Документ взят из кэша поисковой машины. Адрес оригинального документа : http://imk.msu.ru/Structure/Linguistics/yakubovich/download/palaic1.pdf
Дата изменения: Fri Apr 30 15:11:40 2010
Дата индексирования: Mon Oct 1 20:04:54 2012
Кодировка:

WERE HITTITE KINGS DIVINELY ANOINTED? A PALAIC INVOCATION TO THE SUN-GOD AND ITS SIGNIFICANCE FOR HITTITE RELIGION ILYA YAKUBOVICH Abstract
The aim of this paper is to improve our understanding of a difficult Palaic invocation to the Sun-god, and to elucidate its implications for the study of Hittite religion. The first part of my account contains linguistic and philological discussion that concludes with a new translation of the scrutinized fragment. According to my interpretation, the Sun-god is requested to anoint the Hittite king and to exalt him. This is the only clear evidence that the gods were thought to be personally responsible for the anointment of Hittite kings. A counterpart to this nontrivial concept is well known from the Hebrew Bible and is inherited by the Christian tradition. The second part of this paper, which is meant to be accessible to all the historians of religion, discusses the anointment as a rite of passage among the Hittites, as well as the relevant parallels in other Ancient Near Eastern cultures. I am arguing that anointment with oil was extended to both Hittite priestly kings and certain other categories of Hittite priests, and that the underlying purpose of this act was ritual cleansing. The spread of this rite to those cultures where kings were not at the head of the religious hierarchy boosted the secondary association of divine anointment with empowerment rather that purification.

I. Philological Analysis The Palaic ritual CTH 751 contains a series of invocations to local deities.1 The gods and goddesses are invited to come and taste the
1 Subject to the usual disclaimers. I would like to thank Theo van den Hout, who contributed to the composition of this article in more than one way, Dennis Campbell, who advised me on the matters Hurrian, Craig Melchert, who made detailed comments on the advanced version of the manuscript and was receptive to new ideas that sometimes contradicted his older views, and Petra Goedegebuure, who contributed to my understanding of Palaic syntax. I am also obliged to Onofrio Carruba, John Nielsen and Thomas Wier, who took pains to help me with the style of this paper, and to the lively audience of the Sixth International Congress of Hittitology ( Rome, September 2005), where a version of it was first presented.

© Koninklijke Brill NV, Leiden, 2006 Also available online www.brill.nl

JANER 5

108

ilya yakubovich

food prepared for them by the king and the queen. One of these invocations is dedicated to the Sun-god Tiyaz, the third in rank in the Palaic pantheon. Its beginning is preserved in two mutually complementing parallel versions KUB 35.165 Obv. 21-24 ( Middle Script) and KUB 32.17 7'-10' ( Late Script), and begins as follows: 21. = I 22. = II 23. = III 24. = IV [(nu-ku)] pa-a"-hu-ul-la-"a-a" ti-[ia-]az ta-ba-ar-ni LUGAL-i pa-a-pa-az-ku-ar ti-i [(a-an-na-)]az-ku-ar ti-i i"-ka[ n]u-u"-"i-ia-am-pМ ti-i a-ri nu-u"-"i-ia-am-pМ ti-i [(a-ru-u-)]na-am-pМ ti-i З-i-te-"i2 [x?]-a-an-ta-na-an ti-i З-i-te-"i [(ki-i-a)]t [(wuu-З-l)]a-"i-ni-ki-e" wa-"u-ki-ni-e" "i-miya-a-a" ki-i-tar KI.MIN

At this point, the invocation is cut off in the first version, while the second one, which apparently did not contain the final KI.MIN `ditto', carries on with the stereotypical description of the offerings ( KUB 32.17 11'-13'), which can be reconstructed based on the parallel passage occurring earlier in the text ( KUB 35.165 Obv. 13-15). wa-a-"u З-i"-ta-a" "a-a[(m-lu-wa-a" wuu-la-"i-na-a" kii-ta-ar)] 12'. = VI [(w)]a-a-"u З-i"-ta-a" ma-l[(i-ta-an-na-a" wuu-la-"i-naa" ki-i-ta-ar)] 13'. = VII [(ku)]-i-"a t]u-З wuu-la-"i-n[(a-a" ka-a-ar-ti a-ti-a-paan az-zi-ki-i)] The final two lines of the invocation are preserved only fragmentarily, and will not be discussed in the present paper, except for the clarification of what precedes them. Carruba provided a tentative translation of the quoted passage:3
`Und nun, Tijaz Pashullasas (= Sonnengott der GЖtter?), dem Tabarna, dem KЖnig, Vater (sage ich / sagt er) du, Mutter (sage ich / sagt er) du; du sollst . . . en, und bei ihm (stehen), du sollst (herbeikommen (?) und bei ihm (stehen). Den arun(an) (das Meer?) baust/bringst du; den antan(an) baust/bringst du. Die brotlichen Kostbarkeiten liegen auf den simija-'. Gut liegt wista-GebДck auf den p.-Broten aus /mit samlu-

11'. = V

2 3

KUB 32.17: З-i-ti-x[ Carruba 1972: 29-30.

were hittite kings divinely anointed?
Gut liegt wista-GebДck auf den p.-Broten aus /mit HЖnig. Welches Brot dir im Herzen (ist), das isst du (wiederholt)'.

109

Starke suggested an alternative translation of the last four lines:
GebДckartiges, wertvolles "imiiaGutes an Иui"ta(-GebДck) liegt da, Gutes an Иui"ta(-GebДck) liegt da, Welches GebДck dir am Herzen lich".

4

liegt da GebДck mit Apfel(-FЭllung) GebДck mit hЖnighaltiger ( FЭllung) ist (= deinen Gefallen findet), das iъ reich-

One can see that the general sense of the last four lines is agreed upon, while points of disagreement mostly concern grammatical matters. The description of various breads prepared for the ritual meal concludes with inviting the Sun-god to eat the bread he likes the most. In this article I am not going to discuss this stereotypical formula in any further detail, concentrating instead on lines I-III that are specific to this invocation. Carruba's translation of its first sentence can be immediately refined on two points. First, C. Melchert has convincingly argued that Pal. =ku(w)ar should not be analyzed as a sequence of the coordinative enclitic, =ku= and the quotative enclitic =war, but rather represents an indivisible emphatic particle, functionally similar to Hitt. =pat.5 Thus, Carruba's parenthetical (sage ich /sagt er) is grammatically unnecessary. Second, the epithet pa-a"-hu-ul-la-"aa" is unlikely to be related to Hatt. waa-a"-ha-ap `gods', where the final p is a part of the root, but may be connected with the Hattic waa-a"-hu-З-l[i (1x) and the theonym dwaa-a"-hu-З-li-li.6 Unfortunately, we cannot translate the Hattic hapax *fashЫli, and so the epithet *Fashullasas must also remain without translation for the time being. In fact, both these points of refinement are already reflected, albeit without commentary, in Archi's Italian translation "E, divinitЮ Solare ... al Tabarna, al re, (sei) padre tu, (sei) madre tu".7 Being not content with these improvements, Melchert undertook a more far-reaching syntactic re-analysis of the same sentence. Starting with a natural assumption that every occurrence of the enclitic nu= in Palaic must be sentence-initial, he concluded that iska in line II ought to be a sentence-final verbal form. Since tЛ is

4 5 6 7

Starke 1990: 73. Melchert 1984: 23-24. Soysal 2004: 897 with ref. Archi 1979: 37.

110

ilya yakubovich

the subject of the preceding sentence, the nominatives pЮpaz and Юnnaz can only be predicative, which means that, if iska indeed belongs to the same sentence, it is a form of `be', or a similar linking verb. Emboldened by this conclusion, Melchert took iska < *eske as the 2 sg. imv. of the iterative of Anat. *es- `be'.8 Below I will try to show that this interpretation runs into morphological and textual difficulties. The most serious problem concerns the ablaut. The iterative stems were normally formed from the zero grade of ablauting roots in Hittite (e.g. appisk- from Иp-/ap- `take' or akkusk- from Иku-/aku`drink'). Palaic 2 sg. imv. of the iterative az-zi-ki-i `eat' in line VII must also contain the historical zero grade of Anat. *Иd-/ad- `eat'. Under these circumstances, it is completely unclear why the Palaic iterative of the root Иs-/as- `be', which belongs to the same morphological class as `take', `drink' and `eat' in Anatolian, would contain the full grade of the root.9 But even if we accept this interpretation, it remains to be explained why the form /Иska/ was rendered as i"-ka and not **(e)-e"-ka in a Middle Script text.10 An attempt to circumvent these difficulties came from C. Watkins, who reinterpreted iska as a middle imperative going back to the IE *h1s-sk'С.11 He adduced the Tocharian B 3 pl. copula skente, suffixed skentar-ne, as an evidence for the middle conjugation of this stem in Proto-Indo-European. The speculative character of this suggestion is clear from the fact that the direct reflexes of middle imperatives in *-С are not attested anywhere else in Anatolian. More cogently, the hypothetical reconstruction *h1s-sk'С would imply the final stress in the Palaic form, in which case it remains completely unclear why it is not written **i"-ka-a, with an expected vowel lengthening in a stressed open syllable.

Melchert 1984: 30. The imperative forms of *es- `be' attested in Palaic are 2 sg. a-a" ~ Hitt. e-e"; 3 sg. a-a"-du ~ Hitt. e-e"-du; 3 pl. a-"a-an-du / a-"e-en-du (?) ~ Hitt. a-"a-an-du. The change *Иs- > *Аs- attested in Palaic 2/3 sg. imperative forms may be caused either by a sound law (thus Wallace 1983: 169), or by an analogy with 3 pl. As for the morphological zero grade of the same root, attested in Hitt./Pal. 3 pl. imv. /as-Аndu/, it does not show traces of the e-vocalism in any Anatolian language. 10 The original form of the Anatolian imperfective suffix was *-sk'e-. The change *e>a in the final syllable of iska could occur only in unstressed position (see Melchert 1994: 202, 216). 11 Watkins 1993: 477-78.
9

8

were hittite kings divinely anointed?

111

In addition, if one accepts Melchert's readjustment of clause boundaries, accepted by Watkins, one will wind up with an unusual formula inviting the Sun-god to be the father and the mother of the king. While there is nothing a priori impossible about such a request, other texts indicate that the Hittites tended to take it for granted that Sun-deities were "fathers and mothers" of the human race. Thus, one encounters the following statements in the Hittite prayer to the Sun-goddess of Arinna, and in the Hittite adaptation of the Mesopotamian hymn to Shamash respectively:
(1) KUB 24.3, CTH 376, NS I.46': nu-za KUR-e-a" hu-u-ma-an-da-a" at-ta-a" an-na-a" zi-ik `You are father and mother to all the lands'. (2) KUB 31.127, CTH 372, NS I.21: da-an-ku-wa-ia-a" KUR-e[-a" ]at-ta-a" an-na-a" zi[-ik] `You are father and mother of the dark land'.

The second text further specifies istanus dammishandas kurimmas=a antuhsas attas annas zik "Son-god, you are the father and mother of the oppressed and orphaned person".12 Finally, a fragmentary prayer in which a king addresses the Sun-deity and the Storm-god, contains the following passage:
(3) KUB 43.68 + 36.91, CTH 389, NS Obv. 8'-9': [nu-mu A-BU-YA] NU.Gа[ L] nu-mu AMA-IA NU.Gа[ L] nu-mu "u-me" DINGIR.ME"-i" A-B[U AMA-YA] `[ I do not have] a father. I do not have a mother. You, gods are my fa[ther and mother]'.13

There are no reasons to think that the author of the prayer above is trying to literally trace his genealogy to gods. Calling someone "father and mother" was a usual way to express respect in Hittite, attested beginning with the Anitta text. Palaic construction, under the discussed interpretation, would be quite special in treating close ties between Sun-deity and humans not as a matter of fact, but rather as a desirable state of events. On the contrary, if we follow Archi's analysis, it would fit in well with what we know about Hittite prayers. Now, one must see which of the two ways to set up clause boundaries yields a more plausible syntactic interpretation for the rest of
12 13

Quoted after GЭterbock 1958: 230 (cf. fn. 19). Reconstructed in van den Hout 1998: 74.

112

ilya yakubovich

line II. Melchert agrees with Carruba that the beginning of line III marks the beginning of the new clause,14 and I accept the same analysis in this article. Therefore, one has to choose which of the two passages is more likely to contain well-formed clauses:
Carruba Melchert i"-ka[ n]u-u"-"i-ia-am-pМ ti-i a-ri nu-u"-"i-ia-am-pМ ti-i [ n]u-u"-"i-ia-am-pМ ti-i a-ri nu-u"-"i-ia-am-pМ ti-i

Carruba's translation of the second part of line II implies that we are dealing with a sequence of two compound sentences, i.e. i"-ka | nu-u"-"i-ia-am-pМ ti-i || a-ri | nu-u"-"i-ia-am-pМ ti-i. The second clause of each sentence is elliptical, `and thou (must stand) by him'. I believe, however, that such a clause fragment, consisting of a mere subject tЛ `thou' and an enclitic chain, is impossible in a cohesive text. The necessary condition licensing an ellipsis is the recoverability of an omitted constituent, but Carruba does not explain to us how the missing predicate `must stand' can be inferred from its linguistic context. Yet, Carruba's segmentation can be salvaged if we assume that the discussed passage contains only two imperative clauses i"-ka nuu"-"i-ia-am-pМ ti-i | a-ri nu-u"-"i-ia-am-pМ ti-i. Such a solution has been already advocated by Kammenhuber, who has not, however, clarified the sequence nu-u"-"i-ia-am-pМ.15 The difficulty that a Hittitologist faces is the obvious one: if this is indeed a clitic chain beginning with nu=, why does it not occur clause-initially?16 One must, however, always keep in mind that Palaic and Hittite are two distinct languages, and Palaic clitics need not obey Hittite distributional rules. While Palaic nu-З / nu does frequently occurs in clause-initial position, its second occurrence in the invocation [nu-(З)] d[(zapА)]r-waa nu-З dza-pАr-waa-a `Oh, Zaparva, oh Zaparwa!' ( KUB 35.165 Obv. 1) shows a different pattern. The same example also illustrates the fact that Palaic nu, unlike its Hittite counterpart, was occasionally written plene, which indicates that it could be stressed.

Melchert 1984: 39. Kammenhuber 1959: 21. 16 An unusual example of Hittite n=asta occurring after the topicalized(?) subject is attested in the Hittite and Hurrian bilingual ( KBo 32.14 ii 26). This case, however, can be alternatively explained by the translator's belated decision to insert the enclitic complex into the sentence (cf. the discussion in Neu 1996: 128).
15

14

were hittite kings divinely anointed?

113

These aberrations make one suspect that nu did have a chance to fully develop into a clause-demarcational particle in Palaic, preserving at least in some cases its syntactic mobility. It is possible that, even when this particle was encliticized, it could occur both in the first and in the second position, after the fronted constituent.17 The fronting of the narrow predicate focus in imperative constructions represents a cross-linguistically frequent phenomenon.18 This syntactic analysis is, however, impossible if we accept Melchert's segmentation of line 2. Melchert did not attempt to provide a translation of this passage, but it is clear that, under his analysis, at least one of the two sequences nu-u"-"i-ia-am-pМ ti-i must be treated as a separate clause with an elided predicate. Unless one provides a satisfactory account for what this predicate might be and how one can retrieve it, Melchert's hypothesis cannot be maintained.19 Now it is time to proceed to the lexical analysis of the two imperative clauses. Apparently we are dealing with a set of requests addressed to the ( personified) Sun. The active imperative form i"ka /Мska/ cannot be connected with Anat. *Иs-/as- `be', for the phonological reasons mentioned above, but it represents a straightforward match with Hitt. iska/i- ( Late Hittite iskiya-) `to anoint' that

17 A typological parallel for such word order fluctuation is offered by the behavior of the Lycian particle (=)me=. When the sentence word order is not influenced by its information structure, it appears in the first position. If the first position is occupied by a topicalized constituent, me= appears in the second position as a stressed word. But if the first position is aligned with the contrastive focus, then, more frequently than not, =me= behaves as a second position clitic. Garrett 1992 provides an exhaustive description of the pertinent rules, although he inappropriately refers to contrastive focus as "verb topicalization". Unfortunately, it is difficult to say with certainty whether (=)nu= as a Wackernagel clitic, or as a stressed word in our Palaic sentence. The word space in front of nu observable in the autography of KUB 35.165 is of little probative value here because the Hittite scribes who copied this Palaic ritual would hardly have resisted the temptation of separating this element, which was familiar to them as a first position particle, from the preceding word. 18 Compare Belfast English commands such as Read you quickly that book or Throw us you that rope, or a Russian mild curse Idi ty k chertu, lit. "Go you to the Devil". Both Belfast English and Russian normally exhibit SVO word order. 19 Professor Melchert indicates in his letter from August 18, 2005 that the only conceivable way to interpret the sentence nu-u"-"i-ia-am-pМ ti-i a-ri | nu-u"-"i-ia-ampМ ti-i would be to take it as a figure of repetition "and also raise him up, and also [raise] him up". He agrees, however, that gapping in such a sentence would be no less strange in Hittite than it is in English.

114

ilya yakubovich

is discussed in detail in the second part of this paper.20 The hypothesis that the Sun-god is asked to anoint the king in our Palaic passage can be supported by the interpretation of the second predicate as `exalt (him)'! The stem ari- can be compared with Luw. ari(=HLuw. PUGNUS-ri-) `rise, raise' and, more remotely, with Hitt. arЮi-/ariya- `id'. When used with the king as an object, this root can be translated `to raise, exalt' in Luwian, as the following examples illustrate:
(4) KARKAM(" A11a, § 4 wa/i-mu-ґ DEUS-nМ-zi mi-ia-ti-ґ <">IUSTITIA"-wa/i-nМ-ti PUGNUSmi-la/i/u |PUGNUS-ri+i-ta `The gods exalted me in strength because of my justice'. KARKAM(" A6, § 20 wa/i-ta ( DEUS)TONITRUS-hu-ti-i ( DEUS)SOL-ti-i ( DEUS)ku+AVISpa-pa-ia-ha ta-ni-mi-i-ha-a-wa/i || DEUS-ni-i |FRONS-ti-i |SUPER+ra/i-a-ta5 PUGNUS-ri+i-wa/i `I will raise (him) up in front of the Storm-god, the Sun-god, Kubaba, and every god'.21

(5)

Another occurrence of a similar imperative arie in Palaic is also compatible with the suggested etymology. The sentence given below appears in a context that describes the ritual offerings prepared for the god Zaparwa:
(6) KUB 35.165, CTH 751, MS Obv. 7'-8': ma-a-an-ti mar-ha-a-an-za ma-a-a-an-ti [x]-ia-an-za Б-pa a-ri-e `When you have come as a guest?, when you are . . .-ed, lift them up!'.22

In an alternative analysis, Carruba took both ari and arie as imperatives from В ar `reach(?)', possibly attested as 3 Pl. Pret. a-ra-an-

20 Note that i"-ka must be reconstructed with the stress on the root, whereas the imperative az-zi-ki-i `eat!' occurring in the same passage has a stress on the thematic vowel. This discrepancy correlates with the fact that the latter form contains the synchronic imperfective suffix -ski-, while the former one resists synchronic segmentation. 21 Differently Hawkins (2000: 125): `I shall raise (?) them (?) up . . .'. Hawkins seems to have overlooked the possibility of analyzing the graphic wa/i-ta as (a)=wa=an=ta, where =ta is a sentential particle, and therefore had to take =ta as 3 pl.obl. enclitic pronoun `them'. The identity of "them", however, remains obscure since the context of this sentence addresses the favors that the author of the inscription, the tyrant Yariri, bestowed upon Kamani, the legitimate heir to the to the throne of Karkamish. 22 Carruba (1972: 29) leaves marha- without translation.

were hittite kings divinely anointed?

115

ta-(mu-ut) in a broken context ( KUB 35.167, 22').23 He did not clarify, however, the origin of the -i/-ie suffix. So far as one can judge, active imperatives consist of a bare stem in Hittite and Luwian, and Palaic Юs `be!' does not show any deviation from the standard Anatolian pattern. Melchert (1984: 36), following Carruba, analyzes the form mЫsi as an imperative from В mЫs `satiate/enjoy oneself ', but the same form might be as well taken as 2 sg. pres. with a future function, i.e. /mЫs+si/ `you will satiate yourself '. The same type of explanation is not available for ari / arie, and so setting the stem ar(i)ye- `raise' apart from the stem ar- (with an unknown meaning) appears to be the only acceptable solution. Presumably, the final -e in a-ri-e is phonetically real, while a-ri reflects a secondary contraction *-iye > -Л. The suggested interpretation of line II is supported by Melchert's analysis of line III.24 Objecting to Carruba's interpretation of arЫnaas `sea', Melchert justly remarks that the mention of the sea makes no sense in the context, which concerns the Sun-god's close relationship with the king. Therefore, Melchert analyzes the first clause of line 3 as arЫn=am=pi tЛ widesi `You shall also build him (the king) high'. The adjective aru- `high' is a Caland formation related to ar(i)ye- `raise' discussed above. For the second part of line III, Melchert tentatively suggests the restoration [A.]A-antan=an tЛ widesi `you shall build him strong'.25 Melchert recognizes that the restoration of a heterogram in a Palaic text is a risky matter, but the presence of a positive adjective in this clause is borne out by its syntactic parallelism with the previous one.26 Both clauses of line III, in their turn, seem to be related to the two commands expressed in line II (note that the contrastive focus construction is employed in each of the four cases). This symmetrical relationship can be schematically represented by the following tableau:
iska(=)nu=ssi=am=pi tЛ arЫn=am=pi tЛ widesi ari(=)nu=ssi=am=pi tЛ [A?.]A-antan=an tЛ widesi

23 Carruba 1970: 51. For the analysis of the sequence -mu-ut, see the discussion of example (7) below. 24 Melchert 1984: 39-40. 25 Ibid. fn. 28. 26 Given the general sense of the passage, one can hardly think that Pal. *a-an-ta-an represents a cognate of Hitt. ( Lз) Юnt- `equal', whose existence was persuasively argued for in Goedegebuure 2002.

116

ilya yakubovich

At this point, one must observe that the translation of wide- as `build', based on the comparison with Hitt. wede- `id.', and entertained by all scholars of Palaic at least since Kammenhuber, is not the only option possible.27 "Building someone high" is certainly a plausible metaphor for exalting someone, but "seeing someone high/exalted" would be a straightforward result of the same action. Likewise, seeing someone powerful (vel sim.) can be viewed as a result of anointing someone to the kingship.28 Therefore, a consideration must be given to the comparison between Pal. wide- and Lat. vidХre `to see'. Anatolianists have not considered this possibility so far, perhaps because the cognates of IE В weid `see' are not otherwise attested in Anatolian. It is important, however, to remember that Hittite and Luwian express the notion of seeing in two different ways, i.e. by the stems (a)u(s)- and m(a)nЮ- respectively. If we suppose that these are two independent innovations, then there would be nothing strange in assuming that Palaic might have kept the original Indo-Hittite verb of seeing, which was lost elsewhere in Anatolian. As for the stative suffix -Х-, present in Lat. vid-Х-re, Watkins has demonstrated the existence of the same suffix in Hittite, even though his examples are limited to (etymological) denominative statives.29 The last point that deserves a special discussion is the composition of the particle chains nu-u"-"i-ia-am-pМ in line II. Kammenhuber was reluctant to choose between the division =ampi vs. =am=pi, but tentatively favored the second solution.30 Carruba shared the same view, taking the first component of =am=pi as a sandhi form of the clitic =an.31 The existence of the particle =pi in Palaic is

27 I deem phonetically impossible the comparison between Pal. wide- and Hitt. wida-hi (later wida(i)-mi) `bring' or uwade-mi `conduct', offered as a remote possibility by Carruba (1972: 20). 28 "Bezeichnet aber die Salbung des (hethitischen--I.Y.) KЖnigs bei seiner Inthronisation die Mitteilung der neuen WЭrde, dann ist sie--wie auch die israelitische KЖnigssalbung --als Zuwendung von pondus, auctoritas, Macht verstanden; ihrem VerstДndniss liegt also die krДftigende Wirkung des жles Zugrunde" ( Kutsch 1963: 37). See more on this topic in the second part of the present paper. 29 See Watkins 1969. In many instances, the stative suffix *-Х- is secondarily extended by "inceptive" -s- in Hittite, eventually yielding forms like tebawes(s)- `to become little', alpues(s)- `to become blunt' etc. It is quite possible, in my opinion, that Hitt. kanes(s)- `to recognize' contains the same ingressive suffix *-Х-s- added to В g'nР- `to know'. 30 Kammenhuber 1959: 22. 31 Carruba 1970: 67.

were hittite kings divinely anointed?

117

supported by the alternation ha-"i-i-ra-am-pМ and GмR-an-pЛ in the two copies of the same text KBo 19.152 and KBo 19.153.32 This clitic may be compared with the problematic Cuneiform Luwian element =pi that also occupies final position in the chain of enclitics.33 The translation of =pi as `also' is a mere guess, but the fact that three of the four parallel clauses in lines II-III contain this particle is compatible with the supposition that it could have a coordinative meaning.34 Once we separate the particle =pi, the element =am= must needs be taken as a sandhi variant of =an=, an anaphoric pronoun `him' referring in this case to the king. This raises a question about the position and the function of the clitic =ssi= in our passage. On the one hand, this pronominal clitic occupies a position before the accusative clitic pronoun =am=, which is the reverse of what can be observed in Hittite. On the other hand, it appears to lack a plausible antecedent. It is easy to answer a question about the syntactic position of =ssi=. The following apotropaic formula provides the independent evidence that dative pronominal clitics preceded their accusative counterparts in Palaic:
(7) KBo 30.39 + ta-ha-a-mu-ut um-ma-ri-kat `It is taha-ed, extinguished! KUB 35.164, CTH 752, MS ki-i"-ta-a-mu-ut tar-ta-kat ki-i"-ta-a-am-mu [h]a-al-za-luki-i"-ta-a-am-mu waa-a-ar-ra-kat ki-i"-ta-a-mu. it is extinguished! This curse is extinguished! This h. is This f. is extinguished'!

Carruba tentatively assumed that the final -ut in the participial clauses taha(n)=mu=t and kista(n)=mu=t is a variant of the sentential particle =ta,35 but this solution is phonetically impossible since no other instances of apocopated *-a are attested in Palaic. On the contrary, if one takes -mu-ut as representing the enclitic chain =mu=(a)t, the contraction *-uwa- > *-Ы- can be compared with a

32 Ibid. 22-23. For reasons I do not fully comprehend, Carruba transliterates pЛ as pat in one of the duplicates. The enclitic =pat is not attested anywhere else in the Palaic corpus. 33 See on this form Melchert 1993: 175. 34 Cf. also KBo 19.153 III 6'-8' ID-an-a"-ta ni-i te-eka-an-za GмR-an-pЛ ni-i [p]ata-am-ma-an, which Carruba (1972: 51) tentatively translates `das Wasser nicht getrocknet (?), der Dolch nicht ausgestreckt (?)'. In this context, the particle =pi appears in the second of the two coordinated phrases. 35 Carruba 1970: 64.

118

ilya yakubovich

similar contraction *-iye- > -Л- that left abundant traces in the Palaic verbal system.36 Futhermore, the presence of the enclitic =at `it' in unaccusative clauses with ellipted nominal subjects may be viewed as an instance of the Watkins-Garrett rule, whereas the function of =ta in our passage would be completely unclear.37 The same sequence =mu=at following a finite form can be seen in a-ra-an-ta-mu-ut ( KUB 35.167, 22'), except that in this case =(a)t must represent an accusative clitic. The word order where the nominative / accusative pronominal clitics follow the dative pronominal clitics is well established for Luwian, and the presence of the same pattern in Palaic constitutes a significant grammatical isogloss between the two languages.38 The function of =ssi= in our passage cannot be discussed separately from the other occurrences of the clitic =si= in Palaic and related languages. Carruba seems to indicate that it can function both as a reflexive and as a non-reflexive anaphoric pronoun in Palaic.39 Yet the scrutiny of the relevant examples does not yield a single instance where =si= has to be matched with a non-local antecedent. Two examples, whose interpretation is reasonably clear, suggest that =si functioned as a reflexive pronoun with В mЫs, irrespectively from its person or number:
(8) KUB 32.18, CTH 752, MS I. 6'-7' (cf. I 17'-18') 8: a-an-ti-en-ta ma-a-ar-ha-a" a-ta-a-an-ti ni-ippa-"i mu-"a-a-an-ti `The guests? enter. ( They) eat, (but) do not satiate themselves'.40

36 The particle =mu= probably functions here as an "ethical dative", i.e. kista(n)=mu=(a)t can be literally translated as `It is burned (in my interest)'! Note that in the second through fourth clauses of our passage it does not appear in Wackernagel position, but is uniformly attached to the predicate. This affix-like behavior must be compared with that of the "ethical dative" marker -si in Hieroglyphic Luwian (cf. examples (12)-(13) below). C. Melchert ( pers. comm.), suggests an alternative clause-division and translation "It has been taha-ed for me, it has been extinguished for me, this curse. This h. has been extinguished for me. This w. has been extinguished for me, extinguished for me." This interpretation allows him to analyze =mu as a regular second-position clitic at the cost of complicating the sentence communicative structure. 37 The join KBo 30.39 + KUB 35.164 was not yet available when Carruba compiled the Palaic lexicon. This is the likely reason why Carruba failed to produce the correct syntactic analysis of this sentence. On the applicability of the Watkins-Garrett rule to Palaic, see Garrett 1990: 143-4. 38 Going slightly ahead, let us note that the Luwian word order places the direct object pronominal clitics also after the quasi-reflexive particle =ti= (see Melchert 1993: 226 for examples). This observation is important since, as I will argue below, the pronominal clitic =si developed a similar meaning in Palaic. 39 Carruba 1970: 70, Carruba 1972: 22-23. 40 mЮrhas is conventionally translated as `gods', but `guests' would be equally

were hittite kings divinely anointed?
(9) KUB 35.165, CHT 751, Rev. 9': la-la-an-ta ku[-i"] `You, who crush the curse yourself as much? (as you

119

MS ta-ar-ta-an ma-ri-i"-"i ma-a"-pa-"i mu-З-"i (and) the (evil) tongue, you will satiate/enjoy can)?'.41

The examples given above indicate that =si= can refer to 2/3 person singular and plural subject. The subject clitic does not appear in the second clause of (8), and this would allow one to formally analyze =si as a direct object reflexive pronoun in this case. Yet there is no evidence that mЫs- was ever used as a transitive verb in Palaic without this clitic. In any case, direct object reflexivization could not be the only function of Pal. =si, as the following examples illustrate:
(10) KUB 32.18, CTH 752, MS I.10': [MU"E]N ha-ra-a-a"-ku-wa-ar-zi pa-na-a-ga-an-zi (11) KUB 35.164, CTH 752, MS II.12': -z]i ki-tar-zi ar-"u-З-kat [

However poorly these contexts are understood, we can say that =zi, a sandhi variant of =si does not represent a direct object reflexive pronoun in either of these two cases. Clause (10) probably contains the participial form */panЮgants/ (vel sim),42 and therefore is unlikely to be transitive, while ki-tar in (11) is usually taken as an intransitive verb `lies'. Perhaps one can think of a "transformative" (ki-tar-zi `lies down??'), emphatic, or another secondary function of formally reflexive predicates, but ultimately, one must say

consistent with the Hittite version of the Disappearing Deity Myth. I connect this noun with the Palaic verbs marha- `present oneself, come as a guest' in (6) and marhina- `treat as a guest', as well as with the Luwian verb marha- `to present oneself ' that appears as a Glossenkeil word in the Apology of Hattusili III ( KUB 1.1 IV 5-6 with duplicates). The Palaic word for `god' was possibly tiuna- (cf. Melchert 1984: 40-41). 41 Cf. Melchert 1984: 36. 42 The interpretation of this form much depends or not one assumes that the alternation of verbal forms tАk-ku-wa-a-ti and tАk-ku-wa-ga-ti, discussed in Watkins (1975: 211), is explained by the addition of a -ga- suffix, or the inconsistent rendering of a peculiar sound, such as /"/. If one assumes that pa-na-a-ga-an-zi could likewise reflect a phonetic sequence /panЮ"ants/ (vel sim.), one can perhaps venture a comparison with Hitt. penna-/penni- `drive, hurry', on the assumption that the dropped intervocalic /j/ in /panЮjants/ was replaced by a hiatus-breaker /"/. It would make a perfect sense that the eagle hurries (or is driven, sent) to look for a disappearing deity. Yet, given phonetic uncertainties connected with this word, this etymology remains very hypothetical.

120

ilya yakubovich

non liquet. Note that ki-tar-zi represents a special case, in that =zi is appended to the predicate. This elusive meaning of = si may come as a surprise to a Hittitologist since the cognate and homophonous clitic =si `to him' always functioned as a dative non-reflexive anaphoric pronoun in Hittite. Yet, Melchert proposed that the reflexive value must be posited also for its Lydian cognate -is/-, which was used for both direct and indirect objects. He also showed that in some cases this Lydian particle does not appear in the expected second position, and argued that the "stranded -" following a noun could have an emphatic value.43 It is no accident, remarks Melchert, that the cognates of Hitt. =si develop a reflexive sense in Lydian and Palaic, languages where the third person singular dative pronoun is replaced by =(m)l and =du respectively. Finally, the sifting of the Hieroglyphic Luwian corpus (where the function of =du was extended just in the same way as it was in Palaic) yields a group of seven examples where the element =si is added to what appears to be a perfectly good verbal form.44 Only two verbal stems, the intransitive hwihwisa- `to march', and the transitive iziya- `to make', are attested in conjunction with this element, but restrictions seem to be placed on the person or number verbal forms to which =si can be added. As the examples below illustrate, =si in such cases is normally left without translation.
(12) KARKAM(" A11b+c, § 8 pa-tА-za-pa-wa/i-ta-ґ ( TERRA+LA+LA)wa/i-li-li-tЮ-za mi-i-zi-ґ |tА-tii-zi AVUS-ha-ti-zi-ha |*348(-)la/i/u-tЮ-li-zi-ha |NEG2-ґ ( PES2)HWIHWI-sЮ-tА-si `To those fields my fathers, grand-fathers, and ancestors? have not marched'. ALEPPO 2, § 8. wa/i-mi-ґ ( DEUS)SOL-ni-za ("LIGNUM")ta-ru-sa |i-zi-i-ha-si. `I myself made an image of the Sun-god'.

(13)

The predicates of (12-13) are reminiscent of Pal. ki-tar-zi. These forms have been alternatively explained as medio-passives, or as active verbs with the reflexive pronoun -si. Yet this would-be reflexivisation does not interfere with the clause argument structure

Melchert 1991. Payne 2004: 25. The references to гinekЖy §§ 6, 7 and Tell Ahmar 6, §§ 7, 17 have been swapped.
44

43

were hittite kings divinely anointed?

121

since in neither of the two instances does -si appear to fill in a valency slot of the respective predicate. The evidence of Palaic, Luwian and Lydian, taken together, is enough to demonstrate that the clitic =si developed a quasi-reflexive meaning in this subset of Anatolian languages, and in a number of cases it did not function as a syntactic complement. Given the scarcity of our data, the exact development of =si in each language cannot be precisely reconstructed, but the typological comparison with the Hittite quasi-reflexive clitic *=ti > =z(a) may prove to be illuminating. In Late Hittite, =z/=za could function as a real reflexive pronoun, replacing both the direct and the indirect object, but in Old Hittite it mostly marked "a general, unspecific reference to the subject's active involvement, interest, intent, or mental effort, as the case may be, in the verbal content".45 If we assume that =(s)si could have developed a similar meaning in Palaic, then its usage in the sequence nu=ssi=am=pi simply indicates that the Sun-god is solicited to be actively involved in the task of promoting the king.46 This "ethical dative" reflexive =ssi cannot be adequately rendered in the English translation. Based on the discussion above, I suggest the following translation of the Palaic invocation to the Sun-god:
nu=ku fashullasas tiyaz tabarni LUGAL-i pЮbaz=kuwar tЛ Юnnaz=kuwar tЛ iska(=)nu=ssi(y)=am=pi tЛ ari(=)nu=ssi(y)=am=pi tЛ arЫn=am=pi tЛ widesi [A??]A??-antan=an tЛ widesi Now, Faskhullassas Tiyaz, to tabarna the king you are indeed the father (and) the mother. Anoint him, and exalt him now! You will both see/build him high (and) see/build him strong??.

45 Boley 1993: 2005. See also Oettinger 1997: 410, fn. 3 for the other definitions of the function of =z(a). 46 There are reasons to think that the Palaic clitic =ti, cognate with Hitt =za did not represent a grammaticalized reflexive either. The scrutiny of its attestations (Carruba 1970: 74) indicates that, in all cases except one, it emerges in sentences with an overt accusative complement. In sentence VII of our passage (a=di aban azzkЛ `eat that one'), it appears to function as an intensifying particle accompanying the verb ad- `eat', as the cognate clitic =za frequently does in Hittite. In example (6), the only instance where we lack a direct object, =ti marks the 2 sg. subject of a nominal sentence, in accordance with the same rule that Hoffner (1969) established for Hitt. =za. It is reasonable to assume that "Hoffner's clitic" developed from some sort of an ethical dative. This raises an intriguing question about the nature of semantic distinction between the clitics =si and =ti in Palaic.

122

ilya yakubovich II. Religious Interpretation

It is a well-known fact that the Hittite kings lacked divine genealogy and enjoyed only posthumous deification. While on earth, they had to content themselves with being mere stewards of gods, deriving their power from the divine mandate.47 The following passage from a ritual incantation is frequently cited as an illustration of this doctrine: "The land belongs only to the Storm God; Heaven and Earth together with the army belong only to the Storm God. And he made labarna, the king, his deputy and gave him the whole land of Hattusa. Labarna rules the whole land with (his own) hands. Whoever approaches the person of the king or the borders (of the land), may the Storm God destroy him'! (CTH 821).48 The textual evidence also indicates that the Sun Goddess of Arinna, the consort of the Storm God of Hatti, had her own part in delegating power on earth to the king since the times of the Hittite Old Kingdom.49 Besides serving as the gods' representatives among men, Hittite monarchs had to represent the humans before the gods. In their capacity as the high priests of the Hittite kingdom, they probably had to spend several months a year touring important cult centers and participating in periodic festivals aimed at ensuring the favorable disposition of their divine overlords toward the land of Hatti. In a prayer for the removal of plague from Hatti, Mursili II did not dare call himself a king of the land, but humbly referred to himself as the priest of the gods. Trying to placate them with his piety, he argued: "When I celebrated the festivals, I busied myself

47 One should distinguish this relationship, exercised ex officio by both parties, from less formal ties of the client-patron type that Hittite rulers could establish with individual gods, even before their official ascension to the throne. We are well informed about the divine protectors of the Empire Period kings: Mursili II ascribed his military victories to the patronage of the Sun Goddess of Arinna, Muwatalli II sought the intercession of the Luwian Storm God of Lightning, Hattusili III relied on Ishtar of Samuha at every stage of his career, Tuthaliya IV can still be seen embraced by Sarruma on the Yazilikaya relief. The same practice lurks in the titles of the Neo-Hittite rulers: thus Azatiwada, tyrant of Hiyawa (Que) referred to himself as "the Sun-blessed man", while Yariri, tyrant of Carchemish, called himself tasparuwantis `beloved by Tessub and Runtiya (?)'. 48 For the transliteration of this passage see Archi 1979: 31-2 ( IBoT 1 30 2-8). The same paper can be consulted for numerous other passages illustrating the (desired) relationship between the Hittite king and the gods. 49 Haas 1994: 190.

were hittite kings divinely anointed?

123

for all the gods, I did not pick out any single temple" (CTH 378, II ).50 He was sure that the gods would not hesitate to punish the whole land for his personal negligence in ritual matters. A role of almost equal importance in the affairs of religion belonged to the Hittite queen (tawananna), who participated in religious festivals as the high priestess of the Hatti land.51 On the other hand, Hittite monarchs were free to choose their own heirs, even though the pool of potential candidates was eventually restricted by the Telibinu Proclamation. We do not have any information about divination or oracular inquiries leading to the choice of a royal successor. Since the gods apparently had little say in the original designation of their next representative, a special inauguration ritual was necessary for the post factum divine investiture of a new Hittite king.52 Several oracles from the time of Tuthaliya IV testify to the importance attached to his festive inauguration ceremony. In particular, we learn that it had to be postponed several times because of the fear that an untimely enthronement may cause the illness of the king, or even a coup d'Иtat.53 Presumably, these were among the ways the Hittite gods could veto the election of their unwanted deputies and priests. Unfortunately, no Hittite script of an inauguration ceremony has been identified so far in the royal archives of Hattusa.54 The two texts containing Hurrian incantations that are possibly related to this ceremony remain poorly understood due to our insufficient

Singer 2002: 57-8. Haas 1994: 196. 52 This ceremony could happen after the death of the previous king, as well as during his lifetime. In the latter case, it resulted in a co-regency between the two kings. Such a co-regency is attested in the case of Tuthaliya I/II and Arnuwanda I, and probably in the case of Hattusili III and Tuthaliya IV (see van den Hout 1991: 277-8). For the general discussion of the inauguration ceremony, as well as the interpretation of most of the examples collected in (14-19), see KЭmmel 1967: 43-9. 53 Van den Hout 1997: 106-19. See especially KUB 18.36 12'-18'. 54 Groddek (2002) labeled the recently published ritual for Mursili II ( VSNF 7) as "a purification ritual for Mursili II on the occasion of his enthronement". Further in the same article, he clarified, however, that the purpose of this ritual can be defined more exactly as the attainment of success in military affairs. The text appears to refer back to the time of the co-regency between Suppiluliuma I and Mursili II, and the latter is explicitly called an armantalliya- king in this context. Therefore, one has to conclude that VSNF 7 does not represent the minutes of the royal coronation, since this ceremony must have occurred earlier.
51

50

124

ilya yakubovich

knowledge of the Hurrian language.55 We have to rely on the description of a mock coronation preserved as a part of the ritual for a substitute king.56 In the event an oracle should foretell imminent danger for the king of Hattusa, one had to temporarily crown a prisoner of war in the hope that the wrath of the gods would befall the substitute body natural of the king instead of the real one. The following ritual actions are prescribed:
(14) KUB 24.5 Obv. 19'-22' + 9.13. 7'-9', CTH 419, NS nu Lз"U.DAB I"-TU л.DыG.GA LUGAL-UT-TI i"-ki-ya-an-zi nu ki"an [t(e-ez-zi ka-a)]-"a-wa ka-a-a" LUGAL-u" "UM LUGAL-UT-TIya-w[a]-kАn ke-e-da-ni [te-eh-hu-un TзG LU ]GAL-UT-TI-ya ke-eda-ni wa-a"-"i-ya-nu-un [TзG]lu-pa-an-ni-na-wa-kАn ke-e-da-ni "i-ya-n[u-u]n. `They anoint the prisoner with the fine oil of kingship and (he) says: "Behold, this one is the king! [ I have bestowed] the title of kingship upon this one, I have clothed this one in the [garments] of kingship, and I have put the lubanni-cap on this one" '.

Each of the actions mentioned in this passage was, no doubt, significant in its own right, yet it is the anointment of the king that probably constituted the most significant salient element of the ceremony in the eyes of the Hittites. This is evident from the fact that it was mentioned in the two Middle Hittite passages as a synonym for the inauguration itself. The following example is taken from an instruction for military officers.
(15) KBo 16.24+25 I 64'-66', CTH 251, MS 54' LUGAL-u"-"a-an ku[-in (circa 8 signs) d]a?-a-i? 55' [na-an - kАn] A-NA LUGAL- TIM i " -ki-iz-zi na-an-za [ " ume-e"? x x] x x [x x-]KU-NU 56' ["e-ek-t]e-en `Whoever the king takes?.... and anoints to the kingship, you and your... (must) recognize him'.

The idiom Hitt. *hassuwizni iski- `to anoint for kingship' is repeated in the following fragmentary text dealing with the enthronement of Tuthaliya I/II or Tuthaliya III:57
These texts are published in transliteration in Haas 1984 (#11 and #41). For the connection of the first one with the royal inauguration, see Haas 1994: 192 and the discussion of example (17) below. A passage from the second text that probably deals with the enthronement of the new king is discussed in Salvini 1991: 128. 56 Edited in KЭmmel 1967. 57 For the latest edition of this fragment, see Otten 1990: 224-6 (with references to earlier literature).
55

were hittite kings divinely anointed?
(16) KUB 36.119. 2'-4', CTH 271, MS 2' [mDu-ut-]ha-li-ya-an "E"[(-) ... 3' [x x-]an mDu-ut-ha-li-yaЕ[-... 4' [n]a-an-kАn LUGAL-u-iz-ni i"-k_-i[r ...

125

A colophon to a Hurrian ritual (CTH 784) may also contain the reference to the king's anointment, even though the word sarrassi`kingship' is severely mutilated, and its reading is not assured. The word for "anointment" in (17) is not completely preserved either, but its reconstruction is confirmed by the Hurrian text of the tablet. The first part of this text refers to a number of ritual actions that appear to be necessary for the preparation of consecrated oil, while its second part describes the anointment ceremony, during which the ritual practitioner addresses both the king and the queen several times, and invokes both "Tessub of kingship (and) lordship" and Hebat.58 On the whole, the Hurrian text remains imperfectly understood, but Haas tentatively identified it as a purification ritual performed in connection with the inauguration of Tuthaliya I/II and Taduheba.59
(17) KUB 27.42 Rev. 27'-8', CTH 784, MS 27' DUB 1.KAM QA-TI "Еr?-rЕ?-Е"?-"_-yЕ-Е"?-"[a?-a"60 i]"-ga-u-wa-a" 28' x(-)!x61 LзSANGA DUMU.LUGAL `One tablet, finished, of the [a]nointment [for] kingshi[ p(?). [. . .] the priest, son of the king'.

Another colophon fragment indicates the existence of a special ceremony for the queen. Note that allassi- `queenship/ladyship', like

58 In addition to example (22) below, it is useful to consider the following phrases from this text: du-i-du-ma hЕ-Е-"Е-a-ri e-pМ-_r-"_-ha-a-i (Obv. 26) `they prepared(?) oil with one of lordship', and [d]u-_-d#u-mЕ ha-a-"a-a-ri al-lЕ-Е-"a-a-e `they prepared (?) oil with queenship/ladyship' (Obv. 27). The predicate of these sentences is discussed in RЖseler 1999: 399 (see especially fn. 26), who opts for the translation `to cook'. I believe that Hurr. du- is cognate with Urart. du- `to do, make', and therefore translate it as `to prepare'. The transliteration e-pМ-ir-"i ha-a-i ( Haas 1984: 115) is not supported either by word spacing in the autography or by obvious grammatical considerations. It is true that that the standard stem for `one of the lordship' is ib-ri-i"-hi- rather than e-pМ-ir-"i-hi-, but our form can be easily derived by analogy based on e-pМ-ir-"e `lordship'. 59 Haas 1994: 192 60 Haas (1984: 119, fn. to 27') suggests an unlikely restoration [ma-a-a]n mdu-u[t. 61 Haas (1984: 119) suggests a restoration INIM kАn-li `the words of Kantuzzili'. This restoration is very plausible semantically, since we know that Kantuzzili had a title `the priest, son of the king', but does not seem to match the sign traces on the photograph.

126

ilya yakubovich

sarrassi- `kingship,' represents a transparent Hurrian borrowing. The text referred to by this colophon has been almost completely lost.
(18) IBoT II 120, CTH 500, NS 2' [ DUB x+?]1.KAM QA-TI 3' [ ]x-ni al-la-a"-"i-a" i"-ki-ya-[u-wa-a"]. `( X+)1 tablet, finished. [. . .] of anoint[ing] for queenship'.

Finally, a Hittite prayer to the Storm-god of Nerik mentions an act of ritual anointment in a somewhat different context. Here we are dealing not with the coronation, but with the ordination of Tuthaliya IV, which will occur in two provincial towns. KЭmmel was of the opinion that "anointment for priesthood" in this case is synonymous with the royal inauguration.62 Such a hypothesis, however, leaves open the question of why the prayer on this occasion is addressed to the Storm-god of Nerik, and not to the Storm-god of Hatti. I follow a more straightforward suggestion of Singer, according to which Tuthaliya IV is about to be appointed to the office of the high priest of the Storm-god of Nerik.63 Since the Storm-god of Nerik is a son of the Storm-god of Hatti, the head of the Hittite pantheon, this appears to be a suitable appointment for an heir to the Hittite throne.
(19) KUB 36.90 Obv. 15-18, CTH 386, NS lu-uk-kat-ta-ma-a"-kАn UD.KAM-ti mDu-ut-ha-li-ya-an tu-e-da-a" a-a""i-ya-an-ta-a" pИ-e-da-a" URUHa-ak-mi" URUNe-ri-ik A"-"UM LзSANGA-UT-TIM i"kАn-zi. nu lu-uk-kat-ti UD-ti A-NA EZEN4-KA e-hu `At dawn one will anoint Tuthaliya for priesthood in your favorite places, Hakpis and Nerik. Come tomorrow to your festival'!

The passages (14-17) naturally bring to mind the analogy with the ritualized anointment of kings practiced by the Ancient Hebrews, which was imitated in the Christian tradition, and in part survives in it up to now.64 This analogy is even more significant since the anointment of kings appears to be unknown in Egypt or Mesopotamia. The closest parallel to passage (19) is the Akkadian ritual for the

KЭmmel 1967: 43 with fn. 3. Singer 2002: 106. 64 An integral part of crowning an English king or queen is the anointment performed by the Archbishop of Canterbury acting as the supreme hierarch of the Church of England. Yet, the anointment is no more a universal requirement for the European royalty. Thus, Professor van den Hout informs me that the Dutch queen is not considered to be a divinely anointed individual.
63

62

were hittite kings divinely anointed?

127

installation of Baal's high priestess at Emar ( Emar 369), which mentions the anointment with oil twice.65 The ritual actions prescribed in this text contain, in their turn, numerous parallels with the rites performed during the ordination of Hebrew priests, as described in the Bible, one of which is their anointment.66 In this case, too, neither Egyptian, nor Mesopotamian texts contain any reference to this ritual act.67 The fact that anointment was used in the rites of passage in both Hittite and Syro-Palestinian traditions is certainly significant and interesting, but one has to discuss the implied meaning of this ritual in each of the two cultures before making any statements about possible influences from either side. The only comparative monograph on this subject known to me is Kutsch 1963. Although the author of this book strove to describe all types of ritualized anointment in the Ancient Near East, it is natural that he, as an Old Testament scholar, treated the Biblical passages in a more systematic fashion. Ernst Kutsch drew a sharp line of distinction between the anointment of kings and priests in the Hebrew world. Starting with an observation that all mentions of this ritual act in connection with the investiture of high priests occur in those passages from the Torah that are normally ascribed to the priestly ( P) source, he concluded that they reflect post-exilic realities, and the ritualized anointment of Hebrew priests cannot be proven for an earlier period. Now that we have evidence for the anointment of priestesses in Late Bronze Age Syria, this conclusion would appear to be in need of revision. The purpose of

65 Edited in Fleming 1992. See lines 3-4 and lines 20-21 of Text A for the mention of anointment. 66 Klingbeil 1998: 556-7. 67 Akkadian pa"Л"u `a type of low-ranking priest' is traditionally interpreted as "the anointed one" based on the etymological connection of this form with Akk. pa"Л"u `to anoint'. Yet there is no textual indication that this type of priest had to undergo a special ritual of anointment. Archi (1996) has convincingly shown that the obviously related Eblaite professional name pa4-"e" can have only the active meaning `the anointer'. The only obstacle to the similar interpretation of Akk. pa"Ю"u is the claim that the substantivized adjectives with the morphological structure CaCЛC- have the passive meaning in Akkadian ( Fox 2003: 188). Yet this morphological type is rare and unproductive in East Semitic, and so statistical generalizations have little value here. In other Semitic languages, forms in CaCЛCcan be also used as active adjectives (Arabic, Fox 2003: 189) or active participles (Ge'ez, Fox 2003: 190). For possible Hebrew active adjectives that follow the same pattern see Fleming 1993 (esp. fn. 57), for a different opinion about these forms compare Huenegard 1999.

128

ilya yakubovich

anointing priests, according to Kutsch, was ritual purification, and here he is followed by other scholars.68 As for the anointment of kings of Israel and Judea, it is mentioned many times in historical books, and so can be as old as the respective kingdoms. The general purpose of anointing kings was to increase their power/authority. This is argued based on the fact that the act of anointment is sometimes accompanied by the request to accomplish certain mighty deeds, mostly of military nature.69 In addition to distinguishing various "logical objects" of the anointment, Kutsch classified the same act according to its "logical subjects". The anointment of kings in the kingdoms of Israel and Judea could proceed either from the Hebrew God, in which case it was performed by a prophet as his representative, or from the Hebrew people. Thus we learn that, after David was initially anointed as a child by the Prophet Samuel ( I Sam. 16.13), his claim to kingship was confirmed by the "men of Judah" ( II Sam. 2.4) and by the "elders of Israel" ( II Sam. 5.3) through repeated anointments.70 In this case, multiple ritual acts must reflect the fact that the divine recognition of King David came about much earlier than his public claim to the throne. In the case of legitimate royal succession, one anointment was probably sufficient since the newly invested monarch could rely both on divine and popular support, at least before he commits major misdeeds. The paragon case here is the anointment of Solomon, son of David ( I Kings 1:39, 45). Although it was administered by the priest Zadok with the assistance of the prophet Nathan, the stated purpose of this act was to ensure the loyalty of the people of Israel to the proclaimed heir apparent in the face of an attempted usurpation.71 When Kutsch applies this classification to Hittite data, he comes to rather unexpected conclusions. He claims that the anointment practiced among the Hittites, unlike the one known from the Bible, always had people as its "logical subject". His reasoning was based on the lack of information about Hittite priests administering the royal anointment in the sources available to him. He remarked, furthermore that, if the direct speech in (14) was uttered by a priest,

68 69 70 71

Kutsch 1963: 22. Cf. similar remarks of Klingbeil (1992: 244). Kutsch 1963: 55-6. Kutsch 1963: 53-4. Cf. Kutsch 1963: 66.

were hittite kings divinely anointed?

129

then it follows from the text that this priest could not be in charge of the anointment in the Substitute Ritual. He concluded his argumentation by referring to the prominent role that Hittite aristocracy played in determining the successor to the throne in the times of the Old Kingdom, and hypothesized that in those days the anointment signified "empowering" the king by the nobility, while later it was administered as a homage to tradition.72 Given our present knowledge of Hittite texts, the arguments of Ernst Kutsch do not appear convincing. In example (15), not discussed in Kutsch 1963, the anointment is administered by the king, who also happens to be the high priest of the Hatti land.73 Example (18), equally unknown to Kutsch, introduces the incantations associated with the royal anointment as a ritual text composed by a priest. In (19), Tuthaliya III is ordained for priesthood, and it would be counterintuitive to assume that such a ceremony was performed by laymen rather than clerics. Likewise, it is difficult to understand why Hittite aristocracy would be interested in anointing the queen, who played an active role in the Hittite state cult, but did not have a chance of inheriting the throne. As for (14), we do not have any information on whether one or more ritual practitioners had to participate in the mock coronation, and so no argument can be made regarding the identity of the person in charge of the anointment. Finally, one must keep in mind that no mention of "noblemen", "elders", or "warriors", has been made in connection with the ritualized anointment in any of the texts cited above, and so assigning to them the primary role in the inauguration ceremony would represent a completely unfounded speculation. Summing up, Hittite texts yield no philological support to the "popular" anointment of kings, and provide some evidence for their anointment mediated by priests. The theory of divine anointment would, furthermore, fit well with the Hittite ideology of kingship, as outlined above. Yet, its only decisive proof would be a textual passage where a god is mentioned as the "logical subject" of this ritual act. Such a passage is not known to me from the Hittite sources, but emerges in the Palaic context discussed in the first part of this paper. We do not have evidence that CTH 751 has any

Kutsch 1963: 38-9. The inconsistence of this example with the hypothesis of Kutsch has been already noted in KЭmmel 1967: 45.
73

72

130

ilya yakubovich

connection with the inauguration ritual, and so the request to the Sun-god to anoint the king is best taken metaphorically, as a rough equivalent to the following request to exalt him. Yet the nature of such a metaphor would be completely unclear unless we assume that it refers to a ritual that was administered to a king on behalf of the gods. This is not to say that the Hittite enthronement ceremony ought to be entirely devoid of propagandistic impact. The texts are silent on this matter, but what we know about royal coronation in other traditions strongly suggests that Hittite monarchs would not miss a chance to use it in order to enhance their own popularity, or that of their designated heirs. In other words, the distinction between the two types of royal anointment is no more useful in our case than it is with reference to King Solomon's coronation. Neither does the rigid distinction between anointing kings and priests appear to be methodologically grounded in Anatolia. Since both the king (tabarna-/labarna-) and the queen (tawananna-) had pivotal functions in the Hittite state cult, the anointment for kingship/queenship (1418) can be regarded as a special type of anointment for priesthood (19). Using a quasi-linguistic turn of phrase, one can say that the opposition between the two types of "logical subjects" and "logical objects" appears to be neutralized. In more mundane language, it appears to be simply unnecessary in the Hittite milieu. The elimination of artificial distinctions makes it easier to come up with a plausible scenario for the development of ritualized anointment of kings and priests among the Hittites. We know that the ritual purity of cult officials was a matter of enormous concern in the Hittite kingdom. The instruction for the temple personnel (CTH 264) begins with the detailed descriptions of hygienic rules that must be observed by the temple officials in order to prevent the ritual contamination of the temple, and threatens them with an imminent divine punishment should any of these rules be violated.74 Even more serious were the measures taken in order to ensure the purity of kingship. Even the king's shoe and chariot could be made only from the leather of animals slaughtered in the palace precincts and prepared under the strictest conditions of hygiene in the palace kitchen (CTH 265).75 This pervasive obsession with ritual cleanli-

74 75

SЭel 1985: 22-6, cf., ibid.: 66-72. Bryce 2002: 15-16.

were hittite kings divinely anointed?

131

ness correlates with the existence of special rituals performed in order to ensure the transition of a person from the profane to the consecrated state. A representative of this group is the Hurrian ritual itkalzi (translated into Hittite as "purifying the mouth"), which was performed in a provincial city Zithara on behalf of Tuthaliya III and Taduheba ( part of CTH 777). Purifying substances used during this ritual included water, cedar, tamarisk, oil and silver.76 In particular, anointing with oil is mentioned in the following contexts:77
(20) KUB 29.8 iii 21-2 pa-a-hi-ip "e-hur-ni-pa-a-e hu-u-da-an-na-a-e ha-"a-ra-a-e ha-a-"u-lee-e" `Let your head be anointed with the oil of life and destiny(?)'. KUB 29.8 iii 39-40 ku-du-ni-ip i-ni-pa-a-i ha-a-"a-ra-a-i ha-a-"u-le-e-e" `Let your neck be anointed with the divine oil'.

(21)

Frequent references to an abstract "man" or "woman" for whom the itkalzi ritual can be performed indicate that it was not regarded as appropriate for the king and queen alone.78 Yet, a similar formula appears also in CTH 784, the colophon to which was given in (17). This ritual was definitely restricted to the royal couple, and possibly constituted a part of the inauguration ceremony:
(22) KUB 27.42 Rev. 20'-21', MS. nam-ma A-NA LUGAL te-ez-zi ka-pu-u-"i-ni-ip e-ne-pa-i he-e"-mi-ir-["i]ni-pa-a-i ha-"a-ra-a-i ha-a-"u-le-e-e" Then he says to the king: "Let your k. be anointed with the divine oil of h.'. ( Then the same thing is said to the queen).

The parallelism between the itkalzi ritual and CTH 784 is not limited to these anointment formulae. In both texts, they are followed

76

This ritual is edited in transliteration in Haas 1984 (for its outline see pp.

2-3).
77 The morphological analysis of examples (21-3), as well as the discussion of some other passages from the itkalzi ritual involving anointment, is given in Wilhelm 1997: 180. 78 See, for example, the Hittite script for a part of the itkalzi ritual ( Haas 1984: 86-91). Note also that the Hurrian word sarri- `king' is used throughout the ritual only with reference to Tessub, the head of the Hurrian pantheon, and never with reference to the patient. The references to Tasmisarri (=Tuthaliya III ) and Taduheba that do occur in the text appear to have been inserted there on a particular occasion when the ritual was performed for them.

132

ilya yakubovich

with the wish for the right and left sides of the ritual patron(s) to be purified with certain "male" and "female" entities (tulpura) respectively.79 It appears that the author of CTH 784 adapted the ritualized purification with oil, as described in the itkalzi ritual, for a special ceremony involving the royal couple.80 Much depends, of course, on the interpretation of the colophon (17), and the poor state of its preservation makes this interpretation uncertain. But even if we assume that CTH 784 did not have anything to do with the enthronement, and described some kind of periodic anointment that was administered to the royal couple in order to maintain their ritual purity, it may nevertheless be considered as an intermediate step between an ordinary purification ritual involving oil and the anointment for kingship. Anointment as a passage rite appears to represent a derived phenomenon within Anatolian culture. The fact that the ritualized anointment is mentioned five times in connection with royal inauguration, and only once in Hittite texts dealing with priestly installation, probably reflects no more than the skewed nature of our evidence. The archives of Hattusa were collected in order to accommodate the needs of the royal family, and the only time we hear of "anointing for priesthood", this ritual concerns a prominent member of this family. One may, however, legitimately wonder how far down the social hierarchy the anointment would be extended. The preserved fragments of EZEN4 pulas ("festival of the lots"), celebrated in the Hatti land in connection with installing new priests, do not contain any reference to oil. The potential servants of gods had to content themselves with a more egalitarian purification with water, accomplished immediately before manipulating the lots would determine their new appointments.81 On the other hand, a new high priestess of Baal at Emar had to be anointed with oil immediately after the lots had indicated her selection, and once again after she had been

79 See Wegner 1995: 122-4 for the discussion of these passages in both texts. Yet another Hurrian ritual that contains parallel anointment formulae has been published in Wilhelm 1995. It is probably connected with the agricultural cycle, since anointment is applied there to the plough and the yoke, among other objects. 80 It does not seem possible to prove on paleographic or historical grounds that CTH 777 was composed earlier than CTH 784. Yet, the opposite scenario, according to which a ritual for the royal couple is modified for general application, appears to be less likely since any king could view such an activity as laesio majestatis. 81 Taggar-Cohen 2002: 132-7.

were hittite kings divinely anointed?

133

consecrated through shaving. One wonders whether ritual purification with oil could not constitute a privilege of priests having a certain rank in Late Bronze Age Anatolia and Syria, but the scarcity of the available evidence does not allow me to answer this question in the positive. In any case, the parallelism between the anointment of newly appointed clerics practiced in Nerik and Emar, in two opposite corners of the Hittite Empire, indicates that this ritual was not restricted to a certain narrow geographic area. In this case, too, it is logical to assume that anointment as a rite of passage arose out of purification rituals. We know that passage rites tend to be cross-culturally administered by those already initiated to a particular ritual state. This conforms to the evidence of (15), where the king pledges to personally anoint his heir. Accordingly, one can hypothesize that Tuthaliya IV was expected to be anointed to the priesthood with the blessings of other priests in (19). But what if the king / high priest died before he had a chance to initiate his successor, or if a new ruler came to power by usurping the throne? In this case, the doctrine of divine anointment, consistent with the Hittite ideology of kingship, would provide the priestly elite with an opportunity to perform the investiture of the new king on behalf of the supreme gods, thus avoiding potential disastrous effects of a profane rule. This doctrine would naturally imply that all the kings, including those chosen by their predecessors, needed to obtain the final approval from their divine overlords in order to exercise the duties of kingship. The anointed monarch would be regarded both ritually fit to celebrate the gods, and provided with special powers to carry out their will on earth. The first function united the king with the queen and the other priests, and probably constituted the original raison d'Йtre for the anointment as a ritual act. The second function was unique to the sovereign and, unlike the first one, finds a close parallel in the Hebrew ideology of kingship. Although the second function is formally independent from the first one, the fact that they pertained to the same person could trigger its secondary association with the act of anointment. This association, however, must have been very old since we encounter it already in our Palaic fragment, where the Sun-god is expected to make the king high and strong. The interpretation of various types of anointment discussed in this section is summarized in the table below.

134

ilya yakubovich
Anointment of priests ( Hatti, Syria, Israel) ( Hatti) Anointment of kings / high priests + + Anointment of kings ( Israel) +

Purification Empowerment

+

All the relevant texts coming from the archives of Hattusa show the Middle Hittite or later ductus, and the only preserved ritual of royal anointment is composed in the Hurrian language. Therefore, one can advance a hypothesis that anointment as a religious ceremony originated in the Hurrian or Syrian milieu, and then found its way to Hatti in the Middle Hittite period. Several additional considerations can be mentioned in support of this hypothesis. Oil appears to be fairly important for the Hurrian religious tradition: the so-called "Song of Oil", which may represent a particular tale of the Kumarbi cycle preserved only in the Hurrian transmission, contains passages where Oil appears to act as an animate being, participating in a conversation with Sawoska of Niniveh.82 Interestingly enough, oil is also mentioned in connection with the city of Niniveh in the ritual of royal anointment (CTH 781), which was composed in Central Anatolia.83 Finally, the interaction between oil and various gods is described in the Hurrian purification ritual of Allaiturah(h)i, which was composed in Syria (CTH 780).84 The only piece of evidence that can be used to argue that anointment represented an indigenous Anatolian tradition is our Palaic invocation. It is frequently assumed that all the Palaic texts were recorded in the Old Hittite Period, and that the Palaic language became extinct soon after that. Yet, since none of the fragments of CTH 751 shows the Old Hittite ductus, one is not obliged to take this aprioristic claim as a proven theory. On the other hand, one should not discard a possibility that both our Palaic ritual and Hurrian texts mentioned above reflect some Anatolian cultural tradition that predates the arrival of IndoEuropeans and Hurro-Urartians to this area. Jared Miller has recently advanced a similar substrate hypothesis in connection with substitution and scapegoat rites that are reflected in certain Old Hittite /
82 83 84

See texts 10 and 21-25 published in Salvini and Wegner 2004. KUB 27.42 Obv. 9-14 transliterated in Haas 1984: 114. RЖseler 1999: 396-400 with references.

were hittite kings divinely anointed?

135

Old Script texts, but occupy a more prominent position in various ancient cultures situated along the northeastern Mediterranean coast, from Greece to the Levant.85 If one is willing to extend this unconstrained hypothesis to ritualized anointment with oil, the re-dating of CTH 751 becomes unnecessary. We do not have enough data to state whether the anointment of kings in Hatti represents a secondary extension of the anointment of priests, or vice versa. It is remarkable however, that no model of internal development seems to be able to account for all the instances of ritualized anointment among the Hebrews. The methodological nuances introduced in Kutsch 1963 reflect substantial distinctions between the anointment of kings and high priests, which were always two distinct offices in Israel and Judea, as well as between the underlying ideologies of these two ceremonies. It is reasonable to follow Ernst Kutsch when he analyzes the anointment in the Bible as a set of heterogeneous traditions some or all of which could be borrowed wholesale from neighboring cultures. The Biblical tradition itself regards kingship as an institution introduced in Israel as a way of imitating other nations ( I Sam. 8:5, 19-20). Therefore it is likely that ancient Hebrews imported various elements of the royal inauguration ceremony, including the anointment ritual, for reasons of prestige, but modified its ideological background so as to suit the local needs. In this connection, one can think about the influence of Hurrians on Canaanites in the second millennium B.C., or the cultural impact of Neo-Hittite states in Syria in the early first millennium B.C.86 References
Archi, Alfonso. 1979. "Auguri per il labarna". In Studia mediterranea Piero Meriggi dicata, ed. O. Carruba, pp. 27-51. Pavia: Aurora. ----. 1996. "Eblaita: Pю"I"U `Colui che И addetto all'unzione; sacerdote purificatore, cameriere al servizio di una persona' ", Vicino Oriente 10: 37-71.

Miller 2004: 464-68. Compare the opinion of Taracha 2000: 207-12, according to whom the origin of both traditions, as well as that of the holocaust sacrifice, is to be sought in Syria. 86 I do not have anything to say about the custom of pouring oil on high officials in Egypt at their investiture, discussed in Kutsch 1963: 34-5. Yet, since it is the pharaoh, and not a god, who assumes the responsibility for this act, I believe that it does not represent a close parallel to the divine anointment of kings among the Hebrews.

85

136

ilya yakubovich

Boley, Jacqueline. 1993. The Hittite Particle -z/-za. Innsbruck: Institut der Sprachwissenschaft der UniversitДt Innsbruck. Bryce, Trevor. 2002. Life and Society in the Hittite World. Oxford University Press. Carruba, Onofrio. 1970. Das Palaische. Texte, Grammatik, Lexikon. StBoT 10. Wiesbaden: Harrasowitz. ----. 1972. BeitrДge zum Palaischen. Istanbul: Nederlands Historisch-Archaeologisch Instituut. Fleming, Daniel E. 1992. The Installation of Baal's High Priestess at Emar. A Window on Ancient Syrian Religion. HSS 42. Atlanta, GA: Scholars Press. ----. 1993. "NюBш and MUNABBIюTU: Two New Syrian Religious Personnel". Journal of American Oriental Society 113/2: 175-83. Fox, Joshua. 2003. Semitic Noun Patterns. HSS 52. Winona Lake, IN: Eisenbrauns. Garrett, A. 1990. The Syntax of Anatolian Pronominal Clitics. Harvard PhD dissertation. ----. 1992. "Topics in Lycian Syntax". Historische Sprachforschung 105 (2): 200-12. Goedegebuure, Petra M. 2002. "Remarks on the Hittite Royal Substitution Ritual". Journal of Ancient Near Eastern Religions 2: 61-73. Groddek, Detlev. 2002. "Ein Reinigungsritual fЭr Mur"ili II anlДsslich seiner Thronbesteigung". Hethitica 15: 81-92. GЭterbock, Hans G. 1958. "The Composition of Hittite Prayers to the Sun". Journal of American Oriental Society 78 (4): 237-45. Haas, Volkert. 1984. Die Serien itkahi und itkalzi des AZU-Priesters, Rituale fЭr Ta"mi"arri und Tatuhepa sowie weitere texte mit bezug auf Ta"mi"arri. CHS 1:1. Rome: Multigrafica. ----. 1994. Geschichte der hethitischen Religion. HbO I 15. Leiden: Brill. Hawkins, J. David. 2000. Corpus of Hieroglyphic Luwian Inscriptions. Volume I. Part I, II: Texts. Berlin-New York: W. de Gruyter. Hoffner, Harry. 1969. "On the use of Hittite -za in nominal sentences". Journal of Near Eastern Studies 28: 225-30. van den Hout, Theo. 1991. "Hethitische Thronbesteigungsorakel und die Inauguration Tudhaliyas IV". Zeitschrift fЭr Assyriologie 81 (2): 275-300. ---- 1998. The Purity of Kingship. An Edition of CTH 569 and Related Hittite Oracle Inquiries of Tuthaliya IV. DMOA 25. Leiden: Brill. Huenegard, John. 1999. "On the etymology and meaning of Hebrew NюBн' " Eretz-Israel 26: 89-93. Kammenhuber, Annelies. 1959. "Das Palaische: Texte und Wortschatz". Revue hittite et asianique 17 (fascicle 64). Klingbeil, Gerald A. 1998. A Comparative Study of the Ritual of Ordination as Found in Leviticus 8 and Emar 369. Lewinston-Queenston-Lampeter: Edwin Mellen Press. Kutsch, Ernst. 1963. Salbung als Rechtsakt im Alten Testament und im alten Orient. Berlin: A. TЖpelmann. KЭmmel, H. M. 1967. Erzatzrituale fЭr den hethitischen KЖnig. StBoT 3. Wiesbaden: Harrassowitz. Melchert, H. Craig. 1984. "Notes on Palaic". Zeitschrift fЭr vergleichende Sprachforschung 97(1): 22-41. ----. 1990. "Adjectives in *-iyo- in Anatolian". Historische Sprachforschung 103: 198207. ----. 1991. "The Lydian emphasizing and reflexive particle -/is". Kadmos 30: 131-2. ----. 1993. Cuneiform Luvian Lexicon. Chapel Hill: Self-published. ----. 1994. Anatolian Historical Phonology. Amsterdam-Atlanta: Rodopi. ----. 2004. Dictionary of the Lycian Language. Ann Arbor-New York: Beech Stave Press.

were hittite kings divinely anointed?

137

Miller, Jared. 2004. Studies in the Origins, Development, and Interpretation of the Kizzuwatnean Rituals. StBoT 46. Wiesbaden: Harrassowitz. Neu, Erich. 1996. Das Hurritische Epos der Freilassung I. Untersuchungen zu einem hurritisch-hethitischen Ensemble aus Hattusa. StBoT 32 Wiesbaden: Harrassowitz. Oettinger, Norbert. 1997. "Der Partikel -z des Hethitischen' (mit einem Exkurs zu den Medialformen auf -t, -ti )". In Berthold DelbrЭck y la sintaxis indoeuropea hoy, ed. E. Crespe and J. L. GarcМa-RamСn, pp. 407-20 Madrid: Ediciones de la UAM / Wiesbaden: Reichert. Payne, Annick. 2004. Hieroglyphic Luwian. Wiesbaden: Harrassowitz. RЖseler, Ingeborg. 1999. "Hurritologische Miszellen". Studies on the Civilization and Culture of Nuzi and the Hurrians 10: 393-99. Salvini, Mirjo. 1991 "Betrachtungen zum hurritisch-urartДischen Verbum". Zeitschrift fЭr Assyriologie 81 (1): 120-32. Salvini, Mirjo and Wegner, Ilse. 2004. Die mythologischen Texte. ChS 1:6. Rome: Istituto do Studi sulla CiviltЮ dell'Egeo e del Vicino Oriente. Soysal, O