Tocharian and Indo-European Studies · tively plausible Indo-European etymologies. TA . kaś (TB ....

Tocharianand

Indo-EuropeanStudies

Founded by Jörundur Hilmarsson

Edited by

Birgit Anette Olsen (executive editor)

Michaël Peyrot · Georges-Jean Pinault

omas Olander (assistant editor)

volume 16 · 2015

Museum Tusculanum PressUniversity of Copenhagen

2015

© Museum Tusculanum Press and the author(s) 2015

Tocharian and Indo-European Studies, Vol. 16© Museum Tusculanum Press and the authors, 2015

issn 1012 9286isbN 978 87 635 4398 9

editorial advisory board

Douglas Q. Adams (Moscow, Idaho)Gerd Carling (Lund)Olav Hackstein (Munich)Jay Jasanoff (Harvard)Ronald Kim (Poznań)Frederik Kortlandt (Leiden)Jens Peter Laut (Göttingen)Melanie Malzahn (Vienna)H. Craig Melchert (Los Angeles)Donald Ringe (Pennsylvania)Guðrún Þórhallsdóttir (Reykjavík)

Museum Tusculanum PressBirketinget 6dk 2300 Copenhagen Swww.mtp.dk


Contents

douglas q. adamsRe-examining the tradition: e case of two hapax legomena in B-522 1

Václav Blažek & Michal SchwarzTocharian A kopräṅk ‘antelope, deer’ 9

Ilya ItkinPour la réinterprétation de quelques formes verbales tokhariennes A 17

Ronald I. KimAn explosive etymology 25

Frederik KortlandtTocharian ē-grade verb forms 51

Melanie Malzahn & Hannes A. FellnerLifting up the light: tläś and lkäś in Tocharian A 61

Ogihara HirotoshiKuchean secular documents in the Pelliot Koutchéen Nouvelle Série 81

Michaël PeyrotNotes on Tocharian glosses and colophons in Sanskrit

manuscripts II 107

Georges-Jean PinaultTocharian nostalgia 131

Michael Weisse rite stuff: Lat. rīte, rītus, TB rittetär, TA ritwatär, and

Av. raēθβa- 181

Werner WinterTocharische Metrik (1957) 199


Contentsiv

Fanny MeunierCompte rendu de Emil Sieg, Tocharologica: Selected writings on

Tocharian 211

Michaël PeyrotReview of Dieter Maue, Alttürkische Handschriften Teil 19:

Dokumente in Brāhmī und tibetischer Schrift Teil 2 215


Lifting up the light: tläś and lkäś in Tocharian A1

Melanie Malzahn & Hannes A. Fellner

is paper treats the meaning and origin of the two enigmatic Tocharian A words tläś and lkäś. Based on our new edition and translation of the fragment THT 318/319 we propose that TA tläś belongs to the root tälā- ‘to lift up, carry, bear’ and is a verbal noun meaning ‘the lifting up’. Simi-larly, relying on a Sanskrit parallel text and arguments from Tocharian derivational morphology we argue that TA lkäś is a verbal noun meaning ‘light’ which is built to the root luk- ‘light up, be illuminated’. We trace the formation of these two TA verbal nouns back to the PIE suffix *-nt-i-, comparable to the Hittite tukkanzi-type.

1 Tocharian A has a number of nominals in -ś. TG 51 lists kaś ‘number’, kārāś2 ‘jungle’ (DTA 115a), kuñaś3 ‘fight’ (DTA 148a), kuraś adj. ‘cold’, lkäś (in TG without meaning; see below), śoś ‘small cattle’. To this list can be added ākāś ‘ether, atmosphere’ (DTA 26a), āmāś ‘royal minis-ter’ (DTA 40a), kaliṅkasparś ‘touching of the Kāliṅga textile’ (DTA 105b), kroś ‘krośa; measure of distance’ (DTA 174bf.), kleś ‘affliction, depravity’ (DTA 179a), tläś (in TLT 135 without meaning; see below), and lepäś*, which only is referenced as an animal name in TLT 272.4

1 We would like to thank Georges-Jean Pinault, Olav Hackstein, Michaël Peyrot, Bernhard Koller, and Laura Grestenberger for discussion and valu-able comments on earlier drafts of this paper. e usual disclaimers apply.

2 In TG this word is listed with the outdated meaning ‘poverty’.3 In TG this word is listed without meaning. 4 Not included in this list are words with Sanskrit spelling, e.g., caturdaś, name

of 14th day in the lunar calendar, or proper names that are borrowed from Sanskrit, e.g., Kuś < Skt. Kuśa, name of several kings and heroes (MW 296c), or aviś < Skt. avīci- ‘waveless; Avīci-hell’, name of the lowest level of hell (DTA 17).


melanie malzahn & hannes a. fellner62

Morphologically, these nominals in -ś do not form a coherent class in Tocharian. Only for āmāś (nom. pl. -āñ, obl. pl. -ās), kaś (obl.sg. kaś, nom. pl. -añ), kārāś (obl. sg. kārāś, obl. pl. -äntu), ākāś (obl. sg. ākāś, loc. sg. -aṃ), and kleś (obl. sg. kleś, nom. pl. -āñ, obl. pl. -ās) inflected forms other than the identical nominative/oblique singular are attested.

2 Some of the nouns in (1) are straightforward loanwords. e follow-ing are borrowed from Sanskrit: ākāś (TB akāś ‘id.’) < Skt. ākāśa- ‘region, place’ (BHSD 87a), ‘Raum, Luftraum’ (SWTF I 222b); āmāś (TB amāc ‘id.’) < Skt. amātya- ‘companion of the king, minister’ (MW 81b); kaliṅkasparś < Skt. kaliṅgasparśa-* ‘touching of Kāliṅga’;5 kroś < Skt. krośa- ‘kos; In-dian league’ (MW 322b); kleś (TB kleś ‘id.’) < Skt. kleśa- ‘impurity, deprav-ity’ (BHSD 198a).

TA lepäś*, which means ‘jackal’ (Burlak and Itkin 2010: 354–356),6 seems to be a borrowing from Indo-Aryan as well, cf. Skt. lopāśa- ‘jackal, fox’ (Malzahn 2014: 92f.).

TA kārāś ‘jungle’ (presumably via TB karāśe* ‘id.’) is borrowed from Khot. karāśśä- ‘creeper, twig’ (DKS 54b).

3 Among the nouns in (1) that are not loans only two have compara-tively plausible Indo-European etymologies.

TA kaś (TB keś) ‘sequence, number, calculation’ is traced back to a preform *ḱos-ti- by Hilmarsson (1991: 159f.) who compares Skt. śaś-vant- ‘numerous’ (for which see Klingenschmitt 1975=2005: 149–157). e pho-nological development of the sequence *-sti- > TA -ś TB -ś(c-) is par-alleled by TA päś(ś-) TB paś(c-) ‘breast’ < PIE *psten- in the duals TA päśśäṃ TB paścane.7

5 See BHSD 181a for kāliṅga-. 6 is meaning can be supported by cemetery scenes from Qizil paintings (see

Howard and Vignato 2015: 109 on Qizil Caves 116 and 220 depicting jackals and vultures devouring corpses).

7 For TB cf. also maś(ce) ‘fist’ < PIE *mus-ti-, Skt. muṣṭí- ‘fist’. e form TB maś is attested in THT 597a5.


Lifting up the light: tläś and lkäś in Tocharian A 63

TA śoś ‘small cattle’ goes back to late-PIE *pḱēu̯-d-es according to Pinault (1997: 209–211) and thus is related to Lat. pecūs, pecūd- ‘head of cattle’.

e origin of kuñaś ‘fight’ and the hapax kuraś ‘cold’ are notoriously difficult and etymological interpretations so far remain unconvincing, cf. Hilmarsson (1996: 192f. and 194 with literature).

e two remaining words on the list in (1), tläś and lkäś, even lack an established meaning and will be treated next.

4 TLT 135 lists an adverb tläś without meaning attested once in THT 319a4 with the verb kätā- ‘disperse’. Sieg and Siegling in TochSprR(A) 174 read tläśtont instead. A second co-occurrence of tläś with the verb kätā- ‘disperse’ in the small fragment THT 3333b2 now makes the existence of this form certain since tläś is not followed by tont. While the preserved context of THT 3333b2 /// tläś knānträ-ṃ s· /// ‘… they disperse the tläś for him/her’ is insufficient to determine the meaning of tläś, things look much better concerning THT 319.

It turns out that THT 319 and fragment THT 318 can be joined to-gether into one leaf. e possibility of a joint has already been assumed by Sieg and Siegling in TochSprR(A): 174, but without establishing the details.8 e calculation below is based on the size of the original leaf (based on the almost complete leaf A 312) and on the metrical analysis; unfortunately, the original manuscript is missing.

Transcription

(bhi)-a1 -ṇḍipālyo kärsä(smāṃ9 mā ā)lsanträ ṣome yu(kneñc) /// (approx. 8

syll.) /// yp(e)ñc ce{ṣ} ṣome kumpa-kump klyanträ neṣinäs /// (approx. 10 akṣaras)

8 Siegling in his personal copy of TochSprR(A) also did not make an attempt at joining the two fragments.

9 According to the context, kärsä must be derived from a root kärs- ‘to shoot an arrow’ being the equivalent of TB kärsk- ‘id.’, also attested in TA pärra-krase ‘(distance of an) arrow shot’; the precise grammatical form is uncertain, but



a2 ceṣ penu ṣome kropa-krop ñäktaññ oki tsärk ts· /// (9 syll.) /// tuṅk arseñc || praṇāde träṅkäṣ ehe näṣ nu vidy(ādharñās śomi)-

a3 -nāśśi10 kräntsonās pikārās lkāmāṃ märso oki ñä- /// (8 syll.) /// kuyalte || maṃndhottarinaṃ || lukäśnunt wmāri ,11 mrācaṃ näm ; /// (7 syll.)

a4 ; pärkont tākeñc 1a klośnaṃ tarkañ , /// (7 syll.) /// , hāräsyo ; tläś tont knāṣ-äm 1b – r·äṃ wāmpunt , k· /// (7 syll.)

a5 -keṃ nam ; ṣokyo ciñcraṃ 1c yetunt tim me(m)12 , /// (3 syll.) /// · ; y(e)tunt wiraṃ kulewāñ wrokṣiṃ13 /// (3 syll.) 1d (12 syll.) (yi)-

a6 -ñc pre ṣontac 2a ṣomaṃ nu rpeñc , kispar wic ; ṣomaṃ tsärk /// (4 syll.) /// ; (kä)l(ns)eñc14 ṣulas 2b triwont ṣom(aṃ) /// (10 syll.)

a7 ; vidyādharñāñ 2c not timaṣiṃ , ciñcrone ; märsneñc oki , wire(ś)ś(i) ; /// (4 syll.) 2d /// (vidyādha)rñāñ prusaṃ15 yāmunt16 /// (10 syll.) 3a (2 syll.)

a8 -ṣ tsäṅkrā , (kuya)ll aśśi ; kātkmāṃ waṅkmāṃ , karemāṃ ; ṣmeñc pyāppyāsaṃ 3b /// (18 syll.) 3c (7 syll.)

b1 ; wäry(o) (ṣul)y(iṃ)17 , himavant ; cäñcär käln{ä}ṣ 3 – – – ryo , kä- /// (13 syll.) 4a (11 syll.) , (kä)-

a present VIII m-participle would make some sense, morphologically and syntactically.

10 e restoration (śomi)nāśśi is proposed by Siegling in his personal copy.11 e symbol < ; > marks colon end, < , > marks subcolon end, and the symbol

< # > marks pāda end in the transcription.12 For metrical reasons, me(m) ‘measure’ seems the only possible restoration,

maybe an obliquus of manner. Even though the reading of the preceding word nim or tim cannot be checked, the dual pronoun tim is the only syntac-tically possible solution.

13 e colon division of pāda 1d after the lacuna is unclear; note that this would still be true without accepting the joint.

14 (kä)l(ns)eñc is proposed by Siegling in his personal copy with two question marks, restoration to ṣom(aṃ) at the end of the line is proposed without question mark.

15 e hapax prusaṃ can be a locative sg. or pl. of a noun pru or prus, respec-tively.

16 e colon division is unclear.17 Since himavant is exclusively attested with ṣul ‘mountain’ or the adjective

ṣulyiṃ, we propose to restore (or correct the reading) yā – ·y· to wäry(o)



b2 -rtkālsaṃ ; śomināśśi 4b waṅkmāṃ wraṃn āñc , siñanträ ; nkä(ñci)18 mātä(r) , /// (7 syll.) 4c (3 syll.) /// kr·s , m(ā)trā tim ; päśśä(ṃ) /// (11 syll.) 4d (smi)-

b3 -mā(ṃ)19 akmalyo indre träṅkäṣ wäṣpā nu ñäkciṃ ṣuläṣ (himava)nt ṣul (kuya)lte kinnares vidyādha(res) ///

b4 wsäryās pyāppyāsyo ṣuliñi tsäṅkrunt wā(wluṣ20 jim)utaketuy vidyādhareśśi lā(nt) /// (kuya)-

b5 -lte toṣ tri-wäknā lāñci-waṣtantu /// (7 syll.) /// -ñcä(ṃ)n oki toṣ penu tri lāntsañ (tri)-w(ä)knā ñäkteñ(ñ)ā(ñ oki) /// (7 syll.)

b6 worpunt näṃ21 || citraśokaṃ || # ñäkci wäll oki säs ; vidyā(dhareśśi wäl) ; (jimutake)tu22 1a yetu ñemintuyo ; worpu antiṣpuryo ; triko ⸗ki /// (2 syll.) 1b (5 syll.)

b7 -ryo ; lāntsas āntiṣpurṣās ; skenaṣ kātkässi 1c p·i23 /// (9 syll.) /// lāntsac ; lyutār potaträ 1 || some träṅkäṣ sās ·i (– – – –)

(ṣul)y(iṃ) with the instrumental wäryo. Note that the following 3sg. kälnaṣ for kälnäṣ is certainly a misspelling (thus Hilmarsson 1996: 64) – it is quite certain that the original manuscript indeed had the reading kälnaṣ, because Siegling, who still had access to the original, marked the form with a question mark in his personal copy.

18 nkä(ñci) ‘silver’ is the only known possible form, while restoration to mātä(r) is suggested by the perlative mātrā in the following line – most likely not a variant of mātār ‘sea monster’, but rather derived from the Sanskrit water plant mātar- ‘Asian Watermoss’ [Salvinia cucullata Roxb.].

19 e restoration to (smi)mā(ṃ) is proposed by Siegling in his personal copy.20 e restoration wā(wluṣ) is proposed by Siegling in his personal copy.21 e reduced form of the copula, which surfaces with pronominal suffixes in

näṃ and näm, is indifferent to number, as was shown by TG 167 and Burlak and Itkin (2009: 47). erefore, the agreement between näṃ and the plural participle worpunt is unproblematic. We thank Ilya Itkin (p.c) for pointing this issue out to us.

22 Restoration to (jimutake)tu is already proposed by Siegling in his personal copy.

23 One may restore to p(ñ)i(ntuyo) ‘with merits’.



b8 -r klyomiṃ lālaṃṣkā (pälska-pā)ṣe24 lyutār śkaṃ lā(laṃṣkā)25 /// /// (lāntsa)c lyutār potaträ kuyalte pälska-pāṣe wrasom ///

Translation

a1 … shoot(ing) with a (jave)lin they do (not) hold … in check. Some overcome …, [some] do … ese stand [together] in individual groups. e previous [ones] …

a2 ese [ones], single group by single group, also (make) [lute] music like gods, … [and] they evoke love. || Praṇāda says: “Oh! Now I, look-ing at the beautiful gestures of the

a3 [female] Vidyādharīs, as if forgetting … because of … || In [the tune] Mandodharinaṃ || Shining jewels are on top of their head[s] [i.e. of the Vidyādharīs] …

a4 they [fem.] will arise (in the air?). Earrings [lit. hanging on both ears] (are on them) … with pearls, … (wind?) spreads up the [adornment] put on them. e decorated …

a5 (make a) … bow towards the very lovely [women]. 1c. Both [ears] are adorned according to measure (with) …, the young women [are] adorned (with) … [made of] pearls. 1d.

a6 … they (go) out onto the street. 2a. Now some [women] play the kis-par wic, others [play] the lute, … they (let resound?) the mountains. Some [women] mixed (with) …

a7 the [female] Vidyādharīs. 2c. As if they forget the loveliness of both not, … of the young [masc.]. 2d. e [female] (Vidyādha)rīs having done … in the prus … 3a. …

24 Restoration to (pälska-pā)ṣe is suggested by what follows in the line; what is more, only this kind of verbal governing compounds can end in TA °ṣe.

25 omas (1967: 179) proposed to join the two fragments and to restore to lā(ntsa)c, but according to the calculation, there is a lacuna of at least 8–9 syllables between them. Siegling in his personal copy proposes to restore to lā(laṃṣkā). In total, about 20 akṣaras are missing in this line.



a8 over the mountain peak, why then are they sitting in the flowers being joyful, chatting [and] laughing? 3b. …

b1 (pools filled? with) water from the Himalaya (mountain), it resounds sweetly. 3d. With … 4a. …

b2 in the pools of the girls. 4b. Chatting down in the water they will be satiated, the sil(ver) watermoss, … over the watermoss the two breasts (are floating ?) … 4d.

b3 With a (laugh)ing face Indra says: “But indeed Mount Himalaya [is] the [most] divine mountain, because Kiṃnaras [and] Vidyādharas …

b4 With …, grassland [and] flowers the peaks of the mountains of Jimūtāketu, ki(ng) of the Vidyādharas [are] covered …

b5 (be)cause these royal palaces in three-fold [manner] … like … . Like-wise three queens in a (three-)fold [manner] (like) goddess(es) …

b6 are surrounded by … . || In [the tune] Citraśok || is one, (Jimūtāke)tu, (king of the) Vidyā(dharas) [is] like a divine king. 1a. Adorned with jewels, surrounded by [his] harem, as if confused … 1b. …

b7 With … he strives to please the queens of the harem. 1c. (With merits ?) … he pays more homage to the queens. || Soma says: “is [queen] …

b8 … [is] noble, tender [and] (regardful about the mi)nd and more te(nder) … pays more homage to the (queen) because a being [which is] regardful about the mind …”

e leaf contains a detailed description of a Vidyādhara procession, demigods and -goddesses who according to Indian tradition inhabit the Himalaya region (Liebert 1976: 336). In line a4 the adornment of the Vidyādharīs is described. Since pärkont ‘arise’ at the end of pāda 1a makes either an allusion to (part of) the decor or to the fact that Vidyādharas are themselves able to fly, the same image ought to be referred to in pāda 1b by the phrase tläś kätā-. e subject here may be the wind spreading the



adornments upwards. It may even precisely refer to garlands by means of which the Vidyādharas fly in the air as often depicted in Indian art.26

Syntactically, taking tläś as an adverb is indeed the only likely possible analysis (as already proposed by Poucha).27 e form itself patterns with the root TA tälā- ‘to lift up, carry, bear’ (TVS 659f.; *telh2- 2LIV 622f.),28 so that it is possible to envisage an oblique of a verbal noun ‘the lifting-up’, i.e. ‘upwards’. For adverbs made from oblique forms in TA, cf. for instance kātse ‘nearby’ (Pinault 2008: 486). A closer parallel for the use of the oblique of tläś as an adverb ‘upwards’ is provided by development of the ablative suffix TB -meṃ < CT *-mæn from the PIE acc.sg. *món-m̥ from a root noun *món-/men- ‘height’, i.e.*món-m̥ meaning ‘to the height, upwards’ (Pinault 2006: 277f.).

5 TA lkäś is glossed as ‘visum’, i.e. ‘appearance, sight vision’, in TLT 274. It occurs in A 249a2.29 e text is part of the Tocharian version of the Buddhastotra Varṇārhavarṇa by Mātṛceṭa (see Schmidt 1980, 1983, 1987, and Pinault 2008: 281–291) which is a close rendition of the San-skrit original (see Hartmann 1987, 1988, 2009, and Hartmann and Maue 1991). Light can be shed on the correct meaning of lkäś by comparing the relevant passages of the Sanskrit original and the Tocharian version. e Sanskrit text and its translation are based on Hartmann (1987).

Sanskrit text (VAV)

II.57 sati pradīpe saty agnau satsu tārāmaṇīnduṣu | apūrṇa evārkam ṛte prakāśo bhavati kṣitau ||

26 Cf. the Flying Vidyādhara couple from Gwalior, 5th century A.D, National Museum, New Delhi; see http://en.wikipedia.org/wiki/File:Vidyadhara.jpg (retrieved 30.03.2015).

27 Unless tläś can also be used as a plural form agreeing with the plural of the participle tont.

28 Note that Peyrot (2013: 756) argues that there are no certain base verb forms and that the verb seems to be causative only in TA.

29 e other possible occurrence in A 40a3 /// räṃ kākmart lkä /// is too frag-mentary to make a firm judgment.



Mag eine Lampe vorhanden sein, mag Feuer vorhanden sein, mögen Sterne, Juwelen und der Mond vorhanden sein, ohne die Sonne ist das Licht auf der Erde nur unvollständig.

II.58 tāsu cānyāsu cābhāsu tadviśiṣṭatamāsv api | satīṣv eva nirālokaṃ bhavati tvadṛte jagat ||

Selbst wenn es diese Lichtquellen gäbe und andere, die noch vor-züglicher sind als diese, ist die Welt lichtlos ohne dich.

II.59 ity avidyāndhakārograpaṭalāvṛtacetasaḥ | tamobhūtasya lokasya jyotirbh(ū)tāya te namaḥ ||

In diesem Gedanken (sei) dir Verehrung, der du zum Licht ge-worden bist für die verfinsterte Welt, deren Geist bedeckt ist von dem schrecklichen Schleier, (nämlich) der Blindheit der Unwis-senheit.

Tocharian text (A 249)

a1 (wi)nāsam neñc penu koṃ mañ ñäkcyāñ swāñcenāñ wāwlu nu säs ārkiśoṣi ākntsuneyo ptukk orkäm tñi

a2 (ka)pśiñño nu tri wältseṃ lyalyku knānmuneyo puk traidhātuk wināsam cu lkäś nāṃtsunt

a1 … I venerate … Although there are sun, moon [and] divine rays now this world is covered with ignorance [and] there is only(?) darkness.

a2 And with your body and your wisdom you have illuminated the three thousand-fold traidhātuka completely, I venerate you having become the light.

TA wināsam cu lkäś nāṃtsunt is the exact translation of Skt. jyotirbh(ū)tāya te namaḥ with lkäś corresponding to jyotiṣ- ‘light’ (cf. already the trans-lation ‘lumière’ in Pinault 2008: 288). us there is no doubt that lkäś means ‘light’ and belongs to the root TA luk- ‘light up, be illuminated’ (TVS 855f.; *leu̯k- 2LIV 418f.). is is further corroborated by the fact



that lkäś ‘light’ served as the derivational basis of the adjective TA lukäśnu ‘shining, luminous’30 and is related to TA l(y)ukśone ‘illumination’.31

6 Having established the correct meanings and synchronic patterning of tläś and lkäś as verbal nouns belonging to luk- and tälā-, respectively, it is now possible to address their origin. Phonologically, there are different potential sources of TA -ś:

1 e pre-TA (or rather underlying) sequence -äṣäc gives -ś as is illus-trated by the 2pl. pret. VIII wätkäṣäc > wätkäś (A 95a4, A 342a2).

2 Pre-Proto-Tocharian *-KV[+pal]-, e.g., TA aś- [TB eś-] in the dual TA aśäṃ [TB eśane] ‘eyes’ from PIE *h3okw-ih1.

3 Pre-Proto-Tocharian *-dV[+pal]-, e.g., TA śoś < late-PIE pl. *pḱēu̯-d-es (cf. Lat. pecūd-; Pinault 1997: 209–211).

4 Pre-Proto-Tocharian *-stV[+pal]-, e.g., TA päś(ś-) [TB paś(c-)] ‘breast’ < PIE *psten- in the dual TA päśśäṃ [TB paścane].

5 Pre-Proto-Tocharian *-ntV[+pal]- > TA -ñc > -ś, e.g., nom. pl. of TA wäl [TB walo] ‘king’, lāś < lāñc < *-nt-es.32

e first possibility (1) can be excluded right away since it has been estab-lished that tläś and lkäś are nominal forms.

While the development in (2) could be entertained for an equation of lkäś with Ved. rúci- ‘light’, tläś would then have to be explained differ-ently. Separating tläś and lkäś, however is uneconomic, since tläś and lkäś are morphologically and semantically similar.

Georges-Jean Pinault (p.c.) opts for (3) tracing lkäś back to an abstract noun pre-PT *luk-id-i- built to an adjective *luk-id-o- that could be re-

30 For adjectives in -nu in Tocharian A, cf. ākärnu ‘tearful’ ← ākär ‘tear’, orkämnu ‘dark, gloomy’ ← ‘darkness, gloom’. Note that lukäśnu has a variant lukśanu.

31 TA l(y)ukśone is probably reshaped for *luktsone, the exact match of TB läkutsauña ‘light’, after lukäśnu (Georges-Jean Pinault, p.c.). e attested forms of l(y)ukśone with palatalization are probably due to influence of the averbo of luk- (in which the preterite forms show ly-).

32 Cf. also TA nom. pl. m. kraś ‘good’ [TB kreś] (next to krañc [TB kreñc]) and nom. pl. m. arkaś ‘black’ [TB erkent-] (next to obl. pl. m. arkañcäs).



flected in TB lakutse ‘shining, bright, brilliant’. However, d-stems are a vanishingly small category in Tocharian and at most marginal in other Indo-European branches.33 In the case of a *-id-i- at least two deriva-tional steps would have to be assumed that cannot be corroborated inde-pendently in or outside Tocharian.

Option (4) presupposes secondary *ti-derivatives to *s-stems. is kind of formation is usually associated with Caland system adjectival ab-stracts of the shape *-os-ti- (Melchert 1999a: 365ff.) that were part of a derivational chain; e.g., *h2emǵhos-ti- ‘narrowness’ (OCS ǫzostь ‘id.’) ← *h2emǵhos-to- ‘narrow’ (Lat. angustus ‘id.’) ← *h2émǵho/es- ‘narrowness’ (Ved. áṁhas- ‘distress, constriction’, YAv. ązah- ‘constriction, distress’), and shows full grade both of the root and the s-suffix. Only *leu̯k- is as-sociated with the Caland system and if lkäś continued a secondary *to/i-derivative to an *s-stem it would go back to a formation with zero-grade of the root and the s-suffix.34 However, continuants of s-stems in Tochar-ian always have full grade of the root,35 which indicates that a secondary derivatitive with zero-grade has to be old.36 While this could well be true

33 For another possible reflex of a *d-stem in Tocharian besides TA śoś < late-PIE pl. *pḱēu̯-d-es, see TB kektseñe TA kapśañi ‘body’, see Pinault (1999).

34 According to Pinault (2009: 241), TB laks reflects a secondary s-stem de-rivative with zero-grade in the root and the s-suffix: TB laks < PT *läksä < *luk-s-i- ← *luk-s-o- (Ved. rukṣá- ‘shining, radiant’) ← *leu̯k-es- (Av. raocah- ‘light’), cf. Gk. λεῦκος name of a fish, Lat. lūcius ‘pike’, ON lýrr a kind of cod, for the semantics.

35 E.g., TB āke TA āk ‘end, tip’ < *h2eḱ-os- (cf. Lat. acus, -eris ‘husk of grain, chaff ’, OHG ahir, ehir ‘ear (of corn)’), TB (ñem)kälywe TA (ñom)klyu ‘(name and) fame’ < *ḱléu̯-os (cf. Ved. śrávas- ‘fame’, Gk. κλέος ‘id.’, OIr. clú ‘id.’ etc.), TB ṣalype TA ṣälyp ‘oil, ointment’ < *sélp-os (cf. Gk. (Hsych.) ἔλπος ‘oil, fat’, Alb. gjalpë ‘butter’); for other potential examples cf. Höfler (2012: 132–138 with discussion and literature).

36 It is possible that Tocharian reflects the cognate of Lith. akštìs ‘skewer, spit’, Russ. ostь ‘awn’ < *h2eḱ-s-ti- (: *h2eḱ-os- in TB āke TA āk ‘end, tip’) with full-grade of the root and zero-grade of the s-suffix in TB āśce ‘head’; see 2DoT 61f. with literature. For a potential phonological obstacle for this etymology see the next footnote.



for lkäś,37 it cannot be for tläś. e latter would then have to be an ana-logical creation on the basis of lkäś, which is not very likely given the dif-ferent morphological and semantic profiles of the roots luk- and tälā- and their predecessors.

A way of deriving tläś and lkäś from the same morphological fore-runner is to envisage the phonological development in (5). PIE had *i-stem abstracts that were derived from *nt-stems of the type found in Hittite tukkanzi- ‘cultivation, breeding’ (Melchert 1999b: 23) which are also reflected in Hittite nouns in -anzan-, e.g., laḫḫanzan- ‘shellduck’ < *laḫḫanti̯ō/on- ‘migrating one’ ← *laḫḫanti̯o- ‘migrating’ ← *laḫḫanti- ‘migration’ ← *laḫḫant- ‘migrating’ (Melchert 2003: 135) and presup-posed by the Latin type absentia ‘absence’ ← absēns ‘being away from’ (Nussbaum apud Melchert 1999: 23 n. 32; see Pinault 2014 for possible Vedic reflexes of the tukkanzi-type). As Pinault (2012: 189) shows, the latter type also occurs in Tocharian in the formations with TB -ntsa, TA -nts* (e.g. TB wapāntsa, TA wāpants*38 ‘weaver’ to TA/TB wāpā- ‘weave’ and TB sarāntsa ‘planter’ to TA/TB sār- ‘plant’. e semantics of tläś and lkäś are exactly as expected from an abstract built to a verbal adjective. It is thus not far-fetched to assume that tläś and lkäś also go back to an

Note that āśce has a nom. sg. -e for expected -ø and an obl. pl. āstäṃ with lack of palatalization, which is prima facie unexpected if it continues a *ti-stem. However, other alleged continuants of *ti-stems in Tocharian B behave in the same way, e.g., plāce ‘word, talk, speech’ with obl. sg. plāc, nom. pl. plāci, obl. pl. plātäṃ. If these forms really do continue *ti-stems, nom. sg. -e could have been introduced from the ubiquitous TB e-stems to specifically mark the nominative vis-à-vis the oblique in -c. e lack of palatalization in the oblique plural in these formations might be analogical to nouns having stem-final palatalization in the nominative plural, but lack of stem-final pala-talization in the oblique plural, e.g., lyak ‘thief ’ with nom. pl. lyśi and obl. pl. lyakäṃ, by the following proportion: nom. pl. -śi : nom. pl. -ci :: obl. pl. -käṃ : x, where x was solved as -täṃ.

37 However, a virtual *-k-s-ti- would probably have been affected by the change of the sequence pre-PT *-ks- > TA -ps-, cf. TA kapśañi [TB kektseñe] ‘body’, nom. pl. TA opsi [TB okso], TA klepsā- [TB klaiksā-] ‘dry up, wither’ (TVS 629); see Pinault (1999 and 2008: 49).

38 TA wāpants* is presupposed by the abstract wāpäṃtsune ‘(art of) weaving’.



*nt-i-type formation preserving the old zero-grade of the *nt-suffix in the weak stem since in PIE *nt-stems originally inflected according to the amphikinetic type R(é)-S(o)-E/R(ø)-S(ø)-E(é). A secondary *i-stem regularly built to the weak stem of the participle39 of *telh2- and *leu̯k-, *tl̥h2-n̥t-i- and *luk-n̥t-i-, would give PT *täläñcä40 and *läk(w)äñcä41 which with syncope would develop to pre-TA *tläncä and *lk(w)äncä42 and further to tläś and lkäś. e isolated forms tläś and lkäś would thus represent a morphological archaism in Tocharian A.

Appendix

It is now possible to turn to the two remaining words from the list of Tocharian A nominals in -ś (1), kuñaś ‘fight’ and kuraś ‘cold’, which also are isolated in form and function.

TA kuñaś is attested four times in A 238a3, A 353a5, A 375b5, and PK NS 1b1.43 Besides occurring three times as part of an idiom with the verb yām- / ya(p)- ‘do’ (TVS 788f.; TVS 783) in A 353a5, A 375b5, and A 238a3,

39 For other examples of a secondary derivative from the non-renewed weak stem of amphikinetic *nt-stems, cf. Goth. hulundi ‘cave’ < *ḱl̥ -n̥t-ih2 or Ved. satyá- ‘true, real’ < *h1s-n̥t-i̯o-.

40 e loss of the intervocalic laryngeal is regular.41 e form pre-TA *lkwäncä is seen in the derived adjective lukäśnu / lukśanu.

e labial element of the labiovelar was prone to fluctuation in this root, cf. TA 3sg. prs. lukuṣ / lukäṣ and the adjective TB lakutse / laktse ‘shining, bright, brilliant’. Furthermore, and are difficult to distinguish in Tocharian Brahmi, so that the form lkäś may well reflect expected lukäś.

42 e sequence pre-TA -äñc(-) does not turn into TA -iñc universally as is proved by forms like nkäñci adj. ‘silver’ (next to nkiñc subst. ‘silver’), cäñcär ‘pleasant, lovely’ (next to ciñcär), obl. sg. m. päñcäṃ ‘fifth’, and in the sub-junctive stem (class II), läñc-, of the root länt- ‘go out, emerge’.

43 In PK NS 1b1 the spelling kuñāś is found. is is an error by the scribe, who seems to have placed the virāma stroke a bit too high on the . Interest-ingly, there is a cross between the lines 1 and 2 in the vicinity of kuñāś. Usually this sign marks an addition made by a corrector of a manuscript to the line underneath it, but no akṣara is added.



it appears twice in a collocation with wac ‘struggle’ in A 238a3 and PK NS 1b1. e meaning ‘fight’ for kuñaś is assured by this collocation and in addition by a Sanskrit parallel that exists for A 353. is text is a rendition of the final section of the Prātimokṣasūtra (Schmidt 1989) in which TA mā kuñaś ypamāñ(cs)ā perl. pl. ‘by not doing a fight’ corresponds to Skt. avivada(māna)ir inst. pl. ‘by not fighting’.

TA kuraś is a hapax attested in A 257b2. e fragment belongs to the eleventh act of the Maitreyasamiti-Nāṭaka for which the Old Uyghur ver-sion is attested in MaitrHami XI, 11a6–12a12 (see Geng, Klimkeit, Laut 1988). Accordingly, TA kuraś corresponds to OUygh. soγuq ‘cold’.

Concerning the etymology, Van Windekens already (1941: 48 and 1976: 245) connected kuñaś with the PIE root *gwhen- ‘beat’ (2LIV 218f.)44 and kuraś with the PIE root *gwher- ‘become warm’ (2LIV 219f.).45 He did so, however, by proposing now outdated morphological and phonologi-cal developments (cf. also Hilmarsson 1996: 192f.).46

Based on the morpho-phonological solution proposed for tläś and lkäś, it is possible to take a new look at kuñaś and kuraś. As Peters (2004: 267 n. 4) pointed out, the original locative of the amphikinetic singular paradigm of *nt-stems in PIE ended in *-ent-i or *-ēnt.47 Assuming that the suffixed and the lengthened-grade variants of the locative could be blended (as happened, for instance, in the loc. pl. *pēd-su > OIr. ís ‘un-

44 For the semantics cf. Modern Engl. fight and NHG fechten ‘fence’.45 For the semantics cf. expressions like Modern English ‘burning cold’. Also,

note the possibility that the two roots 2*ḱel- ‘become warm’ (2LIV 323), e.g. in Lat. caleō ‘be warm’, and *ḱelH- ‘become cold, freeze’ (LIV 323), e.g. in Lith. šaliù, might be one and the same (Yamazaki 2009: 450ff.).

46 Already Hilmarsson (1996: 182f., 192f.) argued against the connection of TA kuraś with TB krośce TA krośś- ‘cold’. e latter are probably related to the family of Gk. κρύσταλλος ‘ice’ (as already suggested by Duchesne-Guillemin 1941: 155) and Lat. crusta ‘hard surface-layer’.

47 Peters (loc. cit. and 2006: 344 n. 48) argues that the leveled locative allo-morph *-ēnt- is at the origin of the stem-final palatalization in the Tocharian nt-participle (cf. the contrast between Gk. ἄγω ‘drive, lead’ and TB/TA āk- ‘id.’ in the nt-participle ἄγοντ- vs. TB aśeñca, TA āśant); see Fellner (2014) for a different view.



der’, Alb. për-posh ‘id.’), such forms of the roots in question, *gwhn-ēnt-i and *gwhr-ēnt-i, would result in PT *kwäñæñcä and *kwäræñcä. With syn-cope in pre-TA these would give *kuñañc and *kurañc resulting in attested kuñaś and kuraś.

Semantically, these forms would first have developed to adverbial phrases ‘at X-ing’ (cf. the semantics of *pēd-su originally ‘at the feet’ > ‘under’) functioning as converbs. On the way to Proto-Tocharian the pre-decessors of kuñaś and kuraś were then probably interpreted as belonging to the continuants of the old *nt-i-type abstracts of the lkäś-type with a semantic development from ‘at fighting’ to ‘fight’ and ‘at being cold’ to ‘cold’, respectively, cf. loc. *h2us(s)-er (Ved. uṣar-) ‘at dawn’ → *h2us(s)-ḗr (Gk. ἀήρ) ‘mist, dimness’ (Nussbaum 1986: 237).

[received: may 2015]University of Vienna

Department of LinguisticsSensengasse 3a

1090 ViennaAustria

[email protected]@univie.ac.at

References

BHSD = Edgerton, Franklin (1953) Buddhist Hybrid Sanskrit grammar and dic-tionary. New Haven: Yale University Press.

Burlak, Svetlana A., and Ilya B. Itkin (2009) “e Tocharian A forms naṣ=äṃ, naṣ=äm and n=äṃ, n=äm revisited”. Tocharian and Indo-European Studies 11: 43–48.

Burlak, Svetlana A., and Ilya B. Itkin (2010) “Yreki et autres addenda et corri-genda – 2”. Tatjana M. Nikolaeva et al. (eds.), Issledovanija po lingvistike i semiotike. Sbornik statej k jubileju Vjač. Vs. Ivanova, 342–358. Moskva: Jazyki slavjanskix kul´tur.

CEToM = A comprehensive edition of Tocharian manuscripts, at URL: http://www.univie.ac.at/tocharian



DKS = Bailey, Harold W. (1979) Dictionary of Khotan Saka. Cambridge Univer-sity Press.

2DoT = Douglas Q. Adams (2013) A dictionary of Tocharian B. 2nd edition. Am-sterdam & New York: Rodopi.

DTA = Gerd Carling, Georges-Jean Pinault and Werner Winter (2009) Dic-tionary and thesaurus of Tocharian A. Volume 1: letters a–j. Wiesbaden: Harrassowitz.

Duchesne-Guillemin, Jacques (1941) “Tocharica”. Bulletin de la Société de Lin-guistique de Paris 41: 140–183.

Fellner, Hannes A. (2014) “Tocharian special agents: e nt-participles.” Tocha-rian and Indo-European Studies 15: 53–67.

Geng Shimin, Hans-Joachim Klimkeit and Jens Peter Laut (1988) “„Das Er-scheinen des Bodhisattva.“ Das 11. Kapitel der Hami-Handschrift der Maitri-simit”. Altorientalische Forschungen 15: 315–366.

Hartmann, Jens-Uwe (1987) Das Varṇārhavarṇa des Mātṛceṭa (Sanskrittexte aus den Turfanfunden 12). Göttingen: Vandenhoeck & Ruprecht.

Hartmann, Jens-Uwe (1988) Neue Aśvaghoṣa- und Mātṛceṭa-Fragmente (Nach-richten der Akademie der Wissenschaften zu Göttingen, Philologisch-His-torische Klasse Nr. 2). Göttingen: Vandenhoeck & Ruprecht.

Hartmann, Jens-Uwe (2009) “Neues zum Varṇārhavarṇa”. Martin Straube et al. (eds.), Pāsādikadānaṃ. Festschrift für Bhikkhu Pāsādika, 229–241 (Indica et Tibetica 52). Marburg: Indica et Tibetica.

Hartmann, Jens-Uwe and Dieter Maue (1991) “Neue Spuren von Mātṛceṭas Varṇārhavarṇa”. Zeitschrift der Deutschen Morgenländischen Gesellschaft 141: 69–82.

Hilmarsson, Jörundur (1991) “Tocharian etymological notes 1–13”. Tocharian and Indo-European Studies 5: 137–183.

Hilmarsson, Jörundur (1996) Materials for a Tocharian historical and etymo-logical dictionary. Ed. by Alexander Lubotsky and Guðrún Þórhallsdóttir (Tocharian and Indo-European Studies, Supplementary Series 5). Reykjavík: Málvísindastofnun Háskóla Íslands.

Höfler, Stefan (2012) Untersuchungen zum Ablaut der neutralen s-Stämme des Indogermanischen. MA thesis, University of Vienna.

http://othes.univie.ac.at/18591/1/2012-01-24_0501399.pdfHoward, Angela F. and Giuseppe Vignato (2015) Archaeological and visual sourc-

es of meditation in the ancient monasteries of Kuča. Leiden: Brill.Klingenschmitt, Gert (1975) “Altindisch śaśvat-”. Münchener Studien zur Sprach-

wissenschaft 33: 67–78 = Klingenschmitt (2005: 149–157).



Klingenschmitt, Gert (2005) Aufsätze zur Indogermanistik. Hg. von Michael Jan-da et al. Hamburg: Kovač.

Liebert, Gösta (1976) Iconographic dictionary of the Indian religions: Hinduism, Buddhism, Jainism. Leiden: Brill.

Malzahn, Melanie (2014) “Tocharian A śorki ‘fear’ and two other TA scary words”. Tocharian and Indo-European Studies 15: 87–94.

MW = Monier-Williams, Monier (1899) Sanskrit-English dictionary. London: Clarendon.

Melchert, H. Craig (1999a) “Two problems of Anatolian nominal derivation”. Heiner Eichner and Hans Christian Luschützky (eds.), Compositiones In-dogermanicae in honorem Jochem Schindler, 365–375. Prague: Enigma.

Melchert, Craig H. (1999b) “Hittite tuk(kan)zi- ‘cultivation, breeding’”. Ktema 24: 17–23.

Melchert, Craig H. (2003) “Hittite nominal stems in -anzan-.” Eva Tichy et al. (eds.), Indogermanisches Nomen. Derivation, Flexion und Ablaut. Akten der Arbeitstagung der Indogermanischen Gesellschaft vom 19. bis 21. September 2001 in Freiburg, 129–139. Bremen: Hempen.

Nussbaum, Alan J. (1986) Head and horn in Indo-European. Berlin & New York: Walter de Gruyter.

Nussbaum, Alan J. (1999) “*Jocidus: an account of the Latin adjectives in -idus”. Heiner Eichner and Hans Christian Luschützky (eds.), Compositiones In-dogermanicae in honorem Jochem Schindler, 377–419. Prague: Enigma.

Peters, Martin (2004) “On some Greek nt-formations”. John H. W. Penney (ed.), Indo-European perspectives. Studies in honour of Anna Morpurgo Davies, 266–276. Oxford: Oxford University Press.

Peters, Martin (2006) “Zur morphologischen Einordnung von Messapisch klaohi”. Maria Teresa Laporta (ed.), Studi di antichità linguistiche in memoria die Ciro Santoro, 329–353. Bari: Caccuci.

Peyrot, Michaël (2013) The Tocharian subjunctive. Leiden & Boston: Brill.Pinault, Georges-Jean (1997) “Terminologie du petit bétail en tokharien”. Studia

Etymologica Cracoviensia 2: 175–218.Pinault, Georges-Jean (1999) “Tokharien A kapśañi, B kektseñe” Heiner Eich-

ner and Hans Christian Luschützky (eds.), Compositiones Indogermanicae in honorem Jochem Schindler, 457–478. Prague: Enigma.

Pinault, Georges-Jean (2006) “Morphologie de l’ablatif tokharien”. Gerd Carling (ed.), GIŠ.HURgul-za-at-ta-ra. Festschrift for Folke Josephson (Meijerbergs Ar-kiv 32), 248–283. Göteborg: Meijerbergs institute.

Pinault, Georges-Jean (2008) Chrestomathie tokharienne. Textes et grammaire. Leuven & Paris: Peeters.



Pinault, Georges-Jean (2009) “On the formation of the Tocharian demonstra-tives”. Elisabeth Rieken and Paul Widmer (eds.), Pragmatische Kategorien. Form, Funktion und Diachronie. Akten der Arbeitstagung der Indogermani-schen Gesellschaft vom 24. bis 26. September 2007 in Marburg, 221–245. Wies-baden: Reichert.

Pinault, Georges-Jean (2012) “Tocharian -nt-participles and agent nouns”. Olav Hackstein and Ronald I. Kim (eds.), Linguistic developments along the Silk-road: Archaism and innovation in Tocharian, 179–204. Wien: Verlag der Ös-terreichischen Akademie der Wissenschaften.

Pinault, Georges-Jean (2014) “Vedic reflexes of the Hittite tukkanzi-type”. Craig Melchert et al. (eds.), Munus amicitiae. Norbert Oettinger a collegis et amicis dicatum, 262–275. Ann Arbor & New York: Beech Stave.

Schmidt, Klaus T. (1980) “Zu einer metrischen Übersetzung von Mātṛceṭas Bud-dhastotra Varṇārhavarṇa in osttocharischer Sprache”. Wolfgang Voigt (ed.), XX. Deutscher Orientalistentag vom 3. bis 8. Oktober 1977 in Erlangen, 341–343 (Zeitschrift der Deutschen Morgenländischen Gesellschaft, Supplement 4). Wiesbaden: Steiner.

Schmidt, Klaus T. (1983) “Zum Verhältnis von Sanskritvorlage und tochari-scher Übersetzung. Untersucht am Beispiel osttocharischer Stotratexte”. Klaus Röhrborn and Wolfgang Veenker (eds.), Sprachen des Buddhismus in Zentralasien. Vorträge des Hamburger Symposions vom 2. Juli bis 5. Juli 1981 (Veröffentlichungen der Societas Uralo-Altaica 16), 125–131. Wiesbaden: Har-rassowitz.

Schmidt, Klaus T. (1987) “Zu einer metrischen Übersetzung von Mātṛceṭas Bud-dhastotra Varṇārhavarṇa in tocharischer Sprache”. Tocharian and Indo-Euro-pean Studies 1: 152–168.

Schmidt, Klaus T. (1989) Der Schlußteil des Prātimokṣasutra der Sarvāstivādins. Text in Sanskrit und Tocharisch A verglichen mit den Parallelversionen anderer Schulen (Sanskrittexte aus den Turfanfunden 13). Göttingen: Vandenhoeck & Ruprecht.

SWTF = Ernst Waldschmidt et al. (1973–) Sanskrit-Wörterbuch der buddhistisch-en Texte aus den Turfan-Funden. Göttingen: Vandenhoeck & Ruprecht.

TG = Emil Sieg, Wilhelm Siegling and Wilhelm Schulze (1931) Tocharische Grammatik. Göttingen: Vandenhoeck & Ruprecht.

omas, Werner (1967) “Zu wortverbindendem toch. A śkaṃ / B ṣpä”. Zeitschrift für Vergleichende Sprachforschung 81: 161–180.

TLT = Pavel Poucha (1955) Thesaurus linguae Tocharicae dialecti A. Praha: Státní Pedagogické Nakladatelství.



TochSprR(A) = Emil Sieg and Wilhelm Siegling (1921) Tocharische Sprachreste. I. Band. Die Texte. A. Transcription. Berlin & Leipzig: de Gruyter.

TVS = Melanie Malzahn (2010) The Tocharian verbal system. Leiden & Boston: Brill.

Van Windekens, Albert J. (1941) Lexique étymologique des dialectes tokhariens. Louvain: Muséon.

Van Windekens, Albert J. (1976) Le tokharien confronté avec les autres langues indo-européennes. Volume I. La phonétique et le vocabulaire. Louvain: Mu-séon.

VAV = Hartmann (1987)Yamazaki, Yoko (2009) “e Saussure effect in Lithuanian”. Journal of Indo-Eu-

ropean Studies 37: 430–461.


tocharian and indo-european studies (abbreviated ties) was founded in 1987 by Jörundur Hilmarsson and appeared under his editor-ship in Reykjavík, Iceland, until his death in 1992. Vol. 6 appeared post-humously in 1993 under the editorial finish of Guðrún Þórhallsdóttir who also supervised the continuation of the Supplementary Series up to vol. 5 in 1997. Vols. 7 to 10 were published by C. A. Reitzel Publishers Ltd., Co-penhagen. Subsequent volumes are published by Museum Tusculanum Press in Copenhagen.

Editors: Birgit Anette Olsen (executive editor, Copenhagen) Michaël Peyrot (Berlin) Georges-Jean Pinault (Paris) omas Olander (assistant editor, Copenhagen)

Orders and correspondence concerning distribution should be directed to: Museum Tusculanum Press Birketinget 6 dk 2300 Copenhagen S Denmark www.mtp.dk [email protected]

Manuscripts for publication and other correspondence on editorial mat-ters should be directed to:

Mail: Georges-Jean Pinault École Pratique des Hautes Études. À la Sorbonne. 54 rue Saint-Jacques. CS 20525. 75005 Paris France

Email: [email protected]


Tocharian and Indo-European Studies · tively plausible Indo-European etymologies. TA . kaś (TB ....

Documents

Transcript of Tocharian and Indo-European Studies · tively plausible Indo-European etymologies. TA . kaś (TB ....