Cuneiform Luvian Lexicon Online

Cuneiform Luvian Lexicon

FOREWORD

Like its predecessor, this lexicon has a very modest aim: to furnish a provisional index, as exhaustive as possible, of all attested Cuneiform Luvian lexemes. It is intended to be complete for the CLuvian corpus as established by Frank Starke, Die keilschrift-luwischen Texte in Umschrift (StBoT 30), Wiesbaden: 1985 (with the addenda and corrigenda in StBoT 31 (1990) 592-607). I have followed Starke's readings unless explicitly noted otherwise. The coverage of Luvianisms and Luvian loanwords in Hittite contexts is necessarily selective: see further below. Once again, I warn readers that this work is not a proper dictionary with full critical apparatus. Bibliographical references are haphazard. I welcome help in filling important lacunae, but I expressly disavow any attempt at truly adequate coverage. I also repeat my entreaty to readers to give full weight to all qualifiers: many interpretations remain uncertain and should be treated accordingly.

The starting point for this collection is naturally the seminal work of Emmanuel Laroche, Dictionnaire de la langue louvite, Paris: 1959. It should also be self-evident that this lexicon would not have been possible in its present form without the excellent new organization of the CLuvian corpus by Frank Starke cited above. As will be clear from numerous entries, I have also profited greatly from Starke's published studies on CLuvian morphology, especially his monumental Untersuchung zur Stammbildung des keilschrift-luwischen Nomens (StBoT 31), Wiesbaden: 1990. Unfortunately, however, I find Starke's approach to CLuvian grammar seriously flawed in certain fundamental respects. Specific cases are discussed under the various lemmata, but I have also thought it useful to summarize briefly below important points where my views differ from his. I also refer readers to my review of StBoT 31 in HS 105 (1992) 309-312.

Identification of Luvianisms and Luvian loanwords in Hittite contexts remains a challenging task. First of all, not all words marked with the so-called "Glossenkeil" are Luvian, or even foreign words at all. I categorically reject the implication of Oettinger, KZ 99 (1986) 51⁵⁰, that we should attribute a Luvian origin to words with impeccable Hittite phonology and morphology merely on the basis of the Glossenkeil. Conversely, many words of Luvian origin are not marked with the Glossenkeil in Hittite contexts. As Starke has demonstrated in his Untersuchung and elsewhere, the influence of Luvian on Hittite is much more profound than previously acknowledged. However, Hittite and Luvian are closely related languages, and it remains in principle difficult to distinguish with precision Luvianisms or Luvian loanwords from genuine Hittite cognates of Luvian words. Scholars will inevitably differ in the analysis of particular cases. In collecting Luvian vocabulary I have tried to exploit all readily available sources: Laroche, DLL, Friedrich, HW, Friedrich-Kammenhuber, HW², Puhvel, HED, and of course Starke, StBoT 31. I am well aware that my collection will be far from complete. In attempting to account for all CLuvian lexemes, I have also thought it prudent to make explicit my rejection of certain examples where a Luvian interpretation has been proposed: such lemmata are marked by square brackets ([ ]).

The situation of the Glossenkeilwörter and other Luvianisms in Hittite means that the number of omissions and errors in this lexicon will be even higher than usual in such a compilation. I not only welcome but solicit corrections and addenda from colleagues. I have again chosen the present format to permit the insertion of replacement pages and intend to issue updates at regular intervals.

 H. Craig Melchert

 Chapel Hill

 April, 1993

Cuneiform Luvian Lexicon

CITATION FORMS AND SYMBOLS

For the sake of brevity, I follow the practice of Laroche, DLL, in citing the occurrences of CLuvian words. Texts from Volume XXXV of Keilschrifturkunden aus BoŠazköy (KUB) are cited first, merely by text number; these are followed by other volumes of KUB, cited by volume number alone; finally come instances from Keilschrifttexte aus BoŠazköy (KBo) and other text collections with the usual abbreviations.

Abbreviations of grammatical categories are of the standard sort. C = Common (Animate) Gender. Please note that I use Abs(olute) merely as a convenient label for instances of an uninflected stem, without intending any claim for an "absolute case" as a grammatical category. Note the use of the following symbols:

The citation of nominal stems in CLuvian is complicated by the phenomenon of "i-motion", by which an -i- is obligatorily added to the animate nominative and accusative of most adjectives and nouns (see Starke, StBoT 31.59ff). I use the following conventions. For consonant stems, I place the -i- in parentheses: maššan(i)-. In the case of a-stems, where the -i- replaces the stem-final -a-, I follow Starke in using -a/i-: āra/i-. In the case of stems in -iya- and -aya-, we cannot be certain whether the replacement of the final -a- by -i- results in contraction or not. I assume contraction for the sake of simplicity and list these stems as -i(ya)- and -ay(a)-: ānni(ya)- and ārray(a)-. Since the -i- is not part of the underlying stem, as emphasized by Starke, I ignore it in alphabetization: thus maššan-, āra-, ānniya-, ārraya-. Readers should note that some words traditionally treated as i-stems will thus be alphabetized differently: e.g. ānna/i-'mother' or tāta/i- 'father'.

Establishing "i-motion" with certainty requires an example of the animate nominative or accusative and some other case form. Given our limited corpus, many cases are inevitably indeterminate. Faced with only animate nominative or accusative forms in -i-, I have assumed "i-motion" only in cases where the suffix involved makes such an assumption likely (e.g. diminutives in -anna/i- or adjectives in -alla/i-). Due to the ambiguities of CLuvian case endings, only a neuter nominative-accusative singular can show the difference between a consonant stem and one in -a- (ending respectively 0/(-ša) and -an(-za)). The choice between a stem -C(i)- and -Ca/i- often depends on one's historical analysis and is thus a matter of interpretation, not fact.

Cuneiform Luvian Lexicon

NOTES ON CLUVIAN GRAMMAR

The demonstration of Hawkins, Morpurgo Davies and Neumann, HHL (1974), that HLuvian and CLuvian are closely related dialects means that the best current descriptions of CLuvian grammar are those for HLuvian: see Massimiliano Marazzi, Il geroglifico anatolico. Problemi di analisi e prospettive di ricerca (Rome: 1990) or Rudolf Werner, Kleine Einführung ins HieroglyphenLuwische (Freiburg/ Göttingen: 1991). Readers should note the following points where my analysis differs from that of Starke or other investigators:

I.       Nominal Inflection

1.       I share in the consensus that the particle -ša/-za marks neuter nominative-accusative singular, contra Starke, StBoT 31.46ff.

2.       Nominative-accusative forms in -a beside animate singulars in -iš/in are collectives (Eichner, MSS 45.18f), not reflexes of duals, pace Starke, StBoT 31.29 et al. See e.g. lalama/i- or dušduma/i-.

3.       u-stem adjectives do not show "i-motion" in Luvian, contra Starke, StBoT

31.75 et aliter, who wrongfully emends exx. such as NSgC a-ru-uš. The secondary appearance of -inzi in the animate nominative plural alone (kuwanzuinzi for *kuwanzunzi) does not alter this fact.

4.       Oblique forms of adjectives in -ašša/i- with an inserted element -anz- are genuine, contra Starke, StBoT 31.38ff. As I will argue in detail in the forthcoming memorial volume for Charles Carter, the -anz- in these forms marks plural number of the underlying possessing(!) noun.

II.      Nominal Derivation

1.       Starke, StBoT 31 passim, consistently ignores the possibility that originally deverbative suffixes can secondarily come to be denominative as well. As a result, he arbitrarily reconstructs several dozen unattested and totally unmotivated verb stems. Oettinger, MSS 40.146f, has properly criticized this procedure for Hittite -ant-. The same argument applies to CLuvian suffixes such as -a(i)mma/i-, -ttar/-ttn- and -aæit-.

2.       Starke, StBoT 31, ignores the class of CLuvian deverbative animate nouns in -(a)ma/i- (NB single -m-!): lalama/i- 'receipt' < lala- 'take', dušduma/i- 'manifest, voucher' < *du-šdu- 'make known', etc.. This suffix corresponds to Hittite deverbative -(i)ma-.

3.       Pace Laroche, DLL 137, & Starke, StBoT 31.63f, there is no class of i/ya-stems in Luvian. As shown by Carruba, Fs Neumann 35ff, all alleged cases of ya-allomorphs in nouns actually belong to derived adjectives in -i(ya)-. See also Melchert, HS 103.198ff.

4.       Contra Starke, StBoT 31.384ff, there is no class of Luvian neuter nouns in **-štar, only animate(!) nouns in /-s(t)ra/i-/, as already established by Neumann, Sprache 11.82ff.

III.    Verb Inflection

1.       I follow Morpurgo Davies, KZ 94.106²⁴, in assuming a CLuvian second singular present ending -tiš corresponding to HLuvian -ti-s(a), but one should compare the alternative analysis of Starke, Sprache 31.249ff. The issue cannot be regarded as settled. I do call readers' attention to the likelihood of a CLuvian second singular "æi-present" ending -ti beside third singular -(a)i: see under lāla-, nana-, and waliya-.

2.       CLuvian shows an unexplained preterite third plural ending -aunta matching HLuvian /-aunta/: see under nakkuššā(i)- or warmā(i)-. As shown by HLuvian wa/i-la-u-ta 'they died' to a stem wal(a)-, the -u- is not part of the stem, and one should not set up otherwise non-existent stems in **-aw- (contra Starke, StBoT

31.538).

IV.    Verb Derivation

CLuvian shows four distinct verbal classes which must be kept clearly distinguished (even if the sparse attestation makes certain individual cases ambiguous):

1.       Denominatives in -ā- without "lenition" of the endings: e.g. tūrā- 'pierced' (or simil.) < tūra/i- (a tool or weapon). This type corresponds to Hittite verbs in -aææ- and Lycian non-leniting verbs in -a- (e.g. prñnawa- 'build' < prñnawa-'(grave)-house').

2.       Denominatives in -ā-/-āi- with "lenition": e.g. arunā(i)- '?'. This type corresponds to Hittite verbs in -ā(i)- and Lycian leniting verbs in -a(i)- (e.g. xtta(i)- 'harm' < xtta- 'harm, violence'). Examples such as kiša(i)- 'comb' or wida(i)- 'strike' (sic!) may represent inherited lengthened-grade iteratives (cf. Lat. cēlāre) instead of denominatives.

3.       Denominatives in -i(ya)- without "lenition": e.g. tummanti(ya)- 'hear' < tumman(t)- 'ear'. These correspond of course to Hittite verbs in -ye-/-ya- and Lycian verbs in -i(je)-.

4.       Denominatives/Iteratives in -i-/-ai- with "lenition": e.g. tarmi-/tarmai- 'nail, fasten' < tarma/i- 'nail, peg'; tūpi-/tūpai- 'strike'. These correspond to Hittite verbs in -e-/-a- < *-eye/o- and Lycian leniting verbs in -i-/-ei- (kumezidi/kumezeiti 'sacrifice' < kumaza- 'priest'; tubidi/tubeiti 'strike').

The distinction between types 3 and 4 was already established by Morpurgo Davies, KZ 96.245-270. Starke, StBoT 31 passim, ignores the difference between cases like unlenited tummantitta (with only -i(ya)- forms!) and lenited tarmita (with weak stem tarmai-!) and posits a wide-ranging but non-existent "suppletion" between stems in "-i@i-" and stems in "-ai@i-". This fictitious suppletion and the many consequences drawn from it must be rejected.