The Luwian Title of the Great King

Ilya Yakubovich

The research on this paper was conducted within the framework of the project Digitales philologisch- etymologisches Wörterbuch der altanatolischen Kleinkorpussprachen (RI 1730/7-1) funded by the Deutsche Forschungsgemeinschaft. I am grateful to George Dunkel, Stephen Durnford, Federico Giusfredi, Craig Melchert, Annick Payne, Elisabeth Rieken, Diether Schürr, Turna Somel, and Mark Weeden, with whom I had a chance to discuss various topics pertaining to the content of this paper. Final responsibility remains, of course, my own.

Introduction

  • 1 Artzi/Malamat 1993: 31.
  • 2 Cf. Hoffner 2009: 323 with en. 322.
  • 3 Otten 1981: 22, iii 69-70. Note, however, that Otten understood the Akkadogram HALS. I as ‘fortress (...)

1The title Great King was understood in the Late Bronze Age Near East as a king who is powerful enough to secure the loyalty of the neighbouring kings.1 It was apparently quasi-synonymous with the somewhat later title King of Kings. Thus the 13th century Assyrian king Tukulti-Ninurta I, who was the first one to use the title King of Kings, was also the first ruler of Assyria to extract a grudging recognition of his status as a Great King from the rulers of Hattusa.2 In the Apology of Hattusili III the title ‘great king’ is contrasted with the expression ‘king of one province’. This contrast also speaks for the hierarchical interpretation of the title under discussion.3

  • 4 Whether ura(/i)- ‘great’ should be analyzed as an a-stem or semi-vocalic stem is somewhat unclear b (...)
  • 5 Giusfredi 2010: 81. 6 Payne 2014: 156.
  • 1

2The Hittite or Luwian designations of the Great King are never spelled fully phonetically in the published texts. Usually they are hidden under the spurious Sumerogram LUGAL.GAL in the cuneiform and the complex logogram MAGNUS.REX in the Anatolian hieroglyphic corpus. The conventional Hittite reading of LUGAL.GAL as salli- hassu-, lit. “great king” will be discussed in Section 8. Since the Luwian words ura/(i)- ‘great’ and hantawatt(i)- ‘king’ are both known in syllabic transmission, it is likewise assumed that the Luwian reading of MAGNUS.REX represents their combination.4 It is, furthermore, usually taken for granted that the order of the Luwian constituents is the same as that of the corresponding logograms. Thus Federico Giusfredi affirms that “[t]he reading of the two logograms MAGNUS.REX may be postulated as *ura(zza)-*handawati-”5 and a similar reading ura-hantawat(i)- is offered in the latest manual of Hieroglyphic Luwian prepared by Annick Payne.6

3Contrary to the present consensus, I intend to propose a new reading hantawatt-(an)-ura/(i)-, lit. “the great(est) of kings” for the Luwian title in question. I shall begin my analysis by addressing the syntactic and semantic structure of various classes of compounds containing the morpheme ura/(i)- ‘great’. The proposed interpretation of the complex logogram MAGNUS.REX ‘great king’ identifies this title as a member of a particular compound class described for the first time by Emmanuel Laroche. This identification is corroborated through adducing additional titles belonging to the same class, as well as through combinatory analysis of Luwian compounds containing the elements ‘great’ and ‘king’. In conclusion I will argue that the phonetic interpretation of the Sumerographic title LUGAL.GAL need not impact the proposed understanding of the matching hieroglyphic title MAGNUS.REX, since no Hittite reading of LUGAL.GAL can be regarded as certain.

Compounds beginning in ura/(i)-

  • 7 Most of the relevant examples can already be found in Laroche 1966.
  • 8 Melchert 2013: 41.
  • 9 For the assumed meaning of walkuwa/i- ‘lion’ see Lehrman 1987, but the i-mutation reconstructed in (...)
  • 10 This personal name (KBo 47.11 obv. 7), not yet found in Laroche 1966, is listed in the online suppl (...)
  • 11 For the productive pattern of Luwian “topophoric” names with theophoric interpretation, see Yakubov (...)

4The compounds of the Empire period beginning with ura/(i)- ‘great’ are restricted to personal names.7 Recently the members of this group attested in cuneiform transmission received summary interpretative treatment as descriptive compounds.8 In my opinion, such an analysis is assured only for mUra-walkui- “great lion”, a fitting name for a warrior.9 As for the theophoric names, mUra-dU and mGAL.dIŠTAR-a-10, their interpretations as predicative compounds, respectively “Tarhunt (is) great” and “Šawoška (is) great”, appear to be pragmatically more attractive. The first of these two names may represent a calque of Hurrian Talmi-Teššub “Teššub (is) great”, while the second one can be either Luwian-Hurrian (i.e. Ura-Šawoška-) or completely Hurrian (i.e. Talmi-Šawoška). Although their interpretations as “great Tarhunt” or “great Šawoška” would not require Hurrian inspiration, they seem rather unlikely on semantic grounds, as few individuals would dare give such hubristic names to themselves or their children. Continuing the same line of reasoning, I would propose interpreting mUra-Hattusa not as “great Hattusa”, but either as “Hattusa (is) great” or, more likely, as an elliptic theophoric compound “(the Storm-god of) Hattusa (is) great”.11

  • 12 Here and below, the precise citations of all the Iron Age Luwian forms in hieroglyphic transmission (...)
  • 13 Cf. Yakubovich 2010b: 396, fn. 9., where the name under discussion is erroneously analyzed as Luwia (...)

5The compound names beginning in ura/(i)- ‘great’ that survived into the Early Iron Age are likewise restricted to personal names. The predicative compounds, which continue the earlier pattern of likely Hurrian inspiration, are found in Carchemish. These are Ura-Tarhunt- “Tarhunt (is) great” (KARKAMIŠ A4b, KARKAMIŠ A11b+C, CEKKE), Ura-Sarma- “Šarruma (is) great” (KARKAMIŠ A4a), and its close variant Urahi-Sarma- “Šarruma (is) greatness” (KARKAMIŠ A2+3).12 On the other hand, several personal names coming from other Neo-Hittite states can be analyzed as possessive compounds. The most obvious among them is Ura-muwa- “(having) great might” (KULULU lead strip 1). I had a chance to propose the same syntactic analysis for the name of Ura-hilana- / Ura-hilina-, king of Hama, which, in my opinion, could be synchronically interpreted as “(having a) great gate” regardless of its original Hurrian etymology.13

  • 14 These onomastic compounds can be typologically compared with Italian last names of the type Casa-gr (...)

6A consideration that supports this analysis is the possibility to interpret the name Ura-dam(i)-, which belongs to Urahilina’s son and likewise occurs in Hama inscriptions, as “having (a) great building”. Apparently the kings of Hama had a weak spot for architectural forms.14

Compounds ending in ura/(i)-

  • 15 Melchert (2013: 41) interprets the same name differently, as “great (one) of the gods”. I analyze t (...)
  • 16 Cf. HED H: 238. On the prosopography of mḪa-aš-ta-nu-ri see Singer 2003: 343-344.

7When one turns to compounds ending in -ura/(i)-, a different picture emerges. In this group, too, one occasionally encounters personal names, for example, Massana-ura- “great among the gods”.15 More frequently, however, the determinative compounds of this structure appear to denote administrative titles. It was the honorand of this volume, Emmanuel Laroche, who identified the first Anatolian titles in -ura/(i)-in Akkadian texts from Ugarit. These are RS 11.732 tu-up-pa-nu-ri, RS 11.732, hu-bur-ta-nu-ri, RS 16.180 hu-bur-ta-nu-ru, RS 17.382+380 tu-up-pa-nu-ri and RS 17.227 tu-up-pa-la-nu-ri (var. -ra). Laroche compared these titles with the personal name mHa-aš-ta-nu-ri, occurring in the same corpus (RS 17.251).16 He made the following etymological observation:

  • 17 Laroche 1956: 28. Note also RS 34.126 tu-pal-nu-ri and RS 92.2007 tup-pa-la!-nu-ri (Gordin 2008: (...)

“Il suffit d’attribuer ces noms à la langue hittite pour en apercevoir aussitôt la structure. Ce sont les composés, plus exactement des juxtaposés, d’un nom au génitif pluriel en -an + uri- « grand ». tuppan-uri- est « le grand des tablettes », tandis que tuppalan-uri- est « le grand des scribes » ; le sens de huburtan n’est pas connu. Pour haštan-uri- se rapportant à un personnage de rang royal, on songe à hitt. ḫaššant- « né, e.g. légitime, du sang »”.17

  • 18 Laroche 1965: 37.

8Nine years later Laroche had a chance to reaffirm the same hypothesis.18

  • 19 Malbran-Labat/Lackenbacher 2005: 9. The phonetically sensitive rendering of this Hurrian name would (...)
  • 20 Singer 2006: 244. It is worth noting that this equation represents an argument for the traditional (...)

9Later research demonstrated that Laroche’s basic insight has fully stood the test of time. The late 13th or early 12th-century tablet RS 94.2523, found in Ugarit in 1994, contains the pair of titles tu-up-pi-nu-ra hu-pu-ur-ti-nu-ra that is attached to the name of a certain Penti-Šarruma.19 It turns out that there are bullae from the same region and period containing the hieroglyphic imprints of same name Penti-Šarruma alongside official titles MAGNUS.SCRIBA ‘Chief Scribe’, MAGNUS.AURIGA ‘Chief Charioteer’, and MAGNUS.DOMUS.FILIUS ‘Chief Palace Attendant’.20 On the likely assumption that there was just one high-ranked official named Penti-Šarruma, one can propose a direct equation between the titles tu-up-pi-nu-ra and MAGNUS.SCRIBA. The comparison between the first element of tuppan-uri- / tuppin-ura- and Hittite tuppi- ‘tablet’, implied in Laroche’s reasoning, remains as valid as ever, and so is the comparison between the first element of tuppalan-uri- and Luwian SCRIBA-la- = tuppala- ‘scribe’. One may, however, doubt, that tuppalan-uri- and tuppan-uri- represent two different titles, since the functional distinction between “great of the scribes” and “great of the tablets” is not easy to grasp, and at any rate it would not correspond to any meaningful distinction between Sumerographic or hieroglyphic titles. It is easier to assume that tuppan-uri-and similar forms came about as an abbreviation of tuppalan-uri- ‘Chief Scribe’, lit. “great of the scribes”, while “great of the tablets” may have represented a convenient folk etymology.

  • 21 On the etymology of Luw. hastall(i)- ‘hero’ see Starke 1990: 122, 124.
  • 22 Singer 2006: 244.

10The hypothesis of a morphologically conditioned abbreviation also comes in handy in dealing with the personal name mHa-aš-ta-nu-ri. Laroche’s idea of syncope *hassant- > hast- is not supported by any parallels and therefore appears to have little to recommend itself. On the other hand, the hypothesis that Hastan-uri represents a shortened variant of *Hastallan-uri immediately yields an auspicious name with plausible semantics “great(est) of the heroes”. In this case, too, the shortening of the name might have been mediated by Luw. hast- ‘bone’, of which Luw. hastall(i)- ‘hero’ represents a derivative.21 Only in the instance of the title huburtan-uri are we as much in the dark regarding the first element of this compound as at the time of Laroche. The etymology of the element huburt(V)- remains unknown, while the combinatory method does not offer a way to decide whether ḫu-pu-ur-ti-nu-ra as a title of Penti-Šarruma corresponds to MAGNUS.AURIGA ‘Chief Charioteer’ or MAGNUS.DOMUS.FILIUS ‘Chief Palace Attendant’. In fact, it may correspond to neither of the two hieroglyphic titles, since all three of them may ultimately represent different stages in Penti-Šarruma’s career.22

Language of titles ending in ura/(i)-

  • 23 Laroche 1959: 102.

11Another point where Laroche’s analysis can be improved concerns the language of compounds in -ura/(i)-. It is true that at the time when Laroche originally labelled them as “hittite” little distinction was generally made between Hittite proper and other members of the Anatolian language family. Now, of course, we know that ura/(i)- is the standard term meaning ‘great’ in Luwian, while its standard Hittite equivalent is salli- ‘great’. Accordingly, all the compounds in ura/(i)- can be taken as Luwian formations unless proven otherwise. This is in line with their prominence in the onоmastics of the Empire of Hattusa, which was dominated by names of Luwian origin. It is, however, interesting that Laroche insisted on the Hittite character of the compound tuppalan-ura- not only in his pioneering article but also in the dictionary of the Luwian language, where ura/(i)- in its other occurrences is properly analyzed as a Luwian adjective.23 There are two considerations that could sway him in favour of such a solution: the oddity of a Luwian title embedded in Akkadian discourse and the possibility of interpreting the element -an- in tuppal-an-ura- as a Hittite genitive plural marker.

  • 24 Yakubovich 2010a, Chapter 5.

12The first consideration, perfectly understandable within the context of mid-twentieth-century Anatolian studies, loses its cogency in the face of recent advances in understanding of the sociolinguistic situation in Bronze Age Asia Minor. As long as Luwian was taken as a peripheral language of the Empire of Hattusa, while the Luwian forms embedded in Hittite texts were attributed to the mediation of semi-literate provincial scribes, it would indeed remain unclear why Luwian and not Hittite was chosen for rendering imperial titles in official Akkadian texts. The perspective changes completely if one admits that Luwian and not Hittite was the main spoken language in Hattusa in the thirteenth century BC, while the king and members of the royal family were bilingual in Hittite and Luwian in the period under discussion.24 The titles embedded in Akkadian texts were free of the conventions of Hittite orthography, and as long as the scribes were unwilling or unable to render them in Akkadian, it was only natural for them to fall back upon the main colloquial language of the Empire, which happened to be Luwian. Indeed, it would be the use of Hittite, as opposed to Luwian, to require special pleading under such conditions.

  • 25 Yakubovich 2010a: 45-46.
  • 26 Yakubovich 2010a: 47-53.
  • 27 Cf. Melchert 2012: 275.
  • 28 As a parallel for a genitive case marker developing into an interfix, one can consider the situatio (...)

13The second potential objection is based on the contrastive synchronic analysis of Hittite and Luwian grammars. The genitive plural ending -an is attested in Old Hittite, but the Luwian grammar shows no formal distinction between the expression of singular and plural genitive: the same ending -a-si-i can be deployed for both in hieroglyphic transmission.25 Only in the Luwian dialect of Kizzuwadna, where genitive endings were completely replaced with possessive suffixes, an innovative suffix -assanz- was calqued on a Hurrian model to indicate plural possession.26 But from the historical viewpoint, the Hittite morpheme -an under discussion clearly represents an archaism, because it continues the Early Indo-European genitive plural ending *-om. Furthermore, the Lycian genitive plural ending -assures that the reflexes of *-om also existed in Proto-Luwic, the common ancestor of Luwian and Lycian.27 Therefore the Luwian interfix -an- in determinative compounds can simply be taken as a vestige of the genitive plural maker, which outlived the generalization of genitive singular endings in word-final position.28

Origin of titles ending in ura/(i)-

14A separate question is how the compounds of this type could come into being in Luwian. If one accepts that tuppal-an-ura- and similar forms are not Hittite, then Hittite does not appear to feature the affix -an- as compound linker. Neither can one claim that this suffix represents an Indo-European archaism, since compounds linked by *-om- cannot be reconstructed for Early Indo-European (Indo-Anatolian). The remaining alternative is to treat the compounds of this type as Luwian or Luwic innovations, but in order to make such an interpretation credible, one should first identify a syntactic construction that could provide a source for the new formation.

  • 29 See Yakubovich 2013b for Luwian and GrHL: 273-276 for Hittite. In my opinion, the Luwian function o (...)
  • 30 On the left-hand branching word order in Luwian, see Bauer 2014: 36-37 and passim.

15In my opinion, tuppal-an-ura- and similar titles may represent a morphologization of the Proto-Luwian superlative construction. Although I have argued that the Luwian adjectives secondarily developed comparative and superlative forms in -(a)zza-, it remains very likely that at an earlier stage of its development they lacked morphological expression for the degree of comparison, just as was the state of affairs attested in Hittite.29 Furthermore the genitive case was used for the expression of scope in the Luwian superlative construction. This phrase displays the expected left-branching word order, which is to say the noun expressing scope appears in front of the superlative adjective,30 as illustrated through the following Late Luwian example.

  • 31 Yakubovich 2013b: 161, cf. Hawkins 2000, II: 528.

(1) PORSUK § 531

|u-mi-i

|EXERCITUS-la/i/u-na-sa8

|MAGNUS+ra/i-za-sa

á-sa-ha

PTCL=REFL

army.GEN

greatest.NOM.SG.C

‘(King Masaurahissas(?) was well disposed toward me) and I was the greatest (=commander) of (his) army’.

16According to my reconstruction, at the time when Pre-Luwian still distinguished between singular and plural genitives and had not yet grammaticized the superlative suffix -(a)zza-, the phrase ‘greatest among the heroes’ sounded approximately like *hastallan ura-. The semantic interpretation of such phrases as titles or personal names could easily trigger their morphological reanalysis as compounds. Thus determinative compounds with the last element ura/(i)- appear to be younger than their counterparts beginning in ura/(i)-. This agrees with the fact that models of composition displayed by the compounds of the latter group and discussed in Section 2 are overall typical of the ancient Indo-European languages.

Late Luwian titles ending in ura/(i)-

17To be sure, skeptics among philologists may question etymological considerations as probative arguments for the Luwian origin of ura/(i)-titles, unless their specimens are attested in the actual Luwian texts. The requisite examples come from Late Luwian. Their main difference from the Empire period titles attested in Akkadian texts is the absence of the linker -an-, but otherwise they conform to the structural pattern outlined in Section 3. The most salient example yields us the Luwian title of the Chief Eunuch. Although this title had already been identified in the standard edition, the assumption that it constitutes a noun phrase and not a compound left Hawkins with the transliteration (“*474”)u-[si]-na-SU! MAGNUS+ra/i-sa. The phonological interpretation ussinassura/(i)-, straightforward under the new analysis, immediately explains the function of the <su> sign.

  • 32 Cf. Hawkins 2000, I: 265 and Hawkins 2002: 230.

(2) MARAŞ 14 § 132

[E]GO [

...

]-si-i-sa

|IUDEX-ni-sa

|HEROS-sa

I.NOM

?.NOM.SG.C

of.ruler.NOM.SG.C

of.hero.NOM.SG.C

(“*474”)u-[si]-na-su-MAGNUS+ra/i-sa chief.eunuch.NOM.SG

‘I [am Astiwasu], Chief Eunuch of [X], the ruler, the hero’.

  • 33 Cf. the previous analysis of amurallura/(i)- in Giusfredi 2010: 162: “No etymology can be provided (...)

18The interpretation of the compounds amurallura/(i)- and astaruri(ya)- (vel sim.) as titles in -ura/(i) also imposes itself in the two cases below, even though the precise functions of these titles remains elusive.33

  • 34 Cf. Hawkins 2000, II: 537.
  • 35 Cf. Hawkins 2000, II: 508-509.

(3) ASSUR f+g § 3334

˹|a˺-wa/i

˹|á˺-pi [|DOMUS]-ni-wa/i+ra/i-ia

PTCL=PTCL then

Parniwarri.DAT.SG

[|(X)]á-mu+ra/i-la/i/u+ra/i-´

˹|a˺-sa-ti

Amurallura.DAT.SG

be.3SG.PRS

‘(Send us any kapara, and if you do not have it), but Parniwarri the amurallura has it, (then buy it from him and send it to us)’.

(4) KULULU lead strip 1, #4335

200 “*179”-za-´

Ihu-li-ia-ia

|CUM-ni

|á-˹sa?-tara/i?˺-MAGNUS+ra/i?-ia

200 grain.ACC.SG

Huliya.DAT.SG

for

astaruriya.DAT.SG

‘200 (measures of) grain for Huliya, the astaruri’.

  • 36 Cf. Hawkins 2000, II: 443.

19In purely formal terms, the first component of á-mu+ra/i-la/i/u+ra/i- in (3) appears to be cognate with the noun á-mu+ra/i- of unclear meaning (KULULU 1 § 11).36 It is tempting to reconstruct it as amurall(i)- on the assumption that it is derived from amura/(i)- on the same model as e.g. hastall(i)- ‘hero’ from hast- ‘bone’. In the instance of á-˹sa?-tara/i?˺-MAGNUS+ra/i?- it does not even seem productive to speculate about the etymological connections of its first morpheme in view of its fragmentary state of preservation. But both contexts (3) and (4) are well compatible with the mentions of official titles, since the forms under discussion follow personal names and agree with them in case in both passages. The absence of personal determinatives rules out the interpretation of the same forms as additional names.

  • 37 Compare the title uryall(i)- of a certain Nunuya mentioned as a recipient of sheep in KULULU lead s (...)

20In the case of (4) there are, of course, certain complications. The position of CUM-ni is easy to explain, because this postposition regularly separates personal names from the accompanying titles in the allocation list KULULU lead strip 1. More intriguing is the dative-locative ending -ia attached to the stem á-˹sa?-tara/i?˺-MAGNUS+ra/i?-, which implies an etymological ya-stem adjective derived from base noun astarura- (vel sim.). I prefer to think that this derivation did not radically alter the sense of the form in question and is compatible with its translation along the lines ‘having the function of astarura-’,37 but one cannot altogether exclude the possessive reading ‘belonging to astarura-’. The last interpretation is grammatically more straightforward but yields a unique interpretation, since the recipients of grain in KULULU lead strip 1 are normally defined through patronymics, title or places of origin, but not through the titles of individuals they belong or are related to.

21Despite the difficulties stated above, we have obtained a clear proof that the compounds ending in ura/(i)- are well at home in Late Luwian, which in turn corroborates the Luwian character of the similar compounds attested in Akkadian texts. The disappearance of the linker -an- in the first millennium BC need not amaze us, since we have seen that it represents an isolated archaism even in the forms of the late second millennium BC. Furthermore, the comparison between Bronze and Iron Age forms yielded five titles ending in ura/(i)-, as opposed to two personal names, Massanaura and Hastanuri, having the same structure. This is in stark contrast to the group of compounds beginning in ura/(i)- treated in Section 2, which entirely consists of personal names.

Back to MAGNUS.REX

22Against such a background we can return to the interpretation of the hieroglyphic title MAGNUS.REX ‘great king’. The structure of other Luwian titles ura/(i)- treated in this paper would suggest that if this title was indeed a compound, its Luwian reading must have been *hantawatt-an-ura/(i)- lit. “great(est) among the kings” in the Empire period and probably *hantawatt-ura/(i)- in the Neo-Hittite period. Nonetheless, one should not a priori exclude a theoretical possibility that the Luwian title for the Great King did not form a compound, but represented a noun phrase urazza- hantawatt(i)- “the greatest king”, or something similar. In order to refute such a reconstruction, one should turn to positive arguments pertaining to Anatolian compounds with elements ‘great’ and ‘king’.

23The most important piece of evidence is a passage from the Late Luwian AKSARAY inscription, which contains a unique instance of the title under discussion spelled with phonetic complementation.

  • 38 For the most recent edition of the AKSARAY inscription see Hawkins 2000, II: 475-478, where the ear (...)

24The fact that only one of its stems is provided with an inflectional ending speaks strongly for its compound character, because the inflectional morphology is otherwise overtly expressed in the sentence under discussion. The drawing in (5) below reproduces a fragment that has been read up to now as |MAGNUS-RA/I-REX-zi |REX-ti-zi ‘great kings (and) kings’.38 The proponents of this reading failed, however, to explain why the first stem of the compound MAGNUS.REX acquires a phonetic complement, whereas the second one does not. The contrast between the complements (-)REX-zi REX-ti-zi in the two immediately adjacent words likewise remains begging a question under the traditional interpretation. Therefore I would like to propose that the same group of signs is to be read as |REX.MAGNUS+ra/i-zi |REX-ti-zi. Under such a reading, each of the two coordinated nouns acquires a phonetic complement pointing to the last consonant of its lexical representation and its inflectional ending. The implied phonetic reading of the coordinated pair is hantawatturinzi hantawattinzi.

(5) AKSARAY § 6

za-ti-pa-wa/i-ta

URBS-ni

|REX.MAGNUS+ra/i-zi

this.DAT.SG=but=PTCL=PTCL

town.DAT.SG

great.king.NOM.PL

|REX-ti-zi

¦OMNIS-mí-zi

INFRA-tá-ta

OCULUS(-)zá-ni-ta

king.NOM.PL

all.NOM.PL.C

below

admire ?3PL.PRT

‘And all the great kings and kings admired this town’

  • 39 For the role of aesthetic considerations in determining the order of Anatolian hieroglyphs, which r (...)

25It is true that the order of the signs <REX> and <MAGNUS+ra/i> is a non-canonical one under the new interpretation. This irregularity, however, finds a close match in the divergent order of the signs <zi> and <ti> in the following noun. This does not, of course, imply that the scribe strove for a symmetrical pattern of inversion in the two coordinated nouns. One should rather acknowledge that the order of signs in Anatolian hieroglyphic inscriptions can be inverted for a variety of reasons and generally represents a weaker guide to arriving at their correct transliteration than combinatory constraints. One reason for the graphically inverted sequence of graphemes for ‘king’ and ‘great’ is the fact that the logogram <MAGNUS> is consistently placed on top of <REX> when the two signs form a ligature.39

  • 40 The same name also occurs in Greek transmission as Κενδαβυρα / Κινδαβυρις (Melchert 2004: 109).
  • 41 This is not the only cases when personal names of Lycian B origin are embedded in Lycian A texts. C (...)
  • 42 The first interpretation is advocated by Schürr (2001: 105) and endorsed by Shevoroshkin (2011: 594 (...)

26The other relevant piece of evidence comes from Lycian. Although the personal name xñtabura occurs twice in the Lycian A corpus (TL 103,2; 125b),40 its origin can be safely assigned to the Lycian B (“Milyan”) language based on its formal features.41 The first part of this compound is probably not to be separated from Lyc. (B) xñtaba-, a noun pertaining to the sphere of kingship and cognate with Luw. hantawatt(i)- ‘king’, while its second part is the familiar element ura- ‘great’. Whether Lyc. (B) xñtaba- means ‘ruler’ or ‘rule’ is still a matter of debate,42 but even if the second opinion should be given more weight, this need not fundamentally alter the analysis of the name xñtabura. Since the nouns xñtawat(i)- ‘ruler, king’ and xñtawata- ‘kingship’ coexist in the Lycian A language, it is intrinsically likely that the situation in the closely related Lycian B language was roughly the same. This is to say, even if Lyc. (B) xñtaba- meant something like ‘kingship’, there probably existed also a cognate Lyc. (B) xñtab(i)- meaning ‘king’ or ‘ruler’.

  • 43 Neumann 2007: 240.

27The precise meaning of the compound xñtabura- cannot be determined with certainty. The auspicious name ‘great king’, lit. “(the) greatest of kings” remains a distinct possibility, especially given the fact that the brother of its carrier was called lusãñtra- “Lysander” in the Lycian inscription TL 103.43 As a more mundane alternative, one can envisage the interpretation “great among the kings”, which implies a wish for the favourable disposition toward xñtabura- on the part of the rulers of this world. Under the latter interpretation, the compound under discussion is typologically similar to the Empire Luwian name Massana-ura- “great among the gods”, discussed in Section 3. But whichever of these two solutions one prefers, one winds up with a determinative compound that represents a close formal match of the reconstructed Luwian *hantawatt-ura/(i)-. Given the combined positive evidence of Late Luwian and Lycian B, which is typologically in agreement with the structure of other Luwian compounds for superior officials, the proposed reading of the complex logogram MAGNUS.REX can be regarded as substantiated.

The title LUGAL.GAL

  • 44 Thus e.g. Steiner 1999: 428, Vanséveren 2006: 125. Weeden (2011: 571) reconstructs the Hittite noun (...)
  • 45 On which see Rieken 2006.

28It is appropriate to conclude this paper by preempting a possible objection coming from the side of cuneiformists. The Anatolian hieroglyphic title MAGNUS.REX was used for the Great King of Hattusa alongside the Sumerographic cuneiform title LUGAL.GAL ‘Great King’. It is frequently assumed that the Hittite reading of LUGAL.GAL was *sallis hassus, lit. “great king”.44 The apparent morphosyntactic mismatch between the Hittite and Luwian titles would stand incongruous with the progressive grammatical convergence between Hittite and Luwian in the Empire Period.45

  • 46 The interpretation of this piece of evidence depends, of course, on the date of the Deeds of Anitta (...)

29In order to obviate this difficulty it is necessary to discuss the genesis of the Sumerogram LUGAL.GAL. It was first introduced into Anatolia in the Assyrian colony period as an equivalent of the Akkadian title rubā’um rabūm, which is conventionally translated as ‘Great Prince’. The only sovereign attested with such a title before Anitta’s conquests was a ruler of Purushanda, even though the reading of the relevant passage is not altogether assured (TTC 27 7 ru-ba-im GA[L?]). After the conquests of Anitta, ruler of Kaneš/ Nesa, which culminated in a peaceful submission of the principality of Purushanda, the title ‘Great Prince’ came to be attached to the rulers of Nesa. Thus the genitive form rubā’im rabīm appears next to Anitta’s name in OIP 27 49 (A: 24-25; B: 26-27) while the Sumerographic title LUGAL.GAL accompanies the name ofAnitta’s successor Zuzzu (Kt 89/k 369 1). Finally the Deeds of Anitta in the Hittite language also refer to him once as LUGAL.GAL (KBo 3.22 obv. 41).46 But as far as one can judge, the use of this title never became fully consistent in the Colony period, as we also find the plain rubā’u- ‘Prince’ as the title of Zuzzu, the last known ruler of Nesa (Kt 89/k 370 35).

  • 47 Artzi/Malamat 1993: 30.
  • 48 Steiner 1999: 428-429.

30On the other hand, the metropolitan Assyrian and Syrian traditions apparently connected the Sumerogram LUGAL.GAL with the Akkadian title šarrum rabûm ‘Great King’. Although the first attestation of this title is found in a flattering letter sent to the Assyrian king Šamši-Adad I emanating from Mari (circa 1800 BCE), somewhat later it is attested at Alalakh with reference to the rulers of Yamhad.47 The destruction of the Kingdom of Yamhad and the sack of its capital Aleppo was the accomplishment of Mursili I, King of Hattusa. The later historical tradition of Hattusa, reflected in the preamble to the Talmi-Šarruma Treaty, preserved the recollection of the fact that the rulers of Yamhad were Great Kings, and may have even hinted at the connection between the expedition of Mursili II against Yamhad and the emerging Great Kingship of the rulers of Hattusa.48 It is therefore likely that the Sumerogram LUGAL.GAL in the bulk of Hittite texts does not represent a carry-over from the Old Assyrian colonial tradition but rather reflects the influence of Syrian scribal culture. But whichever of the two scenarios one chooses, it is clear that the spurious Sumerogram LUGAL.GAL represents a calque of an Akkadian royal title, which emerged within a Semitic scribal milieu. Its internal structure is ultimately irrelevant for the issue of reconstructing the underlying Hittite or Luwian forms.

  • 49 For the representative lists of hierarchical titles in GAL and UGULA see Peccholi Daddi 1982: 526-5 (...)

31Can one, then, offer any clues that could help us to approach the structure of the Hittite term for the ‘great king’? On a plausible assumption that this title was coined on the same model as the other Hittite designations of superior officials, one can attempt to use their structure as a typological parallel. A sober synopsis of templates for hierarchical titles can be found in CHD Š: 100. It turns out that the Sumerograms GAL and UGULA, traditionally translated as ‘chief’, and ‘overseer, superintendent’ respectively, were frequently placed in front of plural forms in Hittite texts. Thus one finds GAL LÚ.MEŠA.ZU ‘chief physician’ alongside UGULA LÚ.MEŠA.ZU ‘overseer of physicians’, GAL LÚ.MEŠAŠGAB ‘chief leatherworker’ alongside UGULA LÚ.MEŠAŠGAB ‘overseer of leatherworkers’, GAL LÚ.MEŠ GIŠBANŠUR ‘chief pantler’ alongside UGULA LÚ.MEŠ GIŠBANŠUR ‘overseer of pantlers’ etc.49 Differences in English translation need not obfuscate the fact that both titles in GAL and titles in UGULA adduced above match the structure of the Luw. tuppal-an-uri- ‘Chief Scribe’ and similar determinative compounds, except for the head-dependent word order, which is expected of the Sumerographic syntax. Although the Hittite readings of the titles headed by GAL and UGULA remain, strictly speaking, unknown, the hypothesis that they also represented determinative compounds or possessive noun phrases emerges as the simplest solution.

32As has already been noted at the beginning of this paper, the title LUGAL.GAL had the hierarchical meaning ‘overking’. As such, it was semantically different, for example, from the expression LÚ.MEŠ GAL ‘grandees, notables’, lit. “great people”, but resembled the Sumerographic titles discussed in the previous paragraph. Therefore its Hittite reconstruction as *hassuwan salli- lit. “(the) great(est) of kings” appears to be in no way worse, and perhaps even preferable, in comparison with *salli- hassu-, lit “great king”.

33Being far from claiming that we have enough data to advocate a particular Hittite reading for the title ‘great king’, I maintain that Hittite offers no arguments against the interpretation of MAGNUS.REX as Luw. *hantawatt(-an)-ura/(i)- “(the) great(est) of kings”. It is rather the new Luwian reading of MAGNUS.REX that should be used from now on as one of the considerations in determining the Hittite reading of LUGAL.GAL.

Bibliographie

Archi, А., “How the Anitta Text reached Hattusa”, in: Saeculum. Gedenkschrift für Heinrich Otten anlässlich seines 100. Geburtstags (StBoT 58), Müller-Karpe, A. et al. (éds.). Harrassowitz, Wiesbaden, 2015, 1-13.

Artzi, P. / Malamat, A., “The Great King: A Pre-eminent Royal Title in Cuneiform Sources and the Bible”, in: The Tablet and the Scroll. Near Eastern Studies in Honor of W.H. Hallo, Cohen, М. Е. (éd.). CDL Press, Bethesda, 1993, 28-38.

Bauer, A., Morphosyntax of the noun phrase in Hieroglyphic Luwian (BSIEL 12). Brill, Leyde, 2014.

Eichner, H., “Neues zum lykischen Text der Stele von Xanthos (TL 44)”. In: The IIIrd Symposium on Lycia: Symposium

Proceedings, M. Alparslan et al. (éds.). Institute of Mediterranean Civilizations, Antalya, 2006, 231-238.

Giusfredi, F., Sources for a Socio-economic History of the Neo-Hittite States (THeth 28). Winter, Heidelberg, 2010.

Gordin, Sh., Scribal Families of Hattuša in the 13th Century BCE: A Prosopographic Study, Tel-Aviv University MA Thesis, 2008.

Hawkins, J. D., Corpus of Hieroglyphic Luwian Inscriptions, Volume 1: Inscriptions of the Iron Age, Part 1: Texts. De Gruyter, Berlin, 2000.

Hawkins, J. D., “Eunuchs among the Hittites”, in: Sex and Gender in the Ancient Near East, Proceedings of the 47th Rencontre Assyriologique Internationale, Parpola, S. / Whiting, R. M. (éds.). The Neo-Assyrian Text Corpus Project, Helsinki, 2002, 217-233.

Hoffner, H. A., Jr., Letters from the Hittite Kingdom (WAW 15). Society of Biblical Literature, Atlanta, 2009.

Isebaert L. / Lebrun, R., “Le suffix -(a)za- in Lycien”, in: Calliope: Mélanges de linguistique indo-européenne offerts à

Francine Mawet, Vanséveren, S. (éd.). Peeters, Louvain, 2010, 159-172.

Klinger, J., Untersuchungen zur Rekonstruktion der hattischen Kultschicht (StBoT 37). Harrassowitz, Wiesbaden, 1996.

Laroche, E., “Noms de Dignitaires”, RHA 14, 1956, 26-32.

Laroche, E., Dictionnaire de la langue louvite. A. Maisonneuve, Paris, 1959.

Laroche, E., “Études de Linguistique Anatolienne”, RHA 23, 1965, 33-54.

Laroche, E., Les noms des hittites. Klincksieck, Paris, 1966.

Lehrman, A. A., “Anatolian cognates of the Indo-European word for ‘wolf’”, Die Sprache 33/1, 1987, 13-18.

Malbran-Labat, Fl. / Lackenbacher, S., “Un autre ‘Grand-Scribe’ ”, N.A.B.U. 2005/1, no. 10.

Melchert, H. C., A Dictionary of the Lycian Language. Beech Stave Press, Ann-Arbor, 2004.

Melchert, H. C., “Genitive Case and Possessive Adjective in Anatolian”, in: Per Roberto Gusmani. Studi in Ricordo, Volume 2: Linguistica Storica e teorica, Orioles, V. (éd.). Forum Edizioni, Udine, 2012, I, 273-286.

Melchert, H. C., “Naming Practices in Second- and First Millennium in Western Anatolia”, in: Personal Names in Ancient Anatolia, Parker, R. (éd.). Oxford University Press, Oxford, 2013, 31-49.

Neumann, G., Glossar des Lykischen (DBH 21). Harrassowitz, Wiesbaden, 2007.

Payne, A., Hieroglyphic Luwian: an Introduction with Original Texts. Harrassowitz, Wiesbaden, 2014.

Pecchioli Daddi, Fr., Mestieri, professioni e dignità nell’Anatolia ittita. Ateneo, Rome, 1982.

Otten, H., Die Apologie von Hattusilis III.: Das Bild der Überlieferung (StBoT 24). Harrassowitz, Wiesbaden, 1981.

Richter, Th., Bibliographisches Glossar des Hurritischen. Harrassowitz, Wiesbaden, 2012.

Rieken, E., “Zum hethitisch-luwischen Sprachkontact in historischer Zeit”, AoF 33: 271-285, 2006.

Rieken, E., “Bemerkungen zur Ursprung einiger Merkmale der anatolische Hieroglyphenschrift”, WdO 44, 2014, 162-173.

Schürr, D., “Karische und lykische Sibilanten”, IF 106, 2001, 94-121.

Shevoroshkin, V., “K ponimaniju milijskikh tekstov”, in: Slovo i jazyk. Sbornik statej k vos’midesiatiletiju akademika Ju.D. Apresiana, Boguslavskij, I. et al. (éds.). Languages of Slavonic Cultures, Moscou, 2011, 588-619.

Singer, I., “The Great Scribe Taki-Šarruma”, in: Hittite Studies in Honor of Harry A. Hoffner Jr. on the Occasion of His 65th Birthday, Beckman, G. et al. (éds.). Eisenbrauns, Winona Lake, 2003, 341-348.

Singer, I., “Ships Bound for Lukka: a New Interpretation of the Companion Letters RS 94.2530 and RS. 94.2523”, AoF 33, 2006, 242-261.

Starke, F., Untersuchungen zur Stammbildung des keilschrift-luwischen Nomens (StBoT 31). Harrassowitz, Wiesbaden, 1990.

Steiner, G., “Syrien als Vermittler zwischen Babylonien und Ḫatti (in der ersten Hälfte des 2. Jahrtausends v. Chr.)”, in: Languages and Cultures in Contact: At the Crossroads of Civilizations in the Syro-Mesopotamian Realm, Proceedings of the 42th RAI, van Lerberghe, K., Voet, G. (éds.). Peeters, Louvain, 1999, 425-441.

Vanséveren, S., Manuel de la langue hittite, vol. 1. Peeters, Louvain, 2006.

Weeden, M., Hittite Logograms and Hittite Scholarship (StBoT 54). Harrassowitz: Wiesbaden, 2011.

Wilhelm, G., “Name, Namengebung. D. Bei den Hurritern”, in: Reallexikon der Assyriologie, vol. 9, D.O. Edzard (éd.). De Gruyter, Berlin, 1998, 121-127.

Yakubovich, I., Sociolinguistics of the Luvian Language (BSIEL 2). Brill, Leyde, 2010.

Yakubovich, I., “The West Semitic God El in Anatolian Hieroglyphic Transmission”, in: Pax Hethitica. Studies on the

Hittites and their Neighbours in Honour of Itamar Singer (StBoT 51), Cohen, Y. et al. (éds.). Harrassowitz, Wiesbaden, 2010, 385-398.

Yakubovich, I., “Anatolian Names in -wiya and the Structure of Empire Luwian Onomastics”, in: Luwian Identities: Culture, Language and Religion between Anatolia and the Aegean (CHANE 64), Mouton, A. et al. (éds.). Brill, Leyde, 2013, 87-123.

Yakubovich, I., “The Degree of Comparison in Luwian”, IF 118, 2013, 155-168.

Notes

1 Artzi/Malamat 1993: 31.

2 Cf. Hoffner 2009: 323 with en. 322.

3 Otten 1981: 22, iii 69-70. Note, however, that Otten understood the Akkadogram HALS. I as ‘fortress’. For its correct understanding as ‘province’, see Klinger 1996: 200 w. ref.

4 Whether ura(/i)- ‘great’ should be analyzed as an a-stem or semi-vocalic stem is somewhat unclear because the “mutated” stem uri- is attested only once in a Hittite context (see HEG U: 87-88). This dilemma is ultimately irrelevant for the conclusions of the present paper and will not be addressed here in any further detail.

5 Giusfredi 2010: 81. 6 Payne 2014: 156.

6

7 Most of the relevant examples can already be found in Laroche 1966.

8 Melchert 2013: 41.

9 For the assumed meaning of walkuwa/i- ‘lion’ see Lehrman 1987, but the i-mutation reconstructed in this noun speaks against its being a Hittite cognate of Luw. walw(i)-‘lion’. One wonders whether the alternation /walgwV-/ ~ /walwV-/ might reflect Luwian dialectal variation.

10 This personal name (KBo 47.11 obv. 7), not yet found in Laroche 1966, is listed in the online supplement to Laroche’s work prepared by Marie- Claude Trémouille and known as Répertoire onomastique (http://www.hethport.uni-wuerzburg.de/hetonom/ONOMASTIdata.html).

11 For the productive pattern of Luwian “topophoric” names with theophoric interpretation, see Yakubovich 2013a: 103-107.

12 Here and below, the precise citations of all the Iron Age Luwian forms in hieroglyphic transmission can be found in the Annotated Corpus of Luwian Texts (ACLT), sponsored by a research grant of the Presidium of the Russian Academy of Sciences, at the address http://web-corpora.net/ LuwianCorpus/search/.

13 Cf. Yakubovich 2010b: 396, fn. 9., where the name under discussion is erroneously analyzed as Luwian in origin. For the convincing Hurrian etymology of Ura-hilana see Wilhelm 1998: 124. It is, however, unlikely that the Hurrian language was still spoken in Hama in the first millennium BC.

14 These onomastic compounds can be typologically compared with Italian last names of the type Casa-grande or Casa-nova. I am grateful to Stefano de Martino for this parallel. Compare also the last names the French Anatolianist Olivier Casabonne and Adrien Maisonneuve, the publisher of Laroche 1959.

15 Melchert (2013: 41) interprets the same name differently, as “great (one) of the gods”. I analyze this personal name as reflecting a wish that the gods perceive its carrier as a great person. This interpretation is in line with the likely meanings of other personal names of the same structure, e.g. Late Luwian TONITRUS.HALPA-pa-wasu “dear to (the Storm-god) of Aleppo” or Carian πον-υσωλλος “dear to all”.

16 Cf. HED H: 238. On the prosopography of mḪa-aš-ta-nu-ri see Singer 2003: 343-344.

17 Laroche 1956: 28. Note also RS 34.126 tu-pal-nu-ri and RS 92.2007 tup-pa-la!-nu-ri (Gordin 2008: 158).

18 Laroche 1965: 37.

19 Malbran-Labat/Lackenbacher 2005: 9. The phonetically sensitive rendering of this Hurrian name would be Fendi-Šarruma, literally “Šarruma (is) just” (cf. Richter 2012: 293b with ref.). For the i-vocalism of tu-up-pi-nu-ra hu-pu-ur-ti-nu-ra cf. fn. 14 above.

20 Singer 2006: 244. It is worth noting that this equation represents an argument for the traditional interpretation of <SCRIBA> as ‘scribe’ and against the reinterpretation of this sign as the generic term ‘official’, which was advanced by Theo van den Hout at the Ninth International Congress of Hittitology in Çorum, Turkey (September 2014). This new argument is, however, less strong than the existence of Luw. SCRIBA-lalli(ya)- ‘writing, script’, which Theo van den Hout himself acknowledged as a problem for his hypothesis.

21 On the etymology of Luw. hastall(i)- ‘hero’ see Starke 1990: 122, 124.

22 Singer 2006: 244.

23 Laroche 1959: 102.

24 Yakubovich 2010a, Chapter 5.

25 Yakubovich 2010a: 45-46.

26 Yakubovich 2010a: 47-53.

27 Cf. Melchert 2012: 275.

28 As a parallel for a genitive case marker developing into an interfix, one can consider the situation in German. While the German genitive ending -s is normally restricted to masculine and neuter nouns, its cognate in determinative compounds can also link feminine nouns to their syntactic heads, e.g. Sicherheit-s-dienst ‘security service’, Forschung-s-gemeinschaft ‘research team’.

29 See Yakubovich 2013b for Luwian and GrHL: 273-276 for Hittite. In my opinion, the Luwian function of the suffix -(a)zza- represents an innovation vis-à-vis the situation of Hittite, where the cognate suffix -zziya- < *-ti̯o- was restricted to polar adjectives. A different analysis is offered in Isebaert/Lebrun 2010, where Luw. -(a)zza- is synchonically analyzed as a substantival suffix, although the authors ultimately derive it from the same Indo-European suffix *-ti̯o- forming polar adjectives.

30 On the left-hand branching word order in Luwian, see Bauer 2014: 36-37 and passim.

31 Yakubovich 2013b: 161, cf. Hawkins 2000, II: 528.

32 Cf. Hawkins 2000, I: 265 and Hawkins 2002: 230.

33 Cf. the previous analysis of amurallura/(i)- in Giusfredi 2010: 162: “No etymology can be provided for the word. The stem ending in -(al=)ura/i- is quite surprising, and, although it appears also in the mysterious title a˹satar?˺ura/i-, it provides no clue to an interpretation”. Ibid. sub astaruri(ya)-: “No interpretation of this word can be attempted, since the context is, in this case, no help at all. For the -(al=)ura/i- ending see amuralura/i-”.

34 Cf. Hawkins 2000, II: 537.

35 Cf. Hawkins 2000, II: 508-509.

36 Cf. Hawkins 2000, II: 443.

37 Compare the title uryall(i)- of a certain Nunuya mentioned as a recipient of sheep in KULULU lead strip 2 #4 (Hawkins 2000: 510).

38 For the most recent edition of the AKSARAY inscription see Hawkins 2000, II: 475-478, where the earlier editions of the same text are also cited.

39 For the role of aesthetic considerations in determining the order of Anatolian hieroglyphs, which reflects their erstwhile use for rendering names and titles on official seals, see lately Rieken 2014.

40 The same name also occurs in Greek transmission as Κενδαβυρα / Κινδαβυρις (Melchert 2004: 109).

41 This is not the only cases when personal names of Lycian B origin are embedded in Lycian A texts. Compare, for example, the names Masasa and Masauwẽti (Melchert 2004: 98), which both apparently contain the Lycian B element masa- ‘god’, a cognate of Lyc. (A) maha(na)- ‘id’.

42 The first interpretation is advocated by Schürr (2001: 105) and endorsed by Shevoroshkin (2011: 594, 598, 605), while the second one can be found in Melchert 2004: 136. Given our present level of the knowledge of the Lycian B language, I hesitate about making a choice between these two options. Cf. Neumann 2007: 126-127 and, for a different segmentation of the compound name under discussion, Eichner 2006: 234, fn. 25.

43 Neumann 2007: 240.

44 Thus e.g. Steiner 1999: 428, Vanséveren 2006: 125. Weeden (2011: 571) reconstructs the Hittite noun phrase *salli- hassu- on the basis of KBo 16.45 rev. 5 LUGAL.GAL-uš (OS/MS?). It is not, however, to be ruled out that the heterogram LUGAL.GAL had the plain reading hassu- in this hapax, arguably pertaining to the period when the Akkadogram LUGAL.GAL had not yet been calqued into Hittite.

45 On which see Rieken 2006.

46 The interpretation of this piece of evidence depends, of course, on the date of the Deeds of Anitta. If one assumes that it was first put in writing in Hattusa at the time of Hattusili I or later, then the use of the title LUGAL.GAL in this text may reflect a different tradition coming from Syria (see immediately below). The most recent paper defending the early date of the Anitta text is Archi 2015, which can also be consulted for the history of the debate.

47 Artzi/Malamat 1993: 30.

48 Steiner 1999: 428-429.

49 For the representative lists of hierarchical titles in GAL and UGULA see Peccholi Daddi 1982: 526-528.

Auteur

Ilya Yakubovich

University of Moscow / University of Marburg