Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.imk.msu.ru/Structure/Linguistics/yakubovich/download/upe.pdf
Дата изменения: Fri Apr 30 15:11:58 2010
Дата индексирования: Mon Oct 1 20:49:00 2012
Кодировка:

Ilya Yakubovich
CARIAN MONUMENT
The Carian noun that has the highest frequency in the existing corpus of texts is upe/ue `monument'1. The following inscriptions, all found in Egypt, supply us with its attestations: M5 sdar|upa|З|Зeti|mЗdoni M9 sarnaiupe|quqemimdawn M14 artaЫ:upe:[ M18 ]u|upesa|triel|mrsi M20 saruq|ue|pntmunimЗdoni M28 Зksmu|Зpe|lkorМ|qarpsi M30 swin|upeari[e]ited M33 ·or|Зpe|qdarou|tratar M34 arМom:ue:mЗsat:imudon:i tbridb:i M35 lusi|upe|arquqi|ksolbarliom|mnoi| M43 arlis|psikron|ue MY D pikreuesarЗlМatmsnord MY E terwe|upe|nuo[ ]sarmroЫt MY F psmskЗneit|ue|narМa|suKiq|sara In all the cases listed above we are dealing with actual inscribed stelae. The first group of inscriptions is especially instructive: they were found together in what originally was a temple precinct of the Late Egyptian period located at North Saqqara, near Memphis. It is clear that they were brought from a nearby place as a building material, and a cemetery of Carian mercenaries and their families naturally comes into mind as a source of such an impressive array of Carian monuments. This hypothesis is supported by the apparent funerary motives of sculptural compositions carved on the stelae2. On the other hand, at least one of the stelae of the second group was defined as stХle de donation3.
The Hittitological abbreviations used below are those of the Chicago Hittite Dictionary. I am obliged to Th. van den Hout, A. Kassian, and H. C. Melchert who took pains to read the first draft of this manuscript and made valuable remarks, as well as to Th. Wier who improved its style. Usual disclaimers, of course, apply. Prof. Melchert informs me that he has independently arrived to similar conclusions, although we differ in certain important details. H. C. Melchert's article devoted to the Luwian stem uba- will be published soon in another Festschrift. 2 H. S. Smith, Foreword to (Masson 1979: VVI). The author of the Foreword further hypothesizes that the destruction of the Carian cemetery could be connected with the rise of nationalistic feelings after the end of the Achaemenid Persian rule in Egypt in 404 BC. 3 Masson, Yoyotte 1956: 18.
© Ilya Yakubovich, 2005 Hr d mА n a s . 70- . . . . . -: «», 2005.
1

Carian monument

241

Moving to the inscriptions, in all but two cases the word u(p)e is preceded by a genitive case noun ending in -, and all of these nouns that succumb to identification represent personal names.4 The same holds for the overwhelming majority of words that immediately follow u(p)e; the analysis of other Carian inscriptions makes it very likely that most of them represent extraposed patronymics. Thus the inscriptions begin with a stereotypical phrase `u(p)e of X son of Y'5. It is only natural to assume that the noun u(p)e itself denotes `stele, monument'. Any other suggestion, such as `grave' or `votive gift', would divorce some of the stelae under consideration from their archeological and art historical context, and therefore must be a priori regarded as inferior6. Given the similarity of the sound shape of upe and ue, as well as their identical textual contexts, it is counterintuitive to presume that they represent two different words referring to two different kinds of funerary monuments. The obvious problem one needs to address is reconciling the divergent spellings of upe/ue and reconstructing its phonology. The phonological shapes of the second millenium Anatolian languages exhibit a binary opposition between voiceless/fortis and voiced/lenis consonants7, and there are reasons to think that some of the original lenis consonants were intervocalically realized as voiced fricatives in Lycian and Lydian8. Extrapolating this development to Carian, one can hypothesize that its inventory of labial stops consisted of a voiceless stop /p/ and a voiced fricative /b/ (phonetically [b]~[]). If we had only one grapheme for these two sounds, the reconstruction [ue] would neatly account for the scribal variation upe/ue. In the second case, one would argue, [] was (wrongly) perceived as a secondary epenthetic consonant. In the reality, the situation is more complicated. Carian does have a grapheme b that is normally used to render a sound that correponding to ("beta") in Greek foreign transmission. It is logical to assume that Carians would have used the spelling **ube for a word that sounded [ue]. If one wants to maintain that this word was written as either upe or ue, one must explain why the natural spelling **ube was proscribed in Carian. The personal majority The full explanation of this phenomenon was made available by SchЭrr's discovery of Carian names containing the Egyptian theonym Bastet.9 It appears that these names cover the of cases where the graphic sequence -ub- is used in Carian inscriptions from Egypt. documentation is provided below10:

Cf. e.g. arlis- vs. Gk. (Adiego 1993: 230231), artaЫ- vs. Gk. (ibid.: 231232), lwsi- vs. Gk. (ibid.: 232). 5 Carruba (1998: 55) took ue and upe as enclitic particles going back to *=wa and *=wa=pa respectively. One can make several objections against this theory. First, the enclitic =pa is preserved as a second element of Car. sb `and', which is never written **sbe. Second, the combination |upesa| (M18), convincingly analyzed in Hajnal 1998 : 86 as `this stele', can hardly be explained within this approach. Third, Carruba does not explain why the sequence upe/ue does not appears only on those inscriptions that are engraved on stelae. 6 Similarly SchЭrr 1992: 155 and Vittman 2001: 40. 7 Melchert 1994: 1321. 8 Ibid.: 301, 338. Phonetically speaking, the development was threefold. The etymological voiced stops were preserved as such in a position after nasals. On a phonological level, however, they can be regarded as allophones of voiceless fricatives in Lydian and Lycian (Melchert 1992: 38, fn. 13). 9 SchЭrr 1996: 5571. 10 The list is based on Adiego 1993: 251, complemented by the new readings of Vittman (2001: 5859).

4

242

Ilya Yakubovich 1. 2. 3. 4. 5. 6. 7. MY A piub[a]i-; Ab. 6 F piube Ab. 10 F pdube Ab. 19 F, Ab. 28 Y ttubai; MY A ttbai-[] MY b pduba; Ab. 4 F pdubi- M 41 loubaЗ Th. 55 S ludtuba M 10a tasubt-

The names of the first three groups have been convincingly analysed as Pa-B#st.t ~ Gk. `He of Bastet', P#-dj-B#st.t ~ Gk. `(He), given by Bastet' and T#-djB#st.t ~ Gk. `(She), given by Bastet' respectively.11 The nouns of group 4 can be explained as Carian hypochoristic formations of pdube `(He), given by Bastet'12. Thus 7 out of 10 names containing the sequence -ub- appear to refer to the Egyptian cat-shaped goddess. None of the remaining three names has an established etymology. No Egyptian sign corresponds to the Carian vowel -u- preceding the divine name ba(i), be. Ray thought that it represents a sort of anaptyctic vowel coloured by the following labial13, but this explanation does not square with the fact that it can appear also in Greek transmission of Egyptian names beginning with Bastet14, as well as in the Coptic reflex of the same name OUBECTI. It seems that one would fare better accepting the argumentation of жzing15, who reconstructs *ubХ st° < *wbХ st° < *bwХ st° < *buwХ st° < *bu#Х st.°t . It is possible that Carian ubai, -ube reflects an intermediate stage of this development, where the sequence -ubreflected the combination [wb], or something similar. Under this presumption one can see why Carians would feel uncomfortable using the same sequence -ub- in the word [ue], and had to render it as either upe or ue 16. It is obvious that my suggestion represents no more than an educated guess, unless one can provide external evidence supporting the existence of Carian [ue] `monument'. Below I shall try to demonstrate that this word has very suggestive cognates in the Luwic (South Anatolian) subgroup of the Anatolian group of languages. The latest etymological statement about Car. upe `monument' belongs to I. Hajnal, who compared it on the one hand with Hitt. uppa- `to bring' and, on the other hand, with Lyc. uba`gift' with the verb ube- (presumably `to dedicate, offer')17. A different approach is implied in the Cuneiform Luwian Lexicon of C. Melchert who separates CLuw. uppa- `to bring' (and, by
For the Egyptian correspondences cf. Vittman 2001. Differently SchЭrr (1996: 60), who compares these names with a Lycian PN hrppidubeh (gen). In my opinion, the analysis of this name as hrppi-dube-h is preferrable to that of hr-ppid-ube-h, and so it is no match to the Carian names under consideration. 13 Ray 1998: 129. 14 E.g. B#st.t-о .о r-tj-s, lit. `Bastet is one who gave her' is rendered as ' in Greek transmission (E. LЭddeckens et al. 19802000, 1(2): 136) . 15 жzing 1976, Vol. 2: 856857, fn. 1319.
12 11

Carians also had another way to render a sequence /-u -/ in their writing, namely -wb-, cf. a Carian PN katwbr rendered as in Greek transmission. The second part of this name seems to go back to the substrate verb a ar- `to rule', on which see Yakubovich 2002. The first vowel of this root was apparently labialized under the influence of the following consonant in this Carian name. It seems, however, that the Carian sign w- was never used in Egypt with the value /u/ in initial position. 17 Hajnal 1998: 82. Hajnal's information about Lyc. uba `gift' is based on Melchert 1993a. This is clear from the fact that he repeats Melchert's erroneous reference to TL 44 c 14 for this word, whereas it is in fact attested in TL 44 c 4.

16

Carian monument

243

extension, Hitt. uppa- `id.') from CLuw. upa- `to furnish, grant' (and, by extension, Lyc. ube`dedicate, offer').18 None of the two authors, however, used the Hieroglyphic Luwian data in order to clarify the relationship between the two Luwian verbs and the Carian word for `monument'. The publication of the monumental Corpus of Hieroglyphic Luwian Inscriptions (Hawkins 2000) has helped me in the solution of this "monumental" problem. The following contexts bear witness to the existence of Luw. u-pa- `to place, found, establish': (1) ARSLANTA (Tell Ahmar)19 §2 a-wa/i ha-ta-ta-na(URBS) ||URBS-MI-ni-na "1"-ti-i ("ANNUS")u-si-i ("AEDIFICIUM")u-pa-ha-' `I founded the city Hatata in one year'. (2) SKENDERUN (Gurgum) § 3. a-wa/i | za||-na-i ("*255")ka-ru-na-na ("CAPERE")u-pa-ha-i § 4. wa/i-tu-u-ta-i 4вMILLE 4вCENTUM a-ta ("CAPERE")|| u-pa-ha `I founded this granary, and I placed into it 4400 [here, by zipatani-measure] (3) KARKAMS A11b+c (Building insciption of Katuwas) §12. a-wa/i pa-ia-' | REGIO-ni-ia ("VACUUS")ta-na-tА-ha §13. wa/i-ta-' (SCALPRUM.CAPERE2)u-pa-nМ-zi a-tА | ("CAPERE2") || u-pa-ha `I wasted those countries and erected trophies in them'. (4) KARKAMIS A1a, (Building inscription of Katuwas) §9. | (*349) А-la-ta-ha-na-ha-wa/i(URBS) |ARHA | DELERE-nЗ-wa/i-ha §10. |wa/i-tЗ-ґ | pa+ra/i-i-ha-ґ (SCALPRUM.CAPERE2)u-pa-nМ-na | (CAPERE2) u-pa-ha §11. |9-za-ha-wa/i-tЗ | pi-ia-ha `I destroyed the city Alatahana. I erected a trophy before him (i.e. Tarhunzas) and gave him a ninth '. (5) KARKAMS A2 (Building Inscription of Katuwas) §7 wa/i-ta-' mi-ia-za-' LITUUS+AVIS-ta-ni-ia-za |REGIO-ni-i a-tА (DEUS)BONUS-na (DEUS)VITIS(-)ti-PRAE-ia-ha ||ARHA (CAPERE2)u-pa-ta `In my times he (Tarhunzas) established again the Grain-god and Wine-god in the country'. J. D. Hawkins admitted the existence of HLuw. u-pa- `to found' in (1) but translated the rest of the verbs cited as `to bring', implying the comparison with Hitt. uppa- `id.'. It is difficult, however, to understand how one can "bring" a granary, and bringing trophies to a defeated country also defies common sense. The determinative CAPERE, normally used for Luw. ta- `to take' and representing an image of a grasping hand, is much closer in form to PONERE accompanying Luw. tuwa- `to put' and representing an outstretched hand, than to the foot-shaped determinatives PES and PES2 defining verbs related to motion.20 As for the determinative CAPERE2 consistently employed in Karkamis, it was descibed already by Laroche as `main qui

Melchert 1993b: 242. All the Luwian contexts cited below can be found in Hawkins 2000. The translation, however, is my own. 20 Including transitives, such as e.g. PES2 hinu(wa)- translated by Hawkins as `cause to pass' (Hawkins 2000: 125126).
19

18

244

Ilya Yakubovich

pose'21, and the transcription PONERE2 would probably do more justice to its shape. Last but not least, the noun (SCALPRUM.CAPERE2)u-pa-nМ-, translated as `trophy' also by Hawkins, probably indicates a kind of stone monument commemorating the victory. It is hard to see how a word of such a meaning can be derived from a root meaning `to bring', whereas the derivation from u-pa- `to place' is closely paralleled by that of Gk. `monument, stele', derived from Gk. `stand', but also, `place, establish'22. Going somewhat ahead, I believe that Car. /ue/ was also derived from /u/ meaning `to place'. It would be, however, unfair to claim that my interpretation is entirely unproblematic. For example, § 4 of (2) could be rather translated as `I brought into it 4400...', if not for the determinative CAPERE.23 The following group of examples containing the determinative PES can be used as an additional argument for the existence of u-pa- `to bring' in Luwian. (6) BOYBEYPINARI 1 (Kummuh) §1 [z]a-wa/i (THRONUS)i-sЮ-tara/i-tА-za za-ha MENSA-za mu 1pa-na-mu-wa/i-ti-sa PURUS.FONS.MI-sa IUDEX-ni-sa FEMINA-na-ti-sa PONERE-wa/i-ha §4 (PES)u-pa-tА-pa-wa-tЮ[-' ] IА-za-mi-i-sА PURUS.FONS.MI IUDEX-ni-sa SERVUS-ta5-sa || ``I, Panamuwati, the wife of the ruler Suppiluliuma established this throne and this altar..., while Azamis, the ruler Suppiluliuma's servant, brought them. Cf. BOYBEYPINARI 2 §5 za-pa-wa/i i-sЮ-tara/i-ta-za || za-ha MENSA-za А-mi-sЮ || tА-ti-sa А-za-mi-sА ... || (PES) u-pa-tА §6 wa/i-tЮ PONERE-wa/i-ta `And this throne and his altar my father Azamis brought and established. (7) SULTANHAN (Tabal) §2 | a-wa/i | za-a-na | (DEUS)TONITRUS-hu-zА-na | tu-wa/i+ra/i-sЮ-si-i-na | ta-nu-wa/i-ha || §3 | a-wa/i-sa(!) | А-pi-i | CRUS-nЗ-wa/i-mi-i-na | BOS(ANIMAL)-ri+i-i 9 OVIS a+ra/ima-sa- ri+i-i §4 | a-wa/i-na | u-pa-ha | HWI-i §5 |wa/i-sА | OMNIS-mi-ri+i | [sa]-na-wa/i-sa-tara/i-ri+i | А-wa/i-tЮ-' `I established this Tarhunzas of the vineyard, providing with an ox and nine monthling sheep. When I placed him, he came to me with all goodness'. Cf. below: §10 | ta-nu-wa/i-ha-wa/i-na | REL-i §11 | REL-i-pa-wa/i | (TERRA) ta-sЮ-REL+ra/i | 2 "OVIS"-sa 80 "HORDEUM" CRUS+RA/I || §12 | wa/i-na-' | А-pi-i |zi-na |"AVIS"(-)ta-wa/i-na-ri+i | ("PES")u-pa-ha `When I established him, and 2 sheep costed 80 [measures of] barley, I provided him then with tawani-birds.

Laroche 1960: 43 (sign # 43). E.Laroche explicitly translated (*43)u-pa- as `to found'. It is interesting that Gk. - was apparently borrowed into Lycian only in its technical meaning `to be erected, set down'. Lyc. sttati sttala `a stele will be set down' (TL 44 c 5; Melchert 1993c: 31) represents a figura etymologica comparable to Luw. ubaninzi ubaha in (3). 23 It is possible that a scribal error of perseveration is responsible for the occurrence of CAPERE in this clause.
22

21

Carian monument (8) TOPADA (Tabal) §5. А-mu-ha?-wa/i REX+RA/I-ti (ANIMAL)EQUUS-wa/i-ti u-pa-ha `I provided mysef with the "Royal Horse"'. Cf. below: §15. *274-ia-pa-wa/i FEMINA.MANUS-zi/a-ha SERVUS-sa ("PES")u-pa-tax. `It brought the militia24 and women-and-children into slavery25'.

245

The determinative PES that appears before many attestations of u-pa- in the passages cited suggests that this verb was synchronicaly connected with motion. In some passages the meaning `to bring' makes indeed a perfect sense; thus one can analyze (6) as an account of a queen who sponsored the creation of a new altar, and her servant who actually supervised its transportation, installation, and inauguration. In some other cases, however, u-pa- `to establish' seems to match the context better; thus А-pi-i ... ("PES")u-pa-ha `I provided (vel sim.)' in (7), §12 can be compared with А-pi-i | CRUS-nЗ-wa-mi-i-na `provided (acc. part.)'26 in (7), § 3. If the basic meaning of u-pa in this context was similar to that of (CRUS) tanu(wa)- `place', it is easier to explain why their derived meanings27 are also similar. Summing up, Hieroglyphic Luwian seems to provide the evidence for two originally different verbs, (AEDIFICIUM/CAPERE/ CAPERE2) u-pa `to place, found, establish' and (PES) u-pa `to bring', even though it is not always possible to distinguish between the two meanings in individual cases. This situation can be explained if we assume that both verbs merged in (some dialects of) Late Luwian that underlies Hieroglyphic inscriptions of the first millenium. In fact, the same hypothesis is also entertained by Hawkins (CHLI, 259), even though our analyses of the individual attestations of this verb are different. Another important conclusion is that u-pa- does not seem to mean `to offer, dedicate' in Hieroglyphic Luwian texts. A least, only in (4) has the beneficiary of the erected stele been mentioned. Moving to the Cuneiform Luwian texts that illustrate the state of the same language some five hundred years earlier, we see that the two verbs under consideration were also phonetically different. Laroche (1959 : 1012) was still willing to consider CLuw. З-pa- as a variant of CLuw. up-pa- `to bring', but Melchert (1993b : 242), following Starke (1990 : 520, fn. 1921) set CLuw. З-pa- aside, assigning to it a different meaning `to furnish, grant'. CLuw. up-pa- `to bring', obviously cognate with Hitt. up-pa- `id.' was never regarded as semantically problematic28, but the semantics of CLuw. З-pa- deserves further discussion. All the available contexts are listed below29:

For the translation of 274-ia cf. Marazzi 2002: 509. I tentatively take SERVUS-sa for a free-standing genitive of a word meaning `slavery'. For free-standing genitives in factitive constructions in Hittite cf. e.g. Yoshida 1987: 8. 26 Differently Hawkins (2000: 465466) where CRUS-nЗ-wa-mi-i-na is translated as a 1 pl. form `we presented'. Hawkins does not comment on functional motivation for the alternation between 1 sg. and 1 pl. in the same inscription. 27 I believe that both verbs were modified by the preverb api cognate with Gk. ' (cf. Ivanov 2002: 5) and responsible for the associative semantics of the phrasal verbs. 28 A clear occurrence of CLuw. uppa- `to bring' is available in (10). 29 All the four contexts below are translated in Starke (1985: 132, 226, 278, 65 respectively). Starke's translation is similar, but not identical to mine.
25

24

246

Ilya Yakubovich (9) KBo 9.145. left 2ff , CTH 770 (Unidentified Luwian text), NS 2' [ ]в-in З-pa-an-ti a-an-t[a 3' [du-]pa-in-ti a-ah-ha-ti-ya-a[n 4' [A]N.BAR-ti tar-ma-ti tar-mi-ta hu[ `They u-pa-, ... strike.., when .. it ... (he) nailed with iron nails...'. (10) KUB 35.88 III 11ff, CTH 765 (Luwian Birth Ritual), NS 9' 10' 11' 12' 13'
30

l[a-la-a-at-ta 9-za GISGA.ZUM-za ha-la-ta-pa DZu-li-ya-ya-a[n З-pa-at-ta-pa-wa-du sa-ar-ri-ya-ni-in 2-su 9-u[n-za an-ta-wa-as-ta wa-al-lu-na--sa-an wa-a-ni up-pa-an-t[a sa-ar-ri-wa-tar DUMU-in wa-al-li-it-ta

`He took nine combs and hala-ed? Zuliya. He u-pa-ed? for him sarriyani, two times nine. They brought them to the woman of lifting. She lifted the child up'. (11) KUB 35.133 II 27ff, CTH 772 (an Istanuwian ritual), NS 27' 28' 29' 30' a-a-wa DU-an tu-З-is UD.KAM-is ta-at-ta us-ma-an-ti-in su-na-at-ru-wa-an-ti-in З-pa URUha-at-tu-sa-ya ap-pa-ra-an-ti-en a-ri-in an-na-ru-ma-a-hi hu-u-it-wa-la-a-hi-sa-ha З-pa

`Lo, Storm-god, your day is appointed. Make future day(s) in Hattusa full of wealth and abundance! Establish vitality and life!' (12) KUB 35.54 III 17ff., CTH 758 (Ritual of Puriyanni for the purification of a house), NS. 17. 18. 19. 20. 21. [ [ [ [ [ w]a-a-ar-sa-at-ta мD-ti [ ]в-am-ma-an M]UN-sa-pa a-a-la-a-ti u-wa-a[-ni-ya-ti] З-pa-am-ma-an w]a-a-ar-sa-at-ta zi-i-l[a мD-i] an-da n]a-a-wa i-ti MUN-sa-pa-a[t-ta z]i-la a-a-]li-i u-wa-a-ni-ya na-a[-wa i-t]i

`Rain is ... by the river and salt is provided by the steep saltlick. Rain will not go again into the river and salt will not go again into the steep saltlick'. The difficult passages (9)-(10) do not allow for any positive conclusions, but examples (11)-(12) show the same basic meaning of З-pa-, that is `to cause'. Both `to bring' and `to found, establish' could in principle be used metaphorically with a meaning `to cause', but phonetic considerations tip stakes in favour of the second alternative. One should keep in mind
30

The reconstruction of broken lines is based on a parallel passage KUB 35.89 10ff

Carian monument

247

that the Cuneiform Luwian corpus, almost exclusively limited to incantation rituals, does not represent a direct match to Hieroglyphic Luwian building inscriptions. Both genres illustrate the complementary meanings of Luw. *uba- that is physical construction and metaphysical creation. If this analysis is correct, `to furnish, grant', implying an underlying beneficiary, does not appear to be the best translation of the Luwian verb. In fact, only the obscure example (10) features the dative pronoun as an argument of CLuw. З-pa-. If the verbs upa- `to bring' and uba- `to found' were phonetically distinct in Cuneiform Luwian, then the hypothesis of their merge in (some dialects of) Hieroglyphic Luwian crucially depends on assuming special sound changes at that period. One of the possible solutions in assuming a frontal change of intervocalic -p- to -b- in (some dialects of) Late Luwian. I should like to stress that this change must be considered distinct from the common Luwian lenition, which was dependent on the quality and accentuation of adjacent vowels31. It goes without saying that, in the lack of independent evidence, this suggestion remains purely hypothetical. What one can claim with much more certainty, is the fact that CLuw. uba- does not result from a recent dialectal lenition of upa-. In my opinion, Luw. b `to found' is attested in Early Luwian ubdi(t)- `estate, holding, demesne' that was borrowed in Old Assyrian, Hittite (OS+), and, later, in Ugaritic. The different attestations of this group of words were discussed at length by Beal (1992: 539549), and there is no need to reiterate this discussion here. It is enough to say that Hitt. ubdi- , and its Luwian counterpart u-pa-ti-t° can be used either with reference to a land holding, which is large enough to be mentioned in a political document, or to a military unit, presumably levied in a particular area32. The Ugaritic ubdy `land, plot, farming (in a taxed leasing system)'33 shows that the same Luwian word could also be used on a smaller scale. Finally Akk. ubadinnu, attested exclusively in Old Assyrian economical texts, apparently designates a service-unit, a group of people living in a paricular area and having certain common responsibilities. This semantic amalgam can be plausibly explained if we assume that the original meaning of Luw. ubdi(t)- was simply `estate'. In its ordinary use, this word could designate a plot of land that was regarded as an agricultural unit, and the maintenance of which required some communal labour (for our present purposes it is irrelevant whether it was free or corve'e labour). The same term could be also metonymically used for a group of people who were living on this plot of land. When used in a political or diplomatic context, the same word could designate the land holding of an Anatolian ruler. The personal army of such a ruler would be formed from the men of his ubdi(t)-, and, therefore, could be metonymically designated as ubdi(t)-. Finally, if this prince was a vassal of the Hittite king, his personal retinue could be regarded as a detachment of the Hittite army, and therefore ubБdi(t)- could occasionally function as `military unit'34.

See Morpurgo-Davies 19821983: 245 ff. for the inductive formulation of the Proto-Luwian lenition of dentals. 32 For the list of attestations see Starke 1990: 195 ff. and cf. Marazzi 2002: 509. Starke asserts the Luwian origin of the Hittite word, but does not suggest any etymology for Luw. ubadi(t)-. 33 Olmo Lete and SanmartМn 2003, vol. 1: 7. 34 This analysis is preferable to that of R.Beal (1992), who tentatively assumes that the original meaning of ubadi(t)- was `association / group'. The metonymy from as certain geographic area to the people living in it is common (cf. e.g. H. Nelson's "England expects everyone to do his duty"), whereas the opposite metonymy is extremely unusual. Melchert (1993b:, 243) suggested that Luw. ubadi(t)- originally meant `landgrant < donation',

31

248

Ilya Yakubovich

Now, if the original meaning of Luw. ubdi(t)- was indeed `estate', this word can be derived from Luw. ub `to found' on the assumption of a common semantic change that is exemplified by Lat. fundus `foundation, estate', MPers. bun xБnag and Bactr. `personal estate' and, ultimately, the English word `estate' itself. As for morphology, one can explain Luw. ubdi(t)- by positing an old Anatolian suffix *-at- < *-ot-/-et-, extended by a productive suffix -it- in Luwian35. It is especially interesting that that Hitt. ubdi-, ubadi- is occasionally written with the BA sign (HZL 205) in Old Hittite. The only other Hittite common noun in which this sign occurs is the title tabarna- / labarna- that was probably borrowed from the substrate language(s) of Southern Anatolia via Luwian.36 It is possible that this sign was used in Old Hittite in order to signal a special articulation of the Luwian phoneme /b/, distinct from that of its Hittite counterpart. Alternatively, and perhaps more likely, A.Kassian (pers. comm.) suggests that the BA sign could denotate the same sound combination as PA, but only in markedly borrowed lexemes.37 Thus it appears that a derivative of Luw. ub `to found' was attested already in the first half of the second millenium BC. With Melchert 1993b and pace Hajnal 1998: 82, this root must be kept distinct from Hittite/Luwian uppa- `to bring'. Pace Melchert, this verb did not mean `to dedicate' or `to grant' and, therefore, did not require any explicit beneficiary in Luwian. Having this conclusion in mind, let us consider the other derivatives of ub in Lycian and Carian. The verbal form /ub/ `to place' is attested in two Lycian and one Carian inscriptions cited below: (13) N313m (Xanthos, rim of a marble basin) ]·nida ubet[e (14) N311 (Xanthos, fragment of a limestone block) [erb]bina(j)=ne ubete xruwata ertmi [xer]igah tideimi se(j)=upeneh `Erbbina, the son of Xeriga and Upene, placed it as an offering to Artemis'. (15) *34 ("New York bronze bowl") srquq qtbelem wbt snn orkn ntro pidl `Sarkuk, son of Kotbelem, placed this cup as a gift to Netr'.38
but this interpretation, prompted by his understanding of the meaning of Luw. ub, would make the metonymic usage of ubdi(t)- less likely. 35 For the suffix *-ot-/-et- see Rieken 1999: 100 ff., for the Luwian suffix *-it- see Starke 1990: 151 ff. The only Luwian word that preserves the unextended (Pre-)Anatolian suffix *-ot-/-et- is Tiwad- `Sun-(god)'. Melchert's suggestion to regard Luw. ubБdi(t)- as an extended agent noun *ubata- is less likely since the action noun suffix -ata- never undergoes lenition in Luwo-Lycian. On the substrate origin of the tabarna- / labarna- title cf. Stefanini 2002: 791792; Yakubovich 2002, passim. The former article is also to be consulted for the likely hypothesis about the original function of this title among the Hittites. 37 A. Kassian turns my attention to the fact that the BA sign is regularly used in cult terms (D)abati/e and ambassi- (c), both of Hurrian origin. 38 With Melchert (1993d: 81) and pace Melchert (2002a: 309), I believe that o in -ntro represents a dativelocative marker that can be directly compared with datives-locatives in a in Lycian. The precise phonetic environment for the change */a/>/o/ in Carian (on which see e.g. SchЭrr 2002: 168) remains to be investigated, but it is very likely that the stress was one of its conditioning factors. Even if the Carian genitive in -s indeed assumed
36

Carian monument

249

Since inscriptions (14) and (15) are clearly dedicatory in their nature, the temptation to translate their predicate as `dedicated' is understandable. Yet such a translation would not be compelling in any of the two cases. In both (14) and (15) the beneficiaries of the act of dedication are not expressed by the arguments of the respective predicates, but rather embedded in adjunct phrases `as an offering to Artemis' and `as a gift to Netr'. But even if one opts for the translation `to dedicate' on stylistic grounds, the semantic development `to erect' > `to dedicate' in the inscriptions of this genre is quite natural. One can compare the Greek predicate ' that is translated in inscriptions as either `(he/she) erected' or `(he/she) dedicated', depending on a number of arguments it takes. It is more important to take sides with regard to the original meaning of ub when we consider the following Lycian passages taken from the inscription on the Pillar of Xanthos. (16) TL 44 c35, 67 me=Яne=(e)mu axagЦ: maraza: me ubu=hЦte: kbijti: se utЦna: sttati: sttala:... se xbide: sttati m: sttala: ... `I became a mediator for them (i.e. Persians and Spartans?) . They "issued" (i.e. ordered) a double monument. In Hytenna a stele will be erected ... and in Kaunos likewise a stele will be erected ...'. (17) TL 44 c1314 Яtube: er[e] kizzaprЯna: epriti: se parza: me=Я[ne]: t[a]mЦ: axa: Where a stele? will be prepared? for Tissaphernes and the Persians, I made a building? for them.39 The historical background of the passages cited remains somewhat uncertain, but Keen conjectures that the Xanthian ruler Xeriga functioned as a confidant of the Persian satrap Tissaphernes in his negotiations with the Spartans40. It appears that two stelae were erected as a result of these negotiations. Presumably, they featured the text of an agreement between the Persians and the Spartans, or something similar. Given the existence of HLuw. u-pa-ni`trophy' and Car. [ue] `monument', as well as Lyc. ub- `to place', it is only natural to assume that Lyc. ubu (acc.) also refers to a kind of a monument, that in the next close is designated by the Greek borrowing sttala `stele'. Melchert's provisional translation of me ubu=hЦte: kbijti: as `they issued a double(?) guarantee (?)' can be immediately improved. It is more difficult to provide a probative interpretation for (17) due to a number of unknown lexical items. For Melchert (1993a: 83), Яtube is the predicate of the first clause of this sentence, expressed by 3sg. of an (Yi) conjugation verb. Since the existence of (Yi) conjugation verbs in Lycian has never been proven, I find it safer to assume that the sentence predicate was expressed by a normal 3sg. epriti, even if the exact meaning of this verb is unknown to us. If this assumtion is correct, then the potential candidates for the subjects of the first close are Яtube and kizzaprЯna: ... se parza:. The second option is clearly inferior because
the dative function, as Melchert (2002a) claims, this was a development within Carian that does not preclude the existence of an earlier dative in -o/-e. 39 The translation of the predicate epriti is conjectural. I tentatively derive this verb from *epre < *apara`later, after' on the assumption of a semantic derivation similar to that of Hitt. appai- / appija- `be finished, be done' (if this verb is indeed derived from Hitt. `behind', as per Puhvel 1984: 95). For tama- `building', cf. Melchert 2002b: 249, fn. 35. 40 Keen 1998: 137. This suggestion ultimately goes back to Melchert 1993c: 3135.

250

Ilya Yakubovich

kizzaprЯna: ... se parza: in the nominative can only mean `Tissaphernes and the Persian', and it is utterly unclear why Xeriga would refer to someone by his nationality, as opposite to his personal name. On the contrary if we take this group as a dative, it can be translated as `for Tissaphernes and the Persians', an obvious reference to a Persian satrap and his troops/subordinates. With regard to the nominative Яtube, I am glad to accept the suggestion of Theo van den Hout to regard this noun as a kind of architectural term, formally resembling Яtata- `sarcophagus'(?)41. The two nouns were probably formed in an identical way from the roots ta `to put' and ub `to place, found'. I believe that n~tube designates a certain part of the monument(s) introduced several lines above as *ube/sttala, but only further study of TL 44 as a whole may provide us here with a definite answer. We have seen that the existence of the South Anatolian ub `to place, found' is supported by the combined evidence of Luwian, Lycian and Carian. Yet the impact of various dialects was not identical. The meaning of ub in Cuneiform Luwian is obfuscated by the genre of texts available to us in this dialect. The Hieroglyphic Luwian data are difficult to analyze because two verbs, `to bring' and `to found' were not distinguished in Hieroglyphic orthography. The Lycian (and Carian) verbal forms appear only in dedicatory inscriptions, while Lycian nouns *ube and Яtube belong to the most complicated and controversial text in the Lycian corpus. By contrast, Carian upe / ue `monument' can be immediately linked to the set of objects it denotates, and this prevents us from a skewed interpretation of this word and its cognates. For the first time in Carian studies, the tiny Carian corpus can provide a tangible contribution to the understanding of other Anatolian languages. *** I am offering this Carian monument as a votive gift to Professor L. Gercenberg, who answered the prayers of my heart by training a new generation of Russian Indo-Europeanists. Not being formally his student, I nevertheless also had a chance to experience his generosity at the very beginning of my scholarly career. Against all odds of the transitional period in Russia, L. Gercenberg continues his relentless pedagogical efforts, laying the foundation of the St.Petersburg School of Indo-European Studies.
BIBLIOGRAPHY Adiego Lajara, Ignacio-Javier. 1993. Studia Carica. Investigaciones sobre la escritura y lengua carias. Barcelona: Promociones e publicaciones universitarias. Beal, Richard. H. 1992. The Organization of the Hittite Military (Texte der Hethiter 20). Heidelberg: Carl Winter. Carruba, Onofrio. 1998. "Zum Stand der Entzifferung des Karischen". Kadmos 37: 4756. Hajnal, Ivo. 1998. ""Jungluwisches" *s und die karische Evidenz: Versuch einer dialektologischen KlДrung". Kadmos 37: 80108. Hawkins, J. David. 2000. Corpus of Hieroglyphic Luwian Inscriptions. Berlin; New York: Walter de Gruyter. Ivanov. Vyacheslav V. 2002. Review of Hawkins 2000. Indo-European Studies Bulletin 10/1: 117. Keen, Anthony G. 1998. Dynastic Lycia: a political history of the Lycians and their relations with foreign powers. Leiden: Brill. LЭddeckens, Erich et al. 19802000. Demotisches Namenbuch. Wiesbaden: Reichert. Laroche, Emmanuel. 1959. Dictionnaire de la langue louvite. Paris: A. Maisonneuve. ----. 1960. Les hieroglyphes hittites. Paris: CNRS.

41

Melchert 1993a : 48.

Carian monument

251

Masson, Olivier. 1979. Carian Inscriptions from North Saqqra and Buhen. London: Egypt exploration society. Masson, Olivier and Yoyotte, Jean. 1956. Objets pharaoniques Ю inscription carienne. Cairo: Institut d' archИologie orientale. Marazzi, Massimilano. 2002. "Esercitazione di carri da guerra". Anatolia Antica: Studi in Memoria di Fiorella Imparati (Eothen 11). Florence: Lo Gisma, pp. 50718. Melchert, H. Craig. 1992. "The third Person present in Lydian". Indogermanische Forschungen, 97: 3154. ----. 1993. "Some Remarks on New Readings in Carian". Kadmos 32: 7786. ----. 1993(a). Lycian Lexicon. Chapel Hill: self-published. ----. 1993(b). Cuneiform Luwian Lexicon. Chapel Hill: self-published. ----. 1993(c). "A New Interpretation of Lines 39 of the Xanthos Stele". Akten des II. Internationalen Lykien-Symposiums (ed. J. Borhardt and G. Dobesch). Vienna: жsterreichische Akademie der Wissenschaften. ----. 1994. Anatolian Historical Phonology. Amsterdam; Atlanta: Rodopi. ----. 2002(a). "Sibilants in Carian". Novalis Indogermanica. Festschrift fЭr GЭnter Neumann zum 80. Geburtstag (edd. M. Fritz & S. Zeilfelder). Graz: Leykam. pp. 305313. ----. 2002(b). "The God Sanda in Lycia?" Silva Anatolica: Anatolian Studies Presented to Maciej Popko on the Occasion of His 65th Birthday (Ed. by P. Taracha). Warsaw: AGADE, pp. 241252. Morpurgo-Davies, Anna. 19821983. "Dentals, Rhotacism and Verbal Endings in the Luwian Languages". Zeitschrift fЭr vergleichende Sprachforschung 96: 24570. del Olmo Lete, Grigorio and SanmartМn, Joaquin. 2003. A Dictionary of the Ugaritic Language in the Alphabetic tradition. Leiden: Brill. жzing, JЭrgen. 1976. Die Nominalbildung des дgyptischen. Mainz: von Zabern. Puhvel, Jaan. 1984. Hittite etymological dictionary. v. 1. Berlin; New York: Mouton de Gruyter. Ray, John. 1998. "Aegypto-Carica". Kadmos 37: 12536. Rieken, Elisabeth. 1999. Untersuchungen zur nominalen Stammbildung des Hethitischen. (Studien zu den BoazkЖy-Texten 44). Wiesbaden: Harassowitz. SchЭrr, Dieter. 1992. "Zur Bestimmung der Lautwerte des Karischen Alphabets". Kadmos 31: 127156. ----. 1996. "Bastet-namen in karischen inschriften дgyptens". Kadmos 35, pp. 5571. ----. 2002. "Karische Parallelen zu zwei Arzawa-Namen". Kadmos 41, pp. 16377. Starke, Frank. 1985. Die keilschrift-luwischen Texte in Umschrift (Studien zu den BoazkЖy-Texten 30). Wiesbaden: Harassowitz. ----. 1990. Untersuchungen zur Stammbildung des Keilschrift-Luwischen Nomens (Studien zu den BoazkЖy-Texten 31). Wiesbaden: Harassowitz. Stefanini, Ruggero. 2002. "Toward a Diachronic Reconstruction of the Linguistic Map of Ancient Anatolia". Anatolia Antica: Studi in Memoria di Fiorella Imparati (Eothen 11). Florence: Lo Gisma, pp. 783806. Vittman, GЭnter. 2001. "дgyptisch-Karisches". Kadmos 40: 3959. Yakubovich, Ilya. 2002. "Labyrinth for Tyrants". Studia Linguarum 3 (Memoriae A.A. Korolev dicata), Moscow: Languages of Slavonic Culture. Vol. 1, pp. 93116. Yoshida, Daisuke. 1987. Die Syntax des althethitischen substantivischen genitivs (Texte der Hethiter 13). Heidelberg: Carl Winter.