ON POSSIBLE DARDIC AND BURUSHASKI INFLUENCE ON SOME NORTHWESTERN TIBETAN DIALECTS Текст научной статьи по специальности «Языкознание и литературоведение»

Anton Kogan

Institute of Oriental Studies, Russian Academy of Sciences, Moscow; kogan_anton@yahoo.com

On possible Dardic and Burushaski influence on some Northwestern Tibetan dialects

The Northwestern fringe of the Tibetan-speaking area, now forming a part of the Jammu and Kashmir state of India and of Pakistani-controlled Northern Areas, was in the past an area of intensive ethnic and language contact. This contact resulted in the linguistic assimilation of the local pre-Tibetan population by the Tibetans. More than a century ago it was hypothesized that this pre-Tibetan population may have spoken a certain Dardic language. The article attempts to check this hypothesis through the etymological analysis of the vocabulary of Northwestern Tibetan dialects. The results of this analysis suggest the existence of a significant Indo-Iranian, probably Dardic, lexical stratum, as well as of numerous lexemes borrowed from some early form of Burushaski. The author seeks to define the dialectal distribution of Indo-Iranian and Burushaski loanwords in the area under study.

Keywords: language contact; linguistic substratum; Tibetan dialects; Ladakhi language; Balti language; Purik language; Dardic languages; Burushaski language.

Introduction

It is well known that the now vast and extensive Tibetan-speaking area came into being at a comparatively recent date, i.e. in the 7th-9th centuries. Its formation was the result of the expansion of the Tibetan Empire and of the subsequent spread of the Tibetan language into originally non-Tibetan territories. This fact suggests that the existence of pre-Tibetan substrata in different groups of Tibetan dialects may be hypothesized.

Such a hypothesis was probably first formulated more than a century ago by the renowned German Tibetologist August Hermann Francke for the Ladakh region. According to Francke (1907), this area situated on the northwestern fringe of historical and linguistic Tibet was populated by the two ethnic groups, namely the Mons and the Dards, in pre-Tibetan times. The exact origin of the former group remains unclear, while the Dards were considered to be related to the inhabitants of eastern Hindu Kush and western Karakoram valleys, including Gilgit1. Francke also argued that this "Dardic" population had left an ethnic trace in the Ladakh of his times, its last remnant being the dwellers of several remote villages preserving their original language (Francke 1906). These people undoubtedly belonged to the ethnic group that is nowadays called Brokpa. The Brokpas live in Ladakh to this day, residing in certain areas of Leh and Kargil districts2, and speak Brokskat, a Dardic language closely related to Shina. Although Francke's theory was put forth very long ago and is still accepted by some Tibetologists3, no systematic attempts to verify it nor to define its linguistic implications have been made thus far. The purpose of the present article is to make a first step in this direction, and thus to throw light on the ethnic and linguistic composition of Ladakh and adjoining areas in the pre-Tibetan period.

1 A. H. Francke once supposed that the Mons could have been "an Indian tribe" (Francke 1907: 20). The term "Dards", nowadays denoting the speakers of Dardic languages, in Francke's times was used mainly geographically and could be applied e.g. to Burushaski-speakers and even to the Tibetan-speaking population of Baltistan.

2 Both districts are now part of the Jammu and Kashmir state of India.

3 Among recent publications see e.g. Zeissler 2010.

Journal of Language Relationship • Вопросы языкового родства • 17/4 (2019) • Pp. 263-284 • © The authors, 2019

Testing Francke's hypothesis

First of all, it should be noted that the above-cited theory can hardly be accepted in full because it seems to contradict certain linguistic facts. The etymological study of Brokskat vocabulary yields us some hints as to the Brokpa people's past, and what we learn from these data seems to be inconsistent with some of Francke's ideas. The Brokskat lexicon contains a considerable number of Persian and Arabic loanwords, some of which were attested in the earliest descriptions of the language made by European scholars, cf., e.g., qoda (Shaw 1878: 46), khoda (Francke 1905: 94) 'god' < Persian xoda; barkhad 'blessing' (Francke 1905: 101) < Persian barakat < Arabic; "nasib" 'fate' (Shaw 1878: 46) < Persian nasib < Arabic; darbar 'assembly' (Francke 1905: 96) < Persian darbar '(royal) court, assembly'. The presence of such loanwords cannot be attributed to the influence of Urdu, the present-day official language of the area. It was only in the second decade of the 20th century when the actual spread of Urdu in different spheres of social life began in Ladakh. So this language was not widely known and used in the region at the time of Francke's fieldwork (i.e. between 1896 and 1909), not to mention the period when the author of the first grammatical sketch of Brokskat, Robert Barkley Shaw,4 served in Ladakh as the British joint commissioner. Anyway, even the influence of the official language can hardly convincingly account for the borrowing of Persian and Arabic religious terms 5 into a dialect whose speakers are Buddhists. The most plausible explanation for this astonishing fact is probably the assumption that the Brokpas had been, if not Muslims themselves, at least strongly influenced by Islamic culture6 before migrating to Ladakh and converting to Tibetan Buddhism in this region. In such a case, however, the Brokpa migration can by no means predate the Tibetan conquest of the 8th century.

It is noteworthy in this connection that the Brokpas, according to their own historical tradition, migrated to Ladakh from Gilgit (Francke 1907: 37-38). This story, however, should not be taken at face value. Glottochronological calculations show that Brokskat separated from the closely related Shina early in the 1st millennium A.D. (Kogan, Vasilyev 2013), while there is strong reason to believe that the language of the Gilgit valley was at that time Burushaski (Jettmar 1975). The geography and chronology of the Brokpas' migrations is still a puzzling issue, but what can be stated with certainty at the current state of our knowledge is that their movement to the present-day habitat could not have immediately followed the split of Proto-Shina-Brokskat, but must have been separated from the latter by many centuries.

Revisiting the issue of Tibetan-Dardic contact

It can be plainly seen that the above facts are hardly consistent with Francke's hypothesis that pre-Tibetan inhabitants of Ladakh were the ancestors of the modern Brokpas. But does this really mean that the issue of Dardic ethnic and linguistic substratum in the area is no longer relevant? There are some reasons to answer this question in the negative. Archaeological excavations, conducted under Francke's guidance, showed that prior to the expansion of the Tibetans, Ladakh was populated by people belonging to the Caucasoid race and possessing certain

4 This sketch is included in the above-mentioned work (Shaw 1878). R. B. Shaw lived in Ladakh and collected Brokskat material in 1870s.

5 In addition to the foregoing examples we may quote e.g. saltan 'demon' < Persian saltan 'devil, Satan' < Arabic.

6 The use of Persian as a literary language, along with the use of Arabic as the language of religion, is the characteristic feature of Muslim culture in the region in question. That is why the presence of certain Persian borrowings in Brokskat may also be considered as evidence of the erstwhile influence of Islam on its speakers.

cultural traits, e.g. burial practices, similar to those of the Dardic- and Burushaski-speaking population of the Hindu Kush and Karakoram valleys (Francke 1906; 1907). This cultural and physical-anthropological similarity is, strictly speaking, of no relevance to comparative linguistic studies, but nevertheless, it might well have resulted from ethnic affinity, which in many cases implies genetic relationship of languages. In this regard it should also be pointed out that the easiest route of migration into Ladakh is along the Indus, i.e. either from the South-East, where the population is linguistically Sino-Tibetan and belongs to the Mongoloid race, or from the North-West, where the people speak Dardic languages and Burushaski, and are racially Caucasian. This fact alone suggests that the existence of Dardic and/or Burushaski ethnic elements in pre-Tibetan Ladakh as well as Dardic and/or Burushaski substrata in the local Tibetan dialects can by no means be ruled out.

The possibility of Dardic- and Burushaski-Tibetan language contact is also supported by certain typological peculiarities of Northwestern Tibetan dialects. They possess a number of features at different levels of language structure, which bring them closer to the languages of the Hindu Kush-Karakoram region, as opposed to Old and Classical Tibetan. In Ladakhi7 such features include the presence of cerebral consonants (stops and sibilants), the threefold contrast of deictic pronouns (cf. i, di 'this', ote 'that (within sight)', te 'that (out of sight)'8 vs. the twofold opposition 'di 'this' — de 'that' in Classical Tibetan), and the word order where a demonstrative pronoun precedes the noun it modifies (cf. te khi 'that dog' vs Classical Tibetan khyi de). It is worth noting that the aforementioned typological traits are also characteristic of the Tibetan varieties spoken to the North and West of Ladakh and conventionally grouped under the names Balti and Purik9. Moreover, these varieties show certain additional "non-Tibetan" traits shared by the Dardic languages and Burushaski, namely, cerebral affricates10 and the typical position of an adjective before the noun it qualifies (cf. Balti, Purik di chogo nakpo khi 'this big black dog' vs Classical Tibetan khyi nag-po chen-po 'di)11.

It should, however, be understood that the above features alone can hardly be a decisive argument in favor of Dardic or Burushaski influence. Most of them are found in many lan-

7 The term "Ladakhi", as it appears in literature, seems to be somewhat ambiguous. It may be used for the dialect of Leh, the political center of Ladakh, as well as for all the Tibetan dialects spoken in the Leh district of Jammu and Kashmir. The features listed here are, however, found in all or most of the Ladakhi dialects.

8 Similar systems are found in Dardic languages. Cf., e.g., Kashmiri yi 'this', hu 'that (within sight)', su 'that (out of sight)'; Khowar haiya 'this (near, present)', hes 'that (remote, present)', hasa 'that (more remote, absent)'; Pashai ae(m) 'this', (a)sa 'that (within sight)', o(m) 'that (anaphoric)' (Grierson 1919; Morgenstierne 1967).

9 Historically Balti was the name of the Tibetan dialect spoken in and around the town of Skardu, nowadays the capital of the district with the same name in the Gilgit-Baltistan territory of Pakistan. The name Purik (or Purki) was traditionally used for the dialect of the Kargil town and its environs, now in the Indian state of Jammu and Kashmir. The speakers of these two and a number of adjoining dialects are, unlike Ladakhis and Tibetans, Muslims by religion, and at present the names Balti and Purik are sometimes used as umbrella terms for Tibetan varieties spoken by the Muslim population of Pakistani-controlled Gilgit-Baltistan and Indian-controlled Kargil district respectively.

10 Balti, Purik and Ladakhi cerebrals (stops, affricates and sibilants) in inherited words usually reflect certain Old Tibetan consonantal clusters.

11 One more feature, alien to other dialects, is found in Balti and Purik verbal systems. It is the use of the imperative stem for forming the Prohibitive. In the rest of Tibetan the latter is formed from the present stem. The German scholar Bettina Zeissler attributes this peculiarity of Muslim varieties to the substratum interference (Zeissler 2004: 344-345). Anyhow, this feature also unites both Purik and Balti with Burushaski and Dardic languages. It is remarkable that in this case linguistic typological similarity is coupled with the cultural and physical-anthropological one. Balti- and Purik-speakers share with their non-Tibetan western and northern neighbors not only a common religion, i.e. Islam, but also many traits of material culture and to some extant even a common racial type, a significant fraction of them being Caucasians.

guages belonging to different language groups and families, e.g. Indo-Aryan, Iranian or Dra-vidian, while some are characteristic of larger wolrd regions12. The most convincing evidence for intensive contact between particular languages is undoubtedly not typological similarity but the existence of numerous common lexical isoglosses. The issue of possible non-Tibetan influence on the vocabulary of Northwestern Tibetan dialects is still extremely poorly studied13. In the past the main cause of such a situation was the lack of material. Nowadays things have changed. Dictionaries of Balti (Sprigg 2002) and Ladakhi (Abdul Hamid 1998; Norman 2010) dialects, compiled and published within the last 20-odd years, can provide a relatively rich lexical material. In the following section I list the results of the etymological study of this material, conducted with a view to identifying possible Burushaski and Indo-Iranian14 elements in it.

As the main data sources I used Richard Keith Sprigg's "Balti-English English-Balti dictionary" (Sprigg 2002) and "A dictionary of the Language Spoken by Ladakhis" by Rebecca Norman (Norman 2010). The latter dictionary, unfortunately, available only as an unpublished draft, contains a wealth of lexical data not only from Ladakhi but also from the Purik and Balti dialects. For the Purik data, "A grammar of Purik Tibetan" by Marius Zemp (Zemp 2018) was also consulted. Words of supposedly Burushaski or Indo-Iranian origin, found in these sources, were checked for etymological parallels in other varieties of Tibetan. For this purpose Heinrich August Jaschke's Tibetan-English dictionary (Jaschke 1881) was employed. It includes a good number of lexical items, peculiar to western regions of linguistic Tibet. All such items are marked with specific labels, e.g. Bal. (Balti), Ld. (Ladakh), Lh. (Lahaul), Pur. (Purik), W. (Western Tibet). The absence of one of these labels in a particular entry is likely to suggest that the area of geographical distribution of the respective lexeme is not confined to Western Tibet, and its being borrowed from a language of the Hindu Kush-Karakoram region does not thus seem probable. Naturally, such lexemes were not included in the following list. Sources of Burushaski, Dardic and Iranian material are specified in each entry of the list. Indo-Aryan and some Dardic etymologies were taken from "A comparative dictionary of the Indo-Aryan languages" by R.L. Turner (Turner 1966).

A list of possible Indo-Iranian and Burushaski loanwords in Northwestern Tibetan dialects

1. Balti ashi 'responsibility, someone responsible', ashipa 'owner, master, husband, Lord' (Sprigg 2002: 22), Purik ashipa 'owner, husband (said by the wife)' (Norman 2010: 1081). Cf. Burushaski (Nagir) di-asi 'überlegen sein, überwältigen' (Berger 1998b: 23).

12 This holds true for the word order in the noun phrase with demonstratives and adjectives preceding the noun.

13 This is particularly the case with the dialects of Ladakhi. On Balti and Purik some research has been done which demonstrated the existence of a number of lexical isoglosses common to these varieties and Burushaski (Lorimer 1939; Zemp 2018).

14 I intentionally do not confine myself to detecting possible Dardic loanwords only. Dardic is not the only branch of the Indo-Iranian subfamily, to which the languages, that have influenced Northwestern Tibetan dialects, may have belonged. Some Indo-Aryan languages are spoken in Western and Central Himalaya to the southwest of Ladakh, and the population of the western part of the present-day Xinjiang, bordering upon Ladakh and Gilgit-Baltistan, was in the past Iranian-speaking. For this reason, the presence of Indo-Aryan and Iranian (other than Persian) borrowings in Ladakhi, Purik and Balti cannot be ruled out. It should, however, be kept in mind that the Northwestern Tibetan regions are much more easily accessible from the Dardic-speaking area, than from the Iranian and Indo-Aryan-speaking ones, the latter two being separated by the world's highest mountain ranges, the Himalaya and the Karakoram.

2. Ladakhi asur 'mustard seed as a spice used in Ladakhi pickle' (Norman 2010: 1073). Cf. OIA asuri 'Sinapis ramosa', Sindhi ahuri 'mustard seed', Lahnda ohur, ahur, ahur, Punjabi ahur, Hindi-Urdu asuri 'mustard', Kashmiri asoru 'mustard plant, Sinapis ramosa' (Grierson 1915-1932: 51).

3. Balti baan 'man or men who sing religious songs and foretell the future' (Sprigg 2002: 24). Cf. Kashmiri wan-, Indus Kohistani ban- 'to say', Kalasha bandek '1. to teach; 2. to announce so as to inaugurate, to order or command' (Trail, Cooper 1999: 27), OIA vandate 'praises, worships', Av. vand-, Khotanese van-, Parthian wynd- 'to praise, honor, worship'.

4. Balti balbul 'tepid' (Sprigg 2002: 25), Ladakhi mala-mule 'lukewarm, tepid' (Norman 2010: 687). Jaschke 1881: mal-la-mul-le Ld. lukewarm, tepid. Cf. Burushaski bulbulo 'lukewarm (of water)' (Lorimer 1938: 87), Shina bubulu 'tepid' (Bailey 1924), Indus Kohistani bubul 'lukewarm (water)' (Zoller 2005: 315).

5. Balti bar-ban 'window (in a wall), glass-pane window' (Sprigg 2002: 26), Purik barban 'window' (Zemp 2018: 945). Cf. Brokskat barban id. The element bar- is probably connected with PII *dwar- 'door' (cf. OIA dvar- 'door, gate', Av. duuar- 'gate'). In many Dardic languages the derivatives of this root have the meaning 'window': Pashai dari, Gawar-Bati deri, Kalasha durik, Phalura daruri, Indus Kohistani dari, Shina darn, Kashmiri ddr. Cf. also Burushaski ddri 'window' (Lorimer 1938: 114). The element -ban may be etymologically identical to Balti ban 'fence' (Sprigg 2002: 25), which also seems to be of Indo-Iranian origin (cf. OIA bandha- 'border, framework, damming').

6. Balti basanda 'dandelion' (Sprigg 2002: 27). R.K. Sprigg compares this word with Hindi-Urdu basanti 'yellow (the color of spring blossoms, such as those of the mustard-plant)'. This etymology, however, implies the unexplained loss of the final long i by borrowing. From the viewpoint of historical phonology, the more probable source of the Balti lexeme seems to be an Indo-Iranian form related to OIA vasanta- 'vernal, pertaining to spring' (with subsequent semantic change 'vernal' > 'a spring flower' > 'dandelion'). Cf. also OIA vasanta-, Shina bazon, Phalura basand, Bashkarik basan, Torwali basan, Kalasha basun, Khowar bosun, Pashai wahdn(d), Gawar-Bati wasand 'spring'.

7. Balti bat 'boiled mixture of germinated grain flour and ordinary flour (made during Ramzan)' (Sprigg 2002: 27). Cf. OIA bhakta- 'food; boiled rice', Khowar bot 'evening meal', Kalasha batay 'flour taken to be eaten in high pastures' (Trail, Cooper 1999), Shina bat, Bashkarik batt, Torwali bat, Kashmiri bati, Lahnda, Punjabi bhatt, Hindi-Urdu, Nepali, Gujarati, Marathi bhat 'boiled rice', Burushaski brase bat 'plain, boiled rice', alue bat 'mashed potatoes' (Lorimer 1938: 73), Wakhi bat 'wheat flour gruel', Ishkashimi bat, Shughni bat, Munji batdk 'ritual food made of wheat flour and butter, cooked in water or milk' (Steblin-Kamensky 1999: 95).

8. Balti bekar 'court singer and dancer who improvises poems and songs' (Sprigg 2002: 28), Ladakhi beda 'member of the caste that used to be itinerant musicians', bemo f. (Norman 2010: 640-641). Jaschke 1881: 'be-dha a class of itinerant musicians W. Sprigg compares the Balti word with Hindi-Urdu bhik(h)ari 'beggar', which does not seem semantically convincing, and in addition implies the unnecessary separation of Balti and Ladakhi forms. All the three words in question are clearly derivatives of a common root sounding as be-. 15 This root might have been borrowed from some Indo-Iranian language where it was a reflex of PII *wad(i)ya- 'music, musical instrument' (cf. OIA vadya- 'musical instrument, music', vadyakara- 'musician', vadayati 'plays a musical instrument', Punjabi vajja 'musical instrument', Hindi-Urdu baja 'music', Kashmiri waz- 'to sound (of bell, clock etc.)', way- 'to play a musical instrument', Shina (Guresi) baz- 'to strike (of a gong etc.)', Ossetic wadynj 'flute, panpipe' < *vaddnici (Abaev 1989: 35)).

15 In Ladakhi -mo is a productive suffix indicating the female gender.

The phonological development in the donor language seems to have passed through the intermediate stage *baiya- with the change *w > b16 and the drop of intervocalic d. Both processes are characteristic of many Indo-Iranian languages of the region. Burushaski berits 'musician, bandsman; craftsman' (Lorimer 1938: 75) may also contain the aforesaid root.

9. Balti (Khaplu dialect) bosoq 'calf' (Norman 2010: 604). Cf. Burushaski busooso (Hunza, Nagir), boso (Yasin) id. (Berger 1998b: 65). It is evident that the Balti word is phonologically closer to the Burushaski ones than to Tibetan be-to 'calf'.

10. Balti bwaar 'watermelon' (Sprigg 2002: 35). Cf. Shina buar (Bailey 1924), Burushaski (Hunza, Nagir, Yasin) buar id. (Berger 1998b: 60).

11. Balti, Purik cancan17 'naked' (Sprigg 2002: 40; Zemp 2018: 926; Norman 2010: 288). Cf. Khowar, Kalasha can id.

12. Ladakhi cancil, chancil 'the green outer shell or fruit of walnut' (Norman 2010: 267). Jäschke 1881: can-cil W. the green shell of a walnut. Cf. Burushaski chanjil18 (Nagir, Hunza), canjil (Yasin) 'die grüne äussere Schale der Walnuss, Häutchen zwischen den Teilen des Walnusskerns', Shina chaciil id. (Berger 1998b: 96).

13. Balti cangti 'drop', Purik, Ladakhi (Shamskat and Leh dialects) cangti 'leak in a roof' (Norman 2010: 267). Cf. Burushaski chan man-, Shina chachan- '(Wasser) sickern, herabfliessen' (Berger 1998b: 97).

14. Balti ceri, Ladakhi cari 'bedbug' (Sprigg 2002: 41; Norman 2010: 266). Jäschke 1881: ca-ri W. bug. Cf. Khowar cari id., Burushaski cari, ciri 'Heimchen, Grille' (Berger 1998b: 86, 89), OIA, Pali ciri 'cricket'.

15. Balti cha 'millet' (Sprigg 2002: 41), Ladakhi (Shamskat dialect) cha 'a variety of millet, a cereal grain which was grown in Ladakh in the past, but very little now' (Norman 2010: 290). Cf. Burushaski cha (Nagir, Hunza), ca (Yasin) 'Hirse, Kolbenhirse, Setaria italica' (Berger 1998b: 95).

16. Balti, Purik chal 'overflow, spill over', Ladakhi (Leh and Shamskat dialects) chal-ces 'to splash, to spill over' (Sprigg 2002: 41; Norman 2010: 297). Cf. OIA ksarati, ksalati 'flows, trickles', ksalayati 'washes', Proto-Iranian *xsar- 'to flow'19 (> Persian saridan 'to trickle', (ab)sar 'waterfall', Ossetic äxsärdzän 'waterfall' (< *xsar-cana-), Middle Persian Xsart 'the river Jaxartes' (Livshits 2003)), Kashmiri chalun 'to wash', char 'a sprinkle of water etc. from the fingers'), Gawar-Bati char 'rapids in a stream', Phalura, Indus Kohistani char, Bashkarik, Kalasha uchar, Shina char 'waterfall'.

17. Balti chon 'vain' (Sprigg 2002: 43), Purik chon ("tj^on") 'useless, in vain, unfounded' (Zemp 2018: 109), Ladakhi chon 'spontaneously, for no reason, gratuitously; in vain, for nothing, to no avail, useless; free, for free, at no cost (Norman 2010: 310). Jäschke 1881: chon W. useless, to no purpose. Cf. Burushaski chan (Hunza), can (Yasin) 'leer (Schüssel), unterbeschäftigt, ohne Arbeit, frei' (Berger 1998b: 106), Shina chon, Phalura cheniko, Bashkarik chan, Kashmiri chon, Gawar-Bati chenika, Pashai (Kurangali dialect) chani 'empty, void'. R. L. Turner (1966: 189) compares the Dardic words with the reconstructed OIA verb *ksanati 'is sifted, is strained, falls'. Semantically, however, this comparison does not seem convincing. The vowel o in the Ladakhi, Balti and Purik forms does not indicate that the source of borrowing must necessarily

16 This is not the only example of this change in our material. On the same process, see s.v. baan and basanda.

17 For transcribing Nothwestern Tibetan words the standard Tibetological transcription is used with c and ch conveying voiceless palatal and voiceless palatal aspirated affricate respectively, and ts and tsh standing for their dental counterparts. Dardic and Iranian material is transcribed in another way: c, ch mark palatal affricates and c, ch are used for dental ones.

18 In H. Berger's notation c and ch stand for Burushaski voiceless palatal and voiceless palatal aspirated affricate respectively.

19 On this Iranian root see e.g. Rastorgueva,Edelman 2007.

have been the Shina or Kashmiri word. It may be the result of an independent phonological change in the donor language (cf. the change a > o before n in the above-cited Balti word tshon 'injury').

18. Balti (Skardu dialect) chum 'bunch; muscle of thigh' (Norman 2010: 318). Cf. Bu-rushaski chu 'ear (of corn), bunch (of grapes)' (Lorimer 1938: 101), Indus Kohistani cui 'a bunch of grapes or other fruits' (Zoller 2005: 166).

19. Balti, Purik, Ladakhi culi 'apricot'. Jäschke 1881: cu-li, co-li W. 'a fresh apricot'. Cf. Burushaski ju, Shina joroiti, zuri, Khowar zu\i, Gawar-Bati zizori, Kashmiri cer, Kati cird, Askun cirä, Prasun cire, Wakhi cwan, Ishkashimi cbwend, Munji ciray 'apricot'.

20. Ladakhi darak, Purik deraq 'stiff, hard' (Norman 2010: 438). Cf. Kashmiri dor 'firm, hard, strong, compact, durable, solid', darun 'to become steady (of something in motion), to become firm, to stand steady' (Grierson 1915-1932: 238, 246), Pashai d(h)ar-, Wotapuri dar-, Phalura dhaara- 'to remain, stay', Torwali deri 'they remained', Khowar dorik 'hold back, wait, keep', Hindi, Punjabi dharna 'to keep', Gujarati dharvü 'to hold, catch', OIA dharayati, dharati 'holds, keeps', Avestan daraiiehi '(you) hold'. The prototype of the Ladakhi and Purik word in the donor language may be reconstructed as *daraka- (< PII *dharaka-). It is not quite clear if Balti dalaq 'stiff, hard (of leather, cloth)' (Sprigg 2002: 183) could have been borrowed from the same source.

21. Balti den 'sulphur' (Sprigg 2002: 183). Cf. Burushaski dancil 'Schwefel, findet sich in Schwefelquellen, die als heilkräftig gelten', Shina dancal id. (Berger 1998b: 114). The Shina word may be a loan from Burushaski. The final element -cil in Burushaski is probably etymol-ogically identical to chil 'water'. The original meaning of the Burushaski word may thus have been 'water of a sulphur spring'.

22. Balti diangs 'hard (of soft things that have become hard, bricks, food, etc.)' (Sprigg 2002: 183). Cf. Burushaski day 'hard' (Lorimer 1938: 112), Khowar day id.

23. Ladakhi (l)dim, Purik (dim 'trunk (of a tree), stem (of a plant)' (Norman 2010: 473), Balti dim(s) 'trunk (of a tree)'(Sprigg 2002: 183). Jäschke 1881: drim 'stump, trunk, of a tree or plant deprived of top and branches' Ld. Cf. Burushaski -dim 'Körper, Person, Selbst, (Nagir) Halm (des Getreides)' (Berger 1998b: 132), Shina dim 'body, trunk of tree' ([Bailey 1924), Indus Kohistani dim 'body (of a living being)' (Zoller 2005: 223), Khowar dim, OIA dimba- 'body'.

24. Balti (Skardu dialect) dom 'sadness, trouble, difficulty, adversity' (Norman 2010: 493). Cf. Shina damizhar 'adversity, trouble' (Bailey 1924), Burushaski dam(i)jar 'trouble, inconvenience, worry' (Lorimer 1938). The Shina word is most probably the source for the Burushaski one, and may be related to OIA damayati 'tames, subdues', Ossetic domyn 'to tame; exhaust; demand'.

25. Balti duhum, duhung 'large hole in a field, old tree, among rocks' (Sprigg 2002: 183). Cf. Bashkarik don 'hole' (Morgenstierne 1940), Indus Kohistani dung 'a pit, hole (in the earth), depression, hollow' (Zoller 2005: 223).

26. Ladakhi Galcik, galcik, galtsik, rgalcik, (G)alcilik 'lizard' (Norman 2010: 214).20 Jäschke 1881: rgag-cig a large grey species of lizard Ld. Cf. Burushaski garqas, galqas 'Eidechse' (Berger 1998b: 172), galgu (Hunza, Nagir), galgo (Yasin) 'Würmchen, Larve, Made, Raupe' (Berger 1998b: 167). Also cf. the following word.

27. Ladakhi galto, galto 'centipede (Shamskat dialect), a big type of spider (Leh dialect)' (Norman 2010: 214). Cf. Burushaski galtas 'centipede' (Lorimer 1938: 159), Shina galaac, Indus Kohistani galas, Phalura kalighas id. (Zoller 2005: 141).

28. Balti, Purik gat 'knot, joint of body' (Sprigg 2002: 58; Zemp 2018: 64), Ladakhi (Leh dialect) changgat 'knee-joint' (Norman 2010: 313). Cf. OIA granthi- 'knot', granthayati 'ties',

20 In Norman's transcription G is used to denote voiced uvular fricative.

Hindi, Bengali, Marathi, Gujarati gath, Punjabi, Lahnda gandh, Sindhi g'andhi, Middle Persian grih, Persian girih, Khotanese grratha, Soghdian yr'ns, Ishkashimi yurex, Ossetic xlxync', Kalasha, Phalura grhend, Indus Kohistani gär, Brokskat gathi 'knot', Pashai gatana 'joint', gath-'to tie', Burushaski gat 'Knoten (auch im Stengel von Pflanzen), Knöchel (des Fingers)' (Berger 1998b: 150). Ladakhi gat 'obstacle, tripping up' (Norman 2010: 241) may also belong here.

29. Purik gato 'half-roofed daytime pen for cows, horses, etc.', Ladakhi (Western subdia-lects of Shamskat gato 'cattle yard or pen, corral' (Norman 2010: 159). Originally 'a place where cattle are collected'? Cf. Burushaski gnti 'assembled, collected, together' (Lorimer 1938: 164), Shina gati 'together'.

30. Balti gra-a, Purik gra, 'goitre' (Sprigg 2002: 68; Norman 2010: 247). Cf. Shina gaaro, Burushaski gaaro id. (Berger 1998b: 164). It is not quite clear if the Shina and Burushaski words are connected in any way with OIA gadu- 'excrescence on neck, goitre, hump on back', Khowar gurüli, Kashmiri gadur 'goitre'.

31. Balti gzar 'to flow' (Sprigg 2002: 72), Purik zar id., Ladakhi zar-ces, dzar-cas 'to drip, to run down, to trickle out' (Norman 2010: 790). Cf. Av. yzar- 'to flow', Ossetic gzxlyn 'to pour down, drip', OIA jhara- 'waterfall', jhari 'river', Prakrit jharai 'drips', Hindi jharna 'to ooze, trickle away' < PII *gjhar- < PIE *dhgwher-21 (Cheung 2007: 124).

32. Purik yut, Ladakhi gut 'deaf, a deaf person; feeble, weakened', Balti gut 'mute' (Norman 2010: 215). Cf. Burushaski yut 'taub', yot (Yasin) 'stumm' (Berger 1998b: 182), Shina kütu 'deaf' (Bailey 1924).

33. Balti hal 'goal (sports)' (Norman 2010: 1064; Sprigg 2002: 74). Cf. Burushaski hala (Hunza, Nagir), hal, halu (Yasin) 'Tor (beim Polo)', Shina hal, Khowarp ha\ id. (Berger 1998b: 187).

34. Balti hal 'strength', halcan 'fine, strong, well-fed, healthy (animals)' (Sprigg 2002: 74), Ladakhi halcan 'robust, strong, healthy' (Norman 2010: 1058). Jäschke 1881: hal-med W. weak. Cf. Burushaski halel, Shina halal 'Fett' (Berger 1998b: 187).

35. Balti handam 'deaf and dumb' (Sprigg 2002: 74), Purik handay 'deaf, dumb' (Zemp 2018: 72), Ladakhi handang, handang 'stunned, speechless; mute, unable to speak; mentally retarded, mentally disabled' (Norman 2010: 1057). Jäschke 1881: han-ldang W. 1. dumb, mute. — 2. imbecile, weak of mind. Cf. Burushaski hnn dnm 'simple and stupid, honest and simple' (Lorimer 1938: 193), handam 'einfaltig, treuherzig-dumm, Einfaltspinsel', Shina handam id. (Berger 1998b: 189).

36. Ladakhi (Shamskat dialect) harip 'a musical instrument with an oboe-like double reed, used for non-religious music' (Norman 2010: 1055), Balti harip 'special kind of music played in the palace courtyard before a polo match' (Sprigg 2002: 75). Jäschke 1881: ha-rib Ld. music. Cf. Burushaski hdrip 'music, tune' (Lorimer 1938), Shina harip 'Sina music' (Bailey 1924), Indus Kohistani hnriph 'a tune, melody' (Zoller 2005: 409).

37. Balti haritham 'small centipede' (Sprigg 2002: 75). Jäschke 1881: ha-ri-tsam Pur. centipede. Cf. Burushaski hiriman (Hunza, Nagir), hirmen (Yasin) 'Insekt, dem Tausendfüssler ähnlich' (Berger 1998b: 200). The etymological connection between the Balti and Burushaski words is highly probable, although their exact prototype is not quite clear.

38. Balti kadik 'small branches' (Sprigg 2002: 82), Ladakhi katik 'branches and leaves of trees as fodder for animals in the spring' (Norman 2010: 4). Cf. Kati kat 'branch', Kalasha kat 'board' (Trail, Cooper 1999), Shina kat 'wood', Kashmiri kath 'wood', kdth 'small stick', Lahnda, Punjabi, Hindi-Urdu, Kumauni, Nepali, Gujarati kath, Sindhi kathi 'wood', gp.-MHg. kastha-'piece of wood', kasthika- 'small piece of wood'.

21 PII *gjhar- and PII *ksar- (> OIA ksar-, Proto-Iranian *xsar-) may reflect two different variants of the same Proto-Indo-European root (Rastorgueva, Edelman 2007).

39. Ladakhi, Purik kale 'revenge' (Norman 2010: 79; Zemp 2018: 940]. Cf. Kashmiri kal 'longing, regret, yearning; worry, worrying, anxiety; thought, consideration, calculation' (Grier-son 1915-1932: 433), Shina, Burushaski kaalan 'Erwägung, Berücksichtigung' (Berger 1998b: 237), Shina kalyoiki 'count, number, consider as abuse' (Bailey 1924: 146), Indus Kohistani kalav 'to count' (Zoller 2005: 107-108), OIA kalayati 'counts, thinks'. For the semantic change 'to count' > 'to revenge' cf. the English idiom to square accounts or the verb to reckon in the meaning 'to settle scores', cf. also Russian rasschitat's'a (рассчитаться) 'to retaliate' with the same root as schitat' (считать) 'to count'. It cannot be ruled out that the immediate source of Ladakhi and Purik kale in the donor language was semantically influenced by some cognate of Shina kali 'fighting' (Bailey 1924: 146), Indus Kohistani knyp 'a fight; a war' (Zoller 2005: 101) and OIA kali- 'strife, quarrel'.

40. Purik, Ladakhi kares, karas 'a local type of pea with three corners' (Norman 2010: 6). Cf. Burushaski (Hunza, Nagir, Yasin) gark 'Erbsen', yaras 'Lathyrus sativus Linn., eine Saatplatterbse' (Berger 1998b: 148, 170), OIA kalaya- 'a sort of pea', Nepali kerau 'pea', Hindi-Urdu keraw, kirau ' Pisum arvense', Kashmiri kari 'the pea Pisum sativum'.22 Ladakhi kerze 'lentils' (Norman 2010: 16) may also belong here. The final element -ze in this word is probably ety-mologically identical with the Classical Tibetan suffix -ze found, e.g. in bram-ze 'Brahmin' (< OIA brahmana-), zhim-ze 'sweetmeats' (cf. zhim-pa 'well-tasted, sweet-scented').

41. Purik karkati, Ladakhi (Western subdialects of Shamskat dialect) khaskhati, 'ant, ants' (Norman 2010: 111). Cf. Burushaski (Nagir, Hunza) karaato 'Art Insekt' (Berger 1998b: 242), kdrknnns 'a kind of spider (which has a big body and long legs and bites)' (Lorimer 1938: 229), Shina karaato 'Art grosse giftige Spinne' (Berger 1998b: 242), Khowar kroyunu 'large biting bug'. A contamination of two phonetically similar but etymologically different words may have taken place in the donor language.

42. Balti kelak, kyalak 'twisted' (Sprigg 2002: 82). Cf. Tirahi koold, Pashai kola, Shumashti kolanta, Khowar koli, Bashkarik kol, Torwali kol, Phalura kuulo, Shina kölu 'curved, crooked' < Proto-Dardic, PII *kaula- (> Proto-Iranian *kaura- > Khotanese kura- 'crooked' (Bailey 1979: 62)). The prototype of the Balti word in the donor language may have sounded *kaulak or *kolak and reflected an old formation with the -k- suffix. The irregular development of the first syllable vowel is probably due to the influence of some derivative of Tibetan 'khyil-ba 'to wind, to twist'.

43. Balti (Skardu) khila, khela, Purik khila 'sandals' (Norman 2010: 155). Cf. Indus Kohistani kharpa 'wooden sandals' (Zoller 2005:121), Kashmiri khrav, Punjabi, Hindi-Urdu kharau, Nepali kharau id. < *kasthapaduka- (Turner 1966: 159).

44. Balti (Skardu dialect) khudu, (Khaplu dialect) khutu 'small room', Ladakhi (Shamskat and Nubra dialects) khutu 'hut, cottage; thatched roof' (Norman 2010: 112). Cf. OIA kuti- 'hut', Khowar kutu 'a small room, a cabin, an apartment' (Sloan 1981: 102), Kalasha kutu 'temporary small shelter near one's fields' (Trail, Cooper 1999: 179), Indus Kohistani kurki 'a small house or hut close to a mosque in which the Maulvi and his pupils live' (Zoller 2005: 113), Shina gut 'tent' (Bailey 1924: 142), Burushaski (Hunza, Nagir) guti, (Yasin) kutu. 'Hütte' (Berger 1998b: 163), Wakhi kdta, kuta 'hut, shelter, small house' (Steblin-Kamensky 1999: 220). Cf. also the reflexes of Proto-Iranian *kata- 'house', for which the contamination with the aforementioned root seems to be possible: Wakhi ktic, Ishkashimi k(b)ric, Munji kric(a) 'a hut on the summer pasture' (Steblin-Kamensky 1999: 212). The ultimate source of the Wakhi, Dardic and Tibetan words may well have been Burushaski, though for the OIA kuti- a Dravidian origin was also suggested (Mayrhofer 1992: 362).

22 It is unclear if the Ladakhi word can have anything to do with some Dardic words for millet: Gawar-Bati karaz, Savi karaz, Kalasha karas.

45. Balti khut 'short (speech, etc.); small (area)', kot ci, kot ci 'a short while' (Sprigg 2002: 91, 93). Cf. Burushaski khut, Shina khuto, Khowar iskurdi 'short'.

46. Ladakhi kit-ces 'to catch, seize, capture' (Norman 2010: 82). Since the initial k in Central Ladakhi dialects (including Leh) may reflect earlier g (Zemp 2006), the verbal root kit- may be compared with Bashkarik gita 'took, bought', Wotapuri gat 'took', Gawar-Bati gutim 'I took' < *grpta- < *grb-ta- (cf. Vedic grbhnati 'takes, catches', Avestan gdrdfinaiti 'gains, obtains', Vedic grbdha-, Avestan garapta- 'taken'). According to R.L. Turner, in some Middle and New Indo-Aryan languages the past participle reconstructed as *ghropta- has replaced the regular grbdha-(Turner 1966: 244). In certain cases, the reflex of this form began to function as the new present, future or infinitive base (cf. Prakrit fut. ghattissam, inf. ghittum, ghettum, Oriya ghitiba 'to take, carry', Konkani ghetta 'takes').

47. Balti kodos 'rounded stone, cobble-stone' (Sprigg 2002: 93). Cf. Burushaski guyos 'grinding-stone' (Lorimer 1938: 176). In Burushaski y is a voiced retroflex spirant.23 It regularly reflects intervocalic d in Indo-Iranian loanwords (Berger 1998a: 22), and thus may have developed from an earlier retroflex stop.

48. Balti kulak 'meal (quickly made mixture of buttermilk and flour)' (Sprigg 2002: 94), Purik kholak 'a certain dish', Ladakhi (Leh, Shamskat and Nubra dialects) kholak, qholak 'ready-to-eat dough of roasted flour' (Norman 2010: 117). Jaschke 1881: Lh.: dumpling made of rtsam-pa and beer; Ld.: pap of rtsam-pa and tea, called spags in Central Tibet. Cf. Burushaski (d)-q(h)ul-an- (Hunza, Nagir), d-xul-an- (Yasin) '(Teig) kneten' (Berger 1998b: 357). The initial din Burushaski is a verbal prefix and -an- is a fossilized verbal suffix. For details see (Berger 1998a: 107-110, 212).

49. Ladakhi, Purik kulik 'lock; key (also pekulik); the joint of the jaw' (Norman 2010: 12). Jaschke 1881: ku-lig key, lock; also more accurately phe-kulig key, cug-kulik lock, padlock. W. If the meaning 'key' is historically primary, the word may be compared with PII *kaula-'crooked' and its Dardic and Iranian reflexes (see s.v. kelak). For semantic development cf. Greek kAv[lq, Latin clavis, Old Church Slavic kl'ucb (xAWHb) 'key' < PIE *kleu- 'Haken, krummes Holz' (Pokorny 1959: 604-605).

50. Ladakhi kuruk 'donkey foal; foal' (Norman 2010: 12), Balti bong-kuru 'donkey colt' (Sprigg 2002: 30). Jaschke 1881: gu-rug Ld. colt or foal of an ass. Cf. Kalasha kuak (< *kurak) 'child (male or female); the offspring of a human or animal' (Trail, Cooper 1999), Shumashti kur, Dameli kufa 'child', Ashkun kurd 'child, foetus', Kati kiu, kufuk 'young of animals', Prasun kyuru 'young of animals, child', Kurdish kurr 'son', Middle Persian kurrak, Persian kurra 'foal' < Proto-Aryan *kur-. For more details on this Aryan root, its reflexes, and its cognates in other branches of Indo-European see (Forssman 1980; Hegedus 2002). For Dardic (and perhaps for Nuristani) forms with retroflex r and its reflexes, the most probable prototype seems to be *kurta- with the -t- suffix.24 The meaning 'donkey foal' in Ladakhi might have developed under the semantic influence of a previously existing cognate of Eastern (Kham) Tibetan ku-ru 'donkey' (Norman 2010: 12).

51. Balti lashi 'a resinous wood used as a candle or torch because it burns slowly' (Sprigg 2002: 98). Cf. Kashmiri lash 'a torch' (Grierson 1915-1932: 533), Shina (Gilgit dialect) lai 'torch (unlit)' (Bailey 1924). In Shina the change *s > i is regular in the intervocalic position.

52. Ladakhi ldok 'muddy place', Purik ldoq 'mud, liquid dirt' (Norman 2010: 487). Cf. Burushaski toq 'Schlamm', Shina tok, Khowar toq, tuq id. (Berger 1998b: 429). The initial cluster ld may have developed due to the analogical influence of ldam 'mud'.

23 According to Berger, this phoneme can be described as "ein stimmhafter retroflexer Sibilant mit gleichzeitiger palatal-dorsaler Engebildung" (Berger 1998a: 22).

24 On the derivatives of this root with the suffixes -n-, -k-, and -t- see the above-cited paper (Forssman 1980).

53. Balti matu 'coolie's t-shaped stick' (Sprigg 2002: 112), Ladakhi matu 'walking stick; a tool for smoothing soil and breaking clods' (Norman 2010: 681). With vowel metathesis < *muta? Cf. Dameli muth, Gawar-Bati mutha, Kalasha, Phalura mut 'tree'.

54. Balti menze 'lump of dough' (Sprigg 2002: 113), Ladakhi menze 'ball of dough ready to be shaped and cooked, flattened dough ball, or shaped bread ready to be cooked' (Norman 2010: 710). Cf. Burushaski maano 'grösserer Teigklumpen' (Berger 1998b: 272), Shina mino id., Brokskat manili, Dameli man 'bread', Bashkarik man, Phalura mandili 'very soft bread' (Morgen-stierne 1940; 1941), Phalura mand- 'to knead', Romany ma(n)ro 'bread', Sindhi mani 'bread, loaf, food', Punjabi manda 'a thick cake', West Pahari (Bhalesi) manni 'a large cake', Assamese mar-'to knead (dough)', Maithili mar- 'to knead', OIA mrodnati 'crushes, kneads, rubs', Av. mördnddn '(they) destroy'. On the final element -ze in Balti and Ladakhi words see the note on kerze 'lentils'.

55. Balti monthok 'clod of earth' (Sprigg 2002: 118). With dissimilation of middle consonants < *mat-thok? Cf. OIA mrttika- 'earth, clay', Hindi-Urdu, Punjabi mitti, Nepali mato, Dameli mathi id., Phalura methi, Shina mati 'clay'; Burushaski thoko 'Klumpen, Ballen (von Zucker, Salz, Butter)' (Berger 1998b: 452), Indus Kohistani thokh 'a clod (earth, salt)' (Zoller 2005: 220).

56. Balti mulak, mulu, Purik mulaq 'turnip' (Sprigg 2002: 118; Zemp 2018: 944). Cf. Burushaski mulo, Shina muulo id., OIA mula- 'root', mulaka- 'radish', Shina müli, Khowar mü\ 'root', Pashai muluk, Hindi-Urdu muli 'radish'.

57. Balti, Purik, Ladakhi multuk, multuk 'fist' (Norman 2010: 705; Sprigg 2002: 118). Jäschke 1881: mul-t'ug W. fist. Cf. OIA musti-, Av. musti-, Khotanese mustu, Sindhi muthi, Lah-nda, Punjabi mutth, Hindi-Urdu, Gujarati, Marathi muth, Nepali muthi, Persian most, Shughni mut, Wakhi mast, Shina mut(h), Kashmiri moth, Phalura, Kalasha, Khowar musti, Gawar-Bati mustak, Burushaski (loanword) muc 'fist'. Intervocalic -lt- and -lt- in Western Tibetan dialects seem to convey st and st in borrowings (cf. Ladakhi maltar (along with mastar and mastar) 'teacher' (Norman 2010: 693) < Hindi-Urdu mastar < English master).

58. Balti mush 'squeegee type of rake' (Sprigg 2002: 119). Cf. Burushaki maus (Nagir), mausk (Yasin) 'zahnloser Rechen für Erde und Dung' (Berger 1998b: 284), Khowar maus 'wooden hoe' (Strand 2006).

59. Balti muyu, myo 'mouse', munzhur 'small mole-like mouse' (Sprigg 2002: 119). Cf. OIA mus-, musa-, musika- 'mouse, rat', Persian mus, Pashto mazak, Ossetic myst, Pashai muc, Shu-mashti muso, Gawar-Bati musa, Torwali mus, Phalura muso 'mouse', Bashkarik mus 'mouse, rat', Shina (Guresi dialect) müzu 'rat', müzai 'mouse', West Pahari (Jaunsari) musa, Romany muso 'mouse', Kumauni, Nepali muso 'mouse, rat'.

60. Balti paghon, phraghon, Purik phoGon 'spoon' (Sprigg 2002: 127, 130; Norman 2010: 602). Cf. Burushaski khapun, Khowar khipini, Shina khapai, Indus Kohistani khapay 'spoon, ladle' (Lori-mer 1938: 228; Zoller 2005: 123). In Balti and Purik a metathesis of stops must have taken place.

61. Ladakhi pat, paddi, patsi 'totally, completely' (Norman 2010: 533). The word is attested in the Leh dialect where root-initial p regularly reflects earlier b. The prototype may thus be reconstructed as *bad and compared with OIA baddha- 'bound', Gujarati badhü 'whole, entire', West Pahari (Kotgarhi) baddho 'all, entire, (pl.) all together', Kashmiri bod 'handful'.

62. Balti, Purik payu 'salt' (Sprigg 2002: 127; Norman 2010: 558). Jäschke 1881: pa-yu salt Bal. Cf. Burushaski (Hunza, Nagir, Yasin) bayu, Brokskat payu, Shina pazuu id. The intervocalic z in Shina probably appeared due to the influence of the verb paaz- 'to cook, bake' (Berger 1998b: 45).

63. Ladakhi perak 'Ladakhi women's head-dress, covered with turquoise and coral' (Norman 2010: 539). Cf. OIA paridha- 'to put on (clothes)', Sindhi paharanu, Nepali pairanu, Hindi pahirna 'to put on, wear', Kashmiri pdrun 'to put on; to adorn, ornament', Khowar pur-

duik 'to cover oneself, put on a cloak'. The source of the Ladakhi word in the donor language probably reflects PII *paridhaka-.

64. Balti phada 'bald' (Sprigg 2002: 128). Cf. Brokskat phata, Shina (Drasi dialect) phararo, Burushaski (Hunza, Yasin) bata id. According to H. Berger (Berger 1998b: 44), the word is connected with Burushaski bat 'skin'.

65. Balti phading 'dried apricots' (Sprigg 2002: 127), Purik phading, phating id., Ladakhi phating, phating 'apricot of good quality, with sweet kernel' (Norman 2010: 563). Jäschke 1881: p'a-tin, W., sweet dried apricots. Cf. Burushaski bnter 'dried apricot (split and stoned)' (Lorimer 1938: 73), batér (Hunza, Nagir), batôr (Yasin) 'aufgeschnittene und in der Sonne gedörrte Frucht, bes. Aprikosen' (Berger 1998b: 44), Shina phatôor id. Berger compares this word with OIA sphatayati 'splits', but cf. Burushaski phaqis 'entkernte, noch nicht gedorrte Aprikose' (Berger 1998b: 323).

66. Ladakhi phok 'incense or burning juniper leaves or other fragrant burning materials' (Norman 2010: 580). Cf. OIA pavayati 'purifies', Persian pak 'pure' (< *pavaka-). The source of the Ladakhi word is probably a certain reflex of PII *pavaka- 'purifier'. The origin of the initial aspirate remains unclear.

67. Ladakhi phololing 'a local variety of wild mint' (Norman 2010: 579), Purik phopholiy 'wild type of mint' (Zemp 2018: 47). Jäschke 1881: pho-lo-lin W. peppermint. Cf. Burushaski filnl 'mint' (Lorimer 1938: 155), phalâl (Yasin) 'Pfefferminz' (Berger 1998b: 329), Shina philiil id., Indus Kohistani phimili 'a kind of mint' (Zoller 2005: 291-292).

68. Ladakhi poze 'ram, full-grown male sheep' (Norman 2010: 540). On the final element -ze see the note on kerze 'lentils'. The element po- may reflect PII *pasu- 'cattle' (> OIA pasu-, Avestan pasu- id., Pashto psd, Ossetic fis 'sheep'). Cf. also Pashai paswala, Khowar pazâl, Shina päyälu, Bashkarik payal, Kashmiri puhul 'shepherd', Indus Kohistani pnyal, OIA pasupala-'herdsman'.

69. Ladakhi puli, polo 'Ladakhi biscuits of a particular type' (Norman 2010: 541). Cf. OIA pura- 'cake', pauli- 'a cake of scorched grain and ghee', Sindhi, Punjabi, Hindi, Kumauni purl, Gujarati, Marathi purl 'fried cake', Kashmiri pur 'a kind of cake fried in ghee'.

70. Ladakhi (Shamskat dialect) put 'germination, sprouting' (Norman 2010: 538). Cf. OIA sphutati 'bursts open', Lahnda phuttan 'to sprout', Panjabi phuttna, Hindi phutna 'to burst', Kashmiri phutun 'to be cracked', Shina phutôîki 'to break', Phalura phüuta 'to break (apart), crack' (Liljegren, Haider 2011: 121). The loss of initial consonant aspiration in the Ladakhi word remains unclear.

71. Balti (Skardu dialect) rat 'field, ground, level place' (Norman 2010: 911), 'flat, level' (Sprigg 2002: 135), Purik rat 'level' (Zemp 2018: 939). Cf. Burushaski (Hunza, Nagir, Yasin) rat 'flach, eben, glatt' (Berger 1998b: 364).

72. Balti rindi 'lead, bullet' (Sprigg 2002: 139), Ladakhi rindi 'bullet; lead (metal)' (Norman 2010: 911). Jäschke 1881: rin-di W. 1. lead. — 2. musket-ball. Cf. Burushaski ril 'copper' (Lorimer 1938: 303), Bashkarik rld, Torwali zit (z < r) 'brass', Shina rll 'brass, bronze, copper', Gawar-Bati rlt 'copper', OIA rlti- 'stream; yellow brass, bell-metal'. This word is probably etymologically connected with OIA rl- 'to flow, melt' (< PII *rl-). The original meaning may have been '(metal used for) casting'. The semantic as well as phonological development of the word seems to be influenced by Tibetan ra-nye 'lead'.

73. Balti rkat 'to cut down with a sword' (Sprigg 2002: 140). Cf. PII *kart- 'to cut' > OIA kmtati, kartati 'cuts', Av. kdrdntaiti 'cuts', kardta- 'knife', Hindi-Urdu katta 'curved knife', kattl 'sword, knife, dagger', Tirahi katari 'knife', Pashai katare 'spear', Gawar-Bati kataro 'large knife', Kalasha katar, Khowar kuter 'knife, dagger', Bashkarik kater 'knife', Savi katarei, Phalura katöro 'dagger', Shina khätaru 'knife'.

74. Ladakhi sale 'knitting needle' (Norman 2010: 1005). Cf. Burushaski sel 'Nadel, Stecknadel' (Berger 1998b: 377). May be etymologically connected with Phalura sileni 'needle' and OIA sivyati 'sews'.

75. Balti, Purik samdal 'moustache', Ladakhi samdal 'beard, moustache, whiskers (e.g. of a cat or otter)'. (Zemp 2018: 938; Norman 2010: 1007). Jaschke 1881: sam-dal Ld. mustaches. Cf. Burushaski (Yasin) samlatin, (Hunza) salat 'Schnurrbart' (Berger 1998b: 373), Khowar, Kalasha samlat 'moustache'. Balti, Purik, Ladakhi samdal < *samtal < *samlat with consonant metathesis and subsequent voicing of t after a nasal?

76. Purik sarsing, Ladakhi sarsing, sartsing, sasting 'oleaster, Russian olive, Eleagnus an-gustifolia, a tree with extremely fragrant clusters of small yellow flowers, edible but unpleasant olive-shaped fruit, and silvery leaves that stay on the tree into winter' (Norman 2010: 1007). Cf. Kalasha stein 'Eleagnus' (Trail, Cooper 1999: 271), Khowar sinjur, Wakhi sisk, Shughni sizd, Munji sijia, Tajiki sinjid, Persian senjed id. The word seems to be a local Wanderwort. It is, however, remarkable that in the Tibetan-speaking area it is found only in Northwestern dialects. The source of the Purik and Ladakhi words may have been a Dardic form that was pho-nologically similar to the Kalasha one.

77. Balti, Purik, Ladakhi (Leh) shagaran 'polo ground, playing field' (Norman 2010: 969). Cf. Burushaki sabaran (Hunza), sawaran (Nagir, Yasin), Shina shavamn id.

78. Balti shang 'wisdom, sense' (Sprigg 2002: 151), Purik say 'consciousness' (Zemp 2018: 931), Ladakhi shang 'alertness, awareness, caution, prudence' (Norman 2010: 553). The word should be separated from Old and Classical Tibetan spyang-ba, spyang-po 'skill; skillful, clever' because the Old Tibetan initial cluster spy- is preserved in Balti and Purik, and changes to c- in Ladakhi (cf. Balti spyang-mo 'agile' (Sprigg 2002: 159), Purik spyaymo 'agile, attentive' (Zemp 2018: 928), Ladakhi changpo (=cang-po) 'alert, clever, sharp-witted' (Abdul Hamid 1998)). But cf. Burushaski say 'awake, aware; care, heed, attention' (Lorimer 1938: 322), Shina son 'care, anxiety; awake, alert' (Bailey 1924), Khowar sang 'fear, suspicion' (Morgenstierne 1973), OIA sankate 'is afraid, distrusts', sanka- 'fear, distrust'.

79. Ladakhi shanti 'a leafy vegetable' (Norman 2010: 975). Cf. Indus Kohistani sari 'a green vegetable with round leaves' (Zoller 2005: 378). The word seems to reflect some derivative of PII *saka- 'green vegetable' (> OIA saka-, Shina, Indus Kohistani sa, Phalura so, Bashkarik sa, Kashmiri hakh, Kalsha sak, Khowar sax id.).

80. Ladakhi shen (Leh dialect), sen (Shamskat dialect) '(wooden) floor' (Norman 2010: 986). Jaschke 1881: shen, sren floor of a house or room W. The word may reflect PII *srayana- 'leaning, foothold' (< *sri- 'to lean', cf. OIA sri-, Avestan sri- id.). Cf. Indo-Aryan, Dardic and Iranian words for 'ladder, stair' reflecting various derivatives of the same root: Sinhalese hini, HindiUrdu, Punjabi sirhi, Phalura suri, Pashai sur, Shumashti sindi, Gawar-Bati sidi, Yidgha xad, Pashto sal.

81. Ladakhi shen-ces 'to squash' (Norman 2010: 986). Cf. OIA srnati 'crushes, breaks', Khowar sentk 'to crush'.

82. Balti (Skardu dialect) spa 'taste, tasty (neutral), (sexual) enjoyment', Purik, Ladakhi (Nubra dialect) spa 'taste, flavour' (Norman 2010: 542; Sprigg 2002: 159). Jaschke 1881: dpa'(-ba) 1. bravery, strength, courage; brave, strong, courageous. 2. beauty; beautiful. 3. W. taste, agreeable taste, flavour. Contrary to Jaschke, West Tibetan words can hardly reflect Old Tibetan dpa'(-ba). Such a derivation is problematic not only for semantic but also for phonological reasons, because the regular reflexes of the Old Tibetan cluster dp in Balti and Purik are xp and sp respectively (cf. Balti xpera 'talk, conversation' < Old Tibetan dpe-sgra; Purik spa(w)o, spaho 'warrior' (Norman 2010: 542) < Old Tibetan dpa'-bo). But cf. OIA svada- 'taste', Shina ispavu 'tasty'.

83. Purik su 'nose' (used as a synonym of snamtshul id. < Old Tibetan sna-mtshul) (Zemp 2018: 938). Cf. Burushaski sü etAs 'to smell (v.t.)' (Lorimer 1938: 332), Shina su thoiki, Indus Kohistani su karnv id. (Bailey 1924; Zoller 2005: 391).

84. Ladakhi sutun '(a child's or baby's) trousers' (Norman 2010: 1012). Cf. Waigali so(n)ta, Tirahi sdtän, Pashai sutan, Phal. suthan 'a kind of trousers', Indus Kohistani sutny 'loose cotton trousers, shalwar' (Zoller 2005: 402), Sindhi suthana 'pantaloons, drawers', Lahnda sutthan 'women's drawers', Punjabi sutthan, sutthun 'women's loose trousers', Hindi süthan 'trousers', West Pahari (Kotgarhi) sutthdn 'pair of trousers', Niya Prakrit somstani ' trousers (?)'.

85. Balti (Skardu dialect) syar, Purik, Ladakhi sar 'wick' (Norman 2010: 1007). Jäschke 1881: sar wick W. Cf. Burushaski (Hunza, Nagir, Yasin) sar 'gedrehter Wollfaden' (Berger 1998b: 375). Berger compares this word with OIA sarika 'string of pearls'.

86. Balti takalu, Ladakhi takari, takari 'scales, balance' (Norman 2010: 368; Sprigg 2002: 162). Jäschke 1881: ta-ka-ri common scales, Ld. Cf. Shina cakae, S. trakiri, Lahnda trakkan, Punjabi takkari, Hindi-Urdu tarakri 'pair of scales', Burushaski (loanword) jakaai 'Waage' (Berger 1998b: 233). This word is etymologically connected with OIA tarka- 'inquiry, conjecture', tarkayati 'guesses, thinks', Proto-Iranian *tark- > Ossetic txrxon 'judgement, discussion, verdict, trial' (Abaev 1979: 275-276).

87. Purik taki, taki 'flat bread' (Zemp 2018: 934), Ladakhi taki, tagi tagi 'bread' (Norman 2010: 368-369). Jäschke 1881: ta-gir W. bread. Cf. Shina, Khowar tiki id., Phalura tiki 'small, thick cakes of bread' Liljegren, Haider 2011), Kalasha tiki 'bread made by cooking in oil' (Trail, Cooper 1999), Lahnda tikki 'food', Sindhi tiki, Gujarati tiki, Marathi tikli 'cake', Punjabi tikki, Hindi tikiya ' small cake'.

88. Ladakhi tak-tak (Shamskat dialect), tak-tak, (Leh dialect) 'taut, stretched tight, tight (e.g. of curtain, clothing, greenhouse plastic)', Purik taqtaq 'tight' (Norman 2010: 370), Balti tak-tak 'hard' (Sprigg 2002: 184). Reduplicated adjectives are a well-attested phenomenon in modern Tibetan dialects (cf. Balti, Ladakhi kor-kor 'concave, round', Classical Tibetan kor 'root denoting anything round or concave' (Jäschke 1881: 6); Lhasa chung-chung 'small', Classical Tibetan chung-ba id.). Northwestern Tibetan final k and q usually reflect earlier final g (Miller 1956). The earlier prototype for the root of the words under analysis can thus be reconstructed as *tag. This root may be borrowed from some Indo-Iranian source, where it reflects either PII *tank- (> OIA tanc- 'to contract, put together, coagulate, solidify', Persian tanjidan 'to squeeze; to twist, roll together; to draw tight', tang 'narrow, tight', Balochi tanc- 'to fasten strongly, roll strongly, squeeze', Shughni taz-/tizd- 'to pull, draw, haul, drag, stretch'25) or PII *tangh-(> Proto-Iranian *&anj- 'to pull, draw' > Avestan §anj- id., Persian sanjidan 'to measure; reflect; compare, put in balance', Ossetic tinjin 'to spread, stretch out; to crucify' (Cheung 2007: 391392)). The latter root may have also reflected in Dardic (cf. Kashmiri tanz 'extreme and urgent desire',26 Khowar tonjeik 'to destroy, pull down'). Absence of the nasal in the Tibetan forms may be due to its regular drop before an old voiced guttural in the donor language. A similar process is attested in Shina: agar 'fire' (cf. OIA angara- 'coal'), agui 'finger' (cf. OIA anguli- id.).

89. Balti tangka, dangga 'shelter for the night, room boarded all round' (Sprigg 2002: 183, 184), Purik tangka 'partially roofed daytime pen for animals; open or partially roofed courtyard on the upstairs of a house' (Norman 2010: 359). Cf. Burushaski danoo 'oberirdischer Speicher, mit Tür zum Haus' (Berger 1998b: 131), Shina danö 'native storehouse' (Bailey 1924). But cf. also Tibetan dang-ra 'stable for cattle' (Jäschke 1881).

25 More Iranian reflexes of this root see in (Cheung 2007: 377-378).

26 Semantically cf. Russian t'aga (тяга) 'pull, traction; attraction, craving, strong desire'.

90. Balti tek 'button' (Sprigg 2002: 184). Cf. Burushaski tak, Shina täk, Indus Kohistani tiku, Khowar, Kalasha tak id.

91. Balti thing 'upright, straight up' (Sprigg 2002: 184). Cf. Burushaski tiy 'peak (of mountain), top (of hill); stretched up, erected', Khowar thing 'high place; steep slope' (Lorimer 1938: 352).

92. Balti thup 'darkness' (Sprigg 2002: 166), Purik thup 'dark' (Norman 2010: 435). Cf. Burushaski thap 'night'; Shina (Chilasi dialect), Indus Kohistani thaph 'dark; extinguished (light, fire)' (Zoller 2005: 113). Balti thap-thup 'dawn, dusk, twilight' (Sprigg 2002: 184) may also belong here.

93. Balti tok 'swelling, bump on the head, small hill' (Sprigg 2002: 185). Cf. Burushaski thoko 'Klumpen, Ballen (von Zucker, Salz, Butter)' (Berger 1998b: 452), Shina thokü 'hill' (Bailey 1924), Indus Kohistani thokh 'a clod (earth, salt)' (Zoller 2005: 220).

94. Ladakhi tsapik 'a little, a bit, a little while' (Norman 2010: 731). Jäschke 1881: tsa-big, tsha-big Ld. a little. Cf. Shina cap-, Indus Kohistani cap-, Gawar-Bati cep- 'to bite', Pashai cip-'to bite off', Kashmiri cop 'a bite'. The Dardic forms probably reflect the Proto-Indo-Iranian root *kap-/cap- 'to catch, snatch, pick, pinch'. For the Iranian reflexes of this root see e.g. (Ras-torgueva, Edelman 2003: 221-226). For the semantic development in Ladakhi cf. English bit (etymologically = bite). The final element -ik is a reduced form of the numeral cik 'one'.

95. Ladakhi tsele '(jocular) small, short (esp. of people).; (Shamskat dialect) a small variety of tick' (Norman 2010: 735). Cf. Burushaski chilum, pl. chiliiko, chiliiko 'fein gemahlen, fein (Pulver, Mehl usw.); feinkörnig; klein, winzig (Schriftzüge), in kleinen Buchstaben geschrieben; minderwertig, weniger gut' (Berger 1998b: 77).

96. Ladakhi tsiri 'the small intestine' (Norman 2010: 733-734). Cf. Burushaski -chir (Hunza, Nagar), -cerin (Yasin) 'Darm, pl. Eingeweide' (Berger 1998b: 78), Kalasha c(i)raki, tsiraki 'large intestines' (Trail, Cooper 1999: 63).

97. Ladakhi tshele 'thatched roof, shade made of branches and leaves, thatched-roofed house or hut' (Norman 2010: 773), Balti tshele 'hut' (Sprigg 2002: 170). Cf. Kashmiri chey 'a kind of thatching grass' (Grierson 1915-1932: 1066), Lahnda, Punjabi chatt, Hindi chat 'roof', OIA chadayati 'covers', chadis- 'cover, roof', Proto-Iranian *sad- 'to cover' (cf. Pashto psolal 'to adorn' < *upa- or *pati-sad-) < PIE *sk'ed-. The source of the Ladakhi and Balti words may have been the derivative with the suffix *-l- (*chadila- < *scadila-).

98. Ladakhi tshitu, tshitu 'contamination, causing dirtiness or violation in a way that will cause illness or spiritual pollution; menstruation' (Norman 2010: 766), Purik tshe(r)tu 'dirty' (Zemp 2018: 932). Cf. Kashmiri chetun 'to become impure, to become unclean, to become refuse and untouchable', chetarun 'to render impure, to cause something previously pure to become impure, to defile', cheth 'remnants of food, leavings of a meal' (Grierson 1915-1932: 1063, 1064). Shina usuto (=usuto?) 'foul, unclean' and Burushaski zutu (Hunza, Nagir), zetu (Yasin) 'unrein durch Pollution, Menstruation, oder wenn man nach dem Koitus nicht gebadet hat' (Berger 1998b: 486) with unexplained voicing of the initial consonant may also belong here. Turner's comparison of the Kashmiri and Shina forms with OIA ucchista- 'left over, stale, spat out; leavings' (Turner 1966: 85) doesn't seem impeccable because the OIA cluster st regularly corresponds to the aspirated th in Kashmiri.

99. Balti tshon 'injury' (Sprigg 2002: 171). Cf. OIA ksanoti 'injures, hurts', ksata- 'wounded', ksataka- 'wound', ksanana-, ksati- 'injury, damage', Pali khanati 'destroys', Khotanese vasanaurau 'destructive' (Bailey 1979: 379) < *vi-sana-bara-, Greek ktelvw 'I kill' < PIE *tk'en- (LIV: 645).27

27 In "Lexikon der indogermanischen Verben" the reconstructions *tken- and *tk'en- are considered equally possible, but the single non-geminate intervocalic s in Khotanese indicates higher probability of a prototype with the PIE palatal *k'. The prototype with the plain dorsal *k, *tken- (or *kpen- in earlier reconstructions), would have

Old Tibetan mtshon 'pointed or cutting instrument', proposed by Sprigg as the source of the Balti word, is semantically more distant. To the same Indo-Iranian etymon may also belong Balti tshak 'rheumatism' (Sprigg 2002: 169), Purik tshaq 'pain with difficulty of breathing', tshak yong 'to ache', Ladakhi tshak 'sprain, pulled muscle, sudden cramp, sudden sharp pain' (Norman 2010: 759-760).28 Cf. Kashmiri chokh 'wound', Khowar cay 'illness' (= OIA ksati-?), chek 'illness, pain'.

100. Ladakhi tso 'standing on the hind legs (e.g. khi tso coduk 'the dog is standing on its hind legs')' (Norman 2010: 735). Cf. Burushaski, Shina coko 'senkrecht, ansteigend, aufrecht, steil' (Berger 1998b: 90), Indus Kohistani cukh 'vertical, upright, erect' (Zoller 2005: 168).

101. Ladakhi tsolo 'anus' (Norman 2010: 735). Jaschke 1881: tsho-lo W. vulgar = 'phong-tshos 'buttocks'. Cf. Kashmiri coth 'anus, vulva', Indus Kohistani coth 'buttock' (Zoller 2005: 181), Lahnda, Punjabi cuttar, Hindi cutar 'rump, buttock'. The intervocalic l in Ladakhi may reflect the retroflex r or d (< *t, cf. the Indus Kohistani form) of the donor language.

102. Ladakhi tul 'powdered dung' (Norman 2010: 448). Norman proposes the earlier prototype *dul. Since in the dialects of Upper Ladakh, where this word is attested, one of the sources of the initial t- is d-, Norman's proposal sounds reasonable. The prototype *dul seems to have cognates in Indo-Aryan and Dardic: OIA dhuli- 'dust, powder', dhulika- 'pollen, fog, mist'; Prakrit dhuli, Hindi dhul, Punjabi, Lahnda, Gujarati dhur, Sindhi dhuri, Kumauni dhuli, Bengali dhul, Marathi dhu\ 'dust'; Nepali dhulo 'dust, powder'; Tirahi "duda", Kalasha udhru, Shina udu, Phalura duri 'dust'; Pashai (Wegali dialect) duri 'dust-storm'; Torwali dur 'mist'.

103. Balti tul 'to patch, mend (by sewing)' (Sprigg 2002: 172), Purik tul 'to stitch a thick or hard object with many stitches', Ladakhi (Shamskat, Leh and Nubra) tur-ces 'to darn, mend by stitching' (Norman 2010: 374). Jascke 1881: tur-ba W 'to darn (stockings)'. Cf. Burushaski (Hunza, Nagir) tol 'awl (cobbler's)', cuk etas tol 'a sewing awl' (Lorimer 1938: 355), Burushaski (Yasin) tul 'Ahle' (Berger 1998b: 428).

104. Balti tyaku 'branch used as a hockey stick' (Sprigg 2002: 185), Ladakhi taku, teku, taku 'polo mallet, hooked stick' (Norman 2010: 368). Jaschke 1881: ta-ku W stick with a hook, hooked cane, crutch. Cf. Burushaski (Yasin) daku 'Spazierstock' (Berger 1998b: 129), Burushaski (Hunza, Nagir), Shina dako 'wooden post supporting roof, pillar, tent-pole' (Lorimer 1938: 108).

105. Balti zanzos, dzanzos (Turtuk dialect), Purik zanzos 'wife; family' (Norman 2010: 842; Sprigg 2002: 180). Norman derives this word from Persian zan 'woman'. Her etymology cannot be discarded because massive borrowing from Persian into Muslim Tibetan dialects (directly and through Urdu) is an established fact. This process took place at a comparatively recent date and was mainly mediated through written language. That being said, it seems somewhat strange that the word in question is preserved only as a part of a fossilized compound, the second component of which cannot be identified with certainty. 29 It is, however, possible that the element (d)zan-, being etymologically related to the above-mentioned Persian lexeme, was adopted from some other Indo-Iranian source. Cf., e.g. Phalura jeeni 'female person' (Liljegren, Haider 2011: 76), Kashmiri zan', Sindhi, Lahnda, Punjabi jani 'woman', Bash-karik jin kar- 'to marry', Indus Kohistani zhal 'marriage' (Zoller 2005: 202), OIA jani-, Av. faini-'woman, wife'. The initial dz in Balti may point to the initial affricate of the donor language.

reflected as *xsan- in Proto-Iranian, and the regular Khotanese continuant of its derivative *vi-xsana-bara- would have sounded **vassanaurau.

28 Etymologically = OIA ksataka-?

29 Norman mentions the hypothetical variant dzantshos in the Turtuk dialect of Balti. If such a form really exists, the second part of the compound may be compared with Tibetan tshos 'color'. The original meaning of the word in this case should be something like 'female appearance'.

Data interpretation, conclusions and issues for future research

The above list contains a hundred-plus probable loanwords. They can be classified on the basis of their dialectal distribution, as well as of their etymological sources. If the former criterion is applied, the greater part of the list may be subdivided into two groups, i.e. lexical items found in the whole area and those peculiar to Muslim dialects (Balti and Purik). These groups include 35 (circa 33.3%) and 37 (circa 35.2%) words respectively. In addition, 11 words (circa 10.5%) are attested in Ladakhi and Purik only, and 22 (circa 21%) are peculiar to Ladakhi.

Etymologically, the list is far from homogeneous. Words of Burushaski and Indo-Iranian origin seem to be almost equally numerous. The delimitation of these two strata, however, poses a problem. Some Indo-Iranian languages were and still are in close contact with Burushaski. Since a considerable part of modern Indo-Iranian vocabulary remains unetymologized, the direction of borrowing is not clear in all cases. For this reason, at the current state of our knowledge the most accurate etymological classification should include the following classes:

1) words with parallels in Burushaski only;

2) words with both modern Indo-Iranian and Burushaski parallels, not etymologizable in Indo-Iranian (possible Burushaski loanwords in Indo-Iranian);

3) words with both modern Indo-Iranian and Burushaski parallels, etymologizable in Indo-Iranian (possible Indo-Iranian loanwords in Burushaski);

4) words with parallels in Indo-Iranian only.

These classes comprise 11 (circa 10.5%), 38 (circa 36.2%), 19 (circa 18%) and 37 (circa 35.2%) words on the list respectively.

The most intriguing conclusions can, however, be drawn if the two classifications are "superimposed" on each other, i.e. if we try to establish the distribution of loanwords belonging to the four above etymological classes, in different dialects. Out of 37 words, attested only in Balti and/or Purik, 6 belong to class 1, 15 belong to class 2, 7 belong to class 3, 9 belong to class 4. That is to say, the majority (circa 57%) of loanwords peculiar to Muslim dialects are unquestionable or probable borrowings form Burushaski. In the light of this fact, Zemp's claim that "... the deepest mark on Purik appears to have been left by Burushaski" (Zemp 2018: 4) should be considered realistic. A totally different picture emerges based on the etymological analysis of lexical items peculiar to Ladakhi. In this group the predominace of Indo-Iranian borrowings is quite evident, 16 words (circa 73%) belonging to classes 3 and 4. A somewhat similar distribution show the words, attested in Ladakhi and Purik only. Classes 3 and 4 cover 3 (circa 27%) and 4 (circa 36%) of such words respectively, their total number being 11. Among the borrowings common to all the analyzed dialects, i.e. Ladakhi and Muslim varieties, Bu-rushaski and Indo-Iranian loans are numerically almost equal, classes 1 and 2 comprising 18 words (circa 51%), classes 3 and 4 — 17 words (circa 49%) out of 35.

The situation just described looks somewhat confusing and needs some explanation. To explain it in full, copious additional lexical material on Ladakhi, Purik and Balti will, no doubt, be necessary. Nevertheless, some preliminary hypotheses may already be proposed. Dissimilarity in distribution of loanwords between Indo-Iranian and Burushaski strata in different Northwestern Tibetan dialects may well have been caused by local differences in ethnic and linguistic composition in the pre-Tibetan period. If this be the case, it should be assumed that the pre-Tibetan population of the present-day Kargil and Skardu districts spoke some form of Burushaski, while in Ladakh the predominant language was Indo-Iranian. This is, howerver, a somewhat simplistic scheme, and the true picture of language contact in the region seems to be more intricate. The route of Tibetan migration to the present-day Muslim areas must undoubtedly have passed through Ladakh. This fact explains why a large percentage

of the loanwords that are common for Ladakhi and Muslim dialects is of Indo-Iranian origin. On the other hand, the presence of numerous Burushaski loans in Ladakhi is also a certainty. It cannot be ruled out that some of them were not borrowed directly but rather came via Balti or Purik.30 Yet one should bear in mind that Indo-Iranian languages were not autochthonous in the region, even though they seem to have been spoken there before the spread of Tibetan. The pre-Indo-Iranian inhabitants of Ladakh may well have been speakers of some local variety of early Burushaski. It implies the possibility that the Burushaski elements in Ladakhi may, at least partly, have been adopted via some pre-Tibetan Indo-Iranian dialect from this earlier substrate language.

Another fact yet to be explained is the considerable number of Indo-Iranian borrowings attested in Muslim dialects only. 31 It seems quite possible that some of them are in fact present in Ladakhi, too, but for some reasons were not included in Norman's dictionary. A certain part of such words, however, may belong to dialects other than those of Skardu and Kargil and stem from their substrate languages, which could have been Indo-Iranian rather than Bu-rushaski. Unfortunately, this hypothesis can hardly be tested because of an extreme lack of data. The situation may change in the future, provided field linguistic research in the region makes more progress.

The above wordlist raises a number of issues for further research. For example, it still remains unclear whether Indo-Iranian words in it were borrowed from a single source or several sources. If there were more than one donor languages, were they all substrata for the Tibetan dialects or at least some of them were adstrata? Which branches of Indo-Iranian do they represent? All these questions will remain unanswered as long as the historical-phonological peculiarities of the Indo-Iranian lexical items on the list remain unclear. This problem will be addressed at length in a separate article. Here I will limit myself to several most conspicuous features.

Some loanwords in Ladakhi show devoicing of historical voiced stops (cf., e.g. kit-ces 'to catch, seize, capture' < PII *grbh-ta-; pat, paddi, patsi 'totally, completely' < PII *baddha- < *badh-ta-). This phonological change affected also the inherited vocabulary of several Ladakhi dialects including the dialect of Leh. Since in more cases Indo-Iranian word-initial voiced consonants are preserved, it may be hypothesized that the process of borrowing started before devoicing and finished after it had already taken place. A synchronically similar, but probably historically different phenomenon can be seen in loans form Burushaski, where voiceless and voiceless aspirated consonants sometimes correspond to voiced ones in the donor language: Balti kddos 'rounded stone, cobble-stone' (cf. Burushaski guyos 'grinding-stone'), Balti, Purik payu 'salt' (cf. Burushaski bayu), Balti phada 'bald' (cf. Burushaski bata), Balti tyaku 'branch used as a hockey stick', Ladakhi taku, teku, taku 'polo mallet, hooked stick' (cf. Burushaski daku 'Spazierstock'). As is evident from the above examples, this type of sound correspondence is not confined to Ladakhi but found also in Balti and Purik, where Old Tibetan root-initial voiced stops were in most cases not devoiced. The cause of this phenomenon is not quite clear, but most likely it has something to do with the fact that in Burushaski itself many roots with initial mediae have variants with aspirated and unaspirated tenues. These variants often appear after prefixes.

Another noteworthy historical-phonological feature of Indo-Iranian loanwords is the den-talization of old palatal affricates: Ladakhi tsapik 'a little, a bit, a little while' (cf. Shina cap-

30 In this regard, it is worth noting that a number of such words are attested only in the Shamskat and Nubra dialects, spoken in Northern Ladakh and staying in close contact with Purik and Balti.

31 They make up circa 43% of loanwords peculiar to Balti and Purik.

'to bite' (< PII *cap-),32 Balti (d)zanzos, Purik zanzos 'wife; family' (cf. OIA jani-, Av. faini- 'woman, wife'). This change, however, is not observed in some borrowings from Burushaski: Ladakhi c(h)ancil 'the green outer shell or fruit of walnut' (cf. Burushaski c(h)anjil), Balti, Ladakhi cha 'millet' (cf. Burushaski cha).

The central question arising in the context of Indo-Iranian loanwords in Northwestern Tibetan dialects is certainly that of the exact position of the donor language(s) in the Indo-Iranian subfamily. In the present article I confine myself to making just a few points in this connection. First of all, in the above material there is not a single example of typically Iranian phonological development. 33 The presence of phonological innovations peculiar to Indo-Aryan is highly questionable, the only possible example being perhaps the change of PIE *sk' to palatal aspirated affricate ch with subsequent dentalization (cf. Ladakhi, Balti tshele 'hut', etymologically connected with OIA chad- 'to cover' < PIE *sk'ed-). This example is, however, hardly indicative because the change *sk' > ch is attested in a number of Dardic languages where it is an independent development of Proto-Dardic *sc (Kogan 2005).

On the other hand, there are several reasons to believe that the souce language of at least some of the words on the list may have belonged to the Dardic group. In this regard, one fact is remarkable. In Indo-Iranian loanwords we find two different coorespondences to OIA cluster ks, i.e. tsh (voiceless dental aspirated affricate) and ch (voiceless palatal aspirated affricate). Cf., e.g. Balti tshon 'injury', tshak 'rheumatism', Purik tshaq, Ladakhi tshak 'sharp pain' (= OIA ksanoti 'injures, hurts', ksataka- 'wound'); Balti, Purik chal 'overflow, spill over', Ladakhi chal-ces 'to splash, to spill over' (= OIA ksarati, ksalati 'flows, trickles', ksalayati 'washes'). Iranian etymological parallels to these two words do not show uniformity of correspondences too (cf. Proto-Iranian *vi-sana-bara- 'destructive' > Khotanese vasanaurau; Proto-Iranian *xsar- 'to flow' > Ossetic axsardzan 'waterfall'). This picture clearly shows that we are dealing here with the re-flexation of two different old consonant groups. Iranian is known to preserve the distinction between Proto-Indo-European clusters containing palatal dorsal consonants (i.e. *k's, *tk' etc.) and those containg plain or labial dorsals (i.e. *ks, *gs, *kws, *gws and some others), the reflexes being *s and *xs respectively, while in Indo-Aryan the merger took place. Dardic in this respect is somewhat close but not identical to Iranian. Here the reflexes did not merge, and clusters of the two aforesaid types developed in two divergent ways, i.e. to palatal ch and to cerebral ch (Kogan 2005). 34 As we can see, this Dardic reflexation is not only very similar to the one detected in the above-metioned Nothwestern Tibetan words, but may well represent the earlier stage of the latter. The dental voiceless aspirate tsh may have evolved from palatal ch in the wake of dentalization, and the cerebral ch may have been decerebralized into palatal ch.35 Summing up, the system of reflexes of the Indo-European clusters under analysis in the two Indo-Iranian loanwords on our list, being totally different from both Indo-Aryan and Iranian, is at the same time derivable from the Proto-Dardic one. It can, howerver, be stated with certainty that the source of these loans cannot be any of the present-day Dardic languages spoken

32 See also other examples with initial ts and tsh in the previous section.

33 The only feature that is found in both Iranian and the loanwords on the above list is the deaspiration of Proto-Indo-European voiced aspirates. This sound change may, however, be attributed to the fact that voiced aspirated consonants are absent from the phonological systems of Ladakhi, Balti and Purik. In addition, it should be noted that deaspiration is characteristic not only of Iranian but also of Dardic and Nuristani.

34 The palatal aspirated affricate *ch is the Proto-Dardic reflex of clusters containing palatal dorsals, while the groups with non-palatal dorsals are reflected as the cerebal aspirated affricate ch. The latter phoneme may have developed from the earlier Proto-Dardic cluster *ks.

35 The exact typological parallel to this process is attested in the history of Kashmiri (Kogan 2016).

adjacent to Balti, Ladakhi and Purik, i.e. Brokskat, Shina and Kashmiri. All these languages belong to the Eastern Dardic subgroup, where the earlier ch and ch have merged word-initially (Kogan 2016).

Thus, historical phonology gives us some evidence in favor of Francke's hypothesis. Nonetheless, as noted above, this hypothesis is undoubtedly unacceptable in its original version and should be reformulated. The pre-Tibetan population of Ladakh may well have been linguistically Dardic but its language could in no way have been an early form of Brokskat. What this language really was is an interesting problem in itself. It is hoped that further study will throw some light on it.

Abbreviations for language names

Av — Avestan; OIA — Old Indo-Aryan; PIE — Proto-Indo-European; PII — Proto-Indo-Iranian

References

Abaev, Vasily I. 1979. Istoriko-etimologicheskij slovar' osetinskogo jazyka. Tom III. Leningrad: Nauka. Abaev, Vasily I. 1989. Istoriko-etimologicheskij slovar' osetinskogo jazyka. Tom IV. Leningrad: Nauka. Abdul Hamid. 1998. Ladakhi-English-Urdu dictionary with an English-Ladakhi index. Leh: Melong Publications. Bailey, T. Grahame. 1924. Grammar of the Shina Language ("Royal Asiatic Society, Prize Publication Fund", vol. VIII).

London: Royal Asiatic Society. Bailey, Harold W. 1979. Dictionary of Khotan Saka. Cambridge University Press.

Berger, Hermann. 1998a. Die Burushaski-Sprache von Hunza und Nager. Teil I. Grammatik. Wiesbaden: Harrassowitz Verlag.

Berger, Hermann. 1998b. Die Burushaski-Sprache von Hunza und Nager. Teil III. Wörterbuch Burushaski-Deutsch,

Deutsch-Burushaski. Unter Mitarbeit von Nasiruddin Hunzai. Wiesbaden: Harrassowitz Verlag. Cheung, Johnny. 2007. Etymological Dictionary of the Iranian Verb. Leiden Indo-European etymological dictionary series. Leiden / Boston: Brill.

Forssman, Bernhard. 1980. Hethitisch kurka-Comm. 'Fohlen'. Zeitschrift für vergleichende Sprachforschung 94(1/2): 70-74.

Francke, August H. 1905. The eighteen songs of the Bono-na festival (Dard text, with translation, notes and vocabulary). The Indian Antiquary XXXIV: 93-110. Francke, August H. 1906. The Dards at Khalatse in Western Tibet. Memoirs of the Asiatic Society of Bengal 1(19): 413-419. Francke, August H. 1907. A history of Western Tibet: one of the unknown empires. London: S. W. Partridge & Co. Grierson, George A. 1915-32. A Dictionary of the Kashmiri Language. Vol. I-IV ("Bibliotheca Indica". Work No. 229).

Calcutta: Royal Asiatic Society of Bengal. Grierson, George A. 1919. Linguistic Survey of India. Vol. VIII, pt. 2. Specimens of the Dardic or Pisaca Languages (including Käshmiri). Calcutta: Superintendent Goverment Printing, India. Hegedus, Iren. 2002. Proto-Nuristani *kurka 'young of an animal'. In: Fabrice Cavoto (ed.). The Linguist's Linguist.

A Collection of Papers in Honour of Alexis Manaster Ramer. München: Lincom Europa. Jäschke, Heinrich A. 1881. A Tibetan-English dictionary with special reference to the prevailing dialects. London: Rout-ledge and Kegan Paul.

Jettmar, Karl. 1975. Die Religionen des Hindukusch (= Die Religionen der Menschheit. Bd. 4, 1). Mit Beiträgen von Schuyler Jones und Max Klimburg. Stuttgart: Kohlhammer Verlag. Kogan, Anton I. 2005. Dardskie yazyki. Geneticheskaya kharakteristika. Moskva: Vostochnaya literatura. Kogan, Anton I. 2016. Problemy sravnitel'no-istoricheskogo izuchenija jazyka kashmiri. Moskva: "Fond razvitija funda-

mental'nyx lingvisticheskix issledovanij". Kogan, Anton I., Mikhail E. Vasilyev. 2013. K voprosu o vostochnodardskoj jazykovoj obshchnosti. Journal of Language Relationship 10: 149-178. Liljegren, Henrik, Naseem Haider. 2011. Palula vocabulary. Islamabad: Forum for Language Initiatives.

Livshits, Vladimir A. 2003. Drevneje nazvanije Syrdar'ji. Vestnik drevnej istorii 1(244): 3-10.

Lorimer, David L. R. 1939. The Burushaski language. Vol. III. Vocabularies and index. Oslo: H. Aschehoug & Co. (W. Ny-gaard).

Mayrhofer, Manfred. 1992. Etymologisches Wörterbuch des Altindoarischen. I Band. Heidelberg: Carl Winter Universitätsverlag

Miller, Roy Andrew. 1956. Segmental diachronic phonology of a Ladakh (Tibetan) dialect. Zeitschrift der Deutschen Morgenländischen Gesellschaft 106(2): 345-362.

Morgenstierne, Georg. 1940. Notes on Bashkarik. Acta Orientalia XVIII(3/4): 206-257.

Morgenstierne, Georg. 1941. Notes on Phalura, an Unknown Dardic Language of Chitral. Oslo: J. Dybwad.

Morgenstierne, Georg. 1967. Indo-Iranian Frontier Languages. Vol. III. The Pashai Language. 1. Grammar. Oslo: Univer-sitetsforlaget.

Morgenstierne, Georg. 1973. Sanskritic Words in Khowar. In: Georg Morgenstierne. Irano-Dardica: 256-272. Wiesbaden: Dr Ludwig Reichert Verlag.

Norman, Rebecca. 2010. A Dictionary of the Language Spoken by Ladakhis. Ms. (unpublished draft).

Pokorny, Julius. 1959. Indogermanisches etymologisches Wörterbuch. Bern / München: Francke Verlag.

Rastorgueva, Vera S., Dzhoy I. Edelman. 2003. Etimologicheskij slovar' iranskix jazykov. T. 2 (b—d). Moskva: «Vostochnaja literatura» RAN.

Rastorgueva, Vera S., Dzhoy I. Edelman. 2007. Etimologicheskij slovar' iranskix jazykov. T. 3 (f—h). Moskva: «Vostochnaja literatura» RAN.

Shaw, R. B. 1878. Stray Aryans in Tibet. Journal of the Asiatic Society of Bengal XLVII/I: 26-62.

Sloan, Mohammad Ismail. 1981. Khowar-English dictionary. Peshawar: Printing Corporation of Frontier Ltd.

Sprigg, Richard K. 2002. Balti-English, English-Balti dictionary. New York / London: Routledge.

Steblin-Kamenskiy, Ivan M. 1999. Etimologicheskij slovar' vaxanskogo jazyka. Sankt-Petersburg: Peterburgskoje vostokovedenije.

Strand, Richard F. 2006. Khowar lexicon. Available online: https://nuristan.info/lngFrameL.html (accessed 27.11.2019).

Trail, Ronald L., Gregory R. Cooper. 1999. Kalasha Dictionary with English and Urdu compiled by Ronald L. Trail, Gregory R. Cooper. Islamabad: National Institute of Pakistan Studies, Quaidi-Azam University, Summer Institute of Linguistics.

Turner, Ralph L. 1966. A Comparative Dictionary of the Indo-Aryan Languages. Oxford University Press.

Zeissler, Bettina. 2004. Relative tense and aspectual values in Tibetan languages. A comparative study. Berlin / New York: Mouton de Gruyter.

Zeissler, Bettina. 2010. East of the moon and West of the sun? Approaches to a land with many names, North of Ancient India and South of Khotan. The Tibet Journal XXXIV (3/4) / XXXV (1/2): 371-463.

Zemp, Marius. 2006. Synchronic and diachronic phonology of the Tibetan dialect of Kargil. Lizentiatsarbeit im Fach Historisch Vergleichende Sprachwissenschaft. Bern.

Zemp, Marius. 2018. A grammar of Purik Tibetan. Leiden / Boston: Brill.

Zoller, Claus P. 2005. A Grammar and Dictionary of Indus Kohistani. Volume 1: Dictionary. Berlin / New York: Mouton de Gruyter.

А. И. Коган. О возможном влиянии дардских языков и языка бурушаски на северозападные диалекты тибетского языка.

Северо-западная окраина тибетского языкового ареала, ныне входящая в состав индийского штата Джамму и Кашмир и подконтрольных Пакистану Северных Территорий, в прошлом являлась зоной инстенсивных этнических и языковых контактов. Результатом этих контактов стала языковая ассимиляция тибетцами местного дотибетского населения. Согласно гипотезе, выдвинутой более века назад, это население, возможно, говорило на одном из дардских языков. В статье делается попытка проверки данной гипотезы путем этимологического анализа лексики северо-западных тибетских диалектов. Результаты анализа свидетельствуют о наличии значительного индоиранского (вероятно, дардского) лексического пласта, а также многочисленных лексических еди-

ниц, усвоенных из некой ранней формы языка бурушаски. Автор пытается выявить диалектное распределение заимствований из индоиранского источника и бурушаски в исследуемом ареале.

Ключевые слова: языковые контакты; языковой субстрат; тибетские диалекты; язык ла-дакхи; язык балти; язык пурик; дардские языки; язык бурушаски.