CC BY-NC-ND
22
Miguel Valério New University of Lisbon
Hani-Rabbat as the Semitic name of Mitanni*
This article argues in favor of (re-)reading the Semitic name of Mitanni as Hani-Rabbat against the presently consensual Hanigalbat. In connection with this, an etymology (so far lacking) for this geographical name is proposed. It is argued that Hani-Rabbat is a West-Semitic (Amorite) compound meaning 'Great Hani', in contrast with the Middle-Euphratean country of Hana. Hence, Rabbat marks the opposition between two Hanaean lands set off by the Euphrates. Hani / Hana ought also to be linked with Akkadian hanu and Humanizing haniahhe, all being designations of probable West-Semitic origin used for certain human groups that followed a (semi-)nomadic lifestyle on both sides of the river. The scenario is one of linguistic contact between Amorite, Akkadian and Hurrian. Finally, the toponym Hanu AN.TA 'Upper Hanu', attested in a Middle-Assyrian letter, is examined as a possible Sumerographic spelling of Hani-Rabbat.
Keywords: Hani-Rabbat, Hanigalbat, Mitanni, Hana, Amorrite, West-Semitic, Assyrians, Middle Assyrian, Akkadian, Middle Euphrates, nomads.
When the Assyriological discipline was still taking its first steps, Hanigalbat, today the commonly accepted Assyrian name for the Hurrian Mitanni1, was read as Hani-Rabbat (KUR Ha-ni-rab-bat) and taken to mean 'Great Hani'2. Already after the first stage of the decipherment of cuneiform was completed, Rawlinson (1857) read it as "Khani-Rabbi"3. Two of the earliest instances of "Khani-Rabbat" in the literature appear with Budge (1881: 3) and Sayce (1890). Winckler (1896: 38)4 read "Hanigalbat" and offered "Ha-ni-rab-bat" only as an alternative, but
* I wish to express my gratitude towards Ilya Yakubovich (Oxford University), Joshua Anderson (Oxford University), Seraina Nett (University of Copenhagen) and Ignacio Márquez Rowe (CSIC, Madrid), who read different drafts of this article and made various invaluable remarks, criticisms and suggestions. I am also thankful to Francisco Caramelo (New University of Lisbon) and Robert Whiting (University of Helsinki), as the writing of this article was ultimately possible due to their help in more practical matters. The responsibility for the views here upheld remains mine alone. The main abbreviations used in this article are CAD = Chicago Assyrian Dictionary and EA = Letters from Tell el-Amarna
1 That Mitanni and Hanigalbat are synonyms is proved by the fact that king Tusratta refers to himself as LUGAL KUR Mi-i-ta-a-an-ni (e.g. in EA 19: 03 and EA 25: iv 63), but also names his country as KUR Ha-ni-gal-bat (e.g. in EA 20: 17 and EA 29: 49) (see already Astour 1972: 103, with references). See also the detailed account in Wilhelm (1982: 34 and passim).
2 Hani-Rabbat/Hanigalbat was formerly thought to be located in Eastern Cappadocia, with the capital in Meliddu (the Roman Melitene, modern-day Malatya, in Turkey; see e.g. Sayce 1890).
3 "Khani-Rabbi" and "Khanni-Rabbi" appear already in Inscription of Tiglath Pileser I, King of Assyria, B.C. 1150, as translated by H. Rawlinson, Fox Talbot, Dr. Hincks and Dr. Oppert (1857), which gathered the four different translations of each of these scholars. Oppert, on the other hand, preferred the reading "Khanigalmit" (cf. also Oppert 1862: 317). Naturally, this was a time of variable trial readings.
4 Instances and respective readings as cited by Winckler (1896: 38): "Hanigalbat. Hanigalbatt gentilic. 1, 38. Hanigalbatü 15,22. 18,17. Ha-ni-gal-ba-ti 15 RS. 1. Ha-ni-gal-bat 21, 49. Ha-na-kal(rib)-bat 256, 10.20. od. Ha-ni-rab-bat cf. rab-bu-tí?"
Journal of Language Relationship • Вопросы языкового родства • 6 (2011) • Pp. 173-183 • © Valério M., 2011
he did not state the reasons for his preference. This is not the place for an exhaustive overview of the occurrences of different readings in past bibliography, but it is enough to say that in the late 19th and early 20th century the spelling Hani-Rabbat was frequently preferred, or at least mentioned as a possibility.
Later on this reading was fully replaced with KUR Ha-ni-gal-bat, which became universal in the Assyriological literature produced in the past one-hundred years. Independently, the communis opinio is that Hanigalbat is etymologically connected to a family of words including Hana (KUR Hana), a land of nomadic peoples in the Middle Euphrates, and Akkadian hanu and Hurrianizing Akkadian haniahhe, both roughly meaning 'Hanaean'. If so, Hani-galbat ought to be analyzed as a compound, but its second element remains frustratingly opaque. The attestation of the phrase URUKUR Ha-nu AN.TA, i. e. 'the city (cities) of the land of Upper Hana' in a letter retrieved at the Middle-Assyrian site of Tell Seh Hamad/Dur-Katlimmu (Rollig, 1997: 290-291) bears witness to a macrotoponym that was intended to be distinguished from the traditional Middle-Euphratean country of Hana. These data could apparently all be harmonized if one identified the 'Upper Hana' of the Middle-Assyrian text with the old reading of Hani-Rabbat 'Great Hani'. Intriguingly, however, any scholar drawing this conclusion will be struck by the disappearance of Hani-Rabbat from specialist literature. The motivation for such a revision was the optional spelling of this geographical name with the cuneiform sign GAL9 (Borger's sign no. 496, phonetically readable as kal), as opposed to the more standard orthography with GAL (Borger's sign no. 553, read phonetically as qal, kal, gal and rab in Akkadian and Sumerographically as GAL = rabum 'great').
The present essay intends to rehabilitate Hani-Rabbat as the accurate normalization of the Assyrian name of Mitanni, by showing the unmotivated nature of the alternative Hanigalbat as opposed to the more substantiated reading of GAL as rab in the spelling of this toponym. The following consists of a three-folded argument. I will begin by reviewing the evidence that Hani-galbat and Hana are etymologically connected, on the one hand, and correspond to a geographic opposition, on the other. Secondly, I will revisit early 20th-century historiography and the arguments that motivated the replacement of Hani-Rabbat. Thirdly, I will argue for reinstating the older reading based on inscriptional evidence and propose an etymology for it. Finally, I will devote some lines to discussing the possible identification of the reinstated Hani-Rabbat with the isolated occurrence of Hanu AN.TA in the Middle-Assyrian letter from Dur-Katlimmu.
Scholarship agrees in placing the country of Hana (KUR Hana), attested in inscriptional sources from the Old Babylonian period on, in the area of the Middle Euphrates and its confluence with the Habur (see Groneberg 1980: 88, Podany 1991-1993: 60, Rollig 1997: 290, n. 62, inter alia). In post-Old Babylonian times, the same land is definable as the territory to the north of Suhu and to the south of Mari (i.e. the Middle-Assyrian vassal kingdom with the capital in Tabetu).
Charpin argues that the term Hana originally denoted a human group rather than a physical territory per se. He suggests (Charpin 1995: 23) that the title "king of the land of Hana", used by the rulers of Mari and Terqa, intended to be a sign of their dominance over the "Bedouins" of the Middle Euphrates and Habur areas, whereas Rollig (1997: 291, n. 67) adds that in the Mari documents LU.Hanu refers to "nomadic population groups in the area of influence of the Mari kings". Whitting (1995: 1235) stresses the Amorite ethnicity of the Hanaeans, but duly points out that Hana "could also refer to any non-sedentary population and could be simply rendered 'nomad'". The itinerant character of these populations is perceptible from phrases such as nawum sa LU HA.NA.MES 'the camp of the Hana people' (ARM 3: 15, 11 apud CAD H: 82). The word in question is the Akkadian adjective hanu 'coming from Hana' (see CAD H: 82-83), i.e. 'Hanaean', often masked under Sumerographic appeareance.
CAD lists haniahhe (pl. haniahhena) as another related word. The term is analyzed as having been formed from Akkadian hanü plus the Hurrian adjective suffix -hhe5. The resulting Hurrian word was subsequently borrowed back into Akkadian and is attested in the Alalakh texts. The contexts of two of its instances hint at its semantics: 1-en LÚ sa KUR Mi-ta-an-ni ha-ni-a-ah-hi 'one [man] native of Mitanni, of the Hana class' (Wiseman 1953: 135, 12 apud CAD H: 82); and DUMU.MES ekudu haniahha attested in a "census list among sabe name people living outside of villages and towns" (Wiseman 1953: 143, 24 apud CAD H: 82). The first passage shows the haniahhena as groups of people also dwelling in the Mitannian territory, whereas the second example emphasizes their (at least) semi-nomadic character.
It is possible, however, that haniahhe is not derived from Akkadian hanü, but rather from the West-Semitic root *ny 'poor one'6, whose semantics agrees with the presumable (semi-)nomadic lifestyle of these populations. A similar proposal was already made by Green (1983: 192, n. 45). The merit of this etymology is that haniahhe is more easily derived from *ny than from Akkadian hanü. The former is plausibly vocalized as **aniy- > *haniy-, the Hurrian adaptation of ^ayin being unproblematically represented by <h> in cuneiform writing. In that case, this stem is a suitable base for haniahhe, since the Hurrian qualitative suffix -hhe is normally appended to words by means of a connective -o-/-a- that tends to elide stem-final -e/-i: cf. Hurrian ast(i)=o=hhe 'feminine' (< asti 'woman'), sin=z(i)=o=hh(e)=a 'second-ranking' (< sin=zi 'second'), etc. (see Giorgieri 2000: 208). Thus the final -y that was originally part of the West-Semitic root would regularly be deleted: *hani(i)=a=hhe 'poorish' < *haniy- 'poor one'7.
I believe that this etymology does not endanger the link implied above between the Hanaeans of Hana and the haniahhena of Mitanni. A compromise is possible if we also assume a West-Semitic source for Akkadian hanü and Hana itself. In this case, the root in question would be *nh, a variant of *ny (attested, e.g., in Old Aramaic, see DNWSI: 874; cf. also Ugaritic *nw 'to be/remain depressed, humble', Olmo Lete and Sanmartín 2003: 172)8, which in Akkadian would regularly yield the forms Hana and hanü. In summary, the entire family of words considered here may be Amorite borrowings in Akkadian and Hurrian. I will return to this issue below.
Another noteworthy point is that the country of Hana may be equated with the Astata of Hittite sources. The Treaty between Suppiluliuma I of Hatti and Sattiwaza of Mitanni (KBo 1.1, §10, A rev. 14-21) mentions a city of Tirga in the country of Astata. The equation between Tirga and the Terqa from Mesopotamian records is widely accepted9, and Astata is definable as a territory
5 The Hurrian suffix in question is -(h)he. Unlike -he, the variant -hhe is used mostly as a qualitative suffix (see Wegner 2007: 54).
6 See e.g. the entry ^ny3 in DNWSI: 877.
7 In addition to Hurro-Akkadian haniahhe, one also finds at Nuzi a number of personal names with the element Hani-: Hani-ku(ya), Hani-kuzzi, Hani-izza and Hani-Ashari. The last of these (attested in text EN 9/2 36) is formed with the divine name dAshar, mentioned in the seal impression of king Ithi-Tesup (see Albright 1940: 21 and Lacheman 1949: 53). Cf. also Hanuqala and Hanaqqala (Lacheman 1949: 50). I cannot, however, pursue here this path of research, and any etymological link between these personal names and our toponyms must remain in the realm of possibilities.
8 It is also remarkable that two related Old Testament passages of genealogical content (Gen. 36: 20-21; I Chr. 1: 38-39) mention a certain Anah (Hebrew ?anah), son of Seir, a Horite (Hebrew hori) (cf. already Green 1983: 192, n. 45). A son of Lotan, one of the brothers of Anah, goes also by the name of Hori. The connection between the biblical Horites and the historical Hurrians seems unlikely on geographical and etymological grounds. What is interesting to note is that these Horites (who are mentioned in connection with eponymous Anah) were a nomadic or semi-nomadic people that dwelled in a mountainous and semi-deserted area (Gen. 14: 6; Deut. 2: 12).
9 This equation is safe even from a linguistic perspective: Tirga and Terqa contrast only in the graphic oppositions between Hittite <i> and <g> and Akkadian <e> and <q>, respectively. Regarding the first opposition, one should note that the Late Assyrian name of Terqa is spelled Sirqu (cf. RIMA 2: A.0.100.5.90, 92, 93), adding consis-
that included Hittite Karkemis and Emar to its north and Terqa (the political head of Hana) to its south. Since Astata and Hana for the most part coincided10, I suggest that the former was the local non-Semitic name of the country. Although this cannot be proved for the time being, Astata could be a Hurrian designation borrowed into Hittite.
Rollig (1997: 290-291) considers the possibility that a second region could also have been known as Hana in both Old and Middle Babylonian times. He mentions the occurrence, in a Dur-Katlimmu letter (BATSH 4, n° 3), of a KUR Ha-a-ni located "between Nihria and siddi Haburi (the course of the Habur)", i. e. roughly the area formerly occupied by Mitanni (on the location of Nihria see below). Also according to Rollig, his non-Middle-Euphratean Hana is perhaps to be placed at the feet of the Kasiyari or Hasumu/Hasume mountains. In this regard, it is remarkable that, in the map produced by Cancik-Kirschbaum (1996: Abb 7)11 based on the toponymic evidence of the Dur-Katlimmu records, mat Hanu (i.e. the non-Middle-Euphratean Hana) and mat Hanigalbat almost overlap.
Regardless of whether these two geographical names are to be equated or not, the reference in the Alalakh texts to (semi-)nomadic Hanean populations in Mitanni provides a motive for the initial Hani- in the Assyrian name of the country. The old reading "Hani-Rabbat", interpreted as 'Great Hani (or Hana)', would imply an intent to differentiate the territory west of the Assyrian heartland and northeast of the Euphrates from the "traditional" Hana located in the west bank of the Middle Euphrates. The use of the river as a natural border even seems to be calqued in the frontiers established by Suppiluliuma I between Mitanni and Astata:
"All the cities which are situated in the land of Ashtata, on the west bank (of the Euphrates) of the land of Mittanni — Ekalte, [ . . . ] , Ahuna, and Terqa — these cities belong to the land of Ashtata. Since Prince Piyas-sili crossed the Euphrates with [Prince] Shattiwaza and penetrated to the city of Irrite, all the cities on the west bank which Piyassili, [my son], holds — these belong to Piyassili." (Treaty between Suppiluliuma I of Hatti and Sattiwaza of Mitanni, KBo 1.1, §10, A rev. 14-21, in Beckman 1996: 41)
The soundness of this interpretation now depends on our ability to confirm the accuracy of the reading Hani-Rabbat and demonstrate the incorrectness of its shifting to Hanigalbat.
The spectacular number of Assyriological scholarly works published in the last hundred and fifty years makes it difficult to detect with precision the roots of this radical replacement. One of the triggers seems to have been A. Olmstead's (1916) review of L. W. King's (1915) A History of Babylon. On page 284 we read:
"We may well ask, however, whether with King we should say Hani-rabbat, Hana the great, or whether we should not rather read Hani-Galbat, and connect it with the Galbatha which Isidore of Charax knew as a deserted village on the Euphrates four hours below Nicephorium. It is true that Galbatha is rather far north for the capital of Hana, Tirqa, which is to be located at 'Isharah. Curiously enough, the document which proves it here is from a certain Zim, the son of Ish, who rules as king of Mari; in other words, Hana is the later representative of Mari. That this is actually the site of the most important city in the middle Euphrates region in
tency to the Hittite vocalism. Furthermore, cuneiform Akkadian <q> reflects the Semitic emphatic consonant /k/ which in Hittite would be neutralized to /k/, first, and consequently voiced to /g/ in the vicinity of the sonorant /r/ (I. Yakubovich, pers. comm.). Indeed, Beckman (1996: 41) opts for translating Tirga as "Terqa".
10 Astour (1972: 106-107, citing Astour 1969) affirms that the "state of Astata, so called for its capital city, was formed by the Bedouin tribe of Rabbu, an eighteenth-century BCE group living on the western bank of the Euphrates". However, also according to Astour, the capital of these adversaries of Yahdun-Lim, king of Mari and Terqa (reigning over the Hanaeans), was called Abattum during this period. In any event, the Rabbu mentioned by Astour were a Hanaean tribe (Whitting 1995: 1238; see below).
11 This map is reproduced in Rollig (1997: 294) and in Tenu (2009: 352, map 6).
early times is further confirmed by a neat case of dovetailing. In apparent ignorance of the recent discovery of Herzfeld, Clay has identified Mari with the Merrha of the Parthian Stations of Isidore of Charax, and Isidore locates Merrha, as exactly as one can locate on the hour basis, at 'Isharah."
Apparently, Olmstead's intention was to etymologize "Hani-Galbat" by connecting the land of Hana in the Middle Euphrates with the Late Antique town of Galabatha (as it is spelled in Isidore of Charax), a Parthian station in the confluence of the Balih with the Euphrates, based solely on the graphic similitude between two place names. This probably predisposed him to write "Galbatha" instead of using the accurate spelling in the original source. The equation that this scholar proposed in his days is, of course, not valid in front of the data available to us today. Nonetheless, this idea became determinant in Olmstead's work, and in subsequent publications on different topics of the history of Mesopotamia (1917, 1918, 1920, 1922) he made consistent use of "Hani Galbat" or "Hani-Galbat" and put aside "Hani-Rabbat".
The turning point, however, was the publication of Die El-Amarna-Tafeln (1907-1915) by J. A. Knudtzon, a Norwegian scholar who, because of variant spellings with GAL9 / kal, favored the reading Ha-ni-kal-bat. This was even before the aforementioned works by Olmstead, who himself quoted from Knudtzon. Nevertheless, Die El-Amarna-Tafeln is still the standard edition in use today and has been one of the key documental sources for Assyriological and related research for the last one-hundred years. It would not be an overstatement to say that nearly all the literature on the political history of the Late Bronze Age Near East produced since then drew something from Knudtzon's work and, hence, it is mostly to his edition of the Amarna tablets that we owe the current popularity of Hanigalbat.
Fifteen years after the second volume of Knudtzon appeared, E. A. Speiser (1930: 95, n. 33), the influential Assyriologist who specialized in Mitanni and Hurrian themes, recorded the following comment on a footnote to his book Mesopotamian origins: the basic population of the Near East:
"That the name was Hanigalbat and not Hanirabbat (which Smith, EHA, 210 considers as a possibility) is proved by the writing (mat) Ha-ni-kal-bat in the Amarna Letters (Knudtzon, 255. 10), and by the form (mat) Ha-ni-kal-bat, HSS V. 63.3."
Speiser's argument and his subscription to Knudtzon's reading certainly played a role in perpetuating Hanigalbat, to which scholars vastly adhered in subsequent decades. One rare instance of maintenance of "Hani-Rabbat" is found in the work of F. Cornelius, who not only believed Hanigalbat to be an incorrect form (1967: 306) but also defended that the alternative "Gross-Hani" was "im Gegensatz zu dem Hani der Mari-Urkunden" (1958: 2). His opinion, however, was not followed.
Thus Hani-Rabbat, a relatively transparent Semitic compound, ended up replaced by the obscure Hanigalbat whose second element has defied analysis through the decades. The socio-linguistic situation of the Syro-Mesopotamian area in the Late Bronze Age is still rather obscure, and numerous geographical names could belong to unknown substrata, but one is nevertheless struck by the opacity of -galbat. On the one hand, the ending -at is suspiciously reminiscent of the Semitic feminine ending. On the other hand, if one attempts connecting the expected meaning of 'great' with galbat, Sumerian incongruously comes to mind, as if it echoed a partially Sumerographic spelling of an Akkadian word. I suggest that the Assyrian name of Mitanni is to be read as Ha-ni-rab-bat, wherein the sign GAL could a priori be used logographi-cally or phonetically (with the value rab). Below I will argue for the latter option. But first, other problems need to be resolved.
The first obstacle to be tackled is Speiser's 1930 argument. This scholar presented as evidence variant spellings which apparently corroborate the phonetic reading of Hanigalbat. His
examples refer to the spelling of Ha-ni-kal-bat, or rather Ha-ni-galg-bat, known from at least six documents:
ha-ni-galg-bat-i (EA 1: 38; sent by Amenhotep III to Kadasman-Enlil I) ha-na-galg-bat (EA 255: 10; provenance unknown)
ha-ni-galg-bat (KBo X 1 obv. 11, the Akkadian version of the Acts of Hattusili I) ha-ni-galg-bat (KBo I 14, letter sent by Hattusili III to Adad-nerari I) ha-ni-galg-bat (HSS V 63. 3; SMN12 2065:7, at Nuzi)
All these instances are found in texts written in Peripheral Akkadian, as shown by their provenance: two from Nuzi, two from Hittite Anatolia, one from Egypt, and one from an unknown location (possibly Syria). They stand alone in an ocean of spellings with GAL. In Nuzi alone, GAL9 is used rather exceptionally: Lacheman (1940) contrasts seven instances of Ha-ni/e-GAL-bat and thirteen of Ha-ni-GAL-bat with only two examples of Ha-ni-GAL9-bat. In the Amarna letter EA 255, ha-ni-galg-bat even coexists with standard ha-ni-gal-bat.
I consequently propose that the six cases above consist of misspellings by foreign scribes. Mitanni was known under that name in Anatolia, Egypt and in the Hurrian land itself, and it is possible that the authors of at least four of these texts never actually heard its Semitic name. It is remarkable that the signs GAL and GAL9, albeit graphically distinct, have much in common with respect to their readings:
Sign Value
Logographic Phonetic
GAL9 = dannum 'strong' gal9, lablp, rib/p (Akkadian) kal, dan, tan (Hittite)
Mh GAL = rabûm 'great' gal, kal, qal, rab
As we can see, equivalences and similarities between the various possible values (marked in bold in the table) of these otherwise different signs might have motivated some confusion in their use, especially for scribes that used Akkadian as a written, not spoken language. For frequent errors in the spelling of Akkadian at Nuzi, see the survey by Berkooz already in 1937.
The first piece of evidence in favor of this view is the lack of spellings with **-ga-al- that would substantiate the phonetization as Hanigalbat. Apart from the above examples with GAL9, all instances of this place name and derivatives of it known to me are consistently spelled with GAL:
ha-na-gal-bat (EA 255: 20; provenance unknown)
ha-ne-gal<-ba>-tum, adj. 'native of Hanigalbat (technical term for a member of the chariot team)'(equated to lu.ki.zu.u and tas-li-su in lexical series HAR.gud = imru = ballu, B.VI.145, cited in CAD H: 80)
ha-ni-gal-bat (RIMA 1: A.0.77.1.58, 60; RIMA 2: A.0.87.1.34; A.0.99.2.39, 42, 45, etc.; A.0.101.1.22, inter alia)
12 SMN is used in specialist literature for unpublished Nuzi texts in the Semitic Museum of Harvard University.
ha-ni-gal-ba-tu-u (EA 16: 22) '[king of] Hanigalbat' ha-ni-in-gal-bat (HSS 15:5)
ha-ni-gal-ba-tu-ti-su, MB Alalakh s. '[his] citizenship of, or social status as a native of, Hanigalbat' (Wiseman 1953: 13, 4, apud CAD H: 80)
Of course, this is also true of the reading rab, for which we lack a corroborating **-ra-ab-. But it is at least possible that gal is not the correct reading. If one sets out to explore the alternative, it must be decided whether GAL is being used Sumerographically (in representation of the Akkadian indeclinable form rab) or if it directly notates phonetic rab. The answer to this question is intrinsically connected to the etymology of the word underlying GAL-bat.
The first possibility is suggested by a number of spellings in Akkadian where GAL is used logographically to notate words that are cognate to rabum 'great'. These include GAL-bu-ti = rabuti (CAD R: 16, with ref. to CT 17: 4.i.9ff) and GAL-bu-te.MES = rabute 'officials, dignitaries' (EA 164: 33 mentioned in CAD, R: 36). More relevant is the occurrence of GAL-bat-dGu-la (BE 15: 188, v.22) vs. Ra-bat-dGu-la (BE 15: 185, v.22) in Middle Babylonian records from the Temple of Nippur at Babylon. This last form could imply that an indeclinable rabat underlies the second part of Ha-ni-GAL-bat. This uncommon Akkadian form is nevertheless regularly employed in personal-names or references to deities: cf. the abovementioned pair GAL-bat-dGu-la ~ Ra-bat-dGu-la 'Gula, the Great' and fIna-Akkadi-ra-bat 'the Great (one) in Akkad' from the same Middle Babylonian source (BE 14 / Clay 1906: 55). These are examples of predicative constructions with a verbal adjective base, which are referred to as the stative (see Huehnergard 2005: 222). In this case, rabat is a feminine indeclinable form of the verb rabu with its secondary meaning of 'to become great, superior', and, accordingly, one finds it glossed under that same lemma in CAD (R: 41-42). The original verbal adjective base for rabu would be *rabi?-, which would yield the 3rd fem. sg. stative in rabPat. Presumably, in Post-Old-Babylonian times (after the loss of the intervocalic aleph had already taken place), vowel contraction would have occurred, resulting in rabat (< rabiat) — even though sequences of i and long or short a are expected to remain uncontracted (see Huehnergard 2005: 24 and 39). That GAL-bat, ra-bat, ra-ba-at and ra-ba-at are all possible spellings of rabat, with the long vowel, is ascertained (see CAD R: 38-39, 42). But it remains a disadvantage of this explanation that, as we have seen, Hani-GAL-bat is exclusively spelled with the sign bat, and no variants with -ba-at are attested.
The second hypothesis is one in which GAL is phoneticized as rab, producing the reading Ha-ni-rab-bat = Hani-Rabbat. In this scenario, the Akkadian solution is to be excluded in favor of a West Semitic, i. e. Amorite, solution. Accordingly, West Semitic *rabbat would be the indeclinable absolute form (or status absolutus) of the adjective *rabbatu, the regular feminine of *rabbu, in obvious contrast to Akkadian rabi (m.) / rabitu (f.)13. The shape in question is well-attested in the Middle-Euphratean area: in Mari texts, the numeral rabbatu '10.000' (< *'large [amount]') appears in Akkadian as a West Semitic loanword (CAD, R: 14-15)14; and, blended in Akkadian texts from Emar, where a West Semitic dialect was also spoken, one finds the substantivized feminine adjective NINDAra-ba-tu4 /rabbatu/, 'large; a kind of bread' (Pentiuc 2001: 152); for the masculine form, cf. the Hanaean tribe's name Rab(a)bu (Whitting 1995: 1238)15.
13 Thanks to Ilya Yakubovich, who suggested this solution (pers. comm.).
14 The variant ribbatu (with the indeclinable form ribbat) for this Mariot numeral is also attested (CAD, R: 314). Its existence renders theoretically possible the reading of Ha-ni-gak-bat as Ha-ni-rib-bat. However, it is not likely that Hittites, Egyptians and Hurrians would make use of a marginal form of a toponym that is otherwise unat-tested in Semitic sources.
15 See also footnote 10 of this article.
Moreover, the use of absolute forms of adjectives in Amorite personal names, attested in texts from Mari, was a common practice. The greatest merit of this solution is that it removes the problem of the inconsistent spelling of a presumable long d raised by the Akkadian explanation. But it also makes more sense in historical terms, since ethnically the Hanaeans were themselves Amorite (Whitting 1995: 1235), at least originally. That is, Hani-Rabbat could have been the autoethnonym of West Semitic-speaking Hanaeans. The distinction between the traditional Hana and the Great Hani (see map 1), across the Euphrates, would then be an old one, possibly dating back to, at least, the 18th century BCE. In the light of the West Semitic etymology proposed for Hana itself (see above), I underline the coherence of the present analysis, which renders the compound Hani-Rabbat fully Amorite16.
Map. 1. Geographic opposition between Hana and Hani-Rabbat and main sites mentioned in the text.
Finally, we come to the issue of whether the hapax KUR Ha-nu AN.TA is an alternative Sumerographic strategy for spelling the recurrent KUR Ha-ni-rab-bat or whether we are dealing with two distinct geographical names.
In the Middle-Assyrian text DeZ 3281 from Dur-Katlimmu, URUKUR Ha-nu AN.TA 'city (cities) of the land of Upper Hanu' appears after URUNi-ih-ri-a (Nihria) and before URUHu-um-na-hu-sa. Assyriologists generally agree that Nihriya is to be found to the north of the Assyrian heartland, and Rollig (1997: 290, with references), who has published and studied DeZ 3281 in detail, argues for a location "not too far from Harran". This latter scholar even cites a suggestion by J. Mellaart: "the copious springs at Urfa could very well elicit a [Semitic] town name Nihria". With regard to Humnahusa, Rollig (1997: 291, citing Fincke 1993: 103f) cannot propose a setting for the city, but duly notes that initial Hum- hints at a Hurrian origin of the toponym, if one compares the place names Humella and Humpurse in Nuzi texts. This would place Humnahusa somewhere in Hurrian land. In the end, not only are the locations of both cities unclear, but it is also not guaranteed that the collocation of place names in a given text will reflect their actual geographical position. Whatever the case may be, this insufficient evi-
16 The presumed Amorite opposition of Hana and Hani-Rabbat is somewhat echoed in the Canaanite place-names Arad of Beth Yeroham ('Arad of the House of Yeroham'; cf. "the South [Negev] of the Jerahmeelites" and "the cities of the Jerahmeelites" in I Sam. 27:10 and 30:29) and Arad-Rabbat, attested in the list of cities (Megiddo stela) conquered by Pharaoh Sheshonq I in Canaan, dating to the second half of the 10th century BCE (see Roth 1972: 244, 250) (I. Yakubovich, pers. comm.).
dence does not impede — in reality, it allows — the proposed equation of Ha-nu AN.TA with KUR Hani-Rabbat. The fact that, out of the entire list of place names contained in DeZ 3281, the former is the only one preceded by the determinative KUR indicates that the scribe meant a group of cities from an entire country, not an individual settlement. Therefore, I propose, as a working hypothesis, that KUR Ha-nu AN.TA in this letter is an exceptional spelling for the then Middle-Assyrian pahatu of Hani-Rabbat, whose political center was precisely Dur-Katlimmu.
I hope that I have provided fellow Assyriologists with sufficient arguments to consider rehabilitating the old form Hani-Rabbat as the Semitic name of the Hurrian land. From a historical viewpoint, the present interpretation throws new light on Mesopotamian geographical conceptions. Once more in Near Eastern history, emphasis is put on the role of the river Euphrates as a natural frontier and psychological barrier, in this particular case for being the raison d'être of the dichotomy Hana vs. Great Hani / Upper Hanu.
References
Albright, William F. 1940. New Light on the History of W. Asia in the 2nd Millennium B.C., BASOR 77, 20-31.
ARM 3 = Kupper, Jean Robert. 1950. Correspondance de Kibri-Dagan, gouverneur de Terqa. Archives royales de Mari 3. Paris: Paul Geuthner.
Astour, Michael C. 1969. The Partition of the Confederacy of Mukis-Nuhasse-Nii by Suppiluliuma: a study in political geography of the Amarna age. Orientalia N.S. 38, 381-414
Astour, Michael C. 1972. Hattusilis, Halab and Hanigalbat. Journal of Near Eastern Studies, 31/2, 102-109.
BE = Hilprecht, V. (ed.). 1893-1914. The Babylonian expedition of the University of Pennsylvania. Series A: Cuneiform texts. Philadelphia: Department of Archaeology, University of Pennsylvania.
BE 14 = Clay, Albert T. 1906. The Babylonian Expedition of The University of Pennsylvania. Series A: Cuneiform Texts — vol. XIV: Documents from the Temple Archives of Nippur — Dated in the Reigns of Cassite Rulers. Philadelphia: Department of Archaeology, University of Pennsylvania.
Beckman, Gary. 1996. Hittite Diplomatic Texts. Atlanta: Scholars Press.
Belck, Waldemar. 1897. Hanigalbat und Mitilene. Zeitschrift der Deutschen Morgenländischen Gesellschaft (ZDMG) 51, 855-868.
Berkooz, Moshe. 1937. The Nuzi Dialect of Akkadian: Orthography and Phonology. Language 13/1 (Language Dissertation No. 23), 5-64.
Borger, Rykle. 2003. Mesopotamisches Zeichenlexikon, Münster.
Budge, Sir Ernest A. T. W. 1881. The History of Esarhaddon (son of Sennacherib), King of Assyria, B.C. 681-668, translated from the Cuneiform Inscriptions upon Cylinders and Tablets in the British Museum Collection together with the Original Texts, a grammatical analysis of each word, explanations of the ideographs by extracts from the bilingual syllabaries and list of eponyms, &c. Boston: J. R. Osgood & Co.
Cancik-Kirschbaum, Eva. 1996. Die mittelassyrische Briefe aus Tall Seh Hamad, BATSH 4. Berlin: Dietrich Reimer Verlag.
Charpin, Dominique. 1995. Ä propos des rois de Hana. N.A.B.U. 23.
Cornelius, Friedrich. 1958. Zur Hethitischen Geographie: Die Nachbarn des Hethiterreiches" Revue Hittite et Asianique 16, 1-17.
Cornelius, Friedrich. 1967. Geistesgeschichte der Frühzeit. II. Teil Band 2 — Der Aufstieg der Indogermanen und der Orient bis zum Untergang des Hethiterreiches. Leiden: Brill.
CT 17 = Cuneiform Texts from Babylonian Tablets, &c., in the British Museum. Part XVII. London: Trustees of the British Museum, 1903.
DNWSI = Hoftijzer, J. and Jongeling, K. 1995. Dictionary of the North-West Semitic Inscriptions. With Appendices by R. C. Steiner, A. Mosak Moshavi and B. Porten. Part Two: M-T. Leiden / New York / Köln: Brill.
EHA = Smith, Sidney. 1927. Early History of Assyria to 1000 B.C. New York: E.P. Dutton and Co.
Giorgieri, Mauro. 2000. Schizzo grammatical della lingua hurrica. In: S. de Martino, M. Giorgieri, N. Par-megiani, P.E. Pecorella, M. Salvini, M.C. Trémouille, La civiltà dei Hurriti (= La parola del Passato 55), Naples: Macchiaroli Editore, 171-277.
Green, Alberto R. W. 1983. Social Stratification and Cultural Continuity at Alalakh. In: H. B. Huffmon, F. A. Spina, and A. R. W. Green, The Quest for the Kingdom of God. Studies in Honor of George E. Mendenhall. Winona Lake: Eisenbrauns, 181-204.
Groneberg, Brigitte. 1980. Die Orts- und Gewässernamen der altbabylonischen Zeit unter Benutzung der sammlungen von J.R. Kupper und W.F. Leemans, Wiesbaden, RGTC 3.
HSS 15 = 1955. Excavations at Nuzi. Vol. VI: The Administrative Archives (Harvard Semitic Series, 15). Cambridge: Harvard University Press.
Huehnergard, John. 2005. A Grammar of Akkadian. Second Edition. Harvard Semitic Museum Studies 45, Winona Lake, Indiana: Eisenbrauns.
Knudtzon, Jorgen A. 1915 Die el-Amarna-Tafeln. Anmerkungen und Register bearbeitet von C. Weber und E. Ebe-ling. (Vorderasiatische Bibliothek, 2.) Leipzig. (First volume: 1907)
Lacheman, Ernest R. 1940. Nuzi Geographical Names I: Names of Countries, BASOR 78 , 18-23.
Lacheman, Ernest R. 1949. Nuzi Personal Names: Review Article, Journal of Near Eastern Studies 8/1, 48-55.
Olmo Lete, Gregorio del and Joaquin Sanmartín. 2003. A Dictionary of the Ugaritic Language in the Alphabetic Tradition (transl. W. G. E. Watson). HdO 1:67, Leiden: Brill.
Olmstead, A. T. 1916. Review: A History of Babylon, by L. W. King (1915). The American Journal of Theology 20/2, 277286.
Olmstead, A. T. 1917. Tiglath-Pileser I and His Wars. Journal of the American Oriental Society, 37, 169-185
Olmstead, A. T. 1918. The Calculated Frightfulness of Ashur Nasir Apal. Journal of the American Oriental Society, 38, 209-263.
Olmstead, A. T. 1920. Kashshites, Assyrians, and the Balance of Power. The American Journal of Semitic Languages and Literatures, 36/2, 120-153.
Olmstead, A. T. 1922. The Fall and Rise of Babylon. The American Journal of Semitic Languages and Literatures, 38/2, 73-96.
Oppert, Julius. 1862. Expédition scientifique en Mesopotamie exécutée par ordre du gouvernement de 1851 à 1854 MM. Fulgence Fresnel, Félix Thomas et Jules Oppert. Vol. 1 Relation du voyage et résultats de l'expédition. Paris: Imprimerie Impériale.
Pentiuc, Eugen J. 2001. West-Semitic vocabulary in the Akkadian Texts from Emar (Harvard Semitic Studies 49). Winona Lake: Harvard Semitic Museum / Eisenbrauns.
Rawlinson, Henry Sir; Talbot, Fox; Hincks, Edward; Oppert, Julius. 1857. Inscription of Tiglath Pileser I, King of Assyria, B.C. 1150, as translated by H. Rawlinson, Fox Talbot, Dr. Hincks and Dr. Oppert (Published by the Royal Asiatic Society). London: J. W. Parker and Son.
RIMA 1 = Grayson, A. Kirk. 1987. The Royal Inscriptions of the Mesopotamia Assyrian Periods. Vol. I. Assyrian Rulers of the Third and Second Millennia BC (to 1115 BC). Toronto: University of Toronto Press.
RIMA 2 = Grayson, A. Kirk. 1991. The Royal Inscriptions of the Mesopotamia Assyrian Periods. Vol. 2. Assyrian Rulers of the Early First Millennium BC (1114-859 BC). Toronto: University of Toronto Press.
Röllig, Wolfgang. 1997. Aspects of the Historical Geography of Northeastern Syria from Middle Assyrian to Neo-Assyrian Times. In: S. Parpola & R.M. Whiting (eds.), Assyria 1995, Helsinki: Neo-Assyrian Text Corpus Project, 281-293.
Roth, Cecil (ed.). 1972. Encyclopedia Judaica, vol. 3. Jerusalem: Keter Publishing House.
Sayce, Archibald H. (ed.). 1890. Records of the Past: Being English Translations of the Ancient Monuments of Egypt and Western Asia. Vol. 3. London: S. Bagster and Sons.
Speiser, Ephraim A. 1930. Mesopotamian Origins: the Basic Population of the Near East. Philadelphia: University of Pennsylvania Press.
Tenu, Aline. 2009. L'expansion médio-assyrienne. Approche archéologique. BAR International Series 1906. Oxford: Ar-chaeopress.
Wegner, Ilse. 2007. Hurritisch — Eine Einführung. 2. überarbeitete Auflage. Wiesbaden: Harrassowitz Verlag.
Whitting, Robert M. 1995. Amorite Tribes and Nations of Second-Millennium Western Asia. In J. M. Sasson, ed., Civilizations of the Ancient Near East, vol. 2. New York: Charles Scribners Sons, 1231-1242.
Wilhelm, Gernot. 1982. Grundzüge der Geschichte und Kultur der Hurriter. Darmstadt: Wissenschaftliche Buchgesellschaft.
Winckler, Hugo. 1896. Die Thontafeln von el-Amarna (Keilinschriftliche Bibliothek V). Berlin: Reuther & Reichard.
Wiseman, Donald J. 1953. The Alalakh Tablets. London: British Institute of Archaeology at Ankara.
В статье отстаивается чтение семитского наименования государства Митанни как Хани-Раббат, в противоположность общепринятому Ханигальбат. В этой связи предлагается этимология данного топонима, чье происхождение не было до сегодняшнего дня прояснено. Согласно гипотезе автора, Хани-Раббат — это западносемитский (аморейский) композит со значением «великая Хани», противопоставленный области Хана в среднем течении Евфрата. Таким образом раббат «великая» маркирует противопоставление между двумя хананейскими землями, отделенными друг от друга Евфратом. Хани / Хана оказывается также связанным с аккадским hanü и (заимствованным) хурритским haniahhe, использовавшимися для обозначения (полу)кочевых западносемитских племен по обе стороны Евфрата. Предлагаемый сценарий опирается на гипотезу лингвистических контактов между аморейским, аккадским и хурритским языками. Наконец, топоним Hanu AN.TA «верхняя Хана», засвидетельствованный в одном среднеассирий-ском письме, рассматривается как возможный кандидат на шумерографический эквивалент Хани-Раббат.
Ключевые слова: Хани-Раббат, Ханигальбат, Митанни, Хана, аморейский язык, западно-семитские языки, среднеассирийский диалект, аккадский язык, Средний Евфрат, кочевники.
