SBORNI'K PRACI FILOZOFICKE FAKULTY BRNFNSKE UNIVERZITY STUDIA MINORA FACULTATIS PHILOSOPH1CAE UNIVERSITATIS BRUNENSIS A 43, 1995 VACLAV BLAZEK INDO-EUROPEAN „APPLE(S) 0. There are two different denotations of „apple" in Indo-European lan­ guages attested in more than one branch: (1) *abol- (*H2eboI-); (2) *malo- (*meH lo-). 2 1.0. IE *abol- has been usually demonstrated as an illustrative example of I- stem (Beekes 1990: 217 reconstructs the following paradigm: nom. *H eb-6l, 2 acc. Hib-el-m, gen. *Hib-l-6s); u-stem (n.) was used for a designation of the fruit while n-extension (f.) served to denotate the tree how can document the following projections on the late plE level: East Baltic Old Prussian Slavic Celtic Germanic fruit *abol(y)o- woble *ablu- *ablu-/ wabelcke *ablu-ko- *ablyo- tree *abel- wobalne *abol-ni- *abl-na- *abol(n)- See Adams 1985: 79-82, supposing an original athematic root noun. On the other hand, Hamp 1979: 158-66 prefers the originality of u-stem. He recon­ structs it also in Baltic, supposing a metathesis *abul- < *ablu. Both the authors accept the explanation of the initial long vowels in Slavic and Baltic as the out­ come of Winter's law. 1.1. Also the Oscan place name Abella from Campania described by Vergil (Aeneid 7.V.740) „...et quos nioenia maliferae Abellae..." and analyzable as *abelna- or the like, may be an Italic reflex of this etymon (but it may not - Adams I.e.). Cf. yet Latin nux avellana „hazelnut" (Cato) and late Latin (7-8 cent. AD, Closes) acer-abulus < French erable „maple" (Mann 1984-87: 1). On the other hand, all the Italic examples can be of Celtic origin (cf. Hamp 1979: 158 for Abella). 1.2. There are possible Paleo-Balcanian cognates in the following plant names: (Dioscorides Medicus 4, 182 RV) bryonia leuke" hoi de madon [=ma!on ?]... hoi de m6lothron..., RhomaToi notiam, hoi de koukourbita 6natika, Dakoi 16 VACLAV BLAZEK kinouboila...and (Pseudo-Apuleius 67, 8) A Graecis dicitur brionia,...,Itali vitis alba, alii coriaria, alii apiastellutn Bessi dinupula... and A Graecis dicitur botracion..., Romani apiastellum..., Des[si] sinupyla. Thracian (Bessian) dinu­ pula & sinupula „niilothron, Bryonia alba, wild pumpkin" and Dacian ki- nouboila id. designate the same plant as Lithuanian siin-obuolas, -obele „buckthorn, crab-tree", lit. ,,dog's apple" (Georgiev 1977: 14). Hamp 1979: 160-1 reconstructs (he following development: *0inupula < *@unipula < *Ounepula < *kyun-abolo-. The unexpected ..centum" reflex in Dacian kin- represents perhaps the proto-A I banian rule which neutralized IE palatals before resonants (Hamp, KZ 76, 1960: 275-80). 1.3. A hypothetical continuant of IE *abol- can be reflected also in Ossetin fatk'u ,,apple". Abaev (Etimologija 1966[68]: 242) separates the suffix -tk'u/- ck'u known from more plant names as nyma-tk'u „Viburnum lantana", ma- ck'u „Vaccinium vitis idaea" and reconstructs the proper root *fal-. The loss of -I- has an analogy in ma-ck'u vs. Georgian marc'qu „straw- berry", lit. „Erdbeere". The dropping out of the expected initial vowel is not rare too, cf. e.g. nyx ..forehead" < *anika- 1.4. A derivation of Pashai (Dardic) wall „apple" from the same root is only speculative without any proof (Gamkrelidze & Ivanov 1984: 639). 1.5. The most wide-spread Iranian name of „apple" can be reconstructed as *amarna- It is diffused esp. among East Iran - ian languages: Sodgian cmn'k [*amarna-ka-] .,apple-tree" (Erdal 1993: 32-3), Parachi amar, Yidgha amuno, Pashto mana „apple", maraGune pi. ,,the bitter apple", lit. „apple-like", manii ,.a fruit like sloe", Munjan amenygya, Waxi miir, Shugni itiun, dial. Bajui mun m. „apple", f. „apple-tree" (cf. Burushaski bait ,.apple" [class x] vs. „apple-tree" [class y]), Roshani mawn, Xufi, Bartangi, Oroshori mown, Sarikoli mon, Yazghulami main, Wanji mayn, Zeba men, Ishkashimi mend, Sanglichi mej < *amarnti- Ormuri milTz, milTc < *(a)mrnaci „apple" (Morgenstierne 1927: 45; Id. 1974: 44; Efimov 1986: 100; Steblin-Kamenskij 1982: 103; Edel'tnan 1986: 181). Shina phala-manra ,,apple" (Morgenstierne 1927: 45 < Tomaschek) is probably composited from Indo-Aryan and Iranian components resp. 1.6. The East Iranian origin is very probable for Turkic *alimla, attested be­ ginning 10 cent. AD in Old Uyghur alimla, similarly by Mahmud of Kashgar (11 cent. AD) alimla & almila, Khalaj alimla & alumla, Salar alima etc., Chu­ vash omla, ulma (< *alma > Hungarian alma) > Udmurt ulmo, umo, Komi ulmo, Mari olma „apple" (Joki 1964: 18). Mongolian alima(n) id. is probably borrowed from Turkic (cf. Salar alima). On the other hand, it can reflect a more archaic form yet before the supposed assimilation *alimna > *alimla. The hypo­ thetical source of Turkic and Mongolian ,,apple" was apparently some Scythian dialect where the development *amarna- > *amalna- or *amlna- was quite regular (Erdal 1993: 27-36). The Fenno-Volgaic forms as Finnish omena, omen(n)us, olma (only in Lonnrots dictionary), Wotic euna, Estonian oun, uvin. 17 INDO-I:UROPL;AN ..APPLE(S)" ubin „applc", omin ,.Pferdapfel", Livonian umar ,,apple", Mordvinian (Erzya) umar ,,id., fruit, berry" (containined with mar „berry" ?), umarina „apple-tree" (Joki 1964: 10) suggest much more archaic borrowing of the Iranian word ..apple" very near to reconstructed protoform *amama- < *amrna-. Comparing this reconstruction with West Indo-European *abl-na/i- „apple- tree", we get the forms differing only in b/m. The change *b > *in caused by the following nasal suffix looks quite naturally (Paxalina 1983: 68; Erdal 1993: 33). The name of the fruit corresponding with the West Indo-European *ablu- can be seen in Modern Persian amrud ,,pear", Pehlevi urmod, Yidgha abriio (Steblin- Kamenskij 1982: 107-8; Gamkrelidze & Ivanov 1984: 642) and perhaps „Kafir" amlukey „sloe" (Tomaschek 1880: 791). 1.6. Palaic samluwa- and Hittite samluwanza- (= GlSHASHUR-lu- wa-an- za) denotate ,,apple". Gamkrelidze & lvanov 1984: 639 try to demonstrate the connection of the Anatolian *samlu- and West Indo-European *ablu- via s- mobile (cf. Hittite sankui- „claw" vs. Latin unguis) and the change *-ml— > *- mbl- > *-bl-. Probably more promising is their comparison of the Anatolian forms with Haltic sa-a-waa-at „apple" and Kartvelian *wasl- id., suggesting the local origin (p. 641). This point of view is in a good agreement with the hy­ pothetical relationship of Anatolian *samlu- and Greek helon. brabylon, kok- kymelon. Furnee reconstructs here *vvelon (Furnee 1972: 246) but the original protoform *s\velon or *svvalon is also possible, cf. Greek hedus vs. Latin suadus ,.sweet". There is also Badaxshi (Tajik dialect from Pamir) samlu „walnut" (Tomaschek 1880: 790). But it is compatible with the (Greek-)Anatoliam iso- gloss only as Indo-Aryanism (Iranian *s- signalizes *ky-). 2.1. The second IE name of.,apple" is attested safely only in Greek (Horn., Alt.) melon, cf. pi. inela „cheeks", (Dor., Aeol.) malon. 2.2. Albanian molle „apple" (cf. pi. [in some context] molla „cheeks", mol- lishte ..apple orchard") is presumably Doracism (Muld. KZ 95,1981: 303; Mar- key 1988: 52). Accepting it, the initial protoform could be *malno-. This con­ clusion indicates a similar archetype for proto-Greek, cf. Horn, stele vs. Dor. stala, Aeol. stalla „stele" < *stalna (Chantraine 1977: 1055). Erdal 1993: 34 sees in the hypothetical proto-Greek *malno-a counterpart to Iranian forms and explains it as a borrowing from Scythian *amalna- or sim. 2.3. Similarly Latin malum n. „apple" (Plaulus), mains f. „apple-tree" (Varro), mala „cheekbone" is borrowed from Doiic malon, while the later melum (Pelronius) has an origin in Ion-Attic dialects (Walde & Hofmann 1954: 18-9, 64). An interesting fusion of Romance (malum or mela) and Celtic (Gaulish avallo) appears in Occitanian amelanco and Northern Italian (Tridente-Venetian-Istrian) amolo ,.plum" (Hamp 1979: 51). 2.4. Frequently compared Tocharian A malan, B meli reflect either *molo- or *melo- but not *malo- which have given **malan7**mali. The correct 18 VACLAV BLAZEK. meaning is only „nose", never „cheeks". That is why it cannot be a cognate (Adams 1985: 82)! 2.5. The apparent similarity of Hittite mahia- ,,branch of grapevine", Palaic mahlanza- and Greek malon & melon „apple" is not so strong not only for se­ mantics but also for the different internal structure, accepting the reconstruction above (2.2.). The comparison with Lydian molax.eldos oinou.apo tou molou hos tines. Lydoi ton oinon (Hsch.) (Tischler 1990: 89-90) looks better. 2.6. Latvian malabols „a kind of apple (serinka)" represents probably a com­ pound of the otherwise unattested *mal- and the usual abols (Karulis 1987: 135). Formally similar compounds appear in Greek melapion „a kind of apple", lit. „apple" & „pear" and Shina (Tomaschek) phala-manra „apple", cf. Old In­ die phala- „fruit" & Iranian *amarna „apple" (see 1.5.). 2.7. Yatviangian aim „apple" is probably of Turkic origin (Orel & Xelimskij 1987: 123). Crimean Tatars live in Polish- Lithuanian borderland till the present time. Cf. also (isolated) Finnish olma id. - see 1.5. Conclusions: (1) The oldest and probably only denotation of „apple" in IE with a wider distribution is *abVI-. One of possible archetypes *abul- is in a good agree­ ment with Semitic *"abul- & *'ubal- denotating various kinds of fruits and cultural trees or plants: Arabic 'ubullat- „fruits verts de I'arak, figues seches comprimees en masse", 'abal- „frais, vert, fourrage", Tigre 'obal „tamarisc", Amhara 'abalo „plante dont on se sert pour parfumer les vases a lait" (D. Co­ hen, Dictionnaire des racines Semitiques. Paris & Le Haye: Mouton 1970, 3^4). In spite of the semantic difference (but cf. Akkadian serkum „a string of (dried) fruit, normally figs, less often apples" - see Markey 1988: 54) both the terms can be connected. A cultural character gives at least an undirect evidence for a borrowing from some Near Eastern source (Semitic ?) into Indo-European lan­ guages. (2) The compound attested in Latvian malabols can represent a key to the solution of the connection of the terms 1. & 2. The second component abols means „apple". It would be natural to expect that the first component expresses some quality, cf. e.g. Greek glukumalon, -melon „sweet apple". The question of the meaning of Latvian *mal- is open. A connection with mals „clay", Lithuanian molis id., muolas „slime" is semantically more than problematic (Karulis 1987: 135). The Russian equivalent 'serinka' is apparently derived from 'seryj' „grey". It is well-known that just colors frequently represent a semantic motivation in a terminology describing „mud, marsh, clay" etc., cf. Lithuanian balas „white" : bal, „swamp, pool" or baltas „white" : Slavic *bolto „mud" etc. There is really a related color term in Baltic: Lithuanian moliugas & molinas „fallow, yellowish" (Fraenkel 1962: 463). A similar color term ap­ pears also in Greek *malos/*mallos, cf. malos 'epith. of a goat' (Theoc. Ep. 1.5.), explained as „white" (Hsch.), malloparauos leukopareios, acc. malo- 19 INDO-EUROPEAN ..APPLE(S)" parouan & maloparauan „white and chestnut" (about mare) and malouros leuk- ouros & malouns leukokerkos (Hsch.) or melis & mall's ,,yeIlow pigment" etc.; also the place name Meline designating 'earth characterized by grey clay used by painters and physicians' (Chantraine 1968: 694) (cf. Baltic!). The preceding arguments allow to identify in Proto-Greek *malo-/ *malno- some light color, most probably „ycllow" or „\vhite" describing a sort of apples. Ft is possible to imagine a development *abVI malom/malnom „[a sort of] yellowish fruit [apples ?]" > *malom/malnom „apple" comparable with Latin malum sacrum > Basque sagarr ,,apple" (Markey 1988:63-4, ftn. 5). From this point of view, Hittite mahla- „branch of grapevine" can represent originally also some sort of grape characterized by its color, esp. with regard to a distinction of vine according to color. REFERENCES: ADAMS. D.Q. 198?: The Indo-European Word for 'apple" Again. Indogermanische Forschungen 90: 79-82. BEEKF.S. R. 1990: Vergelijkende taalwetenschap. Utrecht: Aula. CIIANTRAINF. P. 1968 & 1977: Dictionnaire etymologique de la langue grecque, III & IV,I. Paris: Kliucksieck. EDEL'MAN. D.I. 1986: Sravniternaja grammatika vostocnoiranskix jazykov. Fonologija. Moskva: Nauka. RFIMOV, V.A. 1986: Jazyk ornuni v sinxronnom i istoriceskom osveSCenii. Moskva: Nauka. ERDAL, M. 1993: Around Turkic 'Apple'. J1ES 21: 27-36. FRAENKEL. E. 1962-65: Lilauisches ctymologisches WOrtcrbuch. Heidelberg: Winter. EURNEE, E.: 1972: Die wichtigsten konsonantischen Erscheinungen. The Hague-Paris: Mouton. GAMKREI.IDZE, T. & Ivanov. V. 1984: Indoevropejskij jazyk i indoevropejey. Tbilisi: Izd. Tbi- lisskogo universiteta. GEORGIEV, V.I. 1977: Trakite i texnjiat ezik. Sofija: Izd. na BSlgarskata akademija na naukite. HAMP, E. 1979: The North European word for 'apple'. Zeitschrift fur cellische Philologie 37: 158-166. JIES Journal of Indo-European Studies. .IOKI. A. 1964: Der wandernde Apfel. Sludia orientalia 28. KARUI.IS. K. 1987: K jatyjazskomu slovariku. Balto-slavjanskie issledovanija 1985: 134-140. MANN. S. 1984-87: An Indo-European Comparative Dictionary. Hamburg: Buske. MARKEY. I.E. 1988: Eurasian 'Apple' as Arboreal Unit and Item of Culture. JIES 16: 49-66. MORGENSTIFRNE. G. 1927: An etymological vocabulary of Pashto. Oslo: Dybvvad. MORCJENSTIERNE. G. 1974: Etymological Vocabulary' of the Shugni Group. Wiesbaden: Reichert. OREL. V. & XEI.IMSKIJ. E. 1987: Nnbljudenija nad baltijskitn jazykom pol'sko-jatyjazskogo slo- vaiika. Balto-slavjanskie issledovanija 1985: 121-134. PAXAMNA, T.N. 1983: lssledovanie po sravnitel'no-istorifeskoj fonetike pamirskix jazykov. Moskva: Nauka. STEHLIN-KAMENSKIJ. l.M. 1982: Oceiki po istorii pamirskix jazykov. Moskva: Nauka. TiSCin.ER, .1. 1990: Ilethilisches etymologischcs Glossar. Lief 5-6. Innsbruck. 20 VACLAV BLAZKK TOMASC'IIEK, W. 1880: Cenlralasiatische Sludien. II. Die V'amir-Dialekle. Sitzungsberichte der philosophisch-historischcn Klasse der kaiserlichen Akadcmie der Wissenschaften zu Wien 96: 735-900. WALDE. A. & HOFMANN, J.B. 1954: Lateinisches clymologisches Worterbuch. 11. Heidelberg: Winter.