Jaan Puhvel Los Angeles
CDU 801.52: 803.99-54
ON THE ORIGIN OF GOTHIC ULBANDUS CAMEL'
Great are the merits of Bojan Cop in Hittite lexicography and etymology. Glancing merely at volume 3 of the Hittite Etymological Dictionary, the latest to leave my workshop, his ideas have either provided the etyma or pointed the way to the interpretation of such important items as halkuessar, halluwa-, halzai-, hattalkesna-, hazzila-, huek- (both verbs!), and hu(wa)nhuessar. May the small token of homage that follows be worthy of his endeavors!
The Semitic word for 'camel', exemplified by Hebrew gamal, makes its Greek debut as KajiriXoq in Aeschylus (Hiketides 285); via Latin camelus it has conquered the modern west. In the other direction it appears, with some apparent interference from the verb kramate 'stride', as Sanskrit kramela-, supplementing the old Vedic term ustra-, of Indo-Iranian attestation (cf. Avestan Zarad-ustra-, either 'Old-camel' or 'Camel-driver'). Unlike such relatively neat and short terms, the ungainliness of the beast is more picturesquely expressed by the designations for it in Germanic, Slavic, and Baltic. Starting with Gothic ulbandus, the Germanic attestations are ON ulfalde, OE olfend, OS olbundeo, and OHG olbento. The Slavic forms (OCS velibQdu, Russian verbljud, Polish wielblqd) appear to have been folk-etymologically twisted in the direction of velii 'big, great', whereas the Baltic variants (OPr. weloblundis, misren-dered as 'mule' by the Elbing Vocabulary [line 437]; Lith. velbliudas) are patent borrowings from Slavic.
Received wisdom1 presumes for ulbandus and company "same origin" as OE elpend, OHG elafant, i.e. 'elephant' borrowed from Latin elephas {elephant-) which in turn comes from Greek eteqxxg {¿XeqxxvT-). A "transfer of meaning" ('elephant' > 'camel') is assumed, but the best parallel produced is a dialectal Arabic bil for 'camel', allegedly a variant of the old word for 'elephant' which ranges from Akk. piru or pilu 'elephant' to Old Persian piru- 'ivory' to Aramaic and Persian pil, Sanskrit (late) pilu-, Arabic ./W, and Old Norse fill 'elephant'2. Even this is no parallel at all3, for the Ger-
1	E.g. W.P. Lehmann's recasting of S. Feist's Gothic Etymological Dictionary (Leiden, 1986), 375. In the same vein still C. Sandoz, Latomus 48 (1989), 753-764.
2	Also mpi(J<7CCg 'elephant' (Hesychius), perhaps 'elephant cow' like Akk. pirati or Skt. pillaka.
3	Weaker still is the alleged derivation of Russian slon 'elephant' from Turkic a(r)slan 'lion'; Russian Ruslan is well taken as such in origin, but slon is better traced back to Austroasiatic words for 'elephant' (cf. T. Gamkrelidze and V.V. Ivanov, Indoevropejskij jazyk i indoevropejcy [Tbilisi, 1984], 524-525). Not so, however, Tocharian B onkolmo (A onkalam) 'elephant', whose source V.V. Ivanov (Etimologija 1980 [Moskva, 1982], 166) also surmised in Austroasiatic. In onkolmo may lurk rather an Indo-European
187
manic words for 'camel' are no stray dialectal semantic deviates but engrained across early Germanic before the intrusion of the Greco-Latin word for 'elephant'. One would have to assume instead that some remote proto-designation existed from which evolved semantically 'elephant' on the one hand, 'camel' on the other. But the earliest attainable meaning of éXéqxxvz- is rather 'ivory': e-re-pa-(te-) is well attested as an ornamental material in Mycenaean Greek, but the beast itself was unknown in Europe, being mentioned first for Ethiopia and Libya by Herodotus (3.114,4.191), and described for India by Aristotle (Hist. Anim. 2). Syro-Phoenician sources of Euphratic ivory were accessible to the Greeks earlier than the animal made a mental imprint, even as in Italy Punic mediation produced the ebur (from a Hamitic term seen in Egypt. 3bw 'elephant' and 'ivory')4 way before the Luca bos or bos Luca that was militarily introduced to Italy ahead of Hannibal by Pyrrhus in 280 b.c.e. and later described in lumbering detail by Lucretius (5:1301-2). Clearly, therefore, éXécpavT- entered Greek as a term for ivoiy, rather than as a zoonym; only later did metonymy include the animal source of the product5.
Additionally, the discovery of lahpa- as the Hittite term for 'ivory' (matching Akk. sinnu at Ugarit)6 has clarified the sources of éÁéqxxvr-, It did not stop Gamkre-lidze and Ivanov from Indo-European etymologizing (still drawing in ulbandus, etc.)7, and may have led Laroche8 to overstress similarities to the Semitic terms for 'ox'
compound 'snake-handed' (cf. Lucretius 2:538 anguimanos elephantos, 5:1301 boves Lucas ... anguimanos), made up of cognates of Lat. anguis and palma (Oír. lam 'hand'), i.e. *anghwi-p¡Amon. This makes more sense than A.J. Van Windekens' *anq-olmo(n) 'bent-being' (i.e. with curved trunk; Le tokharien 1 [Louvain, 1976], 337) or J.E. Rasmussen's *h2mbhi-kwlhimah2 (allegedly elephant as 'servant', similar in form and meaning to Gk. ájMpínoXog. Tocharian and. Indo-European Studies 2 [1988] 166-183; Acta Lingüistica Hafniensia 21 [1988], 92). Cf. J. Puhvel, HS 106.
4	Also in the Hebrew hapax sen-ha-bblm 'tooth of elephants' (usually simply sen 'tooth' or qarnotsen 'horns of tooth' for 'ivory'). On the increasingly dubious relevance of Skt. ibha- 'elephant' see M. Mayrhofer, Etymologisches Worterbuch des Altindoarischen (Heidelberg, 1988), 194.
5	In languages closer to elephant habitats the distinction was clearer from the start: Vedic for 'elephant' was mfgóhastín- 'beast with hand', and 'ivory'was Skt. hástidantam. Sumerian had am.si 'elephant' {am sign seemingly made up of gud + kur 'mountain ox', si 'horn'; thus 'horned mountain ox') and ka+ud am.si 'ivory' (literally 'tooth of elephant', like Akk. sinnu [pin]). In modern languages such derivation has reasserted itself (e.g. German elfenbein, Danish filsben).
6	See Chicago Hittite Dictionary 3.1 (1980), 12-13; J. Tischler, Hethitisch.es etymologisches Glossar 2 (1990), 14-15. J.-C. Billigmeier (A Linguistic Happening in Memory of Ben Schwartz [Louvain, 1988], 321-331) suggested that Lat. Luca bos is in reality lucabos, a late Asianic ("neo-Hittite", perhaps Lycian) survival of lahpa- (presumably meaning 'elephant' in addition to 'ivory'), and that bos Luca was a Roman folk interpretation. This hunch puts a lot of faith in Pyrrhus' Hellenistic army employing eastern mahouts who instantly taught their exotic term for 'elephant' to the Italic natives under battle conditions. More probably bos Luca is a genuine "nativism", like Sumerian am.si (see fn. 5); cf. R. Arena, Rivista difilología 116(1988), 185-197.
7	Loc.cit. in fn. 3. Lehmann (see fn. 1) erroneously claims that Th. von Grienberger (SOAW 142/8 [1900], 225) denied an early relationship of ulbandus and elephant-; G. did exactly the opposite, postulating an Indo-European prototype of both.
188
(Akk. alpu, etc.), but it established a likely term for 'ivory' in the Syrian-Anatolian-Helladic orbit of the latter half of the second millennium, one that has no plausible Indo-European origin or connection with 'camel' words such as Gothic ulbandus.
What then of the latter? Otto Schräder9 had a thought on ulbandus which deserves to be quoted: "An einem Zusammenhang mit lateinisch elephantus ist schwerlich zu denken. Vielleicht gehört got. ulbandus wie andere undeutbare Kulturwörter dieser Sprache (z.B. peikabagms 'Dattelpalme', smakka 'Feige', bairabagms 'Maulbeerbaum') zu dem Sprachgut, das die Goten nach ihrer Ankunft am Schwarzen Meer von ihren Beutezügen in Kleinasien heimbrachten".
This suggestion may have been an inspired one, for it appears that Hittite supplies what may be either a borrowing source or a cognate for ulbandus. There is the word huwalpant- which means 'humpback', as in the teratogeny lists of birth omina (KBo XIII34 IV 3-4 takku SAL-za häsi n-as [...] huwalpanza 'if a woman gives birth and [the baby] [is] a humpback...'). There is further the derived noun hu(wa)lpanzina-, hulpanzana- 'hump, protuberance, embossment', both in an anatomical sense (KBo XIII3116) and referring to metal ornamentation (e.g. KUB XLE 69 Vs. 25-26 GfR vmHatta arahzan[da... [...] 30 hulpanzanas GUS[KIN 'swords of Hatti, all around thirty embossments of gold'; these were of the type of the Homeric %i(pog [or (pacrya-vov] äpYüpörjXov 'silver-studded sword'). From a denominative verb hulpanzinai-'emboss' there exist both Hittite and Luwian-type passive participles, e.g. KBo II 1 I 38 1 urudiqjr TA GUSKIN huwal<pa>zinan 'one bronze dagger embossed with gold'; KBo XX 47, 11-12 nu-za LUGAL-ws ... [hu]lpanzinanda wasiyazi 'the king dons embossed regalia'; KUB XII1IV 34 GU]SKIN hulpanzinaimes '[shoes] embossed with gold'10.
The external similarities of hu(wa)lpant- and ulband- require little comment. The Gothic «-stem is unexpected; considering the Luwian elements such as the participle hulpanzinaima-, one would have looked for an i- stem *hu(wa)lpanti- as perhaps underlying the derivative hulpanzina-. The camel being called 'humpback' is self-explanatory. As Indo-European reconstruction can serve *Hwlb(h)o- related etymologically to Skt. ülbam, Lat. volba 'uterus, womb', and perhaps to Hittite huelpi- 'newborn'. The question whether ulbandus is a borrowing from Anatolian into early Germanic or cognate Indo-European remains open; perhaps the typically Anatolian -nt- suffix tends to tilt the balance in favor of borrowing, thus giving Schrader's intuition its full due.
8	Revue de philologie 39 (1965), 56-59; cf. E. Masson, Recherches sur les plus anciens emprunts sémitiques en grec (Paris, 1967), 80-83.
9	Reallexikon der indogermanischen Altertumskunde, 2nd ed. (Berlin, 1917-29), 1:553.
10	For full data, see J. Puhvel, Hittite Etymological Dictionary 3 (Berlin, 1991), 424-426.
189
Povzetek
IZVOR GOTSKEGA ULBANDUS 'KAMELA'
Slovanska in germanska beseda za kamelo (npr. gotsko ulbandus in starocerkvenoslovansko velibodu) se navadno povezujeta z besedo elefant 'slon' (grško ¿Xe(pavT-). Ta povezava je sporna: ¿Xi(pavT- spada k hetit-skemu lahpa- 'slonova kost', vtem ko je sorodnik besede ulbandus hetitsko himalpant- 'grbavec' in njegove izpeljanke s pomenom'grba'ali'izboklina'.
190