The onomastic evidence for Bronze-Age West Semitic.

• Author: O'Connor, M.

1. BRONZE-AGE WEST SEMITIC

   The history of the West Semitic family begins in earnest in the late second millennium, more or less around the start of the Syro-Palestinian archeological period known as the Iron Age. (1) It is around 1200 B.C.E. that the Canaanite branch of the family (most importantly, Hebrew and Phoenician) begins to be documented in the epigraphic record, and not long thereafter are dated the earliest Aramaic and Old South Arabian texts; the Arabic and Ethiopic branches follow during the first millennium C.E. In all these cases, except Old South Arabian, archeologically recovered texts show some degree of correlation with traditions preserved in manuscript form. Prior to 1200 B.C.E. West Semitic presents a different picture: all the sources are archeologically recovered, and none of them shows a straightforward relation to literary traditions. The linguistic evaluation of the remains of Bronze-Age West Semitic (BAWS) is thus an area of intensive research and contentious discussion. Off in a quiet corner by themselves sit several heaps of onomastic evidence: they are assumed to belong somewhere and in some way. The evaluation of them as evidence for Bronze-Age West Semitic is not wildly controversial, but neither is it settled.

   The primary sources from which these several heaps are drawn are the publications, some of them over a century old, of texts excavated or extracted from Old and Middle Babylonian sites in Western Asia and of synchronous texts from Egypt. The pride of place here must go to the series Archives royales de Mari, a set produced over the span of half a century (so far) by French and Francophone scholars. The heaps themselves are variously gathered up in major publications. The most imposing is Assyriological Studies 21, surely the largest single volume regularly used by any West Semitist, one of I. J. Gelb's several late masterworks. Situated close to AS 21 are Buccellati's dissertation on Ur III Amorites, Huffmon's dissertation on Mari Amorite, Zadok's contribution to the Hallo festschrift, and Streck's Habilitation, along with various other studies of so-called Amorite material. (2) Although the term Amorite is generally reserved for names from the Old Babylonian and earlier periods (roughly, the first half of the second millennium), there is no distinct break between such names and names from the Late Bronze Age (roughly, the second half of the second millennium). Other gatherings of BAWS names are Grondahl's dissertation on the names from Ugarit, Hess's dissertation on the Amarna names, Sivan's dissertation on Late-Bronze West Semitic vocabulary, Pruzsinszky's dissertation on Emar names, and various other works. (3) The heaps betray devoted labor and careful attention, and yet the evaluation of the material in them remains incomplete.

   In order to pursue this evaluation it will be useful to review the entire range of materials relevant to Bronze-Age West Semitic. The bulk of the evidence is Late Bronze or Middle Babylonian, although there is ample material from the Old Babylonian period and some from the Ur III period (late third millennium). (4) The largest single body of evidence comes from the Late Bronze site of Ugarit. (5) The evidence comprises texts written in the Ugaritic alphabetic script, as well as texts written in Mesopotamian cuneiform, primarily in a form of Middle Babylonian, and some texts written in other writing systems. (6) The alphabetic texts are of various sorts, including poetic texts and ritual texts that include poetic material, as well as letters and administrative documents. It is the working assumption of students of Ugaritic that the corpus of alphabetic material represents more or less a single language, although there are disparities across the corpus.

   The study of Ugaritic alphabetic texts is aided by some cuneiform material, and the pattern of the cuneiform contribution to Ugaritic study is one that is repeated over and over for Bronze-Age West Semitic. In cuneiform texts from Mari, Amarna, Ugarit, Emar, and other Bronze Age sites, we find isolated data relevant to local West Semitic languages. (7) The data include (a) loan words into Akkadian, some perhaps of long standing and some nonce loans, made for a particular occasion and leaving no enduring trade in the language; (b) glosses and other recognizably alien material provided in the Akkadian for the sake of intelligibility; and (c) substrate influences on morphology and syntax. The distinction between (a) and (b) is not always easily made. (8) The loans and some of the glosses are recorded as words of Akkadian by the Akkadian dictionaries, and in the case of the Amarna material many of them are also recorded in Hoftijzer and Jongeling's dictionary of Northwest Semitic. Texts in Egyptian and Hieroglyphic Luwian yield a very modest crop of similar material. (9)

   These data are linguistically isolated on the West Semitic side, but not, it is important to note, on the side of Akkadian or Egyptian. They usually have a linguistic context provided by the language material they occur in. Sometimes that linguistic context is straightforward, as when a West Semitic gloss is provided after an Akkadian word in the Amarna texts. Generally the Akkadian word is known and the gloss suggests that the West Semitic-speaking scribe used this West Semitic word in his own language to correspond to the Akkadian word. In the area of morphology and syntax the linguistic context is harder to grasp. Morphological deviations and adaptations can reveal different degrees of distance from an Akkadian target: sometimes a purely West Semitic form is used, while in other cases a form that is not quite Akkadian but not really West Semitic is used. It is in the morphology that the West Semitic scribes' imperfect grasp of Akkadian shows up most plainly. In the syntax, the verb-final word order of Akkadian is often replaced by verb-initial or V2 West Semitic orders. Similarly, an Akkadian dependent clause usually precedes the clause it depends on, while in these texts a dependent clause may follow its main clause. The most elegant form of linguistic context for this West Semitic material is perhaps that furnished by the lexical texts from Ugarit, where words of Ugaritic are lined up with words of Sumerian, Akkadian, and Hurrian in a quadrilingual format. (10) Despite the enormous bulk of the Old and Middle Babylonian corpora in question, the data they provide remain largely isolated on the West Semitic side. Despite all the information furnished by these texts, there is no continuous text in West Semitic.

   Another facet of the isolation is geographical. The various texts derive from different sites and in many cases the texts were not found where they were written. Texts written at Mari and found at Mari can be presumed to belong together, but texts written at Terqa or Aleppo and found at Mari need to be kept separate from Mari texts proper and from each other. In the case of the Amarna material subcorpora have regularly been studied in terms of their city of origin, but the principle is valid for all relevant texts, even though it often cannot be carried through, since many of the texts cannot be assigned a city of origin. (11)

   There are, moreover, difficulties in the interpretation of the writing systems: cuneiform and hieroglyphics, both Egyptian and Luwian (Hittite), were not designed or ever fully adapted to record West Semitic languages, so at every juncture there are uncertainties in interpreting the ways the elements of the writing systems are being used. The questions of interpretation are both global and local, i.e., we must consider not only what a particular writing reveals but also how systematic the patterns behind the writing are. (12)

   Thus far we have two major bodies of Bronze-Age West Semitic: the Ugaritic alphabetic texts and the isolated language material found in cuneiform and Egyptian texts. There is a third body of Bronze-Age West Semitic, the onomastic evidence. It is scattered across an enormous number of texts, most of them documentary, i.e., legal, economic, and administrative texts. The major areas from which these texts derive, going roughly south to north and east to west, are listed below:

   (1) 1. Mesopotamia proper, including Ur, Uruk, Larsa, Isin, and Babylon; AS 21

2. the Diyala region, including Esnunna (Tell Asmar); AS 21 (13)

3. Assyria; AS 21

4. Mari; the volumes of ARM, especially ARM XVI/1, APN, AS 21

5. Emar; SCCNH 13 (14)

6. the Syro-Palestinian cities of the Egyptian Empire under the Eighteenth Dynasty; names in Hess, Amarna Personal Names

7. Ugarit; names in Grondahl, Die Personnenamen der Texte aus Ugarit (15)

   The Egyptian names from both the Middle and New Kingdom sources tend to refer to Syria-Palestine. (16) It is conventional to refer to the earlier names, those in groups 1-4 (and the corresponding Egyptian names, those in the Middle Kingdom execration texts), as Amorite. Some scholars also apply that term to later Bronze-Age material. (17) For the later names, most refer simply to Emar, Amarna, Ugarit(ian), etc. names. The materials from groups 1-4 are largely Old Babylonian, the others Middle Babylonian. There are smaller sites that yield relevant names. Southern Syro-Palestinian sites have themselves yielded only a small number of names. (18)

   Various features of names and name-giving are crucial to the study of Semitic onomastica. (19) These include ancient awareness of names ([section]2), linguistic transparency ([section]3), and name-giving semantics ([section]4). These topics are all well known, but a closer look at them will reveal some important features of the BAWS onomastic material. (20) After reviewing them, I will discuss the question of what languages are behind the names ([section]5).

2. ANCIENT AWARENESS OF NAMES

   Ancient awareness of the character of Semitic names and name-giving manifests itself in various ways. The best known is the literary use of names...