis of -y- is not as common as in Pashtōщлπtʲʣʳʳʣdлa is often lengthened, unstressed ā is shortened, word-final unstressed a and ə usually merge in pronunciation. Vowels -Д- and - - tend to be prepalatalized /-yХ-, -y -/ and - - and - - prelabialized /-ʷū-, -ʷō-/, but word-initial ⁽ -, ⁽ - are often delabialized. Ḥ‘jh l vowels ē, are not kept apart from m‘ › ” Д, . Consonants have comparable development with Pashtōśл voiceless consonants were voiced word-internally (*-p-, *-k-, *-č-, *-f-, *- - > b, g, ŋǰ > ʒ, w, ); ŋč and ŋǰ were depalatalized; * and ŋu merged as w; * , *ϑ > l as in The Wa etsХ-speaking πpХnлу τʦʧtʣ флρaʲХnлtʲʧϐʣлʤoʲmʳлaлmʧnoʲʧtyлoʤлaлὕaʲʥʣʲлρaʲХnлtʲʧϐʣл – its major group are tʦʣлρōʲлу Bὕaϑ₎ флρaʲХnʳчлanotʦʣʲлʳmaὕὕлʥʲoupлaʲʣлtʦʣлBōʲлу Bʲoʷn флρaʲХnʳщлρʦʣлBὕaϑ₎лandлBʲoʷnлρaʲХnʳлaʲʣлaὕὕл Pashtōлʳpʣa₎ʣʲʳлуELFENBEIN 1984a, 56). 95 ·55· Pashtō;л ϐutл *-t- > y, ø aʳл ʧnл κun₍Хл уoʲл ʧnл aл ʷayл aʳл ʧnл ἵaʲāchХл andл ŌʲmuṛХфщл αʧfferent from Pashtōлʧʳлʲʣtʣntʧonлoʤлr in ŋ›ž > ›ž; development ŋ k, *ft > k, w (Pasht. č, (w)d) or retention of *nd in andəm ʷʦʣat л <л ŋgántum‘-, Pasht. anəm. Often *CrVC > C›C. Cerebral * , * merged with , ž in Wa etsХчл ϐutл tʦʣyл mayл ϐʣл oϑϑaʳʧonaὕὕyл ʲʣʧntʲoduϑʣd л ʧnл ʳpʣʣϑʦл undʣʲл Pashtōлʧnfluence. As in colloquial Pashtōчлtʦʣʲʣлaʲʣлnoлъʤъчлъqълandлъʦълʳoundʳлʧnлτa etsХчлtʦʣʳʣл sounds are usually realised as p, k, ø (rarely x) respectivʣὕyчлtʦʣyлϑanлappʣaʲлonὕyлʧnл ʣduϑatʣd л speech. Phonological changes show that *Proto-Wa etsХл dʣvʣὕopʣdл quʧtʣл ʣaʲὕyл ʤʲomл хἵʲotoPaṭʦānл anϑʣʳtoʲл andл хἵʲoto-Wa etsХʳл ʷʣʲʣл pʲoϐaϐὕyл ʤoʲʣʲunnʣʲʳл oʤл tʦʣл Pashtūnʳл toʷaʲdʳл tʦʣл East. (MORGENSTIERNE 1983a; ELFENBEIN 1984a, 56-57; ELFENBEIN 1984b; MOSHKALO 2000, 150) consonants stops bilabials pb affricates td postalveolars retroflexes ṭ c čǰ (f) s z š ž (š) (ž) uvulars kg (q) glottals nrl vowels iХ uū eē oō ə ṛ y у ф palatals velars sonorants mw labiodentals alveolars fricatives a ā x (h) Table 27 Sound system of Wa etsХ. Wa etsХлnouns distinguish gender (masculine and feminine), number (singular and plural) and three cases (direct, oblique and vocative), vocative is usually the same as the direct case. There are eight inflectional classes – five for masculines and three for feminines, only masculine and feminine class I nouns have different vocative forms. As a fourth case can be considered ablative formed by agglutination of -(ē)₍‘. Wa etsХл ʦaʳл ʤoʲmʳл ʤoʲл aὕὕл tʦʲʣʣл pʣʲʳonʳ;л first and second persons singular and third person for both numbers distinguish direct and oblique cases, forms of the third person also retain gender distinction. Unlike Pashtō, Wa etsХлdʣmonʳtʲatives have triple deixis. Verbal system is based on two stems – present and past, past stems are formed from old past participles as in other Iranian languages. The past tense is formed by means of ergative construction. (MORGENSTIERNE 1983a; ELFENBEIN 1984a; MOSHKALO 2000) Wa etsХлpʦonoὕoʥyлandлmoʲpʦoὕoʥyлʧʳлfrom historical point of view very similar to Pashtō, many forms were also influenced by language contact. Main differences between both languages can be seen in syntax and lexicon. I.1.1.4.12. Parāchī ἵaʲāch is one of the New Iranʧanл ὕanʥuaʥʣʳл ϑὕoʳʣὕyл ʲʣὕatʣdл toл ŌʲmuṛХ, its accurate classifiϑatʧonлʦaʳлnotлϐʣʣnлʳuϑϑʣʳʳʤuὕὕyлʣʸpὕaʧnʣdśлʳomʣлʳϑʦoὕaʲʳлϑὕaʧmлἵaʲāchХлуandлŌʲmuṛХфлaʳл ·56· Eastern Iranian, some other as (North) Western Iranian language (see MORGENSTIERNE 1929; KIEFFER 1989; EFIMOV 1999a). ἵaʲāchХлʧʳлʳpo₎ʣnлϐyлappʲoʸʧmatʣὕyл5000 speakers in the Shotol, Pachaghānл and Ghochūὕān valleys 96 in λʣ₍ʲāϐл district south-ʣaʳtʷaʲdʳл ʤʲomл ἰāϐuὕ (EFIMOV 1999a, 257). The oldest reference about the language quoted as p‘›āǰД comes from the 16th century from the Bāϐuʲnāma, memories of Mughal sultan aʦХʲuddХn Mu ammadл Bāϐuʲ (KIEFFER 1989, 445-446). The language does not have its own written tradition. consonants bilabials stops affricates p pʰ b ϐʰ t tʰ d dʰ postalveolars retroflexes c č čʰ ǰ ǰʰ f (v) s ʳʰ z š žлžʰ ṭ ṭʰ k ₎ʰ g ʥʰ uvulars (q) у‚ф glottals vowels Х ǖ n nʰ r ʲʰ l ὕʰ ṛ (y) у ф palatals velars sonorants m mʰ w labiodentals alveolars fricatives ū i ē ȫ e u ō o ə ä a ā x h Table 28 Sound system of ἵaʲāchХ. Historical phonology shows some similarities with Pashtō and with Saka dialects, but preservation of word-initial voiced stops is similar with the North Western Iranian languages. Word-initial (voiceless) fricativʣʳлϑʦanʥʣdлtoлvoʧϑʣὕʣʳʳлaʳpʧʲatʣdлʳtopʳлуaʳлʧnлπa₎aлoʲлBaὕōchХфśл*f-, *ϑ-, *x- > pʰ, tʰ, kʰ. Characteristic changes are ŋu-, ŋ > *g -, ŋǰ > -, ž- (with certain similarities in KhōtanʣʳʣлandлBaὕōchХф; *s(t)r, *ʦr > ; sonorization of intervocalic voiceless stops *-p-, *-t-, *-k- and their merger with voiced fricatives (< Old Iranian voiced stops) *- -/*-ƀ-, *- -/*-đ-, *- -/*-ǥ- > w, ø~⁽~₍~ʰ97, ; ŋ t > ṭ, *st following *i > t, *rt, *rd > ṛ, ŋu - > u-. Intervocalic *- - is lost. There is often metathesis of h and subsequent development of aspirated consonants. Consonantal system is very similar to PashtōчлmoʲʣovʣʲлἵaʲāchХлʦaʳлaʳpʧʲatʣdлʳoundʳлʧnϑὕudʧnʥл nasals, sonorants or sibilants. In vowels there are following significant changes: stressed *a > ȫ~ ; ŋ , ŋ‘ > Д; ŋ‘u > ; ŋ ‘ > ē; ŋā u‘ > ȫ~ ; ŋ› > ur; *a, ŋā under i-Umlaut > , *a under a-Umlaut > a. ἵaʲāchХлὕonʥлϐaϑ₎лʲoundʣdлvoʷʣὕʳлtʣndлtoлbe fronted: , > ǖ, ȫ; ā is strongly rounded and often raised in front of a nasal. Long vowels are shortened in unstressed position. Diphthongs tend to reduce its non-syllabic part, mainly in fast speech. Besides oral vowels there are also nasalized vowels. (MORGENSTIERNE 1929, 22; KIEFFER 2009, 694-695; SἰηÆοσØ 1989a; EFIMOV 1999a, 258) Each of these valleys has its own dialect – ShotoὕХчлἵachaghānХлandлGhochūὕānХщ ζnлʳomʣлϑaʳʣʳл oὕd лandл nʣʷ л* /ŋđ continued as *h, it is preserved as aspiration oʤлϑonʳonantʳчлϑʤщлἵaʲāch. dʰД < *dǖh < ŋd йđ‘- < *d ta- ʳmo₎ʣ ;л ’ʰām < ŋ’uhām < bu йđām < ŋ’udām‘- ʳmʣὕὕ ;л Avʣщл bao a- (MORGENSTIERNE 1929, 36). 96 97 ·57· ParāchХл nouns do not distinguish gender, plural is formed by adding an ending -ān, but there is also a elliptic dual in -hā and numerative in -a. There are three cases: direct, oblique and ablative, other cases (accusative, dative, locative-directive and instrumental-comitative) are formed analytically with pre- or postpositions. Adjectives are not morphologically marked. Pronouns distinguish five cases: direct (nominative), oblique, dative, accusative and possessive. Verbal system is based on an opposition of present and past stems (past stems are alike in other Iranian languages formed from past participles in *-ta-). Verbs have perfective and imperfective aspect, past tenses transitional and intransitional verbs are formed with ergative construction (KIEFFER 2009, 696-708). I.1.1.4.13. rmuṛī ŌrmuṛХ уōʲmuṛ чл ōʲmuḷ чл ϐaʲʥʧʳtā, ϐaʲa₎Хфл is a New Iranian language variously treated as a member of the Eastern or Western Iranian group (see MORGENSTIERNE 1929; HALLBERG 1992, 53-66; EFIMOV ьŚŚŚϐфщл ζtл ʧʳл ϑὕoʳʣὕyл ʲʣὕatʣdл toл ἵaʲāchХчл ϐotʦл ὕanʥuaʥʣʳл aʲʣл noʷл mutuaὕὕyл unintelligible. ŌʲmuṛХлʧʳлʳpo₎ʣnлϐyлʳomʣ 100-200 people of city of BarakХ-Barak (Ōʲm. ё›ām; Pasht. ›məṛ, ›muṛ) in Afghan province of Lōgar and approximately of 5000 speakers in city of KānХgurām 98 in South WazХristān, Pakistan (EFIMOV 1999b, 276). The language has been mentioned for the first time in the 16th century as ’Д›kД together with some other regional ὕanʥuaʥʣʳл ʧnл vʧϑʧnʧtyл oʤл ἰāϐuὕл ʧnл the Bāburnāma of Mughal sultan ahХruddХn Mu ammad Bābur (KIEFFER 1989, 445-446). ŌʲmuṛХл ʦaʳл noл ʷʲʧttʣnл tʲadʧtʧonчл ʧnл ʲʣϑʣntл tʧmʣл tʦʣʲʣл ʷaʳл ϑʲʣatʣdл anл aὕpʦaϐʣtл ʤoʲл ἱōʥaʲл ŌʲmuṛХ based on Pashtō variety of the Arabic alphabet (BURKI 2001). consonants stops bilabials pb affricates labiodentals alveolars td postalveolars retroflexes ṭ c čǰ fricatives sonorants ϕw (f) sz šž řšž m palatals velars uvulars labiouvulars glottals kg q ʸ x vowels Х nrl ū i e ö ē u ə o ō ṛ (ḷ) ä y у ф a ā a ʸ° ° h Table 29 Sound system of ŌʲmuṛХ. ŌʲmuṛХл voʷʣὕʳл dʣvʣὕopʣdл aʳл ʤoὕὕoʷʳśл *a > *a, ā (labialized > u, ; palatalized > ē); ŋā > ā (unstressed > a; palatalized > ē); *i > i, e (unstressed > a; before ṛ > ē); Д > Д; *u > u (unstressed > ŌʲmuṛХл ʦaʳл tʷoл vaʲʧʣtʧʣʳл – ἰānХʥuʲāmл dʧaὕʣϑtл oʤл Pakistan andл ἱōʥaʲл dʧaὕʣϑtл oʤл Baʲa₎Х-Barak, both vernaculars differ quite a lot, there are differences in phonology, morphology and in lexicon. 98 ·58· a); ŋ > ŋ ; * (a) > ŋē (before *n > Д; unstressed > i); word-final *-‘ ‘- > *-Д; * u > ŋ (before *n > ); * u‘ > ŋā, (word-finally > ); ŋ› > ar, ᵃr, ur. There are some differences between Afghan and Pakistani vaʲʧʣtʧʣʳл oʤл ŌʲmuṛХл – in Afghanistan there is under the influence of αaʲХъἰāϐuὕХл tʣndʣnϑyл toл ʲʣaὕʧzʣл ʳʦoʲtл i, u as e, o and ā is labialized . Development of consonants shares some simʧὕaʲл ʤʣatuʲʣʳл ʷʧtʦл ἵaʲāchХчл andл ʧnл aл ʷʧdʣʲл ʲanʥʣл aὕʳoл ʷʧtʦл πa₎aл dialects or North Western Iranian. Word-initial voiceless fricatives *f-, *ϑ-, *x- were probably preserved (attested is only x-), *f, *x were preserved word-internally, but *-ϑ-, *-ϑu-, *-ϑ - > ø. Voiceless stops (except *k) were sonorized, they later merged with * , * , ŋž and then changed to w, ø, ž/ʒ; ŋč, ŋǰ were often depalatalized > c~č, ʒ~ǰ. Word-initial ŋu- changes to (°)- or to ǰ- when palatalized; *-fr-, *-ϑr-, *-xr- > ṛ99; *ft, *xt > *tt > ø (but *xt sometimes > k); *rt, *rϑ, *rd > l; ŋ⁾ , *rʦ, ŋ› > ; ŋ⁾ n > ṇ; intervocalic *- - > y, ø; *h is lost, but initial *h- may be pʲʣʳʣʲvʣdлϐʣʤoʲʣлaлʳtʲʣʳʳʣdлvoʷʣὕлʧnлtʦʣлdʧaὕʣϑtлoʤлἰānХʥuʲāmщлуMORGENSTIERNE 1929, 322-339; EFIMOV 1999b, 278; SἰηÆοσØ 1989a). Except sibilants, there are no retroflex sounds in genuine ŌʲmuṛХл ʷoʲdʳчл ϐʣʳʧdʣл paὕato-alveolar affricates there are also alveolar affricates (the second mʣntʧonʣdлaʲʣлnotлpʲʣʳʣntлʧnлἵaʲāchХфщлρoлtʦʣлʳoundл in the dialect of KānХʥuʲāmлϑoʲʲʣʳpondʳл ⁾ ʧnлἱōʥaʲлŌʲmuṛХл(EFIMOV 1999b, 278). ŌʲmuṛХл dʧaὕʣϑtл oʤл KānХgurām distinguishes masculine and feminine gender; in plural animates and inanimates are distinguished. Umlaut or palatalization occurs quite frequently in inflection, cases are often expressed syntactically. Personal pronoun of the first person has direct and oblique cases, other persons have just one form for both cases. Demonstratives are used also for the third person pronouns, they are declined in three cases: nominative, accusative-objective and possessive. σʣʲϐaὕлmoʲpʦoὕoʥyлʧʳлʧnлϑommonлvʣʲyлʳʧmʧὕaʲлtoлἵaʲāchХлandлPashtō – there are two verbal stems: present and past. (EFIMOV 1999b, 281-296; KIEFFER 1989, 454-451). Morphologyл oʤл ἱōgar ŌʲmuṛХл ʷaʳл ϑonʳʧdʣʲaϐὕyл simplifiʣdл ʷʦʣnл ϑompaʲʣdл toл tʦʣл ἰānХgurām variety (MORGENSTIERNE 1929, 313). I.1.1.5. Other Eastern Iranian languages Apart from the above mentioned languages, various scholars mention some other languages that can be considered as members of the Eastern Iranian branch. Project Ethnologue lists WaʲdōjХ 100 – a language of the Wardō₍л ʲiver valley noʲtʦʷaʲdʳл ʤʲomл ḳēϐā₎л ʧnл Afghan This sound can be transcribed also ʳ, the sound should be similar to Czech voiceless ř (BURKI 2001), phonetically [ ]: voiceless retroflex non-sibilant fricative. Similar sound but voiced occurs also in the λūʲʧʳtānХл languages. 100 However, it is possibὕʣлtʦatлtʦʧʳлʧʳлmayлϐʣлanotʦʣʲлnamʣлoʤлtʦʣлḳēϐā₎Хлὕanʥuaʥʣл– tʦʣлϑʧtyлoʤлḳēϐā₎лὕayʳлonлtʦʣл rivʣʲлτaʲdō₍щлἴnлtʦʣл ʷʣϐʳʧtʣл http://globalrecordings.net there is given a record of biblical story about the Great Flood inлτaʲdō₍Хл(with an alternative name ḳēϐā₎Х; URL: http://globalrecordings.net/en/language/3400, cit. 24. 3. 2012, 13:37). When I compared this recording with Ish₎āshmХлandлπanʥὕēchХлζлϑanлtʣὕὕлtʦatлτaʲdō₍Хлʳoundʳлmuϑʦл different from Ish₎āshmХ-πanʥὕēchХл уʷʦʧϑʦл ʳʦouὕdл notл ʦappʣnл ʧnл ϑaʳʣл oʤл ḳēϐā₎Хл aʳл aл dʧaὕʣϑtл oʤ Ish₎āshmХл andл πanʥὕēchХфщлρoлmyлʣaʲʳлτaʲdō₍Хлʳoundʳлmoʲʣлὕʧ₎ʣлaлὕanʥuaʥʣлoʤлtʦʣлShughnХ-οōshānХлʥʲoupщ 99 ·59· Badakhshānлʷʧtʦлappʲoʸʧmatʣὕyл 4000 speakers. The language is not classified precisely, but it mayл ϐʣὕonʥл toл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳ (ETHNOLOGUE, 318). Based on analysis of toponyms of Tajik ἶaʲōtʣʥХnлandлDarvōz and Afghan Darwāz can be assumed that also in these regions there ʦaʳлϐʣʣnлʳomʣлἵāmХʲлὕanʥuaʥʣʳлoʲлdʧaὕʣϑtʳлуoʲлὕanʥuaʥʣʳъdʧaὕʣϑtʳлϑὕoʳʣὕyлʲʣὕatʣdлtoлtʦʣmфлʳpo₎ʣnл in the past (PAYNE 1989, 420), in case of *αaʲʷāzХ we can analyse toponymy of both Tajik and Afghan αaʲvāzлϐutлaὕʳoлtʦʣʲʣлaʲʣлʳomʣлʳuϐʳtʲatʣлʷoʲdʳлʧnлαaʲvōzлdʧaὕʣϑtʳлoʤлρā₍Х₎Х, some other substrate words appear in QarōtegХn Tajik dialects101. Georg Morgenstierne lists a hypothetical group of Southeast Eastern Iranian languages, from which could have developed *Proto-ἵaʲāchХл andл *Proto-ŌʲmuṛХ, relicts of this language may be observable in lexical borrowings in Pashtōл andл ʧnл tʦʣл λūʲʧʳtānХл ὕanʥuaʥʣʳ (MORGENSTIERNE 1926, 14-39; MORGENSTIERNE 1983b; KIEFFER 1989, 451-454), There is also an assumption that the 3rd and 5th version of inscription from Afghan Dasht-e Nāʷoʲ could have been attempt to write this unknown language with an adaptation of the Kʦaʲō ṭʦХ script (MORGENSTIERNE 1983b; FUSSMAN 1974)чл δéʲaʲdл γuʳʳman suggests for this hypothetical Southeast Eastern Iranian language label Kambojian (K‘m’oǰД), after Iranian tribe of the Kambojians, ʷʦoлpʲoϐaϐὕyлdʷʣὕὕʣdлʧnлaʲʣaлoʤлʷʣʳtʣʲnлεʧndū₎ush (FUSSMAN 1974, 32-34). I.1.2. Classification of the (Eastern) Iranian languages The Iranian language family is conventionally divided into two basic groups – Eastern and Western Iranian. Differences between these two groups begun to appear probably in the Old My assumption was confirmed by ShughnХл ʳpʣa₎ʣʲл γōkhʧʲл Ḳūʳuʤϐē₎ovл уʳonл oʤл ρā₍Х₎л ὕʧnʥuʧʳtл ShōdХkhōnл Ḳūʳuʤϐē₎ovчлʷʧtʦлʷʦomлʷaʳлtʦʧʳлmattʣʲлϑonʳuὕtʣdфлandлοāshāʲvХлʳpʣa₎ʣʲлGhuὕōmshōлAὕХnazaʲovл– the informants have stated that the language of the record is ShughnХлmʧʸʣdлʷʧtʦлοōshānХл– tʦʧʳлτaʲdō₍Хлcan be characterized as ShughnХлʷʧtʦлοōshānХлaϑϑʣntлandлʳomʣлοōshānХлvoϑaϐuὕaʲyчлonлtʦʣлotʦʣʲлʦandлϐotʦлʧnʤoʲmantʳлʳtatʣdлtʦatлtʦʣyл ʦavʣлnʣvʣʲлʦʣaʲdлaϐoutлτaʲdō₍Хлуϐotʦлoʤлtʦʣmлϑomʣлʤʲomлtʦʣлρā₍Х₎лϐan₎лoʤлtʦʣлʲiver Panj), according to words of GhuὕōmshōлAὕХnazaʲovлtʦʣʲʣлaʲʣлʳomʣлvʧὕὕaʥʣʳлonлtʦʣлShughnХ–οōshānХлὕanʥuaʥʣлϐoʲdʣʲлʷʦʣʲʣлtʦʣлpʣopὕʣлʳpʣa₎лʧnл a mixed language that is not so diffʣʲʣntл ʤʲomл Ba₍ūХл уγōkhir YŪπςγBῦἰἴσ and Ghuὕōmshōл AἱФλAḳAοἴσ, pers. comm., 24.-26. 3. 2012). Another informant – Ish₎āshmХ speaker Mu ammadл Bōduʲϐē₎ovл – stated that the language of the record is quite similar to Ish₎āshmХлoʤлρajikistan, but there are diffʣʲʣnϑʣʳлmaʧnὕyлʧnлὕʣʸʧϑonчлʷʦʧϑʦлʧʳлϑommonлʧnлπanʥὕēchХл and Yidghāлуʳʧϑмчлtʦʣлʧnʤoʲmantлpʲoϐaϐὕyлmʣantлκun₍Х;лκu ammad BŌαςοBῦἰἴσ, pers. comm., 2. 4. 2012). ζʤл tʦʧʳл tʦʣoʲyл ʧʳл ϑoʲʲʣϑtл tʦʣnл τaʲdō₍Хл ʧʳл notл aл ShughnХ–οōshānХл mʧʸʣdл ὕanʥuaʥʣл ϐutл ʧtл ʧʳл ʲatʦʣʲл aл Ish₎āshmХ-πanʥὕēchХлὕanʥuaʥʣлʷʧtʦлShughnХлandлοōshānХлadmʧʸtuʲʣчлʳuϑʦлtʦʣoʲyлmayлϐʣлʳuppoʲtʣdлϐyлʷʧtnʣʳʳлoʤл George Abraham Grierson, who stated that: «Th“ t›‘ct o” Zē’‘k s on“ o” th“ most pol₍glot spots n th s p‘›t o” As ‘.» (GRIERSON 1920, 3). Based on the above mentioned facts, it is necessary to critically examine the source of the recording; a question is how credible is the source published on the Web, how reliable was the informant уʣʳpʣϑʧaὕὕyлʷʧtʦлʲʣʥaʲdлtoлtʦʣлdʣʳʧʥnatʧonлoʤлḳēϐā₎Хлaʳлanлaὕtʣʲnative name), or to what extent was the author of the recording competent in linguistics. 101 αaʲvōzл dʧaὕʣϑtʳлaʲʣлϑὕoʳʣл toл otʦʣʲл ρā₍Х₎лdʧaὕʣϑtʳлoʤлtʦʣл ἵāmХʲлaʲʣaлуʣщʥщлdʧaὕʣϑtлoʤлσan₍л oʲлσakhʧyōyʧлBōὕō;лʳʣʣл RASTORGUEVA 1964, 4, 162). Question of classifiϑatʧonл oʤл хἶaʲātʣʥХnХл ʳuϐʳtʲatʣл ʷʧtʦʧnл tʦʣл βaʳtʣʲnл Iranian languages – ἶaʲōtʣʥХnлρā₍Х₎лϐʣὕonʥʳлtoлπoutʦʣʲnлρā₍Х₎лdʧaὕʣϑtʳлуRASTORGUEVA 1964, 5, 161), it has some ties with Upper Mastchōʦлdʧaὕʣϑtʳлoʤлρā₍Х₎луKHROMOV 1962, 16). ·60· ἵāmХʲ / Central (?) Eastern Iranian South Eastern Iranian *b-, *d-, *g*- mountain fish arrow dog North Eastern Iranian *ft, *xt Southeast Eastern Iranian *-b-, *-d-, *-gŋ- North Western Iranian *ʦ, *ʣ, *ʦu South Western Iranian Iranian period and became more distinctive in the Middle Iranian period. Each of these groups later split into two subgroups – South and North subgroup. In the North Western Iranian languages we can find e.g. Median (Old Iranian period), Parthian (Old and mainly Middle Iranian period), Old ĀzarХ, BalōchХ, Kurdish, Zâzâki (Dimli), GōrānХ, dialectʳл oʤл πʣmnān (SemnānХ, SangesarХ), dialects of Central Iran (ĀshtХyānХ, VafsХ, KhwānsārХ/KhūnʳāʲХ, NatanzХ, Borū₍erdХ, YazdХ, KermānХ, πХvandХ, KhūrХ etc.), Caspian dialects (κāzandʣʲānХ, δХla₎Х, GorgānХ), ρālyshХъρāὕəshХ, ρātХ, Khō‚ХnХ and many others. South Western Iranian languages are represented by Old Persian, *Old ShХʲāzХ (in the Old Iranian period), Middle Persian–PahlavХ (in the Middle Iranian period); in the New Iranian period there are varieties of Modern Persian (Classical Persian (Fā›s -yi dar ), Contemporary Persian of Iran (F rs ), Afghan Persian (P ›s -ye Dar ) and Tajik Persian (T jДkД), and non-literal or sub-standard forms of Persian such as HazāraʥХ, (Chā›-)AymāqХ, HerātХъεaʲavХ, ἰāϐoὕХчл πХstānХ, Bukhāʲ(āy)Хчл ἵāʲʳХ oʤл ἵāmХʲл etc.), dialects of Fāʲʳ (ρā₍Х₎Х of Iran, BūshehrХ, Dashta₎Х, KondāzХ, MāsaramХ, SamghānХ/SomghūnХ), ἱāʲХъἱārestānХ, ShХrāzХ, LurХ/LorХ, BakhtiyārХчлBandaʲХ, umzārХ, ἰāzerūnХ and others. Among the North Eastern Iranian are classified Scythian dialects and *Sauromatian (in the Old Iranian period), Sarmatian, Alanic, Sogdian (Middle Iranian period) and Ossetic and YaghnōϐХ (New Iranian period). South Eastern Iranian languages are represented by dialects of the Saka (mainly Khōtanese and Tumshuqese), Bactrian (Middle Iranian period), the ἵāmХr languages (ShughnХ-RōshānХ group, YazghulāmХ, Wan₍Х, WakhХ, IshkāshmХ-πanʥὕēchХ and Mun₍Х-Yidghā), Pashtō and Wa etsХ (New Iranian period). Questionable is classification of the Avestan language – it is probably one of the South Eastern Iranian, Khʷāʲezmian is variously classified as North or South Eastern Iranian; the most complicated is classification of ParāchХ and ŌʲmuṛХ – some scholars claim them as North Western Iranian but some other hive off new – Southeast branch within Eastern Iranian. *ϑ, *d, *s *s, *z, *sp *s, *z, *sp *s, *z, *sp *s, *z, *sp *s, *z, *sp * >*b, * >*y, * >*g ŋǰ * ,*,* ŋǰ *y * ,*,* *y * ,*,* *y * ,*,* *y * ,*,* *y *ft, *xt *ft, *xt * d, * d * d, * d * d, * d * d, * d *b, *d, *g ŋ * ,*,* ŋ *b, *d, *g *b, *d, *g ŋ ŋ ŋk‘u”‘ʦ ‘*tigra*ʦuą-ka- * ,*,* ŋž *gariŋpāϑa- ? *kuta-, ŋkutД Table 30 Basic isoglosses of the Iranian languages. ·61· * ,*,* ŋž *ʦuą-ka- South Western Iranian languages and dialects differ from other Iranian languages by significant isogloss Ir. *ʦ, *ʣ, *ʦu > *ϑ, *d (< * ??), *s; such isogloss, however, does not separate North Western Iranian languages from Eastern Iranian, cf. development of Ir. *ʦ, *ʣ, *ʦu > *s, *z, *sp 102 . Differences between the (North) Western Iranian and Eastern Iranian have to be looked up within other features. Some basic isoglosses between the branches of the Iranian languages are summarized in Table 30. However, according to the isoglosses shown in Table 30, distinctive features cannot be found only on phonological level. There were not many phonological differences between the Eastern and Western Iranian in the Middle Iranian period, one of the essential features was development of word-initial voiced stops *b-, *d-, *g- and development of clusters *ft and *xt. To establish a border between the Eastern and Western Iranian, lexical (e.g. in many works presented example *gari- ×лŋk‘u”‘- mountain and *kap - ×л ʦ ‘- fish ; cf. SIMS-WILLIAMS 1989a, 168-169) and grammatical differences should be also taken into account. There can be mentioned some other words from lexicon that can be considered typical for the Eastern Iranian area: √ yr √’₍› /√ ХʲъчлKhʷāʲ. yr-, Bactr. *abi-ar- to find, to obtain > Sogd ʾ yr-, Yazgh. vir-, Yagh. vД›-; *(h)ánd‘- blind > Khōt. hana-, Sogd ʾnt ʾnd /aṁd/, Munj. ₍āndəy, Pasht. ṛ nd, Ōʲm. h nd (but cf. Parth. hand); ŋ‘u‘-sú⁾t‘-(ka-) clean, purified > Khōt. Tumshuq. vasuta-, Sogd ʾwsw ty, ʾ⁽s(ʾ)⁽ tʾk ʾwsw tyy Bactr. ъōʳu d/, Oss. (without prefix) s dæg ‖ su dæg, Khʷāʲ. (with other prefix) (ʾ)”s d; w-y йžə⁽íй, Yagh. d‘›áu ‖ dⁱ›áu, Oss. æ›du ‖л ŋd›áu‘- hair > Khōt. drau-, Sogd æ›do, Shugh. cД⁽, οōsh. c ⁽, Yazgh. c Ōʲm. d›Д ×лἵʣʲʳщлm < ŋm‘ud‘-; ŋgá› - mountain > Khōt. ggara-, ggari-, Sogd r-y ъ əʲíъчлBactr. чл / Хʲчл aʲ/, Yagh. ar, Shugh. žД›, Wakh. ar, Munj. ā›, Pasht. ar, Ōʲm. g›Д, Parāch. gir ×лἵʣʲʳщлk h < ŋk‘u”‘-; - fish > Khōt. k‘vā-, Sogd kp-y /kəpíъчл Khʷāʲ. k b, Scyth. ( ) , Oss. kæ”, Wakh. k p, Munj. k p, Pasht. kab ×лἵʣʲʳщл < ʦ ‘-; *kąta- house > Sogd. ktʾ₍, ktʾk qt, qty(y), ktyy qṯy /kə , Bactr. у ф /kad(a)g/, Yagh. kat, Shugh. čДd, οōsh. Khūʤ. čod, Bart. čȫd, οāshrv. čǖd, SarХq. č“d, Yazgh. k d, Munj. ḱay, Yidgh. k i, Ave. kata- (+ Parth. Pahl. kdg) ×л ἵʣʲʳщл ⁾āná < -ka- (but Sogd. ʾnʾk(h) ⁾ʾnʾ Wakh. xun, Ishk. xon, Sangl. ⁾ān); ŋkút‘-, ŋkutД- dog > Sogd ʾkwt-y kwt-y, qwt-y / ₎{ətíъчл Bactr. /kud/, Yagh. kut, Oss. k ʒ ‖лkuy, Shugh-οōsh. kud, SarХq. k d, Yazgh. k°od, Ishk. kьd ×лἵʣʲʳщл sag < *ʦuą-ka-, Med. луbut Khōt. śv“, Wakh. ‘č, Pasht. spay (f. spə ), Wa . spa (f. spД), Ōʲm. ᵊspuk, Parāch. ᵊspȫ); 102 But in Wakhī *ʦu > and in Khōtanʣʳʣ *ʦu > śś [ʆ]. ·62· ŋmá ϑa- day > Sogd my m(ʾ)₍ my , my(y) myϑ, mỿϑ, myd ъmē /, Khʷāʲ. my ъmē /, Yagh. mēs ‖ mēt, Shugh. mēϑ, οōsh. Khūʤ. Bart. οāshrv. mДϑ, πaʲХq. maϑ, Yazgh. miϑ, Ishk may, Sangl. mē , ḳēϐ. mД, Munj. Yidgh. mД⁾ ×лἵʣʲʳщл› ₎ < ŋ›áuč‘- (but Pasht. wraʒ, rwaʒ, Wa . wrez, Ōʲm. wriez, wrioz); *pati-gaʣ- to accept > Khōt. p‘jā₍s-, Sogd √pc ʾ(ʾ)₎ ъ√pə Khʷāʲ. pc ʾz-; *ʦ - enemy > Khōt. Tumshuq. sān‘-, Sogd sʾn ъʳānъчлOss. son ×лἵʣʲʳщл du mán < ŋdu -mana- (cf. SIMS-WILLIAMS 1989a, 169; SIMS-WILLIAMS 1996b [online]). Eastern Iranian languages also borrowed some Indo-Aryan words (in this case old loans are meant, not loaned Buddhist terminology, which appear in several Eastern Middle Iranian languages): Skt. ākāś‘- heaven > Khōt. ātāś‘-, Sogd. ʾʾkʾc(h) maraṇa- death > Khōt. maraṇa-, Bactr. (adj.) /mara ʧ ʥъ; Skt. markaṭa- (Prkt. makkaḍa-) monkey > Khōt. makala-, Sogd. mkkr(ʾ) mk›ʾ /makká ṛ уáф/, Khʷāʲ. mrk; Skt. puṇya- merit > Khōt. puñ‘-, Sogd. pw(r)n₍ʾn(h), p⁽(›)n₍ʾn₍h /pu , Bactr. pwwn /pu /. Some of the above mentioned Indo-Aryan words are found in North Western Iranian Parthian (ākāś‘- > Parth. ʾʾgʾc ъāʥāčъ; maraṇa- > Parth. mrn /mara /; puṇya- > Parth. pwn /pu /; SIMS-WILLIAMS 1989a, 169), Parthian also borrowed Eastern Iranian word *pāϑa- arrow > Parth. pʾẖ ъpāʦъл(SUNDERMANN 1989, 112) – such fact is probably due to a long-time contact of historical Parthia (modern date South-western Turkmenistan and North-eastern Iran) with Khʷāʲezm, Bactria, Sogdiana andлδandʦāʲa. Division of Eastern Iranian languages into Northern and Southern branch (and eventually South-eastern ϐʲanϑʦлʧʤлʷʣлʷʧὕὕлϑonʳʧdʣʲлŌʲmuṛХлandлἵaʲāchХлaʳлmʣmϐʣʲʳлoʤлtʦʣлβaʳtʣʲnлIranian languages) is often used by many scholars, only few of them explain the criteria of such classification, so it seems that this division was more based on (modern) geographical distribution of the Eastern Iranian languages. For example Vera Sergeevna Rastorgueva lists the following criteria for dividing the Eastern Iranian languages: «Basic features of the North Eastern [Iranian] languages: 1) ending of plurals of nouns -t (in Kh⁽ā›ezmian -c < -t): Sogd. ›ʼt›t brothers , Kh⁽ā›. nikanc stakes , a ›āc eyebrows , Oss. ⁾æʒæ›ttæ houses ж ’ælæstæ trees , Yagh. odamt people , žutot sons ; 2) preservation of Iranian post-vocalic d; e.g. Ir. pād‘ foot is reflected as Sogd. pʼ , Yagh. podáж Oss. fad; 3) preservation of Old Iranian cluster dv word-initially; e.g. Ir. dvara door , is reflected as Sogd. wr, Yagh. dⁱv‘›ж Oss. dwar; One of the basic features of the South Eastern group is sonorization of Old Iranian consonant ; e.g. Ir. word g‘u ‘ ear , is reflected as Shugh. , sh. ⁽ж Pasht. wa , Yazgh. əvón and other.» (RASTORGUEVA 1966, 198) From the above mentioned characteristics only two can be confirmed – typical feature for the North Eastern languages is formation of plural with originally abstract suffix *- (such suffix ·63· can be found also in YazghulāmХл andл ʧnл ʳomʣл non-productive forms in Ish₎āshmХ) and sonorization of intervocalic *- - in South Eastern Iranian. Other presented features are not distinctive for both groups. Comparation of selected sound changes and other features can characterize some isoglosses in the Eastern Iranian languages. As can be seen in Table 31, some changes are common for many of these languages regardless to their ranking to the Northern or Southern branch. Based on a comparison of isoglosses listed in Table 31, instead of classification of the Northern and Southern branch, there can be better postulated a dialect continuum than two different branches; the only (?) branch that seems to show more distinctive features is the South-eastern ϐʲanϑʦлʷʦʧϑʦлϑontʧnuʣʳлʧnлtʦʣлŌʲmuṛХ-ParāchХлʳuϐʥʲoup. As distinctive features of the South Eastern Iranian branch can be considered 1) preservation of archaic formation of plural (i.e. absence of innovation of plural form by adding an abstract suffix *- ); 2) sonorization of intervocalic *- -; 3) change of Ir. *rd, *rt; 4) change of Ir. *rʣ, *rʦ and 5) emergence of innovated form of the second person plural personal pronoun from combination of forms of the second person singular and first person plural. All the above mentioned changes have not emerged in all South Eastern Iranian area: feature 1) have not took place in YazghulāmХ (and except some non-productive forms in Ish₎āshmХ); intervocalic *- - has not been sonorized in Bactrian and probably also in SarghulāmХ; changes under the point 3) have not taken place in Bactrian and WakhХ; in MunjХ, Yidghā and WakhХ (and probably also in Bactrian) has not taken place change point 4); innovated forms of plural the second person plural (point 5)) are present in all South Eastern Iranian languages, ϐutлʧnлἵaʲāchХлtʦʣyлϑomʣл from different source than from the above mentioned. ἵaʲāch.-Ōʲm. V Pasht.-Wa . - IV Khōtщ - II Wakh. preservation of *rz, *rs < *rʣ, *rʦ - III Ishk.-Sangl. + III Bactr. + Munj.-Yidgh. + Sargh. Khʷāʲщ + Shugh.-οōsh. Oss. ː Yazgh. Yagh. preservation of *rd, *rt, Wanj. Sogd. III Ave. I + - + + - - + ː + + - - - + ? - + ? - - innovated form of 2 pers. pl. pers. pronoun - - - - - + + + + + + - + ː preservation of ŋV V + + + - - - - + - + - ː - - - ā-Umlaut - - - - - + + ? + + + plural ending in ŋ-t(u -/*-ϑu - - + + + ː - + - - - ː ? - - - ŋu- > *gw- - - - - - - - - - - - - + - + preservation of diphthong in *ϑr a- + ? + * ? + + + + + + - * - * , ŋu > *w - - - - - - - - - - - - - + + palatalization of *t - ː ː + + - + - ː - - - ː ː ː second palatalization of velars - - - ː ː - + + + - - - - ː - sonorization of *p, *t, *k, ŋč - - - + + + + - + + + ː - + ː depalatalization of ŋ , ŋž, ŋč, ŋǰ - - - + + ? - + - - + ː ː ː + emergence of cerebral sounds - ː - - - ? * * ː - + + + + + augment + ː + - ː - - - - - - - ? - - nd ·64· - ? + ἵaʲāch.-Ōʲm. - ː - - + + + ? + ː + + - - + + + + + + + + + - - + + + + + + + - - ː - + + + ? + + + + + - - - - - - - - + + + - - - + - - - + + + + + + + + + + + - ? - - ː ː - - - ː - - + + + ː ː - + ː ː - - - - - - - - - ː - + - + + + + + ː + - + - - - - - - - - + + - + - + - - - - - - + - - - - + - - - - - - - + + + + + - - - - + - - - + + + - + - + + + - + preservation of gender + + - - + *ʦu > *sp + + + + + preservation of ŋdu + + + + i-Umlaut + + + ? u-Umlaut + + - preservation of *ϑ + + - - + *b, *g, ŋǰ > * , * , *ž ː + + + * > *d - - + + * > *l - ː - preservation of word-initial *b, *d, *g, ŋǰ + - - ŋ m > *m - ː voicing of initial *fr, *ϑr, *xr - - - 3 pl. verbal ending *-ā›- + - preservation of initial *h- + vocalic outcome of *- - -stems - ? - Bactr. - ː - ergative Sargh. - + Khʷāʲщ - ː Oss. - ː Yagh. - ː Sogd. + Ave. Pasht.-Wa . V Khōtщ IV Wakh. II Ishk.-Sangl. Munj.-Yidgh. III Yazgh. III - labialization of tectals rd Shugh.-οōsh. III Wanj. I ː ː Table 31 Isoglosses in the Eastern Iranian languages (plus (+) or minus (-) signs mean operation/absence of such change; asterisk (*) means that this change can be observed only with regard to the historical development of the language(s); plus-minus sign (±) indicates, that such change has not operated in full extent; question mark (?) means that according to attested material it is impossible to judge about operation of such change; text in gray indicates innovation when compared to the older state). I have outlined new classification in the note nr. 48 (Chapter I.1.1.4.b.). The Eastern Iranian languages can be divided into five branches: I Northern (Sogdo-Scythian) group; II Northeastern (Saka) group, III Central (PāmД›) group, IV Southern (Paṭhān) group and V Southeastern (ђ nd kush) group. Group I can be defined by innovated plural ending *- - (comparable to YazghuὕāmХфчлpʲʣʳʣʲvatʧonлoʤлʧntʣʲvoϑaὕʧϑл*- - (shared with Bactrian and WakhХлϐutлʣʸϑὕudʧnʥл Ossetic). Groups III, IV, V have undergone common change of form of the second person personal pronoun, in languages of these groups there are innovated forms of plural, they may be influenced by Indo-Aryan or Dardic pronouns. Innovated forms of the second person plural often comes from combination of personal pronoun of the second person singular with form of the first person plural *ta/u-*ah -(k/xam-), or *ta/u- ma- copied from Indo-Aryanлуϑʤщлκaʧyāл tus; i āлtsa/o;лἱaʦndāлtus) diffʣʲʣntлʤoʲmлʧʳл₍uʳtлʧnлἵaʲāchХщ Groups II and IV share sonorization of word-initial *fr-, *ϑr-, *xr-. ·65· II. Archaism and innovation in Sogdian and Yaghn bī According to the outline of the Eastern Iranian languages presented in the previous chapters one can state that there are four dozen extinct or living Eastern Iranian languages and dialects. Majority of those languages can be studied mainly from synchronous point of view – these languages and dialects are attested as individual stages of the Eastern Iranian branch but with some exceptions we do not know their older development stages. There is exception within the North Eastern Iranian branch – in this case both YaghnōϐХлandлἴʳʳʣtʧϑлϑanлϐʣлϑompaʲʣdлʷʧtʦл their closely related ancestors. The development of Ossetic can be continuously observed from the Old Iranian period – there are many similar features in the Scytho-Sarmatian dialects and in Alanic that can be compared with Ossetic and we can even suppose that Ossetic is a modern descendent of one of Alanic (or Sarmatian or even Sauromatian) dialects. Similar situation applies for Sogdian and YaghnōϐХ – these two languages are very similar from many points of view, YaghnōϐХл ʦaʳл ϐʣʣnл ʣvʣnл ὕaϐʣὕὕʣdл ḥ“o Sogd ‘n by some authors (BOGOLYUBOV 1956; KLIMCHITSKIY 1935; SἰηÆοσØ 1989a, 375-376), nowadays many scholars are inclined to believe that YaghnōbХл may come from some non-attested non-literary dialect of Sogdian (BIELMEIER 1989, 480; SIMS-WILLIAMS 1989b, 173), Aὕ bert Leonidovich Khromov expressed an opinion that YaghnōϐХлϑouὕdлʦavʣлoʲʧʥʧnatʣлʧnлaлnon-attested Sogdian dialect of Ustʲōshana (KHROMOV 1987, 645), unfortunately there is no relevant data to confirm this hypothesis. Some other New Eastern Iranian languages share several isoglosses with the Middle Iranian languages: Khōtanese and Tumshuqese share some isoglosses with WakhХлandлʳpoʲadʧϑaὕὕyлaὕʳoлʷʧtʦлtʦʣлotʦʣʲлἵāmХʲл languages; Bactrian shares many isoglosses with MunjХ and Yidghāл andл aὕʳoл ʷʧtʦл Pashtōл andл Wa etsХл or even with the ShughnХ-RōshānХл ὕanʥuaʥʣʳ. Khʷāʲezmian (whose affiliation to the North or South Eastern Iranian languages remains unsolved; see ÈDEL MAN 2000a, 95; ÈDEL MAN 2008, 6; ÈDEL MAN 1986, 6) is similar to Ossetic from one point of view and to Pashtōл andл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл ʤʲomл anotʦʣʲ;л anл ʣndʧnʥл oʤл tʦʣл tʦʧʲdл pʣʲʳonл pὕuʲaὕл oʤл subjunctive connects Khʷāʲʣzmʧanлtoʥʣtʦʣʲлʷʧtʦлtʦʣлπa₎aлdʧaὕʣϑtʳлandлʷʧtʦлḲaghnōϐХлʣndʧnʥлoʤл the third person plural of present, imperfect and non-durative preterite (SἰηÆοσØ 1989a). On the basis of the above mentioned data we can declare that a thorough diachronic and synchronic study of the Eastern Iranian languages is possible in its Northern branch – but in the case of Ossetic comparable material lies mainly in lexicon, development of grammar and syntax is blurred (cf. ABAEV 1949). It is of course possible to outline historical development of other (New) Eastern Iranian languages, but in these cases it is necessary to deal only with methods of historical and comparative linguistics because there are not attested direct ancestors of these languages. Based on the above mentioned facts the main theme of this thesis will be the comparison of Sogdian and YaghnōϐХ – information on Sogdian are available in a large corpus of texts from which we can learn about Sogdian grammar, lexicon and syntax; YaghnōϐХлaʳлaлὕiving language is so far undrawn repository of knowledge – to linguists YaghnōϐХлʧʳл₎noʷnлaлὕʧttὕʣлϐʧtлmoʲʣлtʦanл ·66· hundred years, within that period of time some texts, grammars and lexicons have been published, at the present time a research on the YaghnōϐХлὕanʥuaʥʣлandлʣtʦnoʥʲapʦyлʧʳлundʣʲл patronage of the Academy of Sciences of the Republic of Tajikistan, where under the Rudakî Institute of Language and Literature falls the Department of Yaghn ’Д Studies (Tjk. ёu› h yaghn ’sh n sî). The study of the Sogdian and YaghnōϐХлὕanʥuaʥʣʳ certainly cannot be separated from study of the other Eastern Iranian languages therefore I will also pay attention to interpretation of relevant innovations and archaisms in other languages and dialects of the Eastern Iranian branch. In case of YaghnōϐХ (and the other Modern Eastern Iranian languages except Ossetic) it is also necessary to follow development of Modern Persian, mainly its varieties in Tajikistan and Afghanistan103. A comparison of the Sogdian and YaghnōϐХ material can solve the issue of the relationship of both of these languages. It can be supposed that both languages developed from one common North Eastern Iranian proto-language or proto-dialect, such proto-language will be labelled *Proto-Sogdic (i.e. aлCʣntʲaὕлAʳʧatʧϑлvaʲʧʣtyлoʤл πϑytʦʧanъπa₎a лoʤл the late Old Iranian period) here. Later *Proto-Sogdic split into two (or even more) main dialects – *Proto-Sogdian and *Proto-Yaghn ’Д. Both *Proto-Sogdian and *Proto-YaghnōϐХлaʲʣл reconstructed as predecessors of the attested languages – Sogdian and YaghnōϐХ, besides those two languages there may have been Sogdian dialects of Bukhāʲāчл Ustʲōshana and Zhetisu – *Bukhāʲanл Sogdian is attested by several short texts, *Zhetisu Sogdian is attested on several inscriptions and from historical sources while хςʳtʲōshanian remains to be a hypothetical Early Mediaeval ancestor of YaghnōϐХчлхςʳtʲōshanian is also thought to be an ancestor of hypothetic *ZarafshānХлlanguage/dialect which remained as substrate in Tajik dialects of Mastchōʦчлγaὕghar andлγōn. 103 Development of Persian as a member of the South Western Iranian branch is surely not the theme of this work. For simplification the development of Persian will be observed on basis of following works – general development of Persian and its vernaculars was described by Valentin Aleksandrovich Efimov, Vera Sergeevna Rastorgueva and E. N. Sharova (EFIMOV – RASTORGUEVA – SHAROVA ьŚřэф;л ρā₍Х₎л ʥʲammaʲл ʧʳл tʦoʲouʥʦὕyл dʣʳϑʲʧϐʣdл ϐyл ηoʦnл PERRY (2005), grammar of Afghan αaʲХлʧʳлdʣʳϑʲʧϐʣdлϐyлἱʧdʧyaлλʧ₎oὕaʣvnaлKISELEVA (1985). Thorough description oʤлρā₍Х₎лdʧaὕʣϑtʳлʷaʳлpuϐὕʧʳʦʣdлϐyлσʣʲaлπʣʲʥʣʣvnaлRASTORGUEVA (1964). ·67· II.1. Historical phonology104 The *Proto-Sogdic language split into two reconstructible dialects – *Proto-Sogdian and *Proto-YaghnōϐХщл γoʲл dʣʳϑʲʧptʧonл oʤл tʦʣл ʦʧʳtoʲʧϑaὕл pʦonoὕoʥyл oʤл Sogdian it is necessary to outline several stages of development of the Sogdian language (see Table 32). th th th th 4 -5 cent. 7 -9 cent. th half of the 11 (?) cent. (middle ages) up to cca. 1900 from cca. 1900 *Proto-Sogdic *Proto-Sogdian / *Proto-YaghnōϐХ *Old Sogdian Preclassical Sogdian Early Classical Sogdian Classical Sogdian (& Bukhā›‘n d ‘l“ct) Postclassical Sogdian (& Zhetisu dialect) (death of Sogdian) *ZarafshānХ Early Modern YaghnōϐХ Contemporary YaghnōϐХ language of Sogdian translation of A əm voh the Ancient Letters Christian document C 2 majority of texts B›āhmД documents, Christian document C 5 preserved only in central Tajik dialects p›“s“›v‘t on o” m‘jh l ‘nd ϑ Table 32 Relative chronology of *Proto-Sogdic dialects. YaghnōϐХл andл πoʥdʧanл pʦonoὕoʥyл ʷʧὕὕл ϐʣл outὕʧnʣdл ʧnл aл comprehensive view. I will try to present all phonological changes of both languages. The main sources for the study of historical phonology of Sogdian and YaghnōϐХлʷʣʲʣлoutὕʧnʣʳлoʤлπoʥdʧanлandлḲaghnōϐХлʦʧʳtoʲʧϑaὕлʥʲammaʲл (LIVSHITS – KHROMOV 1981, 373-116; KHROMOV 1987, 653-66ыфл andл δκπл έřэ-530. In many case I have tried to find same responses both in Sogdian and in YaghnōϐХл ʤoʲл ϐʣttʣʲл demonstration of similar development of both languages. Before I start with historical phonology I will describe Sogdian orthographical system in order to explain possibilities of reconstruction of Sogdian phonology. (excursion 4) Sogdian orthographical systems Sogdian texts have been written in three various graphic systems: in the Sogdian, Manichaean and Syriac alphabets (see Table 33 to compare transliteration of the alphabets). The Sogdian script was a locally developed variety the Aramaic alphabet, this script was used in Sogdian (so-called Ancient Letters documents from approximately the first third of the 4th century th th found at Dunhuang in China) up to the 9 -10 centuries. The Manichaean alphabet was also a In the presented work the majority of Sogdian and YaghnōϐХл ʷoʲdʳл ʷʧὕὕл ϐʣл ʳuppὕʣmʣntʣdл ϐyл tʦʣʧʲл хἵʲotoIranian form – in this reconstruction I will transcribe continuants of some sounds in rather archaic state: for continuants of *Proto-Indo-European vocalic nasals, *ʦ, *ʣ for *Ide. ŋk, ŋg(ʰ) and sometimes I will use *ʜ for *Proto-(Indo-)Iranian continuant of *Proto-Indo-European laryngeals. Stress will be shown on majority of examples, but stress will usually marked in position of St›“ss II (see chapter II.1.1.), only in several cases position of St›“ss I (i.e. *Proto-Iranian stress) will be marked – such only in cases where it was known to me. I decided for such notation of stress for two reasons – 1) original position of stress in *Proto-Iranian is not marked in majority of reconstructed forms, and 2) marking of the position of Stress II is preferable for explanation of *Proto-Sogdic development. 104 ·68· modification of the Aramaic alphabet, according to legends the creator of this script was a pʲopʦʣtлκānХлуэь6-276 ), founder of Manichaeism; the Manichaean alphabet differs from the Aramaic original by number of new consonant graphemes – this alphabet was quite widespread, apart from the Sogdian texts there are attested also κʧddὕʣл ἵʣʲʳʧanл уἵaʦὕavХфчл ἵaʲtʦʧanл oʲл Bactrian (or even non-Iranian Tokharian and Old Turkic) documents written in the Manichaean script. Sogdian translations of Christian texts were written in Eastern (Nestorian, Esṭʲanʥēὕā) variety of the Syriac script, Sogdian adapted Syriac script was supplemented by three new consonant graphemes. All three scripts originated in the Aramaic alphabet so Sogdian orthographies were based on the model used for Aramaic and for other Semitic languages – alphabets of Semitic origin do not have special signs for vowels, vowels were either not written or written with consonant graphemes ( m‘t›“s l“ct on s – in Sogdian ʾ, y, w; and also ( ), h, k ( )). In the Syriac script diacritic vowel signs occasionally appear. Besides documents written in the Sogdian, Manichaean and Syriac scripts, there are also some Sogdian documents written in North Turkestan variety of the BʲāʦmХ script – reading of the Sogdian documents in the BʲāʦmХ script can considerably help with reconstruction of Sogdian sound system. In Aϐū-r-Ray ānл κu ammad bin A mad al-BēʲūnХ ʳл Ŭ tā’ ‘l-āthā› ‘l-’ā‹ ₍‘ẗ an al-‹u› n al-khāl ₍‘ẗ there are some Sogdian glosses written in the Arabic alphabet, also in an unnamed manuscript from the 13th century by Mu ammad bin Man ūʲл ϐʧn πaƒХdл κuϐāʲa₎л Shāʦл (Fakhr-i ḤudД›) we can find Sogdian adaptation of the Arabic alphabet together with several Sogdian glosses (ROSS – GAUTHIOT 1913), moreover Sogdian letters are also transliterated (in this case rendered for Old Turkic) by Ma mūdл ϐʧn usayn bin Mu ammad al-ἰāshghaʲХл ʧnл Ŭ tā’u dēvānu lughāt ʼt-t ›k. Aramaic alphabet ‚āὕap ϐēṯ Sogdian alphabet <> // ʾ , ə, f( ) ʥāmaὕ ,f f , x, h, q Manichaean alphabet <> // ʾ , ə, b (’) g b g Syriac alphabet <> // ʾ , ə, 105 b ,b g g, ( ) dāὕaṯ d - d d d ʦē h - ,ø h h, x h h, (- waw w w w, , u, , , z zayn ₎ w, , u, , , z, ž, ž z, ž, ž z, ž, ž w z ž (₎) w, , u, , , z ž, ž z , (d) z In the Syriac script can be observed some differences in reading of the letters ṭēṯ and taw: ṭēṯ is usually used for writing t (eventually d), but in several cases it is used also for ϑ <ϑ>; taw normally serves as a grapheme for ϑ, but it can be used also for t (d) <t>. Whether one of the other variant was used, it was consistent throughout the document, i.e. if ṭēṯ = t/d, thus taw = ϑ and vice-versa, if ṭēṯ = ϑ, then taw = t/d (the second variety is not common according majority of Christian texts). 105 ·69· (ž‘₍n) j ž, ž, ǰ ž ž, ž x, h, q (q) - ẖ - ,ø ẖ h ṭēṯ x (ẍ) ṭ ṯ t, d t, d (ϑ) yuḏ y y, , , , ə, ₎āp k k, g, - , - y k ⁾ (k) l y, , , , ə, k x l ṯ (ϑ) y k x l y, , , , ə, k x l ēṯ ὕāmaḏ l ,ϑ чл ( āl‘ṯ) mim m m, ṁ m m, ṁ m m, ṁ nun n n, ṁ n n semkaṯ s s, ( ) - s n, ṁ s n, ṁ s , , p, b, f f p f (ṗ) c () p f c ƒayʧnлълƒē āḏē p p c qop q č, ǰ, (ʦ), (ʣ) - ʲēš r l (ṛ) r, ṙ, ʳ, l l pē šʧn taw ( ām‘ḏ) t, d p, b f q č, ǰ, (ʦ), (ʣ) k q k, g r r, ṙ, ʳ r r, ṙ, ʳ , t p, b f s , t t, d č, ǰ, ʦ, (ʣ) , ϑ (t) ϑ, (t, d) ,ϑ Table 33 Overview of transliterations of Sogdian from the scripts derived from the Aramaic alphabet (after SIMS-WILLIAMS 1989b, 176 and KÜκκβἱ 2006; edited). Sogdian orthography of the Ancient Letters (written in an archaic non-cursive variety of the Sogdian script) corresponds to a rather archaic ( P›“-Cl‘ss c‘l ) form of the language, in which the *- -stems were neither contracted yet nor there have been change *ϑr, * rл şл ъšчл ž/ occurred. From Aramaic ductus was adopted writing of word-final - with letter hē, but it cannot be judged whether already in the language of the Ancient Letters operated Stress III and the Rhythmic Law. Younger (or Cl‘ss c‘l ) Sogdian texts from the 8th-9th centuries come from the orthography similar to the orthography of the Ancient Letters, but in these younger texts appear some orthographic doublets – word-final - (originally masculine aka-stems) was written either archaic as <-(ʾ)kш or phonetically as <-(ʾ)₍ш and word-final - (from originally unstressed feminine ākā-stems) was written as <-(ʾ)kh> or according to its pronunciation as <-ʾ(ʾ)> or <-(ʾ)hш, even - (in forms of adverbs, and accusative of masculines and nominative/accusative of neuter) was written as <-(ʾ)k⁽ш and word-final - of old ā-stems is often written as <-h> in endings of later heavy stems106; also sounds , (< *ϑr, * r) were often written archaically as < r> or by phonetically similar graphemes < ж ₎й й₎>. Texts in the Manichaean and Syriac alphabet It means that the grapheme hē had two functions: 1) it marked word-final - in forms of the light stems, and 2) it was used as a common marker of feminine nouns and adjectives (with no phonetic value); later also the third function was emerged – it was used as filler at the end of the line. 106 ·70· use rather phonetic spelling (if we can really use the term pʦonʣtʧϑл ʳpʣὕὕʧnʥ л ʧnл a case of a consonant script which does not have separate vowel graphemes) – reflexes of the unstressed ākā-stems were written by use of the letter ālap and reflexes of the aka-stems with the letter yuḏ; continuants of old *ϑr and * r were written with the letters n and ž‘yn . Interesting is an adoption of a grapheme for (and ϑ) – In the Sogdian alphabet /ϑ was written by Aramaic letter lāmaḏ, in the Manichaean script with the letter ālaṯ , which is morphologically derived from the letter lāmaḏ, but in the Syriac alphabet the sound is written as dāl‘ṯ and ϑ as taw (i.e. only in the Syriac script there are two separate graphemes for and ϑ), problem of Sogdian л<л *d : <l> ( lambda Sogdica ) will be discussed in excursion 5 in chapter II.1.3.6. With the exception of sibilants there were no different graphemes for opposition of voiced and voiceless consonants in the Sogdian script – voiced stops (which have been rather rare in Sogdian) were written with graphemes for voiceless stops; on the contrary voiceless fricatives were written with graphemes for voiced fricatives, an exception presented only x and , which had two separate graphemes: letters gāmal and ēṯ, these graphemes slowly merged and their forms were distinguishable only word-finally, word-initially and word-internally was the difference in shapes of gāmal and ēṯ hardly evident. Labial fricative f was written with two graphemes – with the letters bēṯ and pē, the first mentioned was used also for , the second letter was used also for labial stops p and b; occasionally the letters ’ēṯ and pē were supplemented with diacritics to spell f – bēṯ was supplemented by a subscribed dot or hook beneath the original letter, pē could have two dots written over the original letter (such way was used in Manichaean texts written in the Sogdian script). The letter zayn could have been also supplemented by diacritics – by either one dot/hook or two dots beneath the letter – these diacritic marks (without a distinction of < > and <z>) had two meanings – they either distinguished ž-sound or they kept apart the letter zayn from the letter nun (nun was always written without diacritics). In a later period a subscribed hook under the letter rē for l appears, this new grapheme is of Turkic origin and in Sogdian it has been used rarely (as there was no l in Sogdian). The Syriac alphabet has special graphemes for voiced and voiceless fricatives; and also the voiced velar stop g had its own grapheme gāmal (but g could have been written as qop), the other voiced stops were written either as voiceless stops or as voiced fricatives (i.e. d = ṭēṯ or dāl‘ṯ; b = pē or bēṯ). Only the Manichaean script had quite a full range of graphemes to represent Sogdian consonants (but the letter n was used for and and ž‘yn for ž and and except the letter āl‘ṯ which served both for and ϑ, but occasionally double āl‘ṯ < şлʷaʳл used for ϑ107), it was possible to distinguish stops clearly in writing, but voiced stops were often written as their voiceless counterparts. Moreover Aramaic had some phonemes that do not appear in the Iranian languages, mainly emphatic ṭ, , q and pharyngeal , . Letters for those sounds were used in different ways in Compaʲʣлʳʧmʧὕaʲлʷayлoʤлʥʲapʦʧϑлʲʣpʲʣʳʣntatʧonлoʤлъ ълandлъ ъ in Modern English – both sounds are written with a single digraph <th>. 107 ·71· Sogdian. The letter āḏē was used in all three alphabets for č and ǰ (and possibly for ʦ and its allophone ʣ). In the Sogdian alphabet the letter ēṯ was used for x, in the Manichaean alphabet ēṯ served as a line-filler and in the Syriac script it was used for h. The letter ṭēṯ was not used in the Sogdian script, in the Manichaean script it was interchangeable with the letter taw and in the Syriac script it was used for t (as taw has been used for ϑ). The letter qop had no use in the Sogdian alphabet, in the Manichaean and Syriac scripts it was interchangeable with the letter kāp (while kāp was used rarely in the Syriac script). The letter ayin was used in the Manichaean script for vowels , , ; in the Sogdian alphabet it was not used and in the Syriac alphabet it was used for . The Sogdian alphabet had no use for the letter dāl‘ṯ108. The alphabet order is known from the attested material – the order was the same as in Aramaic109. The collation of the Sogdian alphabet was found on an ostracon from Panjakent and on a fragment from the Ōtani collection from Japan (LIVSHITS 2008, 305), the alphabet order was as follows: ʾ d h w z x ṭ ₍ k l m n s p c ‹ › t 110. The alphabet order of the κanʧϑʦaʣanлaὕpʦaϐʣtлʧʳлattʣʳtʣdлʧnлtʦʣлκʧddὕʣлἵʣʲʳʧanлуἵaʦὕavХфлandлἵaʲtʦʧanлdoϑumʣntʳ: ʾ b g d ẖ w z j h ṯ y k l m n s p c ‹ › t – the graphemes < >, < >, <žш, <x> and <f> were not considered as separate letters of the alphabet, but as varieties of <b>, <g>, <z>, <k>, <p>, from which they differed only by supplemented diacritic marks (cf. BOYCE 1952). Unlike its Semitic original the Manichaean alphabet differed in collation of the letters hē and ēṯ which switched their positions. In the Syriac script the alphabet order is the same as in Aramaic, the collation of the Sogdian letters ž‘₍n , xāp and fē is not known but it can be suggested that they followed after the letters zayn, kāp, pē from which they were derived. The Sogdian alphabet was not used only for recording Sogdian language – it served also for Old Uyghur and later for Mongolian, Oyrat, Manchu or Sibe (Xibe) who use it up today. In In the Sogdian alphabet the non-used letters dāl‘ṯ, ṭēṯ, ayin and ‹op appear only in Aramaic ideograms. Thus ʾ b g d h w z ṭ ₍ k l m n s p ‹ › t. 110 Another interpretation of the collation is also … ₍ k m n … t l, by analogy after the Old Uyghur alphabet, where the collation is as follows: ʾ, v (Sogd. ), , (h ъøъфч w, z, q (Sogd. x) /q/ (rarely /x/), y, k /k, g/, d (Sogd. ) ъ ~d?/, m, n, s, p, c, r, , t /t, d/, l (Sogd. ṛ). In case of Old Uyghur digraph <nk> should be mentioned, which was used for a velar nasal . The Uyghur variety of Sogdian script used some other letters supplemented by diacritics – ēṯ, and kāp could have been written with two superscript dots, n and zayn with two subscribed dots and nun used single superscript dot – <‹> was used to distinguish the letter ēṯ from the letter gām‘l; < > to distinguish n form semkaṯ; <ṅ> distinguished nun from āl‘p; <₎> was used for ž in Sogdian loans; the use of <k> is not known to me. The Uyghur variety of the Sogdian alphabet has been adopted by the Mongolians, who changed the collation as follows: ‘ ( ʾ(ʾ) ), “ ( ʾ ), ( (ʾ)₍ ), oйu ( (ʾ)⁽ ), öй ( (ʾ)⁽(₍) ), n ( nйṅ ), ( nk ), ’ ( p ), p (n“⁽ g›‘ph c v‘› “t₍ o” p ), q ( ⁾ ), ( ẍй⁾ < ), k/g ( k ), m, l ( ṛ ), s, ( s < ), t/d (according to a shape of surrounding letters shape of the letter is ’‘s“d “ th“› on o› g n‘l t o› ), č ( c ), ǰ ( ₎ ), y, r, v/w ( ), f ( ṗ ), ḳ (n“⁽ g›‘ph c v‘› “t₍ o” k ) and also letters c, ʒ and h were probably adopted from the Tibetan script for Tibetan and Sanskrit words. The Sogdo-Uyghur alphabet has spread from the Mongolians to other nations such as the Oyrats, Manchus or Sibe; the Mongolian variety of the Sogdo-Uyghur alphabet and its local varieties are used even in the present time. 108 109 ·72· Sogdian translation of the Buddhist text Av‘lok t“śv‘›‘s₍‘nāmā ṭ‘ś‘t‘k‘stot›‘ a Sanskrit quatrain is recorded in the Sogdian script (Figure 6): Figure 6 Sanskrit inscription in the Sogdian script уBʧϐὕʧotʦèquʣлλatʧonaὕʣлdʣлγʲanϑʣчлἵʣὕὕʧotлϑʦʧnoʧʳлn° 3520, lines 53-54; http://gallica.bnf.fr/ark:/12148/btv1b8305780t/f2.image.r=pelliot+3520.langEN, cit. 12.9.2012, 10:16) (53) sr n sarvaṃ (54) sr n sarvaṃ tʾnʾn dānaṃ p›ʾn balaṃ trm tʾnʾn dʰaʲmadʰānaṃ k ʾnt₍ p›ʾn k āntʧϐaὕaṃ c₍nʾt₍ sr n rtym ₍ʧnātʧ | sarvaṃ ratiṃ c₍nʾt₍ t›ʾ₍ nʾ k ʾʾ₍ ₍ʧnātʧ | tʲ ā₎ aya trm rtym dʰaʲmaʲatʧṃ sr swkk ʳaʲvaʳu₎ʰa c₍nʾt₍ ₍ʧnātʧ c₍nʾt₍ ₍ʧnātʧ «The greatest of gifts is the gift of the law; the greatest of delights is delight in the law, the greatest of strengths is the strength of patience; the greatest happiness is the destruction of desire.» (GAUTHIOT 1911, 94). In this example characteristics of the Sogdian script can be seen – by comparison with Sanskrit whose sound system is well known, reading of individual graphemes can be verified – an effort to mark vowels i and u regardless their quantity is evident, but a similarly as in Sogdian is marked rarely; voiced stops were written with graphemes for their voiceless counterparts. Neither aspiration was marked (orthography <kk> for kʰ cannot be interpreted as an effort to mark aspiration – the first <k> probably marks velar, the second <k> probably stands for vowel -a). In case of the word t› ṇāk aya <t›ʾ₍ nʾ k ʾʾ₍ш we can presume that it is a scribal error 111 for ŋ<t›ʾ₍ nʾ k ʾ₍ʾш . The sound l was written with the letter rē , in many Sanskrit loans in Sogdian there is l often written with the letter lāmaḏ: Sogd ⁽kʾ, rwk / / world, loka < Ved. loká-. Sogdian texts written in the BʲāʦmХ script are quite different from the text in Aramaicderived scripts – Sogdian has adopted Central Asian variety of BʲāʦmХ as it has been used by the ancient Uyghurs, but in the case of Sogdian cannot speak about Sogdian literature in this script, only a dozen texts are known. The main advantage of the BʲāʦmХ script is its ability to mark vowel quality, however quantity is not marked. The Sogdian BʲāʦmХ documents are not dated well, but they can come from the later period of Sogdian and thus they can bring valuable information about the development of the language. In case of Sogdian written in the BʲāʦmХ script we cannot speak even about developed orthography, it is rather an effort to record Sogdian words in an orthography created for some Turkic language, presumably Old Uyghur, but there are several features that can tell more about the Sogdian sound system; reading of the BʲāʦmХ Sogdian documents have to be compared with The orthography of this word informs also about pronunciation of Sanskrit <›> /ri/ sometimes after the 8 century. 111 ·73· th other records in the Sogdian, Manichaean and Syriac alphabets. The North Turkestan BʲāʦmХ script used nine graphemes (ak ara112) for vowels and diphthongs: a /a, ə, (ɨ)/, ā ъāчл- , i 113, u 114 , › /əw/, e , ai йā /, o , au ъāu, u/. Other 33 ak aras were used for consonants: ka /k/, kʰ‘, ga, gʰ‘, ṅa, ca ъčъ, cʰ‘, ja ъǰчлčъ, jʰ‘, ñ‘, ṭa /ṭ, ṭʰ/, ṭʰa /ṭʰ/, ḍa / /, ḍʰa, ṇa /n, , ṁ/, ta /t/, tʰ‘ / -/, da, dʰ‘ ъ ъ, na /n, ṁ/, pa /p/, pʰ‘, ba, ’ʰ‘, ma /m, ṁ/, ya /y/, ra /r/, la, va /w/, ś‘ ъšчлšчлžчлž/, a ъšчлšчлžчлž/, sa /s/, ha ъʸчл ъ and there were 13 new graphemes: ъ ъ, ъ ъ, w ъ члʤъ, z /z/, ź /žчл ž/, ḵ /k/, ṯ /t, /, p /p/, m /m/, ṟ /r/, ḻ, u /w/, ъšчлš/ and three diacritic marks – anusvāra (ṃ /n/), virāma (sign that marks that after a consonant ak ara does not follow a vowel) and ä /-iчлø/. Beside the above mentioned ak aras there were used some digraphs, e.g.: ar /ar, əw/, ccʰ ъčъчлtt /t/, yu /ü /, yueлъü veлъü hk /q~x/, hv /ʸ°/, h ъʸšъчлhu /ʸ°/, wt ~ ⁽dʰлъ d/, wṯлъ d/, wv /f/, u“лъü etc. Based on the present state of knowledge we can hardly talk about literature in the Sogdian BʲāʦmХ script, yet even there we can trace certain orthographic conventions; e.g. for ə (and/or its allophone ) existed two different spellings – 1) in an open syllable of a disyllabic word the nd vowel ə/ was not marked: Sogd knā k⁽nʾ ‹⁽nʾ) /₎у{фənā/ do! (2 pers. sg. imper. pres.); Sogd mdʰu m w mdw) /mə úъ ʷʧnʣ ; Sogd prau ( p›ʾ(y)w p›ʾ(₍)⁽, pryw pryw prw) /pər u/ уtoʥʣtʦʣʲфл ʷʧtʦ ; Sogd hji /xəčíъ ψуʳфʦʣъʧtϊл ʧʳ ; Sogd n ā-m n₍dʾm) /nɨ husk, bark ; 2) in a closed syllable it has been written as a: Sogd h a ⁽dʰ , h a wṯi ʾ⁾ ₍ t-y) / ʸšɨ dí/ mʧὕ₎ ; Sogd dʰ‘ ⁽dʰ -k, dʰ‘ wṯi-k ( tyk tyk, tʾ₍k(⁽) t(t)yk, tyq, tyk dbṯyq) / ɨ d aʥaʧn ; Sogd p‘ t₍ā-p pt₍ʾp) /p paʲt ф. Interesting issue presents pronunciation and orthography of ϑ – it is written as the letter nd tʰ‘-kā›‘ (Sogd tʰ‘u / āu/ shoot (2 pers. sg. imper. pres.) ), in other positions it is written as F›“md₎“ ch“ ṯa-kā›‘, which is used either for or for t: Sogd me-ṯ ʾmy ʾ)mʾ₍ mʾ₍ mʾ₍ ( ) myϑ, myϑ) /mē / tʦuʳ ×л Sogd pcai-ṯ, pcā-₍tä, pcā-yt /p č ʧt/ √pcʾ₍ /p č ʧ/) is beneficial – it is possible that in Sogdian dialect recorded by the BʲāʦmХ script ϑ changed to t (i.e. similarly as e.g. in Sogdian dialects of Zhetisu; see Excursion 1), or it is determined by the fact that there was no ak ara for the voiceless dental fricative ϑ in the Old Uyghur BʲāʦmХ and thus this sound has been written with an ak ara for voiceless dental stop t (ta-kā›‘). In some words we find i and u instead of (etymologically) expected and , Nicolas Sims-Williams explains this change with an assumption that there was a stress shift to the last syllable (see Stress IV in chapter II.1.1.4.) and newly unstressed and were shifted towards and (SIMS-WILLIAMS 1996a, 310). Moreover, the BʲāʦmХ script shows pronunciation of the numeral one in Sogdian – it is attested as Sogd ʾyw ʾyw, yw yw, ẏw in the Aramaic 112 After each ak ara will be shown its phonetic value as it was pronounced in Sogdian. 113 114 ·74· derived alphabets but written yau in BʲāʦmХ, so it could have been pronounced as /| u/ 115 (SIMS-WILLIAMS 1996a, 313); also reading of Sogdian digraph <wy> have been corrected – Nicolas Sims-Williams originally suggested reading either or ȫ, after BʲāʦmХ orthography <yu, yueж v“ж u“>, the reading has been corrected as a rising diphthong ē or “ (SIMS-WILLIAMS 1996a, 313-314). Sogdian documents in BʲāʦmХ still wait for a thorough study, since just one Sogdian– Sanskrit bilingual document has been published (MAUE – SIMS-WILLIAMS 1991) together with some words quoted by Nicolas Sims-Williams to evaluate Sogdian phonology (SIMS-WILLIAMS 1996a). II.1.1. Stress Development of stress in the *Proto-Sogdic language is essential to understand phonology of Sogdian and YaghnōϐХ and also to discover differences between both languages. It is not necessary to focus on position of stress in *Proto-Iranian because there was a stress shift in *Proto-Sogdic from which both languages developed. The reconstruction of *Proto-Iranian stress is complex – it can be supposed that the *Proto-Iranian stress was mobile and its position was similar to Vedic. For the reconstruction of Old Iranian stress is essential to study stress in Pashtō (GRYUNBERG – ÈDEL MAN 1987, 38-39). Position of stress changed also in the other Eastern Iranian languages, maʧnὕyл ʧnл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл ʷʦʣʲʣл ʳtʲʣʳʳл ʳʦʧʤtʳл ϑauʳʣdл ʣʧtʦʣʲл syncopation of unstressed vowels or changes of stressed vowels under operation of ā- or i-Umlaut;лnoʷadayʳлaὕὕлἵāmХʲлὕanʥuaʥʣʳлoʤлBadakhshānлʦavʣлʳtʲʣʳʳлonлtʦʣлὕaʳtлʳyὕὕaϐὕʣщ It seems that predecessors of both YaghnōϐХл andл Sogdian underwent the same or very similar stress shifts, the results of operations of stress slightly differ in both languages. Some Sogdian words point to original *Proto-(Eastern-)Iranian stress, the place of this stress (Stress I) can be reconstructed after operation of i-Umlaut, e.g. Sogd. zyrn /zeṙn/ < *ʣᛑn ‘- ʥoὕd (SIMS-WILLIAMS 1989b, 181). Stress later shifted to another position (Stress II): the stress fell on penultimate or antepenultimate syllable. Words with penultimate stress were either disyllabic words or words with a penultima containing long syllable i.e. syllable containing either long vowel (long either naturally or rhythmically) or a diphthong (diphthong could have been formed also by a nasal or ŋ ›) in a closed syllable; in other positions the stress shifts on antepenultima. Position of stress in YaghnōϐХлϑomʣʳлʤʲomлtʦʣлʲʣʳuὕtʳлoʤлopʣʲatʧonлoʤл the Stress II, this stress can be observed in Sogdian in results of operation of i-Umlaut of several words. Such stress shift is also probably related with change of its strength – many unstressed vowels (in YaghnōϐХлoʤtʣnл all syllables) were reduced or even syncopated, mainly short vowels directly preceding or following a stressed syllable. Other stress shift (Stress III) took place in Sogdian, and this change is related operation of the Sogdian Rhythmic Law; but no such shift has taken place in YaghnōϐХ. The Rhythmic Law, 115 prau /pər u/ mentioned above. ·75· which was originally only a phonological feature caused many other changes in Sogdian morphology – this problem will be discussed in following parts of this thesis. The Rhythmic Law divides Sogdian words into two groups – in so-called light and heavy stems 116 (cf. SIMS-WILLIAMS 1984; δκπлέяřя-530; TEDESCO 1926). As the heavy stems we can classify words with stressed root syllable, in fact stress falls on the first possible rhythmically long syllable (i.e. either on a long vowel or on a diphthong – in this case diphthongs are considered groups V , Vu, Vṙ, Vṁ in closed syllable), the heavy stems end with a consonant in majority of words. In the light stems stress shifted to the ending – the light stem words do not have rhythmically long root syllables and the stress shifted towards the end of the word, and thus *Proto-Sogdian endings have been preserved. Emergence of the Rhythmic Law also influenced reduction of vowels in unstressed syllables, mainly when they followed stress – in the heavy stems the original endings disappeared but they remained in the light stem forms. Subsequently the last stress shift (Stress IV) appears – this stress shifts to the ultimate syllable (Nicolas Sims-Williams suggests this development after an analysis of Sogdian documents in the BʲāʦmХ script, some evidence of this feature can be found in several vocalized documents in the Syriac script; SIMS-WILLIAMS 1996a, 312-313) As indicated above, mere shifts in stress position presented a significant feature which resulted in further sound changes in Sogdian and in YaghnōϐХ. Both languages probably shared similar changes of stress for quite a long period of time during their common development. YaghnōϐХлʲʣtaʧnʣdлoʲʧʥʧnaὕлʳtʲʣʳʳлonлуantʣфpʣnuὕtʧmaлуʧщʣщлStress II) Sogdian, however, was more progressive and there developed another innovation in stress (Stress III), this shift was motivated by rhythmical weight of a syllable – the operation of Stress III and the Sogdian Rhythmic Law is one of the most important distinctive features distinguishing YaghnōϐХлʤʲomлSogdian. The following parts present analysis of stress operation reflexes from the *Proto-(Eastern-)Iranian state up to (*Proto-)Sogdian and (*Proto-)YaghnōϐХ, and subsequent Sogdian innovation in the form of the Rhythmic Law. We can distinguish three development stages of stress changes: Stress I, Stress II and Stress III – the first two stages can be observed in both languages (there are sources for position of the Stress I mainly in Sogdian, but they can be suggested in YaghnōϐХ), Stress III is just Sogdian development – in the scientific literature the Stress III is labelled as the Sogdian Rhythmic Law. In the presented thesis I will use the term Rhythmic ŭ‘⁽ just for the outcome of the operation of the Stress III in all its complexity, mainly as a feature influencing Sogdian grammar; the label Stress III means only phonological shift of stress. In Late Sogdian Stress IV followed. A good example of all stress shifts can be seen in the following example: Stress I *aʣám ζ л (Pasht. zə; Wa . ze; Munj. za; Yidgh. zo, zə; cf. Ave. azəm, Ved. ‘hám; Ide. *h “gʰóm, Gre. ἐ ώ) > Stress II ŋázam (Proto-Sogdic ŋá₎u; Yagh. The light stem words can be also labelled o⁾₍ton“s as they had stressed ending; the heavy stem words can be regarded as ’‘›₍ton“s i.e. words with a stressed root. 116 ·76· ŋ ‘₎; Wakh. wuz; Ishk. az(i); Sangl. azə; azi; Yazgh. az; Shugh. (w)uz;лοōsh. az; Khūʤщлοāshrv. πaʲХqщлwaz; Bart. ā₎) > Stress III Sogd ʾzw, /əzúълş Stress IV Sogd. zw /zu/ (?). II.1.1.1. Stress I Stress I corresponds to the position of stress in *Proto-(Eastern-)Iranian. Its responses are preserved only in rare cases, examples can be found mainly in Sogdian words, in YaghnōϐХлtʦʣʲʣл are no direct traces, but its operation can be also presumed. The position of the Stress I is not attested but it can be reconstructed in several words due to reflects of i-Umlaut in some roots, Nicolas Sims-Williams presents several examples: Sogd. zyrn /zeṙn/ ʥoὕd < *ʣᛑn ‘-; -daϑ ‘-. Sogd. ryp - /rep áъ noon < ŋ›áp ϑ ā; Sogd. prʾy pryϑ ъpʲē / toлʳʣὕὕ < In some cases stress can be found even on some nominal prefixes: Sogd py(t) ʾr py ʾr pydʾr / r/ ϐʣϑauʳʣ уoʤф < ŋpáⁱt ā›ⁱ < ŋpát -›ād -; Sogd. pyr nn /péṙ an/ ʳaddὕʣ < ŋpá› -dān‘-; Sogd wz ʾm, ʾw ʾm /uz mъл absolutely, ever < ŋúʣ-gām‘n- (SIMS-WILLIAMS 1989b, 181). The position of the Stress I can be better reconstructed from Pashtō. On the position of Stress I, cannot be presumed much, we can just conclude that under its influence some unstressed vowels were syncopated and/or reduced, this feature can be observed in the first three above mentioned examples – there can be seen a syncopated intermediate stage (ŋ₎á›ᵃn ‘-, ŋ›ápⁱϑ ā-, *pă ᵃϑ ‘-), the syncopation of unstressed vowel subsequently caused i-Umlaut of the stressed root vowel (ŋ₎áⁱṙn( )a-, ŋ›áⁱpϑ ā-, *pə ⁱϑ( )a-). The Stress I can be supposed also in the word ŋ›áupāʦa- ʤoʸ (cf. Ved. lopāśá-, Gre. ἀ ώπη ) > ProtoSogd. rwps / əs/, Yagh. : in this case there was no syncope but shorting ŋ›áopăsa- > Sogd. of ŋā > *a, this change was probably *Proto-Sogdic, in Persian there is < Elam.-OPers. ŋ›‘upāϑa- (cf. MAYRHOFER 1996, 482). Regarding later development in Sogdian and in YaghnōϐХл ʧtл ϑanл ϐʣл ʳuʥʥʣʳtʣdчл tʦatл tʦʣл i-Umlaut occured later, probably in the stage of the Stress II – the crucial reflex of the Stress I was probably result of syncopation of some short vowels preceding a stressed syllable. The position of Stress I is attested from some words in Sogdian, traces of the Stress I can be better found in Pashtōлandлaὕʳoлʧnлκun₍Х-YidghāлandлτakhХ (here the position of the Stress I can be observed in results of Umlaut), many of examples of the Stress I can be compared with Vedic: Pasht. ásp‘ maʲʣ л<лŋáʦuā-; Ved. áśvā-; Pasht. ₍aʷ л<л*ʣám’ā-; Ved. jám’ʰ‘-; Pasht. ϐὕaϑ₎ (f) ; Munj. ; Yidgh. daʲ₎nʣʳʳ < ŋtánϑra-; Ved. tám s›ā-; Pasht. sxər; Wa . xwsar; Wakh. ⁾u›s ʤatʦʣʲлʧnлὕaʷ л<л*x aʦú›‘- ×лσʣdщлśváśu›‘-; Pasht. ē~⁾⁽ ⁾ē; Wakh. ⁾‘ motʦʣʲлʧnлὕaʷ л<л*x áʦr(u)- ×лσʣdщл -; Pasht. d›ē; Wakh. t› (₍) tʦʲʣʣ л<л*ϑ ‘-; Ved. t›á₍‘ ; Pasht. Wa . p‘ night л<л* -; Ved. k -; Pasht. l ná corn, uncer л<л*dān -; Ved. ; Pasht. p‘⁾á cooked, ripe (f) л<л*pax -; Ved. p‘kvá-; ·77· Pasht. zə; Wa . ze; Munj. za; Yidgh. zo, zə ζ л<л*aʣám-; Ved. ‘hám-; Pasht. atə; Wa . otá; Munj. ḱá; Yidgh. ‘ čó; Wakh. at ʣʧʥʦt л<л*a tá-; Ved. a ṭ (MAYRHOFER 1989, 13; MORGENSTIERNE 2003; STEBLIN-KAMENSKIY 1999). II.1.1.2. Stress II Stress shift marked as Stress II characterizes another development in *Proto-Sogdic 117 . The original *Proto-(Eastern-)Iranian stress shifted to penultimate or antepenultimate syllable according to its rhythmic weight: stress was on penultima if this syllable contained naturally or metrically long vowel (i.e. either a long vowel or a short vowel/diphthong in a closed syllable), in other circumstances stress fell on the antepenultima (that implies that rules of stress were similar to those in Latin or Sanskrit). The shift towards Stress II position brought about several significant features, which were characteristic for the development in *Proto-Sogdian and *Proto-YaghnōϐХчлnotaϐὕʣлaʲʣлʤoὕὕoʷʧnʥлʤouʲлphenomena: 1) stress shift was probably related also with its strength, the new Stress II being probably stronger than Stress I; 2) after the operation of Stress II some unstressed vowels (or even whole syllables) were reduced or lost; 3) after the reduction of unstressed vowels the syllabic structure was rearranged, and 4) after loss of unstressed *i (or ŋД and ŋ ) the stressed root vowels and some consonants were palatalized. The results of the changes caused by Stress II have different reflexes in *Proto-YaghnōϐХ (and probably in Ustʲōshanian Sogdian) and in *Proto-Sogdian, it is possible that at this stage the Sogdian dialects of BukhāʲāлandлZhetisu started to split. The majority of dialects developed from *Proto-Sogdic probably retained the position of the Stress II, but clear evidence can be found just for (*Proto-)YaghnōϐХ. For *BukhāʲanлSogdian, *Ustʲōshanian and *Zhetisu Sogdian we can only suppose the preservation of Stress II and no shift towards Stress III. The shift of Stress II resulted mainly in a change of stress strength which led to the reduction of unstressed vowels – short vowels were reduced or changed into Schwa (ə). Long vowels were shortened when unstressed – in YaghnōϐХлʧtлϑanлϐʣлʳaʧdлʷʧtʦлcertainty that *ProtoSogdic unstressed ŋД and ŋ changed to i, u; in Sogdian a similar development can be presumed, but there is no clear evidence due to unsuitable graphic representation of vowels in the Aramaic-derived alphabets. One knows for certain that in Sogdian long vowels ŋД and ŋ were retained in syllables that later bore Stress III; but *Proto-Sogdic ŋā usually remained unchanged, although there some examples of shortening of ŋā > ŋă are attested. The transition from Stress I to Stress II must have been regular, the original Stress I being preserved only in rare cases, mainly in cases of old syncopation of vowels, but also under some other circumstances (see examples given above); and some words have double forms that either preserve an archaic state with Stress I or show Stress II innovations: ŋúʣ-gām‘m- (Stress I) 117 *Proto-Sogdic as a reconstructed language can be interpreted as a development stage of a North Eastern Iranian language just in a period when Stress II operated, but the features caused by effects of Stress II are different for the development of *Proto-Sogdian and *Proto-YaghnōϐХщ ·78· absolutely, ever > *uzStress II) > ŋ ₎> Sogd wz ʾm, ʾw ʾm /{ə m/ (Stress I) × z ʾm /( ) m/ (Stress II) (SIMS-WILLIAMS 1989b, 182). For many forms we cannot exactly conclude whether there was any shift from Stress I towards Stress II, but there is better evidence for the Stress II from a later stage, so I will interpret the position of stress according to position of the Stress II. The fundamental change related to the stress shift has been the above mentioned vowel reductions in unstressed positions, this change can be shown on many examples: Sogd zrʾync zrync /zriṁǰ/ < *uʣ-›ínč‘ ‘- to save, deliver ; Sogd. mʾ (h), mʾ w mʾx / Yagh. m ⁾ ʷʣ < ŋĭ < *ah am. Together with the reduction and loss of vowels a whole syllable can also disappear, such feature is characteristic for YaghnōϐХ, but it can be rarely traced in Sogdian): Sogd wrʾṯy / { < w rʾtyy /w ʾüɨ /, Yagh. aʷa₎ʣ < *u a-(ka-); ‖ tⁱ , t ʤouʲ < ŋč‘ϑu -; Yagh. Sogd ct ʾr ctfʾr cṯfʾr, ṯfʾr /č r/, Yagh. t , t ž‘vá›- ‖ žⁱvá›- to bring, to produce, to ʧnvʣnt < ŋn ǰ-’ᛑ-; moreover, the whole first syllable was reduced in YaghnōϐХ when two short and open syllables preceded the stressed one: Sogd. √pt ⁽ √pt (ʾ)⁽ /p -/, Yagh. d - to ʦʣaʲ < *pati-gáu a- (KhROMOV 1987, 661). The vowel loss is related to an Umlaut of stressed root vowels. Operation of i-Umlaut causes palatalization of a stressed vowel or diphthong after loss of *i or ŋ . Outcomes of palatalization differ in Sogdian and YaghnōϐХ – in Sogdian there are palatalized vowels and diphthong , *u, * u, in YaghnōϐХлtʦʣʲʣлis attested palatalization of and *u: Sogd (y)z-y, ʾ (y)z-y (y)j-y, ʾ j-y / žíл <л ʣžíъл <л * éž < ŋ’á₎d ‘- ϐadчл ʣvʧὕ ; Sogd wyzp- wzpwjp- ʾ⁽ž’-лъüɨžϐáъл<лŋú’ǰ ā tʣʲʲoʲ ; Sogd ”n₍ - ъʤnʣš-/ < ŋ”›‘-nás ‘- to be deceivʣd ; Sogd xwtʾyn xwṯyn hu‘-t unД- quʣʣn ; Sogd. pt y √pṯbyd- /√p t üɨ -/ < p‘t -’úd ‘- to perceive (SIMS-WILLIAMS 1989b); Sogd. √ʾ₎⁽ʾ₍›t √(ʾ)₎⁽ʾ₍›t √₎⁽ʾ₍›t √₎⁽₍›ṯ /√ zwíṙt/, Yagh. zⁱ⁽í›t- to tuʲn < *uʣ-uá›t(‘) ‘-; Sogd ⁽₍ (h) /ʷēšъчлYagh. w ‖л⁽‘ ʥʲaʳʳ < ŋu ʦt› ‘-, Ave. vāst› ‘-; Sogd. frʾʾ⁽₍ c₍ frʾwycyẖ /ʤʲāʷɨуšфčʧъчлYagh. ‖ fⁱ /f oϐὕivʧouʳnʣʳʳ < ŋ”›āmú t -. The issue of syllabic structure transformation in Sogdian and YaghnōϐХлʷʧὕὕлϐʣлthoroughly discussed in Chapter II.1.9., now we need to outline only the basic features of Sogdian and YaghnōϐХлʳyὕὕaϐὕʣ – due to loss of unstressed vowels consonant clusters emerged, in later stages of the language consonant clusters were not allowed in word-initial positions – the clusters have been reanalysed by prothesis (in Sogdian there are reconstructed two prothetic vowels ᵊ and ᶤ), or epenthesis (in YaghnōϐХлa, ⁱ, ); an anaptyctic vowel appeared in YaghnōϐХлʧnлʳʣvʣʲaὕлʷoʲdfinal positions if the word ended in *xm, *xn, * n, ŋ m, ŋ(⁾) n, ŋčn, *fr and *zm: ›á⁾ ⁱn daʷn < ŋ›áu⁾ na-; ⁽á”ⁱr ʳnoʷ < ŋuá”›a-; ⁽á⁾ⁱn ϐὕood < ŋuáhun -; í₎ⁱm fiʲʣʷood < ŋá zma- (cf. KHROMOV 1987, 661), both in Sogdian and in YaghnōϐХлaлʳvaʲaϐʦa₎tʧлvoʷʣὕлʧnлinserted in wordfinal cluster * n: Sogd. rw n ro haṃ, ro aṃ / n/, Yagh. ⁱn, an oʧὕчлϐuttʣʲ < ŋ›áugn‘- [Ave. rao na-, Pahl. › n, Pers. › án, Tjk. ›‘u án, Fārs. ro än]. ·79· II.1.1.3. Stress III and the Sogdian Rhythmic Law The last of significant stress shifts in the languages derived from *Proto-Sogdic is Stress III – this change took place in (literary) Sogdian, where it is generally known as the Sogdian Rhythmic Law; the Stress III has not developed in YaghnōϐХлand probably it did not operate in the Sogdian dialect of Zhetisu, its impact can be excluded less likely in the Sogdian dialects of Bukhāʲā and Ustʲōshana. Together with the operation of Stress III the morphology of Sogdian words was completely rebuilt – stress shifted on the first possible rhythmically long syllable. A long syllable was defined as a syllable containing either a long vowel or a diphthong in a closed syllable. Together with the operation of the Rhythmic Law transformation the loss of unstressed endings took place. If a word contained no rhythmically long syllable, stress shifted to the ending, which under these circumstances remained. According to the position of stress either on the root or on the ending, Sogdian words split into two groups: in the so-called heavy and light stems. The *Proto-Sogdian endings of the heavy stems were reduced or lost due to the stress shift; the light stems when compared to the heavy stems are richer in morphology – in the light stems the original endings were retained as they bore stress. The difference between the light and heavy stems can be demonstrated in the following examples (all forms are in nominative singular): ʾ bʾ / ā ъ < * i< h ʥaʲdʣn × -y b -y / ə íъл<л* á i < ŋ’ág‘h ʥod ; mr h mr /maṙ ъл <л ŋmáṙ i < ŋmá›g‘h ʤoʲʣʳtчл mʣadoʷ ×л ʾmr -y / m(əwф íъл mr -y /məw íъл< ŋm i < ŋm g‘h ϐʧʲd . Apart from the transformation of endings in forms of the heavy stems also other transformations occurred – mainly ŋ and ŋД were shortened in unstressed positions > , ĭ; rarely also ŋā has been shortened to ă: Sogd. ʾʾmʾtʾy ʾʾmʾt(ʾ)k ʾʾmʾṯyy ʾmʾṯy, ʾmṯy / ʾʾmṯyy / ā-ka- ʲʣady . As I have mentioned above, the rhythmically long syllable was every syllable containing rhythmically long vowel – i.e. either a quantitatively long vowel, or a vowel as the first part of a diphthong118 in a closed syllable, or a vowel followed by a labialized velar (uvular) fricative ⁾°; other syllables are considered as rhythmically short (i.e. vowels followed by clusters mt, ny, my, tkw/tk , ⁾ n, and rw). However, if there was a light stem word terminating either in -y, -w, -r or a nasal supplemented by an ending beginning with a stop or an affricate, the light stem changed √ rto a heavy stem (this feature can be observed mainly with verbs), e.g.: Sogd. st 119 √’›- /√ ər-/ to bear, to ϑaʲʲy : ə›ám ψζϊлϐʣaʲ (1 pers. sg. pres.) × aṙt ψ(s)he] bears (3rd pers. st sg. pres.) ; Sogd. √ ⁽- /√šəw-/ to ʥo : əwám ψζϊлʥo (1 pers. sg. pres.) × t ψуʳфʦʣϊлʥoʣʳ (3rd pers. sg. pres.) ; such a feature is not attested in forms of the plural in *-tá: Sogd. wn(ʾk)h wnʾ /wənáъ tʲʣʣ (nom. sg.) : wntʾk(h) wndʾ /wəndá ~ wəntá/ tʲʣʣʳ (nom. pl.) . Apaʲtлʤʲomлtʦʣл ʧnʦʣʲʧtʣd лʤaὕὕʧnʥлdʧpʦtʦonʥʳлtʣʲmʧnatʧnʥлʧnл - and -u also vowels followed by -ṙ and -ṁ were classified as diphthongs, and in such case it is necessary to say that ṙ and ṁ had to be followed by a stop or a fricative (SIMS-WILLIAMS 1984, 206, 209-212). 119 By analogy also Sogd. ə›tí. 118 ·80· II.1.1.4. Stress IV According to analysis of late Sogdian texts/words written in the BʲāʦmХ script it can be supposed that after the operation of Stress III another stress shift (Stress IV) took place in Sogdian – in this case the stress shifted to the ultimate syllable. The position of Stress IV can be seen in the graphic representation of phonemes and versus and in the BʲāʦmХ script: the graphemes <e> and <o> were used only in the last (i.e. stressed) syllable of a polysyllabic word, in monosyllabic words or in proclitics; the graphemes <i> and <u> appear instead of (etymologically) expected ē, in another than the last (i.e. unstressed) syllable of a polysyllabic word or in enclitics. The main evidence for the shift towards the Stress IV comes from the documents in the BʲāʦmХ script, and some indications can also be seen in some vocalized Christian Sogdian texts in the Syriac script 120 (SIMS-WILLIAMS 1996a, 312-313). It is possible that there was a transitional stage between operation of Stress III and Stress IV, when stress shifted from the first rhythmically long syllable towards the last possible rhythmically long syllable – Nicolas Sims-Williams states that according to the analysis of the Sogdian texts written in the BʲāʦmХ script it will be necessary to revise the Sogdian Rhythmic Law (SIMS-WILLIAMS 1996a, 312). Elio Provasi analysed the metre of Sogdian verses in a Sogdian translation of the Middle Persian hymn cycle ђu₍‘d‘gmān, and he supposed that in a heavy stem word with two rhythmically long syllables the stress shifted towards the last rhythmically long syllable (PROVASI 2009, 351-353), which seems to be inconsistent with the definition of the Sogdian Rhythmic Law according to Nicolas Sims-Williams. As is evident from the Sogdian documents written in the BʲāʦmХ script, the shift from Stress III towards Stress IV was not only a stress shift but also a cause of sound system changes in late Sogdian – after the operation of Stress IV the sounds and could not remain in an unstressed position and so they have been changed to and respectively. Unfortunately, in the Sogdian variety of the BʲāʦmХ script (originating in the Central Asian variety of BʲāʦmХ as it has been used for Old Uyghur; SIMS-WILLIAMS 1966a, 309) the quantity of the vowels , , and was not distinguished, therefore it cannot be assessed whether the shift towards Stress IV was related to the change of quality of and witch probably also changed in their quantity. Examples of the Stress IV can be shown in the following examples: Sogd ine ʾỵnỿ / nē/ (×л ʾyny, ʾ₍nʾk yny(y), ʾyny(y) ʾỿnỿ / ) tʦʧʳ < ŋá n‘-ka- [Ved. ena-, Pahl. ēn, Pers. Дn]; Sogd. ₎ā u‘ rkeṃ /zāʷaṙ ₎ʾ⁽›k₍n ₎ʾ⁽›k(ʾ)₍n, ₎ʾ⁽›‹₍n ₎ʾ⁽›‹₍n / ṙ₎ēnъф ʳtʲonʥ ; or enclitics: Sogd ni-st / у× nyst(y) ny(y)st(t) nysṯ(y) nỿsṯ, nysṯ, nsṯ /n stуíф/) ψуʳфʦʣъʧtϊлʧʳ not л< *n‘ (d)-ást [Yagh. nēst, Pers. nēst] a Sogd wu-ṯ / wt bwṯ / ōtъф ψуʳфʦʣϊлʧʳ < ’áu‘-ti [Pers. ’uváđ]. 120 In the documents appear primarily vocalic signs <ỿ, ×> and <ỵ>, i.e. and , vocalization of <ẇ> and been used rarely (SIMS-WILLIAMS 1981a, 356; 1996, 307). ·81· <ẉ> have II.1.2. Vowels and diphthongs The original Old Iranian system of seven vowels (*a, *i, *u, *›, *ā, *Д, * ) and four diphthongs (*‘ , *‘u, *ā , *āu) has considerably changed in course of the development of Sogdian and YaghnōϐХ. In Sogdian there can be reconstructed 17 (or even 19) vowels (a, ə, əʳ [ ], e, i, ʳ ψʧ˞ϊч 121 , o, u, uʳ ψu˞ϊ, ā, ē, Д, , , ṁ122, ṙ123; eventually / [ ] and ą ψãϊ), two super-short prothetic vowels (ᵊ, ᶤ) and eight diphthongs (a , au, ā , āu, ēu, u, 124, 125), to these old diphthongs are added 19 nʣʷ лdiphthongs (aṙ, aṁ, iṙ, iṁ, ṁ, eṙ, eṁ, uṙ, uṁ, āṙ, āṁ, ēṙ, ēṁ, Дṙ, Дṁ, ṙ, ṁ, ṙ, ṁ). In YaghnōϐХ the situation corresponds more to the Middle Iranian stage: in every dialect there are eight (nine) vowels (a126, i, u, ē, Д, , , furthermore ʉ in the Western dialect, and in the Eastern dialect ; peripheral sound is ā), two super-short svarabhakti vowels (ⁱ, ) and one true diphthong (‘ , in the Eastern dialect it is pronounced , in the transitional dialect there is ᵎ [ ᵎ•]) and three newly built diphthongs (‘u, ēu, u). *Proto-Sogdic vowel system developed differently in these two languages, the most significant difference was mainly Sogdian reduction of all historical short vowels in unstressed position (i.e. *a, *i, *u > ə or ), in YaghnōϐХлtʦʣлʦʧʳtoʲʧϑaὕлʳʦoʲt vowels were also reduced in unstressed positions, but not to such extent as in Sogdian (In Sogdian the unstressed short vowels were neutralized, in YaghnōϐХлtʦʣ reduction resulted in emergence of super-short vowels in an open syllable preceding a stressed syllable). Vowel system of Sogdian needs to be based mainly on the study of historical phonology – as mentioned above, Sogdian was written in alphabets derived from Aramaic which was not able to sign vowels properly and thus their appearance have to be reconstructed – as a valuable source here serve s ʤʣʷлdoϑumʣntʳлʷʲʧttʣnлʧnлtʦʣлBʲāʦmХлʳϑʲʧptлandлʳʣvʣʲaὕлvoϑaὕʧzʣdлCʦʲʧʳtʧanлtʣʸtʳлʧnл the Syriac script, on their basis we can evaluate the reconstructed data (see Table 34). Analysis of Sogdian phonology has been studied by Nicolas Sims-Williams, basic outline of Sogdian vowel system can be found in his basic outline of Sogdian grammar (SIMS-WILLIAMS 1989b, The vowel is interchangeable with ə (in majority of occurrences they are allophones; the exception is ( ??) as a reflex of palatalized ŋ ‘u). 122 Sound marked as ṁ is a vocalic nasal prolongation of preceding vowel appearing as the second part of a diphthong, its realisation changed according to the pronunciation of preceding vowel e.g.: aṁл ψa л ~л aã] (SIMS-WILLIAMS 1989b, 181). 123 The sound represented as ṙ is something like syllabic › as second parts of a diphthong, it was realized as rhotacized vowel əʳ, in this case the rhotacized vowel was non-syllabic [ ϊл уoʲл ψʲ]), e.g. aṙ [aəw ~ a ~л aʲ] (cf. SIMS-WILLIAMS 1989b, 181). 124 Or probably monophthong (cf. SIMS-WILLIAMS 1989b, 181). 125 Nicolas Sims-Williams interpreted development of ŋu , ŋu‘ and palatalized ŋu‘, ŋ‘u as > ȫ, eventually o (SIMS-WILLIAMS 1984, 206-эы7фчл ϐutл aϑϑoʲdʧnʥл toл tʦʣл BʲāʦmХл ʳpʣὕὕʧnʥл <yuʣчл yuчл uʣ, ve> he revised his reconstruction towards rising diphthong (SIMS-WILLIAMS 1996a, 313-314). Ilya Gershevitch does not solve this u pʲoϐὕʣmчл onὕyл ʧnл tʦʣл ϑaʳʣл oʤл tʦʣл ʷoʲdл ʤoʲл tʦʣл ʳun л *x á› ‘w(y)r, xwyr xwr he reconstructs reading gh. xʉr. /xuwəʲълуδκπлέээюфчлϑoʲʲʣϑtὕyлъʸü 126 With positional allophone â i.e. half-long a ψa•л~ ȴ•ϊ. 121 ·82· 175-181), in the paper The Sogdian sound-system and the origin of the Uyghur script (SIMS-WILLIAMS 1981a) he compared spelling in the Sogdian script with sound system of Old Uyghur and in the paper Th“ Sogd ‘n m‘nusc› pts n B›āhmД sc› pt ‘s “v d“nc“ ”o› Sogd ‘n phonology (SIMS-WILLIAMS 1996a) he evaluated Sogdian phonology with the help of Sogdian gloʳʳʣʳл ʧnл tʦʣл BʲāʦmХл ʳϑʲʧptщл ἴtʦʣʲл ʳtudʧʣʳл oʤл πoʥdʧanл pʦonoὕoʥyл ϑanл ϐʣл ʤoundл ʧnл ʤoὕὕoʷʧnʥл ʷoʲ₎ʳśл δκπл έ82-483; GAUTHIOT – BENVENISTE 1914-1923; LIVSHITS – KHROMOV 1981, 373416; QAοФB 1383, xxix-xxxii. vowel Sogdian alphabet Manichaean alphabet BʲāʦmХ script Syriac alphabet a ʾʾ-, -(ʾ)-, -ʾ(h), -h, -(ʾ)k(h) ʾ-, -(ʾ)-, -ʾ(-), -(ʾ)ẖ ʾ-, ᴟ-, -(ʾ)-, -×-, -ʾ(-), -ᴟ, (-h) a, -a, -ā ā ʾʾ-, -ʾ(ʾ)-, -(ʾ)h, -(ʾ)k(h), -ʾ(ʾ) ʾʾ-, -ʾ-, -ʾ(ʾ), -(ʾ)ẖ ʾ-, /ᴤ-, -(ʾ)-, - /ᴤ-, -×й×-, -ʾ, - /ᴤ, (-h) ā ə ʾ-, -ø-, -y-, -(w)- ʾ-, -ø-, -y-, -(w)- ʾ-, ᴟ-, -(×)-, -y-, -(w)- ʾy-, -(y)- ʾy-, (y)-, -(y)-, - (y)- ʾy-, ʾỵ-, -y(-), -ỵ(-) a, -ø- ʾy-, ʾỵ-, -y(-), -ỵ(-) i, -i, -ä ʾy-, ʾỿ-, -y(-),-ỿ(-), -×- e, i ʾw-, ʾẉ-, -w(-), -ẉ(-) u ʾw-, ʾẇ-, -w(-), -ẇ(-) o, u i Д e ē u ʾy-, -(ʾ)₍(-) ʾy-, -(ʾ)₍(-), -(ʾ)k ʾy-, y-, -(ʾ)₍(-), -y(y)(-) ʾw-, -(ʾ)⁽(-) ʾw-, -w(w)(-) o ʾw-, -(ʾ)⁽(-), -(ʾ)kw ā ʾʾy-, -ʾ(ʾ)₍(-) ʾʾy-, -ʾ(ʾ)₍(-) ʾy-, -(ʾ)₍(-) ai, ā₍ āu ʾʾw-, -ʾ(ʾ)⁽(-) ʾʾw-, -ʾ(ʾ)⁽(-) ʾw-, -(ʾ)⁽(-) au (-)w(y)- (-)w(y)- (ʾ)⁽₍-, (-)w(y)(-)⁽(ʾ)₍-, (-)w(y)- i yu(e), u“, ve əʳ iʳ uʳ ṙ (ʾ)›-, -r-yr-wrr (ʾ)›-, -r-yr-wrr (ʾ)›-, -r-yr-wrr ar, › ṁ m, n m, n(n) m, n m, ṃ, n, ṇ r, ṟ Table 34 Spelling of vowels in the Sogdian, Manichaean nad Syriac alphabets and in the BʲāʦmХ script. By comparison of Sogdian documents in the Sogdian, Manichaean and Syriac alphabets along with a few fragments in the BʲāʦmХ script and with use of methods of historical linguistics it is possible to reconstruct the Sogdian vowel system. Another important source, which can be used to validate values of reconstructed vowels, are Sogdian words shared with YaghnōϐХчл moʲʣovʣʲл tʦʣл data can be compared also with Sogdian loanwords in some other languages, especially in Persian (primarily in ρā₍Х₎Х Persian and in Tajik dialects), in Old Uyghur (and also in other Turkic languages – some Sogdian words have been recorded for example by Ma mūdлϐʧn usayn bin Mu ammad al-ἰāshghaʲХфщ Nicolas Sims-Williams in his study The Sogdian sound-system and the origin of the Uyghur script compared the Sogdian alphabet (with regard to the Manichaean and Syriac alphabets) with the so-called Uyghur script, which originates from cursive version of the Sogdian script. The ·83· speakers of Old Uyghur adopted the already established Sogdian alphabet to record their language, however, they simplified its (in many aspects archaic) orthographical rules (SIMS-WILLIAMS 1981a; on the contrary the Old Uyghur variety of the BʲāʦmХ script was taken over by the Sogdians from the Uyghurs, see SIMS-WILLIAMS 1996a, 309). Since Old Uyghur vowel system can be quite easily reconstructed by comparation with other Turkic languages, in the following lines I will summarize a short outline of the Old Uyghur vowel system as compared to Sogdian. Old Uyghur had nine vowels: a *[ȵ], ä хψǦϊ ~ ė *[e], ï *[ɯ], i, o, ö, u, – there were four pairs of front/back vowels in mutual opposition and moreover vowel ė, which was a positional allophone of ä; question of quantity of Old Uyghur vowels is unclear, in *Proto-Turkic there are reconstructed also long counterparts of the above mentioned Old Uyghur vowels, reflexes of *Proto-Turkic quantity have remained in languages such as Turkmen or Khalaj (cf. RÓNA-TAS 1998, 69-71). For graphic representation of Old Uyghur vowels Sogdian spelling rules were adopted: OUygh. a has been written <ʾʾ-, -ʾ(-)> i.e. same as Sogd. ā; OUygh. ä <ʾ-, -ø-, -ʾ> = Sogd. a, ə; OUygh. ė, ж ï <ʾy-, -y(-)> = Sogd. , , ; OUygh. o, u <ʾw-, -w(-)> = Sogd. , ; OUygh. ö, <ʾwy-, -wy(-) /in the first syllable of a word/, -w(-) /in other then the first syllable of the word/> = Sogd. ( ), . Apart from the above mentioned spelling rules for vowels, the Old Uyghur spelling took over some Sogdian orthographical conventions, mainly spelling of word-initial a as <ʾ-> prior to a nasal and r; on the other side Old Uyghur took over neither the archaic writing of - and - by the letter kāp, nor spelling of - with the letter hē (SIMS-WILLIAMS 1981a). To precise the reading of the Old Uyghur alphabet (traditional) Mongolian alphabet can help as it has been adopted from the ancient Uyghurs (see excursion 5). By combination of the methods of historical and comparative linguistics with the study of Sogdian orthographies in the Aramaic-derived scripts and in the BʲāʦmХ script together with comparation of the material with Old Uyghur documents and with a study of Sogdian loans (i.e. study of the Sogdian loans in neighbouring languages and also study of Sogdian borrowings from other languages such as Sanskrit and PrakritsчлἵaʦὕavХчлTurkic or Chinese) basic patterns of the Sogdian vowel system can be reconstructed. None of the graphic systems utilized for writing Sogdian for example does not mark vowel quality (with an exception of a ×лā, in this case, as will be seen later, the difference between those two sounds was not in quality but in quantity), but due to operation of Stress III it can be supposed that long Д and have been wr ъ ūʲъл ʤaʲ л <л -, √ ›ʾ₍n √⁾›₍n preserved only in stressed positions: -; otherwise the historical long Д, was shortened in unstressed ъ√ʸʲХn/ to ϐuy л <л positions, similarly in YaghnōϐХлtʦʣʲʣлʧʳлno Д and in other than stressed position, so YaghnōϐХл and Sogdian development are comparable in this case. More complicated is a situation of Sogdian ē and , we can state with certainty that their long varieties occurred in stressed positions, but according to etymology there is attested also and in unstressed positions (in majority of cases in endings of masculine aka-stems; e.g. Sogd. < *ʣ -ka- son ). No texts in the BʲāʦmХ script can help to solve problem of quality of unstressed ē, , but according to the ·84· BʲāʦmХ documents in a late ( Post-Cl‘ss c‘l ) Sogdian it can be surmised that in the Late Sogdian language vowel quantity was not as important as vowel quality, as can be demonstrated on some examples in the vocalized Syriac texts: Sogd. ʾỵnỿ ine / nē/ < Sogd ʾyny, ʾ₍nʾk yny(y), ʾyny(y) ʾỿnỿ / tʦʧʳ < ŋá n‘-ka- – in this case stress just shifted towards the last syllable, but neither Syriac vocalization nor BʲāʦmХ vowels show vowel quantity. The problem of word-final - and - was commented by Walter Bruno Henning in his study Sogdian LoanWords in New Persian – all the Sogdian -stem endings are rendered as -a in Persian (HENNING 1939, 98), i.e. consistently with development of the -stems in Persian (OPers. - - - > Pahl. -ag > Pers. Tjk. AfghP. -a, γāʲʳ. -e (-ä)), and thus Henning suggests that the Sogdian unstressed word-final -e and -a were realized as short vowels (ibid.). Much more evident is the difference between a and ā – both vowels differed not only quantitatively but also qualitatively: a was a front open short vowel, while ā was a back open (rounded) vowel similar to Modern ἵʣʲʳʧanлandлαaʲХ or to long in Scandinavian languages; different quantity of a and ā can be presumed also from the adoption of the Sogdian script for Old Uyghur. The North Turkestan BʲāʦmХ script did not distinguish in quantity of e, i, o, u but retained distinction between a and ā, and similarly vowel diacritics in the Syriac script express rather vowel quality then quantity (i.e. a, ā, , , , ) – it can be assumed that both BʲāʦmХлὕʣtters a-kā›‘ and ā-kā›‘ as well as Syriac ᴤ/× ( /×) and ᴟ/× and Manichaean and Sogdian ʾ-/-ø(-) and ʾʾ-/-ʾ(-) primarily did not distinguish vowel quantity but vowel quantity 127 (cf. SIMS-WILLIAMS 1981a, 355-358; SIMS-WILLIAMS 1996a, 310-311). Just the difference in quality of a and ā motivated adoption of spelling of ä and a in Old Uyghur – Sogd. a (and its allophone ə) and OUygh. ä were front vowels, whilst Sogd. ā and OUygh. a were both back vowels (SIMS-WILLIAMS 1981a, 358). After the operation of the Stress III (as a phonological feature) Sogdian morphology and phonology underwent other changes labelled as the Sogdian Rhythmic Law. From the phonological point of view the Rhythmic Law can be characterized by a change of syllabic structure (this feature will be discussed in chapter II.1.9.) and by a split of vocalic system according to their rhythmic length to short (reduced) and (rhythmically) long (i.e. long vowels and diphthongs) vowels – according to the syllabic weight the Sogdian words distinguished rhythmically light and heavy stems. Words with initial unstressed syllables could start only in reduced vowels ə, or əʳ; words beginning in vowels a, ā, i, Д, e, ē, , u and belonged to the heavy stems as they always bore stress. Word-internally the situation is similar, but the vowels a, i, o, u can stand also in an unstressed position without being considered rhythmically long (i.e. that in such change they do need not to be the first part of a diphthong) – the vowels a, i and u are shortened varieties of originally long ŋā, ŋД and ŋ ; the vowel o comes either from a A similar difference in vowel quality can be observed in continuants of Iranian vowels *a, ŋā in other Iranian ὕanʥuaʥʣʳśлγārs. ä (< ŋ‘) ψǦϊл×л [ȵ”]; Tjk. a ψaϊл×л (< ŋā) [ɔ”], Yagh. a ψaϊл×л (< ŋā) [ɔ”], Os. æ (< ŋ‘) [ȳϊл×л‘ (< ŋā) [ȴ], Kurd. e (< *a) ψǦϊл×л‘ (< ŋā) [ȵ”]; Pasht. a ψaϊл×лā [ȵ”] etc. 127 ·85· diphthong ŋ‘u prior to *xm, ŋ⁾ (u) or from labialization of *a in front of *x or ŋCu. From all the (historically) long vowels only ā can appear also in an unstressed position. In Sogdian there were probably two reduced vowels ə and both originating in *ProtoSogdic short unstressed vowels *a, *i and *u. In the Aramaic-derived alphabets these vowels were usually unmarked, rarely they were written by the letter yuḏ. Both vowels can be considered as allophones of the Schwa sound (ə), I will use the letter (i.e. allophone of ə) in Sogdian words where Schwa is written with the letter yuḏ. Moreover the vowel can originate in palatalization of ŋ ‘u in the Sogdian word ʾync(h), ynch ʾync(h) ync ʾync ъ|ɨṁǰ/, Yagh. nč ʷʧʤʣчлʷoman < ŋ áun -kā- – in such case was probably not an allophone of ə, but it is a separate phoneme. In some words it is difficult to interpret a vowel recorded by the letter yuḏ. Yuḏ often appears instead of expected a in front of a nasal (e.g.: Sogd √ (y)nd : √ (y)sṯ- ъ√ ɨṁd : √ ɨst-/, Yagh. vant- : vást‘ to bind (present : past participle) < ŋ’ánd‘- : ŋ’ásta-(ka-)): here yuḏ appears to be an attempt to record a similar sound change that can be observed in Avestan, where *a is often realized as ə in front of a nasal (cf. Ave. asənga- ʳtonʣ < *aʦáng‘-). In Sogdian there was at least one rising diphthong – , which emerged either from diphthongs ŋu and ŋu or as a result of palatalization of ŋ u or ŋu‘ (SIMS-WILLIAMS 1981a, 206-207; 1989b, 180; 1996a, 313-314). With less certainty, we can assume a second rising diphthong that emerged from palatalization of *u; Nicolas Sims-Williams interprets the result of palatalization of *u as (SIMS-WILLIAMS 1989b, 181; however a development ŋu > can be expected rather than Sims-τʧὕὕʧamʳ ŋu > ). Both diphthongs can be phonetically 128 interpreted as follows: [yʣу”фл~ ʣу”фл~ ʣу”фϊ, [yɨ ~ ɨ ~ ɨ] . In the presented work I will interpret the result of development of ŋu and palatalization of *u as (although the development outlined by Nicolas Sims-Williams can be seen as an alternative), e.g. Sogd √ wys /√xüɨs/ (according to Sims-Williams /√ʸü ) to ʳʷʣat < ŋhu s‘-; Sogd. wyzp- wzpwjp- ʾ⁽ž’-ʾлъüɨžϐá/ (according to Sims-Williams /üžϐáъ) < ŋú’ǰ ā terror (SIMS-WILLIAMS 1989b, 181). One of the reasons for instead of is the spelling of this diphthong in the Aramaicderived alphabets, in which is spelled either as <wy> or <w>, just in the Syriac alphabet there is <‚ʷşчлʳoлʧtлʷaʳлʲatʦʣʲлaлdʧpʦtʦonʥ, when has been marked by the letter waw but the letter yuḏ for has been used inconsistently (similarly as on other occasions). On the contrary, later pronunciation of the diphthong was spelled with i-kā›‘ ʧnлtʦʣлBʲāʦmХлʳϑʲʧpt, this probably means delabialization of either or (delabialization is evident from some younger Sogdian texts): Sogd cā-ṯ /ɨč tл <л üɨžčy tъл comfortable (cf. SIMS-WILLIAMS 1996a, 308; SIMS-WILLIAMS – HAMILTON 1990, 42-43). It can be supposed that the diphthong was later monophthongized into a back vowel (Sims-Williams presumes ; SIMS-WILLIAMS 1981a, 207) – such change is attested in Manichaean and Syriac orthographies written as waw (but its diphthongal character remained in the BʲāʦmХ documents). 128 βvʣnлψyə ~ ə~ ə] if the sounds ə and were allophones in this case. ·86· Nicolas Sims-Williams postulates one more diphthong: u. This diphthong is reconstructed aϑϑoʲdʧnʥлtoлBʲāʦmХлʳpʣὕὕʧnʥ of two words: Sogd yau /| u/ onʣ and (from the previous word derived) Sogd prau /pər u/ уtoʥʣtʦʣʲф ʷʧtʦ щ In the Aramaic-derived alphabets there are diffʣʲʣntлʤoʲmʳлoʤлʳpʣὕὕʧnʥлoʤлtʦoʳʣлʷoʲdʳśл onʣ – Sogd. ʾyw ʾyw, yw yw, ẏw; Yagh. Д; and (together) with – Sogd. p›ʾ(₍)⁽ p›ʾ(₍)⁽, pryw pryw prw (SIMS-WILLIAMS 1996a, 313). However, ʧtл ʧʳл poʳʳʧϐὕʣл tʦatл BʲāʦmХл <au> was not read as / uъл ϐutл aʳл ъau/ or /ɔu/ – according to vocalized record in the Syriac script: Sogd ẏw129 / ~ yɔu/: Sogd prau, yau /pəʲauл ~ pərɔuчл yauл ~ yɔu/ (cf. SIMS-WILLIAMS 1996a, 313). In this thesis I will tend to mark pʲonunϑʧatʧonлoʤлBʲāʦmХ au-kā›‘ as / u/, even though according to the spelling in the Semiticderived alphabets it is possible to read these words as /у|ф uл~ yɔu/ and /pər uл~ pərɔuл~ pəʲāu/. [a] basic pronunciation of the phoneme pad ψǦϊ allophonous pronunciation after aл paὕataὕ лϑonʳonantчлmaʧnὕyлʧnлtʦʣлʷoʲd ǰax ǰax [ȴϊ allophonous variety in vicinity of an uvular xar ψa•ϊ half-long vowel, only in the word vânt vânt [ȴ•ϊ half-long vowel, only in the word â› ā [ȴ”ϊ result of compensatory lengthening in case of loss of ᴥ, h or consonant âʲ tār x < taᴥ < kahd nчлǰām < ǰamᴥ e ψʣ•ϊ half-long variety, in native words or in Tajik loans it originates from *i prior to ᴥ, h and in a closed syllable; pronunciation of e in Russian loans mehm nчлaϐéd ψʣ”ϊ basic pronunciation of the phoneme, in inherited words it appears only in a stressed position pēnчлʳēϐ ψʧ”ϊ in vicinity of , ž or a nasal šēʲчлmēt [ ”ϊ [ ʧ•] ψaʧ] pronunciation of historical diphthong *aı is preserved as a diphthong in the Western dialect, In the Eastern dialect it is pronounced (and often merges with ē); in the transitional dialect it is pronounced rather as half-long semi diphthong ᵎ m n ‖ m ᵎn ‖лmaʧn, w šл‖ w ᵎšл‖лʷaʧš [ɪ] basic pronunciation of the phoneme pit [ɪ] super-short pronunciation (mainly in an open syllable before a stressed 130 vowel) xⁱšíʤtчлtⁱʲáy;лʲ [i] allophonous pronunciation either near to fricatives or in a closed syllable following a palatal k, g ʧšчлʥʧʲdчл₎ıša₎ ψʣ] in unstressed position or in closed stressed syllable [e] allophonous variety word-finally or before a pharyngeal or an uvular ψʧ”ϊ basic pronunciation of the phoneme ψʣ”ϊ allophonous pronunciation between stops rХšчлpХr, tХr [ ʧ”] pronunciation after a stop tХ₎члtХʳ ψʧ ”] pronunciation after a fricative ʤХ₎ a â ē ᵎ aʧ i Х ᵎ ᵎ after a before a n ⁱn tírak, amír, níža₎чл áʸtit m ʲtʧчл áʷʧчлʧʸчлdíhak, qizíq See SIMS-WILLIAMS 1996a, 313 ẏw was a scribal error or an abbreviation for *yẇw (?). Super-short /i/ will be transcribed ⁱ, its pronunciation is consistent with an allophonous realizations of a nonreduced iśлψЧл~ ɪ ~лʣл~ ĕ]. 129 27 130 ·87· o ō u ū ᵎ [ɔ•] half-long pronunciation of appearing only in Russian loans ʤoὕ₎ὕóʲ [ɔ”ϊ basic pronunciation of the phoneme Ḳá nōϐчлya nōϐ ψo”ϊ in a closed stressed syllable or in front of a nasal zⁱv k [ ”ϊ allophonous pronunciation in a closed stressed syllable rōt, dⁱr t ψu”ϊ allophonous variety of in front of a nasal n m, mehm n [ ] basic pronunciation of the phoneme ϐuqqá [ ] super-short pronunciation (mainly in a syllable in an open syllable before a 131 stressed vowel) s [u] allophonous pronunciation near to a fricative šuʤt [o] allophonous realization in closed syllable containing a stop urk, kut, pul, kun [ ] allophonous pronunciation of u, mainly before an uvular sound [u”ϊ sound that emerged from historical ŋ (and ŋ ), in the native words it appears only in a stressed syllable uʸš r pas < *r pas, < *r n ψy”ϊ [yʧ”] allophonous pronunciation of historical ŋ in a stressed syllable – such pronunciation appears only in the Western dialect, in the Eastern and transitional dialects it merged with Yagh. , kab d ‖ kab d < *kab d, xūr ‖ x ᵎr < хʸūʲ Table 35 YaghnōϐХлvoʷʣὕлʳyʳtʣm (NἴσÁἰ 2010, 220-221). YaghnōϐХлvoʷʣὕлʳyʳtʣmлʧʳлϑonʳʧdʣʲaϐὕyлʣaʳʧʣʲлtoлʧntʣʲpʲʣtлduʣлtoлtʦʣлʤaϑtлtʦatлḲaghnōϐХлʧʳлaл living language, but the situation is complicated by number of allophones of the basic vowels. YaghnōϐХл voʷʣὕл ʳyʳtʣmл ʧʳл ʧnл ϑontʲaʲyл toл tʦʣл уʲʣϑonʳtʲuϑtʣdфл πoʥdʧanл ʳtatʣл muϑʦл pooʲʣʲчл however YaghnōϐХлʥives an impression of a more archaic language than Sogdian. I do not want to discuss the phonology of YaghnōϐХл voʷʣὕʳл – this issue has been dealt with by Valentina Stepanovna SOKOLOVA (1953a), a shorter overview is outlined in the grammatical overview attached to the Yaghn ’Д–Czech dictionary (NἴσÁἰ 2010, 220-221 – see Table 35). YaghnōϐХ vowel system is practically the same as the vowel system of the Zarafshān dialects of Tajik (see excursion 3; NἴσÁἰ [in print], Table 1, Table 2; NἴσÁἰ 2009), it may be in a way influenced by a vowel system of literary and colloquial Tajik. The basic difference of YaghnōϐХл and neighbouring dialects of Tajik (i.e. Zarafshān dialects of Mastchōʦ, Falghar and Fōn and Southern Tajik dʧaὕʣϑtʳлoʤлσaʲzōϐ) is pronunciation of short u – in YaghnōϐХ (mainly in rapid speech) there is a tendency of front articulation of u132 (SOKOLOVA 1953a, 69; this feature can explain development of > ʉ), but neither in literal Tajik nor in its σaʲzōϐлdʧaὕʣϑtлtʦʣʲʣлʦaʳл not been described such change. YaghnōϐХлʳʦaʲʣʳлanother feature with the neighbouring dialects of Tajik – rising of (ŋā >) > (in this work marked < >) in front of a nasal. Roland Bielmeier explained this change as a Tajik influence (BIELMEIER 2006 [online]; after him NἴσÁἰ [in print]); similar change appears also in other Iranian languages and dialects – in the Zarafshānл dʧaὕʣϑtʳ (ŋā > _{mж n} > ), in Southern Tajik dialects (ŋā > _{mж n} > ~ ); in Super-short /u/ is transcribed here as , its allophonous pronunciation is similar to a non-reduced u: [ ~лuл~лoл~ ]. 132 There is no tendency of fronted pronunciation of long (< ŋ ) – this feature can be explained as a result of the chain-shift ŋā > | ŋ > | ŋ > ŋ > ʉ | *u > u. 131 ·88· ρʣʦуʣфʲānХл colloquial Persian (ā_{mж n} > ), ʧnл εʣʲātл dʧaὕʣϑtл oʤл Afghan αaʲХ (ā_{mж n} > ; IOANNESYAN 1999, 21), in the Hazāra language (i.e. Persian dialect of descendants of Ǧenghis Khān ʳл κonʥoὕʧanл ʳoὕdʧʣʲʳ; ŋā > _{m, n} > ; EFIMOV 2008, 355), ʳʧmʧὕaʲὕyл ʧnл ρātХ, dialects of Fāʲʳ oʲлʧnлḲazdХ (ŋā_{m, n} > u; GRYUNBERG – DAVYDOVA 1982, 224; KERIMOVA 1982, 319; MOLCHANOVA 2008, 253, 260), in ShughnХ (ŋā > _{mж n} > ) etc., this feature is probably characteristic for development of the Western Iranian languages133 with a partial projection into the Eastern Iranian language area. Other feature borrowed from Tajik is lowering of articulation of > before tautosyllabic h, a ᴥ – this feature is typical for Tajik, but it rarely appears in Uzbek or in ShughnХ; in YaghnōϐХл tʦʧʳл dʣvʣὕopmʣntл ʧʳл attʣʳtʣdл in one example – on the YaghnōϐХ verb díh‘k to ʦʧt л– in forms of the third person singular and in the second person plural there are forms déhč ‖лdéht t respectively déhs t ‖лdéht t (both examples are shown in the present tense), and forms of present and past participles déhn‘ and déht‘, in other cases there are forms with dih- (although in the contemporary language forms derived from the innovated root deh- by analogy begin to appear in all verbal forms; such feature cannot be shown in other YaghnōϐХл ʣʸampὕʣʳл ϐʣϑauʳʣл tʦʣл ʳoundл h appears rarely in genuine YaghnōϐХл ʷoʲdʳ; NἴσÁἰ [in print]); analogical feature is lowering of > й before h, and ᴥ in a closed syllable, which can be observed in Tajik (and Uzbek), but it is not directly demonstrable in YaghnōϐХ – as mentioned above, the h sound is rare in YaghnōϐХ (and the sounds and ᴥ appear only in Arabic loans), so such changes are observable only in Tajik loans in YaghnōϐХлуʧnлḲaghnōϐХлtʦʣл results of lowering are the same as in the Zarafshān dialects; i.e. > but does not change in > u/ in Zarafshānлdʧaὕʣϑtʳ or > ʉ in YaghnōϐХ). Peripheral vowel is long ā (or eventually ), which is a result of compensatory lengthening of a before h, , ᴥ in a closed syllable (e.g. baᴥd > ’ād > Yā Ḳaƒqūb, Jacob ); similar ὕatʣʲ ; kahd n > kād n : k d n moʷчлʦayὕoʤt ; Yaᴥ development can be seen not only in neighbouring Tajik dialects but in other languages/dialects ʳuϑʦлaʳлρʣʦуʣфʲānХ colloquial Persian, Afghan αaʲХчлShughnХ, Uzbekчлςʲdūлʣtϑ. *Proto-YaghnōϐХл ʳʦoʲtл voʷʣὕʳл *a, *i and *u were reduced in an open syllable and they changed into super-short vowels ⁱ and . These ultra-short vowels are also of svarabhakti origin as they were inserted to break word-initial consonant cluster (See chapter II.1.5.). As an epenthetic svarabhakti vowel may appear either super-short vowels ⁱ a or short a (svarabhakti a mainly in the Eastern dialect, instead of svarabhakti a often there is ⁱ in the Western dialect). Super-short vowels ⁱ, and short a thus may have twofold origin: 1) < ŋă, ŋĭ, ŋ : < #C_C- < *#CC-. We can observe some regularities reduction of short vowels and epenthesis of svarabhakti vowels there – they can be better observed mainly in the Western YaghnōϐХ: in majority of examples the super-short vowel is realized as ⁱ, e.g. * > ʣvʣnʧnʥ ; * > Yagh. ϐʲotʦʣʲ ; *ϑ ar- > Yagh. t‘”á›- ‖лtⁱ”á›- to givʣ ; *ϑray > Yagh. sa›á₍ ‖лtⁱ›á₍ tʦʲʣʣ ; but when the reduced or svarabhakti vowel was followed by a labial sound or h and a stressed back vowels (i.e. , , ), the short/epenthetic vowel has been labialized: ŋ(čə)ϑ > Yagh. 133 There is no such change in Afghan αaʲХлуKISELEVA 1985, 23 фчлʳamʣлaʳлʧnлἰāϐuὕХлἵʣʲʳʧanлуDOROFEEVA 1960, 13). 14 ·89· t (also tⁱ ; t , t ) ʤouʲ ; > Yagh. n (also nⁱ ) pʲayʣʲ . The super-short vowels emerged also in loan-words, e.g. Pers. > Yagh. b ʳpʲʧnʥуtʧmʣф ; Arab. ḫabar > Pers. ⁾‘’á› > Yagh. ⁾‘pá› ‖л ⁾ⁱpá› nʣʷʳчл ʲʣpoʲt ; Rus. > Yagh. m t mʧnutʣ ; Rus. > Yagh. tⁱ›áktⁱ› tʲaϑtoʲ . As for articulation of ⁱ and it is qualitatively identical with their non-ʲʣduϑʣd лvaʲʧʣtʧʣʳ i, u, i.e. ⁱ can be realized as [ʧ ~ ɪ ~лʣл~лĕ] and [uъŭл~ / ~лoъŏ]. YaghnōϐХ super-short vowels are basically very similar to the super-short vowels ᵊ (< ă), ⁱ and in Zarafshānл dʧaὕʣϑtʳ (cf. KHROMOV 1958; KHROMOV 1962, 17-26; KHROMOV 1969, 306), and do not considerably differ from pronunciation of short vowels in an open unstressed syllable in Standard Tajik (PERRY 2005, 15-22), the only exception is YaghnōϐХ a, which does not reduce either in quality or in quantity, it remains stable regardless of stress position. In contemporary YaghnōϐХ (probably under influence of Tajik) distinction of opposition ὕonʥл×лʳʦoʲtлvoʷʣὕлgradually disappears which led to quantitative reform of the vowel system – historical long and short vowels in stressed position behave as long vowels, long or short vowels in closed syllable or historical long vowels in open syllable behave as short vowels; and short vowels in an unstressed open syllable are realized as super-short. Thus a new opposition comes to existence: from the historical opposition short ×лlong vowel there is super-short (reduced) ×л short (non- reduced) ×лlong (stressed) vowel, while the difference in the quantity of the latter two is given only by the position of stress. Development of Iranian vowels in Sogdian and in YaghnōϐХлϑanлϐʣлϑʦaʲaϑtʣʲʧzʣdлaʳлʤoὕὕoʷʳ: II.1.2.1. *a, ŋą i. ii. iii. iv. v. vi. (in a stressed position or as a part of a diphthong) > Sogd. a, Yagh. a: Sogd snk(ʾ) sng /ʳáṁg Yagh. sánk(‘) ʳtonʣ < *aʦáng‘-(ka-), Ave. asənga-, OPers. aϑanga-, Pers. sang; (in an unstressed position) > Sogd. ə, Yagh. a: Sogd. ʾsp-y /əʳpíъчл (ʾ)sp-y /( фʳpí/, Yagh. asp ʦoʲʳʣ < ŋáʦu‘-, Ave. aspa-, OPers. asa-, Ved. áśv‘-; (in an unstressed position) > Sogd. ø, Yagh. ø: Sogd. √pnʾ₍ ъ√p Yagh. pⁱn - ‖л pⁱná - toлὕoʳʣ < Ir. *apa- ʦ‘ ‘-; (word-initially in an unstressed position before a syllable containing * or ŋ ) > Sogd. , Yagh. ē: Sogd. ₎₍›t(ʾ)k zyrtyh /zēṙt /, Yagh. yʣὕὕoʷ < *ʣá› t‘-ka-, Ave. zaⁱrita-; Sogd √npʾy /√nəpē /, Yagh. nⁱp d- to ʳὕʣʣp < *ni-pád(‘) ‘-, Ave. nipaⁱ iia-; (word-initially before *nk or *ng under influence of a following syllable containing * or ŋ ) > Sogd. a, Yagh. i: Sogd ʾnkʾyr /áṁgir/, Yagh. ínk › fiʲʣpὕaϑʣ < *ham-gá› ‘-; (word-initially, mainly before a nasal or *s, *šлor after ŋǰ) > Sogd. ~ ə, Yagh. a: Sogd. √ (y)nd : √ (y)sṯ- /√ ɨṁd : √ ɨst-/, Yagh. vant- : vást‘ toл ϐʧnd (pres. : past part.) < ŋ’ánd‘- : ŋ’ásta-(ka-), Pers. ’‘stán : ’‘nd-; Sogd. √jyṯ- /√žɨt-/ to strike (past part.) < ŋǰát‘-, Pers. ₎‘dán : ₎‘n- уδκπлέьы6-113); ·90· vii. viii. ix. x. xi. xii. xiii. xiv. xv. xvi. xvii. xviii. 134 135 (word-initially in an unstressed reduced syllable before a syllable containing * or ŋ ) > Sogd. i, Yagh. i: Sogd. my ʾn m₍d(ʾ)n /m n/, Yagh. bⁱd n mʧddὕʣ < ŋm‘d na-, Ave. maⁱ ān‘-; (under effect of i-Umlaut before a syllable containing ŋ‘ ‘ > ŋ ‘) > Sogd. i, Yagh. i: Sogd. √ʾ₎⁽ʾ₍›t √(ʾ)₎⁽ʾ₍›t √₎⁽ʾ₍›t √₎⁽₍›ṯ /√ zwíṙt/, Yagh. zⁱ⁽í›t- to tuʲn < *uʣ-uá›t(‘) ‘-; (under effect of u-Umlaut) > Sogd. , Yagh. a (?): Sogd. ʾtyxw ʾtw (w), ʾtxw ʾtʾw , ʾtw (w), ʾtwx ʾtwx ʾṯwx / °)/ ʦappy < * ə⁾° < * a-á⁾uā-; Sogd. 134 fswx fsx /f ʳōʸ/ parasang < *fra-ʦá(n)⁾u‘-, Pers. ”‘›sá⁾ ; Sogd kwf / , Yagh. xaf ʤoam < ŋká”u‘-, Ave. kafa-; (before *rt, in verbal stems also before *rϑ, ŋ›č) > Sogd. a (> ā?), Yagh. : Sogd. g man л< ŋmá›t ‘-, OPers. martiya-; Sogd. srt /saṙt/, mrty mrtyy /máṙt /, Yagh. Yagh. s ›t ϑoὕd ; Sogd √ ₍(ʾ)›t, √ʾ ₍ʾ›t √ ₍(ʾ)›t √ yrt √’₍›ṯ /√ yáṙt/, Yagh. to find (past part.) < *abi-ar-ta-(ka-); (word-initially before *n{k, g, x, x } under influence of a following syllable containing or ŋu) > Sogd. a, Yagh. u: Sogd ʾnk⁽ t ʾng⁽ t /aṁg{əštъчл Yagh. unkú t finʥʣʲ л < ŋángu t‘-, Ave. ‘ngu t‘-; Sogd. ʾn wsty ъáṁʸ°ə gh. un⁾‘st‘g < ŋhám-x asta-ka-; (in vicinity of a labial sound) > Sogd. u, Yagh. u: Sogd. /√p uʤʳ-/, Yagh. b dú”s- to attach, to ʥὕuʣ < *upa-dáfʦa-; (result of metathesis) > Sogd. u, Yagh. u: Sogd. ⁽⁾⁽ ⁽, w ⁽ ⁽, ʾ ⁽ ⁽ w ⁽ ⁽, ʾ ⁽ ⁽ ⁾⁽ ⁽ /{əʸšú, ʸ{əšú/, Yagh. u⁾ ʳʧʸ < ŋ⁾ú u < *x á u < ŋ⁾ uá ‘m; *ah > Sogd. i ( ), Yagh. i: Sogd. ʾsp-y /əʳpíъчл (ʾ)sp-y /( фʳpí/, Yagh. asp ʦoʲʳʣ < *áʦu‘-, Ave. aspa-; Sogd. ʾym ym /ɨm/, Yagh. im ψζϊл am < *áhmi, OAve. ‘hmД, OPers. ahmiy, Ved. ásm ; Sogd. ky /ki/, Yagh. ki ʷʦʧϑʦ < *kah, Ave. k ; ŋ‘h ‘ (in an unstressed position) > Sogd. Д (ē ?), Yagh. i: Sogd. ntmy nṯmy / áṁdəmХъчлYagh. ámtun (< ántum ) ʷʦʣat (obl. sg. < gen. sg.) < ŋgántum‘h ‘ (GMS έэыяф; *am (in an ending) > Sogd. u, Yagh. u/ʉ/ø: Sogd. (ʾ)pw pw / pú/, Yagh. pʉᵎ ʷʧtʦout < *apám; Sogd. ʾzw zw /(əфzúъчлYagh. (arch.) az135 ζ < *áʣam, Ave. azəm, OPers. adam; Sogd ⁽⁾⁽ ⁽, w ⁽ ⁽, ʾ ⁽ ⁽ w ⁽ ⁽, ʾ ⁽ ⁽ ⁾⁽ ⁽ /{əʸšú, ʸ{əšúъ, Yagh. u⁾ ʳʧʸ < ŋ⁾ uá ‘m; *anϑ > Sogd. ā, Yagh. : Sogd. tʾrʾk /tāʲēъчлYagh. daʲ₎nʣʳʳ < *tánϑra-ka-, Ave. tąϑraka-, Pers. ; *(a) (in an unstressed position) > Sogd. Д, Yagh. i: Sogd. mʾny(h) mʾny mʧnd (loc. sg.) < ŋm n ‘ < ŋm n‘ ā; ʾ sʾn f(n)sʾx fsx) / ʤʳáṁx/ < fra-ʦán⁾(u)a-, Pers. ”‘s›áng, p‘›sáng. See GAUTHIOT – BENVENISTE 1929, 108-109. ·91· xix. xx. xxi. ŋu‘ (following *x, *h, *r) > Sogd. u, Yagh. u: Sogd. w n- xw n- /ʸu níъчлYagh. xuvn/xumn dʲʣam < *huá”n‘-, Ave. x afna-; ŋu‘ (affected by i-Umlaut) > Sogd. (> ), Yagh. ʉ: Sogd. wyr xwyrлъʸü later wr xwr ъʸōʲъф, Yagh. xʉr tʦʣ ʳun < ŋhuá› ‘-; ŋu (in an unstressed position) > Sogd. u, Yagh. u: Sogd wtʾrnk /ʸutáṙ Yagh. ⁾utánn‘ water-mill < ŋhu‘t(a)-ᛑn‘-ka-; (ad ii.) Reduction of unstressed *a, *i, *u and their merger to ə (/ фл ʧʳл ʳʧmʧὕaʲл toл κun₍Хчл where unstressed ə, ă, (< *i, *a, *u) phonetically all merge to Schwa: Munj. ə) w r-y wfr-y /wəʤʲíъ;лḲagh. ⁽á”ⁱ› ʳnoʷ ; ʳnoʷ л<лŋuá”›a(ad x.) Similar development can be seen in several Avestan examples, e.g. vā əm ϑʦaʲʧot < ŋuárta-, ϑ ā əm quʧϑ₎ὕy л<л*ϑuá›t‘- or ’ā ā›əm ʦoʲʳʣman л<лŋ’á›tā›‘-; comparable can be also Pers. sāl yʣaʲ л<лOPers. ϑarda- < *ʦarda- (MORGENSTIERNE 1973, 46)136; II.1.2.2. ŋā, ŋā i. ii. iii. iv. v. vi. vii. viii. (in majority of cases) > Sogd. ā, Yagh. : Sogd ʾʾph ʾʾp(h) ʾʾp ʾp ā-p /āpъчл Yagh. p ʷatʣʲ < ŋāp-, Ave. āp-; Sogd. pʾ /pā ъчл Yagh. da ʤootчл ὕʣʥ < ₎ʾ₍ ₎ʾ₍(₍) /zāʧ/ ʣaʲtʦ ч Yagh. ₎ ₍ *p da-(ka-), Ave. pā a-, OPers. pād‘-; Sogd. fiʣὕd < *ʣ ‘-; (shortened when following a preceding long vowel) > Sogd. a, Yagh. a: Sogd. rwps / əs/, Yagh. ʤoʸ < ŋ›áup ʦa-, Pers. , Ved. lopāśá-; (unstressed before ŋ. ) > Sogd. ə, Yagh. i: Sogd. s₍ʾʾk(h) s₍ʾk s₍ʾ‹ /sə gh. sⁱ ʳʦadoʷ л<л*aʦā - -(ka-), Ave. asaiia-, Pahl. sā₍‘g, Ved. cʰā₍ā- уδκπлέльэю-124); (before *. ) > Sogd. a, Yagh. a: ʾ ry ʾ ryw (ʾ) ry ʾ ry(y) ₍ / šaʧ/, Yagh. s‘›á₍ ‖лtⁱ›á₍ tʦʲʣʣ л<л*ϑ ‘-; Ave. ϑ›ā , Pers. se; (before *.u ) > Sogd. ə, Yagh. ?: Sogd nwʾʾz nʾwzyy (a scribal error?) /nə / ʳaʧὕoʲ < *nāu ʣa-(ka-), YAve. n‘uuā₎‘-, Parth. nā⁽ā₎, Ved. nāvājá- уδκπлέльэючльэ5ф; (before a syllable containing ŋ or * ) > Sogd. ē, Yagh. ‖лa : Sogd ⁽₍ (h) /ʷēšъчлYagh. w ‖л⁽‘ ʥʲaʳʳ < ŋu ʦt› ‘-, Ave. vāst› ‘-; Yagh. n s ‖лn‘ s noʳʣ < *n sn ‘-; (in original causative stems before an ending ŋ-(‘) ‘-) > Sogd. ē, Yagh. ‖л ‘ : Sogd. √s₍n /√ʳēn/, Yagh. s n- ‖л s‘ n- to ascend, to raise < *ʦ n‘ ‘-; Sogd. √pnʾ₍ /√p /, Yagh. pⁱn - ‖лpⁱná - to ὕoʳʣ < Ir. *apa- ʦ‘ ‘-, (LIVSHITS – KHROMOV 1981, 388); (shortened/reduced) > Sogd. a / ə, Yagh. a: Sogd p nʾ /pə ʦʣʣὕ , Yagh. pá n‘ heel of a shoe < ŋpá n‘-ka- < -ka- ʦʣʣὕ , Ave. pā n‘-, Pers. pā ná, Ved. ṇi-; Sogd. ʾʾmʾtʾ₍ ʾʾmʾt(ʾ)k ʾʾm(ʾ)ṯyy ʾmʾṯy, ʾmṯy ā-ka- ʲʣady ; Georg Morgenstierne quotes also comparable development in Eastern Norwegian: gāḷ yaʲd луλoʲʷʣʥʧanлgar(d), Danish g ›d) < OScand. g‘›ðʀ (MORGENSTIERNE 1973, 46). 136 ·92· ix. x. xi. (after loss of *h) > Sogd. ā, Yagh. : Sogd. √pt ⁽(ʾ)₍ √pt⁾⁽ʾ₍ √pṯ⁾⁽ʾ₍, √pṯ⁽⁾ʾ₍ /√p tʸ° /, Yagh. t ⁾ - to ₎ʧὕὕ < *pati-x h‘ ‘-; Sogd. xwʾr /ʸ°āʲъчлYagh. ⁾ › sister < *hu‘h‘›-, Ave. x ‘ h‘›; ŋām (in ā-stem obl. pl. ending) > Sogd. , Yagh. ?: Sogd. ⁽₍ n⁽ / ə tʦʣy < *auá ‘nām; ŋuā (affected by i-Umlaut) > Sogd. , Yagh. ?: Sogd √ ⁽ʾ₍› √xwyr √h₍u(“)-›, 137 √hu“-›, √hv“-› /√ʸü to ʤʣʣd < ŋhu ›‘ ‘-; II.1.2.3. *i i. ii. iii. iv. v. vi. vii. (in a stressed position or as a part of a diphthong) > Sogd. i, Yagh. i: Sogd ʾyntkʾw /íṁdku/ (< *íṁduk) Indian, ζndʧϑ < *híndu-ka-, OPers. hiⁿduya-, Pahl. h nd g, Pers. ; (in an unstressed position) > Sogd. ə/ , Yagh. i: Sogd ʾ⁾ (ʾ)y t-y h a wdi / ʸšɨ dí/, Yagh. xⁱ í”t mʧὕ₎ < ŋ⁾ uí”t‘-, Ave. ⁾ uu pt‘-; Sogd √ y - √zy -, √ y √j - /√žɨ -/ to ϑʦʣʷ , Yagh. živ- to sew, to stitch < ŋží’‘-; Sogd › k-ʾ, › k-h /rəš₎á/, Yagh. › k nit < *›í kā-, Pers. › k, Oss. l sk liskæ, Skt. lik - уδκπлέлььяф; (in an unstressed position) > Sogd. ø, Yagh. ø: Yagh. ž‘vá›- ‖ žⁱvá›- to bring, to produce, to ʧnvʣnt < ŋn ǰ-’ᛑ-; ŋ ‘ > Sogd. , Yagh. : Sogd. √ yr √’₍› /√ Хʲ/, Yagh. vД›- to find < *abi-ar-; Sogd. mrty mrtyy /máṙ Yagh. man < ŋmá›t ‘-, OPers. martiya-; ŋ ( )i > Sogd. Д, Yagh. Д: Sogd. tys /√tХʳ/, Yagh. tДs- to ʣntʣʲ < *ati-íʦa-; √n₍ √n₍d : √n₍st √n₍sṯ /√nХ śл√nХʳt/, Yagh. *ihi > Sogd. Д, Yagh. Д: Sogd. nДd- to ʳʧt < *nihida-; ŋu > Sogd. , Yagh. ?: Sogd √ wys /√xüɨs/ to ʳʷʣat < ŋhu s‘-, Ave. x Дs‘-; (ad ii.) Reduction of unstressed Ir. *a, *i, *u and their merger to ə (/ ) show similar dʣvʣὕopmʣntлʧnлκun₍Хчлʷʦʣʲʣл stressed or unstressed *i and unstressed *a, *u change to ə, ă, and nowadays they all merge to Schwa: Munj. s(ᵊ)pə₍ă ὕouʳʣ < *ʦuí ‘-; Ave. p ph ъšpəšáъчлḲagh. pú / ⁱpú ; yə ə [› ₍] tʦʣлḲʧdghāлὕanʥuaʥʣ л<л -ka-kaʾ₍ntkʾ⁽, ʾ₍ntʾ⁽k ъíṁd₎uчл íṁduk/ (cf. modern loans: Yagh. ζndʧan ;л κun₍щл ζndʧa < Pers. / ); (ad iv.) The change ŋ ‘ > Sogd. , Yagh. probably took place after lengthening of *a before *rt (see II.1.2.1.x.): Sogd. √ yr √’₍› /√ Хʲъл: √ ₍(ʾ)›t, √ʾ ₍ʾ›t √ ₍(ʾ)›t √ yrt √’₍›ṯ /√ yáṙt/, Yagh. vД›- : to find (pres. stem : past part.) л<лŋ(ă) í₍(ă)›- : ŋ(ă) ₍áṙta(ka)- < *abi-ar- : *abi-ar-ta-(ka-); other explanation of different forms of the present stem and of past participle of the word *abi-ar- can be explained as difference in stress (in such case probably 137 √⁾⁽ỿ›dʾ›ỵm ъ√ʸüēʲ- āʲХmъл ʷʣлʦavʣлϑauʳʣdлyouлtoлdʲʧn₎ луSIMS-WILLIAMS 1996a, 314). ·93· Stress I, which remained on preverb in the present stem: ŋ‘’í-ar-138, but shifted towards the Stress II in the past participle: ŋ‘’í-ar-ta- > *abi-á›-ta-); II.1.2.4. *Д i. > Sogd. Д, Yagh. Д: Sogd. k›Дṇ ; ii. *-Дm (in i-stem accusative ending ?) > Sogd. , Yagh. ?: Sogd (acc. sg.) < ŋuáhu(r)nДm уδκπлέю5ыщiv); √ rʾyn √xryn /√xrХn/, Yagh. xⁱr - to ϐuy < *xr -, Ved. xwrnyy /ʸownХ/ ϐὕood II.1.2.5. *u i. ii. iii. iv. v. vi. vii. (in a stressed position or as a part of a diphthong) > Sogd. u, Yagh. u: Sogd √n (ʾ)wnt /√nə úṁd/, Yagh. n únt- to dʲʣʳʳ < *ni-gúnd‘-; (in an unstressed position) > Sogd. ə/ , Yagh. u: Sogd. ntm nṯm / áṁdəm/, ⃝ py ›ʾk, ⃝p r Yagh. ámtun (< ántum) ʷʦʣat < ŋgántum‘-, Ave. gantuma-; Sogd. ⃝ py rʾk, ⃝p r, ⃝p ₍ ⃝p ₍ (as a part of compounds) /pɨš ʳon , Yagh. púl(l)‘ (?) ϐoyчл child; little, small < ŋpúϑra-(ka-), OPers. puç‘-; Sogd. ʾkwt-y kwt-y, qwt-y / ₎{ətíъчл k ʒ kuy уδκπлέлььŚф; Yagh. kut doʥ л< ŋkút (in an unstressed position) > Sogd. ø, Yagh. ø: Sogd. √ʾ₎⁽ʾ₍›t √(ʾ)₎⁽ʾ₍›t √₎⁽ʾ₍›t √₎⁽₍›ṯ ъ√ zʷíṙt/, Yagh. zⁱ⁽í›t- to tuʲn < ŋu₎-uá›t ‘- < *uʣ-uá›t‘ ‘-; (reduced sound in an initial syllable) > Sogd. ə/ , Yagh. i: Sogd. my rh my r-y /mɨžá, mɨž / m⁽ ʾkk /mužē/, Yagh. mí›d‘ ϐʣadчлpʣaʲὕ < ŋmú ra-(ka-), Ved. -; 139 (in a stressed position under effect of i-Umlaut) > Sogd. (> / i), Yagh. i: Sogd. rd √ny wynṯ /√n üɨṁd/ ψуʳфʦʣϊл dʲʣʳʳʣdл (3 pers. sg. impf.) < *ni-ʜa-gúnd‘ ‘-t, Khʷāʲ. /(əфn ʷʧndъ; Sogd. frʾʾ⁽₍ c₍ frʾwycyẖ /ʤʲāʷɨуšфčʧъчл Yagh. ‖ fⁱ /f oϐὕivʧouʳnʣʳʳ < ŋ”›āmú t -; wyzp- wzp- wjp- ʾ⁽ž’-ʾ /üɨžϐá/ < ŋú’ǰ ā terror ; wr xwr /ʸūʲъчлYagh. xʉr ʳun ŋuu‘ (following *x) > Sogd. , Yagh. : Sogd. Sogd. < ŋhuuár-, Ved. suvá›-; *hu- (followed by more than one consonant) > Sogd. u, Yagh. u: Sogd. √ʾ⁽ t√ʾ⁽’d-, √ʾ⁽”ṯ- /√u d-/, Yagh. ú”t‘ to sleep (past part.) < *húfta-(ka-); (ad ii.) Reduction of unstressed Ir. *u together with *a, *i and their merger to ə (/ ) can be ϑompaʲʣdлʷʧtʦлκun₍Хчлʷʦʣʲʣлunʳtʲʣʳʳʣdл*i, *a and *u changed to ə, ă, and nowadays they all (~ əm) ʷʦʣat < ŋgántum‘ntm nṯm merge to Schwa: Munj. ъ áṁdəm/, Yagh. ámtun/ ántumщлζnлκun₍Хл shifted to ə does not cause labialization of velars as in Sogdian. In the Iron dialect of Ossetic there is merger of *Proto-Ossetic *i and *u (Ir. < , 138 Position of stress on *i can be also caused by operation of Stress II after a syncope of word-initial *a-: ŋ‘’í-ar-/*abi-á›- (with Stress I either on a second syllable of prefix or on a root) > * í₍‘›- > * -. 139 According to Nicolas Sims-Williams (SIMS-WILLIAMS 1989b, 181). ·94· ) > (in Digoron they remained unchanged), but *u following a velar or uvular sound caused labialization of the preceding tectal: *{k, k, g, q, x, }u > {k, k, g, q, x, } ; II.1.2.6. * i. ii. > Sogd. , Yagh. ʉ: Sogd w / ū /, Yagh. ʉs ‖л ʉᵎt ʤaʣϑʣʳ < ϑa-, Ave. g ϑa-; Sogd. wr ъ ūʲъчлYagh. dʉr ʤaʲ < -, Pers. d ›; (in ablaut) > Sogd. , Yagh. : Sogd. (ʾ)k⁽ ʾkw, kʾ⁽ k(ʾ)⁽ / kу{фūъчл Yagh. k ʷʦʣʲʣ , OAve. k ; II.1.2.7. *› i. ii. iii. iv. v. vi. vii. viii. ix. x. Yagh. k‘lpás‘140 (dissimilation r > l) > Sogd. əʳ, Yagh. ar (?): Sogd. krpsʾk /kəwpə ὕʧzaʲd < Ir. ŋk p‘s‘-ka-, Ave. kahrpuna-, Pers. k‘›’ás, Tjk. k‘lpēsá; 141 ϐʧʲdчл ʤoʷὕ < Ir. > Sogd. əʳ, Yagh. ur (?): Sogd. ʾmr -y / m(əw) í/, Yagh. m ʲ *m ga-, Ave. mərə a-, Ved. m›gá-; > Sogd. iʳ, Yagh. ir: Sogd. kyrm-y qyrm-y /kiwmíъчлYagh. kí›(ⁱ)m ʳna₎ʣ < Ir. ŋk›m -, Ved. k m -; (in front of ŋ ) > Sogd. iʳ (ir), Yagh. ir: Sogd. √ṯr- /√tir-/, Yagh. tir- to ʥo < *t ‘-; Sogd. √m₍›- /√mɨr-/, Yagh. mir- toлdʧʣ < *m ‘-; > Sogd. iʳ, Yagh. i: Sogd. *tiʳ , Yagh. tí p‘ ʳouʲ < *t p‘-ka- (KHROMOV 1987, 653); k›m(ʾ)₍›, kyrmyr qrmyr qyrmyr gh. ‖ ʲʣd чл ἵaὕʦщл k‘›mД›; (following a labial sound) > Sogd. uʳ, Yagh. ur: Sogd. √ wrt- /√ uwt-/, Yagh. vú›t‘ √mwrt- √mwrṯ- /√muwt-/, Yagh. mú›t‘ to bring (past part.) < *b ta-(ka-); Sogd. to die (past part.) < *m ta-(ka-); √k⁽n(in present stem of the verb *kar- toл do ) > Sogd. u, Yagh. u: Sogd. √‹⁽n- /√kun-/, Yagh. kun- to do, to ma₎ʣ < *k n‘u‘-, Ave. kərənaoⁱti; Ved. k›nót ʦʣл doʣʳ ; (before *t, *ʦ, *ʣ, ŋ , ŋž, * < *g) > Sogd. ø, Yagh. ø: Sogd. √(ʾ)krt- √ʾkrt- √(ʾ)kt√(ʾ)qṯ- /√ᶤkt-/, Yagh. íkt‘ to do, to make (past part.) л < ŋk t‘-(ka-); Sogd y (ʾ)rt-y, yr t y rṯ-y /yəуwф dí/, Yagh. yaxt (ya d) ʷʧdʣ < ŋu -g t‘-; (before ŋ (< Ide. ŋtk)) > Sogd. , Yagh. ?: Sogd ʾ h / š šáъ ϐʣaʲ < *ʜ a-, YAve. ‘› ‘-, Khʷāʲщлhrs, Pers. xirs, Ved. k a-, Ide. *h ›tko- уδκπлέь55ф; (before *nt/*nd, *nk/*ng, *xt, *xs, *g) > Sogd. ra, Yagh. ?: Sogd. √ʾnkrʾnt √ʾngrnd ṯrnng- / tʲáṁg-/ oppression < /√áṁgraṁd/ to cut < *h‘m-k nt‘-; Sogd. *abi-t nga- (ŋ‘’ -t nka-) уδκπлέль5эф; The YaghnōϐХлʤoʲmлʷaʳл pʲoϐaϐὕyлʧnfluenced by Tjk. k‘lpēsá (this word itself can be of Sogdian origin), Pers. k‘›’ás. 141 A Persian loan: Pers. mur ϐʧʲdчлʤoʷὕ Š 140 ·95· xi. xii. xiii. xiv. xv. (before *ft, ŋnč) > Sogd. ri, Yagh. ri: Sogd. √pt ry t- /√p t ʲʧ d-/ to take (past part.) , Yagh. ⁱ›í”t‘ to know, to understand (past part.) < ŋ(p‘t -)g ”t‘-(ka-) < ŋg›’- to know, to understand, to ta₎ʣ , Ave. gərəbiia-, OPers. g›’ā₍‘-, Pers. g › ”tán : gД›- to ta₎ʣ (GMS έь5юa); (before *fʦ or in the cluster ŋm›ʦ) > Sogd. ru, Yagh. ?: Sogd. √z rw s- /√z ʲufs-/ to be raised < *uʣ-g ”ʦa-; Sogd. √pcmrws- /√p čmʲuʳ-/ to touϑʦ < *pati-m ʦa- уδκπлέь5юb); ŋ›n > Sogd. ʳn, Yagh. n(n): Sogd. pwrn-y /puwníъ142, Yagh. pun(n) ʤuὕὕ < ŋp n‘-; ŋu›- > Sogd. wiʳ, Yagh. ur: Sogd. wyrk-y wyrq-y /wiw₎íъчлYagh. urk ʷoὕʤ < ŋu ka-, Ave. vəhrka-, Pers. gurg, Ved. v k‘-; ŋu› (before *t, *s, *z, ŋ , ŋž, * < *g) > Sogd. u, Yagh. ?: Sogd. ⁽ n-y /{ušní/ ʦunʥʲy < ŋu n‘-, colloq. Tjk. gu ná, Tehr. go né < Pers. gu›usná; II.1.2.8. * 143 i. > Sogd. ē, Yagh. Д: Sogd. ii. ŋ(h)‘ (word-initially) > Sogd. , Yagh. Дйē: Yagh. Д⁾йē⁾ ʧϑʣ < *á ⁾‘-, Ave. ‘ē⁾‘-; Sogd yy n /|ɨ ʸán/ ʥὕaϑʧʣʲ < Ir. *á ⁾‘-dān‘-; Sogd yttkw ytkw ytqw /|ɨt₎úъл<л* túk, Yagh. ētkйДtk ϐʲʧdʥʣ < *há tu-ka-; Sogd. ʙ л ʾzm-y /ɨzmíъчл Yagh. zⁱm fiʲʣʷood < ŋá ₎ma-, Ave. ‘ēsm‘-, Khʷāʲ. ʾzm, Pers. hē₎úm, Ved. idʰmá-; ŋ‘ (a) > Sogd. ē, Yagh. ē: Sogd. ›ʾ₍k(ʾ) ryk ъ ʲ kуăф/, Yagh. ⁱ ϑὕayчлʣaʲtʦ < Ir. ŋg›á ‘-ka-; Sogd. ʾspʾ₍t(ʾk), ʾsp(ʾ)₍tk, (ʾ)spʾ₍tk, (ʾ)spʾ₍t₍ spyṯy /ᶤ ( )/, Yagh. sⁱ ʷʦʧtʣ < *ʦuá t‘-(ka-), Ave. sp‘ēt‘-; *- ‘- (in the word *ϑ ‘-) > Sogd. ‘ (?), Yagh. ay: Sogd. ʾ ry g ʾ ryw (ʾ) ry ʾ ry(y) ₍ / šaʧ/, Yagh. s‘›á₍ ‖лtⁱrá₍ tʦʲʣʣ л< *ϑ ‘-; Ave. ϑ›ā , Pers. se; ŋ‘ t (in the word ŋ’‘ t-) > Sogd. ē, Yagh. ē: Sogd. yk yk, yq byq / ē₎ъчлYagh. vēk ʳʧdʣчлoutʳʧdʣ;лʣʸtʣʲnaὕ < Ir. ŋ’‘ t-k -, Ave. ’ t, ’ā, ’ē; ⁽₍ t› ⁾⁽(₍) t› ⁾⁽ ṯr /ʸüēštər ər/ ϑʦʧʣʤ < ŋu > Sogd. , Yagh. ?: Sogd. *hu ( )- t‘-tara-, Ave. h t‘- tʦʣ ʦʧʥʦʣʲлуonʣф ; iii. iv. v. vi. ʾy /ē ъчлYagh. Дd tʦʧʳ < *á ta-; (ad iv.) Preservation of the diphthong in the cluster *- ‘- in forms of the numeral л ъʦəʲēyъчл ϑ ‘- tʦʲʣʣ , is similar in the other Eastern Iranian languages, e.g.: Bactr. Shugh. aray, οōsh. Bart. ‘›ā₍, πaʲХq. aroy, Ishk. r (y), Sangl. › ₍, Yazgh. c y, Wakh. tr (y), Munj. ⁾ ›‘₍, Yidgh. ⁾ ›‘₍, ⁾u›o₍ ×лOss. æ›tæ, Pasht. d›ē, Khōt. drai, Tumshuq. dre, Khʷāʲ. ₍ ъšēъ; cf. Tjk. dial. of Tafkōn (MALLITSKIY 1924). Such feature is related to shortening of In Sogdian dialect of Zhetisu ŋ›n > n(n): pwn ъpunуnфъл ʤuὕὕ л<лŋp n‘- (LIVSHITS 2008, 351-352). In *Proto-Sogdic there we can observe instead of the expected diphthong ŋ its innovated form ŋ‘ , or even ŋæ ; cf. also spelling of this diphthong in Avestan: ‘ē. 142 143 ·96· ŋā shown in II.1.2.2.iv-v., it is an Eastern Iranian isogloss ŋā > ŋă in front of ŋ or ŋu (cf. MACKENZIE 1988, 88-89), in majority of the Eastern Iranʧanл ὕanʥuaʥʣʳл tʦʧʳл nʣʷ л ŋă has not been contracted ŋ‘ (a) > *ē, the diphthong has been usually preserved, but it could have undergone some later changes in various languages. Aϑϑoʲdʧnʥл toл ʳpʣὕὕʧnʥл oʤл tʦʣл л umʣʲaὕл tʦʲʣʣ л ʧnл tʦʣл κountл κuʥʦл doϑumʣntʳл aʳл ʾ ryw (< *ϑ›á ‘m ??флζtлϑanлϐʣлʳuppoʳʣdчлtʦatлtʦʣлmumʣʲaὕл tʦʲʣʣ лʷaʳлpʲonounϑʣdл*ᵊ ‘ , which may later changed to *ᵊ ē. 144 II.1.2.9. * u i. ii. iii. iv. v. > Sogd. , Yagh. : Sogd. w h w / ō ъчлYagh. v d ʳϑʣnt < ’áud‘-, Ave. baoⁱ i-; Sogd. ⁽k ⁽‹ /šō₎ъчлYagh. k ʳʧὕʣnt < *a-ʦráuk‘-; Sogd. rwps / pas/, Yagh. pas ʤoʸ < *ráup ʦa-, Pers. h, Ved. lopāśá-; Sogd rw n ro haṃ, ro aṃ / n/, Yagh. ⁱn, an oʧὕчл ϐuttʣʲ < ŋ›áugn‘-, Ave. rao na-, Pers. › án, Tjk. ›‘u án; (preceding *xm, ŋ⁾ (u)) > Sogd. o, Yagh. a: Sogd. t m-y ṯxm-y /toxmí/, t my ṯ(w)xmy / Yagh. taxm ʣʥʥчлʳʣʣd < *taoxma-(ka-) < *táu⁾m‘n-, Ave. taoxman-, OPers. t‘umā-, Pahl. t m, Pers. tuxm; Sogd. rw n-y ›⁽⁾ n-y /ʲoʸšní/ ὕʧʥʦtл(of colour) , Yagh. ›á⁾ ⁱn daʷn < *rao⁾ n‘- < ŋ›áu⁾ n‘-, Ave. ›‘o⁾ n‘-; (affected by i-Umlaut) > Sogd. , Yagh. ‖л‘ (?): Sogd √tw(y) , √twz √twj √ṯ⁽ž /√tü > √tōž/ to pay < ŋt‘uǰ‘ ‘-; Sogd √pck⁽₍› ъ√p č₎ü Yagh. čⁱk r- ‖л č ká ›- to ʤʣaʲ < ŋp‘t -k u›‘ ‘- (?); (affected by i-Umlaut after dissimilation or in vicinity of a labial sound) > Sogd. , Yagh. ē: Sogd. wt(ʾ)ynh / ⁾uət quʣʣn < *hu‘-t unД-; Sogd. Sogd. g yp xyp ( ) xỿpϑ h“-p / ə) /, Yagh. ⁾ēp oʷnчл ʳʣὕʤ < ŋ⁾ ϑ < ŋhu -paϑ ‘-, Ave. x ‘ēp‘ϑiia- у×лYagh. xap < *hu -paϑa-); ŋ ‘u (affected by i-Umlaut) > Sogd. , Yagh. i: Sogd. ʾync(h), ynch ʾync /|ɨṁǰ/, Yagh. nč ʷomanчлʷʧʤʣ < ŋ áun -kā-. g ʾync(h) ync II.1.3. Consonants The (Eastern) Old Iranian system of 24 (26) consonants (*p, *t, *k, ŋč, *b, *d, *g, ŋǰ, *f, *ϑ, *x, *x йŋhu, ŋ , ŋž, *m, *n, *r, (*l), *s, *h, *z, *ʦ, *ʣ, ŋ , ŋu, (*ʜ)) underwent a number of changes during its development towards Sogdian and YaghnōϐХ, Sogdian had 25 consonants (p, t, k, č, ʦ, f, ϑ, x, ⁾°, , , , , , ž, ž, m, n, r, l, s, h, z, y, w)145 and YaghnōϐХ has 28 consonants (p, t, k, č, q, Instead of the original diphthong ŋ u we can expect an innovated form ŋ‘ , or ŋ in *Proto-Sogdic; see spelling of this diphthong in Avestan: ao. 145 With five voiced allophones b, d, g, ǰ, ʣ and four labialized allophones (originating from unstressed *u > ə/ ) ᵊ k , ᵊ g , ᵊ ⁾ , ᵊ . In Sanskrit and Turkic loan-words may appear ṭ, ṭʰ, ḍ, ṇ, q; and in Aramaic loans (less) possibly , , sˁ (~ ʦˁ), tˁ, А. Status of l and h in inherited lexicon is unclear. 144 ·97· b, d, g, ǰ, f, x, ⁾, , v, , ž, m, n, , r, l, s, h, z, y, w, , ᴥ)146. For the Eastern Middle Iranian languages (at least in their early stages) is characteristic a phonemic opposition of voiceless stops and voiced fricatives *p : * , *t : * , *k : * ж ŋč : ŋž which emerged after spirantization of voiced stops (*b, *d, *g, ŋǰ > * , * , * , ŋž) 147. This led to threefold opposition of consonants (see Table 36), other *Proto-Sogdic consonants (except voice opposition *s : *z) do not have any phonemic opposition. The threefold opposition was replaced by fourfold opposition (i.e. voiced ×лvoʧϑʣὕʣʳʳлʳtopлandлvoʧϑʣdл×лvoʧϑʣὕʣʳʳлʤʲʧϑative) in both languages, but origin of new voiced stops differ. In Sogdian the new voiced stops emerged from voicing of voiceless stops after voiced fricatives or after a nasal (or better after a vocalic nasal prolongation ṁ), in YaghnōϐХл voʧϑʣdл stops emerged from voice assimilation (certainly for b and d), later YaghnōϐХлvoʧϑʣdлʳtopʳлʷʣʲʣл borrowed from Tajik, Arabic and Uzbek. *p *f * : *t *ϑ * : *k *x * : ŋ ŋč ŋž Table 36 Threefold opposition of consonants in *Proto-Sogdic. In total (i.e. with allophones) Sogdian consonant system consisted of 34 consonants (excluding consonants appearing only in loans), outline of consonantal sounds with their representation in alphabets utilized for Sogdian is presented in Table 37. Real number of consonants was certainly smaller, e.g. voiced stops b, d, g, (g ), ǰ (and ʣ) can be classified only as allophones of their voiceless counterparts p, t, k, (k ), č, ʦ; phonemes k , g , ⁾ , can be also considered as allophonous. Questionable is whether these sounds were labiovelars or (secondary) labialized velars. Stress shift (probably Stress III) caused reduction of historical short vowels in unstressed position, following a velar this historical unstressed reduced *u was still spelled by the letter waw. It can be supposed that *Proto-Sogdic (or *Proto-Sogdian) velars were secondary labialized when they preceded and most likely also before ; later when *u was reduced to ə – the change was not reflected in spelling after velars, and continues to be written with the letter waw, in this case it is either archaic spelling or an attempt to spell labialized characteristics of a preceding velar. There were probably two /ʸ{/ sounds in Sogdian, respectively it was of two sources: 1) it is a continuant of Iranian *x (< *hu < IIr., Ide. ŋsu), and 2) it is a result of secondary labialization of *Proto-Sogdic *x. In documents written in the Sogdian script an indirect result of labialization of velars can be marked word-initially by spelling with a prothetic Schwa ᵊ <‚>: Sogd ʾkwt-y kwt-y, qwt-y) / ₎{ətíъчл Yagh. kut doʥ ; Sogd ʾ⁾⁽ t›-y ʾ ⁽ t›-y ⁾⁽ t›-y) / ʸ{əštʲíъ ϑamʣὕ < ŋu⁾ t›‘-; Sogd ʾ⁾⁽mtʾn / ḱ{ə Khumdān, Xianyang (city in China) [KhōtщлHumdān, Syr. h⁽mdʾn]; prothesis of ᵊ does not appear before vowels beginning in historical *x (ŋhu) √ wr- ъ√ʸ°ər-ъчл notл †ъ√ ʸ{ər-/, Yagh. ⁾‘›- to ʣat . AϑϑoʲdʧnʥлtoлdoϑumʣntʳлʧnлtʦʣлBʲāʦmХлʳϑʲʧptлtʦʣлὕaϐʧaὕʧzʣdлvʣὕaʲʳл later lost their While the consonants q, g, ǰ, l, h, , ᴥ appear only in borrowed lexicon, b is rare in genuine YaghnōϐХлʷoʲdʳщлζtл is possible that l and h can be inherited in some cases. 147 πʣʣлanaὕoʥʧϑaὕлʳʧtuatʧonлʧnлεʣὕὕʣnʧʳtʧϑлδʲʣʣ₎śл лъpълśл ъ ъчл лъtълśл лъ ъчл лъ₎ълśл лъ)/ (BROWNING 1983, 26-27). 146 ·98· labial character, e.g. Sogd knā /kənāъл×лSogd k⁽nʾ ‹⁽nʾ /₎{ənā/, Yagh. kun doмчлma₎ʣ! nd (2 pers. sg. imper. pres.) ; delabialization can be better seen in the case of ⁾° – for the stem ⁾°ər- (Yagh. ⁾‘›-) to ʣat , there are attested following forms: Sogd ho-rt / ṙt/148 √ wrt rd rd /ʸ°aṙt/ (inf.), hu ›ā-ṯ / < /ʸ°ə , Yagh. ⁾‘› (3 pers. sg. sbjn.), hu re /ʸuʲē/ < /ʸ°əʲē/ (3 pers. sg. opt.); similarly delabialization can be seen also on Sogd hu tte /ʸutē/ ×лSogd xwty, wty ʙ ⁽tʾ(ʾ)₍, wty xwty xwṯy, xwdy /ʸ°ətíъ, Yagh. ⁾‘t oʷnчлʳʣὕʤ .149 In Sogdian there may be assumed special treatment of clusters d and following a reduced *u: this phenomenon can be well demonstrated on doublets in following examples 150: w r, w th wxth w t(ʾ) dw ṯ(ʾ) wth dauʥʦtʣʲ л<л*duxtar151 sw (ʾ)yk s w yk πoʥdʧan л<л*ʦug(u)d ‘- - . These examples show probable development *Cu D (D = voiced dental stop or fricative) > *Cə D (certainly not *C əD as there is no prothesis preceding the consonant cluster), i.e. the d/ cluster was probably realised as *C{u } D > *C{ə } D or even *C{u D} > *C{ə D} . Similar development may be assumed also for s ⁽tmʾn aὕὕ ( sw ṯmʾn s tmʾn) I suppose that the letter waw marks labialization i.e. *sə də , unfortunately, etymology of this word is not given in GMS and is neither known to me, Ilya Gershevitch interprets the letter waw as epenthesis of u (GMS έяřэфщ consonant p Sogdian alphabet p Manichaean alphabet p Syriac alphabet p BʲāʦmХ script t, tt, ṭ, ṯ k, ḵ t t t, ṯ ṯ (/t) k k č c k, q c k, q c ʦ (c), ts (c), ts c 152 p, pʰ, p c, ccʰ, j b p, ( ) p, b p, b b d t, ( ) t, ṯ, d, dt, dṯ ṯ (/t), d g k, ( ) k, q, g k, q, g d, dʰ, t, ṯ g εʣʲʣл ʧʳл anл ʧntʣʲʣʳtʧnʥл ʳʧmʧὕaʲʧtyл ʷʧtʦл ρā₍Х₎л уandл αaʲХŠфśл ἵʣʲʳщл x ‘›đán > x u›đán : ρ₍₎щл ъʸuʲdán/, AfghP. ho-rt ṙt/ can be seen a feature similar to lengthening of *a before ›_{čж tж rd to eat (present stem : 3 pers. sg. pres.) щ *ϑ} in YaghnōϐХлvʣʲϐaὕлʲootʳл– e.g. Yagh. ⁾‘›- : ⁾ 149 Delabialization is unclear in these cases – in the word kə < k ə this feature can be seen well. In examples ṙt < *⁾ ‘ṙt; < *⁾ ə ; < ŋ⁾ ə and < *⁾ ətí labial character of ⁾ is of different origin – it continues from Iranian *x and not from secondary labialization caused by reduction of unstressed *u. It is possible that the spelling ho-rt, hu ›ā-ṯ, hu re and hu tte should be read with initial ⁾ -: / ṙtчл ʸ{ə əʲ , ʸ{ətí/, i.e., graphemes ho and hu represented and ⁾ ə; in the second case there thus can be seen again the omission of spelling of Schwa in an open syllable (see excursion 4). 150 In most cases I will put down only spelling varieties in the Sogdian script (i.e. secular texts in the Sogdian script or Buddhist texts), in the Manichaean and Syriac scripts there are no attested such examples of metathesis. In majority of example I will not give phonetic transcription. 151 sw (ʾ)₍k or s w yk can be also explained as development from *ʦugud ‘-ka-, cf. OPers. spelling sw ₍ʾ⁽ s w ₍ʾ⁽. <sᵃ-u-g -dᵃ>, <sᵃ-u-g -u-dᵃ> or <sᵃ-u-gᵃ-dᵃ 152 Here will be given only consonantal part of ak ara. 148 ·99· consonant ǰ (ʣ) f Sogdian alphabet c (c) , fж pж p ϑ c, j (c) f Syriac alphabet c (c) f BʲāʦmХ script j ϑ (/ϑ) x tʰ-, ṭ, ṯ h Manichaean alphabet w, wv x x, ( ) , x ⁾° xw, ( w) xw xw ⁾,( ) ⁾ ⁾ h b v, wv bṯ, bd d wṯ, wt, ⁽dʰ ⁾ d t t 152 , ḍʰ ,h d s t, ṯ, d t s s s z, , ₎ z jж ž z ž s ś, , z ž ź, ś, ś, , r ž z z, , ₎, r m jж ž ž ź, ś, m m m m, m n n n n ṁb mp, mb mp ṁd mp, np nt n, ṇ, (-)ṃ mp nt, nṯ, nd, ndt, ndṯ nṯ ṁg nk n(n)g, nk, nq ṁǰ nc ng, nk, nq nc r r nc, nj r [l] y r, l, w [h] [ṭ] [ṭʰ] [ḍ] [ṇ] [q] {ṭ} {q} {} {ʾ} { } {} r r, ṟ l, l l, ḻ y y y y w w w x x, h h, ẖ r, rt r rn v, u ø (щ) h ṭ, ṭʰ, ṯ ḍ ṇ (hk) rn x, (ẍ) ṯ q c (ʾ) ẖ (c) ⁾ж (ʾ) (x) () Table 37 Spelling of consonants in the Sogdian, Manichaean and Syriac script and in the BʲāʦmХ script. ·100· In YaghnōϐХлtʦʣʲʣлaʲʣлnoлtʲaϑʣʳлoʤлхἵʲoto-Sogdian labialization of velars (but it is possible that this feature was already *Proto-Sogdic), the only possible example of a reflex of labialization can be found in Yagh č ká ›‘k čⁱk rakфл to ʤʣaʲ л– a precise etymology of this verb is not known to me, it may have been derived from ŋp‘t -k u›‘ ‘-?; in Sogdian there is attested √pck⁽₍› ъ√p č₎{ēʲл~ √p č₎üēʲ/, so probably svarabhakti in Western YaghnōϐХлϑanлϐʣлaлʲʣflex of original k (according to development of reduced vowels in Western YaghnōϐХлʷʣлʳʦouὕdлʣʸpʣϑtл †čⁱká ›- as in Eastern YaghnōϐХчлaὕtʦouʥʦлʦʣʲʣлⁱ can be influenced by palatalization of k before ). The Manichaean alphabet as the only Aramaic-derived alphabet could spell voiced stoops b, d, g. The original voiced stops changed in *Proto-Sogdic to fricatives, in Sogdian voiced stops appear from secondary development – either as a result of voice assimilation or from loan words. /p/ /b/ ъt/ /t/ p da ψpʰϊ [p] allophonous pronunciation in front of another consonant íptʧ [b] bⁱd n ψtʰϊ tíʲa₎члmēt, xⁱšíʤt [t] allophonous pronunciation in front of another consonant ēt₎ /d/ [d] dínda₎ ъčъ [mv] ča zчлčХna₎ ъǰъ [jv] ǰáʸa₎ ψ₎ʰϊ /k/ allophonous pronunciation in front of another consonant ϐū₎táʲ ψϑʰϊ allophonous pronunciation in vicinity of a front vowel allophonous pronunciation following a front vowel in front of another consonant tХ₎члʤХ₎члšaʲХ₎чл₎áʸʧ₎ [c] /g/ ₎ōy [k] g záʲa₎ ['] [] allophonous pronunciation in vicinity of a front vowel ψqʰϊ /q/ /m/ /n/ í₎ta ʥʧʲd haq [q] allophonous pronunciation in front of another consonant maqʳád [ʁ] voiced variety of q (not †ψ(]) qōqлψˈqʰo”ʁϊлdayʲ mēʸ₎ [m] [ɱ] čúmʤa₎ allophone of m in front of v, f [n] ψ ϊ n n Хʲán₎a allophone of n preceding a velar /f/ [f] ʤúšma /v/ [v] vⁱʲ t ъʳ/ [s] ʳaʲáyчлʳōʲtчлʳⁱn ya₎ /z/ [z] zōyчлz nk /s/ ъšъ ъžъ šХša [ʃv] [mv] in some loans from Tajik can be pronounced as č [ v] [jv] mainly in non-native words it can be pronounced ǰ ·101· ōčл<лōšчлčapaὕ qл<лšapaὕ q žūta aždaʦ ʲл [a(d) vdaˈʦo”ʲϊчл ʲáuǰnaлψˈʲau(d) vna] /x/ ψ ϊ ъ ъ [ʁ] ъʸ/ /h/ / / /w/ ay₎ ʸáʲa₎члʸatчлʸōʲ ψ {ϊ ψ ϊ often loses its labialized character when followed by ψСϊ ø ψ ] ʸōʲ ʦámmaчлnaʦáʸ [h] [[] /ᴥ/ ʸaʲ archaic pronunciation of merges with h of Arabic origin, in common speech it iss > hiss, su ϐátл şл ʳuʦϐát aʦmáq pronunciation of ayn in Modern YaghnōϐХлʧʳлpʲʣʳʣʲvʣdлonὕyлʧnлaлʷoʲdл šaᴥmá₎ ‘ᴥmák the ayn-sound usually disappears in pronunciation, it often prolongs ᴥúddaл şл údda;л qalᴥáл şл preceding vowel, following a consonant it may be realized as slight stop ψqʰȴὕщˈǦϊ in speech ʷáfir, w nak y [j] dēuлψˈdʣ”u] ʳaudХл ψʳa ˈdʧ”ϊчл ʳaʲ₎ūuл ψʳaʲˈ₎ʰu” ] yau r [r] čaʲʸчлʲ ta l [l] ъu/ ψu] ψ ] allophonous pronunciation of w following a vowel púὕуὕфa Table 38 YaghnōϐХлϑonʳonantлʳyʳtʣmлуNἴσÁἰ 2010, 222-223). Historically YaghnōϐХлϑonʳonant system differs only a little from the state reconstructed for Sogdian. The main differences can be seen in lack of labialization of velars in front of a labial vowels and different development of voicing153. In comparison to Sogdian it can be said that in YaghnōϐХл tʦʣʲʣл aʲʣл noл voiced stops (and affricate) in diachronic view, the voiced consonants emerged from positional allophones. From synchronic point of view YaghnōϐХл tʦʣʲʣл ʧʳл developed opposition of voiceless and voiced stops and affricates, the only exception is uvular stop q which has no voiced counterpart in voiced uvular stop †[(] – voiced counterpart of q is voiced uvular fricative . YaghnōϐХл ϑonʳonantл ʳyʳtʣmл ʧʳл tʦʣл ʳamʣл aʳл ϑonʳonantл ʳyʳtʣmл oʤл tʦʣл ZarafshānлTajik dialects – only voiced alveopalatals fricative ž appears more often in YaghnōϐХлʧnл contrast to ZarafshānлTajik154 (mainly dialects of Mastchōʦ;лcf. KHROMOV 1962, 27). YaghnōϐХл consonant system is also comparable to consonant system of literary Tajik, but the standard Tajik language lacks voiceless pharyngeal fricative which merged with voiceless glottal fricative h ( appeared together with voiced pharyngeal fricative ᴥ both in YaghnōϐХлandлκaʳtchōʦХлʧnл speech of older generations, nowadays is usually realised as h and ᴥ is either lost or it prolongs preceding vowel in YaghnōϐХфщл ζntʣʲʣʳtʧnʥл ʤʣatuʲʣл ʧʳл aл ϑommonл ϑʦanʥʣл oʤл Tajik (and colloquially Russian) ž to ǰ both in MastchōʦХлandлʧnлḲaghnōϐХчлtʦʧʳлʤʣatuʲʣлʧʳлoϐʳʣʲvaϐὕʣлaὕʳoлʧnл And also lack of ʦ in YaghnōϐХчлϐutлʧtлʧʳлaлquʣʳtʧonлʷʦʣtʦʣʲлʦ was a separate phoneme in Sogdian. Nowadays ʦ can appear in some Russian loans in YaghnōϐХчлϐutлʧtл ʧʳл uʳuaὕὕyл ʲʣaὕʧzʣdлaʳл s: Yagh. ›“v‘lús ₍‘ ʲʣvoὕutʧon л <лοuʳщл , g‘stín s‘ tavʣʲn л<лг , k‘nsé›t ϑonϑʣʲt л<л , s gán / ʦ gán δypʳy л<л г . 154 As ž appears rarely also in Persian, in the Zarafshānл dʧaὕʣϑtʳл ž appears only in words of Eastern-Iranian (*ZarafshānХфлoʲʧʥʧnщ 153 ·102· colloquial Tajik and in many Tajik dialects (RASTORGUEVA 1964, 44-45) or Uzbek: Pers. Tjk. m žá ʣyʣὕaʳʦ лşлḲagh. míǰ(ǰ)‘, TMast. m ǰá, Pers. žālá, Tjk. ž lá ʦaʧὕ лşлḲagh. ǰ , TMast. ǰ lá; in YaghnōϐХлaὕʳoл Tajik occasionally changes to č: Tjk. pʧὕaʤ лşлḲagh. č (KHROMOV 1987, 656), Uzb. ‘p‘ł ‹ ʳὕap лşлρ₍₎щл > Yagh. . Development of Iranian consonants in Sogdian and YaghnōϐХлϑanлϐʣлϑʦaʲaϑtʣʲʧzʣdлaʳлʤoὕὕoʷʳ: II.1.3.1. *p i. ii. iii. > Sogd. p155, Yagh. p: Sogd. ʾʾph ʾʾp(h) ʾʾp ʾp ā-p /āp/, Yagh. p ʷatʣʲ > ŋāp-, Ave. āp-; Sogd. pʾ /pā ъчлYagh. ʤoot < -(ka-), Ave. pā a-, OPers. pād‘-; Sogd. ʾ p-h ʾ⁾ p-ʾ(h), ⁾ p-ʾ ⁾ p-ʾ / ʸšəpáъчл Yagh. xⁱ áp nʧʥʦt < ŋ⁾ ap -, Ave. ⁾ ‘pā-; (voice assimilation) > Sogd. p, Yagh. b: Sogd. √p w s-, √p wfs- √p wfs- /√p uʤʳ-/, Yagh. b dú”s- to glue, to ʳtʧϑ₎ < *upa-dá”ʦa-; ⃝ (before ŋ ‘) > Sogd. b, Yagh. ?: Sogd. ’₍ʾmn⁽›₎ / / personal name у×л Sogd. ⃝ √p₍ʾm /√pyāmъ toлʦʣaὕ ) < *upa- уδκπлέюы6ф; II.1.3.2. *t i. ii. iii. ›ʾt ›ʾt, ʾ ›ʾt› ›ʾt ’›ʾṯ / ʲāt(ər)/, Yagh. vⁱ > Sogd. t156, Yagh. t: Sogd. ϐʲotʦʣʲ < -; Sogd. tys /tХʳ-/, Yagh. tДs- to ʣntʣʲ < *ati-íʦa-; Sogd krt /kaṙt/, Yagh. k ›t ₎nʧʤʣ < ŋká›t‘-, Ave. karəta-, Pers. kā›d; Sogd. √pt ⁽(ʾ)₍ √pt⁾⁽ʾ₍ √pṯ⁾⁽ʾ₍, √pṯ⁽⁾ʾ₍ /√p tʸ° /, Yagh. t ⁾ - to ₎ʧὕὕ < ŋp‘t -⁾u h‘ ‘-; (voice assimilation) > Sogd. t, Yagh. d: Sogd. √pt ⁽ √pt (ʾ)⁽ √pṯ ⁽ /√p /, Yagh. d - toлʦʣaʲ < *pati-gáu a-; (palatalized) > Sogd. č, Yagh. č (?): Sogd √pckwyr /√p č₎ü /, Yagh. čⁱk r- ‖л č ká ›- to ʤʣaʲ < *pati- u›‘ ‘- (?); Sogd √pc ʾnt √pcbnṯ /√p č áṁd/ to anʳʷʣʲ < *pati-’ánd‘-; Sogd. g √pc r - /√p č ʲə -/ to aϑϑʣpt < *pati-g›á’‘-; Sogd. cy rd xcy, ycy hji /xəčíчлɨčí/ ψуʳфʦʣъʧtϊлʧʳ < ŋásč < ŋást ; Yagh. -č ending of the 3 pers. sg. 157 pres. < *-ti- ; Later when following a vowel (mainly in younger Christian texts) > Sogd. b (~ ʾb th ʾʾph ʾʾp(h) ʾʾp ʾp ā-p ъāpъфл ʷatʣʲ л<л ʾʾbwx rwc ʲōčъл namʣлoʤлtʦʣльы dayлoʤлaлmontʦ лу×л ʾʾpw ʾ rwc ʾʾpwx уʲōčфъфл<лŋāp‘-u‘ uh-›‘uč‘-, Ave. āp -v‘ uhД уδκπлέюы5фщ 156 xwdy Later in post-vocalic position and after *r (mainly in younger Christian texts) > Sogd. d (~ ?): ъʸ{ə hu tte ъʸ{ət-íъфчлḲagh. ⁾‘t oʷnчлʳʣὕʤ л<лŋhuát‘-, Ave. x ‘t ⁾⁽d(ʾ)⁽лъʸud u ⁽tʾ⁽ ⁽tʾ(ʾ)⁽ ⁾⁽tʾ⁽ ⁾⁽t(ʾ)⁽л ъʸut uъфл ὕoʲd л <л ŋhu‘-t u‘n-, Pers. uđ ( ) mrdxmy ъmáṙ m›tʾ⁾mk mrt mʾk(⁽), mrt mʾ₍ mrtxmy(y) mrtxmy ъmáṙ < ŋm ›t ‘-táu⁾m‘n-ka-, Pers. m‘›dúm уδκπлέэ6Ś-270). 157 Different explanation of development of the ending -č : -č < -č t < *-t t < -t- t (cf. KLIMCHITSKIY 1940, 99100). 155 ·103· iv. v. vi. (rarely after a vowel) > Sogd. d ( ?), Yagh. d: Sogd. ʾy yd /ēdъчлYagh. Дd tʦʧʳ < 158 *á t‘-, Ave. ‘ēta-; Sogd ktʾ(ʾ)m ktʾm, k ʾm ‹dʾm /kə Yagh. ʷʦʧϑʦ < -, Ave. k‘tām‘-, Pers. уδκπлέэ6Śф; (in secondary contact with * < *d) > Sogd. ϑ, Yagh. ?: Sogd √k ʾr /√₎ toл doчл to work (stem of a transitive preterite) < * ktú- - < ŋk t‘m - уδκπлέэ7Śф; ŋt› (word-initially) > Sogd. čə (ʦə ?), Yagh. ?: Sogd tʦʧʲʳt < ŋt n‘-, Pers. t‘ ná; c n-y cn- /čə(š)níл ~ lə(š)níъ II.1.3.3. *k i. ii. iii. iv. v. vi. vii. viii. ix. > Sogd. k159, Yagh. k: Sogd. ktʾ₍, ktʾk qt, qty(y), ktyy qṯy /kətē/, Yagh. kat ʦouʳʣ < *káta-(ka-); Sogd ʾ wkʾ ʾ wk / uk / tʦʲonʣ , Yagh. tk nʣʳt < ϑ -kā-, Ave. gātu-, OPers. gāϑu-; Sogd. ky /ki/, Yagh. ki ʷʦʧϑʦ < *kah, Ave. k ; Sogd wyrk-y wyrq-y /wiw₎íъчлYagh. urk ʷoὕʤ < ŋu k‘-, Ave. vəhrka-, Ved. v k‘-; (in several cases) > Sogd. x, Yagh. x: Sogd √ › -, √ n - √⁾› -, √⁾n -, √⁾ - /√xəwš-, mʾx √ʸąš-/, Yagh. ⁾‘ - to puὕὕ < *k› -, Ave. k‘› -, Ved. ká› ati; Sogd. mʾ (w) / Yagh. m ⁾ ʷʣ < -, OPers. ‘hmā⁾‘m-; (before - ) > Sogd. g, Yagh. ?: Sogd. w ʾstgyʾẖ /wɨ ə / narrative < ⃝ * -k‘- - уδκπлέэя6щэф; (rarely) > Sogd. č, Yagh. ?: Sogd. crks /čáṙkəs/ vuὕtuʲʣ < ŋk k‘-ʦa-, Ave. kəh›kās‘-, Pers. k‘›gás; Sogd c›⁾⁽ ṯ /čáṙʸuštъ wine-press ~ cf. Pahl. k‘›⁾ уδκπлέэяŚф; (before a reduced labial vowel) > Sogd. ᵊ k , Yagh. k: Sogd ʾkwt-y kwt-y, qwt-y / ₎{ətíъчлḲagh. kut dog < ŋkút‘-, ŋkutД-; Oss. k ʒ ‖лkuy; *-a-ka- (denominal abstract suffix *-ka- in ending of masculine a-stems) > Sogd. - , Yagh. -a: Sogd. ʾspʾ₍t(ʾk), ʾsp(ʾ)₍tk, (ʾ)spʾ₍tk, (ʾ)spʾ₍t₍ spyṯy /ᶤ Yagh. sⁱ ʷʦʧtʣ < *ʦuá t‘-ka-; *-a-ka- (denominal abstract suffix *-ka- in ending of neuter and adverbial a-stems) > Sogd. - , Yagh. -a (?): Sogd. cʾn(ʾ)k⁽, cʾnʾ⁽ / aʳчл ʧʤ < ŋh‘č‘-ana-kam; cf. Yagh. č n < ŋh‘č‘-ana-(ka-)); ŋ- -kā- (denominal abstract suffix *-kā- in ending of ā-stems) > Sogd. -ā, Yagh. -a: Sogd. ›₍›ʾkh / Yagh. ʳaὕiva ; Sogd. ʾʾph / /, Yagh. ʷatʣʲ < -kā-, Wakh. yupk; *- -kā- (denominal abstract suffix *-kā- in ending of ā-stems) > Sogd. - k, Yagh. - k: Sogd (ʾ)z ʾ(ʾ)k(h) z ʾk zbʾq / Yagh. zⁱ tonʥuʣчлὕanʥuaʥʣ < hiʣu - -, 158 YaghnōϐХлʤoʲmлmayлϐʣлϐoʲʲoʷʣdлʤʲomлἵʣʲʳʧanщ √ ⁽žtgʾ ə Later in post-vocalic position (in younger Christian texts) > Sogd. g pl. fut.) nwgrwc ( s d‘₍) л<лŋn‘u‘-k‘-›‘uč‘-, Pers. ḥ‘u› члγāʲʳщлNo 159 ·104· nd ask (2 pers. уδκπлέэя6щюфщ x. xi. xii. Ave. h ₎ -, h ₎vā-, hizvah-, Ved. ; Sogd. syʾʾk(h) syʾk syʾq /sə Yagh. sⁱ ʳʦadoʷ < *aʦā - -(ka-), Ave. asaiia-, Pers. sā₍á; *- - - (denominal abstract suffix *- - in ending of -stems) > Sogd. -č (-ǰ ), Yagh. -č: Sogd ʾync(h), ynch g ʾync(h) ync ʾync /|ɨṁǰ/ (< * ṁč), Yagh. nč ʷomanчл ʷʧʤʣ < ŋ áun -kā-; Sogd ʾʾ›ʾ₍nc ʾʾ›ʾnj ʾrync / ɨṁǰ/ (< ṁč), Yagh. ›ínǰ ʣὕϐoʷ < * ϑni-ka-; Sogd nyc /nēčъ noʳtʲʧὕʳ < -kā-, Khʷāʲщлnʾc /nā a/, Ved. -; *-u-ka- (denominal abstract suffix *-kā- in ending of u-stems) > Sogd. -ku, Yagh.-k: Sogd yttkw ytkw ytqw /|ɨt₎úъл <л * túk, Yagh. ētkйДtk ϐʲʧdʥʣ < ŋhá t k‘- < ŋh‘ tu-ka-, Oss. xid xed; Sogd. ʾ₍ntk(ʾ)⁽ /íṁdku/ ζndʧanчлζndʧϑ < ŋ(h)índ k‘- < ŋhíndu-ka-, OPers. hiⁿduya-, Pahl. h nd g, Pers. (> Yagh. ) , Yagh. -k: *- -kā- (denominal abstract suffix *-kā- in ending of -stems) > Sogd. Sogd ʾ ⁽kʾ ʾ wk / tʦʲonʣ , Yagh. tk nʣʳt < ϑ -< ϑ -kā-, Pers. gāh pὕaϑʣ ; Sogd ₎ʾnʾ⁽k, ₎nʾ⁽kʾ znwq / Yagh. ₎ nk ₎nʣʣ < - < *ʣ -kā-, Pers. , Pahl. ₎ān g; Sogd ʾ₍ntʾ⁽k /íṁduk/ ζndʧanчлζndʧϑ л < - < ŋhínd -kā-; (ad iv.) There is a ʳʣϑondaʲy л paὕataὕʧzatʧonл oʤл vʣὕaʲʳл attʣʳtʣdл ʧnл ʳomʣл βaʳtʣʲnл Iranian ὕanʥuaʥʣʳчлmaʧnὕyлʧnлtʦʣлἵāmХʲлϐʲanϑʦчлandлaʳлaлʲʣϑʣntлʤʣatuʲʣлʧnлζʲonлdʧaὕʣϑtлoʤлἴʳʳʣtʧϑщлζtлʧʳл possible that the examples showing secondary palatalization *k > č in Sogdian show possible loans from a Middle IranʧanлἵāmХʲлу?) language. (ad vi.-xii.) A typical feature of the Iranian languages is extension of a nominal stem with denominal abstract suffix *-k - (or its variety *-čД- for feminines). By extending the stem with the denominal abstract suffix the original nomina got a new modified meaning, but most of words did not change their meaning significantly. In individual Iranian languages various reflexes of the suffix *-k - can be observed: in most languages the suffix is more or less maintained (of course, with regard to its further development in various languages). However in some of the Iranian languages it leads to its peculiar transformation – its consonantal part disappears and vowels emerge into new vocalic or diphthongal ending of a nominal stem (such development may be observed in Sogdian, YaghnōϐХ, MunjХ-Yidghā160, Pashtō-Wa etsХ, Saka dialects or in New Persian161). Development of denominal abstract suffixes in *Proto-Sogdic had to start before operation of the Stress II: suffix *-k - became part of the stem and position of Stress II was governed also by presence or absence of the denominal abstract suffix: Sogd. ʾpsʾkkh (ʾ)psʾk ʾpsʾ‹ / For development of the denominal abstract suffix *- - ʧnлκun₍Х-YidghāлʳʣʣлMORGENSTIERNE 1938, 114-115. σoϑaὕʧϑ лdʣvʣὕopmʣntлoʤлtʦʣл*a-ka-suffi (ʾ)stʾ›ʾk (ʾ)stʾ›₍, ʾstry /ᶤ Khōtщлstā›‘‘-, Munj. st ›əy, Yidgh. , Pasht. (f); Pers. s tā›á ×лKhʷāʲщл /(əфʳtāʲʣʥъчлζshk. st› k, Sangl. ustᵊ , Shugh. ⁾ t rʒчлBa₍ūщл⁾ t rʒ, ⁾ t ›ǰ, Khūʤщлοōsh. ⁾ tē›ʒ, ⁾ tu›ǰчлBaʲtщлοāshrv. ⁾ t ›ǰчлπaʲХqщл⁾ tu›ǰ, ⁾ tu›ǰ, Yazgh. ⁾(ə)taragчлŌʲmщлstarrak; Parth. ʾstʾ›g < ŋstā›‘- - ʳtaʲ щ 160 161 ·105· ʷʲʣatʦ, ϑʲoʷn < - < ŋpúʦā-kā- (: Yagh. †p × without the *-k - suffix may be supposed following development: ŋ†púʦā- (Stress II) : Yagh. †pus > †p sá (Stress III) : Sogd. †pəsá). The change of the denominal abstract suffix *- - in an innovated word-stem has two responses: 1) forms preserving *-k-, or 2) contracted forms, in which internal *-k- disappeared and subsequently underwent other sound changes. (1) The original consonant was retained in some feminine ā-stems and in forms of -stems. In case of feminine ā-stem, *-k- was retained when the suffix *-kā- followed a stressed syllable (that emerged from the Stress II shift): Sogd (ʾ)₎ ʾ(ʾ)k z ʾk ₎’ʾ‹ / Yagh. zⁱ tonʥuʣчлὕanʥuaʥʣ < *hĭz - < *hiʣuā- - [cf. Pers. < < *hiʣuā- -]; Sogd s₍ʾʾk(h) s₍ʾk s₍ʾ‹ /sə Yagh. sⁱ ʳʦadoʷ < ŋs‘ -(ka-) < *aʦ ā-kā- [Pers. sā₍á < Pahl. sā₍‘g < *aʦ ā-kā-]. Forms of denominal abstract suffix of the original -stems have a different outcome in Sogdian and in YaghnōϐХ: Sogdian forms retain thematic -u- (for ; for masculines *-u-ka- > -ku, -uk), in YaghnōϐХл *- - was feminines *- -kā- > syncopated: Sogd yttkw ytkw ytqw /|ɨt₎úъл <л * túk, Yagh. ētkйДtk ϐʲʧdʥʣ < ŋhá“t k‘- < ŋh‘ tu-ka- [Oss. xid xed]; Sogd. ʾ₍ntk(ʾ)⁽ /íṁdku/ ζndʧanчл ζndʧϑ < ŋ(h)índ k‘- < ŋhíndu-ka- у×л Sogd ʾ₍ntʾ⁽k /íṁduk/ < - < ŋhíndu-kā-) [OPers. hiⁿduya-, Pahl. h nd g, Pers. > Yagh. ]; ʾ›(ʾ)⁽k(ʾ), ʾ›ʾ⁽kh ʾ›⁽k(ʾ) dʾ›⁽‹ Yagh. d ›k ʷood л<л* -< -ka-; Sogd ʾ ⁽kʾ ʾ wk / tʦʲonʣ , Yagh. tk nʣʳt < * ϑ - < ϑ -kā- [Pers. gāh pὕaϑʣ ]; Sogd ₎ʾnʾ⁽k, ₎nʾ⁽kʾ znwq / Yagh. ₎ nk ₎nʣʣ < - < *ʣ -kā- [Pers. , Pahl. ₎ān g]. Similar ›ʾ₍k(ʾ) ryk / ʲ kуăф/, development can be observed also for other substantives: Sogd. Yagh. ⁱ ϑὕayчлʣaʲtʦ < * ›á“kă- < ŋg›á ‘-ka-; Sogd s w yk, sw (ʾ)₍k /sə { πoʥdʧan < *su - < *ʦug(u)d ‘- - [Pers. su ; cf. OPers. Sug(u)da- Sogdiana ]; Sogd pʾ›s₍k / ἵʣʲʳʧan < - < ŋpā›ʦ ‘- - [Pers. , < Pahl. pā›sДg]. Nicolas Sims-Williams interprets this development as a result of the Sogdian Rhythmic Law (i.e. Stress III) and presents two examples, which show different development as should be expected for the Rhythmic Law: Sogd. ntʾk(ʾ), ntʾkk ntʾ(ʾ)k(ʾ), ntʾkk ndʾk nṯʾq / áṁdā₎ / ϐad < *gand-āk‘- ʳtʧn₎ʧnʥ and Sogd ʾ⁽tʾk ʾ⁽tʾ(ʾ)k ʾwṯʾq / pὕaϑʣ < ŋ‘u‘-tāk-a- (SIMS-WILLIAMS 1981b, 13); both these examples can be systematically explained as a result of the Stress II, and subsequently the stress shifted to the Stress III: ŋg‘ndā-kā- > - > áṁdāk [Yagh. g‘ndá ϐad < Pers. g‘ndá ϐad < ʳtʧn₎ʧnʥ ; Parth. gndʾg; Ved. -> [> Turkic otaq (Uzb. t ‹) > Pers. ot ʲoom ]. gandʰá- ʳmʣὕὕ ]; ŋ‘u‘-tā-kā- > ao (2) Forms of a-stem masculines and ā-stem feminines with stress on antepenultima delete the original *-k- of the denominal abstract suffix, after the loss of *-k- there is a further development which has different responses in both languages: in Sogdian can be observed development *-a-kah (nominative singular) > *-‘ʼ > *- and *-ā-kāh (nom. sg.) > *- > *- ; in YaghnōϐХл tʦʣʲʣл ʧʳл tʦʣл ʳamʣл dʣvʣὕopmʣntл ʤoʲл ϐotʦл a- and ā-stems: *-a-kah (nom. sg.) > *-‘ʼ > -a and *-ā-kāh (nom. sg.) > *- > -a: ·106· Sogd. ʾspʾ₍t(ʾk), ʾsp(ʾ)₍tk, (ʾ)spʾ₍tk, (ʾ)spʾ₍t₍ spyṯy /ᶤ Yagh. sⁱ ʷʦʧtʣ < ŋspá“t‘ʼ < *ʦuá t‘-ka- [Munj. spД, Pasht. Wa . spДn, Pers. , , ]; Sogd ⁽tʾ›nk /ʸutáṙ Yagh. ⁾utánn‘ water-mʧὕὕ < ŋ⁾u‘tá›n‘ʼ < ŋhu‘t(a)-ᛑn‘-ka- [Yazgh. ⁾°‘₍“›g, Shugh. ⁾ d ›ǰ, οōsh. xad ›ǰ, Sangl. xu ri, Wakh. xəd ›g, Munj. ⁾Д›g‘, Yidgh. o]; Sogd. ʾ›ʾk ʾryh ’ʾ›₍ / ʲʧdʣʲ; ʲʧdʧnʥл anʧmaὕ , Yagh. ʲʧdʣʲ < < -ka- [Pahl. ’ʾ›g /ϐāʲaʥъчлἵʣʲʳщл’ā›á, Shugh. v ›ǰ, οōsh. v ›ǰ, Yazgh. v‘›āg, Ishk. vь› k ʦoʲʳʣ ; Oss. ’‘›æg ʲʧdʣʲ ]; Sogd. ›₍›ʾkh / Yagh. ʳaὕiva < ŋ›á“r ʼ < *r ( )r -kā- 162 [Pers. lē›, Pasht. ṛa]163; Sogd. ʾʾph / Yagh. ʷatʣʲ < ʼ < ŋāp -kā- [Khōt. tcā-, Ishk. vek, ā, Yidgh. yow o, Pasht. ’ə, Tjk. ’á ʷatʣʲ ; Oss. avg ʥὕaʳʳ ]. Wakh. yupk, Munj. According to examples of contracted (or aka- and ākā-) stems shown in the unit (2) mentioned above, it can be suggested that words derived from denominal abstract suffix *-k - retained its semantic value in subsequent stages of *Proto-Sogdic. If we did not consider the *- - suffix this way, we would not be able to convincingly explain the development of originally suffixed *-k( )- from the development of *k in all other cases – Iranian (and *Proto-Sogdian) *k is usually retained as k both in Sogdian and YaghnōϐХ (e.g. except Sogdian change *nk > ṁg etc.), . Different but *Proto-Sogdian denominal *-k- disappears between unstressed * development of this suffixed *-k- can be seen in other forms of the aka-stems, e.g. in neuter (and in adverbs): *-a-kam (nom. and acc. sg. neuter and acc. sg. masculine) > *-‘ʼu > Sogd. (cf. Sogd. cʾn(ʾ)k⁽, cʾnʾ⁽ / aʳчлʧʤ < < ŋh‘č‘-ana-kam; cf. Yagh. č n164 < ŋh‘č‘-ana-(ka-)). Some features of development of aka- and ākā-stems will be shown later in analysis of *Proto-Sogdic inflectional system. Apart from the denominal abstract suffix *-k - there was a similar suffix *-ka-, which was used to form diminutives – this suffix did not morphologically distinguish the original stem system and thus its development considerably differs from the denominal abstract suffix: responses of the diminutive suffix give both in YaghnōϐХлandлSogdian regular form in -(a)k. (ad ix.) This suffix belongs also to the denominal abstract suffixes in *- -, in this example can be seen its development with the i-stems. See also an analogical development in the Slavic . languages: Ide. *h “u -keh - > ŋou -kā- > PSl. ŋouĭc‘- > OCS. ь ʳʦʣʣp × Ved. 162 Precise etymology of this word is not known to me. Cf. etymologically unrelated Hebrew ›Д›, Aramaic ›Д›ā of the same meaning. 164 The root - - in č n emerged either from ŋč n < ŋčān уʧщʣщл ʲʣʥuὕaʲ лḲaghnōϐХлϑʦanʥʣл > in front of a nasal) or by labialization of ŋā after disappearance of *- < a-kam or *-u < *-am; but influence of Persian cannot be excluded č n < ŋč -g‘un‘-. 163 ·107· II.1.3.4. *č i. ii. iii. > Sogd. č 165 , Yagh. č: Sogd rwcn / ən/, Yagh. ʷʧndoʷ < ŋ›áuč‘-na // ŋ›áuč‘-ka-; sṯ›₍ ṯ /ᶤ ʷomʣn (pl. from (in front of *k, *t, *n) > Sogd. , Yagh. č, : Sogd /ᶤ < -ta- < *(ʜ) -kā-t -; Sogd ʾ₍ ktʾ₍h ₍ kṯyẖ /ɨš₎ə harem < ŋ áun -kā-kata-ka- уδκπлέэ5Śф; Yagh. čto burn (pres. stem : past part.); ŋč > Sogd. , Yagh. : Sogd. OPers. ‘v-,Ved. cyav-; √ ⁽- /√šəw-/, Yagh. ‘u- to ʥo < *č áu‘-, Ave. ( )‘uu-, II.1.3.5. *b i. > Sogd. , Yagh. v: Sogd. ϐʲotʦʣʲ < -; ›ʾt ›ʾt, ʾ ›ʾt› ›ʾt ’›ʾṯ / ʲātуər)/, Yagh. vⁱ II.1.3.6. *d i. ii. iii. iv. v. ʾ›⁽k(ʾ) dʾ›⁽‹ / уăф/, Yagh. > Sogd. (l?), Yagh. d: Sogd. ʾ›(ʾ)⁽k(ʾ), ʾ›ʾ⁽kh d ›k ʷood < -ka-, Ave. dā›u-; Sogd w h w / ō ъчл Yagh. v d ʳϑʣnt < ŋ’áud -, Ave. baodi-, Khʷāʲ. / ō ъ; (in secondary contact with ŋ ) > Sogd. č (but older ), Yagh. ?: Sogd. c t⁽ʾn, t⁽ʾn d₍ ṯ⁽ʾn /č ɨ ъл pooʲ < *du -tuu - уδκπлέэř6ф; *dr > Sogd. ž, Yagh. dⱽr ‖ dⱽr (word-initially): Sogd. ẓw /žō/, Yagh. d‘›áu ‖ dⁱ›áu ʦaʧʲ < ŋd›áu‘-; my r- /mɨžá/ m⁽ ʾkk *dr > Sogd. ž, Yagh. rd (word-internally): Sogd. my rh /mužē/, Yagh. mí›d‘ ϐʣadчлpʣaʲὕ < ŋmú ra-(ka-), Ved. -; ŋdu > Sogd. , Yagh. d(ⱽ)v: Sogd. r-y dbr-y / əríъчлYagh. d‘vá› ‖ dⁱvá› dooʲ < ŋduá›(a)-, Ave. duuar-; (excursion 5) Lambda Sogdica? In many Eastern Iranian languages there can be seen a development of Iranian voiced dental stop *d: it appears in some of the Eastern Iranian languages and dialects as ŋ l. The development *d > ŋ l is attested already in the Old Iranian period –in Scythian and Cimmerian, in the Middle Iranian Bactrian and in the New Iranian Pashtōчлτa etsХ, MunjХ and Yidghāлуand probably in SarghuὕāmХлandлʧnлʳomʣлwords ʧnлtʦʣлἵāmХʲлaʲʣaф. Several personal names are attested from Cimmerian, one of them was recorded as Tugdammē or Dugdammē in Assyrian, in Greek the same name was recorded either as Δύ or as ύ , (HERODOTUS I, 61). The name of the Cimmerian king 165 Later in post-vocalic position > Sogd. ǰ (often not reflected in spelling, the only example can be spelling in the hji /xəǰ cy xcy /xəčíъфл ψуʳфʦʣъʧtϊлʧʳ л<лŋ(⁾)ásč < ŋh‘u-ást . ·108· Dygdamis/Lygdami(o)s (reigned between the years 660 and 640 ) demonstrates that the change *d > ŋ l took place already in the first half of the 7th century at the end of th the 8 . Similar feature is documented also for the name of the Scythians: in Greek they are known as ύ (and from there Latin Scythae), which is derived from their own ethnic name *Skuda- < Ir. *skuda- aʲϑʦʣʲ (< Ide. *skud-o-, cf. Eng. shoot; Old English sɧēotan; Ger. schiessen to ʳʦoot ; ABAEV 1965, 25). Herodotus quoted that the Scythians called themselves 166 (i.e. *Skula- -) after a king called ύ (i.e. *Skula- A›ch“› ). If we compare the Greek (nom. sg.) ύ πϑytʦʧan andл πϑytʦʧan ύ Scyles , we can see the only difference ϑ ×л , it is the feature we observed already in the Cimmerian name Dygdamis/Lygdamis. The Histories of Herodotus were written in the second half of the 5th In this period the change *d > ŋ l was probably finished already – the Greek name for the Scythians ( ύϑ ) was probably of an older date167, the later names of the king Scyles and the Scythians-Scolotians ( ύλ and λ ) was recorded in innovated forms by Herodotus. If we compare once more the spellings of the Cimmerian name Tugd‘mmē : Dugd‘mmē : Δύ : ύ with the Scythian ethnic names ύ : we can see changing of lateral l with dentals (or less possibly alveolars). Dental pronunciation of Iranian *d can better explain a dichotomy in development of Ir. *d > ŋ / ŋ l in the Eastern Iranian languages. The development can be summarized as follows: (dental) stop > (dental) approximant > (dental~alveolar) lateral approximant ×л уdʣntal~alveolar) fricative, i.eщл хψdϊл şл хψ ] > *ψὕ>ὕϊл × *[d > ]168. Similar development can be assumed not only for dentals, but also for labials and velars: thus we can better explain a shift of *b towards labiodental fricative or labialized velar approximant and *g towards uvular fricative (i.e. *[b]лşлхψ ] > *[v~w]; *[g] > *[ ϊлşл лψʁ]).169 The change *d > l which is typical for some Eastern Iranian languages is nothing unusual when compared with other Indo-European languages. Apart from Iranian Pashtōчл τa etsХчл κun₍Хчл Ḳʧdghāчл πaʲghuὕāmХл уŠфчл Baϑtʲʧanл this change is attested as ʳuϐʳtʲatʣл ʧnл ʳomʣл ἵāmХʲл languages; in other Indo-βuʲopʣanлὕanʥuaʥʣʳлʳuϑʦлaʳлλūristānХ Prāsūn (Pār n/Vāsivari; e.g. ləz -< < IIr. *dáć‘- tʣn , l < IIr. ŋdu‘- tʷo )170; Indo-Aryan Romani (Gypsy; e.g. pʰ›‘l < -; Ved. ’ʰ›āt› brother ;лϑʤщлβnʥщлpal); change *d > l can be partly observed in Latin (in 166 «… ύ л л л ·л л л ύ л Ἕ лὠ » (HERODOTUS IV, 6). For a relatively older origin of the name ύ (and not † ύ лoʲл† ύ ) can testify also an absence of a plural ending in *- - typical for the Scythian language of the period of Herodotus. 168 Other explanation of the development *d > ŋ l offers Ivan MikhaЧὕovich Steblin-Kamenskiy: *d > *ḍ > *ḷ > *l 2 (STEBLIN-KAMENSKIY 1999, 22 ) – he supposes that the intermediate stage was a cerebral sound instead of an approximant. Such explanation does not make sense as cerebral sounds are peripheral in the Iranian languages and when they appeared it was always caused by a contact with *r and they never emerged randomly. 169 ζлʷouὕdлὕʧ₎ʣлtoлtʦan₎лtoлκʥʲщлηanлBʧčovʳ₎ýчлἵʦщαщлʤoʲлʦʧʳлʲʣmaʲ₎ʳлonлpʦonoὕoʥyщ 170 ἵʲoϐaϐὕyлduʣлtoлϑontaϑtлʷʧtʦлnʣʧʥʦϐouʲʧnʥлκun₍Хл(?). 167 ·109· such case in Sabine loan-words 171 (?); PULJU 2000) and in Greek, although it concerns a Mediterranean substrate words172, and in Hittite (e.g. tabarna ×лlabarna ₎ʧnʥ ;лʳʣʣлaὕʳoлnām‘n ×л lām‘n namʣ ). Among non-Indo-European languages a similar feature can be observed in Siouan (North American Indian) Lakota language (which differs from mutually relative Dakota by operation of the shift *d > l; e.g. Lakota ločhí ×лαa₎otaлdočhí ʦunʥʲy ). The problem of development *d > l in Sogdian can be difficult to assess. The Sogdian script used tʦʣл Aʲamaʧϑ л ὕʣttʣʲл lām‘ḏ for a continuant of Ir. *d; this grapheme was used to spell mainly dental fricatives and ϑ, but occasionally it was used also for l in words borrowed from Sanskrit (see excursion 4); although the Aramaic original of the Sogdian alphabet possessed also the letter dāl‘ṯ to spell d ~ , in Sogdian this letter was used only in an Aramaic ideogram Sogd. D ʾt /ət/ toчлtoʷaʲdʳчлʧn . In the Manichaean script the letter āl‘ṯ used to spell (and ϑ), āl‘ṯ was derived from a shape of the letter lām‘ḏ, which was normally used to spell l. Only the Syriac script used the letter dāl‘ṯ to spell . Moreover, the Old Uyghur alphabet used the πoʥdʧan лὕʣttʣʲлlām‘ḏ to spell (or d; in the modern Turkic languages with outcome as y, z or in Chŭvash r). 171 In Latin there are following words showing the (Sabine?) change *d > l: , Gre. αή , Armen. Lat. lēv ›, laevir; ProtoItal. ŋd‘ ⁽ē›, Ide. ŋd“ h -u“›-, Skt. d“vá›-, Pasht. lē⁽‘›, Yagh. s taygr, Lith. dí“v“› s, d “v“›ìs, Latv. diẽveris, CSL. děv“›ь, Rus. ь, Srb.-Cro. // đȅv“›, Balt.-Slav. ŋd‘ʾ u“›-, OHG. zeihhur, OEng. tāco›, ProtoGerm. t‘ ku›‘-(?), Lith. lá gon‘s; Lat. lingva, dingva; ProtoItal. ŋdnχ(u)⁽ā-, Osc. fangvam, fancua < *fənχuā < ŋdʰ-; Ide. ŋdngʰuh , OIrl. , Irl. teanga, Gael. teanga(dh), OWelsh. tauawt, MidWelsh. tavawt, tauaỽt, Welsh tafod, OCorn. tauot, MidCorn. taves, -, Ave. h ₎uuā-, Armen. lezow, Pruss. insuwis, Lith. tavas, tawes, MidBret. te(a)ut, Bret. teod < ŋtngʰ⁽āt-, Ved. l “žuvìs, OCS. ję₎₍k , Goth. tu go, OHG. Zunga, OIcel. tunga < ŋdʰngʰ-, TokhA. käntu, TokhB. kantwo < ŋtänk⁽o; Lat. lacrima, lacruma, dacrima, dacruma; ProtoItal. ŋd(›)(k)‘k›unā-, Ide. ŋd›k-h (“)k›u-, OIrl. , Welsh deigr, ( ), ῡ , Armen. ‘›t‘suk, Lith. ã ‘›‘, OGH. Hitt. ḫaḫru- < *s-h “k›u-, Ved. áś›u-, YAve. ‘s› члδʲʣщл zahar, TokhB. ‘k› n‘; Lat. larix -cis; Ide. *dr-u-; Lat. lautia, dautia, ProtoItal. *dawetio-, Ide. ŋdouʜ-ó-, OIrl. , Skt. dúv‘s- < *duʜ-es-; Lat. lДgo, dus < Ide. ŋu“d-; , Welsh nedd(en), Corn. nedhen, Bret. nez(enn) Lat. lēns -endis; ProtoItal. *dlind-?, Ide. ŋdk(o)n-i-d-; OIrl. < ŋsn dā-;лδʲʣщл члAὕϐanщл уδʦʣʥщфл < ŋkon-id-, Lith. glìnd‘, Latv. gnВd‘, Rus. г , Srb.-Cro. гњ // gnjȉd‘, Sloven. gníd‘ < Balt.-Slav. ŋgn ʾd‘ʾ < *knid- < Ide. ŋkn d-; OEng. hnitu, Eng. nit, OHG. (h)niz < ŋkn d-; Armen. anic; Lat. olor, odor :: ; ProtoItal. *ode/o-, ŋod s, Ide. *h (e)d-, Gre. ὄ : ὄ ώ члὀ , ὀ , Armen. hot < *h ed-, Lith. úost : úodž ‘, Latv. uóst, OCze. jadati (cf. PULJU 2000; WALDE 1906; DE VAAN 2008). 172 ζnл tʦʣл κʣdʧtʣʲʲanʣanл ʳuϐʳtʲatʣл ʧnл δʲʣʣ₎л ʧtл pʲoϐaϐὕyл ʷaʳл aл dʣntaὕл ʳoundл ʷʧtʦл ὕatʣʲaὕл aʲtʧϑuὕatʧonл хψd ]?, its presence shows the d-series of the Linear B script and diffʣʲʣntлoutϑomʣʳлoʤл хψd ] in Greek and other languages: Gre. Ὀ ύ ×лὈ ( ) ύ члὈ ( ) ύ чл ὐ ύ чл ὐ члὈ ύ члὨ ύ ; Etruscan Utʰu₎“, Utʰst“, Ut(ʰ)ustʰ“ ×л Lat. Ul₍ssēs, Ul ⁾ēs ςὕyʳʳʣʳ < *Minoanл ътἴt ulʣл ʾл тἴt ume/, see also Sumerian Utu-zi; further in Mycenaean da-pu -ri-to-joл уʥʣnщл ʳʥщфл ъd aϐuʲʧntʰo₍oъл ×л δʲʣщл ύ : ὕaϐyʲʧntʦ ;л δʲʣщл л ×л dʧʳϑ ; Gre. л×л ;лἱatщлlaurus ὕauʲʣὕ ; Lat. l › c‘ ×лκyϑʣnaʣanлto-ra-keлуnomщлpὕщфлъt ʰōʲā₎ʣʳъл/pl./; Gre. ώ ὕoʲʧϑaчл aʲmouʲ (cf. BAορἴλДἰ 2009, 39). ·110· By comparison with YaghnōϐХлa similar development in Sogdian might be expected: Sogd. *d > : Yagh. *d > (ŋ >) d. So why the issue of lambda Sogdica then? There are several Sogdian (or in common Eastern Iranian) loans in Persian, in which (and also ϑ) appears as l173: ‫ )الفخن( الفغدن‬: ‫ـ‬ ‫‘ الف ج‬l”‘ dán (‘l”‘⁾tán) : ‘l”‘nǰ- to acquire, gain, earn, collect, save < Sogd √ ʾ t(‫باد)ە‬ √ϑ”⁾ ṯ- ъ√ fəʸšt-/ to collect, gain (pret.) < *ϑuą⁾ -ta-; ‫‘’ بایە‬lā₍á contemptible, corrupted, depraved, perverted < Sogd. ʾp ʾt₍ (ʾ)p ʾt₍ p ʾt₍ pdʾṯy / ; (/ ‫ بالیك‬b ) leather shoes < Sogd ‫پالیك‬ pʾ ₍k related to foot < Sogd. pʾ (h) pʾ pʾd ъpā ъ foot, leg ; ‫ پل‬pil heel < Sogd p -y p ( )-y /pə íъ foot ; Yagh. pad ×лἵʣʲʳщлp‘ ; ‫پل دین‬ door-frame, lintel < Sogd treshold < *pati-‘ntā-, Shugh p Дnd; a jug with a wide mouth < Sogd ‫غولن‬ ⁽ ʾk(h) p ynd /pə ṁd/ ⁽₍ vessel, container, pitcher, (a dry) measure ; Ave. g‘oⁱ -ж g‘o ‘n‘-; ‫ـ‬ ‫ـ( ل ج‬ ‫ )لی ج‬l nǰ- to pull, to extract < Sogd √ ₍nc ъ√ iṁǰ/ to pull out < *ϑ‘nǰ‘ ‘- ×л inherited Pers. ha⁾tán (h‘nǰДdán) : h‘nǰ- of the same meaning; mw m ⁽ mwd, mdw mdʰu /mū > mə úъ; Oss. ‫ مل‬mul wine < Sogd m d ‖лmud ʦonʣy ; ‫ نال‬nāl reed(-pen) < Sogd. (?) /nā ъл×лἵʣʲʳщл reed flute ; Yidgh nəl, Wakh. nālč k tube, pipe (HENNING 1939); In addition to the above shown forms there are some other Eastern Iranian words in Persian that show the change *d > l, e.g. m‘lá⁾ ʥʲaʳʳʦoppʣʲчлὕoϑuʳt ; ’ líst span ; l Дdán to milk ; žālá ʦaʧὕ ; also in a place-names ђ lmánd in Afghanistan (cf. Ave. ђ‘ētum‘nt, Greek Ἐ ύ ф and probably Sar and Yaz in BadakhshānлуʳʣʣлϑʦaptʣʲлI.1.1.4.b., note 54). There is also double form with both l and d in the word Badakhshān ʧnлἵāmХʲлуρjk. ; cf. Balas(c)ian and Badas(c)ian mentioned by Marco Polo) and Ām (< *Ā )/Ā Āmūлαaʲyāчлἴʸuʳ (cf. QAοФB 1965, 63). In Persian there in attested a loan that shows đ pʲʣʳʣʲvatʧonл oʤл πoʥdʧan л without any change: Pers. ‫ )خدوك( خذوك‬xađ dʧʳappoʧntmʣnt, grief, anger < Sogd ⁽k ⁾ ⁽k /xə 173 anger (HENNING 1939, 93-94). In Persian l normally originates from OPers. *rd < Ir. *rd, *rʣ. However, in the Early Classical Persian there has been the sound (nowadays realized as d < < *t, *d; only in few words there is z < < *t, *d; e.g. Pers. gu ‘ tán : gu ‘›- şл γāʲʳщл go₎ä tän : go₎ä›-; Tjk. gu₎‘ tán : gu₎‘›-; AfghP. go₎‘ tán : go₎ar- to paʳʳ л ×л ἵʣʲʳщл u şл γāʲʳщл AfghP. od ; Tjk. δod фщлἶuʣʳtʧonлʧʳлʷʦyлἵʣʲʳʧanлϐoʲʲoʷʣdлπoʥdʧanл as l, when the same sound has already ϐʣʣnл pʲʣʳʣntл ʧnл ἵʣʲʳʧanщл ἵʣʲʳʧanл уoʲл ἵaʦὕavХфл l appears unchanged in Sogdian, as can be seen in an example of (Middle?) Persian p‘hl‘⁽ānДg < Parth. *parϑawprϑ⁽ʾ₍‹ ъpáṙ ə in the Sogdian script – p⁾lʾ⁽ʾnʾk, p lʾ⁽ʾnʾk p ›ʾ⁽ʾnʾk) /pə – to the Sogdians there probably was a difference in pronunciation of (Middle) Persian l in contrast to Sogdian . ·111· Given the above mentioned facts, the issue of the nature of Iranian *d in Sogdian is difficult to assess. To make it more difficult, I will show responses of Sogdian sw (ʾ)yk, s ⁽ ₍k /sə πoʥdʧanчлϐʣὕonʥʧnʥлtoлπoʥdʧana and Sogd s ⁽ ₍kstn /Sə ɨstan/ Sogdiana in the neighbouring languages: OPers. sᵃ-u-g -(u-)dᵃ, sᵃ-u-gᵃ-dᵃ /Sug(u)da-/ Sogdiana ; Pers. Su d, su d Sogdian(aф ; Pahl. swt /ʳū ъчлswptyk /ʳu Хʥъл Sogdʧan ; Ave. Su ‘-, Su⁾ ‘- Sogdiana ; Parth. swgd ~ ϐʣὕonʥʧnʥл toл Sogdʧana ; Sogdʧan ; Bactr. ( ) Tumshuq. suḏana-, pl. suḏ‘n‘nā Sogdian(s) ; Gre. ᾱ , (pl.) Sogdiana, Sogdʧanʳ ; Elam. ú-ug-da Sogdiana ; Akkad. Su-ug-du Sogdiana ; Syr. S ḏ, S ḏ ‹ā₍ē (pl.) Sogdian(s) ; Arab. a s - S uᷧd Sogdiana ; ρü. so d‘‹, so du‹, su d‘‹, so (u)d Sogdian(aф ; Armen. S d k πoʥdʧan ; Chin. 粟 特 Sùtè Sogdian(a) ; MidChin. *Siok-dək Sogdian(a) ; Tibet. སོག་ཌག། Sog-dag Sogdian(a) ; × Pahl. swl(y)k /s g/ sw l₍₍ /ʳu ὕХъ Sogdʧan ; Khōt. s lД, pl. s l₍‘ Sogdʧan ; Northwestern Prkt. suliǵa- Sogdʧan ; Chin. 窣 利 Sùlì Sogdiana ; MidChin. *Sɑ(k)-lis Sogdian(a) . As wa can see in the above shown examples (which I have divided into two groups), the name for the Sogdians and for Sogdiana differed variously in neighbouring languages – in some of them there is development *d > and in the other there is *d > l. Interesting is mainly the Bactrian form ( ) (LIVSHITS 2008, 324) – in Bactrian should be excepted a form † λ ( ) . Had the Bactrian form ʤoundлonлanлʧnʳϑʲʧptʧonлʤʲomлἶaὕƒa-yi Afrāsiyāϐ reflected local Sogdian pronunciation? Or was the attested form contaminated by Greek ᾱ ? Bactrian certainly needed to have its own name for the neighbouring countries that was probably inherited from Old Iranian, so why the attested form looks non-Bactrian? According to the above shown examples, there is a majority of forms with attested , not with l, and because of YaghnōϐХлуandлхḳaʲaʤshānХфлd ʧtлϑanлϐʣлaʳʳumʣdлtʦatлpʲonunϑʧatʧonлъ ъл was more common (or standard?) in Sogdʧanчлaὕʳoлtʦʣл ϐoʲʲoʷʣd лBaϑtʲʧanлʷoʲdл ( ) shows development *d > in Sogdian. It is possible that the l-forms attested in Persian may have been borrowed via Bactrian (or) with Bactrian-like pronunciation. How can be lambda Sogdic‘ explained? 1) It is possible that Sogdian loans in Persian with l instead of * mayл ϐʣл ʧntʣʲpʲʣtʣdл aʳл ʳϑʲʧϐaὕл (or copyist) ʣʲʲoʲʳ чл ʧщe. that these words were recorded according to the written form, not according to the spoken language174. 2) In Sogdian 174 See e.g. realisation of Sogdian as f in many Persian words (HENNING 1939) – Sogdian was spelled as ‫ ڤ‬in the Perso-Arabic script, but due to its resemblance with ‫ ف‬this grapheme has been replaced by the letter f : ‫فژ( ف‬, ‫‘” ڤژ‬ž (y)z-y, ʾ (y)z-y (y)j-y, ʾ j-y / žíл<л ʣžíъл ʣvʧὕ л<лŋ’éž < ŋ’á₎d ‘-). Then ʳoʲdʧdnʣʳʳчлʧmpuʲʧtyчлfiὕtʦ л<лSogd. letter ‫ ڤ‬was used also in Classical Persian to write ƀ ъ ъчлtʦʧʳлʳoundлʦaʳлϐʣʣnлὕoʳtлʧnлὕatʣʲлʳtaʥʣʳлoʤлtʦʣлὕanʥuaʥʣлandл changed to b, e.g. ‫زڤان‬ > ‫زبان‬ ὕanʥuaʥʣ . ·112· there were several dialects, from which a majority (delta-dialects) underwent a development *d > , but some dialects (lambda-dialects) changed *d (and perhaps also *ϑ) > l 175 – those lambda-dialects were probably in contact with Persian – this can explain the dichotomy of forms with l not only in Persian but also in Chinese Suli у×лSute), ʧnлἵaʦὕavХл , su lД у×лs , su Дg) and in other languages (cf. QAοФB 1965, 62-64). There is, however, one problem – whether a postulation of the lambda-dialect is not just a purposeful attempt to solve this issue. There is also another explanation: 3) in Sogdian there was retained pronunciation of *d as a dental approximant *[ ], which appeared as (*[ ] ~) *[ ] ~ *[d ] to speakers of some other languages, but as *[d] ~ *[d] or even as *[ὕ] ~ *[l] to speakers of other neighbouring languages. The adoption of the Sogdian dental approximant ŋ in various languages differed according to how it was perceived by non-Sogdian speakers who borrowed Sogdian lexemes. Indeed preservation of ŋ ϑanлʣʸpὕaʧnлtʦʣл pʲʣʳʣʲvatʧon лoʤл pronunciation of *d as such in YaghnōϐХ. Similar example can be found in Danish pronunciation of soft d, i.e. dental approximant (or alveolar voiced sonorant; see HABERLAND 1994, 320) as in mad ψˈma ] food , dydig ψˈdy” i] virtuous , or huset ψˈʦu”Zsə ] the house . «Its auditive impression is quite close to [l] and it is often confused with it by non-native learners of Danish.» (ibid.) Finally a theme for reflection – do we really know what kind of sound has been spelled by the Aramaic letter lām‘ḏ in the period when the Sogdians adopted the Aramaic alphabet for their language176? In the presented work I will not deal with this problem, I will leave it to the Semitic scholars … II.1.3.7. *g i. ii. > Sogd. , Yagh. : Sogd. √ ʾʾr /√ āʲ/ to ʥuaʲd , Yagh. ›- to ὕoo₎ < ŋgā›-, Ave. gā›- to be awake, to pʲotʣϑt ; Sogd. / āu/, Yagh. u ϑoʷ < ŋg u‘-, Ave. gāu ; (before a labial vowel) > Sogd. , Yagh. : Sogd ⁽›ʾṯy ъ { < w ›ʾt₍₍ /w gh. aʷa₎ʣn < *u a-(ka-); II.1.3.8. *ǰ i. > Sogd. ž, Yagh. ž: Sogd. √ʾ₎⁽(-) ž - to livʣ < ŋǰáu‘-, Ave. ǰ(a)uua-; √(ʾ)₎⁽(-) √j⁽(-) √ž⁽(-) /√ žūчл žau-/, Yagh. II.1.3.9. *f i. > Sogd. f, Yagh. f: Sogd. w r-y wfr-y /wəfrí/, Yagh. ⁽á”ⁱr ʳnoʷ < ŋuá”›a-, Ave. vafra-; Sogd rʾʾk /ʤʲā₎ъчлYagh. fⁱ tomoʲʲoʷ < a-, Ave. fr nk-, ”›āk‘-, Ved. p›áṅk-, p›āk-; Sogd rʾn frʾʾn /ʤʲānъ ϐʲʣatʦ , Yagh. fⁱ ʳmʣὕὕ < a-; Skt. p›ān‘- ϐʲʣatʦ ; 175 It is possible that in those lambda-dialects, if we accept its existence, there has been an opposition of voiced and voiceless l. 176 AndлaὕʳoлonлtʦʣлtʧmʣлʷʦʣnлκānХлϑʲʣatʣdлtʦʣлκanʧϑʦaʣanлʳϑʲʧpt. ·113· ii. iii. iv. v. vi. vii. *fn > Sogd. n, Yagh. vn, mn: Sogd. w n-y xw n-y /ʸu níъчл Yagh. xuvn/xumn dʲʣam < ŋhuá”n‘-, Ave. x afna-; ŋ”n > Sogd. m, Yagh. m: Sogd. ⁾ ʾm /ʸšāmъ ʣvʣnʧnʥ , Yagh. xⁱ diner < ŋ⁾ ”n ‘- ʣvʣnʧnʥ , Ave. ⁾ ā”n ‘-, Parth. ʾm уδκπлέюьюф; *ft > Sogd. d, Yagh. ft (vd?): Sogd. ʾ ₍ t(-y) ʾ⁾ ʾ₍ t ⁾ ₍ t h a ⁽dʰ , h a wṯi / ʸšХ dуáфл~ ʸšɨ dí/, Yagh. xⁱ í”t mʧὕ₎ < ŋ⁾ uí”t‘-; Sogd. a wta /a d, ə dáъчлYagh. aft (avd) ʳʣvʣn < ŋhá”tą, Ave. *hapta, Oss. avd, Pers. haft; *fra- (before *s, ŋ , *t, *r, *n and probably before ŋ ) > Sogd. f(ⱽ)-, Yagh. fⱽ-: Sogd. √(ʾ) ʾm, √ʾp ʾm √ʾ ʾm, √ʾp ʾm √” ʾm /√ /, Yagh. fⁱ - to ʳʣnd < *fra-; Sogd. ʾ tm-y, (ʾ)p›tm-y ʾftm-y fṯm(ʾ) / ftəmíъ fiʲʳt < *fra-tám‘-; Yagh. f ‖л f t, fⁱ dayл aʤtʣʲл tomoʲʲoʷ < ŋ”›‘tā-má ϑā-, ŋ”›‘-t‘m‘-má ϑ - уδκπлέюь5-322); √ r- √ϑbr- /√ ər-/, Yagh. t‘”á›- ‖ *fra-b- > Sogd. ϑ -, Yagh. tⱽf-: Sogd. tⁱ”á›- toлʥivʣ < *f( ) á›ă- < *fra-’ᛑ-; *fru- (before ŋ ) > Sogd. f(ⱽ)-, Yagh. ?: Sogd ʾ -ʾh / ʤšáъ flʣa < ŋ”›ú ā-, Pasht. wrəž‘ уδκπлέюэюф; II.1.3.10. *ϑ i. ii. iii. iv. v. vi. vii. m(ʾ)y my > Sogd. ϑ, Yagh. s ‖ t (< Early Modern Yaghn ’Д ϑ 177 ): Sogd. my my(y) myϑ, myd /mē /, Yagh. mēs ‖ mēt day < ŋmá ϑa-; Sogd pʾ ( ) pʾ pʾϑ /pā /, Yagh. p s ‖ p t aʲʲoʷчлϐuὕὕʣt < ϑa-; ʾ ⁽kʾ ʾ wk / uk / tʦʲone , Yagh. tk (before *k) > Sogd. ϑ, Yagh. t: Sogd nʣst < ϑ -kā-, Ave. gātu-, OPers. gāϑu-, Pers. ; (before *n) > Sogd. ø, Yagh. ø: Sogd. pnʾnc /pə ṁǰ/, Yagh. pⁱ co-wife < *hapaϑnД-, Oss. ’Дn[o₍n‘g], Pahl. ʾbwg ; Sogd ʾʾ›ʾ₍nc ʾʾ›ʾnj ʾrync / ɨṁǰ/ (< ṁč), Yagh. ›ínǰ ʣὕϐoʷ < * ϑni-ka-; (before *ʦ) > Sogd. t, Yagh. t: Sogd m›tsʾ› mcʾ, msʾ /máуṙ ṙ)/, Yagh. mástar ʦʣʲʣ < ŋĭmá›ϑă-sā›- < ŋ máϑra-ʦā›-; Sogd ʾ⁽(›)tsʾ› ʾ⁽›tsʾ› ʾ⁽tsʾ› ʾ⁽cʾ, ʾ⁽sʾ / ṙ ṙ)/, Yagh. ⁽ástar tʦʣʲʣ < ŋăuá›ϑă-sā›- < ŋ‘uáϑra-ʦā›- уδκπлέюыьщьф; (before ŋ ) > Sogd. t (> *č), Yagh. ?: Sogd t rwc (₍) č₍(₍) / ətš~ eščí/ namʣлoʤл th the 15 dayлoʤлaлmontʦ < * át < *dáϑu a-, Ave. daϑu уδκπлέюыьщ2); to transmit (impt. 2. os. pl.) (after * < *d) > Sogd. t, Yagh. ?: Sogd √pt⁽₍ ṯ /√p < ŋp‘t -uá d‘ ‘-ϑ - уδκπлέюыэф; (occasionally before *i) > Sogd. , Yagh. ?: Sogd ›ʾn / ṁšъ tʧʣ < *granϑi- (GMS έюыэщvʧф; Around the year 1913 still ϑ (JUNKER 1930, 126, 128-129), the dental aspirate ϑ is attested in YaghnōϐХлϑʣʲtaʧnὕyлʧnл the year 1877, but in this period there are double forms with a sibilant s (DE UJFALVY DE MEZŐ KÖσβπα 1882, 276; TOMASCHEK 1880, 735; cf. JUNKER 1930, 4-5). In this work continuants of *ϑ will be marked s ‖лt. 177 ·114· viii. ix. x. xi. xii. xiii. *ϑr (word-initially) > Sogd. , Yagh. sⱽr ‖ tⱽr (< Early Modern Yaghn ’Д ϑⱽr): Sogd ʾ ry g ʾ ryw (ʾ) ry ʾ ry(y) ₍ / šaʧ/, Yagh. s‘›á₍ ‖ tⁱ›á₍ tʦʲʣʣ < *ϑr ‘-, Ave. ϑ›ā ; Pers. se < sih; ⃝ *ϑr (word-internally) > Sogd. , Yagh. l(l) (?): py ›ʾk, ⃝p r ⃝py ›ʾk, ⃝p r, ⃝ p ₍ ⃝p ₍ (as a part of compounds) /pɨš gh. púl(l)‘ уŠфл ϐoyчл ϑʦʧὕd; small, little л< ŋpúϑra-, Pers. p sá›; √p yr √(ʾ)p yr /√ p to hasten < *ϑu > Sogd. ϑ (ϑf), Yagh. ? Sogd *upa-ϑu ›‘ ‘- уδκπлέэŚюф; *ϑu (after ŋč) > Sogd. tf, Yagh. t(ⱽ)f: Sogd ct ʾr ct”ʾ› cṯ”ʾ›, ṯ”ʾ› /č ṙ/, Yagh. t , t ‖ tⁱ , t ʤouʲ < ŋč‘ϑu -, Ave. č‘ϑ - уδκπлέэŚ5ф *ϑu (word-initially) > Sogd. tf (t , ϑ ), Yagh. ? Sogd. √ ʾyz √ j √ϑ”₍ž, √ṯ”₍ž, √ṯ’₍ž /√tʤēžъ to ϑoὕὕʣϑt < *ϑu ǰ‘ ‘- уδκπлέэŚ6ф; nd *ϑu (occasionally) > Sogd. f, Yagh. ?: Sogd =f(y) /=f(i)/ encl. pron. of the 2 pers. pl. < *=ϑuā, Ave. -ϑ ā уδκπлέлэŚ7ф; (ad ix.) Development of Iranian *-ϑr- > l(l) (instead of expected †›s ‖놛t, cf. KHROMOV 1972, 127) in YaghnōϐХлʧʳлʲatʦʣʲлpʲoϐὕʣmatʧϑл– there are not many attested continuants of *ϑr. This development is for the first time mentioned by Wilhelm Geiger: «ϑr is preserved word-initially in t ›ā (tⁱ›-) th›““ = Ave. ϑ›ā₍ . Word- nt“›n‘ll₍ t s l n āl fi›“ ч Av“. āϑr-ж pul‘ sonж ch ld ч Av“. puϑra-.» (GEIGER 1898-1901, 336). Aὕ ϐʣʲtлἱʣonʧdovʧch Khromov sees such development as less acceptable, he notes, that YaghnōϐХл l fiʲʣ лʧʳлattʣʳtʣdлonὕyлʧnлvʣʲϐл l⁾á - to ὕʧʥʦtлup лandлtʦatл ʧnлaὕὕлotʦʣʲлϑaʳʣʳл fiʲʣ лʧʳлϑaὕὕʣdлϐyлTajik loan l u (Tjk. , lá⁽, ‘lá⁽, Pers. )178. YaghnōϐХл l(⁾á -) ϑanл ϐʣл ϑonnʣϑtʣdл ʷʧtʦл ἰāϐuὕХл l ₎‘d‘n to ʣmʧtл ʦʣat щл Khromov also assumes that YaghnōϐХлpúl(l)a may not be connected with Iranian *puϑra- as in YaghnōϐХлʧtлʧʳлuʳʣdлmaʧnὕyлʧnл tʦʣлʳʣnʳʣл уyounʥфлϑʦʧὕdчлyounʥлϐoy ʲatʦʣʲлtʦanл ϐoy лandлtʦʣлʷoʲdлϑanлϐʣлta₎ʣnлʤʲomлϑʦʧὕd ʳл speech (KHROMOV 1972, 127). The development of *-ϑr- > l(l) in YaghnōϐХ can be confirmed in other example: Yagh. k‘t(t) lá, k‘tt‘lá уьфл ʳtonʣл ʳʦʣὕtʣʲл madʣл ʷʧtʦл noл ʷood;л уэфл ʲuʧnуʳф л <л *kata-āϑra-ka- house-fire л уRASTORGUEVA – ÈDEL MAN 2000, 321) and TMast. щл ἰatōὕaʳл are used by herdsmen in mountains far from their villages – this term is connected with seminomadic life of the YaghnōϐХʳл ʳoл ʧtл ϑanл ϐʣл aʳʳumʣdл tʦatл tʦʣл ʷoʲdл ϑanл ϐʣл oʤл oὕdл oʲʧʥʧn. It is certainly not a borrowing as I have not found similar word in various Tajik, Uzbek and Kyrgyz dictionaries. MastchōʦХл Tajik has, similar to YaghnōϐХчл k‘tól ʤoʲл aл ʦʣʲdʣʲʳ л ʳʦʣὕtʣʲл – the Ghalch‘ (i.e. Mountain Tajik(s)) of Mastchōʦл ʳʦaʲʣл aл ʳʧmʧὕaʲл paʳtoʲaὕл ʳtyὕʣл oʤл ὕʧʤʣчл ʳoл MastchōʦХлk‘tól may be *ZarafshānХлʳuϐʳtʲatʣлʷoʲdлʧnлtʦʧʳлTajik dʧaὕʣϑtщлρʦʣлʷoʲdлʤoʲл fiʲʣ л- l (cf. Sogd. ʾ(ʾ) ъāš/) quoted by Geiger thus can be considered archaic, nowadays replaced by the Tajik word l u. Cf. also development *rt, *rϑ > ( ) *[ɬ] in Avestan (MACKENZIE 1988, 90). 178 I have neither heard l ʤoʲл fiʲʣ лduʲʧnʥлmyлʳtayʳлʷʧtʦлtʦʣлḲaghnōϐХʳщ ·115· II.1.3.11. *x i. ii. iii. iv. v. vi. > Sogd. x, Yagh. x: Sogd. g r-y xr-y /xəríъчлYagh. xar aʳʳчлdon₎ʣy < ŋ⁾ᛑ-, Ave. xara-; Sogd. /ʸāʸъчлYagh. ⁾ k ʳpʲʧnʥ < ha-; (non-etymological intrusive x before ŋ C) > Sogd. x, Yagh. ø: Sogd √np⁾ t- /√nəpɨʸšt-/, Yagh. nⁱpí t‘ to write (past part.) < ŋn pí t‘-(ka-), OPers. n p t‘-; Sogd. ʾ⁾⁽ t›-y ʾ ⁽ t›-y ⁾⁽ t›-y / x{əštʲíъ ϑamʣὕ < ŋu⁾ t›í < ŋu t›‘-, Ave. u t›‘-, Pers. utú› уδκπлέэ56ф; (before a labial vowel) > Sogd. ⁾ , Yagh. x: Sogd (ʾ) w xw(w) xw / xō/, Yagh. ax, he, that < *ah‘u-; *xt > Sogd. d, Yagh. xt ( d ?): Sogd y (ʾ)rt-, yr t y rṯ- /yə dí/, Yagh. yaxt (ya d) √sw t- /√su d-/, Yagh. sú⁾t‘ to burn (past part.) < *ʦú⁾t‘-, ʷʧdʣ < ŋu -g t‘-; Sogd. Ave. -suxta-; ŋ⁾ > Sogd. ⁾ , Yagh. x(ⱽ) : Sogd. ʾ p-h ʾ⁾ p-ʾ(h), ⁾ p-ʾ ⁾ p-ʾ / ʸšəpáъ, Yagh. xⁱ áp nʧʥʦt < ŋ⁾ ápā-, Ave. ⁾ ‘pā-; ŋ⁾ u > Sogd. ⁾ , Yagh. x(ⱽ) : Sogd. ʾ ₍ t(-y) ʾ⁾ ʾ₍ t ⁾ ₍ t h a ⁽dʰ , h a wṯi / ʸšХ dуáфл ~ ʸšɨ dí/, Yagh. xⁱ í”t mʧὕ₎ < ŋ⁾ uí”t‘-, Ave. ⁾ uu pt‘-; Sogd. /уuфʸušuъл у<л *x ‘ u), Yagh. u⁾ ʳʧʸ < ŋ⁾ uá ‘m; ʾ›⁽⁾ (ad ii.) Intrusive x before ŋ is attested also in Avestan: Ave. u›uu ⁾ n‘/əʲüɨʸšъл ϐandaʥʣ члḲagh. , ʷaʸ-ʣndчлtʦʲʣad луδκπлέэ56ф, YAve. ₎ ⁾ n həmn kennen lernen wollend чл ἴAvʣщл ”›ā⁾ nənəm ςntʣʲʷʣʧʳunʥ , YAve. ⁾ uu‘ ʳʧʸ чл ḲAvʣщл ⁾ tāt ψуʳфϊʦʣл ʳtandʳ чл ḲAvʣщл ”›‘⁾ tāⁱt“ er soll hervortreten чл ḲAvʣщл ‘ ‘uu‘⁾ nu den Frommen zufrieden stellend члAvʣщл⁾ n tó zufrieden gestellt лandлʧnлἴὕdлἵʣʲʳʧanл⁾ nāssāt ₍ er wird kennen л (BARTHOLOMAE 1895-ьŚыьчл ю6л έř6фщл πʧmʧὕaʲл ʤʣatuʲʣл ϑanл ϐʣл ʤoundл aὕʳoл ʧnл Baϑtʲʧanл чл ṇ yxt- to write (past part.) л oʲл ʧnл κun₍Хл чл чл чл чл nəwuxt- to write (past stem) (SIMS-WILLIAMS 1988 [online], 348). II.1.3.12. *x , *hu i. ii. > Sogd. ⁾°, Yagh. ⁾: Sogd. xwʾr /ʸ°āʲъчл Yagh. ⁾ › ʳʧʳtʣʲ < *hu‘h‘›-, Ave. x ‘ h‘›; √ wr- ъ√ʸ°ər-/, Yagh. ⁾‘›- to ʣat < *x ᛑ-, Ave. x ara-; Sogd. xwty, wty ʙ Sogd. ⁽t(ʾ)(ʾ)₍ xwty xwṯy, xwdy hu tte /ʸ°ətíъчл Yagh. ⁾‘t oʷnчл ʳʣὕʤ < *huát‘-, Ave. x ‘t , Pers. x ‘đ > ⁾uđ; > Sogd. x, Yagh. x: Sogd. g yp xyp ( ) xỿpϑ h“-p / ə) /, Yagh. ⁾ēp (‖ xap) oʷnчлʳʣὕʤ < *huá p‘ϑ ‘-, Ave. x ‘ēp‘ϑiia-; Sogd wtʾrnk /ʸutáṙ Yagh. ⁾utánn‘ w n-y xw n-y /ʸu níъчлYagh. xuvn/xumn ʷatʣʲ-mʧὕὕ < *hu‘t(a)-ᛑn‘-ka-; Sogd. dʲʣam < *huá”n‘-, Ave. x afna-; (ad i.) See analogical development in the Brythonic branch of the Celtic languages: Ide. ŋsu > Brythonic ŋhu > ŋ⁾u; cf. Mid. and Mod. Welsh chwaer; Mid. Bret. hoer, hoar; Mod. Bret. ·116· cʼhoar // w hoé›; Old Cornish huir; Modern Revived Cornish (Kernewek Kemmyn) hwoer < Ide. ŋsu“s ›, sister; OIrl. ıu ; Manx shuyr; Ir. ŋhu‘h‘›-; Pers. x āhá›; Ved. svás‘›-; II.1.3.13. * i. ii. iii. iv. ⁽ / ōšъчлYagh. ʣaʲ < ŋgáu ‘-, Ave. g‘o ‘-; Sogd. > Sogd. , Yagh. : Sogd. ʾ ₍ t(-y) ʾ⁾ ʾ₍ t ⁾ ₍ t h a ⁽dʰ , h a wṯi / ʸšХ dуáфл~ ʸšɨ dí/, Yagh. xⁱ í”t mʧὕ₎ < ŋ⁾ uí”t‘-, Ave. ⁾ uu pt‘-; (occasionally after ŋč(‘) in front of a nasal) > Sogd. ø, Yagh. ?: Sogd. cm-y cm-y(y) c(y)m-y /čɨmíъ у×л Sogd. c m-y /čɨšmíъф ʣyʣ < ŋčá m‘n-, Khʷāʲ. cm-, cn- /čəníл~ ləníъ cm- /camma/, Khōt. ts“ʼ m‘-, tsaima-, Ishk. com, Ōʲm. c mД; Sogd. у×лSogd c n-y /čəšníл~ ləšníъф tʦʧʲʳt < ŋt n‘-, Pers. t‘ ná, Ōʲm. trunuk уδκπлέюř5386); ŋ t(i) > Sogd. č (simplification of ProtoSogd. č), Yagh. č: Sogd. prch /paṙč/ ʳpʧnʣ , Yagh. pá›č‘ ʲʧmчл ʣdʥʣ < ŋpá› ta-(ka-), Ave. p‘› t -; Sogd. frʾʾ⁽₍ c₍ frʾwycyẖ ‖ fⁱ /f oϐὕivʧouʳnʣʳʳ < ŋ”›āmú t - уδκπлέюřэф; / wɨуšфčʧъчлYagh. ŋ t› (occasionally) > Sogd. č, Yagh. ?: Sogd. ʾzrʾ⁽ c / ₎›⁽ c-y / Zruščíъ Zarathushtra < *ʣaraϑú t›‘- < IIr. * arat-ʜu tra-, Ave Zaraϑu t›‘-, Parth. ₎›h⁽ t, Pers. Z‘›dú t; The development of Ide. *s > ŋ under the operation of the RUKI-rule is recorded not only in the Indo-Iranian languages, it is known also in Slavic (in Slavic later ŋ > * > ~ x) and partially in Baltic and Armenian (cf. BEEKES 2011, 137; MEIER-BοÜδδβο 2003, 102-105; MARTIROSYAN 2008, 536-538). II.1.3.14. *ž i. > Sogd. ž, Yagh. ž: Sogd √ y živ- to sew, to stitch < ŋží’‘-; √₎₍ -, √ y - √j - /√žɨ -/ to bite, to ϑʦʣʷ , Yagh. II.1.3.15. *m i. ii. iii. iv. > Sogd. m, Yagh. m: Sogd. √m₍n /√mēn/ to be ʳʧmʧὕaʲ , Yagh. m (n)ta ‖ má( )nta ʳʧmʧὕaʲὕyчлуὕʧ₎ʣфлaʳ < ŋm n‘ ‘- to ϐʣлʳʧmʧὕaʲ ; (occasionally) > Sogd. m, Yagh. b: Sogd. my ʾn m₍d(ʾ)n / Yagh. bⁱ mʧddὕʣ -, Ave. maⁱ ān‘-; < ŋm‘d (following ŋā in front of a vowel) > Sogd. ā⁽, Yagh. m: Sogd. frʾʾ⁽₍ c₍ frʾwycyẖ / ɨуšфčʧъчлYagh. ‖ fⁱ /f oϐὕivʧouʳnʣʳʳ < ŋ”›āmú t -; Sogd ŋm›ʾ⁽ /mʲāu/ ʷʣʣpʧnʥ < -; *mp, *mb > Sogd. ṁb, Yagh. mp: Sogd. < ŋ kámba-(ka-) (Khromov 1987, ); (ʾ) kʾnp / š₎áṁb/ ʷoʲὕd , Yagh. ⁱkámp‘ ϐʣὕὕy ·117· (ad ii.) cf. Gre. л<л л moʲtaὕ члἄ л ʧmmoʲtaὕ л<лζdʣщлŋ(n)m›to-s; (ad ii.-iii.) cf. opposite development *w or * /*b > m ʧnлḳâzâ₎îлandлἰuʲdʧʳʦśлḳâzâщл ₎ım‘n, ₎ı⁽‘n, zun, Kurd. ₎ m‘n < ŋh ₎’ān- ὕanʥuaʥʣ ; II.1.3.16. *n i. ii. iii. iv. v. vi. vii. viii. ix. > Sogd. n, Yagh. n: Sogd. /nāʤъ ʦumanл ₎ʧnd , Yagh. n ” navʣὕ < -, Ave. nā”‘- navʣὕ ; (in some cases before ŋč, < *g, *k, *m, *ʦ, (< ŋč_t, *-i- -_t), *ϑ, *x) > Sogd. ø ~ n, Yagh. n ~ ø: Sogd ʾ₍ ktʾ₍h ₍ kṯyẖ /ɨš₎ə harem < * ṁč-k‘t‘ʼ < ŋ áun -kā-kata-ka-; x r /ʸą əʲъл у×л Sogd xn r /ʸáṁ ər/) ʳʷoʲd < ŋ⁾áng‘›‘-, Sogd kϑ, qϑ Sogd /₎ą ълу×лSogd. kn (h) kn ( ) knϑ, qnϑ /kaṁ /) ϑʧtyчлtoʷn , Yagh. Ŭánsi Kansi kænt (name of a village in Yaghn ’) чл Ḳagh. [P‘nǰ ]kát ἵan₍a₎ʣnt < ŋkánϑ ϐuʧὕdʧnʥ члKhōtщлk‘nthā-, ka(ṃ)tha- toʷn уδκπлέлююя-341); (non-etymological intrusive *n before *s) > Sogd. n ~ ø, Yagh. ø (?): Sogd ʾns ʾ /áṁs aълу×лSogd ʾs , s ( ʾ), s ( h) /ąʳ , s уá)/), Yagh. (i) ‖л t ψyouϊлaʲʣл(copula of the 2nd pers. pl. pres.) < *sϑá-; *nt, *nd > Sogd. ṁd, Yagh. nt: Sogd. ntm nṯm / áṁdəm/, Yagh. ámtun (< ántum) ʷʦʣat < ŋgántum‘-, Ave. gantuma-; *nt, *nd (occasionally) > Sogd. ṁd, Yagh. nd 179 : Sogd. nt(ʾ)k ntʾkh dnṯʾ / ɨṁd у₎фчл aṁd (k)/, Yagh. dínd‘k tootʦчлtʣʣtʦ < *dántu-ka-; *nk, *ng > Sogd. ṁg, Yagh. nk: Sogd snk(ʾ) sng /sáṁg Yagh. sánk(‘) ʳtonʣ < ʾnkʾyr /áṁgir/, Yagh. ínk › *aʦáng‘-(ka-), Ave. asənga-, OPers. aϑanga-; Sogd fiʲʣpὕaϑʣ < ŋhám-g‘› ‘-; ŋnč, ŋnǰ > Sogd. ṁǰ, Yagh. nč: Sogd. pnc pnc, pnž, pnj⃝ /paṁǰ/, Yagh. p‘nč fivʣ < ŋpánč‘-, Ave. p‘nč‘-, Pers. p‘nǰ; *n + *-ik( )- > Sogd. ṁǰ, Yagh. nč: Sogd ʾync(h) ync ʾync /ɨṁǰ/ (< * ṁč) ʷoman , Yagh. nč ʷʧʤʣ < ŋ áun -kā-; *n + *-ik( )- (rarely) > Sogd. ṁǰ, Yagh. nǰ180: Sogd ʾʾrʾync ʾʾ›ʾnj ʾrync / ɨṁǰ/ (< * ṁč), Yagh. ›ínǰ ʣὕϐoʷ < * ›áϑni-ka-, Pers. ā›ánǰ; II.1.3.17. *r i. > Sogd. r, Yagh. r: Sogd. ϑ(a)-; ›ʾ (h) ›ʾ ( )(h) ›ʾϑ /ʲā /, Yagh. › s ‖ › t patʦчлʲoad < In YaghnōϐХлnd is attested just in one inherited word: dínd‘k tootʦ чл tʦʣл ʤoʲmлϑanлϐʣл ϑontamʧnʣdл ϐyл ἵʣʲʳʧanл of the same meaning (KHROMOV 1987, 659). 180 In YaghnōϐХл nǰ is attested only in one inherited word: ›ínǰ ʣὕϐoʷ л ϐutл ʧtʳл ʤoʲmл ϑanл ʦavʣл ϐʣʣnл ʧnfluenced by Persian ā›ánǰ. 179 ·118· ii. iii. iv. (non-etymological intrusive *r before *n, after < *b or after a long vowel) > Sogd. r (ʳ) ~ ø, Yagh. ø (?): Sogd. wrn-w, wrn-y (y)xwrn-y xwrn-y /(yə)ʸ{əwníчлʸ{əwnúълу× Sogd. y wn-w, w rn-h yxwn-y ywxn-y /yəʸ{əníчлyəʸ{ənúчлyoʸníчлʷəxəwná/, Yagh. ⁽á⁾ⁱn) ϐὕood < *uáhu(r)na-, Ave. vohunД-; Sogd rywr / ər/ tʣnл tʦouʳand < ŋ’á u‘›-, Ave. ’‘ēuu‘›-; Oss. ’ ›æ ‖ be(u)›æ manyчлmuϑʦ луδκπлέ359-362); (in several cases before ŋž, *n, ŋ , *ʦ or after ŋā) > Sogd. ø˞, Yagh. ø: Sogd kj ‹ž /ka˞žълу×лSogd. krz, kr krj ‹›ž /kaṙž/) mʧʲaϑὕʣ < *ká›ǰ‘-; Sogd. pʾ / ˞ълу×л Sogd. pʾ› / ṙ/, Yagh. par) ʤoʲчл ϐʣϑauʳʣл oʤ < *pār-; Sogd =sʾ /= ˞ъл у×л Sogd. =sʾ› /= ṙ/), Yagh. =sa уtoʷaʲdʳфлto < *ʦā›- уδκπлέ354-358); *rn > Sogd. ṙn, Yagh. n(n): Sogd. prn /paṙn/ ʤʣatʦʣʲ , Yagh. pan(n) blade of a wheel of a watter-mill < ŋpá›n‘- ʤʣatʦʣʲ ; Sogd. krn /kaṙn/, Yagh. kan(n) dʣaʤ < ŋká›n‘-, Ave. karəna-; II.1.3.18. *l (?) i. √rys /√ʲēʳ ~ √ὕēʳ/, Yagh. lēs- to ὕʧϑ₎ < > Sogd. l (?) / r (?), Yagh. l (?): Sogd. *›‘ ʣ- (ŋl‘ ʣ-), Ave. ʲaēz-, Pers. lēsДdán : lēs-; Sogd √wyrʾrz √wlrz, √⁽d›₎ /√ʷʧὕáṙz/, Yagh. larz- (< Pers.?) toл tʲʣmϐὕʣ < ŋ(u -)rarʣ- (ŋ(u -)larʣ-), Khōt. ››Д₍s-, Pers. l‘›₎Дdán : l‘›₎-; II.1.3.19. *s i. > Sogd. s, Yagh. s: Sogd ʾst ʾsty sṯy /(əфʳtíчлáʳtуɨ)/, Yagh. ást( ) ψуʳфʦʣъʧtϊлʧʳ л< ŋást , OAve. ‘stД, OPers. astiy, Ved. ást , Ide. *h ést ; ii. *sp (often stem-initially) > Sogd. p, Yagh. p (?): Sogd. √ʾn p› /√áṁšpər/ to ʷaὕ₎ < *hám-spar- уδκπлέю7ыф; *sk (often stem-initially) > Sogd. k, Yagh. ⱽk (?): Sogd (ʾ) kʾ⁽› (ʾ) k⁽› ‹⁽›ϑ / š₎ōṙ / dʧffiϑuὕt < ŋskáuϑra-, OPers. k‘uϑi-; Sogd √(ʾ) kʾ₍› √ ‹₍› /√ to be drivʣn , Yagh. ⁱ - to puʳʦ < *sk ›‘ ‘- (δκπлέлю66-367) ŋsč (outcome of simplification of a clusteru) > Sogd. č, Yagh. č (?): Sogd. cy xcy, ycy hji /xəčíчлɨčíъ ψуʳфʦʣъʧtϊлʧʳ < ŋ(⁾)ásč < ŋást уδκπлέю7эф; ŋsč (in forms of preposition ŋp‘sč‘-) > Sogd. , Yagh. ?: Sogd. p₍ - p₍ -, ʾp₍ p -(ʾ), p ₍₍ p -(ʾ), p ₍ /pɨšуə), pɨšчлpɨší/ aʤtʣʲчлὕatʣʲ < *pásča-, Ave. p‘sča- уδκπлέю7юф; iii. iv. v. II.1.3.20. *h i. Yagh. ax ʦʣ < áh‘u; (in front of , ŋu, ŋ u) > Sogd. x, Yagh. x: Sogd. /(əфʸúчлуə Sogd wrn-w, wrn-y, y wn-w, w rn-h (y)xwrn-y, yxwn-y xwrn-y, ywxn-y /(yəфʸ{əуwфníчлʸ{əwnúчлyəʸ{ənúчлyoʸníчлʷəxəwnáъч Yagh. ⁽á⁾ⁱn ϐὕood < ŋuáhu(r)na-, Ave. vohunД-, vohuna-; Sogd. w(y)r, xwyr xwrлъʸü later ъʸōʲъф, Yagh. xʉr ʳun < ŋhuá› ‘- уδκπлέюřŚ-396); ·119· ii. iii. iv. v. vi. (following a long vowel) > Sogd. x, Yagh. x ~ k (?): Sogd. ʾ h /ʸāʸъчл Yagh. ⁾ k ʳpʲʧnʥ < -; Sogd. mʾ (h) mʾ⁾ /māʸъ moon < h-, Pers. māh, Ved. - уδκπлέюŚя-396); (word-initially, mainly before , ŋ ) > Sogd. ø, Yagh. ø: Sogd (ʾ)z ʾ(ʾ)k z ʾk zbʾq / z ā₎ъчл Yagh. zⁱ ὕanʥuaʥʣ < hiʣu - -, Ave. h ₎ -, h ₎vā-, hizvah-, Ved. - уδκπлέюŚ7ф; (often word-internally) > Sogd. ø, Yagh. ø: Sogd. ⁾⁽ʾ› /ʸ°āʲъчл Yagh. ⁾ › ʳʧʳtʣʲ < ŋhuáh‘›-, Ave. x ‘ h‘›; Sogd. √n₍ √n₍d : √n₍st √n₍sṯ /√nХ лśл√nХʳtъчлḲagh. nДd- to ʳʧt л<лхnihida- уδκπлέюŚř-401); (occasionally when palatalized) > Sogd. , Yagh. : Sogd. ₍ /ɨš/, Yagh. t181 ψtʦouϊлaʲt < *áh , OAve. ‘hД, Ved. ás уδκπлέ405); (in some forms of the verb *ah- to ϐʣ ) > Sogd. x, Yagh. x: Sogd. xnt nt xnd xnṯ /xaṁd/ ψtʦʣyϊл aʲʣ < ŋhánt , OAve. həṇtД, OPers. haⁿtiy, Ved. sánt уδκπл έ77ыф;л Sogd. cy xcy ʙr hji /xəčíъчлYagh. ⁾ást( ) ψуʳфʦʣъʧtϊлʧʳ < ŋást ; Sogd. ⁾ʾ₍ ʾy, rd ⁾ʾ₍ /ʸāʧ/, Yagh. ⁾ ₍ ψуʳфʦʣϊл ʷaʳ (copula 3 pers. sg. impf.) < ŋ ă < < (GMS έ77ы-771); Iranian *h originates from Ide. *s, except when it is followed by another obstruent. Similar development *s > *h can be seen also in Greek, Armenian, Celtic, Phrygian, Lycian or Albanian, and marginally in Vedic. In Greek Ide. *s changed to *h (but remained when adjacent to a stop or word-finally), later on it waʳлʳuϐ₍ʣϑtлoʤлδʲaʳʳmann ʳлἱaʷлʷoʲd-initially or disappears wordinternally. In Celtic original *s following a vowel was lenited to *h when no obstruent followed, in Brythonic there has been the change *s > *h also word-initially 182 , later word-internal *h disappears. In Armenian the development was the same as in Brythonic Celtic; in Albanian *s changes to h between vowels. In Vedic Ide. word-final *s changes to v s‘›g‘ ( ) before a pause (cf. BEEKES 2011, 137; MEIER-BοÜδδβο 2003, 102-105; KÜκκβἱ 2010, 12; MARTIROSYAN 2008, 536). II.1.3.21. *z i. ii. ʾzm-y /ɨzmí/, Yagh. ⁱm fiʲʣʷood < > Sogd. z, Yagh. z: Sogd. Sogd zmy / ŋá ₎ma-(ka-), Ave. ‘ēsm‘-, Khʷāʲ. ʾzm, Pers. hē₎úm, Ved. idʰmá-; (prothesis before *m-) > Sogd. z, Yagh. ø (?): Sogd ₎mʾ⁽›c, ₎mʾ⁽›ʾk / 183 Yagh. ant < ŋ(₎)máu› -ka-(ka-), (₎)máu›‘-ka-, Ave. maoiri-, Tjk. m ›čák, Pers. m ›čá уδκπлέюřыф184; YaghnōϐХл t < ŋ чt < *ʜáh Ō чt (encl. pron. 2 pers. sg.) (GAUTHIOT – BENVENISTE 1929, 52). 182 See OIrl. ; Irl. sean; Gael. sean(n); Manx shenn ×лτʣὕʳʦлhen (hên); Bret. hen; Cornish hen old < Ide. *seno-s; cf. Ved. sán‘ ; Lat. senex; Goth. sineigs; Lith. sẽnas; Latv. sęns ×лδʲʣщл ; Armen. hin. 183 ϑanлoʲʧʥʧnatʣл ʣʧtʦʣʲлʤʲomлρā₍Х₎л m ›čák у×л ἵʣʲʳщл m ›čáфчлoʲлtʦʣл ρā₍Х₎лʤoʲmлoʲʧʥʧnatʣʳлʤʲomлaл YaghnōϐХл Sogdo-YaghnōϐХлdialect. 181 nd ·120· iii. iv. *zd > Sogd. zd, Yagh. zd ‖ st: Sogd. pzt- /pəzd-áъчл Yagh. pazd ‖ p‘( )st ʳmo₎ʣ < ŋpá₎d‘( ‘)-, Ave. pazdaiia-; *zd (palatalized) > Sogd. ž, Yagh. ž (?): Sogd. (y)z-y, ʾ (y)z-y (y)j-y, ʾ j-y / žíл<л ʣžíъл ϐad < ŋ’éž < ŋ’á₎d ‘- уδκπлέю7Śф; II.1.3.22. *ʦ i. ii. iii. iv. v. vi. Yagh. sánk(‘) ʳtonʣ < *aʦáng‘-(ka-), > Sogd. s, Yagh. s: Sogd snk(ʾ) sng /ʳáṁg Ave. asənga-, OPers. aϑanga-; Sogd. rwps / əs/, Yagh. ʤoʸ < ŋ›áup ʦa-, Pers. , Ved. lopāśá-; Sogd. sr-y /səríъчл Yagh. sar ʦʣad < *ʦᛑ-, Pers. sar, Oss. sæ› уδκπлέю6яф; (palatalized) > Sogd. , Yagh. : Sogd. √pnʾ₍ /√p /, Yagh. pⁱn - ‖лpⁱná - to ὕoʳʣ < Ir. *apa- ʦ‘ ‘- (LIVSHITS – KHROMOV 1981, 388 ; δκπлέю7яф; *ʦtr > Sogd. , Yagh. : Sogd ⁽₍ (h) ъʷēšъчл Yagh. w ‖л ⁽‘ ʥʲaʳʳ < ŋu ʦt› ‘-, Ave. vāst› ‘-; ⁽k ⁽‹ /šō₎ъчлYagh. k ʳʧὕʣnt < *a-ʦ›áuk‘- (GMS *ʦr > Sogd. , Yagh. : Sogd. έю7ьф; ʾ⁽ ⁽ /šāu/, Yagh. u ϐὕaϑ₎ < *ʦ āu‘-, Ave. *ʦ > Sogd. , Yagh. : Sogd. s āuu‘-, Pers. (GMπлέьŚяф; *ʦu > *ʦϕ > Sogd. sp, Yagh. sp: Sogd. ʾsp-y /əʳpíъчл (ʾ)sp-y /( фʳpíл ъ, Yagh. asp ʦoʲʳʣ < ŋáʦu‘-, Ave. aspa-; Sogd. ʾspʾ₍t(ʾk), ʾsp(ʾ)₍tk, (ʾ)spʾ₍tk, (ʾ)spʾ₍t₍ spyṯy /ᶤ /, Yagh. sⁱ ʷʦʧtʣ < *ʦuá t‘-ka- уδκπлέю6яф; (ad vi.) Development *ʦu > *sp is common in majority of Eastern Iranian languages, an exception is the South Western ( Pers ‘n ) branch, WakhХл and Saka dialects. In the λūristānХл andл αaʲdʧϑл ὕanʥauʥʣʳл tʦʣʲʣл ʧʳ IIr. ŋću > λūʲщъαaʲdщ ŋ p, in the Indo-Aryan languages there is expected development IIr. ŋću > Ved. śv. The development of Ir. *ʦu (IIr. ŋću, Ide. ŋku) can be demonstrated in an example of Ide. *h ékuo-s ʦoʲʳʣ : IIr. *ʜáću‘-s; Ir. *(ʜ)áʦu‘-h; Ave. aspa-, Sogd. əspí, Khʷāʲ. ʾsb/ʾsp /asp/, Bactr. α π /asp/, Yagh. asp, Oss. yæ”s ‖лæ”sæ, Munj. ₍ sp, Yidgh. yasp, Pasht. ās ((m.) // ásp‘ (f.); WazХrХ dial.: ⁽ s // ⁽ spa; AfrХdХ dial. ⁽ s // ⁽ sp‘ < *Proto-Paṭʦān * spă- // ŋáspā-), Wa . ās, Ōʲm. ₍āsp, ἵaʲāch. ȫsp; Med. *aspa-, Balōch. (h)aps, (h)asp, Kurd. (h)esp ×лOPers. asa- (but Pers. asb/asp and Pahl. asp is probably of Median or Parthian origin 185 ); Wakh. ₍‘ , Khōt. ‘śś‘-186; πʣʣлaὕʳoлδʲʣщл ύ ×л ύ myʲʲʦ ;лδʲʣл ×лπ₎tщлmarakata- ʳmaʲaʥd щ Similarly in other New South Western Iranian languages: BakhtʧyāʲХчлπamghānХчлαavānХлaspчлἱāʲʣʳtānХл(ʾ)‘sp etc. ζnлtʦʧʳлϑaʳʣлtʦʣyлaʲʣлὕoanʳлἵʣʲʳʧanлὕoanʳлуγāʲʳщлäsb). 186 Development of *ʦu > уτa₎ʦХфлълśś уἰʦōtanʣʳʣфлʧʳлʳuʲʣὕyлnotлaʲϑʦaʧϑлpʲʣʳʣʲvatʧonлoʤлpaὕataὕлϑʦaʲaϑtʣʲлoʤлζIr. ŋću, but development */lълşлъšлʾлśълʧʳлϑauʳʣdлϐyлʲoundʧnʥлaʳʳʧmʧὕatʧonлϐyлtʦʣлϐʧὕaϐʧaὕлʤʲʧϑatʧvʣл*/ϕ/, i.e. IIr. ŋću > Ir. *ʦu 184 185 ·121· (Indo-Aryan responses) Ved. áśv‘- , ἵāὕʧ assa, BeṅʥāὕХ ‘ś’‘; cf. ā u[ anni] horse trainer in Mitanni Indic; (ḥ ristānД responses) Kati (BashgalХ): ú p‘ (ἰāmvʧʲʧ) / v‘ úp (ἰātəviri); (Dardic responses) i ā ä po, Kalā a hã ; (other Ide. responses) Gre. луAeolic фчлἱat. equus (m.) // equa (f.) > Romanian ‘pă, Spanish yegua mare ; Celtic *epos ~ *ekuos, OIrl. ęɧ, Irl. Gael. each; OBret. eb; Goth. ‘íƕa-, OEng. ēoh, OIcel. jó›, Tokh. yuk yakwe, Armenщлēš; Lith. ‘ v /“ v mare . II.1.3.23. *ʣ i. ii. iii. iv. v. vi. ₎ʾ₍ ₎ʾ₍(₍) /zāʧ/ ʣaʲtʦ , Yagh. ₎ ₍ fiʣὕd < *ʣ ‘-; > Sogd. z, Yagh. z: Sogd. (palatalized) > Sogd. ž, Yagh. ž: Sogd. √p›₍ž /√pə /, Yagh. pⁱr ž- ‖ pⁱ›á ž- to ʣʳϑapʣ < *upa- ʣ‘ ‘- уδκπлέэыьф; (dissimilated) > Sogd. , Yagh. d: Sogd. g st-y dsṯ-y / əʳtíъчлYagh. dast ʦand < ŋdást‘- < *ʣást‘-, Ave. zasta-, Ved. hást‘-; (before < *g) > Sogd. ž, Yagh. ž (?): Sogd √ʾ⁽₎ √ʾ⁽j ( ) / ə ъ to dʧʳmount < *‘u‘-₎gád-, Ave. zgad- уδκπлέю76щэф; *ʣr > Sogd. ž, Yagh. ž (?): Sogd y n(h) / ən/ ʦaʧὕ < *ʣ -, Ved. h›ādúnД-, cf. 187 Pers. žālá ; *ʣu > *ʣ > Sogd. z 188, Yagh. zⱽv: Sogd. (ʾ)z ʾ(ʾ)k z ʾk zbʾq / z ā₎ъчл Yagh. zⁱ tonʥuʣчлὕanʥuaʥʣ < hiʣu - -, Ave. h ₎ -, h ₎vā-, hizvah-, Ved. (GMS έю77ф; (ad vi.) Development of *ʣu 189 is rather complicated in comparison with the above mentioned development of *ʦu (II.1.3.22.vi.). There are no many examples, the best one */lϕ ~ sϕ/ > ἰʦōtщл śś [ʆу”ф] / Wakh. . Similar development of rounding can be observed e.g. in Avestan: YAve. ⃝ d›‘” ‘- ϐannʣʲ л ×л σʣdщл d›‘psá-; OAve. n‘” u ʥʲand-child (loc. pl.) л <л *nafsu- < IIr. ŋnápt-su- (Reiner LIPP, pers. comm.). 187 Most likely a borrowing from some Eastern Iranian language which changed *d to l, but there was no i-Umlaut of the root vowel; probably a Bactrian loan, see Yidgh. žДlo ʦaʧὕ щ 188 ž ʾq ъž ā₎ъл tonʥuʣ луδκπлέю7řфщ Dialectally also *ʣu > ž 189 It was claimed by Khromov and Livshits that there was also a development *ʣu > ž in Sogdian YaghnōϐХśлπoʥdщл √jʾ₍ √žʾ₍ /√žāyъл to discuss, to taὕ₎ члḲagh. ž ₍- to read, to sing, to ὕʣaʲn л<л*ʣu ‘-, Ave. zbaiia-, Skt. hvayati (KHROMOV – LIVSHITS 1981, 412; KHROMOV 1987, 567). On contrary, Ilya Gershevitch claims Sogd. /√žāyъ can be connected with Pahl. d›ā₍- уδκπл έ 285) < Ir. *dr - / ŋd›‘u- (the same explanation also in RASTORGUEVA – ÈDEL MAN 2003, 464). YaghnōϐХлž cannot come from Ir. *dr so if this root comes from *dr - we would expect Yagh. †d‘› y- ‖ †dⁱr y-. Both etymologies are wrong – there are comparable examples for another source of Sogd. and Yagh. ž in this case, cf. Wakh. ǰoy- or Munj. ₍-. Ivan MikhaЧὕovʧch Steblin-Kamenskiy connects this verb with Ave. gāϑa- ʳonʥчл δāthā л andл σʣdщл gā₍‘t ʦʣл ʳʧnʥʳ л ~ (I)Ir. ŋǰā - < Ide. ŋgē - (STEBLIN-KAMENSKIY ьŚŚŚчл эыыфщл ἴtʦʣʲл ϑompaʲaϐὕʣ л examples with different etymology are Shugh.-οōsh. ⁾ ₍-члπaʲХqщл⁾u₍- to ʳpʣa₎ лandлPasht. ⁾ ⁽əl to ʳʦoʷ лaʲʣлʤʲomл Ir. ŋs›āu‘ ‘- (ibid.). ·122· is an Iranian word for tongue, language, but unfortunatelly its responses are attested from two stems: *hiʣu - and *hiʣ -. *(hi)ʣu -(k -) > Sogd. / z ā₎л~ ž āk/, Yagh. , Oss. æv₎‘g, Ave. h ₎uuā-190, Khʷāʲщл z ʾk, ʾz ʾk ъzu g, əz /, Bactr. л/əz āʥъ, Munj. zəv zəv g, Yidgh. zᵊvД , zɪ’ē , Shugh.-οōsh. ziv, Yazg. zəveg, Ishk. ₎(ь)v k, Sangl. zəv k, Pasht. žəbaчл τazХʲХл žəbba191, Wa . z(i)bə, zə’ ; Ved. ; *hiʣu -n - > OPers. h ₎ānm (acc. sg.), Pahl. ʾ⁽₎⁽ʾn ₎⁽ʾn ъuzʷānчл ʧzʷānъчл ἵʣʲʳщл zab , Parth. ₎’ʾn ъʧzϐānъчлκʣdщлŋh ₎’ān-; *(hi)ʣ -(k -) > Ave. h ₎ -, Wakh. z k, OPers. h ₎ -192; Ved. juh -; Unexplained is Khōtщ ’ śā / ʧźā/, Vera Sergeevna Rastorgueva and Dzhoy Iosifovich Èdʣὕ man claim it can result from methatesis of *ʣuā-n- ?? (RASTORGUEVA – ÈDEL MAN 2007, 405); II.1.3.24. * i. ii. yw-y /yəʷíъчлYagh. ₍‘u ϐaʲὕʣy < ŋ áu‘, Ave. yauua-; Sogd. > Sogd. y, Yagh. y: Sogd. s₍ʾʾk(h) s₍ʾk s₍ʾ‹ /sə Yagh. sⁱ ʳʦadoʷ < *aʦā - -(ka-), Ave. asaiia-; (hiatus) > Sogd. y, Yagh. y: Sogd. ʾʾy /āʧ/, Yagh. ₍ ψуʳфʦʣϊлʷaʳ (3rd pers. sg. impf.) < ŋ ă < ʼa < *ʜa=ʜáha, Ave. h‘ уδκπлέляыьф; (ad i.) ŋ often disappeared in *Proto-Sogdic. The loss of ŋ caused i-Umlaut of , *u, ŋ›, ŋ u (II.1.2.1.iv-v, vii-viii.; II.1.2.2.iii-iv, vi-vii.; II.1.2.5.vi-v.; II.1.2.7.iii-v.; II.1.2.9.iii-v.) or palatalization of *zd, *ʦ, *ʣ (II.1.3.21.iv.; II.1.3.22.ii.; II.1.3.23.ii.). Palatalization of consonants is widespread mainly in Khōtanʣʳʣ. In Sogdian the result of palatalization of vowels and/or consonants might gave different phonetic forms of verbal stems originating either form ŋ-‘ ‘-causative or from ŋ- ‘-passive, thus the difference cannot ϐʣл₍udʥʣdлʤʲomлʳpʣὕὕʧnʥлoʤлπoʥdʧanлʷoʲdʳлуδκπлέ5яř-550); II.1.3.25. *u i. > Sogd. w, Yagh. w: Sogd. w r-y wfr-y /wəfr-áъчлYagh. ⁽á”ⁱr ʳnoʷ < ŋuá”›a-, Ave. vafra-; Sogd ⁽₍ (h) /ʷēšъчлYagh. w ‖л⁽‘ ʥʲaʳʳ < ŋu ʦt› ‘-, Ave. vāst› ‘-; Sogd. √p›⁽(ʾ)y √p›⁽₍ √p›⁽₍d /√pə / to ʳʣʣ₎ , Yagh. - to ϐʣʥ < ŋp‘› -u d‘-; Avestan zbaiia- and Vedic hvayati is connected with Pasht. zwa -, OCS. zьv‘t : ₎ovǫ toлϑaὕὕчлtoлʧnvʧtʣ л<лζdʣщл ŋgʰ“uʜ- / ŋgʰu“ʜ- / ŋgʰuʜ-, Tokh. k⁽ā- (MAYRHOFER 1996, 810). 190 Instead of expected †h ₎’ā- (or maybe †h ₎ ā-). Maybe -zuu- is to be understood as an allophone of *-z - < *-zb-. 191 ž emerged from palatalization of *z < *ʣ: žə’‘йžəbba < *zⁱba < ŋ ₎’ā < *(hi)ʣuā-, but PashtщъτazХʲХлžə’‘йžəbba may be a loan (or influenϑʣŠфлʤʲomлπʧndʦХлjibʰa (RASTORGUEVA – ÈDEL MAN 2007, 404-405) 192 Instead of expected †h d - < *hi - < *hiʣ -. Probably a Median loan. ·123· II.1.3.26. *ʜ193 i. ii. iii. > Sogd. ø, Yagh. ø: Sogd. ʾst ʾsty sṯy /(əфʳtíчлáʳtуɨ)/, Yagh. ást( ) ψуʳфʦʣъʧtϊлʧʳ < *ʜást , OAve. astД, OPers. astiy, Ved. ást , Ide. *h ést ; Sogd. ʾsp-y /əʳpíъчл (ʾ)sp-y /( фʳpíлъ, Yagh. asp ʦoʲʳʣ < ŋáʦu‘- < Ir. *ʜáʦu‘-, Ave. aspa-, OPers. asa-, Ved. áśv‘- < IIr. *ʜáću‘- < Ide. *h ékuo-s; Sogd ⁾⁽ t›-y /ʸuštʲíъ ϑamʣὕ < ŋu⁾ t›í < *ʜu t›‘-, Ave. u t›‘-, Ved. ú ṭra-; (in forms of internal augment) > Sogd. V, Yagh. V ~ ø: Sogd. √pt ⁽ √pt (ʾ)⁽ √pṯ ⁽ : √pt₍ ⁽ , √ptʾ₍ ʾ⁽ √pt₍ ⁽ √pṯy ⁽ /√p : to hear (pres. stem : impf. stem) < *pătĭ- áo ă- : *păt - ao ă- < *pati-gáu a- : *pati-ʜ‘чgáu a- у×лYagh. d - : ad - < *(pă)tĭ- áo ă- : ŋ ч(pă)tĭ- áo ă- < *pati-gáu a- : *pati-ʜ‘чgáu a-); Sogd. √ʾ⁽₎ √ʾ⁽j ( ) : √⁽ʾ₎ √⁽ʾcʾ / ə л: ə ъ to dismount (pres. stem : impf. stem) л< *áo-ž ă - : * ău -ž ă - < *ʜ‘u‘-ʣgád- : *ʜ‘u‘-ʜa=ʣgád-; Sogd. ʾʾy /āʧ/, Yagh. ₍ ψуʳфʦʣϊлʷaʳ (3rd pers. sg. impf.) < ŋ ă < *ʜ (h)a < *ʜa=ʜáha, Ave. h‘; *aʜa > Sogd. ā, Yagh. : Sogd. ⁽ʾt /ʷātъчл Ḳagh. vāt‘- (trisyllabic) < Ide. *h u“h nto-, Lat. ventus. ʷʧnd л <л *u‘ʜąt‘-(ka-), Ave. II.1.4. Syncope and reduction Syncope and reduction are phenomena related to stress changes (see chapter II.1.1.), mainly with the Stress I and Stress II. Examples of old vowel syncope can be observed in a few Sogdian zyrn /zeṙnъл ʥoὕd л < *ʣᛑn ‘ryp - /rep áъл noon л <л ŋ›áp ϑ ā etc. (SIMS-WILLIAMS 1989b, 181) – these examples show loss of an unstressed vowel already in a *Pre-Proto-Sogdic period (i.e. probably in the late Old Iranian period, but dating is really uncertain in this case). Of old date can be also a reduction (shortening) of ŋā > ŋ ‘ in rwps əs/, Yagh. < ProtoSogd. Sogdian and Yagh ŋ›áopăs‘- < ŋ›áupāʦa- ʤoʸ [Ved. lopāśá-, Pers. ]. More certain examples of syncope can be observed in *Proto-Sogdic development – due to shift to the Stress II unstressed vowels (in an open √ʾ₎⁽ʾ₍›t √(ʾ)₎⁽ʾ₍›t √₎⁽ʾ₍›t √₎⁽₍›ṯ ъ√ zʷíṙt/, Yagh. ₎ⁱ⁽í›t- to tuʲn л<л*ø₎⁽áⁱ›tø- < *uʣ-uá›t(‘) ‘mʾ (h), mʾ w mʾ⁾ ъmāʸл<л gh. m ⁾ ʷʣ л<л*ø ø< < *ah am. In YaghnōϐХлʷʦoὕʣлfirst syllable was √pt ⁽ √pt (ʾ)⁽ /p -/, Yagh. d - to ʦʣaʲ < *(pø)tø áo ø- < *pati-gáu a-; Yagh. ž‘vá›- ‖ žⁱvá›- to bring, to produce, to ʧnvʣnt л<л*(nø)ž á›ø- < ŋn ǰ-’ᛑ- (KhROMOV 1987, 661). Vowel reduction continued later on in Sogdian and YaghnōϐХлʧnлdʧfferent ways. In Sogdian all historical short vowels *a, *i, *u (and also Sogd. e from i-Umlaut of short *a (y)z-y, ʾ (y)z-y (y)j-y, ʾ j-y / žíл<л ʣžíъл<л * éž < ŋ’á₎d ‘- ϐadчлʣvʧὕ ) could have been 193 With some exceptions, I will not mark *Proto-Iranian laryngeals in the presented work. ·124· reduced to Schwa (ə) or to its allophone , only old *u after a velar changed to ŋ ə (i.e. old *u caused labialization of a preceding velar; see chapters II.1.3.3.v., II.1.3.11.iii., II.1.3.7.ii.). In YaghnōϐХлʳʦoʲtл*a, *i and *u (of *Proto-Sodgic origin or from loans from or via Persian/Tajik) tend to be reduced in an open syllable when they directly precede a stressed syllable – the short vowels probably changed to *Schwa in (late) *Proto-YaghnōϐХчлtʦʧʳл*Schwa later developed into short (non-reduced) a or ultra-short (reduced) ⁱ or . Ultra-short developed from *Schwa which was followed by a labial consonant or h, , ᴥ and a stressed labialized vowel , , /ʉ (< , , ): Yagh n (also фл pʲayʣʲ л <л [Pers. ]; Yagh. b ʳpʲʧnʥуtʧmʣф л <л ἵʣʲʳщл [TMast. b hó›]; Yagh. m t mʧnutʣ л <л οuʳщл . In other cases *Schwa usually changes to a ‖лⁱ: Yagh. ⁾‘pá› ‖л⁾ⁱpá› nʣʷʳчлʲʣpoʲt л<лἵʣʲʳщл⁾‘’á› < Arab. ḫabar; Yagh. k‘m dá ‖л kⁱm dá Angelica pὕant чл ϑʤщл ρ₍₎щл (there are no many indigenous YaghnōϐХлʣʸampὕʣʳлoʤлdʣvʣὕopmʣntлoʤл*Schwa as the unstressed short vowels have been lost in *Proto-Sogdic or *Proto-YaghnōϐХфщ Another example of reduction in *Proto-Sogdic is loss of * › under several circumstances (see chapters II.1.2.7.vii.-viii., II.1.3.17.iii., II.1.3.9.v.-vii. and for YaghnōϐХлaὕʳoлII.1.2.7.v.): Sogd. √(ʾ)k›t- √ʾk›t- √(ʾ)kt- √(ʾ)‹ṯ- ъ√ᶤkt-/, Yagh. íkt‘ to do, to make (past part.) л< ŋk t‘-(ka-); kj ‹ž krz, kr krj ‹›ž /kaṙžъфл mʧʲaϑὕʣ л<лŋká›ǰ‘-; Yagh. ‖ k›m(ʾ)₍›, kyrmyr qrmyr qyrmyr √(ʾ) ʾm, √ʾp ʾm √ʾ ʾm, √ʾp ʾm √” ʾm ъ√ gh. - to ʳʣnd л<л*fra- etc. In YaghnōϐХлaὕὕлʷoʲd-final vowels were lost, in Sogdian heavy-stem word-final vowels were lost also, but they have been preserved in light-stem endings. As syncope can be explained origin of indicative present and imperfect ending of the third person plural - t. It originates in older - ›- t194 (attested as - › t by JUNKER 1930, 107). The development of the ending can be reconstructed as follows: - ›(-) t > - (₍) t > - ₍ t (attested in speech of village of Marghtimayn; KLIMCHITSKIY 1940, 99-100) > - t (cf. NἴσÁἰ [in print], note nr. 23). II.1.5. Prothesis and epenthesis Syllabic structure of *Proto-Sogdic permitted presence of word-initial consonant clusters, this feature slowly appears to change in *Proto-Sogdian and *Proto-YaghnōϐХл after the split of *Proto-Sogdic – in both of the derived (proto-)languages the word-initial consonant clusters were not allowed so they were transformed: YaghnōϐХлʳʦoʷʳлepenthesis – a svarabhakti vowel a, ⁱ or was inserted to break the original initial consonant cluster; Sogdian shows prothesis rather than epenthesis – the prothetic vowel is spelled as ᵊ in the presented thesis, but in front of s often appears its allophone ᶤ (we can suppose presence of ᶤ according to texts written in the From Iranian perfect indicative active voice ŋ-›( ) > *-ā› ; and (originally) durative ending - t attʣʳtʣdлʧnлσʣʳʳantaʲaлηāta₎aфщ 194 ·125· ʾ tn Manichaean script, where the epenthetic vowel is often spelled by ayin before s instead of more common āl‘p;лϑʤщлδκπлέь57). After the split of *Proto-Sogdic two kinds of prothesis/epenthesis appeared – vocalic and consonantal: 1) As have been mentioned above, vocalic prothesis appears in Sogdian, in YaghnōϐХлtʦʣʲʣлʧʳл vocalic epenthesis observable in analogous positions. Sogdian prothetic ᵊ (and ᶤ) usually appears before inherited word-ʧnʧtʧaὕлϑὕuʳtʣʲʳлуϑʤщлδκπлέь57фчлpʲotʦʣtʧϑлvoʷʣὕлϑanлappʣaʲлaὕʳoлϐʣʤoʲʣл a historical single consonant – this feature is observable mainly for Sogdian k and x уδκπлέь5Ś160), peripherally also for Sogd. preceding historical – the *Proto-Sogdic velars were probably labialized and labialization was then reanalyzed as a consonant cluster (see chapters II.1.3.3.v., II.1.3.11.iii., II.1.3.7.ii.). There are also other examples of prothesis before a historically single consonant – some examples are givʣnлʧnлδκπлέь5Ś-161 – in all those cases the prothetic vowel emerged from secondary built clusters of ŋCu or ŋC : ʾky / < ŋk ă- < ŋkáh( ā)-; ʾ ⁽(ʾ), ʾ ⁽(ʾ) / у{ф ʷāъл tʷo л <л ŋdu‘-; ʾcw / ŋč āk‘m < ŋč -āk‘-. Different example of prothesis before a single consonant may be seen in Sogdian: ʾpk -y ʳʧdʣ л < *upa-k‘ ‘-, Skt. pak a- (claimed as a Sanskrit loan in Buddhist Sogdian by Gerschevitch (δκπл έь6ь), but cf. Yagh. k‘pá ‖л kⁱpá ) – in this case we can assume pronunciation ᵊpkə í (ϑʤщлἶaʲХϐльюřючл5ылέьэ77ф rather than *ᵊpək í уϑʤщлδκπлέь6ь; but see the same example in chapter on metathesis II.1.7.); comparable example may be πoʥdʧanлʷoʲdлʤoʲл ʤatʦʣʲ śл ʾptr-y (ʾ)pt›-y(y) (ʾ)pṯr-y / pt(ə)ʲíъл<лp tá-r195. It should be noted that there are no many examples of prothesis in Christian Sogdian texts. In YaghnōϐХлtʦʣʲʣлaʲʣлtʦʲʣʣлʣpʣntʦʣtʧϑлʳvaʲaϐʦa₎tʧлvoʷʣὕʳлa, ⁱ and . Svarabhakti a appears mainly in Eastern YaghnōϐХчлin the Western and Transitional dialects there is ⁱ instead (but ⁱ appears in many Eastern YaghnōϐХлʷoʲdʳлaὕʳoфщлπvaʲaϐʦa₎tʧл is quite rare, it can be considered as allophonous variant of a or ⁱ. It can be said that svarabhakti ⁱ is a typical epenthetic vowel in ʣvʣnʧnʥ л< * < ŋ‘’ -‘ ā›‘-ka-; YaghnōϐХчлʧtлappʣars in majority of words, e.g. younʥʣʲлϐʲotʦʣʲ л< * etc., see also Russian > Yagh. tⁱ›áktⁱ› tʲaϑtoʲ щ In other cases there appears a in the Eastern dialect and ⁱ in the Western and Transitional dialects – this often happens in clusters beginning in *ϑ and * (< *d): t‘”á›- ‖лtⁱ”á›- to givʣ л< *ϑ ar- < *fra-’ᛑ-; s‘›á₍ ‖лtⁱ›á₍ tʦʲʣʣ л< *ϑ ‘-; d‘›áu ‖лdⁱ›áu ʦaʧʲ < ŋd›áu‘-; ‖лdⁱ ʳʧϑὕʣ л< * ›āϑ < ϑra-. The third svarabhakti vowel – was originally an allophone of ⁱ and a, it appears only when a following syllable contains a labial consonant followed by a stressed back vowel (i.e. , , /ʉ < , , * ): t ‖t ,t ʤouʲ л< ŋ(čə)ϑ < ŋč‘ϑu -; t ‖лt ψуʳфʦʣϊлʥivʣʳ 196 < *ϑ ar-t- t < *fra-’ᛑ-ti-⃝. But emergence of the prothetic vowel can be interpreted also in a different way – the Sogdian root may originate from a stem comparable to Avestan (p)tā (nom.), ptarəm (acc.) or fə › (dat.). 196 For the change a > see chapter II.1.2.1.x. 195 ·126· In YaghnōϐХлʧʳлattʣʳtʣdлaὕʳoлvowel epenthesis in some word-final clusters ending in *xm, *xn, * n, ŋ m, ŋ(⁾) n, ŋčn, *fr, *zm and * n: ›á⁾ ⁱn daʷn < ŋ›áu⁾ na-; ⁽á”ⁱ› ʳnoʷ < ŋuá”›a-; ⁽á⁾ⁱn ϐὕood < ŋuáhun -; í₎ⁱm fiʲʣʷood < ŋá ₎ma-; Yagh. ⁱn, an197 oʧὕчлϐuttʣʲ л<лŋ›áugn‘- (cf. KHROMOV 1987, 661). Anaptyctic vowel in word-final clusters might appear also in Sogdian, but due to Sogdian spelling there are no many clues to prove it, the only example can be seen in a ʷoʲdл ʤoʲл ϐuttʣʲчл oʧὕ л ʷʦʧϑʦл ʧʳл attʣʳtʣdл aὕʳoл ʧnл tʦʣ BʲāʦmХл ʳpʣὕὕʧnʥśл ro haṃ, ro aṃ n/. πomʣлotʦʣʲлʣʸampὕʣʳлoʤлanaptyʸʧʳлaʲʣлʳʦoʷnлʧnлδκπлέяřэ-483: e.g. s (⁽)tmʾn ъʳa у{ə ʣaʲtʦчлʳoʧὕ члbut examples given by Ilya Gershevitch may be also interpreted as metathesis (see chapter II.1.7.) In YaghnōϐХл tʦʣʲʣл aʲʣл notл aὕὕoʷʣdл ϑὕuʳtʣʲʳл Cy, so an anaptyctic ⁱ is inserted to break the cluster: Cⁱ₍: Yagh. ʷoʲld л<лἵʣʲʳщ < Ar. dunyā, Yagh. muϑʦчлmany л<лἵʣʲʳщл , Yagh. samalⁱ t airplane л <л Russ. . Some of the Cy clusters often undergo metathesis yC: , . 2) Consonantal epenthesis (excrescence) is attested only in a few Sogdian words. In several words intrusive x before C is attested: Sogd. √np⁾ t- ъ√nəpɨʸšt-ъл to write (past part.) л <л ŋn pí t‘-(ka-) ʾ⁾⁽ t›-y ʾ ⁽ t›-y ⁾⁽ t›-y / ʸ{əštʲíъл ϑamʣὕ л <л ŋu⁾ t›í < ŋu t›‘- (see chapter II.1.3.11.ii.;лδκπлέэ57ф. Before *n can appear intrusive r wrn-w, wrn-y (y)xwrn-y xwrn-y /(yəфʸ{əwníчлʸ{əwnúъ ϐὕood , intrusive r appears also after < *b in rywr əʲъл tʣnлtʦouʳand л<лŋ’á u‘›- (see chapter II.1.3.17.ii.;лδκπлέю5Ś-362). As intrusive can be considered also n which appears in a form of present copula of the second person plural in Manichaean Sogdian: ʾns ʾ /лąʳ a ~ áṁs a/ (see chapter II.1.3.16.iii. but cf. GMS έ7řя). There are no attested examples of consonantal epenthesis in YaghnōϐХщ II.1.6. Assimilation and dissimilation There can be found some examples of dissimilation or assimilation in Sogdian and YaghnōϐХщл At first should be mentioned really old dissimilation *ʣ–st > *d–st in Iranian *ʣást‘- ʦand (Ved. hást‘-, Ide. ŋgʰés-to-) – it appears as zasta- in Avestan, in Old Persian is attested dasta- (here d- can originate either from *ʣ- or from *d-), but in all other Iranian languages the st-y dsṯ-y ъ əʳtíъчлḲagh. ʷoʲdлʤoʲл ʦand лϑomʣʳлʤʲomлdʧʳʳʧmʧὕatʣdлʳtʣmлŋdást‘dast; Khʷāʲщл st, Khōtщлdasta-члBaϑtʲщл лхъὕʧʳtъчлShugh. ustчлοōsh. ost, Khūʤщл stчлπaʲХqл st, Wakh. ast, Yazgh. st, Munj. lost, Yidgh. last, Pasht. lāsчл ἵaʲāch. dȫst, Pers. dast, Pahl. dst */dast/, Parth. dst. The dissimilated form of the word dást‘- may have been influenced by past participle of the verb *dā- toлʥivʣ – *dad-ta- > *dasta- уtʦʣфлʥiving (one) [= hand] щ *Proto-πoʥdʧϑлdʧʳʳʧmʧὕatʧonлϑanлϐʣлʳʣʣnлʧnлʣʸampὕʣлoʤлtʦʣлnumʣʲaὕл ʳʧʸ лʷʦʧϑʦлϑomʣʳлʤʲomл Ir. ŋ⁾ uá ‘m and which was dissimilated in *(Pre-)Proto-Sogdic as ŋ⁾uá ‘m > ŋ⁾uá u > Sogd. ʾ ⁽ ⁽ ⁾⁽ ⁽ ъʸ{əšúъщ Another example of dissimilation can be seen in Sogdian ŋm›ʾ⁽лъmʲāu/ 197 Yagh. form ro än, Pahl. › an may be a loan from MastchōʦХл ρā₍Х₎л (cf. Tjk. ›‘u án, TMast › n), or the epenthetic a was taken from/influenced by ρā₍Х₎щ ·127· án, Pers. › ánчл γāʲʳщл ʷʣʣpʧnʥ < - (see chapter II.1.3.15.iii.) < -, similar, but opposite development is attested in YaghnōϐХ mʧddὕʣ л<лŋm‘d my ʾn m₍d(ʾ)n II.1.3.15.ii.). Voice assimilation of stops following a homorganic nasal is typical for Sogdian. This development probably stared in *Proto-Sogdic, where groups *{m/(n)}{p/b}, *{(m)/n}{t/d}, *{(m)/n}{k/g}члхʻуmфъn}ʻčъǰъ } changed to *Proto-Sogdic ŋm’, ŋnd, ŋ g, ŋnǰ. In Sogdian these clusters changed to ṁb, *ṁd, *ṁg, *ṁǰ; in YaghnōϐХлtʦʣyлϑʦanʥʣdлtoлmp, nt, nk, nč (see chapters II.1.3.15.iv., II.1.3.16.iv.-ix.). In YaghnōϐХл voʧϑʣὕʣʳʳл ϑonʳonantл ʷʣʲʣл voʧϑʣdл ʷʦʣnл tʦʣyл dʧʲʣϑtὕyл pʲʣϑʣdʣdл aл voʧϑʣdл consonant – such voicing appeared after syncope of unstressed vowels as can be demonstrated in following examples: Yagh. d - to ʦʣaʲ < *d - < *(pᵊ)t - < *pati-gáu a√pt ⁽ √pt (ʾ)⁽ /p -/; Yagh. b z n- ‖ bⁱz n- to ₎noʷ л<лŋ’₎ān- < ŋp₎ān- < *apa-ʣ -; Yagh. b dú”s- to glue, to ʳtʧϑ₎ л <л *b ú”s- < *p ú”s- < *upa-dá”ʦa√p w s-, √p wfs√p wfs- ъ√p uʤʳ-/ (see chapters II.1.3.1.ii., II.1.3.2.ii.). II.1.7. Metathesis There are attested several examples of metathesis in Sogdian and YaghnōϐХщлζлʷʧὕὕлmʣntʧonлonὕyл a few of them – some of the below given examples show interesting development in Sogdian, other given examples are my re-interpretations of phenomena incorrectly interpreted by Ilya Gershevitch in his Grammar of Manichaean Sogdian (GERSHEVITCH 1954). or ŋu and a velar sound: this In Sogdian there is well attested progressive metathesis of 198 phenomenon can be well demonstrated on doublets in following examples w r, w th wxth w t(ʾ) dw ṯ(ʾ) wth dauʥʦtʣʲ л <л *duxtarsw (ʾ)₍k 199 s w yk πoʥdʧan л<л*ʦug(u)d ‘- - or These examples show probable development *CuKC (K = any velar) > *Cə KC ~ *CəŬ C or even ŋCŬ əC, i.e. probably there was no metathesis of but after reduction of ŋ the reduced sound retained its labial character, and later caused s ⁽tmʾn wrwm w ⁽ ⁽ ʳʧʸ члaὕὕлʷʧtʦлʳvaʲaϐʦa₎tʧлvoʷʣὕлʲʣϑoʲdʣdлϐyлtʦʣлὕʣttʣʲлwaw уδκπлέяřэ). I s ⁽tmʾn aὕὕ лʧʳлattʣʳtʣdлaὕʳoлaʳл suppose that those examples show metathesis of or ŋu sw ṯmʾn s tmʾn I suppose that the letter waw marks labialization i.e. *sə də , unfortunately, etymology of this word is not given in GMS and is neither known to me. Sogdian wrwm wrm(h) xrwm, xwrm xwrm ъʸ{ʲúmъл <л *xruma- – in the attested spellings it is certain, that the letter waw does not mark epenthesis but metathesis of *u and/or labialization of x. Sogdian ⁽⁾⁽ ⁽, w ⁽ ⁽ w ⁽ ⁽ ʳʧʸ лʧʳлattʣʳtʣdл 198 In most cases I will put down only spelling varieties in the Sogdian script (i.e. secular texts in the Sogdian script or Buddhist texts), in the Manichaean and Syriac scripts no such examples of metathesis are attested. In majority of example I will not give phonetic transcription. 199 sw (ʾ)₍k or s w yk can be also explained as development from *ʦugud ‘-ka-, cf. OPers. spelling <sᵃ-u-g -dᵃ>, <sᵃ-u-g -u-dᵃ> or <sᵃ-u-gᵃ-dᵃ> sw ₍ʾ⁽ s w ₍ʾ⁽. ·128· ʾ ⁽ ⁽ ⁾⁽ ⁽ /{əʸšúчл x{əšúъл < ŋ⁾ uá ‘m – in this case the letter waw again marks metathesis of ŋu.200 κʣtatʦʣʳʧʳлʧnлϑaʳʣлoʤлnumʣʲaὕл ʳʧʸ лpʲovʣʳлaὕʳoлḲaghnōϐХлu⁾ , compared to Sogdian we can reconstruct following development: Ir. ŋ⁾ uá ‘m > by dissimilation ŋ⁾uá ‘m > *Proto-Sogdic ŋ⁾uá u > Sogd. ʾ ⁽ ⁽ ⁾⁽ ⁽ / ʸ{əšúъл şл ϐyл mʣtatʦʣʳʧʳл şл ŋuá⁾ u ⁽⁾⁽ ⁽, w ⁽ ⁽ w ⁽ ⁽ ъ{əʸšúъчл Ḳagh. u⁾ . Another example of metathesis attested in both languages is ⁽›ʾṯy, Yagh. × Sogd. w ›ʾt₍₍ aʷa₎ʣ л<л*u a-ka-. Gershevitch also mentions insertion of r in Sogd. k⁽› dʧffiϑuὕt луδκπлέю6ьфлʷʦʧϑʦлʦʣл compared with Old Persian k‘uϑi-, but this etymology should be unacceptable because after loss of final -i- ŋ-‘u- should be influenced by i-Umlaut. According to spelling of ŋ k ṙϑ in (ʾ) kʾ⁽› (ʾ) k⁽› ‹⁽›ϑ I suppose a different etymology from Ir. Sogdian scripts: ŋskáuϑra- with metathesis *ϑr > * ṙ ϑ. For other examples of metathesis in Sogdian see GMS έяы6-447. In YaghnōϐХл ʳʦouὕdл ϐʣл mʣntʧonʣdл mʣtatʦʣʳʧʳл oʤл *pk- > *kp- in k‘pá ‖л kⁱpá armpit л <л *øpøk‘ ø < *upa-ká ‘-. This YaghnōϐХлʷoʲdлaὕʳoлpʲovʣʳлʲʣadʧnʥлoʤлπoʥdщл ʾpk - ʳʧdʣ члʷʦʧϑʦлʧʳл interpreted as a word with prothetic āl‘p by Ilya Gershevitch: «B. ʾpk - (əp‘k -ж l ght st“m) s d“ VJ 8, borrowed from Skt. pak a-» уδκπл έь6ьфчл ϐutл dʣvʣὕopmʣntл ʤʲomл *upa-k‘ ‘- seems to be more probable, thus the word should be read *ᵊpkə í instead of *ᵊpək í as may be presumed from the Sanskrit form. Essential example of metathesis presents YaghnōϐХлpʲʣʳʣntлtʣnʳʣлʣndʧnʥлoʤлtʦʣлtʦʧʲdлpʣʲʳonл singular -č (originally ending used only in Eastern YaghnōϐХчлnoʷadayʳлʧtлʳpʲʣadлaὕʳoлʧntoлotʦʣʲл dialects, in the Transitional and Western dialect there is ending -t t). The ending -č in 201 Eastern YaghnōϐХлʧʳлʤʲomлdʧaϑʦʲonʧϑлpoʧntлoʤлvʧʣʷлtʦʣлʳamʣлaʳлτʣʳtʣʲnлḲaghnōϐХл-t t < -t- t . In Eastern YaghnōϐХлtʦʣлʣndʧnʥлundʣʲʷʣntлmʣtatʦʣʳʧʳśл-t(-) t > *-t t > -č t (attested in speech oʤл vʧὕὕaʥʣл oʤл λōmʧt₎ōn;л KLIMCHITSKIY 1940, 99-100) or -č (in the Transitional dialect of vʧὕὕaʥʣлoʤлἶūὕ;лANDREEV – LIVSHITS – PISARCHIK 1957, 236) > -t (dialect of Qūl; ibid.) > -č (cf. NἴσÁἰ [in print], note nr. 23). II.1.8. Analogy I have not found much examples of analogy in the languages derived from *Proto-Sogdic, in Sogdian there is problem with spelling, so I will present two examples I have recorded in YaghnōϐХщ Present stem form of the YaghnōϐХл vʣʲϐ ⁱ›í”- ‖л ⁱ›ív- to know, to undʣʲʳtand л <л ŋg›’- to know, to understand, to take, to ʥʲaϐ лpʲoϐaϐὕyлʣmʣʲʥʣdлϐyлanaὕoʥyлʤʲomлpaʳtлpaʲtʧϑʧpὕʣл 200 In some cases thus can be assumed that in Sogdian appeared also progressive labialization. Orthography of labialized ⁾ or appears as <xw>, < w>, <wx> or <w > depending on spelling customs in each script utilized for Sogdian. Orthography similar to Sogdian <wx> or <w > can be compared with Parthian spelling <wx> or <xw> for ⁾ (RASTORGUEVA – MOLCHANOVA 1981, 178-179). 201 Id est Iranian indicative present ending of the third person singular *-ti and (originally) durative ending - t (cf. ʾ tn attʣʳtʣdлʧnлσʣʳʳantaʲaлηāta₎aфщ ·129· ⁱ›í”t‘ < * ›í”t‘-ka- < ŋg ”t‘-ka- < ŋg›’-ta-ka-. In Sogdian there is attested present stem √ r √ rb- /√ əw -/ (in YaghnōϐХл tʦuʳл ϑanл ϐʣл aʳʳumʣdл pʲʣʳʣntл ʳtʣmл ʤoʲmл † irv- or † arv- if there was no analogy). For the YaghnōϐХл paʳtл paʲtʧϑʧpὕʣл gⁱ›í”t‘ √pt ry t- ъ√p t ʲʧ d-ъл to take (past part.) л– ŋ› normally develops into əʳ (with allophones) in *Proto-Sogdic (see chapter II.1.2.7.i.-vi.), but before *ft it changes to ri (chapter II.1.2.7.xi.; δκπл έь5юa). Analogous is also -f- of the verb in focus in Eastern YaghnōϐХл ʧnʳtʣadл oʤл etymologically expected -v- in Western YaghnōϐХщ Another example of analogy in YaghnōϐХл ʧʳл dʣvʣὕopmʣntл oʤл auʥmʣntщл ζл ʷʧὕὕл dʧʳϑuʳʳл tʦʧʳл problem later in chapter on verbal inflection (chapter II.2.4.), now I will mention the phenomenon briefly. In *Proto-Sogdic the imperfect tense has been formed by prefixation of an augment in front of a verbal root. If a verbal stem contained a prefix and a root, the augment followed the prefix – i.e. there was so-called internal augment. In Sogdian augment was preserved only in reflexes of the internal augment, original augment of non-prefixed verbs disappeared due to operation of stress (probably Stress III as there is a different development in YaghnōϐХфчл ϐutл auʥmʣntл oʤл non-prefixed verbs is preserved in YaghnōϐХщл Aʳл tʦʣл ὕanʥuaʥʣл developed further there have been lost awareness of Iranian (or *Proto-Sogdic) verbal prefixes and by analogy the augment have been placed in front of the original prefix. See following examples to demonstrate this phenomenon: Yagh. d - : ad - to ʦʣaʲ (pres. stem : impf. stem) < *(pă)tĭ- áo ă- : ŋ ч(pă)tĭ- áo ă- < *pati-gáu a- : *pati-ʜ‘чgáu a- у×л Sogd. √pᵊt < *pătĭ- áo ă- : *pătД- áo ă- < *pati-gáu a- : *pati-ʜ‘чgáu a-); Yagh. var- : ‘vá›- to ϐʣaʲ л (pres. stem : impf. stem) < * ar- : ŋ ч á›- < *bara- : ʜ‘ч’ᛑ- у×л πoʥdщл √ ər- : √ ər- < ar- : ă ar- < *bara- : ʜ‘ч’ᛑ-). II.1.9. Syllabic structure Syllabic structure of *Proto-Sogdic was probably very similar (if not identical) to Old Iranian syllabic structure. After stress-influenced changes in phonology (and morphology), mainly after vowel syncopation and reduction, the syllabic structure of Sogdic daughter-languages changed considerably. Unfortunately there are no many clues to reconstruct syllabic structure of *Proto-Sogdian and *Proto-YaghnōϐХчлʷʣлϑanлaʳʳumʣчлtʦatлaὕʲʣadyлaʤtʣʲлʳpὕʧtлoʤлхProto-Sogdic there slowly emerged a tendency to avoid word-initial consonant clusters, however, this development is not typical only for Sogdic dialects as it appears in many other Iranian languages, especially in the New Iranian period. I have not met many attempts to reconstruct Sogdian Syllabic structure – there are probably only two outlines of the Sogdian syllabic structure. The fiʲʳtл outὕʧnʣл ʷaʳл pʲʣʳʣntʣdл ϐyл πoʤ ya Petrovna Vinogradova: «Th“ sp“c fic structure of the syllable: CCVCC: C k⁽›ϑ [ k ›ϑ] difficult ж c”. B (ʾ) kʾ⁽› -, M (ʾ) k⁽› [(ə) k ›ϑ-], CV (probably also CCV, CCCV, CVC, VCC, VC): B ʾr kw [ar uk, ar ku] s nc“›“ , S [ ‘st₍‘] h‘nd (loc‘tive), martə⁾m“t p“opl“ (o’l ‹u“)ж [p›ām‘n‘] ‘nd [p›ām‘nd ] B›‘hm‘n (voc‘tive singular and plural), B [nər e-] sco›p on .» (VINOGRADOVA 2000a, 64). The other outline of Sogdian syllabification was presented by Elio Provasi in his ·130· study of Sogdian versification: «Sogd ‘nж th“nж h‘d s₍ll‘’ fication rules which were quite different from those of Western Middle Iranian. In Sogdian, inside a word, a group of consonants between two syllabic peaks (i.e. vowels or diphthongs) is not divided between the two syllables, but belongs to the second one, constituting its onset.[note 33: Cf. the observations, from a historical-comparative point of view, by GERCENBERG 1980, pp. 48-49. уŞл δʣʲϑʣnϐʣʲʥчл ἱщл δщл ьŚřьśл ἴϐл aʤʥanʳ₎omл udaʲʣnʧʧщ л ζnśл I›‘nsko“ j‘₎₍ko₎n‘n “: “ž“godn k лʸʷк, pp. 48-56.)] In other words, a syllable boundary ($) must be inserted immediately after a (short or long) vocalic nucleus[note 34: Including in the defin t on o” long voc‘l c nucl“us ж ’“s d“s th“ long vo⁽“ls йāēД йж ‘lso th“ compl“⁾ nucl“ ( d phthongs ) /Vṛ/ and /Vṇ/ (where /V/ = any vowel).] whenever it is followed by any number of consonants, followed in their turn by another vocalic nucleus: 0 > $ / V_C(C(C))V (e.g. / ‘Юst“й h‘nd (g“n.) ж й⁽ Юžp₍‘/ t“››o› (‘’l.) ).» (PROVASI 2009, 350). It seems that both descriptions of Sogdian syllabic structure are correct, thus the description given by Provasi seems to be more elaborate. Elio Provasi analyzed Sogdian poetic translation of Middle Persian Manichaean hymn cycle ђu₍‘d‘gmān – by the analysis of metrical text there can be assumed much about Sogdian phonology, syllabification and stress (cf. chapters on Stress, II.1.1.4. ff.)щлAϑϑoʲdʧnʥлtoлἵʲovaʳʧ ʳлdʣʳϑʲʧptʧonлʧtлʳʣʣmʳлtʦatлπoʥdʧanлpʲʣʤʣʲʲʣdлopʣnлʳyὕὕaϐὕʣʳчлʳoл syllables starting in consonant cluster were quite often – this situation can be compared to syllabic structure in *Proto-Slavic (cf. SCHENKER 1993, 67) or in contemporary Belarusian (BIRILLO – BULAKHOV – SUDNIK 1966, 163). I am not aware of a tendency for open syllables in other Eastern Iranian languages, I am not sure whether it may appear in Pashtōл уʧtл ϑanл ϐʣл suʥʥʣʳtʣdл ϐyл ἵʲovaʳʧ ʳл ϑompaʲatʧonл oʤл πoʥdʧanл ʳyὕὕaϐʧϑл ʳtʲuϑtuʲʣл ʷʧtʦл δʣʲtsʣnϐʣʲʥ ʳл ʳtudyл onл Pashtōлʳtʲʣʳʳл– unfortunately I was not able to get this article; cf. PROVASI 2009, 35033). Syllabic structure of YaghnōϐХл ʦaʳл notл ϐʣʣnл dʣʳϑʲʧϐʣdл ϐyл manyл ʳϑʦoὕaʲʳл either, the only description can be found in an outline of YaghnōϐХлϐyлπoʤ ya Petrovna Vinogradova: «Prevailing syllabic patterns: 1) CVCж (C)VCC (”o› monos₍ll‘’ c nouns): k‘t hous“ ж p t ‘››o⁽ ж mēt d‘₍ ж v d sm“ll ж ‘›k ⁽o›k ж u›k ⁽ol” ж ētk ’› dg“ ж pun ”ull ж n‘₍s nos“ ; м) CV-, CVC- (for di- or trisyllabic ú›d‘ “₍“ ж divá› doo› ж ₎iv m ll ж ⁾ ní t‘ ’utt“› ж n pá₍ n n“ph“⁽ .» (VINOGRADOVA 2000b, 293-294). In YaghnōϐХ there are also monosyllabic words like CV, VC or even V (e.g. č ʤʲom ч ax (s)ʦʣ ч Д onʣ ), but they are not so frequent as the above mentioned CVC and (C)VCC monosyllabic words. YaghnōϐХл syllabic structure is the same as syllabic structure of neighbouring Tajik or Uzbek, but I suppose that in this case the similarity is not due to language contact. ·131· II.2. Historical grammar In following chapters I would like to present basic features of Sogdian and YaghnōϐХлʥʲammaʲщл Both languages differ considerably, but from diachronic view they can be seen gradual development towards simplification of Old Iranian system. I will focus mainly on description of nominal and verbal systems – with primary attention to description of features inherited in both ὕanʥuaʥʣʳщл κanyл ʥʲammatʧϑaὕл ʤʣatuʲʣʳл ʷʧὕὕл ϐʣл ϑompaʲʣdл ʷʧtʦл dʣvʣὕopmʣntл ʧnл tʦʣл ἵāmХʲл languages as there may be seen many common tendencies in development of Sogdic dialects and languages of the ἵāmХʲлʥʲoupщ II.2.1. Nominal inflection The Old Iranian system of nominal inflection was radically transformed in majority of the Eastern Iranian languages. In Avestan and in Old Persian original eight cases, three numbers and three genders are preserved. Inflection distinguished two inflectional categories – thematic and athematic nouns. The thematic nouns distinguished vocalic a-, ā-, Д-, i-/‘ -, u-/‘u- and -stems, the athematic stems ended in a consonant (i.e. p-, b-, t-, d-, n-, nt-, s-, -, h-, r-, r-/n-, k- and g-stems). In the Middle Iranian a syncretism of cases emerged, which resulted in three cases system (nominative/direct case : oblique case : vocative)202 and gradual merger of gender (in many languages remained distinction of masculine (< originally masculine + neuter) and feminine, however, some languages do not distinguish gender at all). The three-case system was pʲʣʳʣʲvʣdлκun₍Х-YidghāчлPashtōлandлτa etsХчлʧnлtʦʣлotʦʣʲлλʣʷлIranian languages the vocative case merged with the nominative. Such outlined development of cases and gender is typical almost for all Eastern Iranian languages (except Ossetic 203 ), it can be found in the Western Iranian languages too204. Case syncretism was certainly a gradual process, from the Middle Iranian languages only Old Khōtanʣʳʣлfully preserved a six-case system with series of inflectional classes (however in Late Khōtanese the case system has been reduced). Somewhat simpler six-case system (for the lightstem words) is attested in Sogdian – a gradual reduction towards three-case inflection can be seen. Khʷāʲezmian had three cases, in Bactrian there were just two cases. In all Eastern Iranian languages masculine merged with neuter, only Khōtanese developed a nʣʷ л nʣutʣʲл ʤʲomл oὕd By comparation of preserved inflectional endings in YaghnōϐХчл ʧnл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл andл ʧnл Pashtōл ʷʣл ϑanл suggest four-case system: nominative-accusative : vocative : genitive-possessive : inessive-oblique (by syncretism of old locative, ablative, dative and instrumental) – see Table 40. 203 The development of Ossetic has been different – we can certainly think about emergence of two-case system based on opposition of nominative/direct case : genitive/oblique, original seems to be ablative and inessive (derived from the locative case); other Ossetic cases emerged anew, probably due to contact with Caucassian languages (cf. KIM 2003; 2007; BELYAEV 2010; chapter I.1.1.3.2.). 204 In the Western Iranian languages there is majority of vernaculars with two- or three-case system, some other languages, such as Persian, lost its inflectional endings, but nominal endings show simplification of the two-case system. 202 ·132· n-stems; in Sogdian there are few relicts of a-stem neutres. Dual began to lose its original function too; it was marginally preserved in Khōtanʣʳʣл andл Khʷāʲezmian; in Sogdian dual shifted to numerative. ζnлἵāmХʲлτakhХлaὕʳoлopʣʲatʣdлtʦʣлʳynϑʲʣtʧʳmлoʤлϑaʳʣʳ (in singular there is just one case, in plural there is direct case, oblique and objective), nowadays several relicts of the original inflectional system still can be seen. Reflexes of several old cases are shown by Tat yana Nikolaevna Pakhalina in her comprehensive study of WakhХ (PAKHALINA 1987a) – archaic inflectional system was preserved in reflexes of ā- and i-Umlaut in several WakhХлʷoʲdʳ: dat. sg.: pətr son < ŋpúϑ›‘ (but also < nom. sg. ŋpúϑrah ?); ə d douʥʦtʣʲ < ŋdúgt›‘ ; instr. sg.: ‘nd‘›č ʦuʳϐand ʳлϐʲotʦʣʲ ʳлʷʧʤʣ < ŋ ánt(‘)›ā-kā; k‘ ϐoy < ŋká› nā; ₍ā / hā; ā₍ уʦuфman < war ʲam < ŋuá›nā; loc. sg.: pər-cəng ϐʲaϑʣὕʣt < ŋup‘› -čáng‘ ; pəlingə t ʲʧnʥ < ŋup‘› -‘ngú t‘ ; p lД₎ ʥaʲdʣn < ŋup‘›‘-d‘ ʣ‘ (PAKHALINA 1987a, 444-445, 449). By means of operation of ā- and i-Umlaut in WakhХлtʦʣʲʣлaʲʣлnotлpʲʣʳʣʲvʣdлonὕyлtʦʣлʲʣflexes of the original cases but also reflexes of nominatives of old dual (mainly in appellatives labelling paired entities or things culturally perceived as pair) and plural (for collective number): nom.-acc. du.: ’ā› dooʲ < ŋdu ›ā; pā / pād ὕʣʥуʳфчлὕoʧnʳ < ; ast ʦandуʳф < ŋdástā < *ʣ‘stā; u ₎ʧdnʣyуʳф < *ʦú ; čə(ž)m ʣyʣуʳф < ŋčá m‘ ; v ›əw уʣyʣфϐʲoʷуʳф < ŋ’›úu‘ ; Д ʣaʲуʳф < ŋgáu ‘ ; t ⁾m ʳʣʣd < ŋtáu⁾m‘ ; kak ʣyʣуʳф < ŋkák ā; əw ʦoʲnуʳф < *ʦ›úu‘ / *ʦ›úuД; wəḷtk ὕunʥуʳф < ŋu t‘-k‘ ; bərət ʳpo₎ʣуʳф < ŋdu -‘›áϑni; yu yo₎ʣ < ŋ úg‘ ; nom. pl.: ₎ā ϑʦʧὕdʲʣn < *ʣ ϑāh; ₍opč ʳʦʣʣp (coll.) < *paʦ(u)uā-kāh; yangl finʥʣʲуʳф < ŋáng(u)›āh; (₍)‘₍č ϐonʣуʳф < ŋást‘-kāh (PAKHALINA 1987a, 444-447, 449-450). ·133· As outlined above, the Middle Iranian languages distinguished two genders: masculine and feminine (and some of them relicts of neuter). In the New Iranian period there are many languages which still retain gender (e.g. Pashtōчл τa etsХчл ḲazghulāmХчл κun₍Хл andл Yidghā, languages of the ShughnХ-RōshānХл ʥʲoupл ʣʸϑʣptл πaʲХqōὕХ), but some of them lost gender (YaghnōϐХ, Ossetic, WakhХ, πaʲХqōὕХ, IshkāshmХ and SanglēchХ). The original gender system has been in fact preserved only in Pashtōчл τa etsХл andл κunjХ-Yidghā;л ʧnл ḲazghulāmХ and ShughnХ-RōshānХл ὕanʥuaʥʣʳл tʦʣл dʧfference in gender was replaced by semantic-syntactic distinction. Ir. OAve. ProtoSogdc. Sogd. nom. *aʦu‘h ŋáspaʼ əspí voc. *aʦu‘ aʳpō ‘spā ŋáspa əspá acc. *aʦu‘m aspəm gen. dat. *aʦuā aʳpaʦēч ‘sp‘h ā ‘spā ŋásp m ŋásp‘ʼ ă əspú *aʦu‘h ‘ abl. *aʦuāʼat Sogd. C5 Yagh. sg. instr. loc. *aʦuā *aʦu‘ ə ŋáspā ‘spā(a)t ‘spā ŋáspāʼ ‘sp ā ŋáspă ā əspí asp əspíД áʳpʧ ə ŋáspā du. ə áʳpʧ numv. nom. voc. *aʦuā aʳpā ŋáspā ə *aʦu‘ āh aspaiia ŋáspă āʼ *ə *aʦuā’ ā aʳpaēⁱϐʧʧāч aʳpōʧϐʧʧā ? *aʦu‘ ‘h aspaiiaчлaʳpōʧʧō *aʦuāh(ah) aspa, aspa ʦō ŋáspāʼ ə *aʦu nh aspəṇg ŋáspān *ə acc. gen. dat. abl. instr. loc. ? ásp (щ) pl. nom. voc. acc. *aʦuān‘ʼ ‘sp‘nąm m *aʦu‘ ’ ‘h ‘sp‘ēⁱ’ ж ‘sp ’ ŋásp‘ ăʼ instr. *aʦuā ‘spā loc. *aʦu‘ u aʳpaēšū gen. dat. abl. ? ? ə ? †ə ? †asp v, †áspД” ? Table 39 Development of a-stem inflection of masculines (given on example ŋáʦu‘- ʦoʲʳʣ ) in Avestan, Sogdian and YaghnōϐХ. YazghuὕāmХл maʳϑuὕʧnʣʳл maʲ₎л maὕʣл namʣʳл andл ʧnanʧmatʣл tʦʧnʥʳ;л tʦʣл ʤʣmʧnʧnʣʳл ʧnϑὕudʣл female names and animals regardless their natural gender (the feminines also contain several words that have retained its gender in relict forms). In the ShughnХ-οōshānХлὕanʥuaʥʣʳ (except πaʲХqōὕХфлdʧdлnotлappʣaʲлʳuϑʦлaлʲadʧϑaὕлtʲanʳʤoʲmatʧonлoʤлʥʣndʣʲлaʳлʧnлḲazghuὕāmХśлaʳлmaʳϑuὕʧnʣʳл ·134· are perceived some original masculines, some male names, male animals and geographical names and in means of collective noun; as feminines are considered female names, female animals and majority of substantives perceived as a single unit (see Chapter I.1.1.3.5.). Reflects of Old Iranian gender are morphologically preserved not only in the ShughnХ-οōshānХл ὕanʥuaʥʣʳл уʧnл tʦʧʳл ϑaʳʣл paʲtʧaὕὕyл ʧnϑὕudʧnʥл πaʲХqōὕХфл andл ʧnл ḲazghuὕāmХчл ϐutл some traces of gender have been preserved in WakhХлoʲлζsh₎āshmХщлοʣmaʧnʳлoʤлmoʲpʦoὕoʥʧϑaὕὕyл (i.e. originally with different ending) expressed gender can be observed in outcomes of effect of ā- and i-Umlaut on originally stressed root vowel; such feature can be documented on the following example: Ir. *xara-h ×лŋ⁾‘›ā-ø (nom. sg.) aʳʳл×лʳʦʣ-aʳʳ лşлοōsh. o› ×л ā›, Bart. ȫ› ×л ā›, οāshrv. u› ×л ā›, Wakh. xur ×л*xar (in moč⁾‘›, lit. female-ass); but SaʲХqщл “›, Yazgh. ⁾ ›, Yagh. xar (< m.фл× Munj. xắăra ⁾ă› ), Yidgh. ⁾á› (< f.). (PAKHALINA 1987a, 444-446) Transformation of the inflectional system, gender and number was probably iniciated by stress shifts. Probably a gradual syncretism and loss of inflectional endings emerged as a consequence of stress strength and its shift on a root (?). Simultaneously with the transformation of the inflectional system also masculine merged with neuter (the neuter differed from the masculine only in different endings in nominative and vocative of all three numbers) and with reconstruction of athematic stems as -stems. Case endings of the -stems gradually generalized also in other vocalic (thematic) stems, the original thematic stems were retained marginally. The above outlined development can be demonstrated quite well in an example of masculine a-stem inflection in Sogdian and YaghnōϐХ – by comparison of both languages with Old Iranian and Avestan is possible to reconstruct also *Proto-Sogdic inflection (see Table 39). As a result of ending loss it was necessary to revise inflectional syntax – the loss of forms of cases of location and direction was syntactically replaced with adpositional constructions (it is possible that Old locative and ablative cases of location or direction joined with adpositions already before the loss of inflectional endings in these cases). Development of genitive and accusative was quite different – both cases have an important role in syntax. Accusative as a case of direct object gradually merged with nominative. But genitive in the Indo-Iranian languages gained a new function when compared to the Indo-European proto-language – it became the case of the verb object in ergative construction. The loss of inflectional endings and case syncretism caused two say undesirable morphological phenomena: 1) the nominative plural endings were lost (in case of absence of ā-Umlaut) and thus forms of nominative plural and singular merged; and 2) genitive and accusative cases were reanalyzed. In singular the form of genitive merged with dative, but in plural the difference between genitive ×лdative(-ablative) remained. In case of accusative there is well attested the difference between accusative ×лʥʣnʧtive in singular, but it is possible that in plural both cases started to merge both in function and in pronunciation. Such feature is observable in Sogdian (respectively in texts younger than the Ancient Letters), where the archaic form of accusative plural -ān behaves as oblique. The archaic accusative in -ān resembles to genitive in -ān (in the Ancient Letters still -ānu). The syncretism of genitive with some cases in ·135· singular and with some other in plural probably led to the dichotomy of function of genitive, dative and accusative: in singular there was opposition accusative ×л ʥʣnʧtive-dative, in plural, however, accusative-genitive ×лdative(-ablative). Yagh. Wakh. Shugh. / οōsh. / Bart. sg. pl. sg. pl. sg. -ø -t -ø -Д t -ø pl. οāshrv. sg. pl. πaʲХq. sg. pl. -ø -x yl rec. acc. gen. obl. dat. m. sg. f. Ishk. Yazgh. pl. sg. pl. sg. pl. -Х -ø -ó -ø -áϑ -уyфō | -Х voc. nom. Munj. -i abl. -ti -ēv, -Хʤ -i -ø -əvi -ēn -əv -ēv, -ēʤ / / -Хʤ -ēn -ø | -a -ьn -ān | -Хn -Хʤ -ef -i, -y -ó₍ -en -i -áϑi -āʤ instr. loc. -i Table 40 Summary of endings in YaghnōϐХлandлʧnлtʦʣлἵāmХʲлὕanʥuaʥʣʳ with account of historical development of individual endings (values in italic letters present endings derived from endings other than those derived from a-stem endings in individual cases; values in grey letters mark change of meaning of the ending; underlined letters label archaisms). In singular the three-case system emerged from reanalysis of nominative, vocative and genitive – nominative merged with accusative (> direct case), and genitive merged with all other oblique cases (> oblique). As mentioned above, vocative remained as individual case only in κun₍Х-YidghāчлPashtōлandлτa etsХ, in all other languages it was replaced by nominative. In the ἵāmХʲлὕanʥuaʥʣʳлoʤлBadakhshānлʥʲaduaὕὕyлϑʣaʳʣdлoʲлϑʦanʥʣdлʤunϑtʧonʳлoʤлʥʣnʧtive/oblique – in WakhХ and Ish₎āshmХлit changed to objective case, in YazghuὕāmХл changed to possessive case; and in the ShughnХ-RōshānХ group it disappeared completely. Different changes occurred in plural than in singular. Due to the loss of the original endings of nominative plural there can be observed two tendencies: 1) emergence of new ending of nominative plural (see endings in YaghnōϐХ, WakhХ, IshkāshmХ, YazghulāmХ and SarХqōὕХ in Table 40); 2) there was reanalyzed the original ending of genitive(-accusative) plural, whose ending passed transferred to nominative (see forms of plural endings in ShughnХ, RōshānХ, BartangХ, οāshārvХ and YazghulāmХ in Table 40). After the genitive form began to function instead of the nominative plural, it was necessary to create a new form of the oblique case – this has become the dative-ablative ending. Sogdian inflectional system preserves a rich stem system, however, it was transformed a lot in comparison to the Old Iranian stage; it distinguishes -, -, -, - and -stems, but there are no consonant stems – they were revised and according to their gender they merged with either a- or ā-stems. Inflection of the -stems became dominant and later on many - and -stem words were inflected as -stems. In the North Eastern Iranian languages essential innovation operated, which separates this branch from other Eastern Iranian languages: from Iranian abstract suffix *-ϑu -/*-t(u) - emerged new plural ending *- -. This new ending was added after the thematic vowel in Sogdian and it was inflected as ā-stem singular feminines. ·136· The plural ending in *- - unʧquʣὕyлappʣaʲʳлʧnлπoutʦʣʲnлἵāmХʲлὕanʥuaʥʣʳл – in Ish₎āshmХлandл WakhХśл ζshk. -d in words sь nd ʦaʧʲл (pl.) л <л *ʦā› ‘-g‘un‘- -; mend appὕʣʳчл appὕʣ-tʲʣʣʳ л <л ŋ‘m‘›n ‘- - and čь⁽“nd apʲʧϑotʳчл apʲʧϑot-tʲʣʣʳ л <л *⃝u‘n - - tʲʣʣʳ ;л andл τakhХл ʣndʧnʥл oʤл direct case plural - t originating in Iranian *-Д-ʦu‘- - or *- (n)- -. From Iranian *-ϑu - comes YazghuὕāmХл pὕuʲaὕл ʣndʧnʥл -áϑ; nota bene Persian plural ending -h (> colloq. Tjk. > Ishk. -o) is also of the same origin. *Ir. sg. m. voc. -a nom. -ah Sogd. n. Sogd. C5 Sogd. a-stems light stems n. m.n. m. Yagh. heavy stems m.n. *Ir. - -am -í -ú Sogd. C5 Sogd. Yagh. heavy stems f. - -á -ø -ø -Х -Х -i -i ā-stems light stems f. f. f. -a -aʧ -éчл-á -í -ø -ø -ā -ā -ām -ūчл-ā -Х -āʧāʦ -yā -Х -i -i -āʧуāʧ) -yā -áХ -áД -Х -i -āʧ -yā -áД -Х -i -ú acc. Sogd. gen. -aʦʧa -ē dat. -āʧ -ē -íХ -íД loc. -aʧ -yā -íД abl. -āt -ā -āʧāt -yā -áД -Х -i instr. -ā -ā -уaʧфā -уyфā -áД -Х -i Table 41 Overview of -stem inflection in Sogdian and in YaghnōϐХ. *Ir. Sogd. C2 Sogd. Yagh. *Ir. Sogd. sg. m. voc. -aka nom. -akah n. m. -a-kam - n. m.n. - - - -a -a -akam - gen. -a₎aʦʧa -ē -ē dat. -a₎āʧ -ē loc. -a₎aʧ -ē abl. -a₎āt - -a₎ā - Table 42 Overview of хФʲщ f. f. -ā₎aʧ -e, -a -ā₎ā - -ā₎ām - -ā acc. instr. Sogd. - -a -a -ā₎āʧāʦ -ē -ē -aʧл/ -‘ / - -ā₎āʧуāʧ) -ē -aʧл/ -‘ / - -ē -‘ / - -ā₎āʧ -ē -‘ / - -ā₎āʧāt -ē -‘ / - -ā₎уaʧфā -ē -‘ / - -stem inflection in Sogdian and in YaghnōϐХ. хФʲщ Yagh. -stems Sogd. хФʲщ Yagh. Sogd. -stems m. f. - m.f. m.f. m.f. voc. -au -a - m. f. m. -ʧʧa -ʧʧaʧ -iya nom. -ušчл-āuš -u acc. -ø -ʧʧaʦ -ʧʧā -umчл-āuam -u gen. -aušчл-uaʦ -ʧʧam -ʧʧām -i -i -ʧʧaʦʧa -ʧʧāʧāʦ dat. -уaфuaʧ -ʧʧāʧ loc. -au, -ø - abl. -ʧʧaʧ -autчл-uat -aʧtчл-ʧāt -Д -i -ʧʧāt instr. -ūчл-uā -i -i -i -ʧʧā -ʧчл-aʧ -ø -ʧуšфчл-ā -i -Хmчл-āʧam -Х -i - Table 43 Overview of -, - and -aʧšчл-ʧ -уaфʧaʧ -Х -Д f. - -Х -yā -i -ʧʧāʧуāʧ) -уʧфХ -( )Д -yāуХф -₍ā(Д) -ʧʧ - -ʧʧāʧ -( )Д -₍ā(Д) - -ʧʧāʧāt -₍ā(Д) (- ) -ʧʧуaʧфā -₍ā(Д) (- ) -stem inflection in Sogdian and in YaghnōϐХщ ·137· Yagh. - -stems sg. -уʷфХчл- Yagh. ā-kā-stems a-ka-stems -yā Sogdian, similarly to some other Eastern Iranian languages, preserves peripheral relicts of r-stem inflection. The relicts of r-stem inflection can be observed in a few plural forms continuing from Indo-European nomina agentis in -ter- (or more correctly from continuants of Ide. *ph -tér-, meh -tér-, ’ʰ›éh -ter-, ŋdʰugh -té›- ʤatʦʣʲчлmotʦʣʲчлϐʲotʦʣʲчлdauʥʦtʣʲ ; moreover by semantic analogy also in ŋsu“só›- ʳʧʳtʣʲ > Ir. ŋp tá-r, -r, -r, ŋdú⁾t‘-r, *x áhar-): Yazgh. v(ə)radár, ə dár brotʦʣʲʳчл dauʥʦtʣʲʳ ; Shugh. οōsh. v › dār ϐʲotʦʣʲʳ (> Shugh. a › (d )r! ϐʲoʳ! ); Sangl. vrudár brotʦʣʲʳ ; ʧnлotʦʣʲлὕanʥuaʥʣʳлtʦʣл r-ʳtʣmлpὕuʲaὕ лʧʳлʣʸtʣndʣdлϐyл normal plural ending: Sogd ›ʾṯrṯ w trth, wtrt / aṙtчл ə {dáṙt/ ϐʲotʦʣʲʳчлdauʥʦtʣʲʳ ; Ishk. vrudarьn (sg. vru(d)), ixodarьn (sg. ⁾ó), ⁽ d darьn brothers, sisters, daughters ; Oss. ” dæltæ ‖лfidæltæ (sg. ” d ‖лfidæ), mæd(t)æltæ (sg. mad ‖лm‘dæ), æ›v‘d(t)æltæ (sg. æ›v‘d ‖лæ›v‘dæ) ʤatʦʣʲʳчлmotʦʣʲʳчлʲʣὕatives‖brothers . Plural ending in *- - needs to be reconstructed already for the North Eastern Old Iranian dialects as it is attested in several Scytho-Sarmatian tribal names: ται, ται, Θ ται and τοι/ τοι/ ται. Apart from the innovated (and say unified) plural ending *- - there are marginally preserved old plural forms in Sogdian – these forms are preserved mainly in - and -stem inflection. In the -stem direct cases there is the ending -a (with allomorph (?) -ya), which often appears with animate substantives (e.g. ə ʦoʲʳʣʳ ). Some animate substantives and majority of -stem nouns have plural ending - (< originally probably an agglutination of abstract suffixes *- (n)-205 and *- -; cf. Wakh. - t). In the oblique cases of the -stems (and also masculine aka-stems) there appears a continuant of old genitive(-accusative) ending -ān(u), this ending can be used to express the oblique case of plural of animate nouns. As marginal and really archaic case can be considered dative-ablative plural of the -stems. In Sogdian there is attested the ending -y /-ē ъл in some toponyms: Sogd (ʾ) tmy h /F ъ – present in Falghar ⁽ʾt₍ c čъл oʤл члpʲʣʳʣntлV dí” in Mastchōʦ; other place-names terminating in -ēvй-ē”, - - can be found in a wider area on upper reaches of the Zarafshōn river in historical regions of Falghar and Mastchōʦ, e.g. , gí”, ē₎g ”, dgí”, Ḥ ždí”, P‘k í”, ŭ‘nglí”, Γ ›í”, , , and also in the Yaghnōϐ Valley: , Mʉ t ”, and most likely also Mar < *Mar 206 . Function of the original dative-ablative case changed, topoformant *-ē /*- was used to express locative function, e.g. Compare Slavic *-ьs(t) in a suffix *-ьstvo: OCS. ’›‘t›ьstvo, Cze. bratrstvo ϐʲotʦʣʲʦood луLIVSHITS – KHROMOV 1981, 425). 206 ζnлσaʲzōϐХлуandлʳomʣлotʦʣʲфлρā₍Х₎лdʧaὕʣϑtʳлandлaὕʳoлʧnлtʦʣл ρā₍Х₎лdʧaὕʣϑtлoʤлtʦʣлḲaghnōϐлσaὕὕʣyлthere is often recorded change -b(-) > -w(-)/-v(-)члϐutлʳuϑʦлϑʦanʥʣлʧʳлnotлattʣʳtʣdлʧnлtʦʣлρā₍Х₎лdʧaὕʣϑtʳлoʤлκaʳtchōʦлandлpaʲtʧaὕὕyл in the Falghar dialects. The form Mar ʦaʳл ϐʣʣnл pʲoϐaϐὕyл ʲʣanaὕyzʣdл уandл ρā₍Х₎ʧzʣd фл andл tʦʣnл ʣmʣʲʥʣdл ʲʣvʣʲʳaὕ л ϑʦanʥʣл *-v > -b, probably by analogy with some other YaghnōϐХл pὕaϑʣ-names: Yagh. ⁱ ›t u ×л ρ₍₎щл , Yagh. F‘›k u × Tjk. , Yagh. ŋ Yá (d)n u > Tjk. Ya . Indeed, the original form *Mar / th г ф( ) in Russian maps from the end of the 18 and beginning of *Mar f is indicated in Russian orthography th the 19 century. 205 ·138· Mar < *Mar < dat.-abl. ŋm‘›g‘ ’ ‘h : *marga- ʤoʲʣʳtчл mʣadoʷ i.e. in meadows / in forests . For other place names the etymology is not known, but it can be supposed that the ending *-ē /might have another function – it could have served as possessive, it is that in some place-names can be attested personal names of founders of such villages – e.g. Imbēf can be interpreted as a village founded by a man called *Imb-, i.e. could mean [the] †Im’s (??) (settlement) . In case of Fatmēv ʧtʳл mʣanʧnʥл ϑanл ϐʣл ʳuppoʳʣdл aʳл х уvʧὕὕaʥʣфл oʤл tʦʣл fiʲʳtуʳф л <л *fratamá ’ ‘h andл σōdʧʤл ϑanл mʣanл уvʧὕὕaʥʣфл oʤл tʦʣл ʷʧndуʳф л <л ŋu‘ʜąt‘ ’ ‘h. Place-names terminating ʧnл ʣtymoὕoʥʧϑaὕὕyл tʦʣл ʳamʣл ʣndʧnʥл ϑanл ϐʣл ʤoundл aὕʳoл onл ἵāmХʲ: vǰ , , , B , P ›⁾ⁱn in Tajik Shughnōn and Shākhdara; in Afghan Sheghnān , in Tajik Rushōn and probably also ŭučД⁽ and Γ in the Sarghulām Valley. See also οāshāʲvХл pὕuʲaὕл ʣndʧnʥл -(₍)Д” and inл τan₍Хл tʦʣʲʣл ϑanл ϐʣл ʳuppoʳʣdл pὕuʲaὕл ʣndʧnʥл -ev. The old dative-ablative ending can be found in oblique plural endings ʧnлκun₍ХлandлḲʧdghāл-ā” ‖л-əf, in WakhХл-əvчлʧnлπaʲХqōὕХл-(y)ef, and in Pashtōл- , Wa etsХл- . After the Stress III shift and operation of the Sogdian Rhythmic Law there emerged differentiation of the light and heavy ( -)stem endings. This change can be observed well mainly in -stem endings – according to position of stress there emerged two different declinations – light and heavy; in the other stems there have remained only the l ght endings, the heavy stems morphologically merged with the heavy -stems. Number of case endings was reduced, mainly in the heavy stems, where virtually remained only one ending207 – gen.-dat. sg. /-Хъчлendings of the direct cases was lost, the endings of the oblique cases merged with the original genitive-dative ending(s), the vocative endings of the heavy stems were taken over by analogy from the light stems. The inflectional system was later simplified, e.g. the archaic endings of masculine (and neuter) aka-stems are attested in Christian Sogdian manuscript C 2 (and also in the Ancient Letters), but in all other documents there is attested much simplified inflection (Table 42). Similarly the light -stem declination is preserved in majority of documents in the form developed from Old Iranian -stem inflection, but the Christian Sogdian manuscript C 5 shows a new innovated inflectional system in which oblique ending is agglutinated to a reflex of a thematic vowel. The Christian Sogdian manuscript C 5 presents agglutinative inflection in Questionable is the ending of the feminine h“‘v₍ ā-stems – ʧnл doϑumʣntʳл ʷʲʧttʣnл ʧnл tʦʣл BʲāʦmХл ʳϑʲʧptл tʦʣл ā-stem feminines light and heavy stems do terminate in the letter hē: -h (the light stems also end in -ʾh or just -ʾ). Question is how to interpret the terminal letter hē. There are several possibilities how to explain this orthography: 1) it is an archaic spelling of terminal vowel - in *all forms of nom. sg. of feminine ā-stems, 2) it is a spelling of word-final - adapted from Aramaic orthography, where in Aramaic words ending in - <-h> have been feminines; or 3) it is a combination of both above shown examples. Outcome is the state attested in documents in the Manichaean and Syriac scripts – form of feminine h“‘v₍ ā-stems with no ending. In the documents written in nthe Sogdian alphabet there has been pertained (archaic) spelling with the letter hē in feminine forms regardless whether the thematic ending remained preserved or whether it has been lost due to operation of the Rhythmic Law. The development of Sogdian nominative singular forms of the h“‘v₍ ā-stem feminines can be shown on following motʦʣʲ л şлἵʲotoπoʥdϑщл > after the Stress III shift *m mʾth) şл Cὕaʳʳʧϑaὕ л example: Ir. πoʥdʧan л mʾt mʾth). 207 ·139· Sogdian in a state similar to YaghnōϐХ, the only difference is a presence or absence of operation of the Sogdian Rhythmic Law (see Table 41). In the case of the plural ending - (for the light stems) or -t (for the heavy stems) there is good example of agglutination too – both masculine and feminine plural is declined as singular ā-stem feminine, but the inflectional endings are not added to stem but they follow the plural marker -t-, e.g. Sogd. rəmí pʣopὕʣ (m) : nom. pl. rəmd : obl. pl. rəmd₍ ; Sogd. wə tʲʣʣл(f) : nom. pl. wənd : obl. pl. wənd₍ ; Sogd. āp ʷatʣʲл (f) : nom. pl. āpt : obl. pl. ; Sogd. ēu demon (m) : nom. pl. ēut : obl. pl. utД. Inflectional system of the Sogdian heavy stems in principle does not differ from inflection in YaghnōϐХ: p ʷatʣʲ : pt : ; dēu demon : dēut : d ut (only in YaghnōϐХлoutϑomʣʳлoʤлtʦʣл -stems the terminal -a has been lengthened before -t and it changed to - : ʳon : : < :* -tah : ž t‘ʼa-taʼ ă). As can be seen from the previous lines and from the Tables 39, 41, 42, and 43, Sogdian nouns gradually changed from inflection towards agglutination. Inflection of adjectives is diachronically the same as noun inflection. In the Middle Iranian period also the adjective inflection was rebuilt. Such development is observable in Sogdian. Initially Sogdian adjective corresponded with its noun in gender, number and case. By simplifying of the inflectional system a new phenomenon emerged – so-called group inflection, where the bearer of the main grammatical information remained to be the noun, but adjective corresponded with it often just in number an later on it remained in form of nominative singular. The origin of the group inflection can be seen in the heavy stem endings, but it later spread to the light stems too. πuϑʦлϑʦanʥʣлʧʳлpʲoϐaϐὕyлoὕdʣʲлtʦanл aʥʥὕutʧnativʣ лʧnflection of the light stems as it is attested in the Christian Sogdian document C 5. The emergence of the group inflection caused that the adjectives became uninflected and they have been fossilized mostly in form of nominative singular masculine. This innovation corresponds with emergence of agglutination of substantives and it is comparable with the group inflection in agglutinative languages such as the Turkic languages208. In a reduction of adjective inflection probably for the longest period of time survived gender distinction, which is preserved for some adjectives in Pashtōчлτa etsХчлḲazghuὕāmХ and in the ShughnХ-RōshānХл ὕanʥuaʥʣʳ. In Pashtō-Wa etsХл tʦʣ adjectives are usually distinguished by different ending, ʧnлtʦʣлἵāmХʲлὕanʥuaʥʣʳлʤʣmʧnʧnʣлad₍ʣϑtives can be distinguished by results of ā- or i-Umlaut, e.g. Pasht. ʷʦʧtʣ ; Shugh. kut : kat ʳʦoʲt , › t : › t ʲʣd ; Bart. čȫ⁽ : č‘⁽ motὕʣy . In contemporary YazghulāmХл tʦʣʲʣл ʧʳл noл ʥʣndʣʲл dʧʳtʧnϑtʧonл oʤл ad₍ʣϑtives, but reflexes of gender are attested in several fossilized nominalised forms, e.g. č < čá⁾ ‘- ϐʧttʣʲ (adj. < adj. m.) ×лč‘ ʳaʥʣϐʲuʳʦ (< ŋčá⁾ ā- ϐʧttʣʲ (adj. f.) ф ; ÈDEL MAN 1987b, 384). In Sogdian gender distinction remained in forms of the light stems, e.g. ›- ʥoodл (f) л ×л ›-í ʥoodл (m) ;л oʲл ʧnл See Uzbek nom. sg. ‹ ›‘ s ϐὕaϑ₎лʷatʣʲ луʷʧtʦлtʦʣлʳamʣлmʣanʧnʥлaὕʳoлʧnлaὕὕлtʦʣлotʦʣʲлʣʸampὕʣʳфлśлʥʣnщлʳʥщл‹ ›‘ : āu āpt : āu tД; Yagh. u p : u : s nï : nom. pl. ‹ ›‘ s łar : gen. pl. ‹ ›‘ s łarnï ×лπoʥdщл āu āp : āu u pt : u ti. 208 ·140· endings of the aka-stems which end in - )209 in Sogdian, e.g. ᶤ ʷʦʧtʣл (f) л×лᶤ ʷʦʧtʣл (m) л<л*ʦuá t‘-ka-; ᶤ donʣл(f) л×лᶤ donʣл(m) л<лŋk t‘-ka-. In Sogdian there are two sets of comparative endings: -tər(-íй- ) < *-tara- for the light and heavy stems and -stər for -stems. Both endings may be used also for superlative forms. Special superlative forms are formed with ending *-tama-, occasionally accompanied by Sogdian ending - . (δκπл έьэřы-1296) Formation of comparatives and superlatives is analytic in YaghnōϐХчл ϐutл ϑaὕquʣdл ʤoʲmʳл ʷʧtʦл Tajik -tar < -*tara- may be found. Some forms with ending -star are quoted in the Yaghn ’Д T“⁾ts by Mikhail Stepanovich ANDREEV – Elena MikhaЧὕovnaлPESHCHEREVA (1957): moʲʣлʧnлʤʲont лoʲлsar ádd st‘› ʦʧʥʦʣʲ л< Perso-Arabic sar- ad(d) ϐoʲdʣʲ + Yagh. -star. (KHROMOV 1972, 20-эь;лλová₎лэыьычлээ5-226) II.2.2. Pronominal inflection Iranian pronominal inflection shows many similarities in development in many of the languages of the Eastern Iranian branch. Almost all languages preserve archaic system with forms just for the first and second person personal pronouns singular and plural, separate forms of the third person emerged only in a few Eastern Iranian languages, in majority of them they are expressed by demonstratives. Personal and demonstrative pronouns developed into three- or two-case system (See Tables 44 and 45 ʤoʲлtʦʣлἵāmХʲлὕanʥuaʥʣʳфщ All languages inherited triple deixis of the demonstrative pronouns, such system is preserved in majority of the Eastern Iranian languages, but in some of them the deictic system has been reduced into double deixis (e.g. in YaghnōϐХл oʲл ḲazghuὕāmХ, the tendency may be observed probably also in Sogdian). Enclitic forms the personal pronouns have been widely used as they were employed for personal endings in ergative construction – ʧnл ma₍oʲʧtyл oʤл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл tʦʣл ʣnϑὕʧtics are used no more there, they have merged with forms of copula. The North Eastern Iranian languages differ from the other Eastern Iranian by retention of archaic form of the second person pronoun: Sogd. ᶤ , Yagh. , Oss. s m‘⁾ ‖лsumax < 210 ŋ u má⁾‘m : ŋ ž‘m youл(gen. : nom.) . In the Eastern Iranian languages both the first and the second persons plural pronouns emerged from old accusative, in the Southern branch, probably > after the change ŋ m > m took place, the pronouns phonetically merged: 211 *(ə ⁾ am ʷʣ л×лŋ( > *(ə ⁾ am ʷʣ щлρʦʣлπoutʦʣʲnлϐʲanϑʦлʦadлtoлdʧfferentiate the first and the second person plural, ʳoл ʤoʲл tʦʣл ʳʣϑondл pʣʲʳonл tʦʣл South Eastern-Iranʧan л form has been augmented by prefix *tu-, *ta- taken from the second person singular: *ta⁾ am or *tu⁾ am – both etymologies can be considered correct, but also the etymologies do not tell whether such innovation of the form of the second person plural really γʣmʧnʧnʣл aka-ʳtʣm лad₍ʣϑtʧvʣʳлdʧʳtʧnʥuʧʳʦлlight and heavy stems. See Modern English you which is originally dative-accusative form of ye. 211 Certainly not from as it is claimed by some scholars (cf. GAUTHIOT – BENVENISTE 1929, 115), it in YaghnōϐХчл †⁾ mā⁾ ʧnл πoʥdʧanл andл ʧnл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл tʦʣл pʲoto-form would give something like should be based on ⁾ am so the forms of the first and second persons plural would not merge together. 209 210 ·141· Wakh. Ishk. Sangl. Yazgh. Munj. Yidgh. Shugh. οōsh. Bart. οāshrv. πaʲХqщ took place after the change ŋ m > *m (cf. MORGENSTIERNE 1929, 62). The innovation of the second person pronoun can be explained either as an aʲʣaὕлʤʣatuʲʣлуʣvʣnлϑauʳʣdлϐyлaлBuʲūshaʳ₎Хlike substrate language?) or as a contact with the Indo-Aryan languages (PAKHALINA 1976a). WakhХлʳʦoʷʳлtʦatлtʦʧʳлʧnnovatʣdлpʲonounлϑanлϐʣлoʤлanлʣaʲὕyлdatʣл originates from *təsa (ŋtusą / ŋt‘są < Middle IAr. *tusma < *tu ma-; PAKHALINA 1976a, 80) + Wakh dir. pl. suffix - t or obl. pl. -əv. Also WakhХлpʲonounлoʤлtʦʣлfirst person plural sak shows Indo-Aryan influence < *asma- (gen.; ibid.). Forms of the personal pronoun of the second person in Pashtōл , , Wa etsХл tās andл Ōʲmщл t s, t₍ s can be compared with WakhХ. Exception from the South Eastern Iranʧanл ὕanʥuaʥʣʳл aʲʣл ἵaʲāchХл and πa₎aл dʧaὕʣϑtʳśл ʧnл ἵaʲāchХл ⁽ comes «”›om Av[“st‘n] “ncl[ t c] v ж ⁽ th p“cul ‘› t›“‘tm“nt o” ⁽.» (MORGENSTIERNE 1929, 62); Khōtanʣʳʣлuhu (later uma, amaфл you лʷʦʧϑʦлʦaʳлϐʣʣnлʧnfluenced by Khōtщлmuhu ʷʣ луGERTSENBERG 1981, 269). wəz, (w)uz az(i) azə, azi az za zo, zə wuz az āz waz waz st 1 sg. dir obl. māž poss. žəчлž encl. -(ə)m mak mak mů(n) m уnф mən ni -ьm (-əm) -əm -(y)əm -əm tu tu, tə mən mun, mən mu mu mu mu(n) m (n) -um -um -um -(u)m -(y)am nd 2 sg. dir tu t tōʷ tow obl. tow, taw fak təfak tu poss. ti ti tō ti encl. -(ə)t -ьt -et -(i) -(i) -š sak m ʸ(ó) aməx, amax m č v(o) poss. səpo encl. -(ə)n tu tū tu tʷ tu tā tā tā taчлt ta to, ta -at -(y)ət (-t) -at -at -at -(a)t -(y)at -ay -(y)əš ? -(i) -(i) -i -ø -(y)i mʧčəf mox mōx māš māš māš māš maš m š mʧč moxi āmōx -on -mōn -an -(y)əmōn -ām -am -an -(a)n -(y)an təmox mōf tama tama tamāš tamāš tamaš rd 3 sg. encl. st 1 pl. dir obl. ? nd 2 pl. dir obl. poss. encl. (y)ʧšt sav t m ʸ təməx t m ʸ( v) təməx(əf) t m ʸ təməx təmoxi āmōf -(ə)v -ьv ? -əf -(y)əʤōm (-f) -ēt -af -at, -af -(a)f -(y)af -(ə)v -on -šōn -an -(y)əšōn ? -ēn -an -an, -af -(a)f -(y)af rd 3 pl. encl. Table 44 Personal pronouns of the first and second persons in the ἵāmХʲлὕanʥuaʥʣʳщ Enclitic forms given in italics are used as copula. Another possible archaism can be seen in WakhХл– oblique case of the first person singular pronoun mā can originate 1) either from Ir. dat. ŋmáʣ ‘m < IIr. ŋmá ʰ ‘m < Ide. ŋm“gʰ -om, 2) or it is an Indo-Aryan loan ŋmá ʰ ‘m. If Wakh. mā is Iranian origin, it should be rather archaic feature, even more archaic then Avestan m‘ⁱ’ ā, m‘ⁱ’ which is an innovation (cf. VAσοἴςŠβἰ 2007, 43), or it is an early loan from Indo-Aryan *má ʰ ‘m, Ved. máh₍‘m (see ·142· oϐὕʧquʣл ʤoʲmʳл ʧnл εʧndХл andл ςʲdūл mujчл κaʲāṭhХл maʒ; PAKHALINA 1976a, 83). ρat yana Nikolaevna Pakhalina rather accepts the Indo-Aryan hypothesis, which can also better explain WakhХл poʳʳʣʳʳive forms ə, < *m an < *mə án < *ma án (ibid., 82), other clue for the Indo-Aryan origin can be ignorance of i-Umlaut, i.e. there is no †mǟ . case Shugh. οōsh. Bart. οāshrv. πaʲХqщ Wakh. Ishk. Yazgh. Munj. I. deixis II. deixis III. deixis sg. pl. sg. pl. sg. pl. dir. yam mā (y)id dā yuлълyā ʷā obl. mi / mam mēv di / dam dēv wi / wam ʷēv dir. (y)im mā (y)id dā yā ʷā obl. may / mum muf day / dum duf way / (w)um wuf dir. yim mā yid dā yā ʷā obl. mХлълmʧm mif dХлълdʧm dif ʷХлълum uf dir. yim mā yid dā yā ʷā obl. mi / mam maf di / dam daf wi / wam waf dir. yam yamчлуmo ф yad yad, уdo ф obl. mi / mem mef di / dem dir. obl. yəmʧšуtф yəm yəməv def yətʧšуtф yət yətəv y y члуʷo ф wi / wem wef yáуʷфʧšуtф yow, yaw yáуʷə)v dir. amуíф amónуonф adуíф adóndуonф aʷуíф aʷóndуonф obl. man mán vуoф dan dán vуoф wan ʷán vуoф dir. – obl. dir. obl. du yu ma māy ya yāy wa ʷāy mānлълmāy māʤ yānлълyāy yāʤ ʷānлълʷāy ʷāʤ Table 45 Inflection of demonstrativʣлpʲonounʳлʧnлtʦʣлἵāmХʲлὕanʥuaʥʣʳщ Sogdian and YaghnōϐХл pʲonomʧnaὕл ʧnflection is very similar one to each other, the main differences were caused by operation of the Rhythmic Law in Sogdian, the system is also ϑompaʲaϐὕʣлtoлtʦʣлἵāmХʲлὕanʥuaʥʣʳщлζnлϐotʦлὕanʥuaʥʣʳлtʦʣлfirst and second persons plural were based on forms of accusative and are not inflected. The personal pronouns for the first and second person singular are both inflected in direct and oblique cases, in YaghnōϐХлtʦʣлdʧʲʣϑtлϑaʳʣл form of the first person gave place to its oblique form, it has been attested once ϐyл Émʧὕʣл Benveniste (GAUTHIOT – BENVENISTE 1929, p. 108-109), but all other sources have just one form for both direct and oblique case: man. In YaghnōϐХлtʦʣлoϐὕʧquʣлʤoʲmлʧnfiltrated the direct case probably under Tajik influence (Pers. Tjk. man ζ флandлmaybe also some impact of Turkic can be suggested (cf. Uzbek mėn, colloq. mȧn, Kyrgyz men etc.). There can be seen a tendency to develop distinct inflectional forms for all personal pronouns both in Sogdian and in YaghnōϐХл – by analogy innovated forms of oblique can be formed from the original personal pronouns by adding a heavy stem л oϐὕʧquʣл ʣndʧnʥл уϑʤщл ρaϐὕʣʳл 46 and 47). *Proto-Sogdic has inherited pronominal system without independent forms of the third person personal pronouns – their function has been fulfilled by demonstratives. Such pattern continued in Sogdian and still goes on in YaghnōϐХщ ·143· The Iranian triple deictic system has been inherited from the Indo-European proto-language. Demonstrative pronouns distinguished I., II. and III. deixis (of also ich-, du- and er-deixis or hic-, iste- and ille-deixis), the inflectional pattern has been based on two suppletive forms – nominative in ŋ ‘-/‘ ‘- (I.), *‘ ‘- (II.) and *(a)h‘u- (III.) and oblique stem in *ima-, ŋ‘ t‘- and ŋ‘u‘- (cf. Tables 45, 46 and 47). In YaghnōϐХл tʦʣл oʲʧʥʧnaὕл nʣaʲл I. deixis disappeared so there is only double deictic system (cf. the same development in YazghuὕāmХфщлζnл Sogdian complete system is attested, but according to preserved forms can be judged that forms of the II. deixis started to disappear or were of lesser importance. sg. st nom. nd 1 pers. 2 pers. m əzú t( фú yu ᶤmú acc. gen.dat. f ᶤ II. deixis III. deixis m f m ᶤšú ᶤšā ᶤtú ᶤ st f təʷá encl. acc. -m -ʤуХф encl. gen. -mХ ᶤmén 2 pers. ōчл ʷú ʷénуēф ᶤ ᶤ nd 1 pers. māʸ məná instr.abl. I. deixis ᶤ pl. I. deixis II. deixis yu ᶤšā ᶤmú ᶤ 212 213 -mən -fən, -tən ən ən x III. deixis 214 loc. -šu -tуХф -šуХф -šən Table 46 Inflectional system of personal and demonstrative pronouns in Sogdian. sg. st pl. nd st 1 2 near far 1 pers. pers. deixis deixis pers. 215 tu Д ax dir. (az) m ⁾216 218 219 á⁽ obl. man t‘u -(ⁱ) -(ⁱ)m ⁾ encl. -(ⁱ)m -(ⁱ)t nd 2 pers. near far deixis deixis á⁾t t áut t 217 - nt Table 47 Inflectional system of personal and demonstrative pronouns in YaghnōϐХ. 212 Occasionally nominative , oblique . Occasionally nominative ᶤ , oblique ᶤ . 214 Cf. Pahl. ‘⁽ē ‘n tʦʣy щ 215 The form az ʧʳл quotʣdл onὕyл ϐyл Émʧὕʣл Bʣnvʣnʧʳtʣл ʧnл ʦʧʳ Essai de grammaire Sogdienne, Deuxieme partie, Morphologie, syntaxe et glossaire (GAUTHIOT – BENVENISTE 1929, 108-109); in all other sources there appears only st single form of 1 person singular man for both cases. 216 Occasionally analogically formed oblique can appear. 217 Occasionally by analogy formed oblique . 218 In colloquial speech appears analogically formed oblique mán . 219 The oblique form can per analogiam appear as tá⁽ . 213 ·144· *Proto-Sogdic pronouns started to develop independent pronominal system of inflectional endings with rich suppletive system. Pronominal inflection developed differently in both languages. In Sogdian it can be seen in inflectional forms of the demonstrative pronouns (cf. Table 46) and also on an adjectivʣл aὕὕ л– ⁽ spú (Table 47) andлaлnumʣʲaὕл onʣ л- u (Table 50). YaghnōϐХлdʣvʣὕopʣdлʧndʣpʣndʣntлʳyʳtʣmлϐaʳʣdлonлʣndʧnʥл-tit for direct case and -titi for oblique (originating in reduplication of the plural ending *- ; KHROMOV 1987, 674). Such ending can be added to interrogative pronouns (Table 49). The pronominal plural ending -tit(i) can be also added to numerals to express number of people, e.g. Yagh. ú⁾ -tit(-i) (of) six indivʧduaὕʳ чл s‘›á₍t t ‖лtⁱ›á₍t t tʦʲʣʣлʧndivʧduaὕʳ члϑʤщлἵʣʲʳщл ‘ - ʳʧʸлʧndivʧduaὕʳ члhaft- ʳʣvʣnлʧndivʧduaὕʳ щ sg. m pl. f nom. ʷʧʳpí ʷʧʳpā ʷʧʳpé acc. ʷʧʳpú gen.-dat. wispəné ʷʧʳp šən instr.-abl. wispəná Table 48 Inflection of wisp- aὕὕ щ The demonstratives can be both in YaghnōϐХл andл ʧn Sogdian extended by prefixed or suffixed particle *nah, - -) – in YaghnōϐХлtʦʣлpaʲtʧϑὕʣлʧʳлpʲoϑὕʧtʧϑʳлandлʧtлϑanлϐʣлuʳʣdлʷʧtʦл various forms derived from the demonstratives; in Sogdian the particle is enclitic: - , -nəx; e.g. xwny, xwnx -nəʸ члḲagh. nah-á⁾ (III. deixis) onʣ члϑʤщлḲagh. nah-id- Sogd. rec.sg. obl.sg. rec.pl. obl.pl. – Yagh. who kax ₎aʧчл₎áyʧ ₎áʸtʧt ₎áytʧtʧ Sogd. č – Yagh. what čō čōʧ – Sogd. Yagh. which kə – k m k mi k mtit k mtiti Sogd. Yagh. how much čāʤ – Sogd. čōʤ fi Yagh. where ₎ū ₎ūʧ – – Table 49 Iterrogative pronouns in Sogdian and YaghnōϐХщ Sogdian has also developed a definite article – it was formally the same as the demonstrative pronouns of III. deixis, but in plural all forms of the definite article were inflected as feminine singular. The definite article has been widely used during the development of the Sogdian languages, but in late Christian texts it is inflected only in two cases (dir. < nom., obl. < acc.) and its forms gradually merged. In really late texts there can be no definite article. As it is attested in several Sogdian documents of Zhetisu, there were probably more ways to express the definite article in *Proto-Sogdian, the dialect of Zhetisu shows the definite article based on extended form of the demonstrative pronoun of the I. deixis ŋ‘ ‘- - -. In YaghnōϐХлtʦʣʲʣлʧʳл no definite article, according to known history of the YaghnōϐХл ὕanʥuaʥʣл ʧtл ϑannotл ϐʣл ₍udʥʣdл whether there have been also a definite article that disappeared during the development of the language or if there has been no definite article in *Proto-YaghnōϐХщлζnлtʦʣлἵāmХʲлὕanʥuaʥʣʳлtʦʣл ·145· demonstrative pronouns serve also as the definite article, but they are used also syntactically and grammatically to express gender or subject of a clause. *Proto-Sogdic enclitic pronouns originate from enclitics inherited from *Proto-(Indo-)Iranian, in *Proto-Sogdian and *Proto-YaghnōϐХлtʦʣлʣnϑὕʧtʧϑʳлʷʣʲʣлʳʧmpὕʧfied. In Sogdian the enclitic pronouns distinguished accusative and genitive forms; in YaghnōϐХл tʦʣл enclitics have just one form (see Tables 46, 47). YaghnōϐХлpὕuʲaὕлʣnϑὕʧtʧϑʳлʦavʣлϐʣʣnлʧnnovatʣdл– enclitic pronoun of the first person plural has been taken from original Iranian accusative (i.e. YaghnōϐХлdʧʲʣϑt-oblique). The inherited forms of enclitics of the second person plural were lost in YaghnōϐХл andл ʷʣʲʣл ʲʣpὕaϑʣdл ϐyл ʤoʲmʳл oʤл tʦʣл tʦʧʲdл pʣʲʳonщл ζnʦerited YaghnōϐХл ʣnϑὕʧtʧϑл pronoun of the second and third persons plural - nt originates from the *Proto-Sogdic (or *Proto-YaghnōϐХфлtʦʧʲdлpʣʲʳonлʣnϑὕʧtʧϑл*- an extended by plural ending *- (cf. KHROMOV 1987, 675). See also merger of the forms of enclitic forms of copula (< enclitic pronouns) of the ʳʣϑondлandлtʦʧʲdлpʣʲʳonʳлʧnлBaʲtanʥХчлοāʳʦāʲvХлandлπaʲХqōὕХлуρaϐὕʣляяфщ II.2.3. Numeral inflection *Proto-Iranian numerals were inflected similarly as nouns. The numeral inflection was present also in *Proto-Sogdic, but the inflectional system changed during later development. In Sogdian there are attested inflectional forms just for numerals one and two (Table 50) – for the numeral one both cases were formed analogically (i.e. accusative by adding oblique case ending, genitive-dative ending is taken from pronominal inflʣϑtʧonфчл tʦʣл numʣʲaὕл two л pʲʣʳʣʲvʣʳл inherited genitive ending. Both Sogdian numerals one and two distinguished masculine and feminine forms (feminine form ʾywh /| ъл onʣ лʧʳлattʣʳtʣdлonὕyлʧnлtʦʣлπoʥdʧanлdoϑumʣntʳлʤʲomл the Mount Mugh; cf. BOGOLYUBOV – SMIRNOVA 1963, 21). Some forms of numerals can have old genitive ending in -nu (YOSHIDA 2009a, 295). on“ m t⁽o f m nom. | u | acc. | ʷХ gen.-dat. | wən221 220 f ʷáчл ú ɨ nú Table 50 Inflection of the numerals on“ and t⁽o in Sogdian. YaghnōϐХл numʣʲaὕʳл do not distinguish gender and they are normally uninflected, but in occasional cases they can be inflected the same way as nouns, the numerals can even take plural ending -t when necessary. Inherited YaghnōϐХл numʣʲaὕʳл ʤrom two to ten can also take pronominal plural endings to express number of people (see chapter II.2.2. above). The YaghnōϐХлὕanʥuaʥʣлʦaʳлtʷoлʳʣtʳлoʤлnumʣʲaὕʳл – inherited and borrowed. Inherited are only the numerals from one to ten (see numerals presented in lexical part of the presented thesis, 220 221 γʣmʧnʧnʣлʤoʲmлoʤлtʦʣлnumʣʲaὕл onʣ лʧʳлattʣʳtʣdлonὕyлʧnлtʦʣлκountлκugh documents. In Christian Sogdian oblique ywy /| ʷХ/. ·146· chapter III.2.), the borrowed numerals are taken from the ZarafshōnлTajik dialects. The Tajik numerals are used to count entities of more than ten items, but with words of Tajik origin (as considered by the YaghnōϐХʳчлʧщʣщлaὕʳoлAʲaϐʧϑлandъoʲлUzbek loans) Tajik numerals are used even for entities less than ten. When counting, entities of more than one item are not presented in their plural form, but numerative form is used. Sogdian numerative originates from *Proto-Sogdic (or Iranian) dual (SIMS-WILLIAMS 1979; cf. table 39). In YaghnōϐХлtʦʣлϑountʣdлʣntʧtʧʣʳлaʲʣлʧnлoϐὕʧquʣлʳʧnʥuὕaʲлуʧtл is possible, that the oblique ending comes from (oblique) dual, but due to formal similarity of continuants of both oblique singular and oblique dual > numerative it can be only difficult to judge222). In other (Modern) Eastern Iranian languages counted entities often appear in singular – this can be interpreted as development influenced by a development of group inflection, interpretation as influence of Persian or Turkic seems to be less probable in this case. YaghnōϐХлʦaʳлὕoʳtлʧnʦʣʲʧtʣdлnumʣʲaὕʳлʤʲomлʣὕʣvʣnлupлtoлtʦʣл ʧnfinʧty л– those numerals have been replaced by Tajik forms. Aὕ ϐʣʲtлἱʣonʧdovʧch KHROMOV (1987, 671-672) notes, that elder YaghnōϐХʳлуʧщʣщлʧnлtʦʣлtʧmʣлoʤлʦʧʳлfield-work in the YaghnōϐлvaὕὕʣyлʧnлtʦʣлfiʲʳtлʦaὕʤлoʤлtʦʣльŚ6ы ʳфл counted in vigesimal system (vigesimal system of counting is attested also in the Zarafshōnл Tajik dʧaὕʣϑtʳлoʲлʧnлʳomʣлoʤлtʦʣлἵāmХʲлὕanʥuaʥʣʳфщлλoʷadayʳлtʦʣлvʧʥʣʳʧmaὕлʳyʳtʣmлʧʳлnotлuʳʣdлʧnл YaghnōϐХчл ϐutл ʳomʣл ʳpʣakers use synthetic counting using inherited YaghnōϐХл numʣʲaὕʳчл ʣщʥщл d‘s Д ʣὕʣvʣn л уoʲл ϐoʲʲoʷʣdл ), u⁾ d‘s ʳʧʸty л уvʧʥʣʳʧmaὕл s‘›á₍ ‖ tⁱ›á₍ ’Дst, borrowed ‘st). Sogdian numerals continue from Iranian numerals, but units precede decades, e.g. ə d-⁽Дst tʷʣntyл ʳʣvʣnл уὕʧtʣʲaὕὕyл s“v“n-t⁽“nt₍ ф чл numϐʣʲʳл ϑὕoʳʣл toл aл ʦʧʥʦʣʲл dʣϑadʣл ϑanл ϐʣл ʣʸpʲʣʳʳʣdл ϐyл subtraction, e.g. u káṁ’Д páṁǰās ʤoʲtyлnʧnʣлуὕʧtʣʲaὕὕyл on“ l“ss fi”t₍ ф луYOSHIDA 2009a, 295). Distributive numerals in Sogdian and YaghnōϐХл ʦavʣл ϑompaʲaϐὕʣл ʣndʧnʥśл πoʥdщл -kД 223 , Yagh. -ki. In *Proto-Sogdic there have been archaic forms of ordinal number first , second and third , ordinal numerals higher than four were formed by addition of endings. Such system has been preserved in Sogdian 224 , where ordinal numerals beginning from four were formed by adding an ending -əm(i) ~ -am(i) or -mДk. YaghnōϐХлuʳʣʳлoʲdʧnaὕлnumʣʲaὕʳлϐoʲʲoʷʣdлʤʲomлTajik (and in case of the ordinal first also Arabic form can be used), occasionally ordinals can be formed from YaghnōϐХл numʣʲaὕʳл ʷʧtʦл Tajik ending -(y)um (this Tajik ending is of the same origin as Sogdian -əm(i) ~ -am(i)). II.2.4. Verbal inflection Sogdian preserved complex conjugation system which in the active voice continues from the Old Iranian pattern, but in the middle voice there is attested conjugation only for indicative present See also comparable Ossetic ending - ‖л -i used for counting entities of more than one item which probably comes from Iranian genitive ending (ISAEV 1987, 593). 223 In Sogdian also -k‘nkД ~ -kaṁgД. 224 Iranian *fra-tám‘- fi ʾ tm-y, (ʾ)p›tm-y ʾftm-y fṯm(ʾ) / ftəmíъфл ʧʳл pʲʣʳʣʲvʣdл ʧnл ḲaghnōϐХл f ] ‖лf ], ] dayлaʤtʣʲлtomoʲʲoʷ л<лŋ”›‘tā-má ϑā-, ŋ”›‘-t‘m‘-má ϑ -. 222 ·147· and imperfect. YaghnōϐХлϑon₍uʥatʧonлaὕʳoлϑontʧnuʣʳлʤʲomлtʦʣлἴὕdлIranian pattern, but there has been completely lost the middle voice and also optative present. Moreover both languages lost Iranian indicative perfect. The endings have undergone several changes in both Sogdian and YaghnōϐХл– *Proto-πoʥdʧϑлvʣʲϐaὕлʳtʣmʳлʦavʣлϐʣʣnлaὕὕл tʦʣmatʧzʣd лandлtʦʣлvʣʲϐaὕлʣndʧnʥʳлʷʣʲʣл based on Iranian thematic endings. The Old Iranian endings changed a little bit in *ProtoSogdic, the main change can be seen in spread of ϑ to all forms of the second person plural. In *Proto-Sogdic there were two sets of endings of the third person plural – in the indicative mood there has been used either ending in *-ant- or in *-ā›- < ŋ-›-. The *-ant- forms have been preserved in Sogdian, in YaghnōϐХл tʦʣл ʣndʧnʥʳл aʲʣл ϐaʳʣdл onл *-ā›- (such endings are comparable to Khʷāʲʣzmʧanчл ʳʧmʧὕaʲл *-ā›- endings can be found in Khōtanʣʳʣ 225 ; and in Avestan 226 ) originating in endings of the third person plural of the lost forms of perfect indicative. In YaghnōϐХл tʦʣʲʣл ʲʣmaʧn preserved transformed forms of perfect which continue from endings of peripheral Indo-European middle voice perfect: primary ending *-(o)ror, secondary ending *-(o)ro (cf. BζČἴσπἰÝ 2012, 109-111). Sogdian present and subjunctive forms were contaminated by causative ŋ-‘ a- endings in the first person plural (see also Bactrian endings influenced by ŋ-‘ ‘-causative, such feature links Bactrian with development observed in Middle Persian), there may also be observed tendency to differentiate present indicative ending from other tenses in Christian Sogdian, where the ending of the second pers. pl. is -t(a) in present indicative, and in all other tenses and moods there remained *-ϑ-. In YaghnōϐХл tʦʣл optative mood has been lost, or better: optative has merged with imperfect – in YaghnōϐХл dialect there have up today survived both optative and imperfect endings in forms of the first person plural: in the Eastern dialect the ending -Дm continues from optative ŋ-‘ m‘, in the Western dialect there continues imperfect ending *-ām‘ > - m. (See Table 51) Indicative present *Ir. 1st sg. -ām Sogd. Active Yagh. *Ir. -ám - m -‘ - nd -ahi -у фл‖л-ē -ø -‘h‘ rd -ati -tл‖л-tí -t -‘t‘ 2 sg. 3 sg. st 1 pl. nd 2 pl. rd 3 pl. 227 228 -ām‘h -ēm -aϑa - уaфл‖л- á -anti -Хm 229 -aṁd - -‘m‘d‘ -ʳл‖л-t -‘du‘ 230 -‘nt‘ -ōʲ 225 Sogd. Middle Indicative present middle voice -ā›“ < *-ā›‘ , subjunctive present active voice -ā›o < *-ā›ām. Perfect indicative active voice -arə < -ar, middle voice -are < ŋ-‘›‘ . 227 st From causative ŋ-‘ ‘-conjugation < ŋ-‘ ‘m‘h; or from optative ŋ-‘ m‘. In the Ancient Letters there is attested 1 plural ending -ʾymn /-ēmən/, cf. Khōtщл-amne. 228 From causative ŋ-‘ ‘-conjugation < ŋ-‘ ‘m‘h; or from optative ŋ-‘ m‘. 229 In Christian Sogdian often -t(a) ‖л-tá. 230 rd From perfect indicative active voice ŋ-›( ) > *-ā› ; cf. Khʷāʲщл-ā› (3 pers. sg. present indicative & subjunctive). 226 ·148· Subjunctive present Indicative perfect *Ir. Optative present Yagh. -a 1st sg. nd 2 sg. -ϑa, -ta -a 3rd sg. 1st pl. 2nd pl. 3rd pl. 1st sg. 2nd sg. 3rd sg. st 1 pl. 2nd pl. rd 3 pl. st 1 sg. -ān -āh -āt ? -āϑa -ānt -‘ -‘ t st -‘ m‘ nd -‘ t‘ rd -‘ ‘nt 3 sg. 1 pl. 2 pl. st 1 sg. -am -m‘d‘ -du‘ -›‘ 231 - -ā -āt 232 -ēm - a -aṁd -Хm -ʳл‖л-t -ant -ē -‘ t‘ -‘ m‘d -ēm -ē -‘ du‘m 235 -ēnt 237 -уuфл‖л-ú -øл‖л-á 2 pl. rd 3 pl. -ām‘ -ata -ant -‘ ‘nt‘ 236 -at nd -ānt‘ -‘ ‘ -уʧфл‖л-í 1 pl. -ām‘d‘ -ādu‘ -‘ ‘ 234 -ah st -āt‘ 233 rd 3 sg. -ā( ‘) -āh‘ - m -ø -ōt nd 2 sg. Sogd. Middle ? -‘ -‘ (a)m rd 2 sg. *Ir. -‘ -ma ? -›( ) nd 3 pl. Indicative imperfect Sogd. Active 238 -im - 239 -ø 240 -ēm - уaфл‖л- á ‖л- m 241 -Хm -ʳ л‖л-t 243 -ant -ōʲ 231 242 -‘ -tu -aha -ti -ata -tуaфл‖л-tá -ām‘d -‘du‘m -anta In Christian Sogdian -ām. From causative ŋ-‘ ‘-conjugation < ŋ-‘ ‘m‘h; or from optative ŋ-‘ m‘. 233 From causative ŋ-‘ ‘-conjugation < ŋ-‘ ‘m‘h; or from optative ŋ-‘ m‘. 234 In the Mugh documents for one of persons also ending -₍ʾ /-ya/, the second person also -₍ /-ēšъщлἵʲoϐaϐὕyлʤʲomл the middle forms. 235 Mugh documents -y y /-ē i/. 236 In the Ancient Letters -ʾ₍ʾnt /- y ṁd/, in one Buddhist document -₍ʾnt /-(ə)y ṁd/. 237 Also used as injunctive and irrealis. 238 From optative ŋ-‘ (a)m (KHROMOV 1987, 681). 239 Either from imperfect *-ah or from optative ŋ-‘ (KHROMOV 1987, 681). 240 Also -ēmu (Mugh documents) or -ēmən < optative ŋ-‘ m‘ (?). 241 From optative ŋ-‘ m‘ ? (KHROMOV 1987, 681). 242 From optative ŋ-‘ ϑa (KHROMOV 1987, 681) influenced by present indicative/subjunctive; with metathesis ŋ-‘ ϑa > *-ēϑ > -ϑ > -s ‖л-t . 243 rd From perfect indicative active voice ŋ-›( ) > *-ā›; cf. Khʷāʲщл-ā›‘ (3 pers. sg. imperfect). 232 ·149· Imperative present *Ir. Sogd. Active Yagh. *Ir. Sogd. Middle -ø -ax a 1st sg. 2nd sg. -a -уaфл‖л-á rd -‘tām 3 sg. st 1 pl. 2nd pl. rd 3 pl. -ata - уaфл‖л- á -antu -ʳл‖л-t -‘du‘m -‘ntām Table 51 Overview of Old Iranian thematic conjugation and its development in Sogdian and YaghnōϐХщ *Proto-(Eastern-)Iranian verbal stem system has been simplified in *Proto-Sogdic, there emerged new conjugation system based on the present augmented or un-augmented stem, present and past participle and infinitive stem. The difference between thematic and athematic ʳtʣmʳлʦaʳлϐʣʣnлὕoʳtлandлaὕὕлvʣʲϐʳлʷʣʲʣлʤoʲmʣdлaʳл tʦʣmatʧϑ щлρʦʣлdʧfference between individual verbal stems gradually merged and the stem system has become quite regular, there are only several irregular verbs both in Sogdian and in YaghnōϐХщ The main difference between Sogdian and YaghnōϐХл ʧʳл dʧfferent treatment of augment in forms of imperfect tense. In Sogdian the original augment has been lost in non-prefixed verbs and remained only as so-called internal augment in between verbal prefix and stem. In YaghnōϐХл augment is preserved in all positions, but there is no internal augment, in the contemporary language augment of prefixed verbs is placed by analogy with non-prefixed verbs before the prefix as if the prefix was integral part of a verbal stem (see also chapter II.1.8.)244. According to development of stress in *(Post-)Proto-Sogdic it is probable, that non- nt“›n‘l augment should have been lost both in *Proto-Sogdian and *Proto-YaghnōϐХчл ϐutл pʲoϐaϐὕyл duʣл toл merger of optative and imperfect endings in *Proto-YaghnōϐХлandлtʦʣʧʲлformal similarity with endings of present indicative (cf. ISKHAKOV 1977, 30-31) the augment possibly acquired a secondary stress and thus was not lost due to operation of stress changes (on the other hand later in Christian Sogdian the imperfect has been gradually replaced by periphrastic perfect). The survival of the augment in YaghnōϐХлуʲʣʥaʲdὕʣʳʳлoʤлʧtʳлϑʦanʥʣ by analogy) is a striking archaism within all modern Indo-European languages. Augment is peripherally preserved in Modern Greek – only accented augment is preserved, but it disappeared in unaccented positions: MGre. έ л ζ ὕooʳʣnʣd члø ύ л ʷʣ ὕooʳʣnʣd л×лδʲʣщл ἔ члἐ ύ луϑʤщлSOPHRONIOU 1962, 79). According to R. L. Turner there are some traces of augment also in Dardic ἰʦoʷāʲлandл ἰaὕā a (TURNER 1927, 538-541). Both in *Proto-Sogdian and in *Proto-YaghnōϐХлʣmʣʲʥʣdлʳʣϑondaʲyлʣndʧnʥʳлtʦatлmayлʦavʣл been used with verbs to modify their syntactic or temporal meaning. In Sogdian there are attested several compound formations from present stem – durative in -(ᵊ)skun (see QAοФB 1965, 167-169), future in -kām (ibid., 174), or preterite in -ā₎ (ibid, 179-180). In YaghnōϐХлtʦʣʲʣлʧʳлaл 244 See also Old Persian a=pari-ā₍- toлϐʣʦave (augmented stem) лʷʧtʦлauʥmʣntлpʲʣϑʣdʧnʥлpʲʣʤʧʸлуSἰηÆοσØ 2005, 50). ·150· *durative suffix - t -ʾ tn in Vessantara Jāta₎aфщлαuʲative suffixes further developed in Christian Sogdian and in YaghnōϐХ, where present durative replaced present indicative (QAοФB 1965, 168). In YaghnōϐХл tʦʣл oʲʧʥʧnaὕл duʲative suffix - t (< *hi(-) t‘- < ŋstā- to ʳtand фл was agglutinated to personal endings, and some forms have changed: ŋ -tŌ t -č (t), ŋ - ›Ō t > - (₍) t, ŋ - Ō t > - t; ŋ -øŌ t > - t. The suffix - t is agglutinated also with endings of imperfect tense. The original forms non-suffixed of indicative present and imperfect tense change their meaning: non- t present serves as a so-ϑaὕὕʣdл dʣpʣndʣntлpaʲadʧʥm 245 and non- t imperfect is used as simple past (simple perfect) tense. Infinitive developed different forms in Sogdian and in YaghnōϐХщл ζnл πoʥdʧanл pʲʣʳʣntл infinitive distinguishes light and heavy stems: the light stems have ending -y(y) in nominative and oblique (but in Christian, Buddhist Sogdian and Sogdian in the Sogdian script also abl. -ʾ and acc. -(ʾ)⁽ уδκπл έŚы5-913), the heavy stems have no ending in nominative and -y in oblique уδκπлέŚы5члŚья-921). Past infinitive has ending -y (or -ʾ) in the light stems and no ending (or -y) in the heavy stems уδκπлέŚээ-934). YaghnōϐХлʦaʳлtʷoлʤoʲmʳлoʤлʧnfinitive – short infinitive (i.e. equal to verbal stem) and infinitives in -ak (cf. infinitive endings in other Iranian languages: Ishk. - k; Sangl. - k, -uk, Wakh. -ak, - k;лŌʲmuṛ. -akчлἵaʲāch. -o;лBaὕōch. -ag). st 1 sg. nd -(y)əm -уyфām -(y)ət -d/-t/-ø -уyфām -уyфām nd -уyфāʤ -уyфāʤ 1 pl. -(y)a -ø nd -əf -ef -ōʤ rd -et -et -ōt st -ən -əm nd -Х -i rd -ō -ø st -əm nd -əf rd - st -in -at nd -ay -(ay) rd -d/-t -an st -əm nd -it rd -an 1 sg. - t 2 sg. -on - v - v 2 pl. rd -on st -əm 3 pl. 1 sg. nd 2 sg. rd -(i) 246 -(i) -ən -ən nd -əv 1 pl. 2 pl. rd 3 pl. -ən 1 pl. 3 pl. 1 sg. -ət -d 3 sg. 2 pl. -on -əm st 3 sg. 3 pl. πanʥὕēchХ -on -əv -it -ø - m nd -ət -ōm -i 1 pl. -ë -ø 2 pl. -(i) -ōm -em - m -u -əm -am nd st -em -d/-t 1 pl. -уyфāt 3 sg. itr. st 3 sg. -уyфāt rd tr. rd 2 sg. st 1 sg. past tense nd rd 2 sg. present tense st 1 sg. -уyфāy st 3 sg. 3 pl. Ish₎āshmХ -(y)əy person itr. rd 2 pl. WakhХ -(y)əm past tense tr. YazghuὕāmХ κun₍Х 2 sg. present tense Yidghā person 2 sg. 3 sg. 1 pl. 2 pl. 3 pl. - -əf Dependent paʲadʧʥm is a characteristic feature of YaghnōϐХлʳyntaʸл– dependent forms are used after another verb in sentences like Yagh ‘⁾ ǰá⁾-t- t s‘ á› tⁱ›á₍ dám Д l‘ᴥ ⁱ -t-ø, n n tⁱ -t-øж č ₍ tⁱ -t-ø ʦʣлʷa₎ʣʳлupл and in the morning [he] give[s] dish to three persons, [he] give[s] bread (and) [heϊлʥʧvʣψʳϊлtʣa л(cf. KHROMOV 1972, 42). 246 Ending -i appears only in Western WakhХщ 245 ·151· itr. -um -um -i -at rd 3 sg. st 1 pl. -t/-d -i -ø -ām -ām nd -ēt -ēt rd -ēn -ēn st -um -um 3 pl. 1 sg. nd 2 sg. -t/-d st -an -am nd -at/-af -af 1 pl. 2 pl. -(i) 247 -ø rd -an st -am -am nd -ø -at rd -t/-d st -an 3 pl. 1 sg. 2 sg. 3 sg. 1 pl. nd -it rd -in 2 pl. 3 pl. -af/-an -i past tense -um itr. -um nd -i -at rd -t/-d 1 sg. 2 sg. 3 sg. tr. -i -ø st -am nd -at/-af -af rd -an -an st -um -um nd -ø -at rd -t/-d -ø st -an -an nd -at/-af rd -an 1 pl. 2 pl. 3 pl. 2 sg. 3 sg. 1 pl. 2 pl. -an present tense st 1 sg. -at -(i) rd 3 sg. person οōshānХлс KhūʤХ ShughnХ tr. nd 2 sg. 2 pl. BartanʥХ past tense st 1 sg. πaʲХqōὕХ present tense οāshāʲvХ person 3 pl. -am -af -(i) -an -af Table 52 BaʳʧϑлpʣʲʳonaὕлʣndʧnʥʳлoʤлtʦʣлἵāmХʲлὕanʥuaʥʣʳлуvaὕuʣʳлʧnлitalic represent enclitic endings usually added to a subject of a clause). Sogdic forms of copula continue from Proto-Iranian *(ʜ)ah- (IIr. *ʜas-, Ide. *h es-). Both in Sogdian and in YaghnōϐХл ʳomʣл of the forms changed from the *Proto-Iranian state (see Table 53). Sogdian forms of the second person singular and the first person plural originate either from ‘ ‘-conjugation (GAUTHIOT – BENVENISTE 1929, 60-61) i.e. ‘ ‘-conjunctive forms of personal endings or they can be taken from optative personal endings ŋ-‘ (> Mug -ē фл эnd pʣʲʳщлʳʥщлoptщ 248 and *-aima > -ém ьst pʣʲʳщлpὕщлoptщ щлḲaghnōϐХлpὕuʲaὕлʤoʲmʳлoʤлϑopuὕaлʦavʣлʤoʲmʳл which may be based on *Proto-YaghnōϐХлpʣʲʳonaὕлʣndʧnʥʳлoʤл*-ām < *-ām‘ ьst pers. pὕщлʧmpʤщ л and *-ā› юrd pʣʲʳщл pὕщл ʧmpʤщ л <л *-ā›( ) *-ā› юrd pʣʲʳщл pὕщл pʣʲʤщ л andл ϐyл anaὕoʥyл aὕʳoл *(-)āϑ; analogical form is also in Christian Sogdian ψyouϊлaʲʣ л<лē ψtʦouϊлaʲt лцл-t(a) < *-aϑa эnd pʣʲʳщлpὕщлʧndщлpʲʣʳщ щлρʦʣлdʣvʣὕopmʣntлoʤ some forms of copula from verbal endings shows, that copula was probably more often used as an enclitic form and thus some of its forms were taken from verbal endings in order to regularize conjugation. Sogdian non-enclitic copula of the ʾn- oʤлanлunϑὕʣaʲлoʲʧʥʧnлуδκπлέ7řяфчл second person plural ʾns ʾлъąʳ such stem may be compared with Pahl. ʾnʾd, ʾnʾnd (ibid.). Not only verbal endings affected copula forms – copula was also influenced by pronominal enclitics. The main feature is prefixation of x= to forms of copula of the third persons singular Forms of the third person diffʣʲл ʧnл BaʲtanʥХл dʧaὕʣϑtʳл – ʧnл BaʳХdХл tʦʣʲʣл ʧʳл noл ʣndʧnʥчл ʧnл πʧpān₍Хл -i is used (cf. SOKOLOVA 1966, 379-юřыфчлζлʦavʣлnoлʧnʤoʲmatʧonлϑonϑʣʲnʧnʥлοaʷmēdХлandлBaʲdaʲaХщ 248 Or maybe by occasional palatalization of *h (δκπлέяы5; see chapter II.1.3.20.v.). 247 ·152· andлpὕuʲaὕщлζnлπoʥdʧanлtʦʧʳл pʲonomʧnaὕ лx= appears in present and imperfect indicative and in subjunctive, in YaghnōϐХл onὕyл ʧnл tʦʣл tʦʧʲdл pʣʲʳonл ʳʧnʥuὕaʲл ʤoʲmʳл oʤл ʧndʧϑative present and imperfect. In all forms there this x= ʧʳл optʧonaὕ чл ʧщʣщл tʦʣʲʣл aʲʣл ʤoʲmʳл ʷʧtʦл x= or without it. The x= originates in the third deictic demonstrative ŋ(‘)h‘u уϑʤщл δκπл έьюŚřщϐчл έьяы5ф 249 . YaghnōϐХл t comes from combination of the second person singular copula with the second person singular enclitic, i.e. -t. The use of pronominal elements in forms of copula can be observed in some Eastern Iranian languages such as Ossetic, Pashtōл oʲл τakhХл уKORN 2011). Comparable is also merger of copula with pʲonomʧnaὕлʣnϑὕʧtʧϑʳлʧnлvʣʲϐaὕлʣndʧnʥʳлʧnлtʦʣлἵāmХʲл languages. indicative present Sogdian 1 sg. m 2nd sg. ē 3rd sg. əstí251, (x) čí 1st pl. ēm255 2nd pl. ąsϑ( )256 3rd pl. (x)aṁd257 st imperfect subjunctive YaghnōϐХ *Iranian Sogdian YaghnōϐХ ŋáhm ⁾ān im 250 ŋáh ā ʧšt (i(št)) 252 (x)áʳt(i), =x ŋást (x)ā (x) (x)āt253, ə ŋhmáh ōm ŋstā ōʳл‖лōt yōʳл‖л yōt ŋhánt (x) ṁd ōʲ optative irrealis Sogdian əst ₍āt254 əst ṁd Table 53 Copula. Copula also serves as a verb to h‘v“ – in this issue only form of the third person singular is uʳʣdлʷʧtʦлoϐὕʧquʣлʤoʲmʳлoʤлʳuϐ₍ʣϑtщлπuϑʦлϑonʳtʲuϑtʧonлʧʳлtypʧϑaὕлaὕʳoлʧnлtʦʣлἵāmХʲлὕanʥuaʥʣʳлoʲл in Turkic (see NἴσÁἰ [in print], note 22). Negative forms of copula have analytic forms in Sogdian based on (historical) negative of the nyst nysṯṯ nysṯ, nỿsṯ, nỵsṯ, nsṯ ъnēʳtчл Şnʧʳtъл ψуʳфʦʣϊл ʧʳ not л <л *na-ást [Pers. nēst, cf. Eng. sn t] nystym ɨmъл ψζϊл amл not уδκπл έ7řяф. In YaghnōϐХл 249 Initial x- in forms of copula can be also explained as analogical spread of x- from the third person plural indicative present copula xnt nt xnd /xaṁd/ < *(ʜ)hánt < Ide. *h sént (cf. δκπлέ77ы-774; HORN 1988, 245). I believe that pronominal origin of x= is the most probable explanation. See also Persian forms ast and hast. 250 From ŋ чt, i.e. with suffixed enclitic second person singular pronoun (GAUTHIOT – BENVENISTE 1929, 52). 251 In Sogdian in the Sogdian Script also enclitic əst/ast; in Christian Sogdian sti (cf. QAοФB 1965, 224). 252 From optative (QAοФB 1965, 225). 253 From Ir. *ahat (QAοФB 1965, 225). 254 From Ir. ŋh ‘t (QAοФB 1965, 225). 255 In Sogdian in the Sogdian Script also man; in Christian Sogdian also (x) (cf. QAοФB 1965, 224; VINOGRADOVA 2000a, 89). In Sogdian in the Sogdian script and in Manichaean Sogdian also enclitic =(ə)sϑ ; in Christian Sogdian (cf. QAοФB 1965, 224). Both forms are probably reanalysed forms of the second person singular copula with second person plural ending (cf. KHROMOV – LIVSHITS 1981, 480). 257 In Sogdian in the Sogdian Script also əstáṁd (cf. QAοФB 1965, 224). 256 ·153· negative prefix ná- is added in front of copula, there can be also the third person copula short form na=x. For a more comprehensive study of Sogdian verb see the Analysis of the Verbal System in the Sogdian Language ϐyлBadʲʣzzamān QAοФB (1965). (excursion 6) Ergative So-called ergative construction 258 appears to be one of the most important features of development of the Iranian languages – it gradually developed into a primary way to express past tense(s). Antje Wendtland connects Iranian ergativity with development of periphrastic perfect which is known also in many (Western) European languages (WENDTLAND 2011). Iranian ergative construction is formed with past participle and auxiliary verb to be or to have, the у βuʲopʣan фл pʣʲʧpʦrastic perfect is formed with a passive participle and auxiliary verb to have (ibid., 39)259. The periphrastic perfect formed with -nt-participles and auxiliaries to be (“ -) and to have (ḫar(k)-) is found also in Hittite, similar construction is attested also in Latin and in Old Indic (ibid., 39-42; cf. also GARRETT 1990). intransitive: I … h‘v“ com“ je suis venu 1 sg. ʾʾ tʾ₍m ət- m nd tu es venu 2 sg. ʾʾ tʾ₍ ət-ē transitive: I … h‘v“ giv“n j ‘ donné rt(w) ʾ›ʾm ϑ áṙ[t ] tu ‘s donné rt- ʾ›ʾ₍ ϑ áṙ[t ] ( ) st ət ətət-ēm il est venu 1st pl. ʾʾ t ʾ tʾ ʾʾ tʾ₍m nous sommes venus ϑ’›dʾ›₍m 2nd pl. ʾʾ ts ət-ąsϑ vous êt“s v“nus ϑ’›dʾ›₍ tʾ 3rd pl. ʾʾ tʾnt ət-aṁd ils sont venus ϑ’›dʾ›nt 3rd sg. rt(w) ʾrt elle est venue ϑ áṙ[t ] ṙt ϑ áṙ[t ] ēm ϑ áṙ[t ϑ áṙ[t ϑ áṙ[t ] ] ] ϑ ṁd lй“ll“ ‘ donné nous ‘v“₎ donné vous ‘v“₎ donné ls ont donné Table 54 Ergative construction in Sogdian, forms are given in various orthographies (after Wendtland 2011, 43, Table 1, edited) For the Iranian languages the periphrastic perfect is attested yet in the Old Iranian period (see examples given in CARDONA 1970). The Iranian periphrastic perfect emerged from forms of past participle and copula – as there was no independent form for verb to have it was also expressed by copula with subject in genitive case260. The ergative construction emerged from difference of transitive and intransitive verbs – the periphrastic perfect of transitive verbs emerged from a past participle and verb to be (i.e. subject in nominative + copula that agrees with subject in form), the intransitive verbs emerged from a past participle and verb to have (i.e. «An S[plit]E[rgative] language is one in which some transitive clauses, but not all, are ergative constructions. … I will define an ergative construction as a transitive clause in which a special case-form or adposition marks the semantic agent, or verb-‘g›““m“nt s ⁽ th p‘t “nt n p›“”“›“nc“ to ‘g“nt» (DELANCEY 1981, 627). 259 There are two kinds of periphrastic perfect in the European languages – be- and have-languages, e.g. (Old High) German, Dutch, Frisian, Icelandic, Norwegian, Danish, French, formerly Catalan; and have-languages; e.g. English, Swedish, Spanish, Catalan, Portuguese, Romanian, Albanian (cf. WENDTLAND 2011, 40 Map 1). 260 E.g. in Latin or Latvian the subject of such possessive construction is in dative case. 258 ·154· subject in genitive (oblique) case and copula in form of the third person singular). The difference of case of the subject and form of copula influenced the development of the ergative construction, in many cases e.g. Old Persian forms are very similar to Latin: OPers. m‘ t₍‘ m‘nā k›t‘m, Lat. hoc (est) quod a me factum est ζлʦavʣлdonʣ луϑʤщлCARDONA 1970, 1). The forms oʤл ʣʲʥative-ὕʧ₎ʣ лpʣʲʧpʦʲaʳtʧϑлpʣʲʤʣϑtлʳʣʲvʣdлaʳлaлϐaʳʣлʤoʲлʤuʲtʦʣʲлdʣvʣὕopmʣntлoʤлpaʳtлtʣnʳʣʳлʧnл all other Iranian languages. stage 1 2 3 4 5 6 verbs itr tr itr tr itr tr itr tr itr unacc itr unerg tr itr unacc itr unerg tr patterns of the perfect past participle + form of be attached past participle, no aux., enclitic pronoun past participle + form of be attached past participle, no aux., enclitic pronoun past participle in -w + form of have past participle + form of be attached past participle in -w + form of have regular past participle + form of be attached first intransitive verbs with have past participle in -w or Ø + form of have past participle + form of be attached past participle + form of have past participle without -w + have attached past participle + form of be attached past participle + form of have past participle without -w + have attached example ʾʾ tʾ₍m -m pt ⁽ t ʾʾ tʾ₍m -m pt ⁽ t rtw ʾrt ʾʾ tʾ₍m wytw ʾrʾnt ʾʾ tʾ₍m ›ʾt ʾrt pt ⁽ t⁽ ʾ›ʾm ʾʾ tʾ₍m ž⁽ʾdʾ›t pt⁽₍sdʾ›nt ʾʾ t-ʾym ⁽₍⁽sdʾ›t ⁽₍dʾ›t texts e.g. Ancient Letter II e.g. Ancient Letter V Buddhist texts (mainly in direct speech) Manichaean texts (also used in narrative) Christian texts (begins to the imperfect) the replace Christian Gospels, KG 2 (used as simple past) Table 55 Stages of development of the have- and be-perfect in Sogdian (Wendtland 2001, 50 Table 2) (tr = transitive verb; itr = intransitive verb; unacc = unaccusative; unerg = unergative; aux. = auxiliary). Nearly in all Modern Iranian languages the past tenses are formed with later developments and reanalysis of the ergative construction – ʣщʥщл ʧnл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳл tʦʣл ʣʲʥative construction was reanalyzed for transitive verbs – there has been lost form of copula of the third person singular and oblique forms of subject were gradually replaced by enclitic personal pʲonounʳщл ἱatʣʲл have-pʲʣtʣʲʧtʣ л pʲʣdomʧnatʣdл ʣщʥщл ʧnл ma₍oʲʧtyл oʤл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳ;л ʧnл several instances the enclitic-based endings influenced personal endings of present tense (see Table 52). In the languages of the ShughnХ-οōshānХлʥʲoupчлʧnлḲazghuὕāmХлandлτakhХлtʦʣл paʳtл tʣnʳʣ л ʣndʧnʥʳл aʲʣл uʳuaὕὕyл ϑonnʣϑtʣdл toл ʳuϐ₍ʣct, not to verb, e.g. Shugh. ⁽u₎ l v-um ζл ʳay л ×л ⁽ú₎чum l vd ζл ʳaʧd ;л οōsh. ‘₎ lúv-um ζл ʳay л ×л mu luvd or á₎чum luvd 261 ζл ʳaʧd ;л οōsh. ā₎ ζлʳaʧd ;лḲazgh. az laf-ín ζлʳay л×лm n l‘”t or á₎чəm laft ζлʳaʧd ;лτakh. lúv-um ζлʳay л×л ⁽u₎ ⁾án-əm ζлʳay л×л⁽ú₎чəm ⁾‘t(əy)262 ζлʳaʧd щлζnлζsh₎āshmХлtʦʣлpaʳtлtʣnʳʣлpʣʲʳonaὕлʣndʧnʥʳлmay ζnлοōshānХлtʲanʳʧtive prefect mu luvd (literary m“ s‘ d ) is used only by elder speakers, younger generations use construction á₎чum luvd (literary Iчm₍ s‘ d ~ Iч‘m s‘ d ) similar to ShughnХлoʲлBaʲtanʥХщ 262 Wú₎чəm ⁾‘t in Western WakhХчл⁽ú₎чəm ⁾‘təy in Eastern and Central WakhХлуPAKHALINA 1969, 100). 261 ·155· be connected to a verb or, more often, to subject: ‘₎í áž-ьm ζл ʳay л ×л ‘₎íчm ‘žd or ‘₎í áždчьm ζл ʳaʧd (the personal ending may be even doubled: ‘₎íчm áždчьm); κun₍Хл intransitive verbs show typical ergative construction: zə ž ₍-əm ζлʳay л×лmən ⁱ təm ζлʳaʧd луaὕὕлtʦʣл above presented examples are taken from PAKHALINA 1969). Development of split ergativity can be seen also in Sogdic dialects – in both Sogdian and YaghnōϐХл ʷʣл ϑanл ʳʣʣл dʣvʣὕopmʣntл oʤл tʦʣл oʲʧʥʧnaὕл pʣʲʧpʦʲaʳtʧϑл pʣʲʤʣϑtл ʧnл liv“ ’›o‘dc‘st . As shown by Antje WENDTLAND (2011) there can be observed six stages of development of perfect in Sogdian, in YaghnōϐХлtʦʣʲʣлtoo₎лpὕaϑʣлʲʣanaὕyʳʧʳлoʤлtʦʣлoʲʧʥʧnaὕлʣʲʥative construction quite recently. According to attested personal endings it seems that already *Proto-Sogdic lost inherited forms of Iranian perfect and it was replaced by a new periphrastic perfect based on split ergativity. Sogdian development of periphrastic perfect shows gradual extension of the ergative construction inherited (?) from *Common Iranian. The oldest attested examples of the ergative construction come from the Ancient Letters – in the Ancient Letter II there are simple archaic forms – past participles of intransitive verbs are formed with subject in nominative and with inflected copula, for transitive verbs the subject takes enclitic form of a personal pronoun (see WENDTLAND 2011, 44 – examples 10-11). In all other Ancient Letters (mainly in the Ancient Letter V) also new forms of periphrastic perfect appear – the transitive past participle has ending in - < *(i.e. accusative singular) followed by inflected form of the verb √ ā› toлʦoὕd луşл ʳʣmantʧϑaὕὕyл to ʦavʣ члϐutлtʦʧʳ meaning of the verb √ ā› is used only for transitive forms, in all other cases the verb to have is expressed by subject in genitive/oblique and copula of the third person singular; cf. WENDTLAND 2011, 45 – examples 12-16), but the archaic form of perfect with enclitic pronouns are still attested together with the innovated forms (ibid., 45 – example 17). Later the periphrastic perfect changes its function form direct speech past through narrative past to expression of past tense in common and replaces imperfect (see Table 55; ibid., 46-50). ζtлʳʦouὕdлϐʣлnotʣdлtʦatлtʦʣлoὕdʣʳtлattʣʳtʣdлʤoʲmatʧonлoʤлtʦʣлpʣʲʧpʦʲaʳtʧϑлpʣʲʤʣϑtлуτʣndtὕand ʳл Stage 1) is very similar to (yet rather archaic) formation of periphrastic perfect in YaghnōϐХ;лonл the other hand, the most innovativʣлʤoʲmʳлуʧщʣщлτʣndtὕand ʳлStage 6) shows similar formation of perfect in Ossetic263. Ossetic has two sets of preterite endings – intransitive endings are based on forms of copula, transitive endings come from forms of verb to have, see following scheme for the Ossetic Iron dialect: perfect copula person present (present) tr. itr. st -(t/d)on -(tйd)æn dæn 1 sg. - n nd -(t/d)ay -(tйd)æ dæ 2 sg. - s rd -(t/d)a -(is) u / is / i 3 sg. st -(t/d)am - stæm stæm 1 pl. -æm nd -(t/d)at - stut stut 2 pl. -ut rd nc -(t/d)oy st st 3 pl. (ISAEV 1987, 619) 263 ·156· The intransitional periphrastic perfect is formed from past participle in *- - to which is are added inflected forms of copula, only forms of the third person singular have no copula, instead of copula nominative singular endings are used – masculine light stems add ending -í < *-ah, but heavy stems have no ending, feminine forms add - < *-ā (with no distinction of light and heavy stems). Transitional perfect forms have ending in - and auxiliary verb √ ā› toл ʦavʣ ;л tʦʣл ʣndʧnʥ л - probably comes from accusative singular of preterite in *- . There are attested forms in <-w> and in -ø in Sogdian, Ilya Gershevitch interprets them as light and heavy stem endings respectivʣὕyл уδκπл έř7ř-879), but Antje Wendtland interprets the forms with -w as older than those without -w (WENDTLAND 2011, 43)264. In later development the auxiliary verb √ ā› merges with the past participle stem ending -t into single agglutinated form: ā›- > *-t(=) ā›- > - ā›-, this feature can be clearly observed in Christian Sogdian texts. st 1 sg. nd 2 sg. transitive verb I saw him (lit. by me seen) intransitive verb I went thou saw him (lit. by thee seen) thou went he saw me rd 3 sg. he saw thee he went he saw (him) st 1 pl. I saw them (lit. they by me saw) 2 pl. rd 3 pl. we went t ʲtуaфŞōʳл‖лt ʲtуaфŞōt nd he saw (them) (lit. they by him saw) (a)Şōʲ you went they went Table 56 Ergative construction in YaghnōϐХлaϑϑoʲdʧnʥлtoлBOGOLYUBOV (1966, 354). In YaghnōϐХлtʦʣлdʣvʣὕopmʣntлoʤлtʦʣлʣʲʥative construction was quite different and is more similar to the ergative construction of the Stage 1 as observed by Wendtland. Mikhail Nikolaevich BOGOLYUBOV (1966, 354) quoted typical ergative construction in YaghnōϐХчлonлtʦʣл other hand Aὕ ϐʣʲtл ἱʣonʧdovʧch KHROMOV уьŚ7эчл ю6фл notʣdл onὕyл ʧntʲanʳʧtivʣ л ʧnflection for perfect and in the latest YaghnōϐХлʥʲammaʲлϐyлπayfiddХnлκХʲzōzōdaлandлBaʦʲʧddХnлAὕavîлtʦʣʲʣлʧʳл pʲʣʳʣntʣdл onὕyл intransitive inflʣϑtʧon л уʳʣʣл ρaϐὕʣʳ 56 and 57). Forms of the ergative construction changed a little bit during past fifty (?) years – this state was probably caused by intensive contact of YaghnōϐХлʷʧtʦлTajik. The forms of intransitive verbs retained unchanged form and they are practically identical with (unaccusative) intransitional perfects in Sogdian. The transitive perfects have two forms – the first (nowadays rather archaic) is quite similar with the forms presented by Bogolyubov (Table 56), but I have not met forms such as , 264 Antje Wendtland claims that the non-auxiliary part of the transitive periphrastic perfect originates from a past stem in -tw (WENDTLAND 2011, 43), I suppose that accusative form of the past participle is more accurate interpretation. The interpretation of the origin of the participial ending - from accusative singular *may have analogies in Latin: l tt“›‘m (”) sc›Дptam (”) h‘’“ > l tt“›‘m (”) sc›Дptum (m) h‘’“ уὕoʳʳлoʤлaʥʲʣʣmʣntфл ζлʦavʣлʷʲʧttʣnлaлὕʣttʣʲ л (WENDTLAND 2011, 40). Maybe that the two different form in - (√ ā›) and -ø (√ ā›) are not connected with the light or heavy stems but with gender. Such issue has to be analysed yet, the loss of - then may be interpreted as loss of gender agreement. ·157· (but it does not mean they are not used even today); the other is consistent with system presentʣdл ϐyл κХʲzōzōdaл andл Aὕavîл уρaϐὕʣл 57фл andл ʧʳл moʲʣл uʳʣdл amonʥл tʦʣл ḲaghnōϐХʳл with whom I have spoken – outline of positive and negative forms of ergative construction in contemporary YaghnōϐХл ʧʳл pʲʣʳʣntʣdл ʧnл ρaϐὕʣ 58; it is evident that there is a tendency to simplify the ergative system in contemporary YaghnōϐХщ ʧntʲanʳʧtivʣлϑon₍uʥatʧon л(Khromov) tʲanʳʧtivʣлϑon₍uʥatʧon л(ḤД›₎ ₎ d‘ – Al‘vî) Cyrillic romanized I bought л st I said nd thou said л he said л he bought we said л we bought 1 sg. 2 sg. rd 3 sg. , st 1 pl. nd 2 pl. ʷ ʤtуaфŞōʳл‖лʷ ʤtуaфŞōt rd 3 pl. t‘u ⁾ⁱ› you said л they said л thou bought you bought áut t ⁾ⁱ› they bought Table 57 Periphrastic perfect according to Khromov (1972, 36) and MФοḳŌḳŌαA – AἱAσÎл(2008, 57). positive negative transitive intransitive st 1 sg. =x(ast) tauлʷ nd 2 sg. =x(ast) rd 3 sg. st 1 pl. =x(ast) š nd 2 pl. rd 3 pl. =x(ast) áutʧtʧлʷ =x(ast) I й thou й (s)h“ … s‘⁽ t ʲtуaфŞōʳл‖ t ʲtуaфŞōt transitive intransitive náŞm wḕta=x(ast) ná tṑrta=im náŞt wḕta=x(ast) ná tṑʲtaŞʧšt n᪚ wḕta=x(ast) ná tṑrta=x(ast) náŞmōʸ wḕta=x(ast) ná tṑʲtуaфŞōm n᪚ʧnt wḕta=x(ast) ná tṑʲtуaфŞōʳл‖ ná tṑʲtуaфŞōt (a)Şōʲ I й thou … ⁽“nt I … d d not s““ ná tṑrt(a)Şōʲ I … d d not com“ Table 58 Overview of ergative construction forms of resultative perfect in contemporary YaghnōϐХщ II.2.5. Adpositions There are several prepositions and postpositions both in Sogdian and YaghnōϐХщлπoʥdʧanлʳʦoʷʳл archaic state of pre- and postpositional system, YaghnōϐХл pʲʣʳʣʲvʣʳл onὕyл ʳomʣл ʧnʦʣʲʧtʣdл adpositions: č ʤʲom (Sogd. cy ъčɨ/; Khʷāʲщлcy), =sa toʷaʲdʳчлto чsʾ› чsʾ(r) , ṙ/; Khʷāʲщлsʾ›), =pi ʷʧtʦ (cf. Khʷāʲщлpy), ч›Дti onчлϐy ryty ryṯyy), =nʉt ʧn , чčìntì› c(y)ntr c(y)ndr ъčɨṁdər/); archaic adpositions are Yagh. par ʤoʲ, because of (Sogd. pʾ› pʾ(›) pʾ , ṙ/), and pu ʷʧtʦout (Sogd. (ʾ)p⁽ pw / púъ). Sogdian YaghnōϐХ 1 person 2 person (definite article) 3rd person st to from with about nd ṁn ṁn pə pə Table 59 Prepositions combined with pronouns. ·158· čau Some prepositions can be combined with pronouns – good examples are attested in Sogdian, in YaghnōϐХлtʦʣʲʣлʧʳлattʣʳtʣdл₍uʳtлonʣлϑomϐʧnʣdлpʲʣpoʳʧtʧonлč‘u ʤʲomлtʦʧʳ луʳʣʣлρaϐὕʣ 59). I will not describe here the adpositional system on both languages – comprehensive dʣʳϑʲʧptʧonлoʤлπoʥdʧanлadpoʳʧtʧonʳлʧʳлʧnлδκπлέ1610-1632 and LIVSHITS – KHROMOV 1981, 503510), for YaghnōϐХлʳʣʣлKHROMOV 1972, 53-62. II.2.6. Conjunctions YaghnōϐХл andл πoʥdʧanл pʲʣʳʣʲvʣл Iranian conjunction ŋutā > Sogd. =ət(í), Yagh. =(a)t and щл ζnл Sogdian this conjunction is often used clause-initially standing after another archaic conjunction ʾr / / and л<лŋ› < Ide. *h (e)r [Gre. чл ᾽, ἄ у фчлἱʧth. ›/a›, Latv. ì›/ar; TokhB ra= emphatic particle ] (GAUTHIOT – BENVENISTE 1929, 171): ʾrty (ʾ)rty, rt(ty) (ʾ)›ṯy, ʾrtty, rty ʾrṯ / t(í)/ < ŋ›чutā. In YaghnōϐХлʳamʣлaʳлʧnлTajik, Uzbek and many other languages of Central Asia is widely used Arabic conjunction wa and лşлḲagh. va (occasionally wa). ·159· III. Lexicon In the third part of the presented thesis there will be presented a short comparative dictionary of ϐaʳʧϑ лvoϑaϐuὕaʲyлof YaghnōϐХлandлπoʥdʧanщлρʦʣлὕʣʸʧϑonлʧʳлϐaʳʣdлonлtʦʣлʣʸtʣndʣdлπʷadʣʳʦл List (i.e. list of 207 words) supplemented by a ὕʧʳtлoʤлэьылvoϑaϐuὕaʲyлoʤлtʦʣл πtandaʲd Word List ζtʣmʳ presented in the five-volume Sociolinguistic Survey of Northern Pakistan (see http://www.sil.org/sociolx/pubs/ssnp.asp) by the National Institute of Pakistani Studies, Quaid-i-Azam University and Summer Institute of Linguistics. By combination of both word-lists I have studied 298 lexical items, but some items have not been translated into YaghnōϐХлandъoʲл Sogdian due to cultural and/or historical reasons (e.g. there are presented terms such as eggplant or mango but I have not translated them because there was no need to search meaning of these words in Sogdian as they are non Central Asian origin, also there is no YaghnōϐХлtʲanʳὕatʧonлoʤл such words because there is only a little possibility that the YaghnōϐХʳлʷʧὕὕлʦavʣлtoлnamʣлʳuϑʦл items, and if so, they will be referred to in Russian or less likely in Tajik), the only exception are words for potatoes and tomatoes – potatoes are planted nowadays in Yaghnōϐлandлtomatoes can be bought on markets in centres adjacent to the Yaghnōϐл σaὕὕʣyл уϐutл tʦʣʳʣл ʷoʲdʳл ϑomʣл from Russian via their colloquial Tajik forms). The items are aligned according to the Swadesh List, items of the standard word-list are usually ordered according to their semantic relations with the Swadesh List, in cases when the standard word-list items do not correspond to the Swadesh List I have kept their alignment as in the SIL publications (see BACKSTROM 1992, 273-284; HALLBERG 1992; DECKER 1992, 177-211). For better work with the vocabulary I have split individual words into 21 units which better group their common semantic values. Some of the words (mainly in case of Sogdian) were left untranslated as I have not found their meanings in Sogdian and/or YaghnōϐХлуunʤoʲtunatʣὕyлζл have not made the YaghnōϐХлtʲanʳὕatʧonʳлduʲʧnʥлmyлʳtayʳлʷʧtʦлtʦʣлḲaghnōϐХʳфщ The numbers of individual lexical items respect their number on both lists: words of the Swadesh List are left unmarked, the standard word-list items are given in brackets. The lexical items that have been borrowed into YaghnōϐХл aʲʣл maʲ₎ʣdл ʧnл italics in the vocabulary, but words that appear similar both in YaghnōϐХл andл ʧnл Tajik (and where precise origin cannot be judged) are considered as inherited. Also some parts of a word can be in italics – I marked such way borrowed elements of compounds (e.g. Yagh. vanlinkáл ʳpʧdʣʲ л <л Ḳagh. van(n) ὕonʥ лandлϐoʲʲoʷʣdлlink ὕʣʥ лцлḲagh. suffix -á) or sounds that changed probably due to Tajik influence (e.g. Yagh. díndak tootʦ л – instead of the second d we should except t in YaghnōϐХф. The analysed lexicon is supplemented by etymologies of the translated items, etymology is given in cases when it was known to me. Many words were unfortunately left without their etymologies. ·160· The analyzed word items are as follows, for comparison I have added their translations into modern literary Tajik (the Tajik forms are transliterated as if they were written in the PersoArabic script; î transliterates Tajik Cyrillic word-final st›“ss“d Д <ӣ>): Pronouns 1. (202.) I man 2. (203.) thou tu 3. (205. & 206.) he, she ( )ж v‘ 4. (207. & 208.) we m 5. (209 & 204.) you 6. (210.) they nh ж v‘ h (171.) this Дn (172.) that n (173.) these (174.) those 9. here Дnǰ ( ) 10. there nǰ ( ) 11. (165.) who? 12. (166.) what? 13. (167.) where? 14. (168.) when? k‘ -t‘u› 15. how? (169.) how many? č‘nd (170.) which? 16. not na 17. (181.) all h‘má 18. (180.) many 19. some 20. (179.) little / few kam 21. other kam (176.) different dДgá› Numerals 22. (151.) one yak 23. (152.) two du 24. (153.) three se 25. (154.) four 26. (155.) five p‘nǰ (156.) six ‘ (157.) seven haft (158.) eight h‘ t (159.) nine n h (160.) ten dah (161.) eleven ₍ ₎dáh (162.) twelve duv ₎dáh (163.) twenty ’Дst (164.) (one) hundred sad Adjectives (i) 27. (142.) big 28. (134.) long 29. wide 30. thick afs 31. (144.) heavy 32. (143.) small ⁾u›dж k číkж m‘ dá 33. (135.) short k t 34. narrow tang 35. thin tunúk (145.) light s‘’úk People 36. (103.) woman zan 37. (102.) man mard 38. human 39. (104.) child k dákж ’‘č(č)á 40. (114.) wife zan 41. (113.) husband ‘uhá› 42. (106.) mother m dá› 43. (105.) father p‘dá›ж p dá› (107.) older brother (108.) younger brother d dá› (109.) older sister ‘p(p)á (110.) younger sister x há› (111.) son p sá› (112.) daughter du⁾tá› (115.) boy ’‘č(č)áж p sá› (116.) girl du⁾tá› Animals 44. animal ‘ v 45. (86.) fish m hî 46. bird mur ж p‘››‘ndá (87.) chicken mur ж č ǰá 47. (95.) dog sag (89.) cow g ⁽ (90.) buffalo (94.) goat buz (97.) monkey m‘ m 48. louse pí 49. (96.) snake m › 50. worm kirm (98.) mosquito / fly p‘ á й m‘gás ·161· (99.) ant m ›čák (100.) spider t ›t‘nák Plants 51. (61.) tree d‘›á⁾t 52. forest ǰ‘ngálж ’ē á 53. stick č ’ 54. (66.) fruit mēvá 55. seed d náж tu⁾m 56. (62.) leaf barg 57. (63.) root ›ē á 58. bark p st- d‘›á⁾t (64.) thorn ⁾ › 59. (65.) flower gul (67.) mango (68.) banana (69.) wheat g‘ndúm (70.) barley ǰ‘u (71.) rice ’ ›ínǰ (72.) potato k‘›tó k‘ (73.) eggplant (74.) groundnut (75.) chilli / pepper mu›č (76.) tumeric (77.) garlic sД› (78.) onion (79.) cauliflower (80.) tomato pom dó› (81.) cabbage k‘›ám 60. grass ‘l‘”ж s‘’₎á 61. (36.) rope ar Body parts 62. (84.) skin p stж č‘›m 63. (84.) meat g t 64. (22.) blood ⁾ n 65. (20.) bone ustux 66. (85.) fat č‘›’ (82.) oil ›‘u án (91.) milk ē› 67. (88.) egg tuxm 68. (92.) horn ⁾ 69. (93.) tail d m 70. feather par 71. (3.) hair m ( ) 72. (2.) head s‘›ж k‘llá (4.) face č“h› 73. (6.) ear g 74. (5.) eye č‘ m 75. (7.) nose ’Дnî 76. (8.) mouth 77. (9.) teeth 78. (10.) tongue 79. (17.) fingernail n ⁾ún 80. foot p ( ) 81. (18.) leg ling 82. knee 83. armband dast (14.) elbow nǰ (15.) palm p‘nǰá (16.) finger ‘ngú t 84. wing ’ lж (1.) body badan, tan 85. (12.) belly kámж kám 86. guts › dá 87. neck g‘›dán 88. back pu t 89. (11.) breast sДná 90. (21.) heart dil, qalb 91. liver ǰ gá› (23.) urine (24.) feces g h Verbs 92. (185.) to drink n Дdán : n 93. (182.) to eat x ›dán : ⁾ ›94. (183.) to bite g‘₎Дdán : g‘₎95. to suck m‘kДdán : m‘k96. to spit tuf k‘›dán 97. to vomit ‹‘ k‘›dán 98. to blow pu” k‘›dánж v‘₎Дdán : v‘₎99. to breathe n‘”ás k‘ Дdán 100. to laugh ⁾‘ndДdán : ⁾‘nd101. (201.) to look / to see dДdán : ’Дn102. (200.) to hear / to listen unДdán : un‘v-й un‘u103. to know d n stán : d n104. to think ‘ndē Дdán : ‘ndē -, fik› k‘›dán 105. to smell ’ ( ) k‘›dán 106. to fear t‘›sДdán- : tars107. (187.) to sleep x u”tánй⁾ ’Дdán : ⁾ ’108. to live ₎Дst‘n : ₎ ₍-, zindagî k‘›dán 109. (192.) to die mu›dán : m ›110. (193.) to kill ku tán : ku 111. to fight ǰ‘ngДdán : ǰ‘ng-ж ǰ‘ng k‘›dán 112. to hunt 113. to hit ₎‘dán : ₎‘n114. to cut ’u››Дdán : ’u››- ·162· 115. to split k ”tán : k ”116. to stab k ›d ₎‘dán 117. to scratch 118. to dig k‘ndán : k‘n-ж k ”tán : k v-йk ⁽119. to swim 120. (194.) to fly p‘››Дdán : p‘››121. (195.) to walk g‘ tán : g‘›d-ж › h ›‘”tán 122. (198.) to come m‘dán : ₍-й ( )(196.) to run d‘vДdán : d‘v-йd‘u(197.) to go ›‘”tán : ›‘v-й›‘u123. (188.) to lie (down) x u”tánй⁾ ’Дdán : ⁾ ’-, 124. (189.) to sit n ‘stán : n Дn-ж tán : Дn125. to stand 126. to turn č‘›⁾Дdán : č‘›⁾-ж g‘›d n(Д)dán : g‘›d n127. to fall ‘”t dán : ‘”t128. (190.) to give d dán : d h-/deh129. to hold g › ”tán : gД›-ж d tán : d ›130. to squeeze fi 131. to rub m lДdán : m l132. to wash ustán : ₍-й ( )133. to wipe p k k‘›dán 134. to pull k‘ Дdán : k‘ 135. to push tēlá d dán 136. to throw p‘›t ”tán : p‘›t v-йp‘›t ⁽-ж ‘nd ⁾tán : ‘nd ₎137. to tie bastan- : band138. to sew d ⁾tán : d ₎139. to count umu›dán : um ›140. to say / to speak guftán : g ₍-йg ( )141. to sing su› dán : s‘› ₍-йs‘› ( )-, x ndán : ⁾ n142. to play ’ ⁾tán : ’ ₎-ж ’ ₎î k‘›dán 143. to float n vá› udán 144. to flow 145. to freeze ₍‘⁾ k‘›dán 146. to swell m sДdán : m s(184.) to be hungry gu›usná ’ dán (186.) to be thirsty t‘ ná ’ dán Celestial objects 147. (41.) sun x 148. (42.) moon 149. (44.) star s t ›á Nature (i) 150. (46.) water ’ 151. (45.) rain 152. (47.) river d‘›₍ ж › d⁾ ná 153. lake k l r 154. sea 155. (83.) salt n‘mák 156. (52.) stone sang 157. (54.) sand ›ēgж ‹um 158. (59.) dust č‘ngж ⁾ k 159. earth (58.) mud l ₍ Weather 160. (48.) cloud abr 161. fog t mán 162. (43.) sky 163. (51.) wind 164. snow barf 165. ice yax (49.) lightning ’‘›‹ж t‘ ák (50.) rainbow -i ‘sán-u Fire 166. (56.) smoke d d 167. (55.) fire tá ж ₎ 168. (57.) ash ⁾ k stá› 169. (191.) to burn s ⁾tán : s ₎(29.) firewood hē₎úm Settlement 170. (53.) road / path › h (25.) village (26.) house ⁾ ná (27.) roof ’ m (28.) door dar Tools (30.) broom ǰ › b (31.) butter churn guppî (32.) pestle č‘⁾č b (33.) hammer ’ l á (34.) knife k ›d (35.) axe t‘’á› (37.) thread t › (38.) needle s ₎án (39.) cloth l‘ttá (40.) ring Nature (ii) 171. mountain k h (60.) gold t Colours 172. (150.) red surx 173. green sabz 174. yellow zard 175. (148.) white ·163· usá n ‘lá⁽ 176. (149.) black Time 177. (118.) night ‘’ 178. (117.) day › ₎ (119.) morning sub , sa á› (120.) noon nДm› z (121.) evening / afternoon (122.) yesterday dД› z (123.) today m› z (124.) tomorrow (125.) week h‘”tá (126.) month m h 179. (127.) year s l Adjectives (ii) 180. (136.) hot garm 181. (137.) cold s‘›dж ⁾unúk 182. full pur(r) 183. (129.) new n‘u 184. (128.) old 185. (130.) good ⁾ ’, na z 186. (131.) bad ’‘dж g‘ndá 187. rotten p sДdá 188. dirty , ifl 189. straight › st 190. round gird î 191. sharp tē₎ 192. dull kund 193. smooth su”tá 194. (132.) wet tar 195. (133.) dry ⁾u kж ‹ ‹ 196. correct du›úst 197. (140.) near 198. (141.) far d › 199. (127.) right › st 200. (139.) left č‘p (175.) whole (178.) broken k‘stáж k‘st‘gî Adpositions 201. at ba 202. in (‘n)dá› (146.) above (147.) below 203. with ’ ж kátîж ‹átî Conjunctions 204. and va, -(y/v)u 205. if ‘gá› 206. because -ki Name 207. name n mж sm Swadesh List and standard word-list with YaghnōϐХл andл πoʥdʧanл tʲanʳὕatʧonл andл ʷʧtʦл etymological notes: III.1. Pronouns 1. (202.) I man (arch. az) : man (occ. mánʧ)  ʾzw zw : obl. mnʾ /(əфzú : mənáъ < *aʣám; Ave. azəm, Khōt. aysu, ‘(₍sä), Tumshuq. asu, azu, Oss. æ₎, Shugh. (w)uz, οōsh. az, Khūʤщ οāshrv. Bart. ā₎, πaʲХqщ waz, Yazgh. az, Ishk. az(i), Sangl. azə, azi, Wakh. wuz, Munj. za, Yidgh. zo, zə, Pasht. zə, Wa . ze, OPers. adam, Pers. man, εazāʲщ ma, Kurd. ez, Ved. ‘hám, Ide. *h “gh óm, Gre. ώ, Lat. eg , OCS. az , OCze. já(₎), ORus. ꙗу фч Lith. , OScand. ek, Ger. ich (cf. formally similar but etymologically unrelated Uzb. mėn, colloq. mȧn, Chaghat. mėn, Uygh. män, Kyrg. men, ρüщ ŋ’ėn, ŋmėn, Eynu. män) 2. (203.) thou tu : tau (occ. táʷʧ)  t w t( )w ṯ( )w : obl. t⁽ʾ /t( фú : təʷáъ < tuu‘m; Ave. t , Oss. d , Shugh. οōsh. Khūʤщ οāshrv. tu, Bart. t , πaʲХqщ t w, Yazgh. tow, Ishk. tь, Sangl. t ⁽, Wakh. tu, Munj. tu, Yidgh. tu, tə, Pasht. tə, Pers. t > t , εazāʲщ ·164· Kurd. tu, Ved. tvam, Ide. *tuʜ, Gre. ύ, Lat. tu, OCS. ty, Lith. tù, OScand. OEng. þ , Eng. thou, Ger. du 3. (205. & 206.) he, she ax : áʷʧ  (ʾ) w : ʾw xw(w) : (ʾ)⁽⁽ xw : ʾw, w- /(əфʸú : (əфʷúч ōъ Ir. *(a)h‘u : ŋ‘u‘m, Bactr. /ōъ 4. (207. & 208.) we (inclusive & exclusive) mōʸ  mʾ (h), mʾ (w) mʾ⁾ ъmāʸу{ф/ < < *ahm xam < Ir. *ahm kam, Bactr. у ф /(əфmāʸъ Oss. max, Shugh. οōsh. ἰūʤ. Bart. οāshrv. mā , πaʲХqщ m‘ , Yazgh. mox, Munj. m ⁾, Yidgh. m , x, Ishk. mь⁾(o), Sangl. amax, aməx, Pasht. m (n)ǵ, Wa . mo , Ōʲm. mâ⁾, OPers. ‘(h)mā⁾‘m, Pers. mā, εazāʲщ m ; IIr. 5. (209 & 204.) you (pl. & honorific) (ʾ) mʾ w, ʾ mʾ h ʾ mʾ⁾(⁽), mʾ⁾ mʾ⁾ /ᶤšm xу{ф/ šm x < <ŋum ; Ave. yūžəm, Oset. s m‘⁾ ‖ sumax, Pers. um , ‘ m , Tjk. um , γāʲʳщ AfghP. om , εazāʲщ m 6. (210.) they áʸtʧt : áutʧtʧ  h, h ⁾ʾ / / < ŋ‘h‘u; Yagh. ax : ‘u- + pronominal pl. ending -tit(i) (171.) this yw : (ʾ)m⁽ mw /yu : mú/  < ŋ am, ŋ‘ am : *imam; OPers. iyam : imam ʧšлśлʧtл (ʾ) ⁽ : ʾtw /ᶤšúлśлᶤtúъч < ŋ‘ am : ŋ‘ tam; Ave. ‘ē ‘- : ‘ēt‘-; Bactr. ( ) лъХdъчлἴἵʣʲʳщл‘ t‘- (obl.) Хd  ʾy ъē ъ < ‘ ta- (obl.); Bactr. ( ) ъХdъ (172.) that ax : áʷʧ  (ʾ) w : ʾw xw(w) : (ʾ)⁽⁽ xw : ʾw, w- /(əфʸú : (əфʷúч ōъ < ŋ(‘)h‘u : ŋ‘u‘m; Bactr. /ōъ au  ʾw (ʾ)⁽⁽ ʾw ъōъ < ŋ‘u‘- (obl.); Bactr. /ōъ (173.) these  yw /yu/ < ŋ a-, ŋ‘ a- : *imaíštʧtлуśлítʧtʧфл /ᶤšā : ᶤ / < ŋ‘ a- : ŋ‘ ta-; Yagh. : t- + pronominal pl. ending -tit(i) (174.) those áʸtʧtлśлáutʧtʧл h, h ⁾ʾ / / ŋ(‘)h‘u : ŋ‘u‘-; Yagh. ‘⁾ : ‘u- + pronominal pl. ending -tit(i) 9. here ·165· máʳtar  m›tsʾ› mcʾ, msʾ ъmáṙlaṙ/ < ŋĭmá›ϑă-sā›- < ŋ máϑra-ʦā›10. there ʷáʳtáʲ  ʾ⁽(›)tsʾ› ʾ⁽›tsʾ› ʾ⁽tsʾ› ʾ⁽cʾ, ʾ⁽sʾ ṙlaṙ/ < ŋăuá›ϑă-sā›- < ŋ‘uáϑra-ʦā›-, cf. Tjk. uʳtáʲ 11. (165.) who? kax (: ₎áyʧ, ₎áʸtʧt : ₎áytʧtʧф, -k  (ʾ)k₍, k₍ʾ k₍(ʾ)ж ‹₍(ʾ) ‹₍(ʾ) : k₍ʾ / : ₎yāъ 265 < *káh( ā)- ; Ave. k , Khōtщ kye, kyi, Oss. č ‖ ka, Wakh. k ₍, Shugh. οōsh. Khūʤщ čā₍, Bart. čД, οāshrv. č , πaʲХqщ čo₍, Ishk. k ₍, Sangl. k (₍), Pers. kД, Kurd. ki, Baὕōch. kē, k‘ ; Ved. kás₍‘-, OCS. k to 12. (166.) what? (ʾ)c⁽ cw / č /; cʾ ъčāъ ʣὕὕative prefiʸ čō : čōʧ  < ŋč -āk‘-; Ave. č t, Oss. c ‖ ci, Bactr. /ci/, Khōtщ cu, Khʷāʲщ ciya, Pasht. c k, ca, Shugh. ca, cf. TVarz. čo (only with verb k‘›dán), OPers. č ₍, Pers. čД, Kurd. ç , Baὕōch. č ; Ved. cid, Lat. quid, Gre. 13. (167.) where? ₎ū  (ʾ)k⁽ ʾkw, kʾ⁽ k(ʾ)⁽ / ₎у{фūъ Ave. gen. k ; Pers. k , kuǰ , cf. TVarz. g ǰó; Gre. 14. (168.) when? k (ʾ) k qd ъ₎a ч kə áъ kad  Ave. ka a-; Bactr. /kad/, Oss. kæd, Pasht. kəla, Pers. k‘ ; Ved. k‘dá-(ⁱ)k (encl.)  cf. Pers. ki 15. how? č t(t)i  ʾcwty (ʾ)c⁽t₍ cwty cwṯy / < Ir. ŋč‘h ‘-uti-; Bactr. ч ʾcyd / c d/ (169.) how many? čōʤ  cʾ cʾ” /čāʤъ  cndn /čaṁdan/ č‘nd n, cf. Pers. č‘nd n, , Ave. č(‘)uu‘nt, čuu‘t (170.) which? k m (: k mi, k mtit : k mtiti)  < ŋkām‘-; Khōtщ kām‘-, Wakh. Pasht. kum kad m  ktʾ(ʾ)m ktʾm, k ʾm ‹dʾm /kə < ; Ave. k‘tām‘-, Bactr. /₎adāmъч Ishk. kьdьm; Pers. k‘đ m, TMast. k d m YaghnōϐХлkax is form ŋkáh( ā)- ʷʦo лцлax personal pronoun of the third person singular / demonstrative pronoun of far (< III.) deixis щ 265 ·166· kД, -k (encl.)  < Pers. kД; cf. Gre. у ф 16. not na(ᴥ), ná(ᴥ)a, naʰ, n ‖ naʧ, n  nʾ₍ ny < *na; Bactr. ч Oss. næ, Pers. na, Tjk. colloq. na, na(ᴥ)á, n , n (ᴥ) , Kurd. na 17. (181.) all hám(m)á  < Pers. h‘má, TVarz. hámm‘, h‘má, Uzb. hȧmmȧ. ἶāraqalp. häm‘ b  < Uzb., Tjk. , TMast. b tǘn, p tǘm, πaʲХqщ p t n  wysp-y / wysp-h, wysp-ʾ wysp-y, wysp-ʾ ъʷʧʳpíч ʷʧʳpáъ < ŋu ʦu‘-; Ave. vispa-, OPers. *visa-, Med. *vispa18. (180.) many bis(ⁱ)y r  < Pers. bisy r, Shugh. ’ s₍ ›, Wakh. bəsyor, Uzb. ’ s₍ ›, Eynu. bisyar zⁱy t, zⁱy d  < Ar., Pers. ziy d, εazāʲщ ziy t, ₎ ₍át, Shugh. ₎ ₍ t, Pasht. ziy t, Urd. ₎₍ādā x lē  < Pers. ⁾á lē, Tjk. AfghP. x lē, γāʲʳщ ⁾é lД, πaʲХqщ x yli albalá  r rf ъ aṙf/ < * áṙfu < ŋ”á› u < ŋ”á›u⁽u < ŋ”‘›uu‘m; OPers. paruvam, cf. Wakh. ‘”č, Parāch. alaba p ʲá (arch.)  ʾyw pʾ›ʾ₍k ʾy pʾ›₍k ъХcf. Sogd. pʾ› ъpāʲъ unʧt of liquid volume (120 galons) 19. some cʾ cʾ” /čāʤъ čōʤ  ,  cndn /čaṁdan/ cf. Pers. , , Ave. č(‘)uu‘nt, čuu‘t 20. (179.) little / few k n-y qbn-y /₎a níъч ₎ávⁱn (arch.), kam  < *ká’n‘-; Oss. k næg ‖ kunæg, Pers. kam < *kamna-(ka-); Wakh. kam; Uzb. kȧm, Kyrg. kem, Tr. kem, Urd. kam, NMong. г 21. other áni  ʾnyʾ, ʾnyh, (ʾ)n₍⁽ ʾn₍ʾ, ʾnyh, ʾnyw (ʾ)n₍⁽ /(ə)nyā, (əфnyú/ < ŋán ‘-, Ave. aⁱniia-, Khʷāʲ. ʾny / nХъч Bactr. ( ) ч ч ч Khōt. ‘ñ‘-, Oss. nnæ, ‘nnæ, Ishk. an, Wakh. πaʲХqщ yan, Pahl. Parth. ʾny, OPers. aniya-, Kurd. h“nî; Ved. anya-, PālХ ‘ññ‘ (176.) different ·167· ánʧл ʾn₍ʾ, ʾnyh, (ʾ)n₍⁽ ʾn₍ʾ, ʾnyh, ʾnyw (ʾ)n₍⁽ /(əфnyā, (əфnyúъ < ŋán ‘-, Ave. aⁱniia-, Khʷāʲщлʾny ( ) чл чл члKhōtщл‘ñ‘-, Oss. nnæ, ‘nnæ, Ishk. an, WakhщлπaʲХqщлyan, Pahl. Parth. ʾny, OPers. aniya-, Kurd. h“nî; Ved. anya-, ἵāὕХл‘ññ‘ d gá(›)  < Pers. dДgá›, Tjk. colloq. d gá, TMast. d gá, d ₍á, TVarz. d gá, d gí, cf. Fārs. dДgär, colloq. dДgé, εazāʲщ d gá, Ishk. digar, Wakh. d gā›, Uzb. digȧr, Tr. d ğ“› III.2. Numerals 22. (151.) one yw yw, ẏw yau (m) : ʾywh (f) ъ| u śл| / Х  ʾyw(h) < ŋá u‘-; Ave. ‘ēuu , Khʷāʲ. ʾyw /ēʷъч Oss. iu ‖ yeu, Khōt. śś‘(u), Bactr. ywg /yōʥъч Pasht. yaw (f. ₍‘⁽á), Munj. Yidgh. ₍ , Shugh. οōsh. Bart. οāshrv. ₍Д⁽, yi, πaʲХq. i(w), Wakh. (₍)Д(w), Yazgh. ⁽ g, Ishk. k, Sangl. wok, Pers. yak, Tjk. yak, colloq. ya(g), Fārs. ₍“k, colloq. ₍“( ), ἰāϐuὕХ yak, yag, OPers. aiva-, Pahl. ʾ₍⁽ʾk /ēʷa₎ъ yk /yak/, Parth. yw ъēʷъч Kurd. yek; Ved. ék‘; Eynu. ₍äk, Kyrg. (Southern dial.) ₍äk 23. (152.) two ʾ w(ʾ), w dw(ʾ) (m) : ʾ wy dwy (f) / , ʷā : ʷí/ d ᵎ  ʾ ⁽(ʾ), < ŋd(u)u‘-; Ave. duua-, Khʷāʲ. ʾ w /a ʷʧъч Bactr. ч у ф /lu/, Khōt. d(u)va-, dvi, Oss. d ⁽⁽æ ‖ du⁽(⁽)æ, Shugh. iy n, u, Ba₍ūщ u₍ n, , οōsh. Bart. οāshrv. aw, Khūʤщ ‘⁽(₍ n), πaʲХq. w, a, Wakh. bu(y), Yazgh. ow, Ishk. dь(⁽), Sangl. daw, dow, Munj. lu, Yidgh. loʰ, Pasht. dwa (f. d⁽ē), Pers. d > , Tjk. du, TMast. d , du, Tjk. dial. d , dial. ρa₎ʤōn (arch.) , AfghP. d , d , γāʲʳщ do, Pahl. d , Kurd. du, Baὕōch. d , Ved. duvā(u)-, Lit. dù, Pruss. duai, OCS. dva, dvě, Gre. ύ ч MGre. , Lat. duo, δótщ twai; Eynu. du 24. (153.) three ʾ ryw (ʾ) ry ʾ ry(y) ₍ / šaʧ/ ʳaʲáy, ʳⁱʲáy ‖ tⁱʲáy  ʾ ry < *ϑ ‘-; Ave. ϑ›ā ; Khʷāʲщ ₍ /šēъч Bactr. /həʲēyъч Khōtщ drai, Tumshuq. dre, Oss. æ›tæ, Shugh. aray, οōsh. Bart. οāshrv. ‘›ā₍, πaʲХqщ aroy, Ishk. › ₍, Sangl. › ₍, Yazgh. c ₍, Wakh. t› (₍), Yidgh. ⁾ⁱray, ⁾u›o₍, Munj. ⁾ ›‘₍, Pasht. d›ē, Wa . dre, Ōʲm. ȫ, ṛД, ἵaʲāch. Д, u, Tjk. dial. ρa₎ʤōn (arch.) , Pers. sē, sih > se, Tjk. AfghP. γāʲʳщ se, Kurd. sê; Eynu. si(h) 25. (154.) four ct ʾr ct”ʾ› cṯ”ʾ›, ṯ”ʾ› ъč taʤ r ‖ t f r, tⁱf r  < ŋč‘ϑu -; Ave. č‘ϑ -, č‘ϑ ā› , Khʷāʲ. c”ʾ› /ϑaʤāʲъч Bactr. / Khōtщ tc ›‘-, tcohora-, tcahora-, Oss. c pp‘› ‖ cuppar, Shugh. Ba₍ūщ c‘v ›, c‘v ›, οōsh. c‘v ›, Bart. οāshrv. cav r, πaʲХqщ cavur, Wakh. cə’ r, c ’ r, Yazgh. č“›, Ishk. cь”u›, Sangl. cə” ›, Munj. čfir / č(ⁱ)” ›, Yidgh. č ›, Pasht. cal r, Tjk. dial. Takfōn (arch.) ф ; Pers. č‘h r, ·168· Tjk. č › (lit. č‘h r), Fārs. čäh r, colloq. čār, AfghP. č(‘h) r, Pahl. chʾl chʾl /čaʦāʲъч Parth. /čaʤāʲъч Kurd. ç‘›, Ved. c‘tvā›‘s, Hind. cā›; Eynu. č‘› 26. (155.) five panč  pnc pnc, pnž, pnj⃝ /paṁǰ/ < ŋpánč‘-; Ave. panča-, Khʷāʲ. pnc /pan /, Bactr. /pan /, Khōtщ p‘ṃjs‘, Oss. fonʒ, Shugh. οōsh. Bart. οāshrv. pДnʒ, πaʲХqщ pinʒ, Wakh. pānʒ, Yazgh. p“nǰ, Ishk. p nʒ, Sangl. p n₎, p nʒ, Munj. p nč, p nǰ, Yidgh. pān , Pasht. pinʒə; Tjk. dial. ρa₎ʤōn (arch.) п ; Pers. p‘nǰ, Kurd. pênc, Ir. хpanča-, Ved. p‘ñc‘; Eynu. pänǰ(ä) (156.) six uʸš  ⁽⁾⁽ ⁽, w ⁽ ⁽, ʾ ⁽ ⁽ w ⁽ ⁽, ʾ ⁽ ⁽ ⁾⁽ ⁽ /{əʸšú, ʸ{əšú/ < ŋ⁾ú u < *x á u < *(x) uá ‘m; Ave. ⁾ uu‘ -, Khʷāʲ. /uʸ, uxs-/, Khōt. k ä(täʼ), Oss. æ⁾sæ₎, Shugh. ⁾ , Ba₍ūщ οōsh. Khūʤщ ⁾ w, Bart. οāshrv. ⁾ȫ⁽, πaʲХq. xel, Wakh. ād / ā , Yazgh. ⁾u(w), Ishk. ⁾ ḷ , Sangl. ⁾āḷ, Munj. ⁾ ‘, Yidgh. u⁾ o, Pasht. p‘ǵ, Pers. ‘ , Tjk. ‘ , colloq. , TMast. ‘ , , ‘ , γāʲʳщ ä , colloq. , “ , Kurd. “ , Ide. *s(u)éks, Ved. a ; Eynu. ä (157.) seven ʾ t-ʾ, ʾ t-h ʾ t-ʾ, (ʾ)’t-ʾ aw ta /(əф dáъ avd / aft  < *haftą-; Ave. hapta-, Sarm. у ф-, Khʷāʲ. ʾ d /a dъч Bactr. ⃝ /ē ъч Khōtщ haudo, Oss. avd, Shugh. οāshrv. (⁽) vd, οōsh. Ba₍ūщ ⁽ vd, Bart. vd, πaʲХq. vd Yazgh. uvd, Ishk. vd, Sangl. vd, Wakh. ’, Munj. vdá, Yidgh. ávdo, Pasht. ⁽ə, Pers. haft, Tjk. colloq. haf, TMast. haf(t), Kurd. heft, Ved. s‘ptá-; Eynu. häp(t); cf. Bactr. /ē уuфdaὕъч εʣpʦtʦaὕʧtʣ (158.) eight ašt  ʾ t(ʾ), ʾ th ʾ t(ʾ) (159.) nau  (160.) das  t ṯʾ /aštч (əфštáъ < ŋá ta-; Ave. ‘ t‘-, Khʷāʲ. ʾ t, Bactr. /ata/, Khōtщ ha ṭa, Oss. ast, Shugh.-οōsh. ⁽‘⁾t, πaʲХq. ⁽o⁾t, Yazgh. u⁾t, Wakh. at, Ishk. ot, Sangl. t, Munj. ḱá, Yidgh. á čo, Pasht. atə, Pers. h‘ t, Tjk. colloq. h‘ , TMast. h‘ (t), Kurd. h“ t; Ved. a ṭá(u); Eynu. hä (t) nine nw, n⁽ʾ nwh, n⁽ʾ n⁽ʾ /nau, nōч n(əфʷáъч < ŋn‘u‘-; Ave. nauua-, Sughn. n ⁽, οōsh. Bart. οāshrv. nā⁽, πaʲХq. new, Wakh. nāw, Yazgh. nu(w), Ishk, naw, nu, Sangl. n ⁽, Munj. naw, Yidgh. now, Pasht. nə, Khōt. nau, Pers. nu(h) < n , Tjk. n h, TMast. n , nu, TFalgh. nu, TVarz. nuh, n h, AfghP. noh, colloq. n , γāʲʳщ noh, Pahl. naum, Kurd. ne, Ved. náv‘; Gre. , Armen. inn; Eynu. noh ten s(ʾ), sh s(ʾ) dsʾ / əʳуáфъ < ŋdáʦą-; Ave. dasa-, Khʷāʲ. s, Bactr. /las/, Khōtщ dasau, Oss. dæs, Shugh. Дs, οōsh. os, Bart. οāshrv. us, πaʲХq. es, Wakh. as, Yazgh. s, Sangl. d s, Yidgh. los, Pasht. ·169· Wa . las, ἵaʲāch. dȫs, Pers. dah, TMast. TFalgh. da, TVarz. da(h), Pahl. dah, OPers. *daϑa-, Kurd. deh, Ved. dáś‘; Gre. ч Armen. tasn, OCS. d“sętь, Lat. decem, Goth. tá hun; Hung. tíz < Scyth.?; Eynu. dah, däh (161.) eleven y ₎dáʰ  ywnṯs(nw), ywṯsnw ṁi(nu)/ < ŋ‘ u‘(n)-daʦą-(‘nām-); Ave. ‘ēuu‘nd‘s‘-, Pers. ₍ā₎dáh, TMast. ₍o(n)₎dá, ₍ n₎dá, γāʲʳщ colloq. y ₎ä. Kurd. yanzdeh (162.) twelve d w ₎dáʰ  ⁽ʾts d⁽ʾṯs ъ ʷāl/ < ŋduuā-daʦą-; Ave. duuadasa-, Pasht. , Pers. duvā₎dáh, γāʲʳщ däv zdäh, colloq. däv zä, TMast. d vo(n)₎dá, d v n₎dá (163.) twenty b st  wysṯ / < ŋu ʦ‘tД < ŋu nʦati; Ave. vДs‘ⁱt -, Khʷāʲ. ʾws c /əws(e) , ūʳуʣф /, Bactr. /w st/, Khōt. ’ stä, Oss. ( )ssæʒ ‖ nsæ₍, Sarm. Ἰ [ ], Wakh. wДst, Yazgh. wast, πaʲХq. vist, Sangl. ⁽ t, Yidgh. wisto, Pers. ’Дst, Tjk. , Pahl. vДst, Kurd. ’îst, Baὕōch. gДst; Ved. viṃś‘tí, v ñś‘tí, Armen. ksan, Gre. ; Eynu. bist (164.) (one) hundred (yak)sád  sṯ-w /sətúъч < *ʦatam-; Ave. satəm-, Khʷāʲщ , Bactr. /sad/, Oss. sædæ, πaʲХqщ sad, Pers. ⃝ ч (yak)s‘đ, OPers. ϑata ; Ved. ś‘tám, Ide. *(h )kmtóm, ŋdkmtóm, Lat. centum, Gre. OCS. s to; Cr.Goth. sada; BukhAr. s t, Eynu. säd III.3. Adjectives (i) 27. (142.) big kátta  < Uzb. kȧttȧ, Uygh. katta, Kyrg. kette, Tatar. kättä, ἶašqщ kȧtȧ, Baš₎щ kəttə, Chŭvash k‘čč‘, TMast. TVarz. k‘ttá, AfghP. k‘ttá, εazāʲщ kaṭá, Shugh. οōsh. katta, katanak < IAr. k‘ttā- ???; cf. Gre. Κ ч name of Bactrian nobleman (4th century ), the word can be of Bactrian origin and in tan explain etymology of Tjk. kal n kal n (occ.)  < Tjk. kal n, TMast. k l n, Parth. k‘lān < Bactr. ??? b ₎ú›g  ⁽₎ʾ›k /wəzáṙk/ wzrg /wəzáṙg/ < Pers. ’u₎ú›g, Pahl. wcwlg /wa urg, wa arg/, OPers. v‘₎›k‘-, Māzand. bazarg, Bactr. /wazurk/; Ott. ’ ₎ ›g, Elam. azzaka, haz(z)ak(k)a 28. (134.) long van(n)  ʾ n-ʾy bn / níчл ənуíф/ ·170· ’‘lánd ʦʧʥʦчлὕonʥ л rz ъ əwzíъ ʦʧʥʦчлὕonʥ < ŋ’›ʣa-; Ave. bərəz-, barəz-, Bactr. - / uʲz-/, Khōtщ bulysa-, Yazgh. vəz, Shugh. v ʒ, οōsh. v ₎, Wakh. v ›₎, Ishk. vь d k, Sangl. və duk, Munj. v‘ńǵ, Yidgh. väṇ, Pasht. (⁽) ǵd, OPers. personal name B›d ₍‘; cf. Pers. ’ulánd, Tjk. ’‘lánd, γāʲʳщ ’oländ, εazāʲщ ’ lán < ŋ’›ʣánt- and Pers. [Al]’ú›₎ Aὕϐoʲzл mountaʧnʳ чл γāʲʳщл [λl]’ó›₎ (< Pahl. Harburz < Ave. ђ‘›ā Bərə₎‘ⁱtД), Wakh. bland; Turkm. belend; Ved. ’›hánt-; cf. Khʷāʲщ žk (m) žϑ (f) / ažʣʥ- : ažʣ -/ d ᵎr ὕonʥчлʤaʲ  wr(h) wr dwr / ūʲъ ὕonʥчлʤaʲ < ŋd ›‘-; Ave. d ›‘-, Khōtщ dura-, Wakh. ir, πaʲХq. ar, Pers. d ›, TMast. d ›, dir, TFalgh. dir, OPers. d ›‘-; Ved. d ›á-, Hind. d ›  < Pers. , Tjk. чл γāʲʳщл d“› z, Ave. d›āǰ‘h-чл πaʲХqщл d‘›ú₎чл ἵaʦὕщл Baὕōch. d›āǰ, Kurd. d ›êj mzyx /mə  m₎ʾ₍ (h) mzy(y)x, mzy cf. Ave. maziia29. wide ya d, yaxt  y (ʾ)›t-y, yr t y rṯ-y /yəуwф dí/) < ŋu -g›t‘paʰm  p nʾ₍ /pa nēъ < *paϑana-; Ave. paϑana-, Oss. ”ætæn ‖ ”‘tan, Pasht. plan, Pers. pahn, Pahl. pahan, p‘hnāД, p‘hnāk, Kurd. pan, Baὕōch. patan 30. thick ”‘›’éhж  rpy /frəpХʸ/ Ave. pДv‘h, Pers. ”‘›’íh, TMast. ”‘›’í, Pahl. ”‘›’Дh sw / əʤʳúъ aʤʳ  Tjk. afs 31. (144.) heavy wazm n  < Ar. WZN, Pers. vazn n, TMast. v‘₎mín ›ʾn(h) ›ʾn ъ ʲānъ g‘›áng  cf. Uzb. gȧrȧ , Tjk. g‘›áng, εazāʲщ g ›án(g), g ›án(k) 32. (143.) small púὕуὕфa  < Yagh. púl(l)‘ ϑʦʧὕd л < *puϑra- son , Ave. puϑra-, Sogd. ⃝py ›ʾk, ⃝p r ⃝py ›ʾk, ⃝ ⃝ p ₍ (as a part of compounds) /pɨš , Khʷāʲ. (ʾ)p›, Scyth. *purϑa-, Bactr. p r, ⃝p ₍ у ф /p (h)r/, Khōtщ p ›‘-, Alan. ч Sarm. *furϑa-, Oss. ” ›t ‖ furt, Shugh.-οōsh. puc, πaʲХq. p c, Yazgh. poc, Munj. p ›, Yidgh. p ›, p ḷ, Wakh. pətr, ἵaʲāch. pu , Pers. p sá›, p ›, pus(á›), γāʲʳщ pesär, Tjk. p sá›, AfghP. p“sá›, Pahl. pus, puhr, OPers. puç‘-, ·171· Med. *puϑra-, Kurd. pisir, Baὕōch. pʰus‘ , Parth. p ›; IIr. ŋput›á-, Ide. ŋputló-, Ved. put›á-, Pāὕʧ putta-, Hind. p t, Bengāl. put; cf. Lat. puer m‘₍dá  < Tjk. m‘ dá, Uzb. mȧydȧ, Kyrg. mayda  ›₍ncʾk(k), ryncyk ryncwk(k), ryncwk ›₍nʾ‹ ъʲíṁǰ ʲíṁǰ ʲíṁǰ / < ŋ›‘nǰ -ka-, ŋ›‘nǰ‘-ka-ka-, ŋ›‘nǰu-ka-; Ave. rənǰ ‘-, Khʷāʲщ rnc, Khōtщ raysga, Pasht. rangay 33. (135.) short k‘ltá  < Uzb. kȧltȧ, Tjk. k‘ltá  mwrzk-y ъmuwz₎íъ < ŋm›ʣuka ›ʾ⁽ ъ ʲūš ~ ʲōšъ  snʾ› ъʳnāʲъ < ŋsnā›‘-; Wakh. sən › 34. narrow ’ ›k < Pers. ’ā› k tank, tang  < Pers. tang, Pahl. t‘ng(Дh), Ave. taṇč t‘-, Wakh. πaʲХqщ tang, Kurd. teng, Baὕōch. tank, Chaghat. tä , Uzb. tȧ 35. thin tank, tang  < Pers. tang, Pahl. t‘ng(Дh), Ave. taṇč t‘-, Wakh. πaʲХqщ tang, Kurd. teng, Baὕōch. tank, Chaghat. tä , Uzb. tȧ t núk, t nukák  < Tjk. tunúk, Ishk. tьnьk, Oss. tænæg, πaʲХqщ t‘n k, Kurd. tenik, Baὕōch. tanak; Ved. t‘nú-, tánuk‘(145.) light s‘’úk, s ’úk  < Tjk. s‘’úk, εazāʲщ su’úk III.4. People 36. (103.) woman za(ᴥ) f(a)  < Ar. α[F da Д”‘ẗ, Pers. ₎‘ Д”á ʷʣa₎ > BukhAr. ₎‘ Д”‘, TMast. zaᴥí”, Tjk. dial. Chust, Ůʲōteppa ₎‘í”, Wa . zaypa, zaypə ʷoman ·172·  ʾstʾ₍›ch (ʾ)st›₍c, stryc sṯryc /ᶤʳtʲХčъ < ŋst›Д-kā-; Ave. st›Д-, Ishk. ьc ʤʣmaὕʣлanʧmaὕ члYazgh. ⁽“nǰ, Shugh. ⁽ānДc ϑaὕʤ (f) ; Oss. ⁽æn g ‖ ⁽ænug ϑaὕʤч ϐuὕὕoϑ₎ 37. (102.) man m rti  mrty mrtyy, mrṯyy ъmáṙ < ŋmá›t ‘-; Ave. m‘ ‘-, Khʷāʲ. mrc(y), Bactr. /mard/, Munj. m‘ṛ‘, Pers. mard, OPers. martiya-, Kurd. mê›, Ved. má›tiya- < Ide. ŋm›to- moʲtaὕ ч Gre. ч 38. human dám  < Ar. ād‘m, BukhAr. ādəmi, Hebrew adam, Pers. ādám, Oset. ‘dæm, Ishk. odam, Shugh. d‘m, Tr. adam, Turkm. ād‘m, Tatar ‘däm, Chŭvash etem ʾ ʾm ʾdm ъĀdamъл Adam л<лAʲщлĀd‘m, Pers. Ādám, Yagh. dám etc. m›tʾ⁾mk mrt mʾk(⁽), mrt mʾ₍ mrtxmy(y) mrṯxmy, mrdxmy ъmáṙ mardʉm  < ŋm ›t ‘-táu⁾m‘n-(ka-); Pers. m‘›dúm, Shugh. mardum, Ishk. m‘›dьm 39. (104.) child ⃝ ⃝ púὕуὕфa  py ›ʾk, p r ⃝py ›ʾk, ⃝p r, ⃝p ₍ ⃝p ₍ (as a part of compounds) /pɨš ʳon < *puϑra- ʳon ч Ave. puϑra-, Khʷāʲщ (ʾ)p›, Scyth. *purϑa-, Bactr. у ф /p (h)r/, Khōtщ p ›‘-, Alan. ч Sarm. *furϑa-, Oss. ” ›t ‖ furt, Shugh.-οōsh. puc, πaʲХqщ p c, Yazgh. poc, Munj. p ›, Yidgh. p ›, p ḷ, Wakh. pətr, ἵaʲāch. pu , Pers. p sá›, p ›, pus(á›), γāʲʳщ p“sär, Tjk. p sá›, AfghP. p“sá›, Pahl. pus, puhr, OPers. puç‘-, Med. *puϑra-, Kurd. pisir, Baὕōch. pʰus‘ , Parth. p ›; IIr. ŋput›á-, Ide. ŋputló-, Ved. put›á-, ἵāὕʧ putta-, Hind. p t, Bengāὕщ put; cf. Lat. puer g dák  Pers. k đák, Tjk. k dák, TVarz. gudák, AfghP. k dák, γāʲʳщ k däk, Pahl. k⁽tkʾ qwtk /₎ō aʥъч Uzb. g dȧk, Uygh. gödäk, Ott. k d“k, Tr. (arch.) kûd“k ”‘›₎ánd  < Pers. ”‘›₎ánd, Pahl. frazand, farzand, Parth. frzynd, Bactr. ч ч ч ч Ave. frazaⁱṇti-, Ir. *fra-zanti40. (114.) wife ʾync(h) ync ʾync ъ|ɨṁǰ/ ʧnč  ʾync(h), ynch < ŋ áun -kā-; Yazgh. ⁽“nǰ, Shugh. ⁽ānДc ϑaὕʤ (f) ; Oss. ⁽æn g ‖ ⁽ænug ϑaὕʤч ϐuὕὕoϑ₎ ay l  < Ar. ayāl, Pers. ay l 41. (113.) husband wyr ъʷХʲъ ʷХʲч vХʲ  < ŋuД›á-; Ave. vД›‘-, Pahl. ⁽Д›, Scyth. ч Ved. vД›á-, Ide. ŋu ђ›ó-s, Lat. vir, OIrl. pl. , Irl. Gael. fear pl. fir, Welsh gŵ› pl. g⁽ŷ›, Bret. gour, Lith. vý›‘s, Latv. vД›s, Goth. wair, OEng. wer, OScand. verr, cf. Engl. (arch.) wer(e), Ger. Wehr 42. (106.) mother ·173· čá  < Uzb. ȧčȧ, čȧ, Turkm. “ǰ“, Tjk. čá, ‘čá m dá›  mʾt mʾṯ /māt/ ŋmāt‘›-; Ave. māt‘›-, Khʷāʲщ mʾd /mād-a/, Bactr. /mādъч Khōtщ māt‘, māvä, Oss. mad ‖ m‘dæ, Shugh. Bajū. m d, Bart. Khūʤщ οōsh. οāshrv. mud, Munj. mā₍ā, Pasht. m ›, Pers. māđá›, OPers. ŋmāt‘›-, Pahl. māt, māt‘›-, Baὕōch. māt, Ide. *meh t r-, Ved. ᾱ ч OCS. matь, OEng. m dor, OIrl. má ; Eynu. m‘dä›, mātá›-, Gre. mėdä› 43. (105.) father d d  < Tjk. d d , d‘dá, d‘dá; Uzb. dada; cf. γāʲʳщ b b ’‘’á padá›  ʾptr-y (ʾ)ptr-y(y) (ʾ)pṯr-y / pt(ə)ʲí/ < *pitar-; Ave. (p)tā (nom.sg.), Khʷāʲщ pc /pica/, Bactr. /pid/, Khōtщ pät‘›-, Oss. ” d ‖ fidæ, Shugh. ped, Khūʤщ οōsh. οāshrv. Bart. pДd, πaʲХqщ pit, Pasht. plā›, Wa . p₍ā›, Pers. p đá›, Tjk. p‘dá›, p dá›, TMast. pədá›, γāʲʳщ p“där, pädär, Pahl. pit(ar) > pi ar, ч Armen. hayr, Eng. OPers. pitar-, Baὕōch. pit, pʰ s, pʰ ϑ, Ide. *p(ə)h t r-, Gre. father, OEng. fæder, OIrl. a ; Eynu. p‘dä›, pėdä› (107.) older brother ak , ‘ká  < Uzb. ȧkȧ, Uygh. aka, Tjk. ‘ká, ak , TMast. ‘kó, Shugh. ‘kā, Tr. ‘ğ‘, Kyrg. Kazakh. ἶāʲaqaὕp. a a, BukhAr. ak (108.) younger brother ›ʾt ›ʾt, ʾ ›ʾt› ›ʾt ’›ʾṯ / ʲātуər)/266 vⁱr t  < ŋ’›āt‘›-; Khʷāʲ. ›ʾd / ʲādъч Bactr. у ф / уəфʲādъч Khōt. ’›āt“, Tumshuq. ’›āḏe, Wakh. v› t, Yazgh. v(ə)›ád, Shugh. οōsh. v › d, Ishk. vru(d), Sangl. vrud, Pasht. ⁽› ›, Pers. ’ ›ādá›, Tjk. ’‘› dá›, TMast. b ›odá› (> Yagh. b › dá›), γāʲʳщ ber där, εazāʲщ bir r, Kurd. bera, Ide. ŋ’ʰ›āt‘›-, Ved. ’ʰ›át‘›-, Gypsy pʰ›‘l, OCS. bratr , OIrl. á , Welsh brawd, OEng. b› ðor, Lat. ”›āt“›; Oss. æ›v‘d ʲʣὕativʣ ‖ æ›vadæ ϐʲotʦʣʲч relativʣ ; Gre. ɪ ᾱ mʣmϐʣʲ of a ϑommunʧty d dá› (occ.)  < Tjk. d dá› (109.) older sister ‘p(p)á  Uzb. p‘, Kyrg. apa, Tjk. ‘pá, TMast. ‘pá, BukhAr. ap (110.) younger sister 266 Meaning both older and/or younger brother in Sogdian. ·174· ʸōʲ  (111.)  ⁽ʾ›h ⁾⁽ʾ› /ʸ°āʲъ267 < ŋhu‘h‘›-; Ave. x ‘ h‘›, Khʷāʲ. ʾ xᵃ, Bactr. /ʸ{āʦъч Oss. xo ‖ ⁾ æ›æ, Yazgh. ⁾°‘›ǵ, Ishk. ⁾ó, Pasht. ⁾ ›, Pers. x āhá›, TMast. ⁾ (v)á›; AfghP. ⁾ r, εazāʲщ ⁾( ) ›, Pahl. xwah, Parth. ⁽⁾ʾ›; Ide. ŋsu“s ›, sister, Ved. svás‘›-; Mid. and Mod. Welsh chwaer; Mid. Bret. hoer, hoar cʼho‘› hoé›; OCorn. huir; Corn. hwoer, OIrl. ; Manx shuyr, Ger. Schwester; BukhAr. ḫ ḫar son < past part. of the verb ž - to livʣ ч Sogd. √ʾ₎⁽(-) √(ʾ)₎⁽(-) √j⁽(-) √ž⁽(-) /√ žūч žau-/, Ave. ǰ(a)uua- ???  ₎ʾt(ʾ)k ₎ʾt₍(₍) ₎ʾṯy /zātēъ < *ʣāt‘-ka-; Ave. ₎āt‘-, Khʷāʲщ , Bactr. /zādъч Pasht. ₎ ₍, Pers. ₎āđá, Ide. ŋgnh tópúὕуὕфaл ϑʦʧὕd л ⃝py ›ʾk, ⃝p r ⃝py ›ʾk, ⃝p r, ⃝p ₍ ⃝p ₍ (as a part of compounds) /pɨš < *puϑra- ʳon ч Ave. puϑra-, Khʷāʲщ (ʾ)p›, Scyth. *purϑa-, Bactr. у ф /p (h)r/, Khōtщ p ›‘-, Alan. ч Sarm. *furϑa-, Oss. ” ›t ‖ furt, Shugh.-οōsh. puc, πaʲХqщ p c, Yazgh. poc, Munj. p ›, Yidgh. p ›, p ḷ, Wakh. pətr, ἵaʲāch. pu , Pers. p sá›, p ›, pus(á›), γāʲʳщ p“sär, Tjk. p sá›, AfghP. p“sá›, Pahl. pus, puhr, OPers. puç‘-, Med. *puϑra-, Kurd. pisir, Baὕōch. pʰus‘ , Parth. p ›; IIr. ŋput›á-, Ide. ŋputló-, Ved. put›á-, Pāὕʧ putta-, Hind. p t, Bengāl. put; cf. Lat. puer (112.) daughter ay₎  cf. Yazgh. ‘č‘ǵ, Shugh. āc, οōsh. ac, πaʲХq. oc du⁾tá› (occ.)  w r wxth, wth wth w t(ʾ) dw ṯ(ʾ) / ə {dуáфъ < *duxtar-; OAve. dugədar-, YAve. du ar-, Khʷāʲщ dᵃ / u daъ, Bactr. ъὕu dуaфъч Khōtщ dutar-, Yazgh. ə d, Ishk. ⁽ d d, Ḳʧdʥʦ. lu do, Pasht. l ›, Pers. du⁾tá›, TMast. d ⁾tá›; Ved. duh tá›-; Eynu. tu⁾tä› (115.) boy ⃝ ⃝ py ›ʾk, p r ⃝py ›ʾk, ⃝p r, ⃝p ₍ ⃝p ₍ (as a part of compounds) /pɨš púὕуὕфa  < *puϑra- ʳon ч Ave. puϑra-, Khʷāʲщ (ʾ)p›, Scyth. *purϑa-, Bactr. у ф /p (h)r/, Khōtщ p ›‘-, Alan. ч Sarm. *furϑa-, Oss. ” ›t ‖ furt, Shugh.-οōsh. puc, πaʲХqщ p c, Yazgh. poc, Munj. p ›, Yidgh. p ›, p ḷ, Wakh. pətr, ἵaʲāch. pu , Pers. p sá›, p ›, pus(á›), γāʲʳщ p“sär, Tjk. p sá›, AfghP. p“sá›, Pahl. pus, puhr, OPers. puç‘-, Med. *puϑra-, Kurd. pisir, Baὕōch. pʰus‘ , Parth. p ›; IIr. ŋput›á-, Ide. ŋputló-, Ved. put›á-, Pāὕʧ putta-, Hind. p t, Bengāl. put; cf. Lat. puer 267 Meaning both older and/or younger sister in Sogdian. ·175·  < past part. of the verb ž - to livʣ ч Sogd. √ʾ₎⁽(-) √(ʾ)₎⁽(-) √j⁽(-) √ž⁽(-) /√ žūч žau-/, Ave. ǰ(a)uua- ???  ₎ʾt(ʾ)k ₎ʾt₍(₍) ₎ʾṯy /zātēъ < *ʣāt‘-ka-; Ave. ₎āt‘-, Khʷāʲщ , Bactr. /zādъч Pasht. ₎ ₍, Pers. ₎āđá, Ide. ŋgnh tó(116.) girl ay₎  cf. Yazgh. ‘č‘ǵ, Shugh. āc, οōsh. ac, πaʲХqщ oc w r wxth, wth wth w t(ʾ) dw ṯ(ʾ) / ə {dуáфъ du⁾tá› (occ.)  < *duxtar-; OAve. dugədar-, YAve. du ar-, Khʷāʲщ dᵃ / u daъч Bactr. ъὕu dуaфъч Khōtщ dutar-, Yazgh. ə d, Ishk. ⁽ d d, Ḳʧdʥʦ. lu do, Pasht. l ›, Pers. du⁾tá›, TMast. d ⁾tá›; Ved. duh tá›-; Eynu. tu⁾tä› III.5. Animals 44. animal ayv n  < Ar. YY ‘₍⁽ān, Hebrew ayah, Syr. aywat , Pers. ‘ v n, Oss. ⁾á₍⁽‘n, Uzb. hayv n ayv n t ʤauna  < Ar. YY ‘₍⁽ānāt (sg. ‘₍⁽ān) Pers. pl. ‘ vān t, εazāʲщ ‘₍⁽ n t -ʉ) ǰ  < Tjk. ǰ , TMast. ǰ ndó› ǰ‘›má›  < Tjk. ǰ‘›má› t-w ъ ətúъ  Ave. daitaka ʾstwrp ʾy, ʾstʾ⁽›p ʾʾy, ʾstʾ⁽›p ʾk stwpr y / ə < ŋst‘u›‘-pada-ka-; cf. Sogd. ʾstʾ⁽›(h) ϑattὕʣ ч Yagh. s ʳʦʣʣp 45. (86.) fish  m , < Pers. , Tjk. , , TMast. m (₍)í, TVarz. m í, ἰāϐuὕщ m ₍ , εazāʲщ m í, Pahl. mʾh₍g /māʦХʥъч OPers. *maϑya-(ka-), Shugh. m ₍ , Wakh. mo(h)í, m‘hí, mo₍í, Parth. ʦ ‘-; Ved. máts₍‘mʾs₍ʾg, Kurd. m‘sî, Ir.  kp-y /kəpíъ < -; Khōtщ k‘vā-, Khʷāʲщ , Scyth. ( ) , Oss. kæ”, Wakh. k p, Munj. k p, Pasht. kab 46. bird ·176· mʉr  mr -y (ʾ)m› -y / m(əwф íч məw íъ < ŋm g‘-; Ave. mərə a-, Khʷāʲ. (ʾ)m -, Bactr. /mʧʲ ъч Khōtщ mura-, Oss. mar , Pers. mur , TMast. m › , εazāʲщ murq, Parth. mwrg, Ishk. mь› ; Ved. m›gá-; p‘›(›)‘ndá  Pers. p‘››‘ndá, Wakh. prinda, Shugh. p‘› ndā, p‘›‘ndā ǰ‘›má›  < Tjk. ǰ‘›má› (ǰ -ʉ) ǰ  < Tjk. ǰ , TMast. ǰ ndó› ‹  < Uzb. ‹u , Tr. ku , ρüщ ŋ‹uś; Chaghat. ‹u anʧmaὕ s ča  s₍cʾkk sycy < *ʦДk‘-ka-, *ʦ ‘kā-ka(87.) chicken mr -y (ʾ)m› -y / m(əwф íч məw íъ mʉr  < ŋm g‘-; Ave. mərə a-, Khʷāʲщ (ʾ)m -, Bactr. /mʧʲ ъч Khōtщ mura-, Oss. mar , Pers. mur , TMast. m › , εazāʲщ murq, Parth. mwrg, Ishk. mь› ; Ved. m›gá-; č ǰá  c⁽₎ʾkk / Khʷāʲщ t⁽žk, Yazgh. č kg, Wakh. č č‘, Yidgh. čuž ₍‘, Pasht. ču əka, Pers. č ǰá, γāʲʳщ ǰ ǰé, ǰouǰé; Uzb. ǰ ǰȧ, Tr. c c“, Qashq. ǰ ǰȧ 47. (95.) dog kut  ʾkwt-y kwt-y, qwt-y / ₎{ətíъ < ŋkút‘-, ŋkutД-; Bactr. /kud/, Oss. k ʒ ‖ kuy, Yazgh. k°od (fem. kid), Shugh.-οōsh. kud (f. kid), πaʲХq. k d, Ishk. kьd, Sangl. kud; Tjk. colloq. kučák, εazāʲщ kuṭá, Ir. *kuta-, х₎utХ-; Hind. kuttā, Tokh. ku ʲáužna, ʲáuǰna  < pres. part. of the verb ›‘uǰ-, ›‘už- to ϐaʲ₎ ч Sogd. √› z- /√ʲə ž-/, Munj. rav- : rivd(89.) cow ōu  ʾw w / āu/ < *g ua-; Ave. - (nom. gəu ), Scyth. *g u-, Khʷāʲщ wk / ō₎ъч Bactr. уф ʾʾw / āʷъч Khōtщ gg hД, Oss. qug ‖ og, Shugh. οōsh. Khūʤщ ž ⁽, Bart. ž‘⁽, οāshrv. žā⁽, , Yidgh. πaʲХqщ ž w, ž‘⁽, Yazgh. w, Wakh. ⁽, Ishk. u, Sangl. u , Munj. avo, Pasht. ⁽ā, ἵaʲāch. g , Ōʲm. g Д, Pers. gāv, TYagh. TFalgh. TVarz. gou, εazāʲщ gaw, Pahl. gāv, g , OPers. ŋg‘u- (ё‘u’› v‘- = Γ ύ фч Kurd. ga, Baὕōch. g k, ρāὕysh. gug; Ide. *g u-s, Ved. go-, gau-, gāv-, Gre. ч Lat. ’ s, Armen. kov, OScand. kýr, OEng. c , cȳ, Eng. cow, dial. kye (pl. kine), OHG. chuo, Ger. Kuh, Irl. ’ó, OCS. gov[ędo] ·177· kⁱš k ϐuὕὕ  < ŋk› -āk‘- ϐuὕὕ // ŋk‘u - -/ - - ϑoʷ ; Bactr. ϸ /ʥāʷ ₎ʧšāʥъч Ishk. kьž k, Sangl. kuǰ k, Munj. , , ἵaʲāch. kâ ‘g ; Sarghul. k ó ϑoʷ (90.) buffalo (94.) goat (pl.); ʾ zynch / zХṁǰ/ ₎ʧd v z  ’₎₍ ṯ /( < ʣa-; Ave. ’ ₎‘-, Khʷāʲ. ʾ z /a zaъч Khōt. buysa-, Yazgh. Shugh. οōsh. vaz, Ishk. vь₎, Munj. vəza, Pasht. wuz (f. wuza), Pers. buz, TFalgh. ’ ₎, Pahl. vuz, ḳázáщ ’ı₎“; Thrac. buza (97.) monkey maym n  < Pers. m‘ m n, Oss. maymuli, Kyrg. m‘₍mïł, Tatar. m‘₍mıl, Tatar. dial. mä₍mun, Uygh. maymun, MGre.  mkkr(ʾ) mk›ʾ ъma₎₎á ṛ уáфъ < Skt. markaṭa-, Prkt. makkaḍa- > Khōtщ makala-, Khʷāʲщ mrk 48. louse š púš, šⁱpúš  p h ъšpəšáъ < *ʦuí ‘-; Ave. p -, Khʷāʲ. spʾh, Oss. s st ‖ s stæ, Yazgh. səpaw, Shugh. s pá , οōsh. sipaw, πaʲХq. spal, Ishk. s(ь)puḷ, s(ь)pьl, Wakh. , Munj. s(ᵊ)pəyă, Yidgh. sp o, p o, Pasht. spəǵa, pəǵa, ἵaʲāch. “sp , Ōʲm. sp Д, Kurd. sipi, κāzand. s”Дǰ, Pahl. sp , Tjk. upú , u’ú (k), ‘’ú (k), TMast. s ’ǘs, εazāʲщ pí 49. (96.) snake m › < Pers. mā›, Kurd. mar kíʲуⁱ)m  kyrm-y qyrm-y ъ₎ʧwmíъ < ŋk›m -, Oss. kalm ‖ kælmæ, Pers. kirm ʷoʲm ; Ved. k m 50. worm ₎íʲmá₎  < ŋk›m - + diminutive suffix -ak (98.) mosquito / fly pá ( )á fly  < TMast. TVarz. p‘ á, TBuch. p‘ ‘ fly ; Tjk. p‘ á moʳquʧto púžna, púǰna fly  ǰ nču›ák moʳquʧto  < Tjk. ču›(ču)›ák ʷʦʧz  m⁽⁾ k moʳquʧto < ŋm‘⁾ k‘(99.) ant ·178· m ›č‘k  ₎mʾ⁽›c, ₎mʾ⁽›ʾk ъzmōʲčч < ŋ(₎)máu› -ka-(ka-); Ave. maoⁱri-, Oss. mælʒ g ‖ mulʒug, Pasht. mēǵay, Wa . mē›ž‘ , Pers. m ›čá, Tjk. m ›čák (100.) spider w fkak  < derived from verb ⁽ ”- : to ʷʣavʣ ч Sogd. √⁽ʾ” /√ʷāʤъ : √⁽”t-, Oss. ⁽‘” n, Pers. ’ā”tán : ’ā”vallinkáч vanlinká, vanp da  (Sogd. ) ὕʣʥ , i.e. ὕonʥ-ὕʣʥʥʣd < van(n) ὕonʥ + link (< Turkic?) / t ›t‘nák  < Tjk. t ›t‘nák < Tjk. t › ʷʣϐ III.6. Plants 51. (61.) tree d‘›á⁾t / dⁱ›á⁾t  < Pers. d ›á⁾t, Tjk. d‘›á⁾t, TMast. də›á⁾t, Wakh. d‘›á⁾t, Shugh. d ›á⁾t; Uzb. daraxt, Kyrg. daraq  ⁽n(ʾk)h ⁽nʾ /wənáъ < ŋu‘nā-; Ave. v‘nā-, Shugh. ⁽ān ʷʣʣpʧnʥ ʷʧὕὕoʷ ч Pasht. wəna, ⁽ún‘, ἵaʲāch. an oa₎ ; cf Ishk. [čь]⁽“n apʲʧϑotч apricot-tʲʣʣ 52. forest mr /maṙ ъ mʣadoʷч ʤoʲʣʳt maʲ ʥʲaʳʳ  mr h < ŋmá›g‘- mʣadoʷ ; Ave. marə a-, Bactr. /maʲ ъ mʣadoʷ ч Sangl. mē› , Yidgh. mД› o, Pasht. mar a, Tjk. mar , mar mʣadoʷ ǰ‘ngál  < Pers. ǰ‘ngál, Shugh. ǰ ngāl, Hind. jaṅgal, Pali. Prkt. jaṅgala, Eng. jungle, Ger. Dschungel ⁽ntʾk(h) /wənd  < ŋu‘nā- tʲʣʣ 53. stick šōʸ  ʾ h ъšāʸъ < ŋ ā⁾‘-; Wakh. ⁾, Pers. ā⁾, Parth. ʾ⁾ ʷʦʧp ápp‘  ⁾⁽ ₍p /ʸ° < Pers. ‘ppá, TMast. ‘p(p)á < ŋ⁾ u‘ p‘- ʷʦʧp ; Ave. ⁾ uu‘ē aiiat ʷʦʧp , οōsh. ⁾‘’ēʒ ʷʦʧpч ʳtʧϑ₎ dōʲ₎ ʷoodч ʳtʧϑ₎  ʾ›(ʾ)⁽k(ʾ), ʾ›ʾ⁽kh ʾ›⁽k(ʾ) dʾ›⁽‹ ʷood <* -< -ka- ʷood ; Yazgh. erk, Shugh. ›g, οōsh. ›g, Ishk. dь›k, Sangl. durk, Pasht. largá₍, Wa . l“›gá, Parth. dʾl⁽g, Pers. dā› ʷoodч tree, pillar 54. (66.) fruit ·179· mēv‘g , mēvá  m ʾk /mə < *migda-ka-; Pers. mēvá, Pahl. mē (ag), Parth. mygdg ъmʧ уaʥфъ; Baὕōch. nД⁽‘g, nД’‘g; Uzb. mėvȧ, Tr. meyve, Azərb. meyvə 55. seed táʸуⁱ)m, tú⁾( )m  t m-y ṯxm-y ъtoʸmíъч t my ṯ(w)xmy < *tao⁾m‘-(ka-) < ŋtáu⁾m‘n-; Ave. taoxman-, Bactr. /tuxman/, Wakh. ta m, Ishk. tь⁾m, Pasht. t m‘, Pers. tuxm, TMast. t ⁾m, εazāʲщ tú⁾ m, Pahl. t m, Parth. tw(x)m /tōуʸфmъч OPers. t‘umā-, Kurd. tom; Ved. tókm‘n56. (62.) leaf barg  wrkr wrqr /ʷáʲ₎aʲъч < ŋu‘›k‘-; Pers. barg, εazāʲщ balk, Pahl. barg > Ar. WRQ ⁽‘›‘‹(‘ẗ) paʥʣ (of a book) , BukhAr. u‘›‘ḳa, Pers. v‘›á‹ (> Yagh. ⁽‘›á‹) 57. (63.) root r ‘ < Pers. ʲēšá, Ave. ›‘ē ‘58. bark pūst  pwst(h) ъpōʳtъ < ŋp‘u(a)sta-; Ave. pąst‘-, Shugh. p st, οōsh. Khūʤщ Bart. p st, πaʲХqщ past, Yazgh. pəst, Wakh. pist, Sangl. pask, Munj. p stá, Yidgh. pisto, Pers. p st, Pasht. Kurd. post; Skt. pustaka- ϐoo₎ p q < Uzb. p č ‹, Tjk. p č (64.) thorn ⁾ › < Pers. ⁾ā›, Pahl. ⁾ā›; Skt. kʰ‘›‘- ʳʦaʲp 59. (65.) flower gʉl ʲoʳʣч floʷʣʲ  wr /waṙ ъ ʲoʳʣ < u‘›d‘-, ŋu›d‘-; Ave. varəda-, Oss. wardi, Pers. gul, TMast. g l, gǝl, g l, Wakh. gul, gəl, Kurd. gul; Uzb. Tr. g l, Kyrg. g l, k l, Tatar. göl, NGr. ύ  ʾsp(ʾ)› my(y), ʾspr m(ʾ)k, spʾ› my ʾspr my(y), spʾ› my /ᶤʳpáṙ уə Ave. sparə a-, Parth. Pahl. ʾsprhm (67.) mango (68.) banana (69.) wheat (husked) ntm nṯm ъ áṁdəm/ ámtun, ántum  < ŋgántum‘-; Ave. -, Khʷāʲ. nd m, Bactr. / andumъч Shugh. ž nd‘m, Wakh ədim, Ishk undum, Munj. n)d m, Pasht. anəm, Wa . andəm, Pers. g‘ndúm, TMast. g‘ndǘm, Pahl. /gandum/, gnwm /gannum/; Gre. ч (70.) barley ·180· yau  yw-y /yəʷíъ < ŋ ‘u‘-; Ave. yauua-, Bactr. ч у ф /yaw/, ₍æu mʧὕὕʣt чл Shugh. ǰav, Wakh. ž‘⁽, žo⁽; Munj. you ʥʲaʧn ч Pers. ǰ‘u, Pahl. ǰaw; Ved. ₍áv‘(71.) rice (husky) ʲynϑ / ṁǰ/ bⁱ›ínǰ  < ŋu›Дʣi-; Ave. v“›“nǰ‘, Khʷāʲщ nc, Khōtщ ››Д₍s -, ››Д₍su‘-, Pasht. (⁽)›íž‘, Wa . ⁽›í₎‘, Wakh. gu›unǰ, Ōʲmщл›Дǰ‘n, Pers. ’ ›ínǰ, gu› nǰ, Pahl. ’› nǰ, Tālysh. birz, Sivandi. ’ ›ǰ ; Ved. v›ДhД-, Elam. mi-ri-zi- , Gre. ὄ ч ч Cze. ›ýž“, Eng. rice, Kāmvir. ⁽›új , Qashq. ’ › nǰ (72.) potato k‘›tu ká, k‘›t ká  < Rus. , ρǰ₎щ k‘›to ká, TVarz. k‘(›)tu ká, Kyrg. k‘›tö kö < Fr. cartouche (73.) eggplant groundnut (74.) (75.) chilli / pepper  Tjk. , , TMast. qəl‘m”ǘ›, qəl‘m”í› < Hind. ₎‘nǰ‘’ l  snkrpyl ъʳíṁgəwϐХὕъ ʥʧnʥʣʲ < ρǰ₎щ (regionally) ₎‘nǰ‘’ ʲʣd pʣppʣʲ < Pers. ʥʧnʥʣʲ ч Pahl. sngypyl ъʳʧnʥa ēʲъч Kurd. ₎“nc“”îl, Ujgh. ₎änǰ ⁽ l, Tr. zencefil, Ázʣʲϐщ zəncəfil, Ar. zanǰ‘’Дl, Gre. ч Mediaeval Lat. gingiber, zingiber < ἵaὕХщ siṅgiv ʥʧnʥʣʲ tumeric (76.) (77.) garlic kámčun (arch.)  cf. TMast. k‘mč ʷʧὕd onʧon sД›  < Pers. sД›, Kurd. sî› (78.) onion pⁱy z  p₍ʾk ъpyā₎ъ < Ir. ŋp āk‘-; Yidgh. pД , Wakh. p k, Yazgh. p ₍ē ; Bactr. /pʧyōzъ; Pers. piy z, Pahl. pa āʒ, Kurd. pîv‘₎; Uygh. piyaz, Kyrg. pï₍‘₎ (79.) cauliflower (80.) tomato , p‘m‘dó›  < Rus. п < Ital. pomi dʼo›o; TMast. ”‘m ldo›ǘ (81.) cabbage v z ūšá₎  < vʉz ʥoat + ʣaʲ 60. grass ·181· w š / ʷaʧš  ⁽₍ (h) ъʷēšъ < ŋu ʦt› ‘-; Ave. vāst›‘- paʳtuʲʣч pʲovʣndʣʲ ч Khʷāʲ. ⁽ , Bactr. ϸ ⁽ /w šъч Yazgh. ⁽“⁾ ʥʲaʳʳч ʦay , Shugh. οōsh. οāshrv. Khūʤ. ⁽ ⁾, πaʲХq. ⁽u⁾, Ishk. (⁽)u , Sangl. ⁽u , Wakh. ⁽ , Munj. wə , ⁽ , Yidgh. ⁽u , Pasht. ⁽ā⁾ə, ἵaʲāch. Д , Ōʲm. ⁽ā Д, Parth. ⁽ā pʲovʣndʣʲ maʲ  mr h mr /maṙ ъ mʣadoʷч ʤoʲʣʳt < ŋmá›g‘- mʣadoʷ ; Ave. marə a-, Bactr. /maʲ ъ mʣadoʷ ч Sangl. mē› , Yidgh. mД› o, Pasht. mar a, Tjk. mar , mar z r mʣadoʷ 61. (36.) rope w ta  < ŋuДt‘-ka-, Oss. ’ ₍ n to ϐʧnd vânt  ’ændæg, Pers. band < *banta-, Bactr. / andъч III.7. Body parts 62. (84.) skin pūst  pwst(h) ъpōʳtъ < ŋp‘u(a)sta-; Ave. pąst‘-, Shugh. p st, οōsh. Khūʤщ Bart. p st, πaʲХqщ past, Yazgh. pəst, Wakh. pist, Sangl. pask, Munj. p stá, Yidgh. pisto, Pers. p st, Pasht. Kurd. post; Skt. pustaka- ϐoo₎ crm /čaṙm/, čaʲm  < ŋč‘›m‘n-; Ave. čaʲəman-, Khʷāʲ. crm /carm/, č›m /čaʲmъч Khōt. tcā›m‘n-, Oss. car(m), Pasht. carman, Pers. č‘›m, Kurd. ç“›m; Ved. cá›m‘n63. (84.) meat y ta  y‚t‚₎ч y‚t₎ y‚ty / < ŋ āt‘-ka-; Khʷāʲ. ₍ātt ; cf. etymologically non-related Uygh. Uzb. ėt, Kyrg. it 64. (22.) blood ʷáʸуⁱ)n, ʷáʸ(ⁱ)m  wrn-w, wrn-y, y wn-w, w rn-h (y)xwrn-y, yxwn-y xwrn-y, ywxn-y /(yəфʸ{əwníч ʸ{əwnúч yəʸ{əníч yəʸ{ənúч yoʸníч wəxəwnáъ < ŋuáhu(r)na-; Ave. vohunД-, vohuna-, Khʷāʲ. hwny, Khōtщ h nä, Shugh. οōsh. ⁽ ⁾Дn, Bart. waxin, οāshrv. wax n, Yazgh. ⁾°‘n, Ishk. ⁽ n, Wakh. ⁽ ⁾ən, Munj. ₍Дn‘, Pasht. ⁽ nē, Pers. ⁾ n, TFalgh. xin 65. (20.) bone ʾstk-y, sṯq-y / stə₎íъ sⁱtá₎  < *asta(-)ka-; Khʷāʲ. ʾstk /əstag/, Khōtщ āst‘‘-, Oss. stæg, Ishk. ⁽ st k, Sangl. k, Wakh. (₍)‘₍č, Munj. ₍ostД₍, Yidgh. ₍‘stë, Pahl. astag, cf. Pers. ustux n 66. (85.) fat (of meat) ·182· čáʲpa  crp ъčaṙp/ < ŋč‘›p(‘)-; Khʷāʲщ crb, Oss. carv, Jass. carif, Tjk. č‘›’ (82.) oil r ⁱnчлr an, r na  rw n ro haṃ, ro aṃ / n/ < ŋ›áugn‘-(ka-); Ave. rao na-, Khʷāʲ. r n, Bactr. rwgn, Khōt. ›› ṇ‘-, Yazgh. ro (ə)n, Shugh. › an, Ishk. re (u)n, Wakh. › n, › ən, Munj. › na, Yidgh. › ən, Pers. › án, Tjk. ›‘u án, TMast. › án, TVarz. › án, TYagh. › ín, AfghP. ›‘u án, colloq. r án, Hazāʲ. ru , r , γāʲʳщ ʲou γāʲʳщ ro än, Pahl. › n (91.) milk xⁱšíʤt  ʾ ₍ t(-y) ʾ⁾ ʾ₍ t ⁾ ₍ t h a ⁽dʰ , h a wṯi / ʸšХ dуáф ~ ʸšɨ dí/ < ŋ⁾ uí”t‘-; Ave. ⁾ uu pt‘-, ⁾ uuДd-, Khʷāʲ. xwfcy /x /, Khōt. vДdä-, Yazgh. ⁾°ovd, Shugh. οōsh. ⁾ vd, πaʲХqщ ⁾ wd, Yidgh. x uvd, Pasht. ‘⁽də, Ōʲm. ДpД, Parth. ₍”t, Zâz. t, cf. Pers. ‘”tāl Д›  < Pers. Д›, Oss. æ⁾s r 67. (88.) egg táʸуⁱфm; tú⁾( )m  t m-y ṯxm-y ъtoʸmíъч t my ṯ(w)xmy < *tao⁾m‘-(ka-) < ŋtáu⁾m‘n-; Ave. taoxman-, Bactr. /tuxman/, Wakh. ta m, Ishk. tь⁾m, Pasht. t m‘, Pers. tuxm, TMast. t ⁾m, εazāʲщ tú⁾ m, Pahl. t m, Parth. tw(x)m /tōуʸфmъч OPers. t‘umā-, Kurd. tom; Ved. tókm‘nx ya tʣʳtʧϑὕʣʳ  < Pers. ⁾ā₍á, ʣʥʥуʳф Pahl. ⁾ā₍‘g, Khʷāʲ. ₍ʾk /yāʥъч Ir. ŋāu ‘-ka-, Ave. ‘ēm, Ide. *h u om, OCS. ‘jьc“, Rus. , Cze. vejce; Lat. vum, Gre. ᾠ ч δótщ ada, OEng. ǣᷧ, OScand. egg, Ger. Ei 68. (92.) horn šōʸ  ʾ h ъšāʸъ < ŋ ā⁾‘-; Wakh. ⁾, Pers. ā⁾, Parth. ʾ⁾  krnʾ /kaṙ < ŋká›nā-kā-; Ave. karəna- ʣaʲ ; Ved. ś ṅga-, Lat. cornu, Goth. h‘ú›n 69. (93.) tail d ᵎm, dʉm’á  wnp- / ūmϐ-/; ⁽npʾk / umϐēъ ψʦavʧnʥ a] taʧὕ < -; Ave. -, Khʷāʲ. wm / m/, Khōtщ dumaa-, Oss. d mæg ‖ dumæg, Yazgh. om, Shugh. οōsh. um, Ishk. d m, Munj. lum, Pasht. ləm, Pers. d m, Tjk. d m(’á), TMast. d m’(á), Kurd. duw, dunk, Baὕōch. dummag 70. feather pan(n) (arch.); par ʤʣatʦʣʲч ʷʧnʥ  prn /paṙn/ < Ir. *parna-; Ave. parəna-, Khʷāʲ. pn, Shugh. p n, οōsh. p n, Bart. p nt, πaʲХqщ pun, Yazgh. p n, Wakh. pā›, Munj. p (ǵ), Yidgh. p ṇ‘, Pasht. bəṇa, Perc. par cf. Tjk. p‘››á 71. (3.) hair ·183· daʲáu ‖ dⁱʲáu  w-y йžə⁽íй < ŋd›áu‘- ʦaʧʲ ; Khōtщ drau-, dro, Oss. æ›du ‖ æ›do, Shugh. cД⁽, οōsh. c ⁽, Yazgh. c ; Ōʲm. d›Д; Ved. drav-, Khowār. dro, Ide. ŋd›“u72. (2.) head sar  sr-y, sʾ› /ʳaʲíъ < *ʦᛑ-; Bactr. /sar/, Oss. sæ›, Ishk. sar, Pers. sar, Kurd. s“›î, Hind. sar, sir, Eynu. sä› k‘llá, s ›kállá  < Pers. k‘llá, Tjk. (s ›)k‘llá BukhAr. kalla, Uzb. kȧllȧ, Karakaplak. gelle, Turkm. kelle (4.) face ryt(h) ry(y)t ryṯ ъʲХtъ ʲХt  r  < Pers. › , TMast. › , εazāʲщ ruy, Kurd. ›û, Ave. rao a-; Goth. ludja lunǰ  < Tjk. lunǰ č“ʰ›á  < Pers. č h›(á); Ir. ŋč ϑra- ʳʧʥn ; Khōt. tcira- ʧmaʥʣ ; Ave. č ϑra- pʧϑtuʲʣ ; Pasht. cē› aὕʧ₎ʣ ; Alan. ч tomϐʳtonʣ , Oss. c ›t ‖ cirt; Ved. citra- vʧʳʧϐὕʣ ; Tatar. çı›‘₍ ʤaϑʣ 73. (6.) ear ūš  ⁽ / ōšъ < ŋg‘u ‘-; Ave. g‘o ‘-, Khʷāʲ. wx / ōʸъч Khōt. gguv ‘-, gg , Oss. qus ‖ os, Scyth. ⃝ , Wakh. , Ishk. ḷ , Shugh. , οōsh. ⁽, πaʲХq. awl, Yazgh. əvon, Munj. ₍, , Yidgh. , Pasht. waǵ, Ōʲm. g Д, g ₍, ἵaʲāch. g , Pers. g , Pahl. Parth g , OPers. g‘u ‘-, Baὕóč g , Kurd. goh; Ved. gʰo a- nʣϑ₎ 74. (5.) eye úʲda  < ŋg›d‘-ka-; cf. Ave. gərə a- ʦoὕʣч pʧt čá (ⁱ)m  c( )m-y cm-y(y), c m-y c(y)m-y, c m-y ъčɨуšфmíъ < ŋčá m‘n-; Ave. č‘ m‘n-, Khʷāʲщ cm-, cm- /camma/, Khōtщ tc“ m‘n-, Oss. cæst ʣyʣ ч casm ‖ cans ʷʧndoʷ-opʣnʧnʥ ч Ishk. com, Sangl. cāṃ, ḳēϐщ c m, Munj. č m, Yidgh. č‘m, Shugh. Baj. cēm, οōsh. Khūʤщ cām, Bart. cēm, οāshrv. cДm, πaʲХqщ cem, Yazgh. , Wakh. čə( )m, Ōʲm. c mД, čДm, c m, Pers. č‘ m, TMast. č m, γāʲʳщ č“ m, εazāʲщ čí ⁱm, Kurd. ç‘v 75. (7.) nose n s / naʧʳ  nns /naṁs/, ns /nąʳъ; nyc /nēčъ < ŋnāsn( ‘)-, -kā-; Ave. n h‘n-, Khʷāʲ. nʾc /nā a/, Yazgh. n“ǰ, Shugh. nāʒ, οōsh. Khūʤ. nēʒ, Bart. οāshrv. n ʒ, πaʲХq. noz, Ishk. nic, ἵaʲāch. nē t; Ved. n s kā76. (8.) mouth ·184· rax  r ʾk /rə 77. (9.) teeth díndak  nt(ʾ)k ntʾkh dnṯʾ ъ ɨṁ aṁ < ŋdāntu(-)ka-; Khʷāʲ. nck / an ig/, Khōtщ dandaa-, Oss. dænd‘g, Shugh. nd n, Khūʤщ οōsh. Bart. οāshrv. nd n, πaʲХqщ andan, andun, Yazgh. and, ān, Wakh. dənd k, dendik, Ishk. dond, Sangl. dānd, Munj. lod, Pers. dand n, TMast. dand n, Pahl. d‘ndān, Kurd. didan, Baὕōch. dāntān; Lat. dans, Gre. ὀ ώ ч δótщ tunþus, Ger. Zahn, OEng. t ð, Lit. d‘ntís, OIrl. ę , Irl. dé‘d, Welsh Bret. dant, Ide. *h dónt78. (10.) tongue zⁱv k  (ʾ)₎ ʾ(ʾ)k z ʾk ₎’ʾ‹ / z ā₎ ~ ž āk / < *hiʣu -k -; Ave. h ₎uuā-, h ₎ -, Khʷāʲщ z ʾk, ʾz ʾk ъzu g, əz /, Bactr. /əz āʥъ, Khōtщ ’ śā / ʧźāъч Oss. æv₎‘g, Munj. zəv zəv g, Yidgh. zᵊvД , zɪ’ē , Shugh.οōsh. ziv, Yazg. z(ə)veg, Wakh. z k, Ishk. ₎(ь)v k, Sangl. zəv k, Pasht. žəba, τazХʲХ žəbba, τazХʲХ žəbba, Wa . z(i)bə, zə’ ; Pers. , TMast. , zu , zə , TBukh. zavon, εazāʲщ ₎ ’ , Pahl. ʾ⁽₎⁽ʾn ₎⁽ʾn ъuzʷānч ʧzʷānъч Parth. ₎’ʾn ъʧzϐānъч OPers. h ₎ānam (acc. sg.), h ₎ -, Med. ŋh ₎’ān-, ḳâzâщ ₎ım‘n, ₎ı⁽‘n, zun, Kurd. ziman, κāzand. , ziwan, Baὕōch. ₎u’ān, ₎uvān, ₎‘vān, Talysh. zəvon, Khō‚ХnХ zuan, TātХ zuhun; Ōʲm. ₎o’ n; Urd. ₎‘’ān; IIr. ʰuā-; Ved. , juh -, πʧndʦХ jibʰa; Ide. ŋdngʰ -, ŋdngʰuā-, OCS. ję₎₍k ; Lat. lingua, OIrl. , , Goth. tu g , Armen. lezu, Tokh. käntu käntvo 79. (17.) fingernail ən/ náʸna  nʾ (ʾ)n < -na-; Khōtщ nāh‘n“, Yidgh. anaxno, Pers. nā⁾ún, Ved. n‘kʰá80. foot p da  pʾ ʾk pʾ (y), pʾ h, pʾ (ʾ)k pʾ (y) pʾd(₍) < -(ka-), Ave. pā a-, Khʷāʲ. pʾ , Khōt. pā‘-, Oss. fad, Wakh. p , Shugh. p , Yazgh. pe , Ishk. pud, Munj. pāl‘; Pers. pā , Pahl. pāД, OPers. pād‘-, Kurd. pê; Gre. ύ ч cf. Pasht. c‘l ›’ l‘ ʤouʲ-ὕʣʥʥʣd 81. (18.) leg link  < ρüщ ??, Kurd. ling p (₍)  pʾ ʾk pʾ (y), pʾ h, pʾ (ʾ)k pʾ (y) pʾd(₍) < -(ka-), Ave. pā a-, Khʷāʲщ pʾ , Khōtщ pā‘-, Oss. fad, Wakh. p , Shugh. p , Yazgh. pe , Ishk. pud, Munj. pāl‘; Pers. pā , Pahl. pāД, OPers. pād‘-, Kurd. pê; Gre. ύ ч cf. Pasht. c‘l ›’ l‘ ʤouʲ-ὕʣʥʥʣd 82. knee z nk  ₎ʾnʾ⁽k, ₎nʾ⁽kʾ, znwq < - < *ʣ -kā-, Ave. žnu-, Khʷāʲщ ₎ʾn⁽k, Khōtщ ₍sān -, ₍sānu‘-, Oss. ₎on g, ·185· Ishk. zong, Wa . ₎u g, ἵaʲāch. zanuk, Pers. , TMast. ₎ ní, Pahl. ₎ʾn⁽k /zānūʥъч Parth. ₎ʾn⁽g, Baὕōch. ₎‘n k; Ved. j nu-, Gre. ч Lat. genu, OEng. cnēo(w) 83. armband st-y dsṯ-y / aʳt-íъ dast  < *dásta- (disslimilation or contamination of past part. of verb - to givʣ -< -ta-) < *ʣásta-; Ave. zasta-, Khʷāʲ. st / aʳt-/, Bactr. /list/, Khōt. dasta-, Shugh. ust, Khūʤ. st οōsh. ost, πaʲХq st, Wakh. ast, dast, Yazgh. st, dast, Munj. l st, Yidgh. last, Pasht. lās, ἵaʲāch. dȫst, Pers. dast, Pahl. dst /dast/, OPers. dasta-, Parth. ч Hitt. k“ ‘›, Tokh. tsar dst, Kurd. des; Ved. hást‘-, Ide. ŋgʰ“s-to-; cf. Gre. ar y zna  < pres. part. of the verb ₍ ₎- to ʳtʲʣtϑʦ ч Pers. ₍ā₎Дđán : ₍ā₎(14.) elbow ōʲínǰ, ōʲúnǰ  ʾʾ›ʾ₍nc ʾʾ›ʾnj ʾrync ɨṁǰ/ < ϑni-ka-; Ave. arəϑna-, Khōtщ ‘› ñ“ ϐʣὕonʥʧnʥ to ʣὕϐoʷ ; Oss. [ælm-йæ›m-]æ› n ‖ [cæng-]æ› næ, Shugh. ā›“nǰ, πaʲХqщ yorn; Sangl. , Wakh. ›Дnǰ; Munj. ən, ›á₎“n, Yidgh. ›‘₎ín, Pers. ›‘n(ǰ), Tjk. › nǰ, Northern dial. olínǰ, olǘnǰ, AfghP. ›ónǰ, γāʲʳщ ränǰ; Ved. ‘›‘tní-, Gre. ὠ ч OEng. eln (15.) palm (of hand) p‘nǰá(ra)  < Tjk. p‘nǰ‘›á, p‘nǰá < ŋp‘nč‘- fivʣ kaf  < Pers. kaf nʧš₎  nn k₍ cf. Khōtщ nänā››‘- < Khōtщ nän‘- / nina- ʷʧtʦʧn + ā››‘- < *arma- aʲm páʸуʸфa  cf. Yagh. pax finʥʣʲ (16.) finger un₎úšt, ‘ngú t  ʾnk⁽ t ʾng⁽ t /aṁʥ{əštъ < ŋángu t‘-; Ave. ‘ngu t‘-, Oss. æng lʒ, Khōt. ā ṭia-, Pers. ‘ngú t, Pahl. ‘ngu t, Kurd. “ngu t; Ved. aṅgʰú ṭʰapax  cf. Yagh. p‘⁾(⁾)á paὕm 84. wing qan t, ‹‘nát  < Uzb. qan t, Uygh. Kyrg. qanat < ρüщ ŋ‹āń‘t; Tjk. , ‹‘nát bal  < Pers. ’āl ·186· par  prn /paṙn/ < Ir. *parna-; Ave. parəna-, Khʷāʲщ pn, Shugh. p n, οōsh. p n, Bart. p nt, πaʲХqщ pun, Yazgh. p n, Wakh. pā›, Munj. p (ǵ), Yidgh. p ṇ‘, Pasht. bəṇa, Pers. par  ⁽ʾ₎ ъʷāzъ cf. Sogd. √⁽₎- ъ√ʷəz-/ to fly ; Pers. v‘₎Дđán : vaz-, Ved. vah- to ϐὕoʷ (1.) body tnpʾ›, t‘n’ʾ›, t‘m’ʾ›, t‘mpʾ› ṯ‘n’ʾ›, ṯ‘npʾ›, ṯ‘m(’)ʾ›, ṯm”ʾ› ъtáṁϐāʲъ tan  < Ir. ŋt‘n -(pā›‘-); Ave. tanu-, Khʷāʲ. tn /tan/, Bactr. /tan/, Pers. tan, Parth. tn’ʾ›, Pahl. tn ʾr; Uzb. tȧn, Uygh. tän, Kyrg. ten, BukhAr. tan ’‘dán  Ar. BDN, Pers. ’‘dán ǰ‘sád  Ar. ηSD ǰasad, Pers. ǰ‘sád 85. (12.) belly šⁱ₎ámpaч ʧš₎ampá  < ŋ k‘m’‘-ka-; Pers. kám, Tjk. k‘m’á, k‘m’á, TVarz. kám, kám dáʲa ϐʣὕὕyч ʥutʳ  k ʾ›(ʾ₍), k ʾ›ʾk qϑʾry, kϑʾry /k k < ŋudᛑ-(ka-); Khʷāʲщ ʾw yr /u íʲъч Ishk. dē›, Wakh. d ›; Tjk. d‘›á ʳtomaϑʦ of a domestic anʧmaὕ ч Ved. udᛑ- ʳtomaϑʦ 86. guts k ϐʣὕὕy dáʲa ϐʣὕὕyч ʥutʳ  k ʾ›(ʾ₍), k ʾ›ʾk qϑʾry, kϑʾry /k < ŋudᛑ-(ka-); Khʷāʲщ ʾw yr /u íʲъч Ishk. dē›, Wakh. d ›; Tjk. d‘›á ʳtomaϑʦ of a domestic anʧmaὕ ч Ved. udᛑ- ʳtomaϑʦ bándil ʦʣaʲtч ʥutʳ  < Tjk. colloq. ’‘ndíl < ’ánd-i dil bundle of ʦʣaʲt ǰⁱgá› ὕiver, ʥutʳ  < Pers. ǰ gá›, Pahl. ǰakar, yakar, Ave. ₍āk‘›-, Khōtщ gyagarra-, j‘tä››‘-, Oss. gæ›, Yidgh. ₍ē ən, Pasht. (₍)Дná, Ōʲm. ʒ ṛ; Ved. ₍ák›t-, Ide. *ʜ k ›t-, Gre. ἧ r ta  < Tjk. r dá, οōsh. › d, Yazgh. rəd, Ishk. r č k, Munj. › ₍əy, › ₍Д 87. neck alk  < Ar. HLQ; Tjk. halq, Shugh. alq kámуáфч k m  kʾkh kʾ ʾʾkh, kʾ k ‹ʾ⁾ , οōsh. , Munj. kā əko, Pasht. < ŋkāh-man-, ŋkāh‘-ka-; Oss. kom, gom, Yazgh. kumai; ἵaʲāch. kām‘; Tjk. k m, Pers. kāk 88. back ᛋá  Uzb. ›‹‘, ρüщ ŋ‘›‹ā; Yazgh. Shugh. οōsh. Ishk. Wakh. ‘›‹á, πaʲХq. ‘›‹ó, Yidgh. hark ·187· sⁱtámч ʳatám  pu t  prch /paṙčъ < ŋpá› t‘-(ka-); Ave. p‘› t -, Pasht. pu⁾t, Pers. pu t, Kurd. p t; Ved. p› ṭí89. (11.) breast čʧč, ǰ ǰí  cf. Tjk. čuč, čoč, ǰ ǰ, Oss. ʒiʒi ‖ ʒeʒe, Khōt. tcДjs‘, Ishk. č č , Sangl. č čД, Shugh. ǰ ǰ, Armen. cic, Ger. Zitze, Cze. cecek, Ital. zizza, Gre. ч Georg. ʒuʒu v na  s na  < Pers. sДná, Shugh. sДn 90. (21.) heart rjy(y) / əwžē/ dil  rzy < *ʣ›d‘ ‘-; Ave. zərə aiia-, Khōtщ ₍s䛑-, Pers. dil, TMast. dil, Ishk. dьl, Ved. h d‘₍‘-, Lat. cor, Gre. ᾱ, ῆ ч St.Sl. sьrdьce; Ide. ŋk›d’ánd l ʦʣaʲtч ʥutʳ  < Tjk. colloq. ’‘ndíl < ’ánd-i dil bundle of ʦʣaʲt 91. liver ǰⁱgá› ὕiver, ʥutʳ  < Pers. ǰ gá›, Pahl. ǰakar, yakar, Ave. ₍āk‘›-, Khōtщ gyagarra-, j‘tä››‘-, Oss. gæ›, Yidgh. ₍ē ən, Pasht. (₍)Дná, Ōʲm. ʒ ṛ; Ved. ₍ák›t-, Ide. *ʜ ›t-, Gre. ἧ (23.) urine g‘₎(₎)ák, gⁱ₎(₎)ák  < Tjk.? (24.) feces w / ū / ʳ / ᵎt  < ŋg ϑ -, ŋg ϑ( -; Ave. ʥū a-, Khʷāʲ. wϑ / ū /, Yazgh. °oϑ, Shugh. οōsh. aϑ, Wakh. gi, Munj. ⁽, Pasht. (w)ul, Pers. guh, Tjk. g h xēʲdá₎  Wakh. ⁾ ›dəx; cf. Yagh. verb ⁾ē›d- to ʳʦʧt : Khʷāʲщ -xr -, Shugh. ‘› - : u⁾t-, οōsh. Bart. ‘› -, › - : u⁾t-, πaʲХqщ ‘› -, Yazgh. xaw - : ⁾‘⁾t-, Yidgh. ⁾‘⁽d-, Pasht. xaṛəl III.8. Verbs 92. (185.) to drink žau- уažáu : žáuta : žáuna : žáʷa₎ф  < ŋž ‘u-; Pasht. ž ⁽əl, Baὕōch. ǰā₍‘g, Pers. ǰāvДđán : ǰāv √ʾʾ ʾm Ave. ‘m93. (182.) to eat ·188· ʸar- уaʸáʲ : ʸ rta : ʸárna : ʸárak)  √ wr- hor-, hur- : √ wrt ъ√ʸ°ər- : √ʸ°aṙt/ < *x ar-; Ave x ar-, Khʷāʲ. x(w)r-, Bactr. -: : /ʸ{aʲ- : ʸ{aʲdъч Pers. x ‘›dán : x ar-, Tjk. x ›dán : x ›-, γāʲʳщ x ordän : x or-, AfghP. ⁾ o›dán : ⁾ or-; Eynu. xorla94. (183.) to bite xⁱ - (axⁱ : xⁱ : xⁱ : xⁱ  √ʾ ₍ʾk (inf.) / < ŋ⁾ ‘u-; Ishk. ā⁽- : ā⁽ d, Wakh. ⁽- : ⁽d, Yazgh. ‘⁽-, Munj. ‘⁾ ⁽- : ‘⁾ ēvdživ- уažív : žívta : žívna : žíva₎ф to sew, to ʳtʧtϑʦ  √ y - √₎₍ -, √ y - √j - ъ√žɨ -/ < ŋží’‘95. to suck zamák-члzamáq- уazamák : zamákta : zamágna : zamákak) ‖ zⁱmá₎- (azⁱmá₎ : zⁱmá₎ta : zⁱmágna : zⁱmá₎ak)  < *uʣ-mak-, cf. Pers. m‘kДđán : makdХy- (ad y : d yta : d yna : d yak)  < ŋdā -; Oss. dæ₍ n ‖ dæ₍un : dad; Ved. dʰā₍-, Gre. ч OCS. dojiti, Goth. daddjan 96. to spit : : : to ϑouʥʦ  √ ⁽ʾʾ ъ√ʸ°āʤъ ʸūʤOss. ʸ{ ” n ‖ xufun to ϑouʥʦ ч Yidgh. xof- : ⁾o”ā -, ἵaʲāch. kʰ ”-, Parth. wf97. to vomit qay kun- (‹á₍i ₎áʲak)  < Pers. ‹‘ k‘›dán rt kun- ( rti ₎áʲak)  cf. TMast. ›t k‘ dán; cf. TMast. ᴥ ›(›) dán to ʳʦout k u- (ak u : k uta : k una : ak) to search; to vomit; to touch; to dʧʥ  < Pers. kā”tán : kāv- / kā’-, TVarz. ko”tán : kou wrtsnty /ʸ°ər-ʳáṁ vomʧtʧnʥ  ‹ʾ⁾⁽ ʾṯy vomʧtʧnʥ 98. to blow dam ʷХd- (dámi w dak)  < Tjk. dam ϐʲʣatʦ + Yagh. ⁽Дd- to pouʲ ; cf. Oss. d m n ‖ dumun 99. to breathe dam I ʸaš- (dámi ʸášak)  < Tjk. dam ϐʲʣatʦ + Yagh. ⁾‘ - to puὕὕ 100. to laugh xant- уaʸánt : ʸántta : ʸántna : ʸántak)  √ nt /√ʸaṁd-/ < *xand-; Yazgh. xənd- : xant-, Shugh. οōsh. ānd- : sДnt-, Ishk. xond-, Wakh. kānd-, Munj. ⁾ d-, Pers. ⁾‘ndДđán : xand101. (201.) to look / to see wēn- (aw n : w ta : w nna : w nak)  √⁽₍n : √⁽₍t √⁽₍ṯ /√ʷēn : √ʷēt/ /ʷēn- : < ŋu‘ n‘-; Ave. v‘ēn‘-, Khʷāʲ. wyn- : wyn d, Bactr. -, у ф - wyn- : * ·189· l d/, Shugh. Bart. οāshrv. ⁽Дn- : ⁽Дnt, Khūʤ. win- : ⁽Дnt, οōsh. wun- : wunt, πaʲХq. w yn- : wand, Yazgh. Wakh. Sangl. ⁽Дn- : ⁽Дnd, Pers. dДđán : ’Дn-, Pahl. ⁽ēn- : dДdōʲ- уa r : rta : rna : rak)  √ ʾʾr /√ āʲъ < ŋgā›‘-; Khʷāʲ. ʾr-; cf. Oss. [æn]‹æl n : [æn]‹æld ‖ [æn] ælun : [æn] ald to ʦopʣ 102. (200.) to hear / to listen d š- (ad š : d šta : d šna : d šak) to ʦʣaʲ  √pt ⁽ √pt (ʾ)⁽ √pṯ ⁽ to ʦʣaʲ ъ√p < *pati-gáu aūš dōʲ- / kun- у šʧ d rak / ₎áʲak) to ὕʧʳtʣn  < Pers. g dā tán, TYagh. dó t‘n, Oss. qus d‘› n 103. to know b z n- (ab z n : b z nta : b z nna : b z nak) ‖ bⁱz n- (abⁱz n : bⁱz nta : bⁱz nna : bⁱz nak)  √(pt)₎ʾn √pt₎ʾn √pṯ₎ʾn ъ√(p t) < *apa-ʣān-, *(pati-)ʣān-; Ave. pāt -₎ān‘-, Khōtщ p‘₍sān-, Oss. (’‘)₎on n : (’‘)₎ nd, ‖ zonun : zund, Yazgh. vəzan- : vəzant-, Shugh. ⁽ ₎ n- : ⁽ ₎ nt; οōsh. Khūʤ. ⁽ ₎ n- : ⁽ ₎ēnt, πaʲХqщ wazon- : wazont, Ishk. pь₎ n- : pь₎ nt-, Wakh. pazdan-, Munj. v₎ n- : v₎ d-, Pasht. pēž n-, Pers. dān stán : dān-; cf. Bactr. ?? ⁱʲíʤ- уa ⁱʲíʤ : ⁱʲíʤta : ⁱʲíʤna : ⁱʲíʤak) ‖ ⁱʲív- уa ⁱʲív : ⁱʲíʤta : ⁱʲívna : ⁱʲívak)  √ r - √ rb- ъ√ ʧw -/ < ŋg›’ ‘- to grab, to ta₎ʣ ; Ave. gərəbiia-, Khʷāʲщ i ya-; Khōtщ grauna-, Oss. æ› æv n : æ› ævd ‖ æ› uvun : æ› uvd, Ishk. urv- : urd, Munj. ərv- : ərivd, OPers. g›’ā₍‘-, Pers. g › ”tán : gД›-, Pahl. graftan, Kurd. girtin, Baὕōch. girag : gipt; Ved. g›‘’ʰ- : g›’ʰṇāt , OCS. g›“’ǫ : grabiti 104. to think fikr kun- (”íkri ₎áʲak)  Pers. fikr ₎aʲdán < Ar. fikr mʧndч opʧnʧon kun- ( ʧ ₎áʲa₎ф  < Pers. ‘ndē Дđán : ‘ndē - to tʦʧn₎ √m₍n ъ√mēnъ  Ave. m‘ⁱn ‘-, Pahl. m“nДd‘n 105. to smell vūd ʸašʸáša₎ф  w h w ъ ō ъ ʳϑʣnt < Ir. ŋ’áud -, calque of Tjk. ’ k‘ Дdán < Yagh. v d, Sogd. Ave. baodi-, Khʷāʲщ w / ō ъ, Oss. bud ‖ ’odæ, Wakh. v l, ἵaʲāch. ’ʰām, Pers. ’ , Pahl. ’ ₍; Hung. ’ű₎ + Yagh. ⁾‘ -, Pers. k‘ Дđán to puὕὕ  √pc ⁽ , √pt₎ ⁽ √pc’⁽ ъ√p 106. to fear ·190· čⁱk r- уačⁱk r : čⁱk rta : čⁱk rna : čⁱk rak) ‖ č ₎áʧʲ- уač ₎áʧʲ : č ₎áʧʲta : č ₎áʧʲna : č ₎áʧʲak)  √pck⁽₍› ъ√p č₎ü < ŋp‘t -k u›‘ ‘- (?) tūʲ: : :  cf. Southern Tjk. t › dán : t › √t›s- ъ√təws-чл√tʲəs-/ < ŋt›ʦ-; Ave. tras-, Pers. t‘›sДđán : tars107. (187.) to sleep fs- уa{ūfs : ūfta : ūfsna : ūfsak)  √ʾ⁽ s(-) √ʾ⁽”s(-) : √ʾ⁽ t(-) √ʾ⁽’ṯ(-), √ʾ⁽”ṯ(-) /√ōʤʳ, √uʤʳ- : √ō d, √u d-/ < ŋ(‘u‘-)hufʦa-; Khōt. h s-, Oss. ʸ{ ss n ‖ xussun, οōsh. ⁾o”s- : ⁾ vd-, Yazgh. pəxas- : pəxovd, Pers. ⁾u”tán : x āƀ-, Baὕōch. vafsag-; cf. Pasht. ud ʳὕʣʣpʧnʥ 108. to live žū- уaž : ž ta : ž na : ž ak, ž {ak)  √ʾ₎⁽(-) √(ʾ)₎⁽(-) √j⁽(-) √ž⁽(-) /√ žūч žau-/ < ŋǰ(‘)u‘-; Ave. ǰ(a)uua-, Khʷāʲ. zyw-, Khōt. j zindag kun- (zindag ʧ ₎áʲak)  < Pers. zindag k‘›dán; ὕʧʤʣ ч past part. of verb ₎Дstán : - to livʣ ч OPers. ǰДv-, Pahl. ₎Д⁽‘st‘n; Pasht. ὕʧʤʣ 109. (192.) to die √m₍›- : √mwrt- √mwrṯ- /√mɨr- : √muwt-/ mir- уamíʲ : m rta : míʲna : míʲak)  < ŋm ‘- : m t‘-; Khʷāʲ. (ʾ)m₍- : ʾmȳd, Bactr. -: уmХʲ- : murd), Khōt. mä›- : muḍa-, Pers. mu›dán : mД›-, Pahl. mД›- : murd110. (193.) to kill √pt ⁽(ʾ)₍ √pt⁾⁽ʾ₍ √pṯ⁾⁽ʾ₍, √pṯ⁽⁾ʾ₍ t ʸ y- (at ʸ y : t ʸáʳta : t ʸ yna : t ʸ yak)  ъ√p tʸ° < *pati-x h‘ ‘pakk- уapá₎₎ : pá₎₎ta : pá₎₎na : pá₎₎ak) to cut, to ₎ʧὕὕ  111. to fight ǰang- (aǰáng : ǰángta : ǰángna : ǰángak)  < Pers. ǰ‘ngДđán : ǰang- < Pers. ǰang ʷaʲ ǰang kun- / nōʳ- (ǰángi ₎áʲak / ) < Pers. ǰang k‘›dán, Pers. ǰang ʷaʲ + k‘›dán to do ; cf. Uzb. ǰ‘ ‹ïłm ‹ /g nōʳ- ( t ni / g i n sak)  Tjk. g t n g › ”tán < Tjk. g t , γārs. ko fight, ʷʲʣʳtὕʧnʥ ʤūʲúš- уaʤūʲúš : ʤūʲúšta : ʤūʲúšna : ʤūʲúša₎ф  ϐⁱd n nōʳ- уϐⁱd ni  Yagh. bid n ʷaʧʳt + n s- to ta₎ʣ  √›n -, √›np- ъ√ʲənb-/ 112. to hunt ·191· ⁱk r kun- ( ⁱk ri ₎áʲak)  √ʾ kʾ›, √(ʾ) kʾ› /√ᶤ Bactr. ϸу ф - ч ϸ - : ϸ /əšуəф₎āʲ- : əšуəф₎āʲd-/; Pers. ₎aʲdán (Pers. , BukhAr. k › ʦunt ф nū₎ kun- (n ki ₎áʲak)  113. to hit deh-, dih- (adíh : déʦta : déʦna : díhak)  < ŋdā(h)- to give, to ʦʧt ; Ave. dā-, Khʷāʲ. dah-, dih-, Khōtщ d -, Yazgh. day- : ed-, Shugh. di(y)- : d, d t, Khūʤ. di(y)- : dēt, οōsh. Bart. dē(₍)- : dēt, πaʲХq. de- : det, Ishk. de- : ded-, Sangl. deh- : dē -, Wakh. dē-, di- : dəyt, dē⁾t, Munj. d -, də-, Yidgh. dah-, ἵaʲāch. dah-, deh-, Pers. dādán : dah-, Kābul. dē-, TVanj. deh k‘›dán; Khowār. dik 114. to cut p ʸ y- (ap ʸ y : p ʸ yta : p ʸ yna : p ʸ yak)  √p ⁽ʾ₍ √(ʾ)p ⁽ʾ₍ √p⁾⁽(w)ʾy √p⁾⁽ʾ₍ ъ√p ʸ° < *apa-/upa-x h‘ ‘pakk- уapá₎₎ : pá₎₎ta : pá₎₎na : pá₎₎ak); pá₎₎a kun- уpá₎₎aʧ ₎áʲak)  burr- (a’ú›› : ’ú››ta : ’ú››na : ’ú››ak)  < Pers. ’u››Дđán : burr115. to split / kun- (ǰ ʧ / ʧ ₎áʲa₎ф  ǰ < Pers. ǰ k‘›dán 116. to stab čumʤ- уačúmʤ : čúmʤta : čúmʤna : čúmʤak)  st wnp fsṯxwmp /f stxúṁϐ/ cf. Khʷāʲщ xwmb117. to scratch  kД›(›)- ( : : : ) 118. to dig √kn- : √knt √‹n- ъ√₎ən- : √₎aṁd-/ kan- уa₎án : ₎ánta : ₎ánna : ₎ának)  < *kan-; Pers. k‘ndán : kanak) to search; to vomit; to touch; to dʧʥ  k u- (ak u : k uta : k una : < Pers. kā”tán : kāv- / kā’-, TVarz. ko”tán : kou119. to swim kun- ( ʧ ₎áʲa₎фл < Pers. (< āƀ ʷatʣʲ лцл ʥamʣ флk‘›dán toлʳʷʧm  √”snʾ₍- ъ√ʤ ʳn ʧ/ < *fra-snā ‘-; Khōtщлh‘₍snāt‘120. (194.) to fly ·192· fur(r)- (‘”ú›(r) : ”ú›(r)ta : ”ú›(r)na : ”ú›(r)ak); par- (apá› : pá›ta : pá›na : pá›ak)  √p›nʾ₍ ъ√páṙnāyч √ʤʲə cf. Pers. p‘››Дđán : parrpayw z kun- (payw zi ₎áʲak)  √ rwz √”›⁽(ʾ)₎ √”›⁽₎ ъ√ʤʲəʷāz, pəʲʷāz/ Tjk. parv z k‘›dán; cf. Sogd. p›⁽ʾ₎ ʷʧnʥʣd  √⁽₎- ъ√ʷəz-/ 121. (195.) to walk šau- уašáu : ta, šáuta, š ta : šáuna : šáwak)  √ ⁽- ъ√šəw-/ < ŋč ‘u-; Ave. ṧ(ii)auu-, TYaghn. ‘⁽-, Khʷāʲ. ciyy-, Khōt. tsu-, Tumsh. ccʰ‘m ζ ʥo ч Bactr. ϸ у ф-, ϸ у ф- : ϸ /šaʷ- : šudъч Oss. cæ⁽ n : c d, ‖ cæ⁽un : cud, OPers. šyav-; Pers. uđán : ‘u- / ‘v- to walk > from the 11th-12th century to ϐʣϑomʣ (Tjk. udán : ‘u/ ‘v-, AfghP. odán : ‘u- / ‘⁽-, γāʲʳщ odän : ou- / äv- to ϐʣϑomʣ ), OPers. šʧyav-; Skt. cyavati 122. (198.) to come √ ʾw √’ʾ⁽ ъ√ āʷъ vōu- (av u : v uta : v una : v wak)  (196.) to run d‘u- (adáu : dáuta : dáuna : dá⁽ak)  Pers. d‘vДđán : d‘u- / dav-; Ir. ŋd‘u-, Ave. dauu-, Oss. d‘⁽ n ‖ dawun : dawd, Ved. dʰ v- : dʰāv‘t , Gre. (197.) to go tir- уatíʲ : t rta : tíʲna : tíʲak)  √ṯr- ъ√tir-/ < *t ‘-; Bactr. tʦʣy do not ϑomʣ ; cf. Sogd. √ʾ tyr√”t(₍)›√”ṯ(y)r- ъ√ ftir-/ to go through, to paʳʳ < *”›‘-t ‘-; Pers. gu ‘ tán : gu ar-, Tjk. gu₎‘ tán : gudar- to go through, to paʳʳ < ŋu -t ‘123. (188.) to lie (down) nap d- (anap d : nap sta : nap dna : nap dak) ‖ nⁱp d- (anⁱp d : nⁱp sta : nⁱp dna : nⁱp dak)  √npʾ₍ ъ√nəpē ъ < ŋn -pád(‘) ‘-; Ave. n p‘ⁱ iia-, Khʷāʲ. ʾnbzy-, Khōtщ nuvad-, Munj. nДlv- : nu⁽ st, Pahl. nibastan 124. (189.) to sit √n₍ √n₍d : √n₍st √nysṯ /√nХ : √nХʳtъ nХd- (an d : n sta : n dna : n dak)  < *ni-; Khʷāʲ. nДϑ-, Khōt. näd-, Yazgh. niϑ- : nust, Shugh. Khūʤ. niϑ- : nust, οōsh. niϑ- : n st, Bart. niϑ- : nȫst, οāshrv. niϑ- : nǖst, Ishk. nid- : nьḷ st, Munj. n ⁾- : nⁱ₍ st-, Yidgh. n ⁾- : n st 125. to stand ūšt- уa{ št : šta : šna : štak)  √ʾ⁽ t √ʾ⁽ ṯ /√ōšt/ < ŋ‘u‘-h t‘-; Khōtщ va t-, Oss. ( )st n ‖ istun, Part. ʾ⁽₍ t-, Baὕōch. t‘g, vu t‘g; cf. Pers. Дstādán : Дst, εazāʲщ st d 126. to turn ·193· : : : ‖z - (az :z :z :z  √₎⁽ʾrt √(ʾ)₎⁽ʾrt √₎⁽›ṯ /√ zwāṙt/ *uʣ-uá›t(a)-; Parth. Pahl. zwrd- : ₎⁽ t-; cf. Pers. g‘ tán : gard- < ŋu›t-; Cze. zvrtnout, v›tět √₎⁽ʾ₍›t √₎⁽₍›ṯ zⁱʷíʲt- уazⁱʷíʲt : zⁱʷíʲtta : zⁱʷíʲtna : zⁱʷíʲta₎ф  √ʾ₎⁽ʾ₍›t √(ʾ)₎⁽ʾ₍›t /√ zʷíṙt/ *uʣ-uá›t(‘)- ‘laks- (aláks : láksta : láksna : láksak)  ϑʤщлAʲщлοἶπ, Pers. raqsДdán : ›‘‹s- to danϑʣ t ’лʸaʲ- ( i ʸáʲak)  < Pers. tāƀ ⁾ ‘›dán, TVarz. tou ⁾ ›dán 127. to fall ⁱ - (adⁱ ⁱ ⁱ ⁱ  < ŋdu‘ ‘-; Ave. duuan- toлfly члPasht. l⁽ēǵ- : l⁽ēd-; Ide. ŋdʰun- oⁱʲaуʧфš- (atⁱʲá(ʧ)šл śл tⁱʲá(ʧ)štaл śл tⁱʲá(ʧ)šnaл śл tⁱʲá(ʧ)ša₎фл  √pt›₎-, √pt›ʾ₍₎ : √pt› t- ъ√p trəž-, √p trəšt-/ < *pati-raʣ ‘-, *pati- ʣ‘ ‘-?; cf. Ave. ›‘ē -, Khōt birata √ʾnpt √ʾnp(ʾ)t √ʾmpt √ʾmpṯ, √ʾmpd ъ√áṁbat/; √ʾ⁽pt √ʾ⁽p(ʾ)t √ʾ⁽pṯ < *ham-pata-, ŋ‘u‘-pata-; Khōʲщ ʾnpd128. (190.) to give taʤáʲ- уataʤáʲ : ʳaʲáʤta, taʤ rta : taʤáʲna : taʤáʲak) ‖ tⁱʤáʲ- уatⁱʤáʲ : tⁱʲáʤtaч tⁱf rta : tⁱʤáʲna : tⁱʤáʲak)  √ r- √ϑbr- ъ√ ər-/ < *f( ) á›ă- < *fra-’ᛑ-; Khʷāʲ. hi °›- : hʾ ryd, Khōt. haur-, hor- : hoḍa-, Tumshuq. ror- : rorda129. to hold dōʲ: : :  √ ʾr √dʾ› : √₎ t-, √ t- √ r t-, √₎ t-, √ tj t- ž ṯ- ъ√ āʲ : √žə d-/ < ŋdā›‘- : d›⁾t‘-; Bactr. - : ч у ф ч ч Khʷāʲщ ʾ -, Oss. d‘› n, Pers. dā tán : dā›-, Ide. ŋdʰ“›č‘k dōʲ- (č‘k ad r : č‘k d rta : č‘k d rna : čáki d rak), čágdōʲ- (ačágdōʲ : čágdōʲta : čágdōʲna : čágdōʲak)  < Tjk. č‘k d tán 130. to squeeze ʧὕ č- уa ʧὕ č : ʧὕ čta : ʧὕ čna : ʧὕ čak)  č u- (ač u : č uta : č una : č wak)  cf. TMast. čov dán, čo”tán : čov131. to rub m ll- ( : : : ) < Pers. mālДđán : māl- ?? - ·194· 132. to wash sⁱn y- (asⁱn y : sⁱn yta : sⁱn yna : sⁱn yak)  √snʾ(ʾ)₍√snʾ₍- : √snʾʾt ъ√ʳnāʧ : √ʳnātъ < ŋsnā ‘- : *snāt‘-; Ave. snaiia-, Khōt. h‘₍snā-, Khʷāʲ. snād‘k ʷaʳʦʣd ч Oss. æ⁾sn n : æ⁾sn‘d ‖ æ⁾snun; Yazgh. Yidgh. Ishk. zənay-, οōsh. ₎ n‘₍-, Munj. ⁽ ₎n-, Ved. snā133. to wipe rant- уaʲánt : ʲánta : ʲán(t)na : ʲántak)  Oss. ›ænd n : ›ænsd-, Baὕōch. randag 134. to pull ʸaš- уaʸáš : ʸášta : ʸášna : ʸášak)  √ › -, √ n - √⁾› -, √⁾n -, √⁾ - ъ√ʸ š-, √ʸąš-/ < ŋk› ‘-; Ave. k‘› -, Khʷāʲ. ⁾ -, Oss. ⁾æss n, Yazgh. xə›á⁾- : ⁾‘›á⁾t-, kə⁾án- : kə⁾ánt-, Ishk. ⁾‘ - : ⁾‘ t-, k›“ -, Wakh. ⁾ā - : ⁾ā t-, Munj. ⁾‘ - : ⁾ k-, Pasht. k⁾əl : ⁾kД-, Pers. k‘ Дđán : k‘ -; Ved. ká› ati 135. to push šⁱk l(l)- уašⁱk l(l) : šⁱk l(l)ta : šⁱk l(l)na : šⁱk l(l)lak)  čumʤ- уačúmʤ : čúmʤta : čúmʤna : čúmʤa₎ф  st wnp fsṯxwmp /f ʳtʸúṁϐ/ cf. Khʷāʲщ xwmb136. to throw  √ʾ sʾ₍p /√ ʤʳēp/ cf. Ave. ‘ēv -sipa137. to tie vant- уavánt : váʳta : vánna : vántak)  √ (y)nd : √ (y)sṯ- ъ√ ɨṁd : √ ɨst-/ < ŋ’ánd‘- : ŋ’ásta-(ka-), Khʷāʲ. ncy-, Khōtщ ban- : bast-, Oss. ’ætt n : bast, Yazgh. vand- : v st-, οōsh. vind- : vost-, Wakh. vānd- : vāst-, Yidgh. vad- : v st-, Ishk. vond- : v st-, Pers. ’‘stán : band-, Kurd. bastin, Baὕōch. bandag 138. to sew šХy- I уaš y : š ta : š yna : š yak)  √ ⁽m ъ√šūmъ Munj. žД₍-, Ved. s₍ tá- ʳʣʷn ; Lit. s út , OCS. t živ- уažív : žívta : žívna : žíva₎ф  √ y - √₎₍ -, √ y - √j - ъ√žɨ -/ to ϑʦʣʷ < ŋží’‘139. to count ⁱs b kun- ( ⁱs bi ₎áʲak)  < Pers. k‘›dán; Pers. < Ar. SB sā’, BukhAr. sā’ ϑount √pt m› √pt mʾ› √pcmr ъ√p  -; cf. Pers. umā›á numϐʣʲ < *pati140. to say / to speak √⁽ʾ √⁽ʾ’ /√ʷā / ʷō(v)- (aw (v) : w (v)ta : w (v)na : w (v)ak)  < ŋu -; Ave. uf- to ʳʧnʥ ч Pasht. wayəl : ⁽ā₍gap deh- (gápi díhak)  < Tjk. gap ₎‘dán; Shugh. gāp di(y)·195· 141. to sing žōy- уaž y : ž yta : ž yna : ž yak)  √jʾ₍ √žʾ₍ /√žāyъ < ŋǰā -; Wakh. ǰoy- : ǰoyd, Munj. ₍-, Ide. ŋgē -; cf. Ave. gāϑa- ʳonʥч δāthā ч Ved. gā₍‘t ʦʣ ʳʧnʥʳ 142. to play ’ ₎ kun- (’ ₎ ʧ ₎áʲak)  < Pers. ’ā₎ k‘›dán, cf. Pers. ’ā⁾tán : ’ā₎143. to float 144. to flow 145. to freeze šХy- II уaš y : š ta : š yna : š yak)  Oss. s ₍ n, Yazgh. ‘₍- : “d, Shugh. Baj. ⁾ c (y)- : ⁾ c d, οōsh. Khūʤ. ⁾ c‘₍- : ⁾ c d, Bart. ⁾ cД- : ⁾ c d, οāshrv. ⁾ c‘₍- : ⁾ c d, πaʲХq. ⁾ c y- : ⁾ cud, Ishk. t ⁽- : tud; cf. πaʲХq. ϑoὕd sir- (a{ sir : sirta, ʳōʲta : sirna : sirak)  < Ir. ŋās› ‘-; Ave. sarəta-, Oss. sælun : sald, Wakh. was r- : was rt, Tjk. Wanj s › dán, Pahl. ʾpsʾ›-; Parth sald; cf. Sogd. (p)s₍›ʾ mndyy, ʤʲʣʣzʧnʥ ; cf. Wakh. ʳ r ϑoὕd 146. to swell  √tϑm-, √”tm- ъ√ʤ əm-/ < *fra-dmā(184.) to be hungry daváz ‖ dⁱváz vХ- уdavázʧ ‖ dⁱvázʧ v yak)  Yagh. d‘vá₎ ‖ dⁱvá₎ ʦunʥʣʲ , Sogd. (ʾ) z-y dbz-y / əzíъч Cʦótщ d“’Дś‘, ἵaštщ lwə ‘, Parth. ʾdbz (186.) to be thirsty t‘ ná vХ- (t‘ náʧ v yak)  < Tjk. t‘ ná ’ dán III.9. Celestial objects 147. (41.) sun x ᵎr (arch. ⁾‘›)  w(y)r ⁾⁽(ʾ)› xwyr ъʸü ʸōʲч ʸ°āʲъ < ŋhuá› ‘-; Ave. h › -, Khʷāʲ. ʾx r, xr, Oss. xur ‖ xor, Yazgh. xə⁽ú›, ⁾ ›, Wakh. (₍)Д›, Shugh. ⁾Д›, οōsh. xor, Bart. ⁾ȫr, πaʲХq. xer, Ved. suvá›-, s rya-; cf. Pers. x ‘›[ đ] > x , Tjk. x u›[ d], Pahl. x ‘›[ ēt], Ave. huuare-⁾ ‘“t‘-; Scyth. ψ ϊ; Ide. ŋs(u)uél-, ŋs l-, Gre. ἥ ч Lat. s l, Lit. sáulė, OCS. sl nьc ft b , a  < Pers. ft ƀ, TMast. ‘”tó’, TVarz. ‘”tó⁽, o”tó⁽, εazāʲщ aft w, ”t w, Kurd. extaw, cf. Skt. ā’ʰā-tāp‘- ·196· 148. (42.) moon maht b,  mʾ (h), mʾ⁾ mʾ⁾ /māʸ/ < ŋmāh-; Ave. OPers. māh-, Bactr. у ф /māуʦфъч Khōtщ māstä, Oss. mæ ‖ mæ₍æ, Shugh. m st, οōsh. mēst, πaʲХqщ most, Yazgh. mast, Wakh. m y, Pasht. m₍ā t, Kurd. meh, Ved. mās-; Pers. m h[t ƀ], TVarz. mohtó’, εazāʲщ m t w, m tá⁽, BukhAr. maht b 149. (44.) star sⁱt ra  (ʾ)stʾ›ʾk (ʾ)stʾ›₍, ʾstry /ᶤ < ŋstā›‘-k -; Khʷāʲ. /(əфʳtāʲʣʥъч Khōtщ stā›‘‘-, stā›‘₍, Shugh. ⁾ t rʒ, Ba₍ūщ ⁾ t rʒ, ⁾ t ›ǰ, Khūʤщ οōsh. ⁾ tē›ʒ, ⁾ tu›ǰ, Bart. οāshrv. ⁾ t ›ǰ, πaʲХqщ ⁾ tu›ǰ, ⁾ tu›ǰ, Yazgh. ⁾(ə)tarag, Ishk. st› k, Sangl. ustᵊ , Wakh. s(ə)t ›, Munj. st ›əy, Yidgh. , Pasht. (f), Ōʲm. starrak, Pers. s tā›á, Parth. ʾstʾ›g ϐʧὕdínʥуaф  unknown origin, in Yaghn ’Д this word is known only in dialect of village Q l; cf. Wakh. pi ing (perf.) : pi ic-, pidic- to ʥὕʧttʣʲ III.10. Nature (i) 150. (46.) water p(a), u  ʾʾph ʾʾp(h) ʾʾp ʾp ā-p p / < -(ka-); Ave. āp- (nom. sg. ā” ), Khʷāʲщ ʾb /āϐъч Bactr. ( ) ₍ʾ /āϐч ā ъч Khōtщ tcā-, Oss. avg ʥὕaʳʳ ч Ishk. vek, Sangl. vē(k), Wakh. yupk, Munj. ā, Yidgh. yow o, Pasht. ’ə, ἵaʲāch. ā⁽ə, Ōʲm. ⁽ k; Pers. āƀ, Tjk. ’(á), TVarz. ow, TMast. TFalgh. ob, TYagh. ob, ow, εazāʲщ aw, Pahl. āp > ā , OPers. āp-, Kurd. aw, Baὕōch. āp; Ide. *h ep-; Ved. pa-, ap-, Hit. ḫa-pa-a, ḫa-ap-pa to the rivʣʲ ; Eynu. ab; Gre. ⃝ ⃝ (in geographical names); OIrl. aba, ʲivʣʲ , Irl. Gael. abhainn ʲivʣʲ , Welsh afon ʲivʣʲ ; Lith. ùpė ʲivʣʲ , cf. Cze. (substrate?) hydronyms Op[ava], Úp‘ ʾ h, ⁾ʾ⁾ /ʸāʸъ ʳpʲʧnʥ ʸō₎ ʳpʲʧnʥ  < -, - ʳpʲʧnʥ ; Ave. ⁾āo, Khōt. ⁾āh‘-, Yazgh. ⁾“⁾, Wanj. xik, ʷatʣʲч ʳpʲʧnʥ ч Shugh. οōsh. ⁾‘c ʷatʣʲ ; οāshrv. ⁾ā₍ ϐʲoo₎ , Wakh. k k, Munj. ⁾ g‘; Ōʲm. ⁾āko 151. (45.) rain ’ › n ⁽ʾ› ъʷāʲъ < ŋuā›‘-; Ave. vā›‘-, Khʷāʲщ ⁽ʾ› /ʷāʲъч Oss. wa› n ‖ warun, Pers. ’ā› n, TBuch. ’o›o , Shugh. ’ › n, Ishk. boron, Munj. ’ › n, Kurd. barin, ’‘›î 152. (47.) river , dar(ⁱ уʥʲʣatф river, уʳʣaф  zry /zʲēъ ʳʣa < *ʣ›‘ ‘-; Ave. zraiia- ʳʣa ч OPers. draya-, Pers. ʳʣa, (great river) ; Tjk. уʥʲʣatф river, ʳʣa ч TMast. d‘ ›ó, TVarz. d‘ ›ó, d‘›₍ó, πaʲХqщ d‘›₍ú; cf. Kyrg. d‘›ï₍‘, dayra, Kazakh. d‘›ï₍‘, Uzb. d‘›₍ , Uygh. dä›₍‘, Tatar. dä›₍‘, Eynu. dä›₍‘; cf. BukhAr. ba ar ʲivʣʲ < Ar. B R ba r ·197· naʰr  r d < Ar. NHR nahr, Pers. nahr; BukhAr. nahr уʧʲʲʧʥatʧonф ϑʦannʣὕ < Pers. › đ; cf. Yagh. › ut, ʲavʧnʣ (arch.) ч Sogd. ›ʾ⁽ʾth /ʲāʷatъч rw(w)t rwṯ ъʲōtъч TMast. ›o⁽ǘt, TYaghn. › ud, Pers. (Lu át-i Furs) ›ʾ⁽d /ʲāvуaфdъ < Ir. ŋ›āu‘t -, Ave. rauuan- vaὕὕʣy ч Khʷāʲщ ›ʾ⁽₍n ʣaʲtʦ ; Oss. ran ‖ ›æ⁽æn pὕaϑʣ ; cf. Kyrg. place-names Ravat, Raut nōu daὕʣ  nʾ⁽n n/ < nāu‘-; Shugh-οōsh. n ⁽, πaʲХqщ n w, Yazgh. n w 153. lake ‘u₎, ‘ud  < Ar. awd; Pers. ‘u₎, TVarz. ‘u₎, TMast. h‘u₎, h‘ud, Shugh. awʒ, awz, Tr. havuz, ʾʾ⁽₎ʾk, ʾʾwzyy, ʾʾ⁽ʾ₎h Rus. , , ; cf. (etymologically /un/related?)  ʾʾ⁽₎ʾk, ʾʾwzyy, ʾʾ⁽ʾ₎h k l < Uzb. k l, Kyrg. köl, Tr. göl, ρüщ ŋköl, Tjk. k l, Bulg. гь ó₎ ›‘  < Rus. , OCS. jezero, jezer , Srb-Cro. jȅ₎“›o, Lith. ẽž“›‘s, Tjk. colloq. o₎ ›á 154. sea , dar(ⁱ уʥʲʣatф river, уʳʣaф  zry /zʲēъ ʳʣa ʳʣa ; Tjk. уʥʲʣatф river, < *ʣ›‘ ‘-; Ave. zraiia- ʳʣa ч OPers. draya-, Pers. ʳʣa ч TMast. d‘ ›ó, TVarz. d‘ ›ó, d‘›₍ó, πaʲХqщ d‘›₍ú; cf. Kyrg. d‘›ï₍‘, dayra, Kazakh. d‘›ï₍‘, Uzb. d‘›₍ , Uygh. dä›₍‘, Tatar. dä›₍‘, Eynu. dä›₍‘ ba r  < Ar. B R ba r, Malt. ’‘А‘›, Pers. ba r ʳʣa ; BukhAr. ba ar ʲivʣʲ  smʾ⁽t› smwtr-y, swmtr smwṯr-y, swmdr ъʳumudʲуíфъ < Skt. samudra155. (83.) salt n‘mák  nmʾ k(h) nmʾ k /nəmā k/ < ŋn‘m‘dkā-; Ave. nəma ka-, Khʷāʲщ nmϑk /nama k/, Bactr. /namilg/, Pasht. mālg‘, Pers. n‘mák, Parth. nmydk 156. (52.) stone sank(a), sang  snk(ʾ) sng ъʳáṁ < *aʦáng‘-(ka-), Ave. asənga-, Khʷāʲщ snk /sang(a)/, Bactr. /aʳa ʥъч Ishk. ʳůnʥ; OPers. aϑanga-, Pers. sang, εazāʲщ san(g), san(k); Eynu. s‘ 157. (54.) sand ›ēg  < Pers. ›ēg, Kurd. ›îk, ›êg, Pasht. ›ēg ·198· 158. (59.) dust ⁾ k < Pers. ⁾āk b r < Ar. КBR ᷧ , Pers. ub r č‘nk, č‘ng  < Pers. č‘ng, BukhAr. č‘ng gard  < Pers. gard  wrwm wrm(h) xrwm, xwrm xwrm ъʸ{ʲúmъ < *xruma-; Ave. paxruma159. earth ⁱr k  ›ʾ₍k(ʾ) ryk < ŋg›á ‘-ka-; Khʷāʲ. ›ʾk, Khōt. grika-, g›u k₍ā-, Oss. æl g ‖ æ› æ, Munj. ərəy, Yazgh. xərik; cf. OCS. glina, Eng. clay ⁾ k < Pers. ⁾āk ʣaʲtʦч ὕand  ₎ʾ₍ ₎ʾ₍(₍) ъzāʧ/ zōy fiʣὕd ч < *ʣ ‘-; Ave. zam-, Bactr. ч , TMast. ₎o₍ák, TYagh. ₎o₍ók ϑuὕtivated ὕand Pers. , εazāʲщ ₎ mí, Wakh. zəmin, πaʲХqщ ₎‘mín; Ide. ŋdʰ“gʰ -m : ŋdʰ“gʰm-, Chet. te-e-kán (tēk‘n), Tokh. tkaṃ kaṃ, Gre. ώ ч Ved. k am-, Lat. humus, OCS. zemlja, Lit. žẽmės (58.) mud l ₍ < Pers. lā₍ ₎‘ʰ, za  < Tjk. zah  ›ʾ₍ ъ ʲХъ < ŋg› ‘-; Khōtщ g›Дh‘-, Gre. III.11. Weather 160. (48.) cloud abr  < Pers. abr, TVarz. ‘u›, εazāʲщ ‘ú›, Shugh. ʰá’› , Kurd. awr, Baὕōch. (h)‘u›, Ir. ŋ‘’›( )a-, Ave. a ra-, Khōtщ ora- ʳ₎y ч Oss. arv ʳ₎y ч æv›‘g ϑὕoud ч Pasht. ›ə, Ved. ‘’ʰ›ā my ъmē ъ < ŋmā gá-; Ave. m‘ē a-, Oss. mi ‖ me æ, Pers. Pahl. mē , Ved. m“gʰá161. fog ·199· t mán  < Tjk. 162. (43.) sky s( )m n  smʾnh (ʾ)smʾn smʾn /ᶤ < *áʦ -; Ave. asman-, Khʷāʲщл ₍ʾsmᵃ ъyā-уaфʳmãъл tʦʣл ʦʣavʣn чл Pers. āsm n, γāʲʳщ colloq. sem n, Tjk. colloq. os m n, os món, TMast. osp n, TVarz. osmón, ospón; OPers. asman-, Pahl. āsmān, Kurd. esman; Ved. áśm‘n-, Pruss. asman-, Eng. heaven; Qashq. ssïm n, sm n 163. (51.) wind w t(a)  ⁽ʾt ⁽ʾt, ⁽ʾʾ , ⁽ʾ /ʷātъч < ŋu‘ąta-(ka-); Ave. vāt‘- (trisyllabic), Bactr. /ʷādъч Oss. wad, Pers. ’āđ, Kurd. ba; IIr. *ʜu‘ʜąta- < Ide. *h u“h nto-, Lat. ventus ‘m l  < Ar. Šκἱ ‘māl noʲtʦʣʲn ʷʧnd , Pers. ‘m l, γāʲʳщ “m l, dial. of Khoʲāʳān umol ʷʧnd ; Tjk. ‘m l, ʷʧnd ; Uzb. ‘m ł ʷʧnd , Kyrg. ‘m‘ł ʷʧnd , Kazakh. s‘m‘ł ʷʧnd , Turkm. ämāł ʷʧnd 164. snow ʷáʤуⁱ)r, warf  w r-y wfr-y /wəʤʲíъ < ŋuá”›‘-; Ave. vafra-, Khʷāʲ. wf rk, Khōt. ’o›ā-, Sangl. varf, Munj. vá”›ă, Pasht. ⁽ā⁽›‘, Pers. barf, Kurd. vafr, befir, bafer, berf 165. ice ēx, ıʸ  yxn(w) /yəʸnúч xn(u)/ < ŋ‘ ⁾‘-; Ave. ‘ē⁾‘-, Khʷāʲ. yyx, Oss. ix ‖ yex, Yazgh. yax, Shugh. οōsh. ₍ā⁾, Wakh. ₍ ⁾, Pers. yax; cf. Sogd. yy n < ŋ‘ ⁾‘-dān‘- ʥὕaϑʧʣʲ , Khʷāʲ. /ēʸmēn a/ ʧϑy (f) (49.) lightning tuntuʲá₎ tʦundʣʲч thunder and ὕʧʥʦtnʧnʥ  twntr /túndər/ tʦundʣʲ Pers. tuntú›, tuntu›ák, Tjk. tundá›, tundú›, dial. Shaydan t ndǘ› t‘ ák ὕʧʥʦtnʧnʥ  < Tjk. t‘ ák, TMast. otə ák < Pers. ātá fiʲʣ ч γāʲʳщ té barq ὕʧʥʦtnʧnʥ  < Ar. BRQ barq; Pers. barq raᴥd(ák) tʦundʣʲ  < Ar. R[D ›‘ d, Pers. ›‘ d (50.) rainbow kam n-i ‘sán-at usá n, kam n-i ‘sán-ʉ usá n  < Pers. kam n-i ‘sán-u usá n asan and uss“ n s bow III.12. Fire ·200· 166. (56.) smoke pazdл‖лpaуʧ)st  pzt- /pəzd-áъ < ŋpá₎d‘( ‘)-, Ave. pazdaiia-; ”æ₎dæg; Hung. ” st dʉᵎd  -, ἵaʲāch. < Pers. d d, TMast. d d, did, TFalgh. dǖd, TYagh. d d, πaʲХq. d, Ir. dʰД 167. (55.) fire ōὕ  ʾʾt(ʾ)›(h), ʾ(ʾ) , ʾrt ʾṯr āš/ < ŋāϑr-, *āt› ; Ave. āt‘›-, āϑr-, Khʷāʲщ ʾ(t)rw, Bactr. у фϸ /ā уuфšъч /ātaʲъч Oss. art, Shugh. Bartang. ₍ c, οōsh. ₍ c, πaʲХqщ yuc, Yazgh. yec, Munj. ₍ ›, Yidgh. ₍ ṛ, Pasht. ›, ἵaʲāch. âṛ, Pers. ā á›, ā ú›, ātí , Tjk. l( w), ₎á›, tá , AfghP. l, ₎á›, té , γāʲʳщ ₎är, té , Pahl. ʾtwr /ā uʲъч Kurd. ar; Eynu. ‘tä l u, al u  ʾʾ›ʾ / flamʣ < ρüщ < ŋ₍‘ł‘ , ŋ₍‘ł‘⁽; Uzb. ł u, Tr. alev; Pers. āl v, āláu, Tjk. al u, láu, l u, ‘láu, TVarz. TYaghn. ‘lóu, TMast. ‘ló’, Shugh. ‘l ⁽, 168. (57.) ash ⁾ k stá›  < Pers. ⁾āk stá›, Yazgh. xək st › dʉᵎda  < Pers. d dá ‘ᴥmák  cf. Ar. Šκ[ ‘m ‘ẗ ϑandlʣ , Pers. ‘m , TMast. BukhAr. ‘ᴥm 169. (191.) to burn sūč- (as č : s čtaч s šta : s čna : s čak)  √s⁽c: √s⁽ t- √ʾs⁽ ṯ- ъ√ʳōč : √ʳu d-/ < ʦ‘uč‘-; Ave. s‘oč‘-, Khōt. s tc-, Oss. suʒ n : so d, Pers. s ⁾tán : s ₎suxs- уaʳúʸʳ : ʳúʸta : ʳúʸʳna : ʳúʸʳak)  √sw s- /√suxs-/ cf. Khōt. vasus- : vasut; Pers. s ⁾tán : s ₎(29.) firewood z(ⁱ)m  zmy ʾzm-y /ɨzmíъ < ŋá ₎ma-(ka-), Ave. ‘ēsm‘-, Khʷāʲщ ʾzm, Munj. , Pers. hē₎úm, TMast. (h)“₎ǘm, (h)“₎ím, TYagh. “₎ím, TVarz. “₎úm, δХὕānХ hД₎əm,Ved. dʰmáIII.13. Settlement 170. (53.) road / path ›ʾ (h) ›ʾ ( )(h) ›ʾϑ ъʲā / ʲōʳ / ʲōt  < ›āϑa-, ›āϑi-; Ave. ›‘ⁱϑДm (acc.), Pasht. lā› < ŋ›āl, dial. l₍ā› < ŋ›āϑД-; Ōʲm. ›āД, Pers. r h, r s, TMast. ra, TVarz. ro(h), ra(h), Pahl. ›ās, Kurd. ›ê, Baὕōch. ›ā(h); Ved. ›‘tʰ₍ -, Armen. ṙah ·201· (25.) village m n ‖ m‘ n  mʾ(ʾ)n / mān/ ʦouʳʣч dʷʣὕὕʧnʥ < ŋdmān( )a- ʦouʳʣч dʷʣὕὕʧnʥ ; Ave. dᵊ -, dᵊmąn‘-, nmān‘-; nmān ‘- ϐʣὕonʥʧnʥ to ʦouʳʣ , Bactr. /mānъч Pasht. m na ʦouʳʣч ʤatʦʣʲὕand , Pers. mān, Pahl. mān; Ved. m na-; cf. Gre. ч Lat. domus, OCS. dom , Lith. nām‘s ‹ l q < Uzb. ‹ï ł ‹, Uygh. qï ł‘‹, Kyrg. ‹ï ł‘‹, ‹ï t‘‹, ‹ï t , Kazakh. ‹ïst‘⁽, Tatar. ‹ı l‘‹, Turkm. ğï ł‘ , Azərb. ‹ı l‘‹, dial. ‹ı l‘⁾, Qashq. ‹ï ł , Turk. kı l‘(k), Ott. ḳï l‘(ḳ) < ρüщ ŋ‹ïś-ŋłā йŋłā‹ = ʷʧntʣʲŞpὕaϑʣ ; εazāʲщ ‹ l q, Pers. ‹ l q, Munj. kə l k, Shugh. ‹ l ‹ deh  d₍⁾(ʾ)⁽ / ʣʸ u/ < ŋd‘h āu-; Pers. dih, OPers. ὕandч province, dʧʳtʲʧϑt (26.) house kat  ktʾ₍, ktʾk qt, qty(y), ktyy qṯy /kə < ŋkąt‘-(ka-); Ave. kata-Bactr. у ф /kad(a)g/, Yagh. kat, Shugh. čДd, οōsh. Khūʤщ čod, Bart. čȫd, οāshrv. čǖd, πaʲХqщ č“d, Yazgh. k d, Munj. ḱay, Yidgh. k i, Pasht. kəl‘ vʧὕὕaʥʣ ч Parth. Pahl. kdg; cf. Ide. Ide. ŋknt‘- : *kan- to dʧʥ x n ʳummʣʲ paʳtuʲʣ  ʾn ⁾ʾn ъʸānъ < -; Bactr. /ʸānъч Wakh. xun, Ishk. xon, Sangl. ⁾ān, Parth. ⁾ʾn; cf. TMast. d ⁾ ná ʳummʣʲ paʳtuʲʣ x ná ʲoom  ʾnʾk(h) ⁾ʾnʾ / < -ka-; Pers. ⁾āná, TMast. ⁾ ná, TYagh. x ná, Kurd. ⁾‘nî; Uzb. ⁾ nȧ, Uygh. ⁾‘nä, Kyrg. qana, Ott. ḫān“, Tr. hane, Tatar. ⁾‘nä, Eynu. xani m n ‖ maʧn vʧὕὕaʥʣ  mʾ(ʾ)n / mān/ < ŋdmān( )a-; Ave. dᵊ -, dᵊmąn‘-, nmān‘-; nmān ‘- ϐʣὕonʥʧnʥ to ʦouʳʣ ч Bactr. /mānъч Pasht. ʦouʳʣч ʤatʦʣʲὕand ч Pers. mān, Pahl. mān; Ved. -; cf. Gre. ч Lat. domus, OCS. dom , Lith. nām‘s (27.) roof k s(ar)  ’ m < Pers. ’ām šamp  < *skamb-; cf. *upa-skamb- to attaϑʦ *fra-skamb- to attach, to ϐuʧὕd ; Khōt. kam- to lift up , Munj. k ’- : k‘’əy- to ʲʧʳʣ , Pasht. āčaʷəl to ovʣʲtʦʲoʷ (28.) door daváʲ ‖ dⁱváʲ  r-y dbr-y ъ əʲíъ < duá›(a)-; Ave. duuara-, Khʷāʲ. r-, Khōt. vara-, Oss. dwar, Wakh. ϐāʲч Pasht. war, Munj. lu⁽ ›, Pers. dar, Pahl. dar, OPers. duvar-, Kurd. d“›î, Ir. хduaʲ-; Ved. dvā›-, Armen. duṙn, OCS. dvьrь, Cze. dv“ř“ dooʲ , dv › уϑouʲtфyaʲd , Lit. dù›₍s, Goth. daur, Ger. T ›, Tor, Gre. ύ ᾱ, OIrl. ·202· III.14. Tools (30.) rūpč  broom < *ra-upa-čД-, Yazgh. rə’áǵ, Wakh. d›“pč, Pasht. ›ē’əʒ; cf. Tjk. ǰ › b < ǰ ₍ + ›u”tán : r b(31.) butter churn kupp  < Tjk. guppí, TMast. k p(p)í, k p(p)í, TFalgh. kuppí tú la  cf. TYaghn. tul á (32.) pestle puš₎á₎  cf. TMast. p kák (33.) hammer á < Tjk. ’ l á (34.) knife ₎ōʲtч k ›d  krt(h) /kaṙt/ < *karta-; Ave. karəta-, Khʷāʲ. krc /kar -/, Oss. kard, Wakh. kəž, Yidgh. k ṛo, Munj. , Pasht. čāṛə, Pers. kārd; Eynu. kard; Cze. kord ʣpéʣ , Hung. kard ʣpéʣ (35.) axe t‘’á›  < Pers. t‘’á›, TVarz. t‘vá›, εazāʲщ t‘⁽á›, t‘’á›, Pahl. tabrak; cf. Rus. п , Cze. toporo ʦʣὕvʣч ʦaʤt ч Ar. tabar tē á adzʣ  t ъtašъ < ŋt‘ ‘-, Ave. t‘ ‘-, Tjk. tē á adzʣ (37.) thread p d < Tjk. p d t › < Tjk. t ›(á) (38.) needle s nčⁱn  < *s nčn < *ʦínč‘n‘-; cf. Oss. suʒin ‖ soʒДnæ, Ishk. ьtun, tьn, Munj. Дžn‘, Yidgh. nǰo, Wa . sunzən, sənǰən, Pasht. stən, Kurd. su₎în, ûj n, Pers. s ₎án, εazāʲщ siz , Pahl. s ʒan (39.) cloth lát(t)‘  < Pers. l‘ttá (40.) ring ·203· ‘ngu ták, , ‘ngu (t)p na  < Tjk. ‘ngu ták, ‘ngu t‘› n, ‘ngu tp ná III.15. Nature (ii) 171. mountain r-y ъ əʲíъ âr mountaʧnч mountain paʳʳ  < ŋgá› -; Ave. gaⁱri-, Bactr. ч / Хʲч aʲ/, Khōtщ ggara-, ggari-, Shugh. οāshrv. žД› ʳtonʣ , οōsh. Bart. žē› ʳtonʣ ч Khūʤщ žæ› ʳtonʣ ч πaʲХqщ ž“› ʳtonʣ ч Wakh. ar ʳtonʣ , Munj. ā› paʳʳ ч Yidgh. ar ʳtonʣч mountaʧn , Pasht. ar, Ōʲm. g›Д, ἵaʲāch. gir, Pahl. ar, Ved. girí-, OCS. gora, Ide. *g orʜ-; Alb. gur rock ; Gre. έᾱ northwind (< 268 *mountain wind; MALLORY – ADAMS 2006, 121) ; Lith. g › forest ; cf. Buʲūshaʳ₎Х oro ʳtonʣ kʉʰ  < Pers. k h, TMast. k , TVarz. kuh, Pahl. kwf /₎ōʤъч OPers. k‘u”‘-, Ave. kaofa-, Munj. kifa, Wanj. kub, kup, Ir. ŋk‘u”‘-; Eynu. kox (60.) gold  < Ar., Pers. t zar (occ.)  zyrn /zeṙn/ < *ʣᛑn ‘-; Ave. zaraniia-, Khʷāʲщл zrny ъzʧʲnХъчл Baϑtʲщл л ъzaʲъчл ἵʣʲʳщл zarr, Pahl. ₎‘›ēn, OPers. daraniya-; Ved. h훑ṇya-, Ide. ŋgʰl“n o-;лϑʤщлδʲʣщл ᾱ л[ ] daric – ʾryk III.16. Colours 172. (150.) red kⁱm r, kam r (arch.)  k›m(ʾ)₍›, kyrmyr qrmyr qyrmyr Pahl. k‘›mД›, Armen. karmir; cf. ByzGre. [ ]; cf. Ar. QRMZ qirmiz, γāʲʳщ q“›mé₎, Tr. kı›mı₎ sʉrx  < Pers. surx, Pahl. suxr, TMast. s ›⁾, OPers. ϑuxra-, Kurd. sor, Baὕōch. suhr, s h›; Ir. ⃝ *ʦuxra-; Ave. suxra-, Bactr. /surx/, Khōtщ surai, Oss. s ›⁾ ‖ surx, Wakh. səkr, Ishk. sь›⁾, Munj. sərx, surx, Yidgh. surx, Pasht. s ›, srə, ἵaʲāch. sú›ku, Ōʲm. u , Ved. śuk›á173. green zaʲ na (arch.)  zr ⁽nʾk zr wnyy /záṙ ōn / cf. Sogd. zr wn ъzaʲ ōnъ pὕantч vʣʥʣtaϐὕʣ 268 See also Slovak hora mountaʧnлъълʤoʲʣʳtлъълmountaʧnлϑovʣʲʣdлʷʧtʦлʤoʲʣʳt щ ·204· sabz, s‘u₎  Pers. sabz, TVarz. s‘u₎, εazāʲщ s‘ú₎, Shugh. sāvʒ, Ir. *ʦ‘p‘č ‘kap t(a) ‖ k p t(a) (arch.), kabʉt ʥʲʣʣnч ϐὕuʣ  kpwt(k) /kə ϐὕuʣч ʥʲʣʣn < ŋk‘p‘ut‘-ka- ϐὕuʣ ; Pers. kab d, TMast. kə’ǘd, Pahl. k‘’ t; Armen. kapoit 174. yellow z rta (arch.), zard  Sogd. ₎₍›t(ʾ)k zyrtyh ъzēṙ < *ʣá› t‘-ka-; Ave. zaⁱrita-; Yazgh. Wakh. z rt, Shugh. οōsh. z rd, Ishk. zord, Munj. Yidgh. ₎Дt, ἵaʲāch. ₎Дt , Pers. zard, Kurd. zer; Hung. ₎öld 175. (148.) white , ‖ sⁱ  ʾspʾ₍t(ʾk), ʾsp(ʾ)₍tk, (ʾ)spʾ₍tk, (ʾ)spʾ₍t₍ spyṯy /ᶤ sⁱp ta (arch.), *ʦuá t‘-(ka-); Ave. sp‘ēt‘-, Khʷāʲ. spyd k, Khōt. śśДt‘-, śśД₍‘-, Munj. spД, s‘”Дd, Pers. sip d, isp d, saf d, Tjk. saf d, sap d, TMast. sə”éd, γāʲʳщ sef d, εazāʲщ s‘”ít, Shugh. s‘”ēd, Ishk. safed, Yazgh. səpid, Kurd. spî, Ved. śv“tá-, OEng. h⁽Дt, δótщ ƕeit-s, OCS. svět ὕʧʥʦt 176. (149.) black , sⁱ₍áʰ  Sogd. ʾ⁽ ⁽ ъšāu/ šōu (arch.), < *ʦ āu‘-, Ave. s āuu‘-, Sarm. [ ], Khʷāʲщ sʾ⁽ /ʳāʷъч Oss. saw, Ishk. u, su, Wakh. ⁽, Pers. h, εazāʲщ s ₍á, ; Tr. siyah, Cr.Tatar. siya m‘₎áng  cf. Ḥ‘l‘ng[ ’йu] in Sarghuὕām уʧщʣщл Bὕaϑ₎лʷatʣʲ члtʦʣлʳʣϑondлpaʲtлʧʳлpʲoϐaϐὕyлἵʣʲʳʧanлaʳл ʷatʣʲ л ʧs ⁽ol ké or ⁽ol kí in SarghuὕāmХ, but it is uncertain whether *malang is a SarghulāmХлʷoʲdлoʲлʧʤлtʦʣлʣὕʣmʣntлʲʣaὕὕyлmʣanʳл ϐὕaϑ₎ ) III.17. Time 177. (118.) night xⁱšáp; ‘’, ‘u  ʾ p-h ʾ⁾ p-ʾ(h), ⁾ p-ʾ ⁾ p-ʾ / ʸšəpáъ -; Ave. ⁾ ‘pā-, Khʷāʲщ ʾ , xb, Khōtщ ‘vā-, k ap-, Oss. æ⁾sæv, Shugh. ⁾ā’, < οōsh. ⁾‘’, πaʲХqщ ⁾o’, Yazgh. ⁾əb, Ishk. ‘’, Yidgh. ⁾ ovo, Munj. ⁾ ‘⁽ , Pasht. ⁾‘’, Pers. ‘ƀ, TVarz. šau, Pahl. ‘p > ‘ , Kurd. “v; Ved. k -; Eynu. ä’ 178. (117.) day mēʳ / mēt  Sogd. my m(ʾ)₍ my , my(y) myϑ, mỿϑ, myd ъmē / < ŋmá ϑā-; Ave. m‘ēϑa- unʳtaϐὕʣч changing (with nʧʥʦtф ч Khʷāʲщ my ъmē /, Yazgh. miϑ, Shugh. mēϑ, οōsh. Khūʤ. Bart. οāshrv. mДϑ, πaʲХq. maϑ, Ishk. may, Sangl. mē , ḳēϐāk. mД, Munj. Yidgh. mД⁾ n r уdayфὕʧʥʦtч day  nwr /nūʲъ < ŋn ›‘- уdayфὕʧʥʦtч day ч Ave. n ›əm, Khʷāʲ. nwr /nūʲъч Pers. n ›, TMast. nir ὕʧʥʦt ; BukhAr. n › day ·205· r z rwc ъʲōčъ < ŋ›‘uč‘-; Bactr. /ʲōϑъч Pasht. rwaʒ, colloq. wraʒ, Pers. › ₎, εazāʲщ r z, Pahl. › ʒ, Kurd. roj, Shugh. › ₎ (119.) morning fⁱr nta ‖ fⁱ  ›ʾ(ʾ)k ›ʾ(ʾ)k ”›ʾk ”›ʾ‹ /ʤʲā₎ъ < ŋ”›āk‘-; Oss. rag, Wakh. və› k; cf. Ved. p›ā(ñ)k- ʧn ʤʲont ч Welsh rhag ʧn ʤʲont ч Corn. rag ʧn ʤʲont ч Bret. rak ʧn ʤʲont sa á›  < Ar. S R sa ar, Pers. sa á›, TMast. sə á›, səhá›, Wakh. s‘hā› č t‘gáʰ(i)  Tjk. č tg h,  pag ʰ, < Tjk. pag h, , TMast. pəgá, Ave. upa-gāϑ’ m moʲnʧnʥчлdaʷn;лtʧmʣлoʤлtʦʣлfiʲʳtлmoʲnʧnʥлpʲayʣʲ л ’ʾm ъ āmъ moʲnʧnʥчлdaʷn cf. Pers. ’ām  ⁽₍ʾ⁽s wyws ъʷyūʳъ Ave. viiusa ʾ⁽cʾ‹ Ave. vДt‘›‘, v čāk < u t(‘)›āk(120.) noon nДm› z, nДm› ₎  nymy (h) nymy / < ŋn‘ m‘-má ϑā-, ŋn‘ m‘-›‘uč‘- mʧdday ; Pers. , nДm› ₎ aʲnám  Pasht. ‘›má < Ir. *garma- ʷaʲm  < Tjk.  ryp - /rep áъ < ŋ›áp ϑ ā (121.) evening / afternoon ₍ʾ›ʾk ₍ʾ›₍₍ vⁱy ra  < ŋ‘’ -‘ ā›‘-ka-, Khʷāʲ. ’ ₍ā› < ŋ‘p‘-‘ ā›‘-; Yazgh. biyir, Shugh. οōsh. ’ ₍ ›; ἵaʲāch. ⁽₍ā› ’ēg ʰ  vʣčʣʲ < Tjk. ’ēg h xⁱš m dʧnʣʲ , m ʣvʣnʧnʥч aʤtʣʲnoon  ⁾ ʾm ъʸšāmъ < ŋ⁾ ”n ‘-; Ave. ⁾ ā”n‘ ‘-; Shugh. ⁾ m, Yidgh. ⁾ ēm‘- dʧnʣʲ ч Pasht. ; Parth. ʾm, Pahl. ⁾ ām, Pers. ām; Tatar. ‘⁾ ‘m ʣvʣnʧnʥ pʲayʣʲ ч Georgian v‘⁾ ‘m (122.) yesterday ·206· pⁱy n  p₍ʾʾnʾkh /py nāъ < ŋ‘p‘-‘ ā-na-(ka-); cf. Pasht. par n, Wa . , ən(d), Sangl. pā›u₎d < *para-aʣna-/aʣni(123.) today n(n) r / dn r  < Yagh. Дt / Дd n r tʦʧʳ day ; cf. Shugh. n › (124.) tomorrow ›ʾ(ʾ)k ›ʾ(ʾ)k ”›ʾk ”›ʾ‹ /ʤʲā₎ъ fⁱr nta ‖ fⁱr k  < ŋ”›āk‘-; Wakh. və› k; cf. Ved. p›ā(ñ)k- ʧn ʤʲont , Welsh rhag ʧn ʤʲont , Corn. rag ʧn ʤʲont , Bret. rak ʧn ʤʲont ,  , , TMast. pəgá, Ave. upa-gāϑ< Tjk. (125.) week há”tá  ʾ t(ʾ)m₍ ʾ tʾm₍ /ə də / < h‘”tą-má ϑā-; Ir. ŋh‘”tą-ka- > Pers. h‘”tá > Shugh. ‘”tā, Tr. hafta, Kazakh. apta; cf. Gre. ч MGre. ά ч Fr. semaine (126.) month mʾ (h), mʾ⁾ mʾ⁾ ъmāʸъ m ʰ, m‘ʰ; mōʸ (arch.)  < ŋmāh-; Ave. OPers. māh-, Bactr. у ф /māуʦфъч Khōtщ māstä, Oss. mæ ‖ mæ₍æ, Shugh. m st, οōsh. mēst, πaʲХqщ most, Yazgh. mast, Pasht. m₍ā t, Wakh. m y, Pers. , TVarz. mo(h), ma(h), Kurd. meh, Ved. mās179. (127.) year sr -y sr ( )-y srd-y ъʳʣw íъ s l < *ʦ›d-; Ave. sarəd-, Khʷāʲщ sr /ʳaʲ -/, Bactr. /sarl/, Khōtщ s‘lД-, Pers. sāl, Kurd. sal, OPers. ϑard-, Ved. ś‘›ád- autumn y ʳō (arch.)  < ŋāʦaka-; Oss. az ‖ anz III.18. Adjectives (ii) 180. (136.) hot rm ъ aṙm/ aʲm  < *garma-; Ave. garəma-, Khʷāʲ. rm, Khōtщ g›ām‘-, Oss. qarm ‖ arm, Ishk. orm, Sangl. ›m, Pers. garm, Munj. gərm, Shugh. gā›m, Ishk. garm, Kurd. germ, Baὕōch. garm(ag), Skt. gʰ‘›m‘-, Gre. ч Lat. formus, Eng. warm, Ger. warm, Cze. žá›; Urd. garm 181. (137.) cold ʳōʲt  srt /saṙt/ < *ʦarta-; Ave. sarəta-, Khōt. sāḍa-, Wakh. s ›, Pasht. s ṛ (f. saṛa), Ōʲm. sālᵃ, Pers. sard, ·207· Pahl. sart, Baὕōch. sart, sard, Kurd. sar, Goth. kalds, Eng. cold, Ger. kalt, Rus. , Cze. chlad, Lit. ált‘s; Urd. sard 182. full pun(n), púnna  Sogd. pwrn-y pwn-y ъpuwníъч pwn /pun(n)/ < ŋp n‘-(ka-); Ave. pərəna-, Bactr. /purr/, Khōt. purra-, Pasht. pur, Pers. Kurd. Baὕōch. pur, Ved. p ›ṇá-, OCS. pl n , Rus. п ,п , Cze. pln(ý), Lit. pìln‘s, δótщ fulls, Ger. voll, Eng. full; cf. Lat. plēnus 183. (129.) new náʷa  n⁽ʾk⁽ nwyy /nə < ŋnáu‘-ka-; Khʷāʲ. n⁽ʾk /naʷāʥъч k, Bactr. ч /nug, nag/, Oss. nog (arch. næ⁽æg) ‖ næ⁽æg, Ishk. nu⁽ k, Sangl. nu⁽ k, Shugh. naw, Yidgh. nowo o, Pasht. nə⁽‘ (f. nə⁽ē), Parth n‘⁽āg, Pers. n‘u 184. (128.) old pД› oὕd (of aʥʣф  , Ir. *parya-; Ave. p‘› pʲʣvʧouʳ ; BukhAr. pД› Pers. pД›, Bactr. kúʰná oὕd уʧnanʧmatʣф  < Pers. kuhná, kuhán, Tjk. k hná, kuhán, TMast. k ná, Pahl. kahwan, Uzb. k hnȧ, k hnȧ, Kazakh. könė, Tr. köhn“, Qashq. köhnȧ, kohn qad m(á)  < Ar. QDM ‹‘dДm(‘ẗ), BukhAr. ḳ‘dДm, Malt. qadim, Pers. qad m, ‹‘dДmá, Wakh. qadim > tčn /  wtcny(y), wcny ʾwcny /üɨ < ŋu -t‘č n‘-ka185. (130.) good wp xwp /ʸūpъ ⁾ ’ < Ir. *hu-apa-, ŋhu‘p‘-; Khʷāʲщ xwb /ʸūϐъч Bactr. /ʸūϐъч Pers. ⁾ ƀ, Fārs. colloq. xob, Skt. svapa-s, Uzb. ⁾ ’, ⁾ p na z  Sogd. n z-y /nə zí/ > Pers. na z, TMast. naxs 186. (131.) bad ntʾk(ʾ), ntʾkk ntʾ(ʾ)k(ʾ), ntʾkk ndʾk nṯʾq ъ áṁ gánd‘  < *gand-āk‘-; Tjk. g‘ndá, Ishk. ganda; Parth. gndʾg / ʳtʧn₎ʧnʥ ч Baὕōch. gandag; Ved. g‘ndʰá- ʳmʣὕὕ ; Uzb. gȧndȧ, BukhAr. ganda  (y)z-y, ʾ (y)z-y, (y)j-y, ʾ j-y / ží < ʣžíъ < ŋ’éž < ŋ’á₎d ‘-; Pers. ”‘ž, ‘ž < Sogd. 187. rotten p ta  pwṯky ə < ŋp t‘-ka-(ka-); Ave. p t 188. dirty ·208· aždл cf. TMast ‘žd  , Shugh. č ›k n < Pers.  rym(nyk) rym, rymny(y) rym Parth. Pahl. ›ēm  ʾʾ wst ʾʾ wstk ə Parth. ʾgwd, ʾgwst, Pahl. ʾgwh189. straight razk, › st  › t(h) /rəštъ ʲʧʥʦtч tʲuʣ < ŋ ʣuka-; Yazgh. ›‘₎ǵ, Sangl. › sk, Munj. wurzug, Ide. ŋ›“g-to-, Lat. rectus, Ger. Recht < * ›ā t‘-; Ave. -, Khʷāʲщ › t /ʲaštъч Khōtщ rra ṭa-, Pers. ›āst, Hazār. › s, Pahl. уρuʲʤānф ›ā t, OPers. ›āst‘-, Kurd. rast, Oss. rast; Uzb. › st, colloq. › s, Kyrg. s  ”› ṯy /frə < *fə›á t‘ < *fra-›‘ t‘-ka-, cf. Sogd. √”›ʾ₍ √”›₍ž : √”› t- ъ√ʤʲēž : √ʤʲštъ to ʳtʲaʧʥʦtʣn < *fra-raʣ‘ ‘- : *fra-›‘ t‘ p›₎pʾ› /páṙzpāʲъ 190. round l nda  < Tjk. l ndá la  cf. Yagh. Дl- to ʲoὕὕ л<лρ₍₎щл ēlДdán : ēlk l lá  < Pers. gul lá, Tjk. kul lá; TVarz. kulolá, γāʲʳщ gol lé ʲound  wrs ъ uṙs/ < *gar(t)ʦu-; cf. etymologically unrelated Ar. QR qur > Tjk. qur (ák) > Yagh. ‹u›s(ák)  ʾskwrnkh /ᶤskúṙn / Ave. skarənā pr rsʾy / páṙ əws / < ŋp‘› -g›ʦ-aka-; cf. Ave. gərəsna191. sharp tД› aʲʲoʷ  tr -y ṯr -y ъtʧw íъ < *tigra-; Ave. tigra-, ti ra- ʳʦaʲp ч ti ri- aʲʲoʷ ч Khʷāʲщ č r ъϑʧ ʲъч Khōtщ ttД›‘-, Oss. c › ‖ cir , Ishk. tir , Munj. tər a, Pers. tД› aʲʲoʷ ч OPers. tigratē₎  < Pers. tē₎, Kurd. tûj 192. dull kunt  tupý Pers. kund, BukhAr. kund ·209· 193. smooth ὕʣʸna  hamw r  < Pers. hamv r, Shugh. ‘m⁽ ›, ‘n⁽ › fit  < Tjk. fit 194. (132.) wet tan(n), tar  trn /taṙn/ < t‘u›n‘-; Ave. taorna-, Khʷāʲ. trn /tarn/, Pers. tar 195. (133.) dry ‹ ‹ < Uzb. ‹ ‹, Kyrg. Tatar. qaq, Tjk. Shugh. ‹ ‹ ⁾u k  ʾ k-w (ʾ) k-w, k⁽₍(₍), ‹⁽₍(₍) /ᶤš₎úч ᶤš₎ə < ŋhu k‘-, ŋh ku-, ŋh kuu‘-ka-; Ave. h ku-, Oss. ⁾ sk ‖ ⁾usk, Pasht. ⁽uč, Pers. ⁾u k 196. correct d rʉst  ʳpʲávnЕ < Pers. du›úst razk, › st  › t(h) /rəštъ < ŋ ʣuka-; Yazgh. ›‘₎ǵ, Sangl. › sk, Munj. wurzug, Ide. ŋ›“g-to-, Lat. rectus, Ger. Recht < * ›ā t‘-; Ave. -, Khʷāʲщ › t /ʲaštъч Khōtщ rra ṭa-, Pers. ›āst, εazāʲщ › s, Pahl. уρuʲʤānф ›ā t, OPers. ›āst‘-, Kurd. rast, Oss. rast; Uzb. › st, colloq. › s, Kyrg. s t r  < Uzb. t ›ï, Tr. doğ›u, Kypch. to ru, Kyrg. t ›‘, Kazakh. tura, Karakalp. tu⁽›ï, Tjk. t r , TMast. t ›í, εazāʲщ t ›í 197. (140.) near nzt-w /nəzdúъ nazd k  < n‘₎d ‘h-; Ave. nazda-, nazdiiah-, Bactr. /nuzd/, πaʲХqщ nizd, Pasht. , , Pers. nazd( k), TMast. n‘₎(₎)ík, Pahl. n‘₎dДk, Kurd. n ₎ûk, nê₎îk, nazik, Baὕōch. n‘₎Дk, n‘₎Д⁾, n‘₎Д, Ved. nédД₍‘s , TMast. qə›í’ Ar. QRB ‹‘›Д’, BukhAr. ḳ‘›Д’, Pers.  ʾw ъ āʷъ  n ʾnt n ʾnt, n ʾynth n nd nbndy, nbnṯ(y), nbnt /nɨ áṁd nɨ éṁd/ 198. (141.) far wr(h) wr dwr ъ ūʲъ d ᵎr  < ŋd ›‘-; Ave. d ›‘-, Khōtщ dura-, Wakh. ir, πaʲХqщ ar, Pers. d ›, TMast. d ›, dir, TFalgh. dir, Ved. d ›á-, Hind. d › 199. (127.) right ·210· ʲáz₎уaфч › st  › t(h) /rəštá/ < ŋ ʣuka-(ka); Yazgh. ›‘₎ǵ, Sangl. › sk, Munj. wurzug, Ide. ŋ›“g-to-, Lat. rectus, Ger. Recht < * ›ā t‘-; Ave. -, Khʷāʲщ › t /ʲaštъч Khōtщ rra ṭa-, Pers. ›āst, εazāʲщ › s, Pahl. уρuʲʤānф ›ā t, OPers. ›āst‘-, Kurd. rast, Oss. rast; Uzb. › st, colloq. › s, Kyrg. s  wrzr-w, wyzr-w wyzr-w ъʷʧwzʲúъ Ave. vərəzra200. (139.) left č‘p(p)á, č‘p  Tjk. č‘p, πaʲХqщ čop, Kurd. ç“p, BukhAr. č‘pp‘ sʾpt(⁽) sʾpṯ t/  (175.) whole  Ar. TMM t‘mām, Pers. (178.) broken áṁʸ°ə ʸ°ə unʸaʳta  ʾwxwsty ʾn wsty xwsty ʾwxsṯy / ʸ°ə < ŋ(‘u‘-/ham-)x asta-ka-; cf. Pahl. xwastan; cf. Pers. suffix - , e.g. ϐʲo₎ʣn < k‘stán : k‘n- to ϐʲʣa₎  < Tjk. , TMast. kĭlét, TVarz. k‘lét; Pers. (Lu át-i Furs) klʾt < Sogd. ??? vayr n  Pers. v‘ › , TMast. , v‘ › , εazāʲ. ’ē› , Pahl. ‘pē›ān, Ishk. veron; BukhAr. ’“ ›ān, u‘ › n III.19. Adpositions 201. -sa  pa-  at -sʾ› -sʾ(›) /- ṙ, < *ʦā›-; Khʷāʲ. -sʾ› /-ʳāʲъч Pasht. -sara cf. Sogd. ʾ ’ʾ ъ āъ par  pr /pər/ < *upari-; Ave. upari-, Khʷāʲ. (-)par, Pasht. pər, Pers. bar  kw kʾ⁽ kw, qw qw ъ₎ōъ cf. OCS. k 202. in číntíʲ  c(y)ntr c(y)ndr ъčáṁdər, čɨṁdər/ < ŋh‘čā-antar-; Wa . ₎dᛓ; cf. Pers. (‘n)dá›, Tjk. dar, γāʲʳщ dä›, TVarz. da(r), -da -n t  ·211· (146.) above -ʳáʲуʧфч -ʳáʲaʧ  na, nad, u Oss. -sæ› ⃝ ʾsk-(ʾ) ʾsk-ʾ(ʾ) ʾsk-(ʾ), sky /əʳ₎ā, šХ (arch.)  Ave. uskát, Khʷāʲщ ʾsk, Pasht. hask; cf. Yagh. Дm n uppʣʲ village; upper part of village of Gh‘›mēn in Yaghn ’ cwpr ər/  < ŋh‘čā-upari(147.) below -tá₎ʧч -tági  cf. Tjk. tag ϐʣὕoʷ  cʾ r, cʾ( ›)sʾ› cʾ r(sʾ›) cʾ ›(pʾ›), cʾ( ›)sʾ› cʾpʾ›, cʾsʾ(›) / ər, ə у ər)s < ŋh‘čā-adari203. with py( ) /piуšф/ -pi  Khʷāʲ. py /pi/ ,  < Tjk. , , TMast. qətí, Shugh. qati  p›ʾ(₍)⁽ p›ʾ(₍)⁽, pryw pryw prw prau /pər u/ < *upari-á u‘- at onϑʣ III.20. Conjunctions 204. and -(a)t; -(y)ʉ, -(v)ʉ, -(y)i  ʾt(y) ʾṯ / tуíф/ < ŋutā; Ave. uta, Khʷāʲ. ʾwd /ud/, Bactr. ч ч ʾwd, ʾwṭ /ud, ut/, Oss. -ta, Yazgh. -ata, -at, -a, Ishk. -ьt, Shugh. -(a)t, Pasht. aw, Pers. -u, (-yu, -vu), TMast. - , -i, Pahl. u , OPers. u-tᵃ-a , Kurd. û  ʾr / / < *›; Ide. *h (e)r; Gre. чл ᾽, ἄ у фчл ἱʧth. ›/‘›, Latv. ì›/ar; Tokh raŞл emphatic particle va, wa  w- /wə-/ < Ar. wa, BukhAr. Malt. u, Hebrew ve, Syriac ; Pers. va, Kurd. ve, Pasht. wa, Uzb. va, Tr. ve, Azərb. və 205. if k (ʾ) k qd ъ₎a ч kə áъ ʷʦʣnч ʧʤ kad ʷʦʣn  Sogd. Ave. ka a-; Bactr. /kad/, Oss. kæd, Pasht. kəla, Pers. k‘ ; Ved. k‘dáagá(r)  < Pers. ‘gá›, poet. gar, Tjk. colloq. ‘gá, TYagh. ‘gá(›), γāʲʳщ ägär colloq. äg“, εazāʲщ ‘gá; ·212· OPers. hakaram onϑʣ , Ave. hakərət, Shugh. aga(r), Bart. agar, agi, πaʲХq. ‘gá›; Uzb. ȧgȧr, colloq. ȧyȧr, Chaghat. ägä›, Tr. “ğ“›, Qashq. ȧyȧr, ȧgȧr, Turkm. “ er, Kyrg. eger, ₍ʧžnщ dial. äg“›, Tatar. ägä›, Kypch. ėgä›, BukhAr. ‘gá›, ‘gál 206. because nah piti ’á⁾ ‘ ‖ nⁱh piti ’á⁾ ‘  < Yagh. nah- ʣnϑὕщ particle of demonstrativʣʳ (Sogd. -nax: wn , ⁽nʾ xwnw, -nax/) + ípt tʦuʳ + Tjk. ’‘⁾ ʤoʲ ч AfghP. ’á⁾č-e hwnx ʾwnw nʾ⁾  < Pers. -ki < ŋč -g‘un‘- + ŋkáh ā cʾn⁽ṯ ъč nūt/  pʾ›⁽t₍ pʾ›⁽ṯy III.21. Name 207. name nʾm ъnāmъ n m < n man-; Ave. nąm‘n-, nām‘n-, Khʷāʲ. nʾm /nāmъч nʾm k /n mag : nāméʥъч Bactr. /nāmъч Khōt. Tumshuq. nām‘-, Oss. nom ‖ non, Pasht. n m, πaʲХqщ num, Pers. nām, Pahl. nām, OPers. nām‘n-, Kurd. nav, Baὕōch. nām, Ide. *h néh men-, Ved. n man-, Armen. anun, Gre. ὄ -, Lat. n m“n, Ger. Name, OCS. jьmę, OCze. jmě, Rus. , OIrl. aınm(m), Irl. Gael. ainm, Bret. h‘ñv, Welsh enw, Hitt. lām‘nVocabulary of YaghnōϐХлandлπoʥdʧanлϑonʳʧdʣʲaϐὕyлdʧffer – the difference is caused by several factors such as non-existent contact between both Sogdic dialects for approximately 1000 years, intensive contact of YaghnōϐХлʷʧtʦлTajik (and to a lesser extent contact with Arabic and Turkic, presumably via Tajik) on one hand, on the other hand some Sogdian words show contact with Sanskrit (mainly Buddhist terminology), Aramaic (in Christian and to a lesser extent in Manichaean texts) 269 , and Turkic (which appears in secular texts, namely from documents found at the Mount Mugh). There are also observable Sogdian contacts with Classical Persian, but it seems to me that there was much more Sogdian influence on Persian than Persian influence on Sogdian. In contemporary YaghnōϐХлtʦʣʲʣлʧʳлaлʥʲʣatлamountлoʤлὕoanʳлʤʲomлуoʲлvʧaфл Tajik – there are approximately 48% loan-words and some 6% word are YaghnōϐХ–Tajik compounds and other approximately 19% words are so-called compound verbs (presumably majority of them calqued from Tajik) – remaining 27% of words are genuine YaghnōϐХлуNἴσÁἰ [in print]). Both languages also show similar patterns of word-formation, even YaghnōϐХлϑaὕquʣʳлʤʲomл Tajik show some Sogdic patterns of word-formation. In YaghnōϐХл tʦʣʲʣл ʳtʧὕὕл ʲʣmaʧnл manyл In this case I do not take in account Aramaic ideograms used in texts written in the Sogdian script – such ideograms were very likely read as Sogdian words as they show e.g. Sogdian inflectional endings. 269 ·213· suffixes attested in Sogdian, unfortunately many of such suffixes are unproductive in the contemporary language (cf. δκπл έŚю5-1166; LIVSHITS – KHROMOV 1981, 434-449; KHROMOV 1987, 665-670). Some YaghnōϐХл ʷoʲdʳл ʦavʣл noл πoʥdʧanл ʲʣʳponʳʣʳчл πoʤ ya Petrovna Vinogradova quotes several of them: ú›d‘ ʣyʣ чл ayk dauʥʦtʣʲ чл rax mouth 270 , n s- to ta₎ʣ (VINOGRADOVA 2000b, 310), there are many other words without Sogdian etymology, but some of those words ʦavʣлʣtymoὕoʥyлʧnлtʦʣлἵāmХʲлlanguages, e.g. Yagh. ayk dauʥʦtʣʲчлʥʧʲὕ лmayлϐʣлϑonnʣϑtʣdлʷʧtʦл Yazgh. ‘č‘ǵ, Shugh. ācчл οōsh. acчл πaʲХqщл oc; Yagh. Saponaria Griffithiana Boiss. plant л ~ Khūʤщл wu m; Yagh. p‘›ám Cousina umbrosa Buge plant л ~ Khūʤщл piram, Yagh. ‖л ʦomʣ-made paper-ὕʧ₎ʣл tʦʧnл ϑottonл ϑὕotʦ л ~ Shugh. ⁾ ⁽Дnǰ, Bart. ⁾ ⁽Дnč, Khūʤщл ⁾ ⁽Дnč, ⁾u⁽‘nǰчлοōsh. ⁾ ⁽ nč; Yagh. x ϑʲoʷчлmaʥpʧʣ л~ Shugh. Khūʤщлk ⁾épc and many other. The YaghnōϐХ–ἵāmХʲХлvoϑaϐuὕaʲyлmayлϐʣлϑonnʣϑtʣdлʷʧtʦлlocal ecology and semi-nomadic lifestyle or it may even be associated ʷʧtʦлtʦʣлἵāmХʲ-εʧndū₎ush Sprachbund mentioned in chapter I.1.1.4.b. Some other YaghnōϐХлʷoʲdʳлʦavʣлϐʣʣnлʲʣϑoʲdʣdлʧnлpaʳtлyʣaʲʳчлϐutлtʦʣyлaʲʣлnotлuʳʣdлʧnлtʦʣл ʲʣd чл yʣὕὕoʷ чл u ϐὕaϑ₎ чл ʷʦʧtʣ чл v rʉk modern language: man appὕʣ чл ʣyʣϐʲoʷ чл , ‘l’‘lá muϑʦчл many л andл manyл otʦʣʲл уϑʤщл BOGOLYUBOV 1966, 359; KLIMCHITSKIY 1940; NἴσÁἰ [in print]), some other *Early Modern YaghnōϐХлʷoʲdʳлtʦatлʷʣʲʣл also similar in Sogdian were replaced by their Tajik similar-sounding counterparts: ŋv (Sogd. ā фл ʥaʲdʣn л×лρ₍₎щлşлḲagh. ’ , ŋm ⁾ (Sogd. mā⁾фл moonчлmontʦ л×лρ₍₎щлşлḲagh. m h montʦ (cf. BOGOLYUBOV 1966, 359) or ŋvДm (Sogd. Дmфл ʤʣaʲ л×лρ₍₎щлşлḲagh. ’Дm. 270 Yagh. rax has attested Sogdian form r ʾk /rə . ·214· IV. Conclusion In the presented thesis I tried to present main development features of the Eastern Iranian languages. The main attention was paid to the development and interrelation of Sogdian and YaghnōϐХл – two closely related languages of the Northern branch of the Eastern Iranian languages. YaghnōϐХлandлπoʥdʧanлʷʣʲʣлʳtudʧʣdлtoʥʣtʦʣʲлʷʧtʦлotʦʣʲлβaʳtʣʲnл Iranian languages, primarily with the languages of the ἵāmХʲs. I have compared all documented Eastern Iranian languages to the sketch of contemporary development of the languages in focus – I have tried to outline their basic development in phonology and morphology in the first part of the presented thesis. By a thorough study of the Eastern Iranian languages I have found another phenomenon, which should be carefully investigated – (re)classification of the Eastern Iranian languages. As I have mentioned in the chapter I.1.2. there is commonly accepted grouping of the language group in focus into the Northern and Southern branch, but as I have observed, there are no given criteria for such grouping. In the Table 31 I put down some thirty isoglosses that I have observed among the Eastern Iranian languages, but according to the isoglosses presented in the Table 31 there are no many really distinct features that can differentiate tʦʣл λoʲtʦʣʲn л andл πoutʦʣʲn лϐʲanϑʦʣʳщлAϑϑoʲdʧnʥлtoлaлpreliminary analysis of Eastern Iranian isoglosses there can be defined at least five groups/branches: I Northern (Sogdo-Scythian), II North-eastern (Saka), III Central (PāmД›), IV Southern (Paṭhān) and V South-eastern (ђ nd kush) groups. Problematic is classification of Avestan (cf. ÈDEL MAN 1986, 6-7 with bibliography), KhʷāʲʣzmʧanлуϑʤщлÈDEL MAN 2000a, 95; ÈDEL MAN 2008, 6; ÈDEL MAN 1986, 6) and Bactrian – presented classification was based mainly on Modern Eastern Iranian languages. Some of isoglosses presented in the Table 31 can be demonstrated on following four examples (all examples are supplemented by forms in Classical Persian): ŋčá m‘n- ʣyʣ c( )m-y cm-y(y), c m-y c(y)m-y, c m-y ъčɨуšфmíъ;лἴʳʳщлcæst, casm ‖лcans I ʷʧndoʷ-opʣnʧnʥ II Khōtщлts“ʼ m‘nIII Ishk. com, Sangl. cāṃчл ḳēϐщ c m, Munj. č m, Yidgh. č‘m, Shugh. Baj. cēm, , Wakh. čə( )m οōsh. Khūʤщлcām, Bart. cēmчлοāshrv. cДmчлπaʲХqщлcem, Yazgh. V Ōʲmщлc mД, čДm, c m ? Khʷāʲщлcm-, cm- /camma/, Ave. č‘ m‘n- [Pers. č‘ m] *ϑ› ‘- tʦʲʣʣ I ʾ ry ʾ ryw (ʾ) ry ʾ ry(y) ₍ / šaʧ/, Yagh. s‘›á₍ ‖лtⁱ›á₍, Oss. æ›tæ II Khōtщлdrai, Tumshuq. dre III Yidgh. ⁾ⁱray, ⁾u›o₍, Munj. ⁾ ›‘₍, Shugh. aray, Baj. Baʲtщл οōsh. ‘›ā₍чл πaʲХqщл aroy, Ishk. › ₍, Sangl. › ₍, Yazgh. c ₍, Wakh. t› (₍) {Bactr. α ηι /həʲēyъ} IV Pasht. d›ē, Wa . dre V Ōʲmщл ȫ, ṛДчлἵaʲāch. Д, u ? Khʷāʲщл ₍ ъšēъчлAvʣщлϑ›ā [Pers. sih > se] ·215· ŋum you (ʾ) mʾ w, ʾ mʾ h ʾ mʾ⁾(⁽), mʾ⁾ mʾ⁾ /ᶤ I Oss. s m‘⁾ ‖ sumax II Khōtщлuhu, , umä, LKhōtщлama , Ishk. tьmь⁾, Sangl. təməx, Munj. m ”, Yidgh. III Wakh. οōsh. Khūʤщл tamaчл Baʲtщл οāshrv. t‘mā чл πaʲХqщл t‘m‘ {Bactr. ч ъtōmāʸчлtumāʸчлtamāʸъ} IV Pasht. , , Wa . tās V ἵaʲāch. ⁽āчлŌʲmщлt s, t₍ s ] ? Ave. ₍ žəm, Khʷāʲщлh y [Pers. gh. , , , Shugh. ч ч ŋg‘u ‘- ʣaʲ ⁽ ъ ōšъчлḲagh. , Oss. qus ‖л os, Scyth. ⃝ I Sogd. II Khōtщлgguv ‘-, gg III Wakh. i , Ishk. ḷ , Sangl. ḷ, Shugh. члοōsh. ⁽члπaʲХqщл awl, Yazgh. əvon, Munj. ₍, Yidgh. IV Pasht. ⁽‘ǵ, wa V Ōʲmщлg Д, g ₍члἵaʲāch. g ? Khʷāʲщл wx ъ ōʸъчлAvʣщлg‘o ‘- [Pers. g ] The issue of reclassification of the Eastern Iranian languages was only outlined in this thesis, the question still waits for its thorough examination. Valentina Stepanovna Sokolova studied genetic relations of YazghuὕāmХл and the ShughnХ-οōshānХл ʥʲoupл уSOKOLOVA 1967) and later relations of the ShughnХ-YazghuὕāmХлʥʲoupлʷʧtʦлκun₍Х271 (SOKOLOVA 1973). Studies of genetic ʲʣὕatʧonʳлoʤлκun₍ХлandлḲʧdghāлʷʧtʦлBaϑtʲʧanлandлaὕʳoлʧntʣʲʲʣὕatʧonʳлoʤлBaϑtʲʧanлʷʧtʦлtʦʣлἵaṭʦānл languages can answer the question of position of Bactrian within the Eastern Iranian group. In a similar way can be studied relationship of WakhХл andлtʦʣлπa₎aлὕanʥuaʥʣʳл – WakhХлappʣaʲʳлtoл share several isoglosses with the Saka languages, but the language shows probable adstrate or substrate phenomena that link it closer to the lanʥuaʥʣʳлoʤлἵāmХr. Classification of the language of Khʷāʲʣzmл ʲʣmaʧnʳл toл ϐʣл rather complicated – Khʷāʲʣzmʧanл ʳʦaʲʣʳл ʳʣvʣʲaὕл ʧʳoʥὕoʳʳʣʳл ʷʧtʦл Alano-ἴʳʳʣtʧϑлὕanʥuaʥʣʳлandлʷʧtʦлtʦʣлὕanʥuaʥʣʳлoʤлἵāmХʲлonлonʣлʦandчлonлtʦʣлotʦʣʲлʦandлtʦʣʲʣл are some similarities with North-Western Iranʧanл πanʥʣʳāʲХл уϑʤщл Aƒ AκФ – WINDFUHR 1972), there are also some isoglosses shared with Sogdian (cf. SIMS-WILLIAMS 1989a, 170); summary of possible connections of Khʷāʲʣzmʧanл ʷʧtʦл Avʣʳtanл ʦavʣл ϐʣʣnл pʲʣʳʣntʣdл ϐyл αavʧdл λʣʧὕл MACKENZIE (1988) and by Vladimir Aronovich LIVSHITS (1962, 140). I tried to solve the issue of mutual affinity of Sogdian and YaghnōϐХщлπomʣлʳϑʦoὕaʲʳлassumed that YaghnōϐХлʧʳлaлὕanʥuaʥʣлϑontʧnuʧnʥлanлunattʣʳtʣdлnon-literary dialect of Sogdian, YaghnōϐХл was even labelled Neo-Sogdian by some of them (cf. BOGOLYUBOV 1956; KLIMCHITSKIY 1935; 271 ζnлtʦʧʳлϑaʳʣлaὕʳoлpoʳʧtʧonлoʤлζʳʦ₎āʳʦmХлandлτa₎ʦХлʧʳлdʧʳϑuʳʳʣd. ·216· SἰηÆοσØ 1989a, 375-376), some other scholars suppose that YaghnōϐХ is a successor of (in texts unattʣʳtʣdфл πoʥdʧanл dʧaὕʣϑtл oʤл ςʳtʲōshana (KHROMOV 1987, 645, BUZURGMEHR 2005, 117). Contemporary studies tend to see rather greater differences between YaghnōϐХл andл πoʥdʧanл – the main differences quoted in scientific literature is absence of operation of the Sogdian Rhythmic Law in YaghnōϐХ, different development of augment and YaghnōϐХл уaʲϑʦaʧϑфл vʣʲϐaὕл ending of the third person plural - › instead of Sogdian -aṁd (cf. YOSHIDA 2009a, 327), another thorough study on relationship of YaghnōϐХл andл πoʥdʧanл ʷaʳл ʲʣϑʣntὕyл pʲʣʳʣntʣdл ϐyл λʧϑoὕaʳл SIMS-WILLIAMS (2012). For definition of interrelation of YaghnōϐХл andл πoʥdʧanл ʧtл ʧʳл ʧmpoʲtantл toл dʣfine both languages. Sogdian retains many archaic features in morphology and is, in comparison to YaghnōϐХчлmoʲpʦoὕoʥʧϑaὕὕyлʲʧϑʦʣʲщлγoʲлḲaghnōϐХлtʦʣʲʣлʧʳлnoлdʧʲʣϑtлʣvʧdʣnϑʣлoʤлdʣvʣὕopmʣntлoʤл its morphology during its history, but it can be assumed, that *Proto-YaghnōϐХл poʳʳʣʳʳʣdл similar morphological forms as those attested in Sogdian. I have decided to reconstruct a proto-language common for both Sogdian and YaghnōϐХл ʤoʲл tʦʣл puʲpoʳʣʳл oʤл tʦʧʳл tʦʣʳʧʳщл Reconstruction of *Proto-Sogdic seems to be the best way to answer questions concerning interrelations of YaghnōϐХл andл πoʥdʧanщл ρʦʣл maʧnл dʧfference appears not to be seen in morphology, which is much simplified in YaghnōϐХчл nʣʧtʦʣʲл ʧnл pʦonoὕoʥyчл ʷʦʧϑʦл ʦaʳл toл ϐʣл carefully reconstructed for Sogdian, but it is the development of stress that can the source of divergent features in both languages. In the chapter II.1.1. there is outlined development of stress in languages derived from *Proto-Sogdic. I have outlined four stages of stress: Stress I (chapter II.1.1.1.) corresponds with original position of stress in *Proto-Iranian, Stress II (chapter II.1.1.2.) presents stress shift that defines position of stress in *Proto-Sogdic and subsequent shifts labelled as Stress III and Stress IV (chapters II.1.1.3. and II.1.1.4.) represent development of stress as it can be reconstructed for Sogdian. Position of stress in YaghnōϐХл ϑontʧnuʣʳл ʤʲomл tʦʣл poʳʧtʧonл oʤл tʦʣл Stress II (i.e. YaghnōϐХл ʳtʲʣʳʳл pʲʣʳʣʲvʣʳл aʲϑʦaʧϑл poʳʧtʧonл oʤл ʳtʲʣʳʳл aʳл ϑanл ϐʣл ʲʣϑonʳtʲuϑtʣdл ʤoʲл *Proto-Sogdic), such position of stress can be also reconstructed for oldest stages of Sogdian before operation of the Stress III. The Sogdian language272 can be defined as a language that developed after shift of the Stress III and subsequent operation of the Sogdian Rhythmic Law – it is the operation of the Rhythmic Law that defines Sogdian as against other Iranian languages, such as this innovation has not been attested in other Iranian languages. As *Proto-Sogdic stress remained on the same position in YaghnōϐХчлḲaghnōϐХлandлπoʥdʧanлdʣvʣὕopʣdлdʧfferently. The operation of the Rhythmic Law divided Sogdian words into two groups – so-called light and heavy stems, the light stem words retained rich inflectional system, but the heavy stems developed three-case system (i.e. oblique cases phonetically merged into a single form). Development in YaghnōϐХлʷaʳлϑomparable with development of the Sogdian heavy stems. 272 I.e. its literary form attested in various texts from territory of Sogdiana, Chinese Turkestan, or from other regions of Central and Inner Asia. ·217· There are also several phonetic differences in development of Sogdian and YaghnōϐХл– these features can be considered dialectal and probably they originally led to the assumption that YaghnōϐХл mayл ϐʣл aл dialect of Sogdian. According to the analysis of stress shifts in languages derived from *Proto-Sogdic it can be suggested, that phonological development was also influenced by stress, namely in *(Proto-)Sogdian, where original short unstressed vowels changed to Schwa (ə or its allophone ), but remained unchanged in YaghnōϐХлуʤoʲлdʣvʣὕopmʣntл in phonology see chapters II.1.2. and II.1.3.). In morphology the differences between YaghnōϐХл andл πoʥdʧanл aʲʧʳʣчл maʧnὕyл duʣл toл tʦʣл operation of the Rhythmic Law, but there are also other phenomena that have not been influenced by stress. Fundamental is development of augment in Sogdian and YaghnōϐХл – in Sogdian augment has been lost for all non-prefixed verbs, but it has been preserved as so-called internal augment for prefixed verbs (i.e. reflects of augment can be seen after a verbal prefix, in this case prefix usually changes its phonetic form when followed by augment), but in YaghnōϐХл augment remained as a distinctive feature of imperfect and was reanalysed by analogy for all verbs as a prefix even for those containing historical verbal prefixes (see chapters II.2.4., II.1.3.26.ii. and II.1.8.). Other essential morphological features are two archaisms preserved only in YaghnōϐХл – preservation (and reanalysis) of peripheral preterite ending - › < *-ā› < Ide. *-(o)ro / -(o)ror and preservation of imperfect ending of the first person plural - m < *-ām‘ in Western YaghnōϐХлуʧnлβaʳtʣʲnлḲaghnōϐХлandлʧnлπoʥdʧanлtʦʣлʧmpʣʲʤʣϑtлʣndʧnʥлoʤлtʦʣлfirst person plural has been replaced by original optative ending ŋ-‘ m‘ > Yagh -Дm, Sogd. -ēm; see Table 51). The fact that YaghnōϐХл dʧaὕʣϑtʳл dʣvʣὕopʣdл tʷoл dʧfferent imperfect endings of the first person plural may indicate an early split of *Proto-YaghnōϐХл andл хἵʲoto-Sogdian, and subsequent innovation of imperfect endings in (*Proto-)Sogdian and *Proto-Eastern YaghnōϐХщ During the development of the Sogdian language, Sogdian nominal morphology gradually simplified inflectional cases and light stem nouns changed their case endings and analogically switched to agglutinative inflection as is attested for heavy stems – the light stems formed minority of nominal roots and as there was double system of nominal inflection in Sogdian the language tended to avoid such dichotomy. As the light stem inflection switched by analogy towards the heavy stem inflection, there remained system of three cases – direct, oblique and vocative, i.e. case system similar to *Proto-YaghnōϐХ. This reduced inflectional system is attested in late Sogdian Christian document C 5 (cf. SIMS-WILLIAMS 1982). Also verbal endings tended to be unified for both light and heavy stems. πʧmʧὕaʲʧtyлʧnл aʥʥὕutʧnativʣ лʳyʳtʣmлoʤлὕatʣл Sogdian inflectional system with YaghnōϐХлʧʳлʳtʲʧ₎ʧnʥчлϐutлonὕyлʤoʲmaὕὕyлуoʲлʳayлonлʳynϑʦʲonʧϑл level), but diachronically the development in both languages differ. The late Sogdian (or C 5-πoʥdʧan флʳyʳtʣmлoʤлnomʧnaὕлʧnflection cannot be considered as a source for development of YaghnōϐХлʧnflectional system as there are still different patterns of stress development in both languages – diachronically YaghnōϐХл ʳtʧὕὕл pʲʣʳʣʲvʣʳл ʳtʲʣʳʳл onл ʧtʳл poʳʧtʧonл aʳл ʧtл ʷaʳл ʧnл *Proto-Sogdic (i.e. Stress II), but (*Proto-)Sogdian certainly developed later stress shift – Stress III that influenced also morphology of the language (i.e. so-called Rhythmic Law), and ·218· probably later on another stress shift appeared in (late) Sogdian – Stress IV. The shift towards the Stress IV can be probably connected with the above mentioned simplification of nominal inflectional cases as attested in the document C 5 – the tendency to equalize the three-case system of the heavy stems and the six-case system of the light stems led towards a heavy stem-like agglutinative system. There was probable opposite tendency in stress – it tended to shift towards the end of a word, such tendency can be seen in analysis of Sogdian versification by Elio PROVASI (2009, 351-353) whereas the final state of the Stress IV shift can be seen in the Sogdian doϑumʣntʳлʷʲʧttʣnлʧnлtʦʣлBʲāʦmХлʳϑʲʧptлуSIMS-WILLIAMS 1996a, 312-313). Lexicon of both Sogdian and YaghnōϐХл dʧffers. This fact can be caused by two facts – 1) Sogdian is attested in various documents, but majority of texts are religious texts so the voϑaϐuὕaʲyл oʤtʣnл doʣʳл notл dʣʳϑʲʧϐʣл ϐaʳʧϑ л voϑaϐuὕaʲyл ϑonnected with everyday life of peasants and other common people in Sogdiana, but such vocabulary is well attested in YaghnōϐХлaʳлtʦʣл YaghnōϐХʳл aʲʣл ʳʣmʧ-nomadʧϑл paʳtoʲaὕʧʳtʳл andл tʦʣʧʲл ὕanʥuaʥʣл pʲʣʳʣʲvʣʳл manyл ʧndʧʥʣnouʳ л terminology connected with animal husbandry and life in the mountains 273 ; and 2) there is approximately a thousand years long gap between Sogdian and (Modern) YaghnōϐХчлduʲʧnʥлtʦʧʳл pʣʲʧodлtʦʣл ʷoʲὕdлoʤлthe πoʥdʧanʳ лϑʦanʥʣdлϑonʳʧdʣʲaϐὕyлandлtʦʧʳлdʣvʣὕopmʣntлmayлϐʣлoϐʳʣʲvʣdл in development of YaghnōϐХлὕʣʸʧϑonщ After the fall of Sogdiana and gradual disuse of the Sogdian language (Arabic and) Persian became the lingua franca of Central Asia and Persian strongly influenced not only (Pre-Modern) YaghnōϐХчл ϐutл aὕʳoл manyл otʦʣʲл ὕanʥuaʥʣʳл ʳuϑʦл aʳл tʦʣл ἵāmХʲл ὕanʥuaʥʣʳчл Pashtō, Indo-Aryan ςʲdūчлtʦʣлλūʲʧʳtānХлandлthe Dardic languages or Turkic Uzbek, Kyrgyz etc. Modern YaghnōϐХл preserves approximately 27% of indigenous vocabulary, other parts of lexicon are borrowings, calques, or YaghnōϐХ-Persian (YaghnōϐХ-Arabic etc.) compounds. Sogdian lexicon contains also number of borrowings, mainly from Sanskrit, Old TurkʧϑлandлAʲamaʧϑлуϐutлʣʸϑὕudʧnʥл πoʥdʧan л words written with Aramaic ideograms). YaghnōϐХлʳʦoʷʳлʳomʣлὕʣʸʧϑaὕлʳʧmʧὕaʲʧtʧʣʳлʷʧtʦлtʦʣлἵāmХʲлὕanʥuaʥʣʳчлʣщʥщ ayk dauʥʦtʣʲчлʥʧʲὕ , Saponaria Griffithiana Boiss. pὕant , x ϑʲoʷчлmaʥpʧʣ лand many others (see end of the chapter III) – these words can be connected either with local ecology and comparable seminomadʧϑл ὕʧʤʣʳtyὕʣл oʲл ʷʧtʦл tʦʣл ἵāmХʲ-εʧndū₎ush Sprachbund mentioned in chapter I.1.1.4.b. Unfortunately there are no attested counterparts in Sogdian. From the above mentioned points it thus can be suggested, that Sogdian and YaghnōϐХлaʲʣл closely related languages, but there is no evidence that shows that YaghnōϐХлdʣvʣὕopʣdлdʧʲʣϑtὕyл from Sogdian. If we assume that YaghnōϐХлdʣvʣὕopʣdлʤʲomлaлπoʥdʧanлdʧaὕʣϑtлʷʣлʦavʣлtoлdʣfine such dialect – I tried to sum up our knowledge of possible Sogdian dialects in the excursion 1, Aʳл ḲaʥʦnōϐХл ʧʳл anл unʷʲʧttʣnл ὕanʥuaʥʣл tʦʣʲʣл ʧʳл noл ʣὕaϐoʲatʣл tʣʲmʧnoὕoʥyл ϑonnʣϑtʣdл ʷʧtʦл ʳayл poὕʧtʧϑaὕл andл religʧouʳлὕʧʤʣлʤoʲлtʦʣʳʣлʤʧʣὕdʳлaʲʣлdomaʧnʳлoʤлρā₍Х₎лἵʣʲʳʧanлуϐutлaὕʳoлʧnлἵʣʲʳʧanлmanyлʷoʲdʳлϑonnʣϑtʣdлʷʧtʦлʲʣὕʧʥʧouʳл life are taken from Arabic). 273 ·219· but the evidence of the dialects is quite deficient. It is certain that both Sogdian and YaghnōϐХл developed from the same proto-language, but this proto-language equally differs from both languages in focus – I labelled the proto-language as *Proto-Sogdic which I find appropriate for explanation of development of both Sogdian and YaghnōϐХлʲatʦʣʲлtʦanлхἵʲoto-Sogdian as there has to be suggested a an intermediate development stage between *Proto-Sogdic and (literary) Sogdian. As can be seen in the part II of the presented thesis, YaghnōϐХлappʣaʲʳлʧnлʳomʣлaʳpʣϑtʳлmoʲʣл archaic in comparison to Sogdian – YaghnōϐХл pʲʣʳʣʲvʣʳл aʲϑʦaʧϑл poʳʧtʧonл oʤл ʳtʲʣʳʳчл ʧtл pʲʣʳʣʲvʣʳл augment (though the augment has been innovated in YaghnōϐХфчл ʧtл ϐʣttʣʲл preserves Iranian vowels (i.e. there is no reduction of unstressed vowels to Schwa as there was no Stress III shift) and YaghnōϐХлdʧaὕʣϑtʳлʳʦoʷлtʦatлoʲʧʥʧnʳлoʤлϐotʦлdʧaὕʣϑtʳлϑanлϐʣлoʤлanлoὕdлdatʣщ Archaic is also formation of ergative construction in YaghnōϐХл andл anotʦʣʲл aʲϑʦaʧʳm shared with Avestan, Khōtanʣʳʣл andл Khʷāʲʣzmʧanл ʧʳл pʲʣʳʣʲvatʧonл oʤл aʲϑʦaʧϑл pʲʣtʣʲʧtʣл ʣndʧnʥл oʤл tʦʣл tʦʧʲdл pʣʲʳonл plural *-ā›. On contrary, Sogdian shows archaic features mainly in morphology – the operation of the Sogdian Rhythmic Law preserved archaic inflectional system for light stem words, and also verbal morphology – Sogdian preserves more inherited verbal forms then does YaghnōϐХщ Both languages share some innovations – main similarity is development of nominal inflection in YaghnōϐХл andл ʧnл ϑaʳʣл oʤл tʦʣл heavy stems in Sogdian – development of direct and oblique cases is comparable, moreover, YaghnōϐХлὕoʳtлvoϑative case. Another shared innovation (typical also for other North Eastern Iranian languages) is formation of plural with the abstract suffix *-t(u) -. Sogdian innovated ergative construction as it replaced copula by the verb ŋdā›- to ʦoὕd л ʤoʲл tʲanʳʧtive verbs (cf. similar development in Khʷāʲʣzmʧanфчл anotʦʣʲл innovations can be seen in new suffixed forms of verbal inflection. The most important innovation in Sogdian was the shift towards the Stress III and subsequent operation of the Rhythmic Law – in this case originally phonetic change strongly influenced morphology and phonology of the language (the later shift towards the Stress IV was probably connected with a tendency to simplify inflectional dichotomy between the light and heavy stems). YaghnōϐХл innovations show spread of prefixed augment by analogy to all verbal forms regardless of their original prefixes and also reanalysis of verbal endings – original durative ending - t serves to form simple present and future tenses or as durative marker for the imperfect. Original indicative endings remained in YaghnōϐХчл ϐutл tʦʣyл ϑʦanʥʣdл tʦʣʧʲл ʤunϑtʧonл – they are used as forms of so-called dependent paradigm, i.e. they are used in a clause where appear more than one verb – for indicative present only the first verb is inflected in the present(/future) tense (i.e. historical present + - t), all other verbs appear in forms of the dependent paradigm (i.e. in forms of historical present). YaghnōϐХлʦaʳлὕoʳtлʤoʲmatʧonлoʤлϑauʳatives from Iranian ŋ-‘ ‘-ʳtʣmʳ´чл there are preserved only several verbs in YaghnōϐХлtʦatлoʲʧʥʧnatʣлʤʲomлʳuϑʦлϑauʳatives, nowadays Tajik causative suffix - n- is used. Tajik has influenced YaghnōϐХл vʣʲϐaὕл moʲpʦoὕoʥyл aὕʳoл ʧnл many other aspects, this issue can be considered as contact phenomenon rather as innovation (cf. NἴσÁἰ [in print]). ·220· *** Both YaghnōϐХл andл πoʥdʧanл ʳʦoʷл manyл dʧfferences, some of them are caused by approximately thousand years of discontinuity of development of both language as Sogdian has been replaced by Persian in the 10th and 11th both languages were gradually influenced by Persian, strong influence of Persian is visible mainly in YaghnōϐХщлAʳлϐotʦлὕanʥuaʥʣʳлdʧffer according to their attested forms, it can be said that from diachronic point of view they are two similar dialects/languages, both comparable in historical development as Sogdic dialects within the North Eastern Iranian language group. ·221· V. Bibliography ABAEV 1949: Василий Иванович Абаев: Осетинский язык и фольклор. Москва – Ленинград, 1949. ABAEV 1958: Василий Иванович Абаев: Историко-этимологический словарь осетинского языка. Том 1: A-Kʼ. Ленинград (: Наука), 1958. ABAEV 1965: Василий Иванович Абаев: Скифо-европейские Запада. Москва (: Наука), 1965. изоглоссы. На стыке Востока и ABAEV 1979: Василий Иванович Абаев: Скифо-сарматские наречия. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Древнеиранские языки. Москва (: Наука) 1979, p. 272-364. AKISHEV 1978: Кемаль А. Аркишев: Курган Иссык. Искусство саков Казахстана. Москва (: Искусство), 1978. ALEMANY I VILAMAJÓ 1999: Augustí Alemany i Vilamajó: Els «Cants arimaspeus» d’Arísteas de Proconnès i la caiguda dels Zhou occidentals. Faventia 21/2, 1999, p. 45-55. ANDREEV 1945: Михаил Степанович Андреев: О таджикском языке настоящего времени. In: Материалы по истории таджиков и Таджикистана. Сб. 1-й. Сталинабад, 1954, p. 6680. ANDREEV – LIVSHITS – PISARCHIK 1957: Михаил Степанович Андреев – Владимир Аронович Лившиц – Антония Константиновна Писарчик: Словарь. In: Михаил Степанович Андреев – Елена Михайловна Пещерева: Ягнобские тексты с приложением ягнобско-русского словаря составленного М. С. Андреевым, В. А. Лившицем и А. К. Писарчик. Москва – Ленинград (: Издат ельст во Академии Наук СССР), 1957, 215-391. ANDREEV – PESHCHEREVA 1957: Михаил Степанович Андреев – Елена Михайловна Пещерева: Ягнобские тексты с приложением ягнобско-русского словаря составленного М. С. Андреевым, В. А. Лившицем и А. К. Писарчик. Москва – Ленинград (: Издательство Академии Наук СССР), 1957. AɈẒAMĪ – WINDFUHR 1972: Cheragh Ali Azami – Gernot Windfuhr: A Dictionary of Sangesari, With a Grammatical Outline. Tehrān (: Franklin Book Company), 1972. ،(‫ ﻣﻮﺳﺴ ـ ـ ــﻪ اﻧﺘﺸ ـ ـ ــﺎرات ﻓ ـ ـ ـ ـﺮاﻧﮑﻠﲔ‬:) ‫ ـ ـ ـ ـﺮان‬ .‫ ﺑ ـ ـ ــﺎ ﻣﻘ ـ ـ ــﺪﻣﻪای از دﺳ ـ ـ ــﺘﻮر آن زﺑ ـ ـ ــﺎن‬،‫ واژەﻧﺎﻣ ـ ـ ـ ـ ٔﻪ ﺳﻨﮕﺴ ـ ـ ــﺮی‬:‫ وﻳﻨ ـ ـ ــﺪﻓﻮﻫﺮ‬.‫ﭼﺮاﻏﻌﻠ ـ ـ ــﯽ اﻋﻈﻤ ـ ـ ــﯽ – ﮔﺮﻧ ـ ـ ــﺖ ل‬ .۱۳۵۱ BACKSTROM 1992: Peter C. Backstrom: Wakhi. In: Peter C. Backstrom – Carla F. Radloff: Sociolinguistic ·222· Survey of Northern Pakistan, Volume 2, Languages of Northern Areas. Islamabad (: National Institute of Pakistan Studies, Quaid-i-Azam University), 1992, s. 55-74 (+ Appendix D – Wakhi Survey Data, s. 273-292). BAKHTĪBĒKOV 1979: Тӯпчӣ Бахтибеков: Грамматикаи забони шуғнонӣ. Душанбе (: Дониш), 1979. BARTHOLOMAE 1895-1901: Christian Bartholomae: Vorgeschichte der iranischen Sprachen. In: Wilhelm Geiger – Ernst Kuhn (eds.): Grundriss der iranischen Philologie, Erster Band, 1. Abteilung. Straßburg (: Verlag von Karl J. Trübner), 1898-1901, p. 1-151. BARTHOLOMAE 1961: Christian Bartholomae: Altiranisches Wörterbuch. Berlin (: Walter de Gruyter & Co.), 1961. BARTONĚK 2009: Antonín Bartoněk: Dialekty klasické řečtiny. Brno (: Masarykova universita), 2009. BEEKES 2011: Robert Stephen Paul Beekes: Comparative Indo-European Linguistics. An introduction. [Second edition]. Amsterdam – Philadelphia (: John Benjamins Publishing Company), 2011. BELYAEV 2010: Oleg Belyayev: Evolution of Case in Ossetic. In: Iran and the Caucasus 14, 2010, p. 287322. BIČOVSKÝ 2012: Jan Bičovský: Stručná mluvnice praindoevropštiny. Praha (: Filozofická fakulta Univerzity Karlovy v Praze), 2012. BIELMEIER 1989: Roland Bielmeier: Yaghnōbī. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 480-488. BIELMEIER 2006 [online]: Roland Bielmeier: Yaghnobi. In: Eḥsān Yārshāṭer (ed.): Encyclopædia Iranica [online edition]. Costa Mesa, California. URL: <http://www.iranica.com/articles/yaghnobi> [quot. 23. 07. 2010, 18:14] BIRILLO – BULAKHOV – SUDNIK 1966: Н. В. Бирилло – М. Г. Булахов – М. Р. Судник: Белорусский язык. In: В. В. Виноградов (ed.): Языки народов СССР. Том первый: Индоевропейские языки. Москва (: Наука), 1966, p. 154-193. BOGOLYUBOV 1956: Михаил Николаевич Боголюбов: Ягнобский (новосогдийский) язык. Исследование и материалы. Автореферат на соискание учёной степени доктора филологических наук. Ленинград 1956. BOGOLYUBOV 1966: Михаил Николаевич Боголюбов: Ягнобский язык. In: В. В. Виноградов (ed.): Языки ·223· народов СССР. Том первый: Индоевропейские языки. Москва (: Наука), 1966, p. 342-361. BOGOLYUBOV – SMIRNOVA 1963: Михаил Николаевич Боголюбов – Ольга Ивановна Смирнова: Хозяйственные документы. Согдийские документы с горы Муг. Чтение. Перевод. Комментарий. Выпуск III. Москва (: Издат ельст во вост очной лит ерат уры), 1963. BOYCE 1952: Mary Boyce: Some Parthian Abecedarian Hymns. In: Bulletin of the School of Oriental and African Studies 14/3, Studies Presented to Vladimir Minorsky by His Colleagues and Friends (: University of London), 1952, s. 435-450. BROWNING 1983: Robert Browning: Medieval and Modern Greek. Cambridge (: Cambridge University Press), 1983. BURKI 2001: Rozi Khan Burki: Dying Languages with Special Focus on Ormuri. In: Pakistan Journal of Public Administration; December 2001; Volume 6. No. 2. URL: <http://www.fli-online.org/documents/languages/ormuri/dying-languages.pdf> [quot. 23. 03. 2012, 20:34] BUSHKOV – NOVIKOV 1992: В. И. Бушков – С. В. Новиков: Об интерпретации некоторых документов с горы Муг и местной топонимике. In: Вестник МГУ. Серия VIII, История, 1992 № 3, p. 1425. BUZURGMEHR 2005: Бурҳониддин Бузургмеҳр: Яғнобиёни муқими Душанбешаҳр ва музофоти он. In: Ю. Шодипур – А. Абдуллоев: Душанбе дар масири таърих (Маҷмӯаи мақолаҳо). Душанбе (: Ст удент ), 2005, p. 117-128. CARDONA 1970: George Cardona: The Indo-Iranian construction mana (mama) kṛtam. In: Language, Vol. 46/1, 1970, p. 1-12. DECKER 1992: Kendall D. Decker: Yidgha. In: Kendall D. Decker: Sociolinguistic Survey of Northern Pakistan, Volume 5, Languages of Chitral. Islamabad (: National Institute of Pakistan Studies, Quaid-i-Azam University), 1992, p. 43-66 (+ Appendix B – Chitral Word lists, p. 177-211; Appendix C.2 – Yidgha texts, p. 216-217). DELANCEY 1981: Scott DeLancey: An Interpretation of Split Ergativity and Related Patterns. In: Language, Vol. 57/3, 1981, p. 626-657. DOROFEEVA 1960: Лидия Николаевна Дорофеева: Язык фарси-кабули. Москва (: Издательство вост очной лит ерат уры), 1960. ÈDEL’MAN 1966: Джой Иосифович Эдельман: Язгулямский язык. Москва (: Наука), 1966. ·224· ÈDEL’MAN 1986: Джой Иосифович Эдельман: Сравнительная грамматика восточноиранских языков. Фонология. Москва (: Наука), 1986. ÈDEL’MAN 1987a: Джой Иосифович Эдельман: Шугнано-рушанская язычная группа. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки II. – Восточная группа. Москва (: Наука), 1987, p. 236-347. ÈDEL’MAN 1987b: Джой Иосифович Эдельман: Язгулямский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки II. – Восточная группа. Москва (: Наука), 1987, p. 348-407. ÈDEL’MAN 2000a: Джой Иосифович Эдельман: Хорезмийский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 95-105. ÈDEL’MAN 2000b: Джой Иосифович Эдельман: Язгулямский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 274-290. ÈDEL’MAN 2008: Джой Иосифович Эдельман: Хорезмийский язык. In: Основы иранского языкознания. Среднеиранские и новоиранские языки. Москва (: Восточная лит ерат ура), 2008, p. 6-60. ÈDEL’MAN – DODYKHUDOEVA 2009: Joy I. Edelman – Leila R. Dodykhudoeva: The Pamir Languages. In: Gernot Windfuhr (ed.): Iranian Languages. London – New York (: Routledge), 2009, p. 773-786. ÈDEL’MAN – YŪSUFBĒKOV 2000a: Джой Иосифович Эдельман – Шодихон П. Юсуфбеков: Шугнанский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 225-242. ÈDEL’MAN – YŪSUFBĒKOV 2000b: Джой Иосифович Эдельман – Шодихон П. Юсуфбеков: Рушанский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 242-254. ÈDEL’MAN – YŪSUFBĒKOV 2000c: Джой Иосифович Эдельман – Шодихон П. Юсуфбеков: Хуфский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 254-259. ÈDEL’MAN – YŪSUFBĒKOV 2000d: Джой Иосифович Эдельман – Шодихон П. Юсуфбеков: Бартангский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 259-264. ÈDEL’MAN – YŪSUFBĒKOV 2000e: Джой Иосифович Эдельман – Шодихон П. Юсуфбеков: Рошорвский язык. In: ·225· Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 264-268. ÈDEL’MAN – YŪSUFBĒKOV 2000f: Джой Иосифович Эдельман – Шодихон П. Юсуфбеков: Сарыкольский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 269-274. EFIMOV 1999a: Валентин Александрович Ефимов: Парачи язык. In: Языки мира. Иранские языки II. – Северо-западные иранские языки. Москва (: ИНДРИК), 1999, p. 257-275. EFIMOV 1999b: Валентин Александрович Ефимов: Ормури язык. In: Языки мира. Иранские языки II. – Юго-западные иранские языки. Москва (: ИНДРИК), 1999, p. 276-296. EFIMOV 2008: Валентин Александрович Ефимов: Хазара. In: Основы иранского языкознания. Среднеиранские и новоиранские языки. Москва (: Восточная литература), 2008, p. 344-414. EFIMOV – RASTORGUEVA – SHAROVA 1982: Валентин Александрович Ефимов – Вера Сергеевна Расторгуева – Е. Н. Шарова: Персидский, таджикский, дариɴ. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки I. – Западная группа, прикаспийские языки. Москва (: Наука), 1982, p. 5-230. ELFENBEIN 1984a: Joseph H. Elfenbein: The Wanetsi connexion. Part I. In: Journal of the Royal Asiatic Society 116/1, 1984, p. 54-76. ELFENBEIN 1984b: Joseph H. Elfenbein: The Wanetsi connexion. Part II. In: Journal of the Royal Asiatic Society 116/2, 1984, p. 229-241. EMMERICK 1989: Ronald Eric Emmerick: Khotanese and Tumshuqese. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag) 1989, p. 204-229. EMMERICK 2009: Ronald Eric Emmerick: Khotanese and Tumshuqese. In: Gernot Windfuhr (ed.): Iranian Languages. London – New York (: Routledge), 2009, p. 377-415. ETHNOLOGUE: Raymond G. Gordon (ed.) Ethnolo«ue. Languages of the World. Fifteenth Edition. Dallas (: SIL International), 2005. FAYZ̤OV 1966: М. Файзов: Язык рушанцев советского Памира. Душанбе (: Таджикский ГосУниверситет им. В. И. Ленина), 1966. ·226· FRYE 1972: R. N. Frye: Historical remarks on the two dialects of the Avesta. In: Dr. J. M. Unvala Memorial Volume. Bombay, 1964, str. 30–34. FUSSMAN 1974: Gérard Fussman: Documents épigraphiques kouchans, In: Bulletin de l’Ecole Française d’Extrême Orient 61, 1974, p. 1-66. GARRETT 1990: Andrew Garrett: Hittite Enclitic Subjects and Transitive Verbs. Journal of Cuineform Studies, Vol. 42/2, 1990, p. 227-242. GAUTHIOT 1911: Robert Gauthiot: De lʼalphabet sogdien. Journal Asiatique 17, 1911, p. 81-95. GAUTHIOT – BENVENISTE 1914-1923: Robert Gauthiot – Émile Benveniste: Essai de grammaire sogdienne. Première partie: Phonétique. Mission Pelliot en Asie centrale: Série petit in-Octavo, 1. Paris, 1914-1923. GAUTHIOT – BENVENISTE 1929: Robert Gauthiot – Émile Benveniste: Essai de grammaire sogdienne. Deuxième partie: Morphologie, syntaxe et glossaire. Mission Pelliot en Asie centrale: Série petit in-Octavo, 3. Paris, 1929. GAWARJON 1996: 高尔锵: 塔吉克汉词典 (Tujik ziv – Hanzu ziv lughot). Sichuan (: Sichuan Nationalities Publishing House), 1996. GEIGER 1898-1901: Wilhelm Geiger: Über das Yaghnōbī. In: Wilhelm Geiger – Ernst Kuhn (eds.): Grundriss der iranischen Philologie, Erster Band, 2. Abteilung. Straßburg (: Verlag von Karl J. Trübner), 1898-1901, p. 334-344. GERSHEVITCH 1954: Ilya Gershevitch: A Grammar of Manichaean Sogdian. Oxford, 1954. GERSHEVITCH 1976: Ilya Gershevitch: The Sogdian Fragments of the British Library: Appendix. IndoIranian Journal 18, 1976, p. 75-82. GERTSENBERG 1981: Леонард Георгиевич Герценберг: Хотаносакский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Среднеиранские языки I. Москва (: Наука), 1981, p. 233-313. GERTSENBERG 2000: Леонард Георгиевич Герценберг: Хотаносакский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 46-57. GMS = GERSHEVITCH 1954 GRIERSON 1920: George Abraham Grierson: Ishkashmi, Zebaki and Yazghulami. An Account of Three Eranian Dialects. London (: Royal Asiatic Society), 1920. ·227· GRYUNBERG 1972: Александр Леонович Грюнберг: Языки Восточного Гиндукуша: Мунджанский язык. Тексты, словарь, грамматический очерк. Ленинград, 1972. GRYUNBERG 1987: Александр Леонович Грюнберг: Мунджанский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки II. – Восточная группа. Москва (: Наука), 1987, p. 155-235. GRYUNBERG 2000: Александр Леонович Грюнберг: Мунджанский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 154-170. GRYUNBERG – DAVYDOVA 1982: Александр Леонович Грюнберг – Л. Х. Давыдова: Татский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки I. – Западная группа, прикаспийские языки. Москва (: Наука), 1982, p. 231-286. GRYUNBERG – ÈDEL’MAN 1987: Александр Леонович Грюнберг – Джой Иосифович Эдельман: Афганский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки II. – Восточная группа. Москва (: Наука), 1987, p. 6-154. HABERLAND 1994: Hartmut Haberland: Danish. In: Ekkehard König – Johann van der Auwera (eds.): The Gremanic Languages. London (: Routledge), 1994, 313-348. HALLBERG 1992: Daniel G. Hallberg: Sociolinguistic Survey of Northern Pakistan, Volume 4, Pashto, Waneci, Ormuri. Islamabad (: National Institute of Pakistan Studies, Quaid-i-Azam University), 1992. HARMATTA 1970: Harmatta János (Harmatta János): Studies in the history and language of the Sarmatians. Acta universitatis de Attila József nominatae – Acta Antiqua et Archaeologica, Tomus XIII. Szeged 1970. HARMATTA 1989: János Harmatta (Harmatta János): The Language of the Southern Sakas. Acta Antiqua Academiae Scientiarum Hungaricae 32. Budapest (: Akadémiai Kiadó), 1989, p. 299-307. HARMATTA 2002a: János Harmatta (Harmatta János): Herodotus. Die Schrift bei den antiken Steppenvölkern. In: László Havas – Imre Tegyey (eds.): János Harmatta. Selected writings. West and East in the unity of the ancient world. ΑΓΑΘΑ XII. Debreceni egyetem bölcsészettudományi kar. Klasszika-filológiai Tanszék. Debrecen (: Kossuth egyetemi kiadó, Debreceni egyetem), 2002, p. 40-50. (Acta Classica Universitatis Scientiarum Debreceniensis 28, 1992, 7-16) HARMATTA 2002b: János Harmatta: Herodotus, historian of the Cimmerians and the Scythians. In: László Havas – Imre Tegyey (eds.): János Harmatta. Selected writings. West and East in the ·228· unity of the ancient world. ΑΓΑΘΑ XII. Debreceni egyetem bölcsészettudományi kar. Klasszika-filológiai Tanszék. Debrecen (: Kossuth egyetemi kiadó, Debreceni egyetem), 2002, p. 207-216. (Entretiens sur l’antiquité antique classique. Tome XXXV. VandœuvresGenève, 1990, 115-130) HENNING 1939: Walter Bruno Henning: Sogdian Loan-Words in New Persian. In: Bulletin of the School of Oriental and African Studies 10/1 (: University of London), 1939, p. 93-106. HENNING 1958: Walter Bruno Henning: Mitteliranisch. In: Karl Hoffmann – Walter Bruno Henning – Harold Walter Bailey – Georg Morgenstierne – Wolfgang Lentz (eds.): Iranistik, Erster Abschnitt – Linguistik. Leiden – Köln (: Brill), 1958. HINGE 2006: George Hinge: Herodot zur skythischen Sprache. In: Glotta 81 2005[2006], p. 86-115. HORN 1988: Paul Horn: Grundriß der neupersischen Etymologie. Sammlung indogermanischer Wörterbücher. Hildesheim – Zürich – New York (: Georg Olms Verlag), 1988. HUMBACH 1989: Helmut Humbach: Choresmian. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 193-203. IDŌ 2009: Shinji Ido: An analysis of the formation of the Tajik vowel system. In: Anju Saxena – rd Åke Viberg (eds.): Multilingualism. Proceedings of the 23 Scandinavian conference of linguistics. Acta universitatis Upsaliensis: Studia Linguistica Upsaliensia 8. Uppsala, 2009, p. 65-74. IOANNESYAN 1999: Юлий Аркадьевич Иоаннесян: Гератский диалект языка дари современного Афганистана. Москва (: Вост очная лит ерат ура), 1999. ISAEV 1966: Магомет Измайлович Исаев: Дигорский диалект осетинского языка. Фонетика, Морфологиа. Москва (: Наука), 1966, p. 237-256. ISAEV 1987: Магомет Измайлович Исаев: Осетинский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки II. – Восточная группа. Москва (: Наука), 1987, p. 537-643. ISKHAKOV 1977: М. М. Исхаков: Глагол в согдийском языке: документы с горы Муг. Ташкент (: Фан), 1977. JUNKER 1930: Heinrich Franz Josef Junker: Arische Forschungen. Yaghnōbī-Studien I. Die Sprachgeographische Gliederung des Yaghnōb-Tales. Abhandlungen der Philologischehistorische Klasse der Sächsischen Akademie der Wissenschaften, Bd. XLI, Nr. II. Leipzig (: Hirzel), 1930. ·229· JUSTI 1895: Ferdinand Justi: Iranisches Namenbuch. Marburg (: Elwert), 1895. KELLENS 1987: Jean Kellens: Avesta the holy book of the Zoroastrians. In: Eḥsān Yārshāṭer (ed.): Encyclopædia Iranica [online edition]. Costa Mesa, California. [quot. 19. 02. URL: <http://www.iranicaonline.org/articles/avesta-holy-book> 2013, 20:22] KERIMOVA 1963: Аза Алимовна Керимова: Особенности говора кишлака Рарза. In: Иранский сборник. К семидесятилетию профессора И. И. Зарубина. Москва (: Издательство вост очной лит ерат уры), 1963, p. 24-43. KERIMOVA 1982: Аза Алимовна Керимова: Диалекты Фарса. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки I. – Западная группа, прикаспийские языки. Москва (: Наука), 1982, p. 316-363. KHROMOV 1388: ‫ ـ ـ ــﺎی ﺳ ـ ـ ــﻐﺪی و‬‫ وﻳﮋﻩﻧﺎﻣ ـ ـ ـ ـ ٔﻪ زﺑﺎ‬.۱۳۸۸ ‫ ﭘ ـ ـ ــﺎﻳﻴﺰ‬،۲۴ ‫ ﴰ ـ ـ ــﺎرٔﻩ‬،‫ ﺳ ـ ـ ــﺎل دﻫ ـ ـ ــﻢ‬،‫ رودﮐ ـ ـ ــﯽ‬.‫ واژەﻫ ـ ـ ــﺎی ﺳ ـ ـ ــﻐﺪی در ﮔﻮﻳﺸ ـ ـ ــﻬﺎی ﺗ ـ ـ ــﺎﺟﻴﮑﯽ‬:‫آﻟ ـ ـ ــﱪت ﺧﺮوﻣ ـ ـ ــﻮف‬ .۹-۱۶ .‫ﻳﻐﻨﺎﺑﯽ‬ KHROMOV 1958: Альберт Леонидович Хромов: Особенности вокализма матчинских говоров. Хусусиятҳои вокализми шеваҳои Мастчоҳ. In: Известия Академии Наук Таджикской ССР, Ахбороти Академияи фанҳои РСС Тоҷикистон. Отделение Общественных Наук, 1958, № 1 (16). Душанбе (: Дониш), 1958, s. 7-20. KHROMOV 1962: Альберт Леонидович Хромов: Говоры таджиков Матчинского района. Гӯишҳои тоҷики райони Мастчоҳ. Труды, т. CVII. Душанбе (: Издательство Академии наук т аджикской ССР), 1962. KHROMOV 1966: Альберт Леонидович Хромов: Общая лингвистическая характеристика топонимии и микротопонимии Ягноба. In: Известия Академии Наук Таджикской ССР, Ахбороти Академияи фанҳои РСС Тоҷикистон. Отделение Общественных Наук, 1966, № 3 (45). Душанбе (: Дониш), 1966, p. 83-87. KHROMOV 1967: A. L. Chromov: Zur Gesamtcharakteristik der Tadschik-Mundarten von Falghar. In: Mitteilungen des Instituts für Orientforschung, Bd. XIII, N. 3, 1967, p. 462-465. KHROMOV 1969: Альберт Леонидович Хромов: Историко-лингвистическое исследование Ягноба и Верхнего Зеравшана. Дисcертация на соискание учёной степени кандидата филологических наук. Душанбе 1969. KHROMOV 1972: Альберт Леонидович Хромов: Ягнобский язык. Москва (: Наука), 1972. ·230· KHROMOV 1987: Альберт Леонидович Хромов: Ягнобский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки II. – Восточная группа. Москва (: Наука), 1987, p. 644-701. KIEFFER 1989: Charles M. Kieffer: Le parāčī, lʼōrmuṛī et le grouppe des langues iraniennes du Sud-Est. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 445-455. KIEFFER 2009: Charles M. Kieffer: Parachi. In: Gernot Windfuhr (ed.): Iranian Languages. London – New York (: Routledge), 2009, p. 693-720. KIM 2003: Ronald I. Kim: On the Historical Phonology of Ossetic: The Origin of the Oblique Case Suffix. In: Journal of the American Oriental Society 123/1, 2003, p. 43-72. KIM 2007: Ronald I. Kim: Two problems of Ossetic nominal morphology. In: Indogermanische Forschungen 112. Band, 2007, p. 47-68. KISELEVA 1985: Лидия Николаевна Киселева: Язык дари Афганистана. Москва (: Наука), 1985. KLIMCHITSKIY 1935: С. И. Климчицкий: Ягнобско-согдийские соответствия. In: Записки института востоковедения академии наук · VI. Ленинград, 1935. 15-25. KLIMCHITSKIY 1940: С. И. Климчицкий: Секретный язык у ягнобцев и язгулёмцев. In: Академия наук СССР – Труды Таджикистанской базы, т. IX – 1938 – История – язык – литература. Akademijaji Fanho SSSR: Asarhoji ʙazaji Toçikiston, çildi IX – Tarix – zaʙon – adaʙijot. Москва – Ленинград (: Издательство Академии наук СССР), 1940. 104117. KORN 2011: Agnes Korn: Pronouns as Verbs, Verbs as Pronouns: Demonstratives and the Copula in Iranian. In: Agnes Korn – Geoffrey Haig – Simin Karimi – Poller Samvelian (eds.): Topics in Iranian Linguistics. Beiträge zur Iranistik 34. Wiesbaden (: Dr. Ludwig Reichert Verlag), 2011, p. 53-70. KOZYREVA 1974: Тамара Заурбековна Козырева (Кодзырты Т. З.): Язык первой осетинской печатной книги. Фыццаг ирон мыхуыргонд чиныджы вӕ заг. Орджоникидзе (: Ир), 1974. KÜMMEL 2006: Martin Joachim Kümmel: Mitteliranisch II: Sogdisch. Sommersemester 2006. URL: <http://www.indogermanistik.unifreiburg.de/seminar/pers/kuemmel/umat/sogd.pdf> [quot. 07. 03. 2012, 11:40] KÜMMEL 2008: Martin Joachim Kümmel: Mitteliranisch I: Khotansakisch. 2008. ·231· URL: <http://www.indogermanistik.unifreiburg.de/seminar/pers/kuemmel/umat/khotan.pdf> [quot. 07. 03. 2012, 11:45] KÜMMEL 2010: Martin Joachim Kümmel: Mittelkymrisch. Sommersemester 2010. URL: <http://www.indogermanistik.unifreiburg.de/seminar/pers/kuemmel/umat/mittelkymrisch> [quot. 20. 08. 2012, 08:32] LASHKARBĒKOV 2008: Б. Б. Лашкарбеков: Старованджский язык (vanǰivor). In: Основы иранского языкознания. Среднеиранские и новоиранские языки. Москва (: Восточная лит ерат ура), 2008, p. 61-109. LENTZ 1933: Wolfgang Lentz: War Marco Polo auf dem Pamir? In: Zeitschrift der Deutschen Morgenländischen Gesellschaft 85, 1933, p. 1-32. LIVSHITS 1962: Владимир Аронович Лившиц: Хорезмийский язык. In: С. П. Толстов – Т. А. Жданко – С. М. Абрамзон – Н. А. Кисляков (eds.): Народы Средней Азии и Казахстана I. Москва (: Издат ельст во Академии наук СССР), 1962, 138-140. LIVSHITS 2000: Владимир Аронович Лившиц: Бактрийский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 38-46. LIVSHITS 2003: Владимир Аронович Лившиц: Согдийские документы из замка Чильхуджра. In: Scripta Gregoriana. Сборник в честь семидесятилетия академика Г. М. Бонгард-Левина. Москва, 2003, 77-88. LIVSHITS 2008: Владимир Аронович Лившиц: Согдийская эпиграфика Средней Азии и Семиречья. Филологический факультет Исследования. Санкт-Петербург (: Санкт -Пет ербургского государственного университета), 2008. LIVSHITS – KAUFMAN – D’YAKONOV 1954: Владимир Аронович Лившиц – К. В. Кауфман – Игорь Михайлович Дьяконов: О древней согдийской письменности Бухары. Вестник древней истории 1954 № 1 (47), 1954, s. 150-163. LIVSHITS – KHROMOV 1981: Владимир Аронович Лившиц – Альберт Леонидович Хромов: Согдийский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Среднеиранские языки I. Москва (: Наука), 1981, p. 347-514. LIVSHITS – LUKONIN 1964: Владимир Аронович Лившиц – В. Г. Луконин: Среднеперсидские и согдийские надписи на серебряных сосудах. Вестник древней истории 1964 № 3 (89), 1964, s. 155176. ·232· LOY 2005: Thomas Loy: Jaghnob 1970. Erinnerungen an eine Zwangsumsiedlung in der Tadschikischen SSR. Wiesbaden (: Reichert Verlag), 2005. LURʼE 2004: Павел Борисович Лурье: Историко-лингвистический анализ согдийской топонимии. Диссертация на соискание учёной степени кандидата филологических наук. Санкт-Петербург, 2004. LURʼE 2011: Павел Борисович Лурье: Согдийские документы, открытые в Хисораке и Пенджикенте в 2011 г. Предварительное сообщение. In: Павел Борисович Лурье (ed.): Материалы пенджикентской археологоческой экспедиции. Вхпуск XIV. Санкт-Петербург, 2011. LURʼE 2012: Павел Борисович Лурье: Согдийские документы из раскопок раннесредневекового Мартшката. Предварительное сообщение. In: Н. Н. Казанский (ed.): Индоевропейское языкознание и классическая филология – XVI. Материалы чтений, посвященных памяти профессора Иосифа Моисеевича Тронского 18–20 июня 2012 г. Санкт-Петерсбург (: Наука), 2012. MACKENZIE 1988: David Neil MacKenzie: Khwarezmian and Avestan. In: East and West, Vol. 38, No. 1/4 (: Instituto Italiano per lʼAfrica e lʼOriente), 1988, p. 81-92. MALLITSKIY 1924: Николай Гурьевич Маллицкий: Ягнобцы. In: Известия Туркестанского отдела Географического общества, Том XVII. Ташкент, 1924, p. 174-178. MALLORY – ADAMS 2006: J. P. Mallory – D. Q. Adams: The Oxford Introduction to Proto-Indo-European and the Proto-Indo-European World. Oxford (: University Press), 2006. MARTIROSYAN 2008: Hrach Martirosyan: Studies in armenian Etymology. With Special Emphasis on Dialects and Culture. Indo-European Heritage. Proefschrift ter verkrijging van de graad van Doctor aan de Universiteit Leiden. Leiden (: Faculty of Arts, Leiden University), 2008. URL: <https://openaccess.leidenuniv.nl/bitstream/handle/1887/12604/Front.pdf?sequence=4> [quot. 25. 12. 2012, 23:13] MAUE – SIMS-WILLIAMS 1991: Dieter Maue – Nicolas Sims-Williams: Eine Sanskrit-Sogdische Bilingue in Brahmi. Bulletin of the School of Oriental and African Studies 54/3 (: University of London), 1991, p. 486-495. MAYRHOFER 1989: Manfred Mayrhofer: Vorgeschichte der iranischen Sprachen; Uriranisch. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 4-24. ·233· MAYRHOFER 1992: Manfred Mayrhofer: Etymologisches Wörterbuch des Altindoarischen. I. Band. Heidelberg (: Carl Winter – Universitätsverlag), 1992. MAYRHOFER 1996: Manfred Mayrhofer: Etymologisches Wörterbuch des Altindoarischen. II. Band. Heidelberg (: Carl Winter – Universitätsverlag), 1996. MEIER-BRÜGGER 2003: Michael Meier-Brügger: Indo-European Linguistics. Berlin – New-York (: Walter de Gruyter), 2003. MENGHIN – PARZINGER – NAGLER 2007: Wilfried Menghin – Hermann Parzinger – Anatoli Nagler (eds.): Im Zeichen des goldenen Greifen. Königsgräber der Skythen. München – Berlin – London – New York (: Prestel), 2007. MĪRZŌZŌDA 2008: Сайфиддин Мирзозода: Фарҳанги яғнобӣ-тоҷикӣ. Душанбе (: Деваштич), 2008. MĪRZŌZŌDA – ALAVÎ 2008: Сайфиддин Мирзозода – Баҳриддин Алавӣ: Дастури забони яғнобӣ. Яғнобӣ зивоки дастур. Душанбе (: Девашт ич), 2008. MOLCHANOVA 2008: Е. К. Молчанова: Йезди (зороастрийский дари). In: Основы иранского языкознания. Среднеиранские и новоиранские языки. Москва (: Восточная лит ерат ура), 2008, p. 235-343. MONIER-WILLIAMS 1964: Monier Monier-Williams: A Sanskṛit – English Dictionary. Etymologically And Philologically Arranged with special reference to Cognate Indo-European Languages. Oxford (: Clarendon Press), 1964. MORGENSTIERNE 1926: Georg Valentin von Munthe af Morgenstierne: Report on a linguistic mission to Afghanistan. Oslo (: H. Aschehoug & Co., W. Nygaard), 1926. MORGENSTIERNE 1929: Georg Valentin von Munthe af Morgenstierne: Indo-Iranian Frontier Languages. Volume I. Parachi and Ormuri. Oslo (: H. Aschehoug & Co., W. Nygaard), 1929. MORGENSTIERNE 1938: Georg Valentin von Munthe af Morgenstierne: Indo-Iranian Frontier Languages. Volume II. Iranian Pamir Languages (Yidgha-Munji, Sanglechi-Ishkashmi and Wakhi). Oslo (: H. Aschehoug & Co., W. Nygaard), 1938. MORGENSTIERNE 1973: Georg Valentin von Munthe af Morgenstierne: Orthography and sound-system of the Avesta. In: Georg Valentin von Munthe af Morgenstierne: Irano-Dardica. Beiträge zur Iranistik 5. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1973, p. 31-83. ·234· MORGENSTIERNE 1974: Georg Valentin von Munthe af Morgenstierne: Etymological Vocabulary of the Shughni Group. Beiträge zur Iranistik 6. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1974. MORGENSTIERNE 1983a: Georg Valentin von Munthe af Morgenstierne: Afghanistan. vi. Paṧto. F. Waṇecī. In: Eḥsān Yārshāṭer (ed.): Encyclopædia Iranica [online edition]. London – Boston – Henley (: Routledge & Kegan Paul), 1983. URL: <http://www.iranicaonline.org/articles/afghanistan-vi-pasto> [quot. 06. 01. 2013, 17:58] MORGENSTIERNE 1983b: Georg Valentin von Munthe af Morgenstierne: Afghanistan. vii. Parāčī. In: Eḥsān Yāršāṭer (ed.): Encyclopædia Iranica [online edition]. London – Boston – Henley (: Routledge & Kegan Paul), 1983, p. 522-525. MORGENSTIERNE 2003: Georg Valentin von Munthe af Morgenstierne: A New Etymological Vocabulary of Pashto. Beiträge zur Iranistik 23. Wiesbaden (: Dr. Ludwig Reichert Verlag), 2003. MOSHKALO 2000: В. В. Мошкало: Ванеци язык//диалект. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 150-154. NÉMETH 1959: Julius Németh: Eine Wörterliste der Jassen, der ungarländische Alanen. Abhandlungen der Deutschen Akademie der Wissenschaften zu Berlin. Berlin (: Akademie-Verlag), 1959. NOVÁK 2009: Ľubomír Novák: ‫ﮔﻮﻳﺶ زﺑﺎن ﺗﺎﺟﻴﮑﺊ وادئ ﻳﻐﻨﺎب‬. Гӯиши забони тоҷикии водии Яғноб. URL: <http://www.academia.edu/1443513/_guyesh-e_zaban-e_tajiki-ye_vadi- ye_yaghnab_> [quot. 11. 03. 2010, 22:28] NOVÁK 2010: Ľubomír Novák: Jaghnóbsko-český slovník s přehledem jaghnóbské gramatiky. Яғнобӣчехӣ луғат яғнобӣ зивоки дастури феҳрастипӣ. Praha (: Filozofická fakulta Univerzity Karlovy v Praze), 2010. NOVÁK [in print]: Ľubomír Novák: Yaghnobi: an Example of a Language in Contact. In: Chatreššar 2011. Praha (: Filozofická fakulta Univerzity Karlovy v Praze), 2011, p. XX-YY. PAKHALINA 1966: Тат ьяна Николаевна Пахалина: Сарыкольский язык. Москва (: Наука), 1966. PAKHALINA 1969: Тат ьяна Николаевна Пахалина: Памирские языки. Москва (: Наука), 1969. PAKHALINA 1976a: Татьяна Николаевна Пахалина: Об индоарийских элементах в системе личных местимений восточноиранских языков. In: Иранское языкознание: история, этимология, типология. К 75-летию В. И. Абаева. Москва (: Наука), 1976, p. 79-84. ·235· PAKHALINA 1976b: Татьяна Николаевна Пахалина: О происхождении топонимов Ишкашим, Язгулям и Вахан. In: Иранское языкознание: история, этимология, типология. К 75-летию В. И. Абаева. Москва (: Наука), 1976, p. 178-181. PAKHALINA 1983: Татьяна Николаевна Пахалина: Исследования по сравнительно-исторической фонетике памирских языков. Москва (: Наука), 1983. PAKHALINA 1987a: Татьяна Николаевна Пахалина: Ваханский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки II. – Восточная группа. Москва (: Наука), 1987, p. 408-473. PAKHALINA 1987b: Татьяна Николаевна Пахалина: Ишкашимский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Новоиранские языки II. – Восточная группа. Москва (: Наука), 1987, p. 474-536. PAKHALINA – QURBŌNOV 2000: Татьяна Николаевна Пахалина – Х. Курбанов: Ишкашимский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 196-208. PAYNE 1989: John Payne: Pāmir Languages. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 418-444. PERRY 2005: John R. Perry: A Tajik Persian Reference Grammar. Handbuch der Orientalistik 11. Leiden (: Brill), 2005. PROVASI 2009: Elio Provasi: Versification in Sogdian. In: Werner Sundermann – Almut Hintze – François de Blois (eds.): Exegisti monumenta. Festschrift in honour of Nicholas Sims-Williams. Wiesbaden (: Harrassowitz), 2009, p. 347-368. PULJU 2000: Tim Pulju: Indo-European *d, *l, and *dl. In: John Charles Smith – Delia Bentley (eds.): Historical linguistics 1995. Volume 1: General issues and non-Germanic Languages. Selected Papers from the 12th International Conference on Historical Linguistics, Manchester, August 1995. Amsterdam – Philadelphia (: John Benjamins Publishing Co.), 2000, p. 311326. QARĪB 1383: Badrezzaman Gharib: Sogdian Dictionary (Sogdian – Persian – English). . Tehran (: Farhangan Publications), 2004. .۱۳۸۳ ،(‫ ﻓﺮﻫﻨﮕﺎن‬:) ‫ﺮان‬ . .(‫ ﻓﺮﻫﻨﮓ ﺳﻐﺪی )ﺳﻐﺪی – ﻓﺎرﺳﻰ – اﻧﮕﻠﻴﺴﯽ‬:‫اﻟﺰﻣﺎن ﻗﺮﻳﺐ‬ ّ ‫ﺑﺪر‬ QARĪB 1965: Badresaman Gharib: Analysis of the Verbal System in the Sogdian Language. A Dissertation in Oriental Studies Presented to the Faculty of the Graduate School of Arts and ·236· Science of the University of Pennsylvania in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy. Philadelphia, 1965. RASTORGUEVA 1964: Вера Сергеевна Расторгуева: Опыт сравнительного изучения таджикских говоров. Москва (: Наука), 1964. RASTORGUEVA 1966: Вера Сергеевна Расторгуева: Иранские языки. Введение. In: В. В. Виноградов (ed.): Языки народов СССР. Том первый: Индоевропейские языки. Москва (: Наука), 1966, p. 194-211. RASTORGUEVA – ÈDEL’MAN 2000: Вера Сергеевна Расторгуева – Джой Иосифович Эдельман: Этимологический словарь иранских языков. Том 1: а-ā. Москва (: Вост очная лит ерат ура), 2000. RASTORGUEVA – ÈDEL’MAN 2003: Вера Сергеевна Расторгуева – Джой Иосифович Эдельман: Этимологический словарь иранских языков. Том 2: b-d. Москва (: Вост очная лит ерат ура), 2003. RASTORGUEVA – ÈDEL’MAN 2007: Вера Сергеевна Расторгуева – Джой Иосифович Эдельман: Этимологический словарь иранских языков. Том 3: f-h. Москва (: Вост очная лит ерат ура), 2007. RASTORGUEVA – MOLCHANOVA 1981: Вера Сергеевна Расторгуева – Е. К. Молчанова: Парфянский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Среднеиранские языки I. Москва (: Наука), 1981, s. 147-232. REINHOLD 2006: Beate Reinhold: Neue Entwicklungen in der Wakhi-Sprache von Gojal (Nordpakistan). Berlin (: Harrassowitz), 2006. ROBSON – TEGEY 2009: Barbara Robson – Habibullah Tegey: Pashto. In: Gernot Windfuhr (ed.): Iranian Languages. London – New York (: Routledge), 2009, p. 721-772. RÓNA-TAS 1998: András Róna-Tas: The Reconstruction of Proto-Turkic and the Genetic Question. In: Lars Johanson – Éva Ágnes Csató (eds.): The Turkic Languages. London – New York (: Routledge), 1998 (2006 reprint), p. 67-80. RONG 2005: Rong Xinjiang (Rong Sin-ťiang): The Name of the So-called “Tumshuqese” Language. In: Carol Altman Bromberg – Nicolas Sims-Williams – Ursula Sims-Williams (eds.): Bulletin of the Asia Institute. Iranian and Zoroastrian Studies in Honor of Prods Oktor Skjærvø. New Series/Volume 19, 2005, p. 119-127. ROSS – GAUTHIOT 1913: E. D. Ross – Robert Gauthiot: L’alphabet sogdien d’après un témoignage du XIIIᵉ siècle. Journal Asiatique, 1913, p. 521-533. ·237· ROZENFEL’D 1964: Анна Зиновьевна Розенфельд: Ванджские говоры таджикского языка. Ленинград (: Издат ельст во ленинградского университета), 1971. ŞAMBIZODA 1937: Çamşed ŞamËizoda: AlifËe. Awalөn sol çat. StalinoËod, 1937. ŞAMBIZODĀT 1931: M. B. ŞamËizodāt: Xuꞕnөni alifËǝ. Ƣullajen çāt. SitalinoËod – Toşkand, 1931. SCHENKER 1993: Alexander M. Schenker: Proto-Slavonic. In: Bernard Comrie – Greville G. Corbett (eds.): The Slavonic Languages. London (: Routledge), 1993, 60-121. SIMS-WILLIAMS 1979: Nicolas Sims-Williams: On the Plural and Dual in Sogdian. Bulletin of the School of Oriental and African Studies 42/2 (: University of London), 1979, s. 337-346. SIMS-WILLIAMS 1981a: Nicolas Sims-Williams: The Sogdian sound-system and the origins of the Uyghur script. Journal Asiatique 269, 1981, p. 347-360. (with errata-slip distributed with JA 270, 1982) SIMS-WILLIAMS 1981b: Nicolas Sims-Williams: Some Sogdian denominal abstract suffixes. Acta Orientalia XLII, Copenhagen (: Munksgaard), 1981, p. 11-19. SIMS-WILLIAMS 1982: Nicolas Sims-Williams: The double system of nominal inflection in Sogdian. Transactions of the Philological Society 1982, Oxford, 1982, s. 67-76. SIMS-WILLIAMS 1984: Nicolas Sims-Williams: The Sogdian “Rhytmic Law”. In: Wojciech Skalmowski, Alois van Tongerloo (eds.): Middle Iranian Studies: Proceedings of the International Symposium organized by the Katholieke Universiteit Leuven from the 17th to the 20th of May 1982. Leuven (: Katholieke Universiteit te Leuven), 1984, p. 203-215. SIMS-WILLIAMS 1988 [online]: Nicolas Sims-Williams: Bactrian Language. In: Eḥsān Yārshāṭer (ed.): Encyclopædia Iranica [online edition]. Costa Mesa, California, p. 344-349. URL: <http://www.iranicaonline.org/articles/bactrian-language> [quot. 16. 02. 2013, 23:18] SIMS-WILLIAMS 1989a: Nicolas Sims-Williams: Eastern Middle Iranian. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 165-172. SIMS-WILLIAMS 1989b: Nicolas Sims-Williams: Sogdian. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 174-192. SIMS-WILLIAMS 1989c: Nicolas Sims-Williams: Bactrian. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 230-235. ·238· SIMS-WILLIAMS 1996a: Nicolas Sims-Williams: The Sogdian manuscripts in Brāhmī script as evidence for Sogdian phonology. In: Ronald Eric Emmerick et ali. (eds.): Turfan, Khotan und Dunhuang: Vorträge der Tagung “Annemarie von Gabain und die Turfanforschung” veranstaltet von der Berlin-Brandenburgischen Akademie der Wissenschaften in Berlin (9.-12. 12, 1994). Berlin, 1996, p. 307-315. SIMS-WILLIAMS 1996b [online]: Nicolas Sims-Williams: Eastern Iranian Languages. In: Eḥsān Yārshāṭer (ed.): Encyclopædia Iranica [online edition]. Costa Mesa, California. URL: <http://www.iranicaonline.org/articles/eastern-iranian-languages> [quot. 05. 12. 2012, 12:34] SIMS-WILLIAMS 2012: Nicolas Sims-Williams: Yaghnobi as a Sogdian dialect. [handout presented on 11. 5. 2012 at Symposium in the memory of Manfred Mayrhofer (1929-2011): Iranian and Indo-European Onomastics and Linguistics, Vienna, May 10-12, 2012] SIMS-WILLIAMS – HAMILTON 1990: Nicolas Sims-Williams – James Hamilton: Documentes turco-sogdiens du IXe-Xe siècle de Touen-houang. Corpus inscriptionum iranicarum. London ( : School of Oriental and African Studies), 1990. SJÖGREN 1844: Андрей Михайловичь Шёгренъ: Осетинская грамматика съ краткимъ словаремъ осетинско-россійскимъ и россійско-осетинскимъ. Санктпетербургъ (: Типографія Императской Академіи Наукъ), 1844. SKJÆRVØ 1989a: Prods Oktor Skjærvø: Modern Eastern Iranian. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 370-383. SKJÆRVØ 1989b: Prods Oktor Skjærvø: Pashto. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 384-410. SKJÆRVØ 1989c: Prods Oktor Skjærvø: Yidgha and Munǰī. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 411-416. SKJÆRVØ 2005: Prods Oktor Skjærvø: An Introduction to Old Persian (revised and expanded 2nd version). 2005. URL: <http://www.fas.harvard.edu/~iranian/OldPersian/opcomplete.pdf> [quot. 01. 10. 2008, 00:33] SKÖLD 1936: Hannes Sköld: Materialien zu den iranischen Pamirsprachen. Lund (: C. W. K. Gleerup), 1936. SMIRNOVA 1963: O. I. Smirnova: La carte des regions du haut Zerafchan dʼapprès les documents du Mt. ·239· Mough. In: Труды двадцать пятого международного конгресса востоковедов. Москва 9-16 августа 1960 г. Том II, заседания серий VI-IX, XII. Moscow – Nendeln/Liechtenstein (: Kraus-Thompson Organization Limited), 1963, p. 329-337. SOKOLOVA 1953a: Валетина Степановна Соколова: Ягнобский язык. In: Валетина Степановна Соколова: Очерки по фонетике иранских языков. Выпуск II. Осетинский, ягнобский и памирские языки. Москва – Ленинград (: Издательство Академии наук СССР), 1953, p. 59-79. SOKOLOVA 1953b: Валетина Степановна Соколова: Шугнано-рушанская группа. In: Валетина Степановна Соколова: Очерки по фонетике иранских языков. Выпуск II. Осетинский, ягнобский и памирские языки. Москва – Ленинград (: Издательство Академии наук СССР), 1953, p. 84-175. SOKOLOVA 1953c: Валетина Степановна Соколова: Ишкашимский язык. In: Валетина Степановна Соколова: Очерки по фонетике иранских языков. Выпуск II. Осетинский, ягнобский и памирские языки. Москва – Ленинград (: Издательство Академии наук СССР), 1953, p. 230-240. SOKOLOVA 1966: Валетина Степановна Соколова: Шугнано-рушанская языковая группа. In: В. В. Виноградов (ed.): Языки народов СССР. Том первый: Индоевропейские языки. Москва (: Наука), 1966, p. 362-397. SOKOLOVA 1967: Валетина Степановна Соколова: Генетические отношения язгулямского языка и шугнанской языковой группы. Ленинград (: Наука), 1967. SOKOLOVA 1973: Валетина Степановна Соколова: Генетические отношения мунджанского языка и шугнано-язгулямской языковой группы. Ленинград, 1973. SOPHRONIOU 1962: Sofronios Agathocli Sofroniou: Teach Yourself Modern Greek. London, 1962. STEBLIN-KAMENSKIY 1976: Иван Михайлович Стеблин-Каменский: Два ваханских топонима. In: Иранское языкознание: история, этимология, типология. К 75-летию В. И. Абаева. Москва (: Наука), 1976, s. 182-185. STEBLIN-KAMENSKIY 1981: Иван Михайлович Стеблин-Каменский: Бактрийский язык. In: Вера Сергеевна Расторгуева (ed.): Основы иранского языкознания. Среднеиранские языки. Москва (: Наука), 1981, p. 314-346. STEBLIN-KAMENSKIY 1999: Иван Михайлович Стеблин-Каменский: Этимологический словарь ваханского языка. Ethymological Dictionary of the Wakhi Language. Санкт-Петербург (: Петербургское Вост оковедение), 1999. ·240· SUNDERMANN 1989: Werner Sundermann: westmitteliranische Sprachen. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 106-113. TEDESCO 1926: Paul Tedesco: Ostiranische Nominalflexion. In: Zeitschrift für Indologie und Iranistik, Band 4. Leipzig (: Deutschen Morgenländische Gesellschaft), 1926, s. 94-166. THORDARSON 1989: Fridrik Thordarson: Ossetic. In: Rüdiger Schmitt (ed.): Compendium Linguarum Iranicarum. Wiesbaden (: Dr. Ludwig Reichert Verlag), 1989, p. 456-479. TOMASCHEK 1880: Wilhelm Tomaschek: Central-asiatische studien II. Die Pamir-Dialecte. Wien, 1880. TURNER 1927: R. L. Turner: Notes on Dardic. In: Bulletin of the School of Oriental Studies 4/3 (: University of London), 1937, p. 533-541. DE UJFALVY DE MEZŐ-KÖVESD 1882: Charles-Eugène de Ujfalvy: La langue des Yagnobis. In: Revue de linguistique et de philologie comparée XV, Paris, 1882, p. 271-292. DE VAAN 2008: Michiel de Vaan: Etymological Dictionary of Latin and the other Italic Languages. Leiden Indo-European Etymological Dictionary Series. Leiden – Boston (: Brill), 2008. DE LA VAISSIÈRE 2005: Étienne de la Vaissière: Sogdian Traders. A History. Handbuch der Orientalistik 10. Leiden – Boston (: Brill), 2005. VAVROUŠEK 2007: Petr Vavroušek: O rekonstrukci praindoevropštiny. Praha (: Filozofický fakulta Univerzity Karlovy v Praze), 2007. VINOGRADOVA 2000a: Софья Петровна Виноградова: Согдийский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 58-95. VINOGRADOVA 2000b: Софья Петровна Виноградова: Ягнобский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 290-310. VITCHAK 1992: К. Т. Витчак: Скифский язык: опыт описания. In: Вопросы языкознания 1991, №5. Москва (: Наука), 1992, p. 50-59. WALDE 1906: Alois Walde: Lateinisches etymologisches Wörterbuch. Heidelberg (: Carl Winter’s – Universitätsbuchhandlung), 1906. WENDTLAND 2011: Antje Wendtland: The Emergence and Development of the Sogdian Perfect. In: Agnes Korn – Geoffrey Haig – Simin Karimi – Poller Samvelian (eds.): Topics in Iranian ·241· Linguistics. Beiträge zur Iranistik 34. Wiesbaden (: Dr. Ludwig Reichert Verlag), 2009, p. 39-52. YOSHIDA 2009a: Yutaka Yoshida: Sogdian. In: Gernot Windfuhr (ed.): Iranian Languages. London – New York (: Routledge), 2009, p. 295-335. YULE – CORDIER 1993: Henry Yule – Henri Cordier: The Travels of Marco Polo. The Complete Yule-Cordier Edition. Volume I. Toronto (: General Publishing Company), 1993. YŪSUFBĒKOV 2000: Шодихон П. Юсуфбеков: Сангличский язык. In: Языки мира. Иранские языки III. – Восточноиранские языки. Москва (: ИНДРИК), 2000, p. 186-196. YŪSUFBĒKOV – DODYKHUDOEVA 2008: Шодихон П. Юсуфбеков – Л. Р. Додыхудоева: Сангличский язык. In: Основы иранского языкознания. Среднеиранские и новоиранские языки. Москва (: Вост очная лит ерат ура), 2008, p. 110-234. ZARSHENĀS 1357: v Zohre Zaršenās: Ḵ ārazmī Language. In: Nāme-ye Farhangestān Vol. 2, No. 1 (Ser. No. 5), 1357, p. 53-65. .۵۳-۶۵ ،۱۳۵۷ ،(۵ ‫ )ﻣﺴﻠﺴﻞ‬۲/۲ ‫ﻧﺎﻣﻪ ﻓﺮﻫﻨﮕﺴﺘﺎن‬ ٔ In: .‫ زﺑﺎن ﺧﻮارزﻣﯽ‬:‫زﻫﺮﻩ زرﺷﻨﺎس‬ ٔ ZARUBIN 1924: Иван Иванович Зарубин: К списку памирских языков. In: Доклады Российской Академии Наук, 1924, серия В, p. 79-81. ·242·