Remerciez-le!

Remerciez @Admin pour avoir partagé cet document gratuitement, de la manière la plus simple, en partageant sur les réseaux sociaux.

Melanie Malzahn Surprise at length of Tocharian nouns Abstract: The Tocharian l

Melanie Malzahn Surprise at length of Tocharian nouns Abstract: The Tocharian languages exhibit many examples of the lengthened grade in roots of nominal stems. We find forms that, on the one hand, descend from PIE paradigms in which lengthened grades are expected given our current understanding of ablaut patterns, e. g. TA śanweṃ‘jaws’, which descends ei- ther from an acrostatic u-stem or from a PIE denominative o-stem that denoted appurtenance and was formed by vṛddhisation. On the other hand, there are lengthened grades in Tocharian nouns that do come as a surprise as is the case with TB ñem/TA ñom ‘name’, if this goes back to a protoform PIE *h₁nḗh₃-m. These unexpected lengthened-grade forms will be discussed especially in the context of the so-called Narten system. Keywords: Tocharian, nominal word formation, vocalism, lengthened grade, Narten system Melanie Malzahn: University of Vienna; melanie.malzahn@univie.ac.at When it comes to reconstructing details of PIE nominal morphology, Tocharian does not take pride of place. Nevertheless, there are quite a few Tocharian nouns showing what seems to be a pre-Proto-Tocharian (pre-PT) lengthened grade in the root or in a nominal suffix.1 In this paper, I tackle what seem to be inherited lengthened grades of roots reflected in Tocharian nouns.2 1 Thematic nouns and what may be derivatives from thematic nouns Lengthened grades found in root syllables of both thematic nouns and athematic nouns that may have been derived from thematic nouns (such as *-i-, *-n-, and 1 There are plenty of such pre-PT lengthened grades in the Tocharian verbal system; see Malzahn 2010 passim. 2 As for suffixal ablaut, e. g., TB maśce ‘fist’ evidently forms an equation with Proto-Indo-Iranian *musti- ‘fist’, but seems to attest to a quite unexpected PIE nom.sg. ending *-tē(y) > pre-PT *-tēs instead of the regular *-ti-s met in Indo-Iranian. 10.1515/if-2014-0014 Bereitgestellt von | Vienna University Library Angemeldet Heruntergeladen am | 12.07.17 15:18 260 Melanie Malzahn *-eh₂- stems) come as no surprise, as PIE had denominative o-stems which were formed by vṛddhisation of the root vowel of the base noun (i. e., addition of an e- or o-vowel) and denoted appurtenance. Derivatives of such a morphological struc- ture were probably used in order to substantivize adjectives, as well (as per Weiss 2007: 261). Both TB yente/TA want ‘wind’ < *h₂wēh₁-t-o- (sic, as per Schindler 1994: 399; Widmer 1997: 28; evidently analyzed as a substantivized variant of a participle with Narten root ablaut *h₂weh₁-t- by Schindler; somewhat differently Widmer)3 and TB yerpe (> TA yerpe) ‘orb’ < *h2/3ērbʰ-o- (see Adams 2013: 548; ana- lyzed as the result of a “substantivization by vṛddhi” of a verbal adjective *h₃erbʰ- ó- ‘turning’ by Weiss 2007: 260f.) seem to exhibit precisely this process. As far as leges artis are concerned, TB śer(u)we/TA śaru ‘hunter’ may be interpreted as the vṛddhi derivative of a solid-looking basic noun as well, i. e., for an adjective of appurtenance based on PIE *ḱerwo- ‘stag’, as suggested by Jasanoff apud Nuss- baum 1986: 8 (other etymological analyses have been proposed, however).4 In a similar vein, it is tempting to construe TB ṣpel ‘mud’ (masc.), which, on the claim of Adams (2013: 731), belongs with Gk. πηλός ‘mud, clay, dung’ and is derived from *spēh₂l-,5 as a masculine endocentric substantivization in *-i-6 of a vṛddhi adjective *(s)pēh₂lo- ‘muddy’ based on the noun *(s)peh₂lo- ‘mud’, a direct reflex of which we would possibly have in the Greek word for ‛mud’. The following Tocharian nouns that also seem to show a lengthened grade in the root may have started out as (derivatives from) thematic vṛddhi formations as well, but here such an analysis is less attractive to account for the morphological structure and/or semantics: – TB ariwe* ‘ram, male goat’ (Adams 2013: 24: < *h₁ōreywo- ~ Skt. āreya- ‘ram’), – TB āntse, TA es ‘shoulder’ < PT *ānsæ usually derived from a preform *ōms-o- (Adams 2013: 46; but see also Hackstein 2002: 190f. on the root ablaut), 3 Actually a protoform *h₂weh₁-to- lacking vṛddhi would in my view have resulted in Late pre-PT *wento-; see Malzahn 2011: 97, fn. 32. 4 See Adams 2013: 695; alternatively, Pinault (2006: 179–181; cf. 2008: 588f.) takes the Tocharian word to be a borrowing from a non-Indo-European Central Asiatic language. 5 *-h₂- should be set up because of the Doric evidence pointing to Proto-Greek *-ā-; but note that for the Greek word, Meillet suggested (see Meillet 1905 and Ernout & Meillet 1985: 645) a completely different account. It is also possible to analyze TB ṣpel as acrostatic *l-stem (possibly once basic to the Greek noun, if this had started out as an adjective itself). As for the root involved, one might entertain the possibility that it is the *√speh₂ said to underlie Ved. sphāyate ‘become fat’ and Hittite išpai-i/išpi- ‘to get full, to be filled, to be satiated’ by Nussbaum apud Jasanoff 1994: 160, fn. 19 and Jasanoff 2003: 108f. (in this case, the original semantics of Gk. πηλός may have been “(earth) satiated with water”); alternatively, these verbs could be derived from a root *√spheh₁ “wunschgemäß geraten, gelingen” (LIV²: 584). 6 See Nussbaum apud Vine 2006: 151. Bereitgestellt von | Vienna University Library Angemeldet Heruntergeladen am | 12.07.17 15:18 Surprise at length of Tocharian nouns 261 – TB yepe ‘knife’ < PT *w’æpæ, cf. Goth. wēpn, etc. ‘weapon’ (according to Adams 2013: 547 from a pre-PT *n-stem “*wēb-en-”), – TB yerkwanto ‘wheel’, according to Hilmarsson 1986: 275 a vṛddhi formation *h₂wērg-wt-ōn- with individualizing -ōn-,7 – TB yetse ‘skin’: according to Adams 2013: 549 just like German Aas ‘carrion’ from *h₁ēd-so- *‘that which one eats’, with a semantic development first to *‘flesh’ and then to ‘skin’, – TA yṣaṃ‘trench, moat’: from *sēd-n-o- > PT *ṣænæ according to Pinault 2008: 208, – TB sāle ‘ground, basis’; said to derive from a PIE *sōlo- by Adams 2013: 748. Here may finally also belong TB yerter ‘wheelrim, felloe’, although this noun looks at first sight rather like a stem with a suffix *-tor- (such as *h₂wērg-tor-, as set up by Adams 2013: 548) or *-or- (such as *wērt-or(-), as set up by Widmer 1997: 47f.). According to Pinault (2011: 165) it is in fact a denominative in *-wer-/wen- derived from a lengthened-grade formation *h₂ēr-to- ‘joint’. 2 Athematic nouns As for lengthened grades of roots in athematic nouns that are not obvious deriva- tives from thematic vṛddhi formations, some of them are completely unremark- able (at least for followers of the Schindler School), e. g., the TA dual form śan- weṃ‘jaws’, which evidently attests a lengthened grade *ǵēn-, and can be derived from the paradigm of an acrostatic u-stem that had an *ē-grade rather than an *o-grade of the root in the strong case forms (as per Nikolaev 2010a: 1–18, esp. 4f. with refs.8; for this type of u-stems in general, see again Nikolaev 2010a: 221, 327f. and also Nikolaev 2010b: 195f.). Similarly, the *ē-grade i-stem TB yel, TA wal < *wēl-i- ‘worm’ is reminiscent of the Greek abstract i-stem δῆρις ‘battle, contest’, which also shows an *ē-grade. Elsewhere, however, lengthened-grade nominal roots in Tocharian do come as a surprise, in particular in the case of one of the most famous Tocharian nouns, 7 But see also Malzahn 2010: 17, fn. 21 on the possibility of secondary palatalization. For other ways to cope with the TA equivalent wärkänt having -ä- instead of expected -a-; see Adams 2013: 547f. 8 Here is also made mention of an alternative account by Klingenschmitt, who tried to explain the form as the continuant of a thematic vṛddhi formation based on the very u-stem. This reference had to be omitted in the English summary on pages 300–305. Bereitgestellt von | Vienna University Library Angemeldet Heruntergeladen am | 12.07.17 15:18 262 Melanie Malzahn TB ñem/TA ñom ‘name’. This noun is now often derived from a lengthened-grade protoform PIE *h₁nḗh₃-m(allegedly backed by evidence from Uralic languages), above all by Neri 2006: 213, 236. However, there are also many scholars who deny that Tocharian ñ- has to be explained by reconstructing a pre-PT root vowel *-ē-; see above all Pinault 2008: 194. Now it is certainly true that a lengthened-grade protoform *h₁nḗh₃-mwould be unwelcome, since neuter men-stems typically in- flect proterokinetically and not acrostatically. However, the same can be said of neuter -wer-/-wen- stems, whereas Hittite mēhur ‘time’ and šēhur ‘urine’, as ana- lyzed by Eichner (1973), are acrostatically inflected nouns as well. Eichner’s anal- ysis is bolstered by genitive forms in -unas, i. e., we do not find in these two nouns a genitive in -waš or -wenaš from typically proterokinetic *-wen-(o)s. These two fa- mous etymological analyses have, however, recently been called into question by Kloekhorst (2008: 568) and Kümmel (2011).9 Finally, at least Anatolian seems to attest to the existence of some acrostatically inflected neuter s-stems (see above all Rieken 1999: 187–190 and most recently Melchert 2010), which flies into the face of Schindler’s well-known claim that the neuter s-stems had also inflected proterokinetically only. 3 Narten forms? Whereas there are no principled reasons against assuming acrostatically uploads/Litterature/ surprise-at-length-of-tocharian-nouns-pdf 1 .pdf