CC BY
30
DOI: 10.30842/ielcp230690152655
Alexander Nikolaev
Boston University, USA. alexander.s.nikolaev@gmail.com
LUWIAN LUHANUWA-, LYCIAN B LUGA- 'TO BURN'*
This paper proposes a new etymology for Luw. (FLAMMAE)lax-ha-nu-wa/i-tax/luhanuwata/ 'he burned' and Lyc. B ni... lugatu 'may they not burn': these verbal stems are analyzed as factitives ultimately based on the Proto-Anatolian nominal stem *louH-a(H) < *h2louh2/3-ah2 'burning' (with the loss of the initial laryngeal by the Saussure Effect), made from the PIE root *h2leuh233- 'to burn'.
Keywords: Luwian language, Luwian hieroglyphic script, Lycian B language, Anatolian verbal stem formation, Indo-European etymology.
А. С. Николаев
Бостонский университет, США. alexander.s.nikolaev@gmail.com Лув. luhanuwa-, лик. Б luga- 'сжигать'
В статье предлагается новая этимология для лув. (FLAMMAE)lax-ha-nu-wa/i-tax /luhanuwata/ 'он сжег' и лик. Б ni ... lugatu 'да не сожгут': эти глагольные основы возводятся к фактитивным дериватам от праанат. именной основы *louH-a(H) < и.-е. *h2louh233-ah2 'горение' (с выпадением начального ларингала по закону де Соссюра), от и.-е. корня *h2leuh2/3- 'гореть'.
Ключевые слова: лувийский язык, лувийское иероглифическое письмо, ликийский Б язык, анатолийское глагольное основообразова-ние, индоевропейская этимология.
The modest purpose of this note is to elaborate on a suggestion made in passing in my earlier publication in this journal1 and to discuss the possible Anatolian descendants of PIE *h2leuh233- 'to burn'.
In lines 3-4 of the Luwian hieroglyphic TOPADA inscription2 we find the following passage (see fig. 1 below):
§ 13 wa/i-sa pa+ra/i-zax-tax-wa/i-nix (URBS) "TERRA"-£wa/ix+ra/i *273-ti ("PES2")i+ra/i
* Warm thanks are due to Craig Melchert, Massimo Poetto, and David Sasseville, and Andrei Sideltsev; all errors of fact and judgment are, however, mine alone.
1 Nikolaev 2021a: 957 n. 14.
2 The inscription is notorious for the problems of interpretation it poses, cf. e.g. Weeden 2010: 47: "due in part to its highly idiosyncratic use of archaising or unusual sign-forms, it is extremely difficult to interpret".
§ 14 wa/i5-tax || URBS+M/.AEDIFICIUM-tax-na (FLAMMAE)(-)/ax-hd-nu-wa/i-tax
§ 15 *274-ia-pa-wa/i FEMINA.MANUS-zi/a-Aa SERVUS-sa ("PES")w-pa-tax3
Payne 2012: 58 translates: "he (the Royal Horse?4 the cavalry?) went into the land of Parzata(?) with strength and burnt(?) building(s)(?), and he reduced the region, (including) women (and) children (in) (to) servitude". The form of interest here is (FLAMMAE)iax-hà-nû-wa/i-tax plausibly interpreted by Hawkins 2000: 456 as 'he burnt': by analogy with other cases where the sign FLAMMAE (*477)5 is used as a determinative with words referring to fire6, Hawkins concludes that the "interpretation of the present verb as '(cause to) burn' would fit the context as understood".7 The
3 The transliteration follows the ACLT; for this inscription see Hawkins 2000: 451-61, plate 253; Woudhuizen 2007; Weeden 2010; Payne 2012: 54-9; d'Alfonso 2019. Per convention, determinatives are written in parentheses, while the presence of a logogram marker (*410) is indicated by placing a logogram in quotes. The numbering of hieroglyphic signs follows Laroche 1960; for an updated sign-list see Payne 2014: 161-96.
4 Viz. REX+RA/I-ti (ANIMAL)EQUUS-wa/i-ti (§ 5, 10). Note zila-ldli-tax-'galloped' (Weeden 2010: 51) in § 11.
5 The reversed form of the sign FLAMMAE in the inscription "seems no good reason to doubt its identity" (Hawkins 2000: 456), but see Weeden 2010: 52 for a position of skepticism; however, the vertical cross-bar mentioned by Weeden is likely to represent the division between two "tongues" of the flame.
6 See Laroche 1960: 241: kinu(wa)- 'to burn', marusana- 'fire-brand', and la/i/u-sa-la/i/u-sa 'to burn' (?), on which see the following footnote.
7 Hawkins' reading and interpretation of the form have been widely accepted. A sole dissenter is d'Alfonso (2019: 137) who reads *477 in TOPADA as a logogram: he transcribes FLAMMAE-lax-hd-nu-wa/i-tax as *laslahhanuwata without discussing either his reasons for this reading or the resulting form. The number of phonetic complements following the logogram (five!) makes his reading a priori less convincing. The form *laslahhanuwata posited by D'Alfonso is probably made from the same root as reduplicated la/i/u-sa-la/i/u-sa-ta in HAMA 4, § 11: ("BOS")u-su-pa\\-ta-ta-ha-wa/i ("FLAMMAE")la/i/u-za-li-na NEG2-' ("FLAMMAE") la/i/u-sa-la/i/u-sa\\-ta DEUS-ni-i SUPER+ra/i- ' INFRA-ta "they did not burn the burnt offering, a sacrificial ox, up and down to the god" (trans. Payne 2012: 65). In view of the triple vocalization of the sign *445, the writing system does not allow distinguishing between laslassa-, luslussa-, or lislissa- 'to burn (habitually)'; the reduplicated formation is clearly an iterative of some kind, construed with the likewise phonologically
verb recurs in § 24 of TOPADA in the same spelling and with the same meaning: URBS [(FLAMMAE)?]lax-ha-nu-wa/i-tax 'burnt the city'.8
The first sign in the form under discussion, transliterated by Hawkins as lax, is *455 (the third sign on the bottom line in fig. 1). This rare syllabogram is also found in § 4 of the same inscription where it is used to write the personal name wa/i5+ra/i-pa-* 455-wa/i-sax. This ruler was identified by Hawkins 1979a: 166 and Poetto 1982: 282 n. 22 with Warpalawa of Tuwana known to the Assyrians as Urbala'a and attested in Assyrian sources from 738 BCE through at least 709 BCE. The value <la> was thus tentatively established for the sign *455.9 However, Hawkins' identification of the ruler
ambiguous laz(z)ali- / liz(z)ali- / luz(z)ali- 'holocaust' as its direct object in a quasi figura etymologica (so ACLT). Importantly, the usage of lVslVsta lVz(z)alin in this passage is not immediately comparable with that of lVhanuwata in the TOPALA inscription where the object is 'buildings'. Contra Hawkins 2000: 406, it does not appear possible to directly compare Luwian lVslVs- with Hittite lukki/a- 'set fire to': PIE *leuk- ended in a plain velar, and as Luwian kisa(i)- 'to comb' shows, one would expect Luwian fluk- as the outcome; note that there is no evidence for PIE (or Proto-Anatolian) *luk-ske/o- (which would not have meant 'to light up, to put on fire' anyway). Poetto 1979: 677 n. 19 (whose study of this difficult passage remains influential) chooses laslas- as the transcription and compares Hittite lazzi- 'good, right', while Sasseville 2020: 135 compares unreduplicated Luwian (OCCIDENS)la/i/u-si- (which he transcribes as lusi- and translates as 'to disappear') and Lycian lusalija 'destructive' and proposes a derivation from PlE *leuH- 'to cut off (on which see Nikolaev 2010: 232-6).
8 To judge from the drawing in Hawkins 2000 (plate 253), the edge of the inscription is mutilated and it is impossible to ascertain whether the determinative (FLAMMAE) was present.
9 The only other word written in Iron Age Luwian texts with the sign *455
Fig. 1. TOPADA § 14 (Hawkins 2000, plate 253)
also allowed dating the inscription to the last third of the 8th cent. BCE, and it is known that another sign in the l-series showed alternating vowel values in the Late Period, namely, *445: the sole reading attested for this sign in the Empire Period is <lu>, but in the Iron Age inscriptions it could also be used to write <la> and <li>.10 The vocalization of lax (*455) is therefore uncertain; note that in his commentary on la/i/u-sa-la/i/u-sa-ta in HAMA § 11 (see n. 7), Hawkins (2000: 406) actually suggests luhanuwata as the bound transcription of the form in TOPADA.
Other potentially relevant Luwian forms are too unclear to be useful. One that has the highest chance of being from the same verb is found in MARA§ 8 § 2: wa/i-ta ku-ma-na mi-Ua?' '-tf^-i \FLUMEN.REGIO-ia-ti-i \\ DOMUS-na-x[...] ARHA (FLAMMAE) la/i/u-x(-x?) "when from my river-land the house[s?...] burned down". This morsel of an (allegedly) verbal form is written with *445, which, again, does not allow deciding between <la>, <li> or <lu>.
To sum up, under the assumption (which is not independently supported) that both l-signs, *445 and *455, had triple vocalization <lalilu> in Luwian hieroglyphic script, the sequence *455-hd-nu-wa/i-tax in the TOPADA inscription may be analyzed as 3 sg. pret. luhanuwata 'he burned'.
This verbal form is clearly made with the suffix -nuwa-i (Hitt. -nu-m\ Lyc. -nuwe-) used in Anatolian to make a) causatives from verbal stems (e.g. Luw. ta- 'to stand' ^ tanuwa- 'cause to stand', hwiya- 'to walk' ^ hwinuwa- 'cause to walk') and b) facti-tives from nominal stems (e.g. ura/i- 'great' ^ uranuwa- 'to magnify', arla- 'place' ^ arlanuwa- 'to relocate').11 It will be argued here that Luw. luhanuwata is best analyzed as a denominative factitive formation.
A cognate of the Luwian form was identified by Sevoroskin 2002: 183 in Lycian B (Milyan) 3 pl. ipv. ni ... lugatu (TL 44d.60) referring to some kind of a forbidden action (Neumann 2007: 187) which Sevoroskin translated as 'let them not burn X': this attractive
is *455-li-ia-na, the meaning of which is unknown. The word is attested in KARKAMIS Ala: § 24 ARHA-pa-wa/i kwa/i-i PES-wa/i-i-ha § 25 wa/i-mi-i-' za-i -na DEUS-ni-si-i-na | *455-li-ia-na i-zi-i-ha "when I came forth,
I made myself this X of gods" (Hawkins 2000: 88 translates 'assemblage' with reservations).
10 See Marazzi 1990: 270; Hawkins 2000: 28.
II See Shatskov 2017: 108-234 and Sasseville 2020: 462-88.
interpretation nicely matches the preceding mention of "destructive (viz. hot) autumns" (lusalija zena) which the xruwasa-deities should make quick (nuniti).
Lycian B luga--which appears to be a transitive verb — can
be analyzed as a factitive stem based on a nominal stem *luga-'burning'12, compare Lycian A kumazati 'performs a sacrifice' ^ kumaza- (c.) 'sacrificer' or Lycian B mrssxati 'desecrates' ^ mrssxa- (c.) 'offense, deceit'.13 The putative nominal stem *luga-can be traced back to Proto-Luwic *luHa- with a regular development of *H > g (cf. Proto-Anat. *HauHa- > Luw. huha-, Lyc. B xuga- 'grandfather'); Proto-Luwic *luHa-, in turn, can go back to Proto-Anatolian *louHa(H) - with monophthongization (cf. Luw. u- < *au- 'away, off or kutas(sa)ra/i- 'orthostat' < *ghou-t-) and regular loss of final *H. The same Proto-Luwic *lu a- may have served as the derivational basis of Luw. luhanuwa-: under this analysis, the middle /a/ in the Luwian form is linguistically real and represents the stem vowel of the underlying nominal stem (/luhanu-/, not /luhnu-/). Luw. luhanuwata and Lycian B lugatu can thus be shown to be derived from the same Proto-Anatolian stem.14
12 For Lycian verbs in -a-tl (Luw. -a-, Hitt. -ahh-) in general see Sasseville 2015; Sasseville 2020: 16-77).
13 mrssxa- (c.) ^ *mrsax- 'to desecrate' ( = Hitt. marsahh-), see Melchert 2015: 161; Sasseville 2020: 48.
14 It is tempting to believe that a direct reflex of Proto-Luwic nominal stem *luHa- is, in fact, attested in cuneiform sources as luha- (c.), in two of its three occurrences written with a Glossenkeil and thus likely a Luwianism. The translation 'light', going back to Laroche 1959: 63 ("lumière?", rejected by Melchert 1993: 128 and Tischler 2001: 95), finds no support in the actual texts: as CHD (s.v.) admits: "the suggested meaning "light" is based on the similarity of this word to the root of Hittite lalukkima-, luk-, and lukatta". This view can no longer be upheld: there is no evidence for a systematic relationship of Hitt. -k(k)- to Luw. -h - / -h-. But even if PIE *leuk- is left out of the picture, it is uncertain whether 2nd-millennium Luwian luha- has anything to do with the 1st-millennium Luwian luhanuwa- and Lycian B luga-. The form luh a- is listed in a "Sitzordnung" of deified abstracts in a ritual text (KUB 17.20 ii 9-12; CTH 492.1): EGIR-SU=ma hantantatar=ya ëszi, EGIR-SU=ma dusgarasza ëszi, EGIR-SU=ma :nus :dummanteias ëszi, EGIR-SU=ma kariyasha :lu-u-ha-as MU.KAM SIG5 [...] hattulatar :usass=a asanzi "behind it sits providence, behind it sits joy, behind it sits contentment, behind it sit mercy, luh a-, a good year, [...] health, (length?) of years" (ed. S. Melzer, hethiter.net/: CTH 492.1, trans. Puhvel 2001: 102), and a similar phrase kari]yashi lu-u -ha
It remains to discuss the Indo-European etymology of Proto-Anatolian *louHa(H) - 'burning'. As a "xo^-type" formation with an o-grade of the root, *louHa(H)- can be mechanically back-reconstructed as either *louh2/3ah2- or *hxlouh2/3äh2-, since the root-initial laryngeal would be expected to be lost in an o-grade environment ("Saussure Effect"). The latter option allows aligning Proto-Anatolian *louHa(H)- 'burning' with PIE *h2leuhx- 'to burn' which can be reconstructed on the basis of the following forms: Ved. rüra- 'burning hot' (< *h2luhx-ro-), Iranian *(H)rau- 'to burn' (Chorasm. pcr'wy- 'to heat up', Ossetic (Iron.) arawyn 'to burn in fire, to parch'), Gk. a^sa 'heat' (< *h2leuhx-eh2 ^ *h2leuhx-o-), (-)a^s^g 'hot', a^ssivog 'id.' (< *h2leuhx-es-), OIr. loscaid 'burns' (< *h2luhxske/o-) and possibly Latin lüstrum (if 'fire ritual').15 If Luw. luhanuwata 'he burned' and Lyc. B ni... lugatu 'may they not burn' belong to this root, its root-final laryngeal should be specified as *h2/3, to give Luw. -h- and Lyc. B -g-.
References and abbreviations
ACLT = Ilya Yakubovich (ed.), Annotated Corpus of Luwian Texts, 20132016 (http://web-corpora.net/LuwianCorpus). Beckman, G. 2001: Hantili I. Kulturgeschichten. Altorientalische Studien für Volkert Haas zum 65. Geburtstag, ed. by Th. Richter, D. Prechel, J. Klinger, 51-58. Saarbrücken: Saarbrücker Druckerei und Verlag.
CHD = The Hittite dictionary of the Oriental Institute of the University of Chicago, ed. by H. G. Güterbock, H. A. Hoffner, Th. P. J. van den Hout, & P. Goedegeburre. Chicago, Oriental Institute, 1980-. d'Alfonso, L. 2019: War in Anatolia in the Post-Hittite Period: The Anatolian Hieroglyphic Inscription of Topada Revised. Journal of Cuneiform Studies 71, 133-152.
MU.KAM S[IG5 (dat.-loc. sg.) is found in IBoT 3.83 5. KBo 3.65 rev. 4' (CTH 655) has acc. sg. :lu-u-ha-an in a broken context (see Beckman 2001: 55): the text may be saying that luha-, along with other favorable things, has been taken away from the murderers of Mursili I (nu=smas ... uiyan harkir), but this interpretation is very uncertain. All in all, luha- is unlikely to have a negative meaning related to burning ('heat', 'warmth' vel sim.) and is therefore best analyzed as etymologically unrelated to Luw. luhanuwata and Lyc. B lugatu and homophonous with the reflex of Proto-Anatolian *louHd(H) - posited in this paper. (N. Oettinger apud Melchert (forthcoming) derives luha- 'letting go, acquiescence' from the root *leuh2-, *luhT 'cut off' (Gk. Xura 'let go', Skt. lunati 'cuts', Old Norse le 'scythe' < *leuon-), see n. 7 above). 15 See Nikolaev 2021b for details.
Hawkins, J. D. 1979a: Some Historical Problems of the Hieroglyphic Luwian Inscriptions. Anatolian Studies 29, 153-167.
Hawkins, J. D. 1979b: The Hieroglyphic Luwian Stelae of Meharde-Sheizar. Florilegium Anatolicum: Mélanges offerts à Emmanuel Laroche, 145-56. Paris: de Boccard.
Hawkins, J. D. 2000: Corpus of Hieroglyphic Luwian Inscriptions. Vol. 1: Inscriptions of the Iron Age. Berlin and New York: de Gruyter.
Laroche, E. 1959: Dictionnaire de la langue louvite. Paris: Librairie Adrien-Maisonneuve
Laroche, E. 1960: Les hiéroglyphes hittites. Paris: Éditions du Centre national de la recherche scientifique.
Marazzi, M. 1990: Il geroglifico anatolico. Problemi di analisi e prospettive di ricerca. Con la collaborazione di N. Bolatti-Guzzo e un'appendice di C. Mora. Roma: Dipartamento di studi glottoantropologici, Università «La Sapienza».
Melchert, H. C. 1993: Cuneiform Luvian Lexicon. Chapel Hill, N.C.: self-published.
Melchert, H. C. 2015: Lycian alaha- and aladalehali-. Genre épigra-phiques et langues d'attestation fragmentaire dans l'espace méditerranéen, ed. by E. Dupraz, and W. Sowa, 153-163. Rouen: Presses universitaires de Rouen et du Havre.
Melchert, H. C. (forthcoming): A Dictionary of Cuneiform Luvian. Ann Arbor; New York: Beech Stave Press.
Neumann, G. 2007: Glossar des Lykischen. Überarbeitet und zum Druck gebracht von J. Tischler. Wiesbaden: Harrassowitz.
Nikolaev, A. 2010: Issledovanija po praindoevropejskoj imennoj morphologii. St. Petersburg: Nauka.
Николаев А. С. 2010: Исследования по праиндоевропейской именной морфологии. СПб.: Наука.
Nikolaev, A. 2021a: Etyma Graeca II. Indo-European Linguistics and Classical Philology 25, 953-976.
Nikolaev, A. 2021b: Vedic rurd- 'burning hot', Ossetic arawyn 'to scorch in fire', Greek àXéâ 'heat', Old Irish loscaid 'burns', and Latin lustrum 'ritual purification'. Münchener Studien zur Sprachwissenschaft 74, 109-132.
Payne, A. 2012: Iron Age Hieroglyphic Luwian Inscriptions. (Writings from the Ancient World 29). Atlanta: Society of Biblical Literature.
Payne, A. 2014: Hieroglyphic Luwian: An Introduction with Original Texts. 3rd revised ed. Wiesbaden: Harrassowitz.
Poetto, M. 1982: Osservazioni sull'iscrizione luvio-geroglifica di Aksaray. Serta Indogermanica: Festschrift Günter Neumann, ed. by J. Tischler, 275-284. Innsbruck: Institut für Sprachwissenschaft der Universität Innsbruck.
Puhvel, J. 2001: Hittite Etymological Dictionary. Vol. 5: Words beginning with L. Berlin; New York: Mouton de Gruyter.
Sasseville, D. 2015: Anatolische verbale Stammbildung: das Faktitivsuffix *-eh2-. Münchener Studien zur Sprachwissenschaft 69, 281-297.
Sasseville, D. 2020: Anatolian Verbal Stem Formation: Luwian, Lycian and Lydian. Leiden; Boston: Brill.
Shatskov, A. 2017: Hittite Nasal Presents. Leiden University Ph.D. Dissertation.
Shevoroshkin, V. 2002: Word Combinations in Milyan and Lycian Inscriptions. Studia Linguarum 3/1 ( = Memoriae A. A. Korolev dicata, ed. by A. S. Kassian, A. V. Sidel'tsev), 117-189. Moscow: Languages of Slavonic Culture.
Tischler, J. 2001: Hethitisches Handwörterbuch mit dem Wortschatz der Nachbarsprachen. 2nd ed. Innsbruck: Institut für Sprachen und Literaturen der Universität Innsbruck.
Weeden, M. 2010: Tuwati and Wasusarma: Imitating the Behaviour of Assyria. Iraq 72 ( = In Honour of the Seventieth Birthday of Professor David Hawkins), 39-62.
