Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.imk.msu.ru/Structure/Linguistics/yakubovich/download/ENEMY.pdf Äàòà èçìåíåíèÿ: Fri Apr 30 15:11:19 2010 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:49:36 2012 Êîäèðîâêà:

ILYA YAKUBOVICH THE LUVIAN ENEMY 1. Introduction The purpose of this paper is to define the meaning of the Bronze Age Luvian lexeme 416-wa/i-nÌ-.1 It occurs several times in the inscriptions YALBURT and SýDBURG, which contain the res gestae of the two Hittite kings of the late Empire period, Tuthaliya IV and Suppiluliyama II, and once in the inscription KIZILDA' 4, which commemorates the deeds of a certain Hartapu, possibly a king of Tarhuntassa.2 The frequency of this word in the Bronze Age Luvian corpus implies that its synchronic analysis is both possible and necessary. On the one hand, we have enough contexts where 416-wa/i-nÌoccurs in order to apply the combinatory method. On the other hand, the correct interpretation of this word is likely to have consequences for the general understanding of the relevant Luvian texts. Hawkins (1990: 307­308) suggests in his preliminary edition of the SýDBURG inscription that ta4-wa/i-nÌ-, as this noun was conventionally transliterated at the time, represents a royal title. The same semantic interpretation was advocated in Hawkins 1995, the volume that contains the editions of all the Luvian texts featuring 416-wa-nÌ-. The appendix to this volume addresses the word under discussion in more detail (Hawkins 1995: 114­117). David Hawkins
1

2

Usual disclaimers apply. The preliminary version of this paper has been presented at the Luvian Seminar taught by Theo van den Hout at the University of Chicago in Winter 2007, and I am obliged to the audience of this seminar for their insightful comments. In a later period, it benefited from good advice of Elisabeth Rieken (Marburg) during our joint work on a related topic (Rieken and Yakubovich, forthcoming). Finally, I am grateful to the same Elisabeth Rieken, David Hawkins (London), and Itamar Singer (Tel-Aviv) for their remarks on the advanced version of this paper, and to Aaron Butts and Yaroslav Gorbachov (Chicago), who helped me to improve its style. See Bryce 2007 for a plausible reconstruction of the history of Tarhuntassa in the thirteenth century B.C., which implies the possibility that Hartapu was a contemporary of Tuthaliya IV and Suppiluliyama II. Hawkins 2000: 434 is inclined to place Hartapu in the period shortly after the collapse of the Hittite Empire.

Kadmos Bd. 47, S. 1­19 © WALTER DE GRUYTER 2008 ISSN 0022-7498

DOI 10.1515/KADMOS.2008.001


2

Ilya Yakubovich

cautiously hypothesizes that it was pronounced /diwani-/ or /liwani-/, which represents a phonetic writing of the hieroglyphic title SOL2 corresponding to the cuneiform title dUTU-SI "My Sun", and is etymologically related to the Luvian theonym Tiwaz `Sun-god'. A different etymology was proposed by Massimo Poetto in his edition of the YALBURT inscription. While endorsing Hawkins' interpretation of 416-wa-ni- as a royal title, Poetto (1993: 28­29 with fn. 43) compares it with the title tawananna- born by the Hittite queen and Lyc. tewinaza, the designation of an official. The last comparison is quoted with approval in Melchert 2003: 181, fn. 12. In this paper, I argue that Luv. 416-wa/i-nÌ- is not a royal title, but a common noun denoting `enemy', and therefore none of the etymologies previously suggested for this noun can be maintained. I will illustrate my findings through the updated translation of all the Luvian passages where this Luvian form occurs. I intend to show that the new translation not only yields a better sense of the relevant contexts but also allows one to arrive for the first time at a cohesive understanding of the SýDBURG inscription. I will conclude my account with discussing the likely Iron Age Luvian cognate of 416-wa/i-nÌ-, whose meaning `enemy' can be established on independent grounds, and proposing the new reading and etymology of the word under discussion. 2. 416-wa/i-nÌ-sa in YALBURT It is appropriate to begin my analysis with the discussion of the arguments advanced for the traditional interpretation of *416-wa/inÌ-, which are most clearly summarized in Hawkins 1995: 26. Discussing the occurrences of the inflected form 416-wa/i-nÌ-sa in the YALBURT inscription, Hawkins assumes that -sa must represent a nominative singular ending and therefore we are dealing with the subjects of the clause. Then he observes that this noun occurs twice in the same clause with the 1. sg. predicate mu-wa/i-ha and concludes that 416-wa/i-nÌ-sa represents a term of self-designation by the speaker (Tuthaliya IV), thus a royal title. These arguments, however, are not logically compelling. The final hieroglyphic in a Luvian nominal form may represent not only the nominative singular marker /-s/ of the common gender but also the particle /-sa/ that can be optionally attached to the nominative and accusative singular nouns of the neuter gender in Bronze Age Luvian (Melchert 2003: 186). If so, 416-wa/i-nÌ-sa can function not only as a subject


The Luvian Enemy

3

but also as a direct object in YALBURT. The choice between these two possibilities must depend on the syntactic analysis of individual clauses where the relevant lexeme occurs. Since the YALBURT inscription represents an array of disjoined blocks, whose relative sequence cannot be established in many cases, the syntactic interpretation of this text sometimes presents challenges. The commonly accepted direct join between blocks 12 and 13 yields, however, a continuous fragment dealing with the Hittite expedition against Lycian towns, in which 416-wa/i-nÌ-sa occurs three times. This well-preserved passage provides a convenient starting point for the reinterpretation of our lexeme. (1) YALBURT 12, § 2­4+13, § 1­4a, cf. Hawkins 1995: 70
1.1 a-wa/i pi-na-416(URBS) tu-pi 1.2 a-wa/i-mu (VIR2)416-wa/i-nÌ-sa pi-na-416(URBS) FORTIS.CRUS 1.3 a-wa/i-mu (DEUS)TONITRUS DOMINUS-na PRAE hwi/a-i(a)-tÀ 1.4 a-wa/i-mi (VIR2) 416-wa/i-nÌ-sa mu-wa/i-ha 1.5 pi-na-416(URBS) ARHA DELERE 1.6 a-wa/i À-wa/i+ra/i-na-´ (REGIO) PES2 1.7 a-wa/i-mu (VIR2) 416-wa/i-nÌ-sa IVxMILLE CENTUM MULUS I attacked the town Pinali. The enemy resisted me in the town Pinali. Tarhunt, my lord, ran in front of me. I defeated the enemy, and destroyed the town Pinali. I went to the country Awarna The enemy, (numbered) 4100, resisted me.

The syntactic considerations bear out the hypothesis that 416-wa/inÌ-sa functions two times as a subject and once as a direct object in (1). The overt expression of the subject is necessary in (1.2) and (1.7), where the non-reflexive clitic =mu `(to) me' indicates that the subject of the respective clauses cannot be co-referential with the narrator.3 If I am right in my identification of the ideogram FORTIS.CRUS, lit. `stand fast' with the Luvian verbal stem ta- `stand' in (1.2) and in my assumption that MULUS in (1.7) represents a rebus writing for the same ta- `stand', then the only syntactic relation available
3

Poetto (1993) and Hawkins (1995) take =mu in II as a reflexive clitic, but the functional extension of =mu to the first person reflexive is reliably attested only in the late Iron Age Luvian texts and in the dialect(s) of Istanuwa (Yakubovich 2008: 215­217). A further argument against this analysis is the actual presence of the reflexive clitic =mi in the same passage (IV) and several more times in the YALBURT inscription. Hawkins (1995) finds an independent instance of the reflexive =mu in YALBURT 6.2, but the reconstruction of the sign L 41=CAPERE in this clause does not appear to be assured (see the photograph in Poetto 1993, plate VII and compare the assured L 41= in YALBURT 7.2a, Poetto 1993, Plate IX).


4

Ilya Yakubovich

for 416-wa/i-nÌ-sa is that of the subject.4 By contrast, the transitive verb muwa- (+refl) `to overcome, defeat' implies the presence of a direct object in clause (1.4), while 416-wa/i-nÌ-sa is the only suitable candidate for this syntactic relation (the reflexive clitic =mi is probably used in the telicizing function and does not fill in a valence slot in this clause).5 All three occurrences of 416-wa/i-nÌ-sa in (1.4) speak against its interpretation as a royal title. In (1.2) and (1.7), this noun refers to different groups who resist Tuthaliya IV, while in (1.4) it refers to the party defeated by Tuthaliya IV. The fact that it can function both as a subject and direct object in a clause without morphological modification points to its neuter gender and prompts the segmentation of the nom./acc. sg. n. particle =sa. As for its semantics, the denotate `enemy army' appears to be appropriate for the general context of (1.2) and (1.4), but the determinative VIR2 and the numeral modifier 4100 in (1.7) suggest that we are rather dealing with a neuter abstract noun that is metaphorically used with the meaning `enemy'.6 The metaphor abstractum pro concreto is, of course, cross-linguistically very frequent, but the best parallel within Luvian would be the abstract noun tarwan(a/i)- `justice' (~ Phoen. dq) that has secondarily acquired the concrete meaning `ruler, judge' and was probably borrowed into Greek as térannow `ruler, tyrant' (cf. Yakubovich 2008: 184 with fn. 91).
4

5

6

The identification of FORTIS.CRUS as equivalent to ta- in the narrow meaning `resist' in (1.2) is not crucial for my interpretation. As an alternative, one can suggest that FORTIS CRUS corresponds to a sequence of two words, where CRUS is the usual equivalent of ta- `stand', while FORTIS is an adverbial modifier `fast' (vel sim.). With regard to (1.7), Hawkins (pers. comm.) observes that L 101 = MULUS, unlike L 100 = ASINUS, is not otherwise attested with the syllabographic value (cf. Hawkins and Morpurgo-Davies 1997: 250). Yet, given that L 100 and L 101 are used for the derivatives of the same noun targasna-, an occasional rebus writing MULUS for the verbal stem ta- does not strain credulity. On the other hand, the literal reading of the logogram MULUS as `mule' does not appear to be conducive to a coherent interpretation of the clause as a whole. For the possible telicizing function of the Luvian reflexive clitics, see Yakubovich 2008: 211. The Luvian verb muwa- (+refl) probably represents a functional equivalent to the Hittite telic verb =za ... tarh- `to conquer, overcome, defeat' (Hawkins 1995: 79 with ref.). A formal parallel for the syntactic template of (1.4) is YALBURT 16.2a a-wa/i-mi REGIO 430 (273)mu-wa/i-ha `I defeated all the countries'. Poetto (1993: 29) was the first to recognize VIR2 as a determinative, while Hawkins (1995) took it as a word-divider in his edition of the YALBURT inscription. The transliteration value accepted in the present paper is based on the sign-list in Hawkins 2000: 617­620. I am grateful to Virginia Rimmer (University of Chicago), who turned my attention to Hawkins' change of opinion.


The Luvian Enemy

5

Now one can apply these preliminary conclusions to the analysis of other passages of the YALBURT inscription featuring 416-wa/inÌ-sa. Example (2) is very similar to (1.4), except for the fact that the location of the defeated group remains unidentified. The reconstruction of (3) depends on the new join between blocks 7 and 17. I suggest that these fragments are likely to follow each other because both of them refer to Tuthaliya IV's campaign against the country 511(REGIO), which does not appear in the rest of the YALBURT inscription. If my join is correct, one gains an additional sentence where 416-wa/i-nÌ-sa functions as a direct object, this time governed by the verb DELERE-ha `I destroyed'. The traditional interpretation of 416-wa/i-nÌ-sa as a royal title and a clause subject remains formally possible in both cases, but I fail to see a pragmatic motivation for the appearance of the title "My Sun" in these clauses but not in their immediate environment. (2) YALBURT 11, § 2, Hawkins 1995: 68 a-wa/i-mi (VIR2) 416-wa/i-nÌ-sa? a-ta?-pa-x(URBS?/REGIO?) mu-wa/i-ha `I defeated the enemy in Atapa-...'. (3) YALBURT 7, § 2b + 17, § 1, cf. Hawkins 1995: 68, 70 (VIR2)416-wa/i-nÌ-s[a] ni-pi+ra/i(REGIO) REL-REL-lu-wa/itÀ(REGIO) *511-sa5(REGIO) x x [...] (REGIO) DELERE-ha `I destroyed the enemy in the country Nipira, the country RELREL-luwada, (and) the country ... (which is part) of the country *511 The final relevant passage of the YALBURT inscription, if correctly understood, is extremely important for reconstructing the political history of the Hittite Empire. Bryce (2005: 319­321) discusses the possibility of a temporary loss of Hattusa by Tuthaliya IV as a result of a coup staged by Kruntiya, king of Tarhuntassa. The main argument for this hypothesis is the discovery of seal impressions in Hattusa bearing the inscription "Kruntiya, Great King, Labarna, My Sun". Nevertheless, no direct confirmation of this hypothesis has been forthcoming thus far from written records, and this is why Bryce (2007: 125) looks for alternative scenarios that can account for the Kruntiya's sealings found in Hattusa.


6

Ilya Yakubovich

The translation of (4) given below provides the first piece of evidence that Tuthaliya IV had to fight for power in Hattusa at a certain point in his career. The new analysis is formally superior to what had been offered before because the subject *416-wa/i-nÌ-sa `My Sun' would preclude the appearance of the non-reflexive object clitic =mu `(to) me, my' in this clause, as it would in (1.2) and (1.7).7 The new translation of (4) also fits well within the context, because the next sentence `Tarhunt, [my lord ran in front] of me' (YALBURT 2, § 3) clearly refers to a military engagement, in which Tuthaliya was victorious. On can specifically indicate the parallelism with (1.2)-(1.3), where the description of the enemy actions precedes the reference to the involvement of Tarhunt. (4) YALBURT 2, § 2, cf. Hawkins 1995: 68 *a-wa/i-mu (VIR2) 416-wa/i-nÌ-sa LINGUA+CLAVUS-tusa(URBS) *a-POST URBS+MI-a IUDEX+LA PES `My enemy came back to Hattusa, to the city of the Labarna'. 3. 416-wa/i-nÌ in SýDBURG The difficulties involved in the interpretation of the SýDBURG inscription are of a very different kind. Here one does not need to solve a jigsaw puzzle, because the reconstruction of the inscription appears to be assured by archaeological means. On the other hand, an unfortunate consequence of the extremely logographic character of its execution is the omission of clause-initial clitic chains in most cases, which makes it difficult to establish clause boundaries. Accordingly, a discussion of the general composition of the inscription is called for before one addresses the interpretation of its individual clauses. Below I provide a new translation of SýDBURG, which is substantially different from the one given in Hawkins 1995: 23. The original paragraph boundaries of the editio princeps are, however, preserved in square brackets. The commentary to the translation focuses on the syntactic differences between Hawkins' analysis of the inscription and mine using the paragraph division of Hawkins 1995 as a framework of reference.
7

The right-branching order of the noun phrase URBS+MI-a IUDEX+LA `to the city of the Labarna' finds parallels in divine titles (DEUS) TONITRUS HATTI `Storm-god of Hatti' and (DEUS) TONITRUS EXERCITUS `Storm-god of the Army', attested in the SýDBURG inscription.


The Luvian Enemy

7

Translation [1a] When in all the land of Hatti, I subjected8 the enemy [1b] Wiyanawanda, Masa, Lukka and Ikkuna [2a] were rebellious (?) against the former kings. But now, [2b] to Suppiluliuma the Great King, the Hero, [3] all the gods (the Sun-goddess of Arinna, the Storm-god of Hatti, the Storm-god of the Army, Sawoska, the Sword-god, the Storm-god of Sapini (?) ... stood with favor, [4a] (and) he subjected the enemy: [4b] Wiyanawanda, Masa, Lukka, and Ikkuna. [5] The chieftains in all the land of Hatti and on the frontiers of Hatti, he removed this enemy. [6] Suppiluliuma, Great King, the Hero, subsequently (re)built the land of Hatti. [7] The town A, the town B, the country C, the town D, the town E, the town Tihihasa, the town Tarahna, these (are the towns) he (re)built. [8] In (the land) Mount F., he subjected and conquered the enemy. [9] Suppiluliuma the Great King subjected the enemy, [10] (and) Mount F. surrendered (?).9 [11] The chieftain(s) of Mount F subsequently (began to) serve Hatti. [12] In the land of the town Tarhuntassa, he subjected and conquered the enemy. [13] The ancestors formerly did not ... to anyone, [14] (but) Suppiluliuma the Great King subjected the enemy. [15] He subjected and removed the chieftain(s) of the town Tarhuntassa. [16] He (re)built the town Adana (?).10 [17] In the town Tarhuntassa, the town G and the town H, I gave offerings to the gods. [18] Here, in that year, I constructed a "Divine Earth-Road". Commentary § 1b. I assume that this segment begins an extended quotation from a text that refers to the deeds of Tuthaliya IV in the third person. The conclusion that the SýDBURG inscription quotes from a different source is prompted by the contrast between the third person clause
8

9

10

For the latest etymological discussion of INFRA À-ka `to subject', see Rieken 2007. For difficulties in interpreting the sign that corresponds to the predicate of this clause, see Melchert 2002: 141, fn. 4. Whether or not this sign can be identified with L 66=DARE, the translation `surrendered' appears to be supported by context. For the suggestion that Adana represents a place where the deported population of Tarhuntassa may have been resettled, see Melchert 2002: 140­141.


8

Ilya Yakubovich

in § 7 (quod vide) and the first person clause in § 17, both having the king as the subject. It seems logical to assume that § 1a summarizes the military exploits of Tuthaliya IV, whereas the embedded narrative beginning in § 1b represents an elaboration of this topic. § 1b­2a. For the interpretation of the first clause of the narrative digression, see Melchert 2006: 292 (with fn. 5). For the analysis of CRUS-nÇ-pa as /nanun=pa/ `but now', see Melchert 1997. Here and below, note the absence of the demarcational =wa= in the preserved clitic chains, an archaic feature that unites the SýDBURG inscription with Luvian texts attested in cuneiform transmission. § 4a. I combine this clause into one sentence with the previous one because it shows a clitic chain /a=ta/. I assume that all of the sentence-initial clitic chains are suppressed in writing in the SýDBURG inscription, whereas clause-initial chains in the middle of the sentence can be preserved, but this distribution is very tentative and not essential for my other conclusions. § 4b­5. I uniformly translate CAPUT.VIR as `chieftain(s)' on the assumption that this title always applies to the rebellious or otherwise inimical rulers. This suggestion was first formulated in Jasink 2001, although Jasink believes that CAPUT.VIR is always used with reference to the same adversary, namely Hartapu.11 In my opinion, this title has three different referents in the text of the inscription. In these case the sentence apparently refers to some unspecified rebels, while later in the text the lands of the rebellious "chieftains" are explicitly mentioned. § 7. The sequence ti-sa-a is to be reinterpreted as *a-ti-sa modified by the scribal convention known as "initial-a-final"(Hawkins 2003: 159­61). It corresponds, in my opinion, to the clitic sequence /a=di=(a)s/, where =di= is the third person reflexive, while =as is the resumptive accusative clitic `them, these' of the common gender, which refers back to the towns rebuilt by Tuthaliya IV.12 This identification confirms the hypothesis of Yakubovich 2008: 78 that the clitic =as (acc. pl. c.) existed in Empire Luvian and was analogically replaced with =ada only in the Iron Age. The likely reason for the preservation of this clitic complex in the logographic SýDBURG inscription is its position in the middle of the clause, after the fronted
11

12

Cf. the different hypotheses of Hawkins (1995), who takes CAPUT.VIR as a title of the Hittite king, and of Melchert (2002), who suggests that CAPUT.VIR is not a title but a plural common noun `men, people'. One can either assume a phonetic contraction /adiyas/ > /ad%s/ or, more likely, a simplifying spelling ti-sa-a instead of ti-a-sa-a. It is important to remember that the sign L 210 = had not existed yet at this point.


The Luvian Enemy

9

constituents. The use of =di= instead of **=mi= indicates that this clause has the third person subject, which in turn provides the crucial piece of support for the hypothesis of an embedded quotation. § 8,11,12. The semantic ambiguity of the logogram PUGNUS. PUGNUS is discussed in Rieken and Yakubovich, forthcoming. We have concluded that it can be used for two different verbal derivatives of hutarla/i- `servant, slave': an intransitive verb meaning `to serve, take care of' and its transitive counterpart meaning `to conquer, subdue'. The first meaning of PUGNUS.PUGNUS is apparent in § 11, whereas the second one must be reconstructed in § 8 and § 12. Both meanings are independently attested in the Iron Age Luvian corpus. It is easy to see that in spite of a number of new interpretations, I essentially agree with Hawkins that a large part of the inscription describes the military campaign of Tuthaliya IV. Having this conclusion in mind, one can turn to the analysis of the seven passages featuring the lexeme 416-wa/i-nÌ, two of which are precisely identical (9). (5) SýDBURG, § 1a, cf. Hawkins 1995: 23 HATTI (REGIO) 430 REL+ra/i 416-wa/i-nÌ INFRA À-ka `When in all the land of Hatti, I subjected the enemy'. (6) SýDBURG, § 4a, cf. Hawkins 1995: 23 a-tÀ 416-wa/i-nÌ INFRA À-ka `(and) he subjected the enemy' (7) SýDBURG, § 5, cf. Hawkins 1995: 23 CAPUT.VIR HATTI (REGIO) 430 FINES-zi/a HATTI 416-wa/i-nÌ zi/a ARHA CAPERE `The chieftains in all the land of Hatti (and) on the frontiers of Hatti, I captured this enemy'. (8) SýDBURG, § 8, cf. Hawkins 1995: 23 MONS.IUDEX?.QUINQUE 416-wa/i-nÌ INFRA À-ka PUGNUS. PUGNUS `(In the land) MONS.IUDEX?.QUINQUE, he subjected and conquered the enemy'.


10

Ilya Yakubovich

(9) SýDBURG, § 9 = § 14, cf. Hawkins 1995: 23 PURUS.FONS.MI MAGNUS.REX 416-wa/i-nÌ INFRA À-ka `Suppiluliyama the Great King subjected the enemy'. (10) SýDBURG, § 12, cf. Hawkins 1995: 23 TONITRUS(URBS) REGIO || 416-wa/i-nÌ INFRA À-ka PUGNUS.PUGNUS `In the land of Tarhuntassa, he subjected and conquered the enemy'. Two convincing reasons plead against the identification of 416-wa/inÌ as a royal title in the clauses under discussion. On the one hand, it is easy to see that the predicates of (5­10) are invariably transitive and denote hostile actions. It would be very difficult to explain why the title 416-wa/i-nÌ is compartmentalized in this category of clauses, especially given the contrast with the title "Suppiluliuma the Great King", which does not show any restrictions of usage. On the other hand, transitive clauses normally require an overtly expressed direct object, and at least in (5), (8), and (10), which begin the new sections of the narrative, it is unlikely to be expressed by a suppressed clitic pronoun. Hawkins (1995) assumes that the subjected lands function as direct objects of these three clauses, but this implies the reconstruction of the pragmatically marked word order OSV. No obvious rationale for the use of this syntactic pattern is forthcoming. While the fronting of a salient patient functioning as argument focus is possible in the SýDBURG inscription, this pragmatic configuration can be achieved through a cleft construction, and not through subject-object transposition, as § 4a­5 and § 7 ostensibly illustrate.13 Both difficulties disappear once we assume that 416-wa/i-nÌ functions as a direct object with a negative meaning in (5­10). The toponyms that were formerly taken as direct objects can now be reinterpreted as locative adjuncts, which, of course, can appear at the beginning of a clause. The clitic complex -tÀ in (6), which Hawkins (1995) analyzed as the accusative pronoun /=da/ `them', can now be understood as the sentential particle /=ta/ (which otherwise appears in NEG-wa/i-tÀ, SýDBURG § 13). It remains unclear based on the evidence presented thus far whether *416-wa/i-nÌ of SýDBURG should
13

The general issue of pragmatically-driven syntactic configurations in Luvian has not been properly investigated thus far. It is, however, clear that OSV word order occurs most frequently in those cases where the patient is aligned with the topic (e.g. ERK~LET 2, § 1 za-wa/i |À-la-na IÀ-sa-ti-wa/i-su-sa |tu-ta || `This ala- is erected by Astiwasu').


The Luvian Enemy

11

be taken as a genuine morphological variant *416-wa/i-nÌ-sa, or whether the omission of the final -sa represents a mere orthographic phenomenon, to be considered together with the other truncated spellings attested in this inscription. In order to solve this problem, one has to turn to the analysis of KIZILDA' 4. 4. 416-wa/i-nÌ in KIZILDA' 4 The general content of this short inscription is reasonably well understood, except for the last clause, which had long defied interpretation by scholars due to the bad state of its preservation. The recent identifications of the poorly preserved signs by Hawkins and Poetto are communicated in Hawkins 2006a: 60. On the basis of these new discoveries, Hawkins arrives at the following translation of the clause under discussion: "The men of all the earth (I), Great King, Diwani, Hunter, took by the help of the Celestial Storm-god". My translation, presented below, acknowledges all the recent sign recognitions but implies a different syntactic analysis of the sentence in question. (11) KIZILDA' 4, § 3, cf. Hawkins 1995: 104 (TERRA) ta-sa-REL+ra/i *430 MAGNUS.REX (VIR) 416-wa/i-nÌ CERVUS4.IACULUM CAPERE (DEUS) TONITRUS CAELUM `On all the earth, the Great King hunted (and) captured the enemy by (the help of) the Storm-god of Heaven'. The main reason to prefer the new interpretation is that it eliminates yet another pragmatically unmotivated instance of OSV word order.14 The only potential advantage of the solution proposed by Hawkins lies in the possibility of interpreting CERVUS4.IACULUM (formerly transcribed VITELLUS 285) as a royal title. David Hawkins succeeded in showing that the title "hunter" (vel sim.) represents the most frequent meaning of this group of signs. Yet, the same group appears to correspond to the predicate `hunt(ed)' in the KARAKUYU
14

REGIO 430 REL-sa (273) FORTIS-tÀ `who conquered all the lands' (KIZILDA' 4, § 2) represents another instance of OSV word order in the same group of texts. In this case, however, the subject appears in second position because it is a relative pronoun in a determinate relative clause. The alleged relative pronoun at the end of KIZILDA' 4, § 2a­b (Hawkins 2000: 438) may possibly be relegated to the beginning of the following clause KIZILDA' 4, § 3 as a resumptive /kwi=pa/ `indeed' (for which see Goedegebuure 1998).


12

Ilya Yakubovich

inscription (Hawkins 2006a: 61), as it does in (11) under my interpretation. The use of the same ideogram for nominal and verbal forms has other parallels in Luvian, the most conspicuous of which are the multiple functions of L 93=PES (`foot; come, bring'). In addition to text-internal arguments, the close parallelism between examples (5), (7), and (11) is worthy of mention. If one accepts that the new meaning of 416-wa/i-nÌ can be established on independent grounds for the SýDBURG res gestae, it must have been used with the same meaning of `enemy' in our passage. On a different level, one can compare the use of the determinative (VIR) in front of 416-wa/i-nÌ in (11) with the use of (VIR2) in the same function in the YALBURT text. This comparison supports the hypothesis that the neuter noun 416-wa/i-nÌ was metaphorically used with reference to a group of human beings. Unlike the SýDBURG inscription, the texts of the KARADA'KIZILDA' group normally show inflectional endings with phonetically spelled word-forms. Therefore, the contrast between 416-wa/i-nÌ in KIZILDA' 4 and 416-wa/i-nÌ-sa in YALBURT is unlikely to have a purely orthographic explanation. One should rather assume that the -sa/-za extension could be optionally added to the direct cases of the Luvian neuter singular nouns in the Bronze Age, and was grammaticalized only in the Iron Age. This conclusion has already been made with reference to the Bronze Age Luvian texts attested in the cuneiform transmission (Melchert 2003: 186), but now it can be extended to the hieroglyphic texts of the same period. 5. Phonetic interpretation The occurrences of the sign L 416 are restricted to Luvian texts of the second millennium BC. Before direct evidence was available, scholars frequently abstained from speculations about its reading (cf. Laroche 1969: 219). New data shedding light on the phonological interpretation of L 416 has become available in the last two decades. The name kÀ-416 occurring on the cruciform seal next to the name of Mursili I virtually guarantees that it belongs to the Hittite queen Kali known from the offering lists (DinÃol et al. 1993: 94­95). The identification of pi-na-416 with the Lycian toponym P¤nara/Pinale finds strong support in the juxtaposition of pi-na-416 and Awarna/ Xanthos in the YALBURT inscription (Poetto 1993: 80). Finally, the digraphic sealing NIANTEPE, #3 prompts a direct equation between the hieroglyphic spelling 172-416-mi and the cuneiform


The Luvian Enemy

13

spelling of the same name a-la!-li-mi (written a-ad-li-mi, Herbordt 2005: 114). The correspondence between L 416 and the segment li of cuneiform sources in word-medial and word-final positions has been duly noted in DinÃol et al. 1993: 95 and Hawkins 1995: 114­115. The analysis of personal names beginning with 416 is conducive to the refinement of this conclusion. In a number of cases, they find direct counterparts in Hittite/Luvian names attested in the cuneiform transmission, which begin with the element Ali-. Thus 416-VIR.ZI/A (TARSUS, #13, Laroche 1958: 257), 416-SARMA (NIANTEPE, #98, Herbordt 2005: 130), 416-LEO (NIANTEPE, # 644, Herbordt 2005: 227), and 416-mu-tÀ (NIANTEPE, # 645­646, Herbordt 2005: 227) can be respectively compared with Aliziti (Laroche 1966, #36), AliSarruma (Laroche 1966, #33), Ali-UR.MAH (Laroche 1966, #34), and Alimutta (Laroche 1981, #32b). It is on the basis of these alternations that Hawkins (2005: 289­90) formulated a revised version of his hypothesis, according to which the sign L 416 may have been used in the Empire Period to "write the element ali- with the initial a- either omitted or implicit". A further step in the analysis of this sign was made in Rieken and Yakubovich, forthcoming. Since there are no cases where L 416 demonstrably corresponds to /li-/ in word-initial position, one can transliterate this sign as in all its occurrences. Under this interpretation, the hieroglyphic names of Mursili I's principal wife and the Lycian city can be rewritten as kÀ-ali and pi-na-ali respectively. This is, of course, primarily an issue of scholarly convention, but doing otherwise would be inconsistent with the established practice of transliterating L 133/134 as even in those cases where the initial vowel is strictly redundant, e.g. sÀ-pi-sa-ara/i-ri+i `with health' (KARATEPE (Ho.), § 49), À-pa-ara/i `by that' (MARA 4, § 15), or wa-ara/i `and them' (KARKAMIS A 6, § 26). If one accepts our suggestion, the Luvian word consituting the main focus of the present paper can be now transliterated as ali-wa/i-nÌ(-). Since all of the previous etymologies of the word for `enemy' were based on its faulted syntactic and semantic analysis, the new phonetic interpretation of this word comes at a zero cost. There is, however, more to be said in its defense. Meriggi (1962: 220) saw the graphic continuity between the Bronze Age sign L 416 and the better-attested Iron Age sign L 319, and this idea was endorsed in Hawkins 1995: 26­27. Rieken and Yakubovich, forthcoming, discuss the phonetic value of L 319 in some detail and arrive at the conclu-


14

Ilya Yakubovich

sion that it was .15 After the phonemes /d/, /l/, and /r/ merged into a flap [] in intervocalic position in Late Luvian, the sign L 319 began to be hypercorrectly used for the historical reflexes of */dV/ and */rV/. In the same period, one begins to encounter the use of the same sign for etymological syllables with the u-vocalism. For example, one can observe the alternation between À-ru! (TOPADA, § 31) and À-la/i (TOPADA, § 33), both representing etymological *aru `in high measure'. Juliette Blevins (pers. comm.) suggests that this practice may reflect a near-merger of all Luvian vowels in the position after a flap.16 Another result of Rieken and Yakubovich, forthcoming, is the documentation of the Iron Age Luvian word for `enemy'. The contexts where this lexeme occurs are listed below with the reconstruction of the "initial-a-final". The varying spellings plead for the phonological reconstruction */alun(n)a/i-/ in early Iron Age Luvian. By the time of the attestation of the forms cited, the intervocalic /l/ had already undergone flapping, as example (16) directly illustrates. The delabialization of *u after the flap is confirmed by the orthography of (12­14), while the historical presence of *u in the word under discussion is born out not only by (16) but also by (15), where the sign appears to have been inserted as a vocalic indicator.17 (12) KARKAMIS A23 § 4­5, cf. Hawkins 2000: 119 *a-wa/i-mu *a-mi-zi a-la/i-ni-zi |pi-pa-sa-ta *a-mu-pa-wa/i *a-la/i-na-za |NEG2 |pi-ia-ta `She gave my enemies to me but did not give me to my enemies'.
15

16

17

Rieken and Yakubovich, forthcoming, suggest that the difference in reading between L 416 = and L 319 = reflects the imperfect transmission of the Anatolian scribal lore in the period after the collapse of the Hittite Empire. In particular, the Iron Age period witnesses frequent interchangeable usage between L 319 and a different sign L 172, whose phonetic value in the Bronze Age was , thus distinct from that of L 416. This process can be conceived of as an instance of progressive (perseverative) assimilation. Since a flap is a consonant of a very short duration, speakers may have lacked time to adjust the shape of the vocalic tract to the articulation of the following high vowel, and hence it became close to the [a] that preceded the flap. Although the sign L 445 has the value in Iron Age Luvian, it was used exclusively with the value in the Bronze Age (Marazzi 1990: 270 with ref.). The use of L 445 for the sequence */lu/ also remained by far the most common one in the Iron Age.


The Luvian Enemy

15

(13) TELL AHMAR 6 § 5, cf. Hawkins 2006b: 12 |*a-mu-pa-wa/i *a-mi-[i]-zi |tÀ-ti-zi *a-la/i-nÌ-zi (DEUS.ORIENS) ki-sÞ-ta-ma-si-zi (DEUS.ORIENS)ki-sÞ-ta-ma-ti |ARHA |"DELERE"-nÇ-[sÞ]-ha `I eliminated my father's eastern enemies from the east'. (14) TELL AHMAR 6 § 21, cf. Hawkins 2006b: 14 *a-la/i-ni-zi-pa-wa/i-mi-i |ARHA |("DELERE")mara wa/i-ha `I destroyed my enemies'.
+ra/i

(?)-nu-

(15) BABYLON 1 § 11­12, cf. Hawkins 2000: 392 |ma-pa-wa/i-sa a-tÀ-ti-li-i-sa | *a-la/i-la/i/u-nÌ-sa |ma-pa-wa/i-sa |ARHA-ti-i-li-sa *a-la/i|-la/i/u-nÌ-sa (a curse formula) `... whether he is an internal enemy, or whether he is an external enemy ...' (16) SULTANHAN § 9, cf. Hawkins 2000: 466 |wa/i-tu-u |À-ru-ni-i-zi |À-pa-si-i-zi |("PES")pa+ra/i-za |SUB-na-na |tu-wa/i-ta `And he (Tarhunt) put his (Wasusarma's) enemies under his feet'. For all the phonetic and morphological problems involved in the comparison between Bronze Age Luvian ali-wa/i-nÌ(-) `enemy' and its Iron Age Luvian synonym reconstructed as /alun(n)i-/, it does not seem possible to separate these two lexemes. The new interpretation of L 416, which was suggested on independent grounds, receives confirmation from the analysis of of 416-wa/i-nÌ(-). On the other hand, the interpretation of this word as `enemy', which initially emerged from the syntactic analysis of the Bronze Age Luvian corpus, finds welcome support in the Iron Age Luvian data. Thus, the conclusions of Sections 2­4 and Section 5 mutually corroborate each other. 6. Diachronic Interpretation The historical account presented below represents an attempt to make sense of the relationship between Bronze Age Luvian and Iron Age Luvian words for `enemy', which prima facie appear to be cognate. It is not meant to serve as a proof of the semantics of either of the


16

Ilya Yakubovich

two words, since both of them have been assigned meanings on the basis of synchronic textual analysis. The main morphological problem complicating the comparison between ali-wa/i-nÌ(-) (n.) and /alunni-/ (c.) is the gender mismatch. Since the word for `enemy' is a priori likely to be a common gender noun, it is the Bronze Age Luvian form that requires an explanation. It is furthermore important to observe that ali-wa/i-nÌ(-sa) `enemy' must contain a suffix that is not overtly present in its attested case forms because there are no neuter i-stems in Luvian. I am happy to accept the solution offered to me by Elisabeth Rieken: the form in question is to be analyzed as {aliwann(a/i)}+{it}+{sa}, where -it- is a suffix deriving nouns from other nouns and adjectives (Melchert 2003: 198).18 The extended suffix -ahi(t) is much more frequent in Luvian than the simple -i(t), but the latter can be seen, for example, in the the abstract noun nanuntarrisa `the present' (nom./acc. sg. n.), morphologically {nanuntarr(a/i)}+{it} +{sa}, a derivative of *nanuntarra/i- `present' (Starke 1990: 177, Melchert 1993: 155­156). In a similar fashion, aliwann-it- must have been derived from the base adjective *aliwanna/i-. The original semantics of the derived noun was `enemy power' (vel. sim), hence `enemy troops', and finally the generic `enemy'.19 The Luvian adjective *aliwanna/i- `inimical' can in turn be derived from *ali(ya)- < Indo-Hittite *alyo- `other', whose direct reflex is attested in Anatolian as Lyd. aa- `id.' (Gusmani 1964: 56). Its earliest meaning must have been `foreign, alien', while the semantic shift to `inimical' can be compared with Lat. hostis `enemy' < IE. *ghosti- `foreigner, alien'. The formation of *ali-wanna/i- based on *ali(ya)- would be structurally similar to that of Lat. ali#nus `belonging to the other; foreign, alien' derived from alius `other', although the Luvian and Latin derivational suffixes are not cognate. The productive Luvian suffix -wanna/i- is mostly attested in adjectives formed from toponyms, but there are also cases where it is attached to the stems for animate beings.20 The adjective *aliwanna/i- had
18

19

20

For the morphophonemic rules responsible for the deletion of the final consonant in -i(t)-, see Melchert 2003: 182. For the last stage of the postulated semantic change, compare the evolution of the English `troop'. Originally used to designate a group of people (< Lat. truppa `flock'), this word can now apply to individual soldiers, notably in numerical expressions (e.g. `4100 troops'). Thus, I suggest interpreting TýNP, § 3­4 [SU]B-na-na ("TERRA")ta-sa-REL+ra/isa i-LITUUS-wa/i-ni-sa |SUPER+ra/i-la+ra+a-pa-wa/i ("CAELUM")ti-pa-sÀ |i-tÞwa/i-za as `The earth below belongs to Ea, while the heaven above belongs to El'


The Luvian Enemy

17

apparently coexisted with its nominal derivative aliwann-it-, but eventually also became substantivized with the meaning `enemy'.21 Once this word was no longer an adjective, it probably came to be perceived as a single morphological unit, since the suffix -wanna/ilost its characteristic function. Under such conditions, it could have undergone an optional syncope yielding Iron Age Luvian alunna/i- > [anna/i-] `enemy'. The phonetic side of the contraction *aliwanna/i- > alunna/irequires special comment. While the developments /-iya-/ > /-i-/ and /-uwa-/ > /-u-/ are well-attested in Luvian (Melchert 2003: 183), there is no independent lexical evidence for the change /-iwa-/ > /u/. The intermediate stage of the contraction in our stem can, however be observed in Luv. *alwan-za- `inimical', which I reconstruct as a base of Hitt. alwanzahh- `to bewitch' and secondary Hittite derivatives, such as alwanz!tar, alwanzessar, alwanzahha- `witchcraft', and alwanzena- `sorcerous, sorcerer'.22 Presumably, black magic was conceived as an act of alienating its victim from the gods and rendering him/her abject in their eyes. Perhaps the irregular syncope of -i- in *aliwanna/i- first emerged in the derivatives of this lexeme in the context of intensive contacts between Luvian and Hittite and then spread by analogy to the base form. The subsequent syncope *alwanni- > alunni- finds a close parallel in the development of the Luvian 1. pl. ending *-wanni > *-unni (Melchert 1994: 276).
References Bryce, Trevor. 2005. The Kingdom of the Hittites. Oxford. ­ 2007. The Secession of Tar"untassa, Tabularia Hethaeorum: Hethitologische BeitrÄge Silvin Kosak zum 65. Geburtstag. Ed. D. Groddek and M. Zorman. Wiesbaden. Pp. 119­29.
(cf. Hawkins 2000: 155). For the sporadic use of L 41 = for the etymological sequence */la/ in late Iron Age Luvian, see Rieken, forthcoming. Compare the interchanging use of the noun KUB 54 iii 3 nanuntarri-sa `the present' and the etymological adjective KUB 52 iii 12 [nanuntarri]yan-za `present' in two duplicates of the same Luvian incantation (Starke 1990: 177). I owe the idea of a connection between *aliwanni- and Hitt. alwanzahh- to Elisabeth Rieken, although the morphological details are mine. The attempts to derive Hitt. alwanzahh- and its cognates from the putative adjective **alwant- are phonetically problematic and do not yield a plausible etymology for this family of words (cf. Puhvel 1984: 46 and Kloekhorst 2007: 171). For the Luvian suffix -(z)za- forming adjectives and names of profession, see Melchert 2003: 197.

21

22


18

Ilya Yakubovich

DinÃol, Ali, Belkis DinÃol, David Hawkins, and Gernot Wilhelm. 1993. The `Cruciform Seal' from Bo`azkÆy-Hattusa, Istanbuler Mitteilungen 43: 87­112. Goedegebuure, Petra. 1998. The Hieroglyphic Luwian Particle REL-i=pa, Acts of the IIIrd International Congress of Hittitology (ãorum, September 16­22, 1996). Ed. S. Alp and A. SÝel. Ankara. Pp. 233­245. Gusmani, Roberto. 1964. Lydisches WÆrterbuch. Heidelberg. Jasink, Anna Margherita. 2001. Suppiluliuma and Hartapu: two "Great Kings" in Conflict, Akten des IV. Internationalen Kongresses fÝr Hethitologie, WÝrzburg, 4.­8. Oktober 1999. Ed. G. Wilhelm. StBoT 45. Pp. 235­240. Herbordt, Suzanne. 2005. Die Prinzen- und Beamtensiegel der hethitischen Grossreichszeit auf Tonbullen aus dem NiÂantepe-Archiv in Hattusa. BoHa 19. Mainz. Hawkins, J. David. 1990. The New Inscription from the SÝdburg of Bo`azkÆy/"attusa, ArchÄologischer Anzeiger 1990/3: 305­14 ­ 1995. The Hieroglyphic Inscription of the Sacred Pool Complex at Hattusa (SýDBURG). StBoT, Beiheft 3. Wiesbaden. ­ 2000. Corpus of Hieroglyphic Luvian Inscriptions. Volume I. Part I, II: Texts; Part III: Plates. Berlin­New York. ­ 2003. Scripts and Texts, The Luvians. Ed. C. Melchert. HdO 1/68. Leiden­Boston. Pp. 128­69. ­ 2005. Commentaries on the Readings, Herbordt 2005: 248­303. ­ 2006a. Tud"aliya the Hunter, The Life and Times of Hattusili III and Tud"aliya IV. Proceedings of a Symposium Held in Honour of J. de Roos, 12­13 December 2003, Leiden. Ed. Th. van den Hout. PIHANS 103. Leiden. Pp. 49­76. ­ 2006b. The Inscription, Guy Bunnens. A New Luwian Stele and the Cult of the Storm-god at Til Barsib-Masuwari. Louvain. Pp. 11­31. Hawkins J. David and Morpurgo-Davies, Anna. 1998. On Donkeys, Mules, and Tarkondemos. MÌr Curad: Studies in Honor of Calvert Watkins. Ed. J. Jasanoff et al. Innsbruck: Institut der Sprachwissenschaft der UniversitÄt Innsbruck. Pp. 243­60. Kloekhorst, Alwin. 2007. Etymological Dictionary of the Hittite Inherited Lexicon. Leiden. Laroche, Emmanuel. 1958. ètudes sur les hiÈroglyphes hittites. Part 2, Syria 35: 252­83. ­ 1966. Les noms des Hittites. Paris. ­ 1969. Les hiÈroglyphes hittites. Paris. ­ 1981. Les noms des Hittites: supplÈment, Hethitica 4: 3­58. Marazzi, Massimiliano. Il geroglifico anatolico: problemi di analisi e prospettive di ricerca. Rome. Melchert, H. Craig. 1993. Cuneiform Luvian Lexicon. Chapel Hill: Selfpublished. ­ 1994. Anatolian Historical Phonology. Amsterdam.


The Luvian Enemy

19

­ 1997. Review of Hawkins 1995, BSOAS 60: 347. ­ 2002. Tar"untassa in the SýDBURG Hieroglyphic Inscription, Recent Developments in Hittite Archaeology and History: Papers in Memory of Hans G. GÝterbock. Ed. A. Yener and H. Hoffner. Winona Lake, IN. Pp. 137­143. ­ 2003. Language, The Luwians. Ed. C. Melchert. HdO 1/68. Leiden­Boston. Pp. 170­210. ­ 2006. Indo-European Verbal Art in Luvian, Langue poÈtique indoeuropÈenne. Ed. G. Pinault and D. Petit. Leuven­Paris. Pp. 291­8. Meriggi, Piero. 1962. Hieroglyphisch-hethitisches Glossar. Second, fully revised edition. Wiesbaden. Poetto, Massimo. 1993. L'iscrizione luvio-geroglifica di Yalburt, StMed 8. Pavia. Puhvel, Jaan. 1984. Hittite Etymological Dictionary. Vol. 1: Words beginning with A. Vol. 2: Words beginning with E and I. Berlin. Rieken, Elisabeth. 2007. Lat. #g% ,fÝhrte`, i#c% ,warf` und h.-luw. INFRA a-ka ,unterwarf`, Verba Docenti: Studies in Historical and Indo-European Linguistics Presented to Jay Jasanoff by Students, Colleagues, and Friends. Ed. A. J. Nussbaum. Ann Arbor/New York. Pp. 293­300. ­ forthcoming. Die Zeichen , und in den hieroglyphenluwischen Inschriften der Nachgroúreichszeit, Atti del 6o Congresso di Ittitologia. Ed. A. Archi und R. Francia. Rieken, Elisabeth and Yakubovich, Ilya. The New Values of Luvian Signs L 319 and L 172, to appear in a forthcoming Festschrift. Starke, Frank. 1990. Untersuchungen zur Stammbildung des keilschriftluwischen Nomens. StBoT 31. Wiesbaden. Yakubovich, Ilya. 2008. Sociolinguistics of the Luvian Language. University of Chicago PhD dissertation, available at http://oi.uchicago.edu/research/ library/dissertation. Abstract The lexeme 416-wa/i-nÌ-, occurring in Bronze Age Luvian hieroglyphic inscriptions, was previously interpreted as a royal title. The goal of this paper is to demonstrate that it rather represents a common noun `enemy (force)'. The new hypothesis entails a syntactic reinterpretation of a number of Luvian sentences and allows one to arrive for the first time at a cohesive understanding of the SýDBURG inscription. On the phonetic level, Bronze Age Luvian *aliwanni- `alien' can be now compared with the independently established Iron Age Luvian *alunni- `enemy'.