اللغة الكنعانية

20
http://www.jstor.org Review: A Grammar of Amarna Canaanite Author(s): John Huehnergard Reviewed work(s): Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed Dialect Used by the Scribes from Canaan by Anson F. Rainey Source: Bulletin of the American Schools of Oriental Research, No. 310 (May, 1998), pp. 59-77 Published by: The American Schools of Oriental Research Stable URL: http://www.jstor.org/stable/1357578 Accessed: 24/08/2008 07:17 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=asor. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

description

Grammar of Amarna Canaanite

Transcript of اللغة الكنعانية

  • http://www.jstor.org

    Review: A Grammar of Amarna CanaaniteAuthor(s): John HuehnergardReviewed work(s): Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed DialectUsed by the Scribes from Canaan by Anson F. RaineySource: Bulletin of the American Schools of Oriental Research, No. 310 (May, 1998), pp. 59-77Published by: The American Schools of Oriental ResearchStable URL: http://www.jstor.org/stable/1357578Accessed: 24/08/2008 07:17

    Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

    Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=asor.

    Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

    JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with thescholarly community to preserve their work and the materials they rely upon, and to build a common research platform thatpromotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

  • Review Article

    A Grammar of Amarna Canaanite

    JOHN HUEHNERGARD Department of Near Eastern Languages and Civilizations

    Harvard University 6 Divinity Avenue

    Cambridge, MA 02138 huehnerg @ fas.harvard.edu

    Canaanite in the Amarna Tablets: A Linguistic Analysis of the Mixed Dialect Used by the Scribes from Canaan, by Anson F. Rainey. 4 vol- umes. Handbuch der Orientalistik I/XXV. Leiden: E. J. Brill, 1996. NIg. 604.00/U.S. $390.00.

    For nearly three decades, Anson Rainey has, with W. L. Moran, led the way in elucidating and ex- plaining the curious grammar of the Amarna letters sent to Egypt in the Late Bronze Age by Canaanite vassals. After dozens of insightful articles, Rainey has now given us what may truly be called, in sev- eral respects, a magnum opus on the subject. Finally we have a comprehensive grammatical description of this difficult corpus.

    The four volumes of this set are I, Orthography, Phonology; Morphosyntactic Analysis of the Pro- nouns, Nouns, Numerals (xxi + 204 pp.); II, Morpho- syntactic Analysis of the Verbal System (xx + 415 pp.); III, Morphosyntactic Analysis of the Particles and Adverbs (xiv + 280 pp.); and IV, References and Index of Texts Cited (viii + 198 pp.).

    The first three volumes are to be seen as separate, but companion, sets of essays. The comprehensive citation index in volume IV allows the reader to cross-check the separate volumes, which are not otherwise cross-referenced to one another. Indeed, each of the first three volumes, which were written at different times-volume III first (in 1984) and volume I last-is essentially self contained; there is little coordination of their contents, as Rainey him- self notes at several points in the prefaces to the individual volumes. This has resulted in a consider- able amount of repetition across the volumes; e.g., much of the material in I 38-43, on "Phonological Features," appears again, verbatim or almost verba- tim, in II 19-24 on "The Akkadian Base Language"; the lack of the normative Akkadian subordinating

    morpheme -u (and of Assyrian -ni) is covered in detail in II 195-202 ("The so-called 'Subjunctive"') and again in III 67-71 (under the subordinating conjunctions); similarly the material in I 37-38 is repeated in II 375-76 (on ipesu as the infinitive of verbs I-e); II 211-12 is the same as III 195-96 (on lu); the presentation particle amur is discussed both in II 274-76 (under the imperative) and in III 167- 74 (under presentation particles). Even within indi- vidual volumes material is repeated, for example II 307-8 and 376-77 (on na-aq-sa-pu in EA 82:51); III 10 is the same as III 127-28 (on tumal). Rainey's desire that the individual volumes be intelligible on their own is laudable but since it is not possible to buy the volumes individually, the desire seems mis- placed; a good number of pages could have been reduced to a simple cross-reference. Doing so might have reduced the enormous cost of the volumes, which will be out of the reach of a great many stu- dents and scholars; and that is most unfortunate for such an important and useful reference work.

    Rainey does not intend to present a full grammar of the Amarna texts. As his title indicates, his con- cern is with the Canaanite elements embedded in them (e.g., II 17), the elements that have made these letters so interesting and valuable to students of early Northwest Semitic (NWS) since their first publica- tion. Even here, however, one might have wished for coverage of additional features in which these texts differ from normative Akkadian and seem to provide us with information about early NWS. But Rainey himself is aware of this problem. He states, "One could hardly define the present work as a conven- tional grammar. It is, rather, an essay, or more pre- cisely, a series of essays, on the language" (I xiii).

    One of the most helpful features of Rainey's pre- sentation is the extensive use of full citations to

    59

  • JOHN HUEHNERGARD

    illustrate nearly all points of grammar covered (cf. I xv-xvi), for which he deserves the thanks of all students. Indeed, he cites so many examples that col- lating them would almost provide us with a complete new transliteration of the Canaanite Amarna tablets, something that is very much needed.

    Not surprisingly, Rainey's detailed investigation of the grammar of these texts has resulted in a good number of improved readings and new, hitherto un- recognized, examples of certain phenomena. Further, his intimate knowledge of both the texts and the scholarship on them assures us of a thoroughly up- to-date presentation of the state of the field. The frequent elicitation of parallels with other Peripheral Akkadian dialects is also valuable; the earlier and contemporary "core" Mesopotamian dialects of Ak- kadian are also often cited. There are, however, a number of instances in which the normative nature of a feature (i.e., its presence in a core dialect or di- alects) has been overlooked by Rainey; see below, e.g., ad I 38, 39, 139-40, 141, 145, 151; II 25, 154, 370; III 43, 210, 214. (Volume III, written a dozen years before the publication of this work, was unfor- tunately only slightly updated before publication, so that a good number of recent works-especially on other Peripheral Akkadian dialects-are not referred to in the volume; this has meant less comparison with the attested adverbs and particles of these other dialects that have been investigated recently, and thus has sometimes led Rainey to ascribe a Canaan- ite origin to some forms and features that are widely attested in Peripheral Akkadian and that may reflect a different common tradition. On the other hand, vol- ume III offers more comparisons with other North- west Semitic languages. In general this volume is invaluable in its presentation of what constitutes, after the verbal system, the greatest obstacle to un- derstanding these texts; NWS influence is perva- sive in the morphosyntax of these elements, which Rainey covers admirably.

    Another nice feature is that when more than one interpretation is allowed by the text and the method- ology of interpreting these forms, Rainey does not usually hesitate to cite all possibilities and to state that certainty is not possible (e.g., II 86, 89-90, 162- 68, 310-11, 316, 325-26); he is rarely dogmatic, except-rightly-when ruling out interpretations that are not permitted by the system as he understands it. (On the other hand, the reader does have to put up with the usual quota of Rainey's typically bombas- tic polemicisms. A number of hobby-horses appear

    several times: the Akkadian iptaras form is "mis- named" the perfect [II 81, et passim]; the West Semitic suffix-conjugation should under no circum- stances be called a "perfect" [II 282, 366]; the long- dead belief that "the ancient Canaanite scribes had no sense of time" is refuted [II 228]; and there is the typical Rainey shower of exclamation points.)

    Will these volumes work for Amarna neophytes? They are, happily, written for such-for students of Hebrew, of Semitics, and of Assyriology-and no knowledge of previous Amarna scholarship is as- sumed. As one of the lucky few to have studied with Rainey or Moran, I am not really able to stand back and judge this question properly. Yet my sense is that a novice could in fact use these volumes, al- though, since there are few summaries, it might be necessary to read all of the first three volumes rather carefully (and it may be unlikely that anyone, other than reviewers, will do that). There are also a few general problems of presentation that may prove to be stumbling blocks. Some of these are considered in the following paragraphs.

    A puzzling aspect of the work is that it provides no list of the texts on which this investigation of "Canaanite" is based. Rainey states that the corpus comprises "those EA texts which show West Semitic influence, viz. those from the land of Canaan plus some of the Alashia letters" (III xiii; cf. I xiii); and a definition of Canaan is offered in II 1: "the Levant south of the Nahr el-Kebir and ancient Kedesh on the Orontes." But a listing of the actual texts and their provenances should have been given (see the very useful listing of all Amarna letters by prove- nance in Izre'el 1990), as should an explanation of the criteria for the inclusion of a text in the corpus. For example, only certain Alashiya letters are in- cluded, e.g., EA 33 but not EA 37 (see I 146, II 187; in I 72 reference is made to EA 37:19, but Rainey does not mention that this letter is not part of the cor- pus; similarly in III 222). Amurru texts are sometimes omitted (e.g., I 81), sometimes included (I 87), with- out explanation; those already familiar with Amarna studies may know that only some of the Amurru texts belong to the Canaanizing variety while others belong to the "northern," non-Canaanizing, variety of Peripheral Akkadian (see Izre'el 1991), but many readers will not. Texts of Dagan-Takala (EA 318 et al.) are described as "more N. Syrian in style than they are Canaanite" (II 189), but one misses a general discussion of this important distinction in Peripheral Akkadian texts anywhere in this study.

    BASOR 310 60

  • A GRAMMAR OF AMARNA CANAANITE

    Rainey has also, happily and quite rightly, included the grammatically similar letters found at Tacanek and at Kamid el-L6z, even though these are not part of the "Amarna letters," but their inclusion without any explanation may well confuse some readers.

    Rainey often mentions the provenance of many citations in his commentary; it would have been helpful, however-and would not have taken up much more room-if the provenance of all citations had been given along with their EA references, for example "EA 87:15-19, Byblos," rather than simply "EA 87:15-19."

    In general, the key to interpreting the complex code of the verbal forms-the use of various Akka- dian bases, which are to be ignored in favor of the overlaid Canaanite system of tense-mood-aspect and person markers (so that, e.g., iz-za-az-zu is to be considered a WS suffix-conjugation form) is fully presented, so that, finally, all who wish to read and understand the Canaanite Amarna texts have a full and reliable guide. In this Rainey has done the field a great service. In II 13-15 is a very clear sum- mary of the workings of the verbal system in this corpus, in which "three linguistic strands . . . woven together in the language of the Amarna texts from Canaan" are disentangled and explained: i) the un- derlying Old Babylonian matrix learned by the scribes (and Rainey musters a considerable array of evidence to show that the Akkadian basis is indeed Old Babylonian, unlike in the Peripheral Akkadian dialects to the north, such as those of Ugarit and Emar, which are based primarily on Middle Babylo- nian); ii) a "colloquialism" strand of "independent developments" not based on West Semitic or on other Akkadian dialects (though this is not supported by many examples in these volumes); and iii) "the di- rect influence of the native West Semitic language(s) spoken by the scribes," which was probably Canaan- ite. This discussion is followed by an extremely useful table of forms illustrating "how the system worked."

    There is little discussion of theoretical or meth- odological issues involved in interpreting the gram- matical structures of these texts (although some are considered in volume II, e.g., pp. 17-18, 24-26, 28- 32, 162). Rainey refers to the language as a "code" (II 32); he also mentions "the peripheral jargon" (II 121), "pidgin developments" (II 133), and the important notion of an "interlanguage" (e.g., II 135, 346), first applied to these texts by A. Gianto (1990: 10). Questions of the origin of the language, its re-

    lationship to normative Akkadian, and whether it was a spoken language are also raised briefly (e.g., "When and where the scribes came to a mutual agreement to write in that particular manner is still a mystery," II 15; see also I 182, II 84, 346), but there is no general review of these matters. Nevertheless, the discussion of such issues can now proceed from the firm de- scriptive groundwork that Rainey has produced, and these volumes should stimulate an increased interest in this fascinating grammar.

    There seems to be an assumption of essential gram- matical uniformity among the texts in the corpus; only letters written by scribes in Jerusalem and Tyre are regularly singled out for their unusual features within the "WS" system. Even with these, it is done feature by feature, not in a general discussion at the beginning of the first volume, which would have pre- pared the reader for such deviations from the norm. The assumption of uniformity entails an assumption that the scribes shared a common educational back- ground, and this should perhaps have been discussed; it is broached briefly in the "concluding observa- tions" to the chapter entitled "Suffix Conjugation- Special Hybrids" (p. 346): "The underlying question that recurs with every discussion of a peculiarity like this one in the Canaanites' Akkadian interlanguage is that of origins. When and where did the scribes of Canaan learn to write Akkadian? Obviously, their original teachers had brought to them a very sophis- ticated knowledge of the finer nuances of Akkadian grammar and syntax."

    The arrangement of the material is not always to the best advantage for the reader. For example, vol- ume II chapter 2, "The Akkadian Base Language," should perhaps have been presented early in the first volume, since it deals not only with the verbal sys- tem (the topic of volume II), but also with pho- nology. The "West Semitic" verbal system, which involves the most interesting features of these texts and those most important for our understanding of early Northwest Semitic grammar, is finally presented in volume II chapter 10, after several chapters on the derived verbs. There is a long presentation of the modal uses of yaqtulu forms of the verb (w)ussuru in II 157-68. This atomistic approach is not really use- ful; it would have been more helpful to have a sec- tion before the derived stems on the "uses" of the "modes"-such as appears again more generally in volume II chapter 10-and then to discuss the mor- phology of, e.g., (w)ussuru forms, which is not really covered in a systematic way because the examples

    61 1998

  • JOHN HUEHNERGARD

    are spread over the semantic functions of the many forms. Similarly II 197-202 (on Akkadian modal markers) might have been fronted for the novice's sake.

    Rainey explains that "Throughout all three vol- umes ... the emphasis is on morphosyntax; morphol- ogy is not divorced from syntactical function" (I xiv; see also II 49). Unfortunately this arrangement some- times means that certain syntactic and semantic fea- tures, which one might like to have seen explored and discussed, are not considered because there is no formal place for them. These include, inter alia, the occurrence of double accusatives; the construction(s) used for counted items/materials; and a general dis- cussion of concord, for example, the fact that plural nouns are frequently construed with singular verbs and are referred to by singular pronouns (see Moran 1950: 60-66). The important phenomenon of clause sequences, such as {"indicative" + "indicative"} or {"injunctive" + "injunctive"} as two means of ex- pressing purpose, is discussed only in passing in the commentary on a number of passages, and then finally more generally in the section on the con- junction u (e.g., III 106), which is perhaps not the first place one would normally look for it. Questions such as why the suffix-conjugation appears in some instances rather than the prefix-preterite yaqtul, or vice versa, are not considered. Rather the semantic ranges of qatal(a) are listed in one place, those of yaqtul in another; but the choice of one over the other, and their possible overlap, are not discussed.

    These volumes contain a large number of typo- graphical errors. Fortunately the vast majority are not critical; the few that are will be noted where ap- propriate below. Square brackets and angle brackets are regularly used not only in the transliterations but also in the translations; there is, however, consider- able inconsistency in this regard, a good number of the translations lacking such brackets where they are expected (e.g., I 120, 122, 135, 141, 164; II 92). There are a few examples of what appear to be bad computer "pasting": in II 124 the second paragraph and its example (EA 106:30-34) seem to be mis- placed; in II 275-76 the suggestion that a-mur should be supplied in EA 288:7-8 is made twice; in III 205, the fragment "Youngblood (1961:1950a:70)." be- tween two text citations must be the remnant of a computer glitch.

    The following additional observations are offered in the hope that they might make this indispensable reference work still more useful to students of the Amarna Letters, and to suggest areas in which fur-

    ther research might profitably be done. Although I will also take issue with Rainey at a number of points, this must not obscure the very great service he has rendered in these fine volumes to all students not only of the Amarna Letters, but of the history of Canaanite and Hebrew grammar, of Semitic linguis- tics, and of creoles and interlanguages. At the end of the preface to volume IV, Rainey expresses his hope that these volumes "will inspire and assist a new generation of scholars to carry on" work on the Amarna tablets. He can rest assured that they will; the field is very much in his debt.

    My specific comments are as follows:

    I chapter 1, "Orthographic Problems." This is sketchy, at best; there is no discussion, for example, of the writing of doubled consonants (below, e.g., ad II 58), or of expected final ultralong vowels (frequently not indicated by an extra vowel sign, unlike in core Akkadian; e.g., an-na for expected anna; for references see I 94).

    I 1-6. It is curious to find the very useful "Publica- tion History" of the Amarna tablets presented as the first section of the chapter on "Orthographic Prob- lems" rather than as a separate introductory chapter.

    I 3-4. It is worth repeating Rainey's observation, known to the few Amarna specialists but not more widely, that J. A. Knudtzon's transliterations in his Die El-Amarna-Tafeln mit Einleitung und Eriduter- ungen (1907-1915) are usually "more reliable than the autographs of his predecessors."

    I 9. In EA 73:34', for a-ni-ni-ta read an-ni-ta. In EA 288:28, for uz read u.

    I 10. It is not made clear why "the grammar would require" us to read TE as ti7 in forms such as te-ma- ha-su-nu (EA 271:21); can it be shown with cer- tainty that some scribes did not have, e.g., [ti-] and [te-] as allophonic prefixes (see Izre'el 1987)? Or even if TI and TE were interchangeable signs for some scribes, is it necessary to write te9 and ti7 on the basis of normative Mesopotamian Akkadian grammar?

    I 11. In the final paragraph the writing [S]rAG - qa-di (for qaqqadi; EA 209:16) is cited as "the only example of KA = qa in a text from Canaan"; in the preceding paragraph, however, the writing KA-rZI- ra (EA 244:14; Megiddo) is said to represent either ga14-zi-ra "or, perhaps more likely," qa-si-ra.

    62 BASOR 310

  • A GRAMMAR OF AMARNA CANAANITE

    I 14, line 8. For "BI = pi" read "BI = p'."

    I 18, paragraph 2, line 8. For KURKa-(pa)-Sf read KURKa-((pa))-SKI.

    1 19. Note that SI occurs for si20 also in Old Baby- lonian, e.g., in writings of simmum in CH.

    1 23-24. It is noted that the IA sign normally has the value /ya/l (transliterated ia), but that it will some- times be transliterated yi. Yet there seems to be a general unwillingness throughout these volumes to assign any value to IA other than ia. For example, in I 106, Rainey suggests that the Canaanite personal interrogative pronoun, written mi-IA(-mi/mi), is in- declinable, miyd; but perhaps nominative examples are to be transliterated mi-iu. Of the causative form IA-sf-ni "may he take me out" (EA 282:14), dis- cussed in II 192, Rainey states that "the prefix syl- lable is written ia- so there is little doubt that the form has to be reconstructed *yasi'nf" rather than *yo6si'n < *yawsi'- as expected from Hebrew and elsewhere in Semitic, and suggests that perhaps "the Canaanite H causative was based on the biconsonan- tal root *s'"; but the problem disappears, of course, if we may transliterate iu-si-ni. In III 110-11 the indefinite adverb a-IA-ka4-mi (EA 149:52) is said to be "closer to the Assyrian ayyakamma than the Babylonian ayyikdma"; perhaps we should translit- erate a-ii-ka4-mi; note that in I 24 it is said to be a question whether a-IA-ka-am (EA 254:27) should be represented as a-ia-ka-am or a-yi(=ii)-ka-am.

    I 36, paragraph 2, line 7. For "Ebeling's" read "Bohl's (loc. cit.)."

    I chapter 2. This chapter, "Phonological Features," describes features of the Akkadian level of the Canaanite Amarna tablets, where Rainey presents some of the strongest evidence for the essentially Old Babylonian matrix of this language: "what the Canaanite scribes are using is an attenuated form of Old Babylonian" (I 43). Unfortunately, there is no discussion of Canaanite phonological processes ev- ident in the texts, such as the Canaanite shift (d > o) or the assimilation of n to a following consonant, as in the energic form nu-ub-ba-lu-us-su (EA 245:7).

    I 38. Note that arhis apparently derives from a root I-' (cf. the verb arahu), despite a few rare instances of warhis in Old Babylonian (hypercorrect forms alongside the much more common arhis).

    I 39. Note that "shortened forms" of (w)ussuru, such as yu-us-si-ra, correspond to forms such as u- se-er in Old Assyrian, where they occur alongside full forms such as uwasser (Hecker 1968: 157 ?93f), and Middle Assyrian, where they are the norm (Mayer 1971: 76 ?78.4). One might also note in this section on verbs I-w the unusual Amarna creations i-wa-si-ir (EA 137:8), ni-wa-as-si-ru-su (EA 197:18), and i-wa-si (EA 147:64).

    I 45. For ri-ik-si in KBo 1 2 r 34,35 read ri-ik-si.

    I 47. The "use of the third person pronouns as vir- tual copula" is mentioned here, but it does not seem to be presented in a systematic way in any of the sec- tions on the pronouns or elsewhere, including I 63, where an example appears: mi-nu su-ut PN UR.GI7 "who is PN the dog?" (EA 76:11-15).

    I 48. It is noted here that the usual word order in verbless clauses with pronominal subject is subject- predicate (although this order is frequently reversed in subordinate clauses); this is a significant distinc- tion from normative (Old and Middle) Babylonian, where one has, e.g., sarrum dannum atta; see also III 254-64.

    1 51. For EA 114:9-15, read EA 114:12-15.

    I 53. In EA 74:62 Rainey reads a-na ma-har sar-ri EN-(ia) "before the king, (my) lord"; while the omission of -ia on EN is unusual, it is not unprece- dented; see EA 100:9, 136:21, 139:29, 140:1, and ZA 86 100:16.

    I 54. Rainey cites EA 23:26-27, from Mittanni, as an example of peripheral use of the independent dative personal pronouns without a preceding prep- osition ana, "in the classical manner"; however, the use of the dative in the example given, SE?-ia ut ia-si li-is-sur-an-na-si, is not "classical," since the native Mesopotamian dialects would have accusa- tive pronouns in this clause (e.g., Old Babylonian ydti u ahi lissur(an)nidti). It might also be noted that the "standard usage" in the Canaanite texts, viz., ana + dative pronoun, is also well attested in the early normative Akkadian dialects (see GAG ??41c, 114d).

    1 55. The translation has been omitted in the citation of EA 364:18-20 [sic; not 364:20]: "The king of Hazor has taken three cities from me."

    63 1998

  • JOHN HUEHNERGARD

    I 56. The translation of EA 365:18-19 should read "they are not doing/acting as I."

    I 59. The translation of EA 34:48-49 is awkward; since the first clause is extraposed, as Rainey notes, we might render "As for what you request, I will give it"; ui introducing the main clause need not be translated.

    I 60. "So the diachronic and synchronic influences that might have led to the EA Canaanite use of kdta for all three cases remains a mystery": Since we find kdta as the independent genitive-accusative form al- ready in Old Babylonian, the form itself is obviously not a mystery; its extension in use to the dative is probably an example of morphological simplifica- tion (reduction of allomorphism), a common phe- nomenon in peripheral dialects (Huehnergard 1989: 273-74).

    I 62. The separation of the occurrences of the third person pronouns into two sections-true third per- son and anaphoric-is unfortunate in that it renders assessment of the morphology of these forms diffi- cult, without contributing significantly to the syntac- tic analysis of the occurrences.

    I 65. "As an independent pronoun, only the dative, sasu, is found": Rainey here contrasts "indepen- dent" (a morphological category) with anaphoric (a semantic category); since all the forms under con- sideration are "independent," a preferable contrast is personal vs. anaphoric.

    I 66. The translation of EA 366:22-26 is missing the phrase "with 50 chariots."

    I 70. In Rainey's statement that "the rich repertoire of pronouns in normal Akkadian was either unknown to the scribes of Canaan or else it was consciously rejected in their interlanguage," the first alternative would seem to contradict his claim in II 346 that the scribes' "original teachers had brought to them a very sophisticated knowledge of the finer nuances of Akkadian grammar and syntax."

    I 72. The appearance of the Ics suffixal allomorph -ya on a singular accusative pa-ar-sa-ia (EA 73:39') may be compared with Ugaritic, which likewise has the longer allomorph on singular accusatives (pace Rainey's comment in I 71; see Gordon 1965 ?6.6).

    I 76. "The -hu in qa-ti-hu could very well be tran- scribed u5 or 'u5 since it appears to be meant for the WS 3rd m.s. suffix, -hu"; or we might simply say that here a H sign is used for Canaanite /h/, as also, e.g., in the Canaanite forms hi-ih-bi-e for /hihbi'a/ (EA 256:7) and ha-ar-ri for /harri/ (EA 74:20).

    I 77. Of nominative lib-bu-su Rainey says that "proper Akkadian inflection would call for libbasu"; this is true only of Babylonian, however; in Assyrian libbusu is expected (see Hecker 1968: 110 ?64d). In I 89, as well, Rainey states that the form of the third person plural accusative suffix, -sunu, "departed from the Akkadian norm," when what is meant is the Babylonian norm, since it is noted immediately after- ward that -sunu is normative in Assyrian.

    I 80-81. The first sentence in the last paragraph on p. 80 should appear as a separate paragraph after the citation of EA 74:32-33 on p. 81.

    I 91. The scribes' "choice of suffix pronouns in- cluded some archaic forms such as the duals and per- haps also 3rd m.s. suit"; it would seem that "suffix" should be deleted from this sentence.

    I 92. Here a very convenient listing of all attested Canaanite pronominal suffixes is presented.

    I 96-97. The nominative form sgat as an anaphoric pronoun in lUUR.GI7 su-ut "that dog" (EA 84:35; cf. 90:21) is said to occur "in demonstrative function," but the accusative pronoun in the same construction, viz., {noun + anaphoric pronoun} (EA 138:117), is said to be "in direct apposition" to its noun.

    I 97. The oblique form suata, with final -a, is unusual, even for peripheral Akkadian. Note also the form su-wa-at (discussed immediately below) and the plural (oblique) su-nu-ta5 (EA 148:44, cited on the same page).

    I 98. The unusual anaphoric form su-wa-at (EA 85:72, TT 2:4) is said to appear in "status absolu- tus" (so also in I 178); the latter term, however, re- fers to the lack of a case vowel on nominal forms, whereas the final vowel of normative sudti, etc., is not a case ending, and so the application of the term status absolutus to su-wa-at seems strange; but per- haps it was created on the analogy of such forms among nouns.

    64 BASOR 310

  • A GRAMMAR OF AMARNA CANAANITE

    In EA 85:69-73, before UD.KAM.MES, add URUsidu-na is-tu (line 71).

    I 99. It is stated that there are no examples of a proleptic [sic; not prolectic] suffix; note, however, ERIN.MEg pi-td-(ti)-su / sa LUGAL "the army of the king" (EA 141:31-32), cited in I 135.

    I 102. The translation of EA 152:53-55 is missing a few words: "The king, the Sun, [my] lord knows the one who is a [true] servant. .. ."

    I 105, last paragraph. The normal Akkadian values of minu and mannu should be given as "what?" and "who?" respectively, not vice versa.

    I 106. On whether the Canaanite personal interrog- ative is indeed indeclinable (mtyd), see above, ad I 23-24.

    I 107. In EA 92:41, for sd-ti-tam read sd-ni-tam.

    I 108. Note that all examples of the form miyati "who?", with its obscure -ti ending, appear in the phrase miyati anaku u "who am I that.. ?".

    I 115, lines 2, 3. For mi-im-mi and mi-ma read mi- im-mi and mi-ma.

    I 120, last line. Add to the translation "to my lord" between "Adonis" and "lest."

    1 129. The last of the first group of examples should be deleted; it does not contain dlu/URU.

    1 131. Note the curious fact in EA 92:10-11 that al- though the adjective modifying nukurtu is feminine, the verb governed by it is masculine.

    I 135. Another possible example of non-normative concord is the treatment of eleppu as masculine in EA 114:12-14: GIS.MA.MES ... sal-mu su-nu, if Rainey's translation in I 51 is correct; Moran (1992 ad loc.) translates differently, however.

    I 137. In EA 141:46, for pi-t[d-at] sa] read pi-t[d- at sa].

    I 138. It is not certain that the writing GIR-pi must represent a dual sepl/sepe; it is possible that a singular bound form sepi may have existed along-

    side sep, on the analogy of qat and qati (cf. GAG ?64c).

    I 139-40. In the section on plurality a few forms are listed "in which MEg or HA is applied to the mor- phologically singular form of the substantive"; but the first of these, a-wa-atMEg LUGAL "the king's words" (EA 192:10), may be intended as plural awat rather than singular awat. Conversely, accusative a-wa-ta5MEg (not quite passim, as claimed by Rainey, but rather five examples, all in Jerusalem letters: EA 286:62, 287:4,67, 288:64, 289:49) is cited as an ex- ample in which "the singular and the plural could be represented by the same syllabic orthography"; this form would seem to be singular, given the final -a (vs. expected pl. awati), but it is usually modified by an adjective with a form that is impossible in nor- mative Akkadian, ba-na-ta, which looks like a con- flation of sg. banita and pl. banati.

    I 140. The reason for the note "nominative for dependent case!" (i.e., genitive) accompanying qa- du LU.MES ha-za-ni-ku-nu (EA 117:62) is unclear to me.

    I 141. The writing pa-nu-ia is contrasted with "the normal Akkadian pani'a"; the use of -ia for "my" on plural nominative forms (ma-ru-ia "my sons"), while less common than the use of -a (ma-ru-a), is well attested in Old Babylonian, however (e.g., a-wa-tu-ia "my words" AbB 3 15:20). It is also claimed here that the preservation of a case vowel before a suffix "is also proof of the plural form"; see, however, GAG ?65h6 for forms such as Old Babylo- nian singular nom. marasu, acc. marasu "his son" (e.g., ut-ul ma-ru-ni / at-ta VAS 8 127:24-25).

    I 145. Oblique forms of samu "sky" are discussed: some are written with final -i, as in sa-mi-i and sa-mi, while others have -e, as in sa-me-e and sa- me; Rainey suggests that the latter "may be due to the reduction of a dip[h]thong," comparing Hebrew samayim. But it is the forms in -e that are normative Babylonian, the form of "sky" being a plural formed on the base samd-, thus nom. samiu < samdau, oblique same < samdal (the uncontracted forms also occur, especially in literary texts). Similarly the forms of "water," me'-(e-)ma, may indeed reflect a Canaanite form /mema/, but it is the addition of -ma that makes this so, not the quality of the vowel, since Akkadian too has oblique me.

    65 1998

  • JOHN HUEHNERGARD

    I 146. To the bibliography on -anu/-dni add the important article by G. Buccellati (1976).

    I 151. Note that the Old Babylonian plural of tuppu may be masculine or feminine (see AHw 1394-95); thus, tup-piMEg (EA 74:10) is not an improper form, as claimed here; tup-paMEg (129:28; sic, not 129:8) is probably singular (cf. Moran's translation: "on n'a [pas pre]sente de tablette"), with MEg marking the form as a logogram, as often in Peripheral Akkadian (including Amarna; Rainey I 28).

    I 153. It might be noted that the construction [da]n- nu-rtu4' ibassu "were strong" (EA 81:49) is non- normative; we expect simply dannu.

    I 156. The references for LU.MAgKIM-rtil are missing: EA 371:24,38.

    I 160. The second clause is missing from the trans- lation of KL 72:600:9-13: ". . .; give order that his accoutrements be sent."

    I 162. In EA 73:34', for a-ni-ni-ta read an-ni-ta.

    1 166. The reading of UD as /ta/ in the singular verb hu-mi-UD (EA 102:30) is undoubtedly correct; the value ta5 given here by Rainey does not seem to be recognized in the modem sign lists, however. Like- wise the form ba-la-UD in EA 126:15, given by Rainey as ba-la-ta5 in I 110 and as ba-la-tai in I 119, should be read ba-la-tax. Alternatively we may read hu-mi-tdm, ba-la-tdm.

    I 171-73. The bound forms biti in (acc.) E-ti ha-za- ni (EA 89:49) and sari in sa-ri / sacr-ri (EA 100:36) are not necessarily "errors"; see above, ad I 138 on sep(i).

    I 180. Another example of a declined (attributive) adjective as predicate was noted above, ad I 153.

    I 181. The suggestion that the noun nukurtu "may have been utilized ... as a logogram" for the pred- icate forms (stative) of nakru is insightful and appealing.

    I 190. The suggestion that li-mi-ma (add the refer- ence: EA 205:6) represents Canaanite /li'mima/ "peoples" rather than Akkadian "thousand(s)" finds

    support in the Emar phrase li-im A. AMEg, probably "people of the field" (Emar 30:31, 180:19, 185:17; AuOr 5 14:14; RA 77 20, 3:19; 23, 4:13).

    I 195. For [i]-[u-su-na in EA 91:25-26 read [i]- pu-su-na.

    I 201. On forms of kalu "all" see Huehnergard 1987b: 190, n. 51.

    1 203. For ma-sa-ar-ti in EA 103:48 read ma-sa-ar-ti.

    II, chapter 1. A very thorough and helpful For- schungsgeschichte is given here, titled "In Search of the Canaanite Verb."

    II 19, paragraph 1, line 3. For "IIIrd weak verbs" read "Ist weak verbs"; examples of plene writings of initial vowels in such forms are said to be absent from the Canaanite letters, but note ra-alP-la-ak-mi in EA 191:16 (Rogisu), cited by Rainey in II 41; for the reading, see Knudtzon's n. g at EA 191.

    II 25. Although forms of verbs III-weak without vowel contraction, such as ta-sa-me-ut and te-le- qe-u are probably rightly considered Assyrianisms, nevertheless given that Rainey suggests that the Old Babylonian dialect inherited by the scribes had "some archaic" features, we might note that lack of con- traction is also characteristic of certain early dialects of OB, such as that of Tell Asmar (Whiting 1987: 7). The form li-il-qe-a-ni is normative OB, contraction of ia/ea > a occurring only in later OB texts. There is some confusion in the presentation of forms of semu here: they are mistakenly called Ist aleph rather than IIIrd weak; the five instances of is-me are said to be Babylonian, but three of these are 3ms (note: correct 151:2 to 151:12), for which isme is nor- mative in both Babylonian and Assyrian, while the remaining two are Ics, for which Babylonian has either esme or asme and Assyrian has asme; 3ms is-te-me (sic; not -me) is also said to be Babylonian, but is normative in both Babylonian and Assyrian.

    II 27. The 3mp verbs with prefix ti- (tiprusi) in early Mari texts are said to be the result of West Semitic influence rather than a vestige of common Semitic (thus Edzard 1985: 85-86). Note, however, that the texts in which such forms occur (which are pre-Zimri-Lim) exhibit no other West Semitic fea-

    66 BASOR 310

  • A GRAMMAR OF AMARNA CANAANITE

    tures and few West Semitic names; further, the 3fs forms compared by Rainey have ta-, not ti- (West- enholz 1978: 165).

    II 29. The form in line 7 of the Shechem letter should be read simply tu-sa-ab-ba-lu, rather than re- stored as tu-sa-ab-ba-lu-[ni-ni ], which is impossible (since the verb is 2ms, the form "with 1st c.s. dative suffix" would have to be tusabbalam).

    II 33-47. The review of the morphology of the person markers in the Canaanite prefix-conjugation forms is thorough and helpful. Rainey notes that there are, by coincidence, no 2fs, 2fp, or 3fp forms. One important form that has been overlooked, how- ever, is the third person dual ti7-id-[di]n-na in EA 246 r.6 (the subject is 2 DUMU.[ME?] / ILa-ab-a-ya, which is resumed by the dual suffix -su-ni); similar dual forms are attested in Emar texts; contrast the use of the 3mp with a dual subject in EA 250:5-8.

    II 42. For lcs G forms with prefix 0u- note also the forms of nasaru the author cites on the previous page.

    II 50-61, "The Active G Stem." One of the features of the verbal morphology of the Canaanite Amarna letters that has become increasingly clear through the work of Moran and especially Rainey is that the Akkadian base of a given verb form is not mean- ingful, the morphological information being carried instead by the Canaanite markers that are affixed to the Akkadian base. Thus, for example, a form like yispuru "he writes, will write," though based on the Akkadian preterite ispur, is actually marked as im- perfective by the ending -u; conversely, yisappar, though based on the Akkadian durative isappar, is to be understood as perfective, "he wrote," i.e., as a yaqtul-0 form (see II 225). There is, however, some uncertainty about how the scribes understood the Ak- kadian durative base: in some instances, it may have been treated as a D stem (Moran 1950: 165-66). Rainey considers such forms at a number of places in volume II, but his treatment seems somewhat inconsistent. Of forms such as yi-na-sl-ru "he will guard" (EA 112:14), he says, "The unusual -i- theme in these examples [vs. normative inassar] is difficult to explain except as a contamination from the D stem" (II 134; cf. also II 53, where i-hal-li-iq is said to be "used transitively as if D" and a-kal-li is said

    to have "the thematic vowel of the D preterite"). But Rainey also sometimes suggests that forms with the normative theme vowel might reflect Canaanite D stems, such as ti-ma-ha-su-ka "that they may strike you" (EA 252:27, see II 55). Such an interpretation becomes difficult, I believe, with intransitive verbs. For example, ti-ma-qu-tu "you fall" (EA 73:10) is said "probably to represent a D stem nuance" (II 56; cf. II 169-70); but this is difficult to accept, since the meaning is clearly intransitive, without any obvious "D stem nuance." On the other hand, Rainey says nothing about, e.g., ti-da-ga-lu "that they may see" (EA 100:34 [sic; not 101:34]) or ti-sa-pa-ru "you write" (EA 93:8) being anything other than G forms (II 52, 60). It seems likely that some scribes at least did not construe iparras forms as D stems, but sim- ply used them as an alternative base to which to affix their NWS modal morphemes. (A similar problem arises in the interpretation of verbs based on Akka- dian t-forms; see below, ad II 82-113.)

    Having reviewed a large number of G verbs writ- ten by the scribes, with an eye to whether they are based on the preterite iprus or the durative iparras, Rainey concludes (II 61) that "for most verbs, only one theme or the other is used"; but his review on the preceding pages in fact shows a significant number of verbs whose forms are based on both (akalu, epesu?, kasadu, malaku, nasdru, qabu, qalu, ra'dmu?, sadalu?, sakanu?, saparu, semu), roughly a third of the total he lists (note also lequ, with both preterite- and perfect-based forms; below, ad II 82-113). This does not, however, vitiate the well-established rule for Amarna Canaanite verbs that "the indicators of tense and mode [are] other than the standard Akka- dian distinction between the iparras and the iprus patterns."

    II 51-52. Under (w)asu add also the Tyre form i- wa-si (EA 147:64).

    II 53, line 1. For "suffix form" read "prefix form." Similarly, under idu, line 4, for "consonantal suffix" read "consonantal prefix." On p. 54, fourth line from bottom, for "WS suffixes" read "WS affixes."

    II 58. Rainey suggests that the Ics forms us-su-ru, u-sur-ru "I am guarding" are based on the core Akkadian imperative usur. While this does seem a likely explanation, it is curious that in a number of them doubled -ss- is written; a conflation of the

    67 1998

  • JOHN HUEHNERGARD

    imperative usur and the preterite vssur? confusion with forms of (w)ussuru?

    II 59. Under ra7amu, second line, for "consonantal suffix" read "consonantal prefix."

    II 60. It is unlikely that yi-se20-me and the four examples of i-se20-me/me are all scribal errors, with si for is; those two signs are not similar in shape; fur- ther, the spelling i-is-me would be highly unusual.

    II 61. Under tarasu, for yi-it-ru-us read yi-it-ru-us.

    II 61-75. This discussion of the Barth-Ginsberg Law, a phenomenon that is observable in only a handful of Amarna forms, is overly long. The review of "the ancient Semitic pattern" (pp. 62-63) is quite out of date, several important articles having been written on the subject in the last two decades (see, e.g., Hetzron 1973-1974; Schub 1974; Testen 1992). Still, the discussion of the Hebrew and Amarna forms at the end of this section (pp. 73-75) is important.

    II 65. In EA 228:18-24, for "May the king" read "May my king."

    II 78. In EA 83:20, for ti-te-u read ti-le-u.

    II 82-113. As with the verbs based on G durative iparras, which Rainey believes were frequently con- strued by the scribes as NWS D forms, so too in this presentation of forms based on Akkadian iptaras, Rainey is at pains to establish a semantic value for the infixed -t- in most of the Amarna Canaanite ex- amples. He divides the presentation of iptaras forms into three sections: "Lexical Gt's from Akkadian"; "West Semitic Lexical Gt's"; and "Verbs with no Lexical Gt." In all of this there is an unwillingness to allow that most of these forms may simply be in- tended as G forms, with the Akkadian perfect, rather than the preterite or the durative, serving as the base for the Canaanite form, especially in verbs II- and III-weak. But if the scribes, having learned the three Akkadian aspectual forms-iprus, iparras, and ip- taras-were able to use either iprus or iparras as the base for their Canaanite forms (above, ad II 50-61), then undoubtedly they could also have used iptaras as a base, without intending any semantic nuance as- sociated with the Akkadian (or their own) Gt stem. In his discussion of -t- forms of semu (pp. 95-98), Rainey himself admits that "there are some contexts

    in the Canaanite Amarna letters where the form with infixed -t- is probably nothing more than the G stem -t- preterite [i.e., the perfect] as theme"; of lequ (which has no Gt in core Akkadian), Rainey notes that "all the forms of lequ with infixed -t- may sim- ply represent the use of the G t-preterite as theme" (p. 101; so also on some-but not all-examples of epesu with -t-, pp. 111-13). If this is true for these verbs, it is difficult to understand why other verbs with -t- must be seen as Gt. Indeed, much of Rainey's effort to find semantic value in certain of these t-forms seems forced. For example, the forms of elu with -t-, discussed pp. 82-86, do not have their normal Akkadian separative nuance, and it is not clear what Rainey believes the "intentional use of the Gt nuance" in these forms (p. 84) to entail semantically. Rainey seems to suggest that the two examples of erebu with -t- reflect the Akkadian Gt meaning of this verb (pp. 86-87), which von Soden gives as "fur die Dauer eintreten" (AHw 236a; cor- rect Rainey's reference); but the Gt of erebu is very rare, and it seems unlikely that the Amama scribes would have been familiar with it. The same is true of the Gt of ezebu (only one example in AHw), of kas- adu, and of a number of other verbs. In these in- stances, at least, it is much more probable that the Amarna forms are simply based on the perfect, i.e., that they are lexical G stems.

    II 83, 10 lines from bottom. For "previous" read "next."

    II 91, line 2. For til-ta-tt read til-la-ti. In EA 149:57-63, quoted further down the page, part of lines 61-62 has been omitted from the translitera- tion and the translation: for GIMA.MES GIR.ME?- su-nu read IMA.ME-su-nu/GI~GIGIR.ME~-su-nu ERIN.MES GIR.MES-su-nu "their ships, their char- iots, and their infantry" (on the last term, see Moran 1992: 140-41 n. 5 to EA 71).

    II 103-4. In EA 276:9 and EA 305:16, a-wa-at is not necessarily singular; if pl. awat is intended, the use of the Gtn is-tap-pdr/is-tap-pa-ra-an-ni in these passages is not a difficulty.

    II 108, paragraph 2, line 6. For "Gulba" read "Gubla" (i.e., Byblos).

    II 117-18. Rainey proposes that the two Amarna attestations of kasasu (EA 94:67; 138:37) be under-

    BASOR 310 68

  • A GRAMMAR OF AMARNA CANAANITE

    stood as N forms. In EA 138:37, however, Rainey's interpretation entails taking the accusative DUMU IR-[A-si-ir-ti] as an agent, "by the son of 'Abdi- [Ashirta]," which seems an unlikely construction. (In Rainey 1989-1990: 63 he translates EA 138:37 as follows: "How long must we be burdened with debt (to) the son of 'Abdi-Ashirta?"; this is some- what more plausible, but still assumes that a prepo- sition such as ana was omitted by the scribe.)

    II 120. In EA 227:8, for an-tu-ta5 read an-nu-ta5. Of EA 227:8-13, Rainey states that "The N stem pro- vides a similar solution to another anomalous form," apparently reading yi-sa-ma in line 10 as N stem; but the form under discussion is unfortunately omitted in the translation (presumably something like u ki yi-sa-ma "and when it was heard"); on the whole, Moran's interpretation seems preferable: u ki-yi-sa- ma "and accordingly."

    II 124. The second paragraph and its example (EA 106:30-34) are out of place here, apparently the re- sult of improper computer "pasting"; it should go at the end of the discussion of epesu/nenpusu.

    II 125. The verbs nu-da-bir5 and ti-ne-pu-us in EA 74:34-35, with their -0 ending, are surely jussives rather than futures: "so let us exterminate . . . that all the territories may go over. . ."; cf. Moran 1992: ad loc.

    II 132. Since an N stem of the stative verb saqui "be high" is not attested (or expected) in core Akkadian (although an obscure Ntn form is noted by von Soden in the Nachtrage to AHw: 1589), it seems more likely that yi-[sWa-q( in EA 144:15, despite the final -i (for -u) is simply a G form with an Akkadian durative base and Canaanite preterite (yaqtul) markers.

    II 138-76. The problems associated with presenting the derived verbs (here, the D) by "semantic catego- ries" may be illustrated by the examples of bullutu cited on p. 139, with the meaning "to obtain sup- plies," which is hardly "factitive" but rather essen- tially the same as the G; similarly, bu`^ (II 140-46) is not a factitive verb, but lexically D, as is kullu, which Rainey (II 151) lists under "causative" D verbs.

    II 141. It is curious that Rainey, who is willing to read the ?E sign as six passim, reads tu-rbil-ru^5'-na

    rather than the proper tu-rbe'-r'u5l-na (and so also in the normative Middle Babylonian examples he cites in the same place); in fact, the latter transliter- ation, with -be-, appears on the following page.

    II 144. In EA 79:27-29, for is-tu read is-tu; add "from the king" to the end of the translation.

    II 145. The transliteration sa!(A)-ba-at in EA 244:41 suggests an emendation; E. I. Gordon believed, however, that sa- was actually the more likely read- ing than a- (cited in Moran 1987: 469 n. 6; 1992: 299 [sic; correct Rainey] n. 6]).

    II 147. In EA 76:38-41, since the verb following the imperative us-si-ra is a yaqtul form (tu-da-bi-ir), we should probably understand a purpose clause, "that it may expel," rather than Rainey's "and it will expel."

    II 148. Under qubbulu, sixth line, for "Though," "Since" is meant.

    II 150. In the translation of EA 197:31-34, for "Itatama" (which Rainey unfortunately follows with a "sic!") read "Itatkama" (Ii-ta-at-ka-ma).

    II 152. It is unfortunate that the biforms mussuru and wussuru (II 157-68) have been listed separately. In the translation of EA 285:7-10, for "the king" read "the kin[g my lord]."

    II 154. The use of D turru in letters for "to send back (word, reply)" does not seem to be "common" (so Rainey) in normative Akkadian, to judge by the number of examples in AHw; one suspects that its use in the Canaanite Amarna texts may be a calque on a Canaanite expression (cf. Hebrew hestb).

    II 156. At the end of the translation of EA 189:13- 17, for "you" read "him."

    II 160. In the translation of EA 117:24-27, for "and he se[n]t" read "and the king se[n]t."

    Rainey's reluctance to read SAR as sir9 (see also II 177) is puzzling; the latter is a well-established value in these texts, as shown by Moran 1975: 150-51.

    II 169. The -u- theme-vowel in yi-ma-qu-ta "let him (not) fall" (EA 81:32) makes it unlikely that "the

    69 1998

  • JOHN HUEHNERGARD

    scribes thought of such action as properly D stem"; more likely this is simply another form based on the Akkadian G durative stem (imaqqut) rather than on the preterite; see above ad II 50-61.

    II 171. It is not certain that we must interpret li- na-sa-ru-su in Idrimi 100 as a D form, since the theme-vowel suggests either D or G durative; note that other precatives based on the durative are found in Idrimi, such as li-ik-ta-na-ra-bu in line 104.

    In EA 327:1, for [i ]s-te-me MAS[KIM] read [i ]s- te-me [a]-wa-ti MA?[KIM].

    II 175. It seems unnecessary to take li-ra-hi-is-su in PRU 3 76a:9,13 as a West Semitic verb "to smash"; the subject, after all, is Adad, who does "inundate" things (see AHw 942, rahasu A 1).

    In the last paragraph, the reference for su-ut mu- ul-ka is EA 333:20.

    II 177, first line. For i-na-m( read i-na-mi.

    II 179. In the translation of EA 126:18-23, add "from the palace" between "sent" and "money."

    II 180. It is unlikely that the KL form tu-wa-as-sa- ru-na can be used uncritically "as evidence for the original Canaanite D passive vocalization of the suffix conjugation ... tuqattaluna"; see Huehnergard 1992: 221 n. 53.

    II 182. In EA 292:27, for LUMASKIM read LUMA~KIM-ia.

    II 189. The gemination in the Amurru imperative se-ez-zi-bd (EA 62:30) and related forms may reflect Hittite Akkadian orthography (see Durham 1976: 381-82).

    II 192. On the causative form IA-sf-ni (EA 282:14) see above ad I 23-24.

    II 192-93. It seems preferable to me to take ui-sa-ka in EA 82:29 as a lcs causative yaqtula form (*osa- ka < *aw.s'ia-ka), "that I may cause to go forth," despite the likelihood, noted by Rainey, that the speaker is not in the place from which the goods are to be shipped.

    II 195. On the "so-called 'subjunctive' marker" see the brief article by von Soden (1973), in which he

    called for the abandonment of the term "subjunc- tive" in favor of "subordinative" (so now also in the third edition of his GAG ?83).

    II 195-220. This chapter, "Prefix Conjugation- Moods and Tenses," and the next, "West Semitic Modes/Moods and Tenses" (pp. 221-64), are among the most important in the work: they offer the reader a clear presentation of the forms of the prefix- conjugation used by the Canaanite scribes, illustrated with excellent examples.

    II 200, 201. In the first line of both pages, "temporal clauses" would be preferable to "circumstantial clauses."

    II 202-11. In this long section, as in his JCS study (1992-1993), Rainey suggests that many of the verb forms with final -a in the Canaanite texts reflect the Akkadian ventive rather than the Northwest Semitic yaqtula ending, as was proposed by Moran (1960). It should be pointed out, as Rainey has failed to do here, that Moran himself recognized a few forms as bearing the Akkadian ventive morpheme. Most of Rainey's additional ventive readings, however, are both forced and implausible. See also below, ad II 254-63.

    II 207. In the Middle Babylonian text cited here (PBS 13 6:16-18), the form ut-zu-uz-za cannot be an- alyzed as lcs plus ventive; the form is a predicate verbal adjective ("stative", as is the 3mp ui-zu-uz-zu two lines earlier), and if it is correct as it stands must be 3fp (uzuzza; so hesitatingly Aro 1957: 115-16); if the form is indeed Ics, as understood by Rainey, we must read u-zu-uz-za-(ku).

    The forms sa-ma-ma and [i]z-rzil-za in EA 104:43-52 are better taken as suffix-conjugation 3fp forms, with Moran 1992: 177, n. 3, than as infinitive and Ics with ventive, respectively.

    II 212. The statement that "a semantic need . .. can override the usual phonetics if the latter would lead to obscurity" is linguistically naive; no examples are given to support this dubious claim.

    II 217. Rainey's nice observation that the precative in these letters is most often used "as the first in a chain of injunctive clauses, the subsequent verb forms being either WS jussives or volitives" has not, to my knowledge, been made previously.

    70 BASOR 310

  • A GRAMMAR OF AMARNA CANAANITE

    II 221. The chapter heading reads "West Semitic Modes and Tenses," but the running headers of the chapter read "... Moods . . ."

    II 228. The relevance of the Driver quotation to the terminology used for the imperfective form of the verb (Driver 1892: 28-29) escapes me.

    II 231. In the translation of EA 195:31 and 201:21, for "to wherever you say" read "to wherever he says."

    II 232. In EA 109:6, for a-bu-tu-ka read a-bu-ti-ka.

    II 233. In the translation of EA 104:17-36, at line 29, for "your towns" read "his towns."

    II 237. In EA 112:13, for yi-na-si-na-an-ni read yi-na-st-ra-an-ni.

    II 239. In EA 121:9, for u-sur-mi read ut-sur-mi.

    II 241. Rainey suggests that EA 251:11-15 is an in- stance of "a conditional sentence without summa," but neither his translation nor that of Moran (1992: 305) reflects a conditional sentence.

    II 241-44. In this section on the energic forms, the semantics of which are admittedly very hard to de- termine, the asseverative and argumentative nuances are not clearly distinguished by Rainey; both are said to be used for passages "charged with emotional content" (p. 242)/"charged with emotion" (p. 243).

    II 245. Under "Morphology," in the second line, for "2nd f.s." read "3rd f.s."

    II 248. The interchangeability of the negatives ul and la is said to be "a peculiar feature of the EA texts from Canaan"; it is, in fact, widely attested among the peripheral dialects of Akkadian.

    II 249. Add to the translation of EA 299:21-22 "from the cApiru."

    In the translation of EA 119:55-59, for "another tablet" read "another tablet of mine."

    II 250. Rainey cites Moran's well-founded rule of "modal congruence" according to which a purpose clause contains a verb in the same mood (viz., indic- ative or injunctive) as the verb of the preceding clause; unfortunately, no examples of the indicative

    sequence are provided either here or earlier in the discussion of indicative forms.

    Other examples of lu niprus for the lcp precative in Old Assyrian appear in Larsen 1988: 116 text 84a:66, 70: lu nu-bi4-il5-su-nu-ma "let us carry them," lu nu-sf "let us leave." In line 8 from the bot- tom of II 250, the reference to Hecker is incorrect: read "Hecker 1968:128-129, ?77."

    II 251. The translation of EA 289:29 (actually, 27- 29 is quoted) is missing the first clause: "grant to the men of Qiltu all their desire."

    II 254-63. Criticism of Rainey's view that the final -a of most yaqtula forms is the Akkadian ventive marker rather than a Northwest Semitic modal marker has already been voiced above (ad II 202-11). It is worth noting here, however, that if such forms are indeed ventive, we should expect at least a few ex- amples of plural verbs with the ventive allomorph -ni (iprusuni), which the scribes would presumably have learned at the same time as the singular; yet such forms are lacking, the plural counterpart to singular yaqtula being, in fact, the expected tiqtulu, which is also the plural of the apocopate yaqtul, just as in Ugaritic and Arabic. In other words, we find injunctive singulars written both yaqtul and yaqtula, and injunctive plurals written tiqtulii; this is just what we expect if both yaqtul and yaqtula represent NWS modal forms, and contrary to what is expected if yaqtula forms are ventive, viz., plurals written tiq- tuluni. The same considerations apply to the forms of the imperative with final -a, which Rainey (II 265-73) also considers to be ventive (although ad- mitting that the purely Canaanite form ku-na prob- ably contains the NWS modal -a; p. 266): plural imperatives regularly end in -u, while the ventive -uni is quite rare. (It might be added that it is not clear, to me at least, why Rainey is so keen to rid Amarna Canaanite of a form that is both perfectly understandable and, indeed, expected.)

    II 262, 5th line from bottom. For 'ilf,a read 'illma.

    II 272, end. Note that the use of the accusative suffix -anni to express the dative is also attested in contemporary core Middle Babylonian, e.g., ta-ba- an-ni-ma in EA 7:9,12,14 (from Babylon).

    II 298. In the translation of EA 289:37-40, for "verily departed" read "verily departed from me."

    1998 71

  • JOHN HUEHNERGARD

    II 302. Referring to rba-ni-til "I have built" (EA 292:29) and sa-mi-ti7 "I have heard" (EA 362:5), Rainey says that these verbs are "vocalized in ac- cordance with phonetic and thematic considerations prevailing in certain of the NWS languages, viz. the Barth-Ginsberg law"; He goes on to remark that verbs II- and III-guttural take "either qatil or qatul in the suffix conjugation pattern." But rba-ni-til, of course, does not have a guttural radical; further, the qatil pattern exhibited in Hebrew bdantl might, in view of Arabic banaytu and Aramaic banet, be due to a relatively late analogical development on the ba- sis of other III-y verbs that were originally (i - a), such as sattit < *satiy-tl. Finally, note also the Ca- naanite III-guttural form [la]-qa-hu, i.e., /laqahu/, in EA 287:36, with its qatal pattern (quoted by Rainey in II 290).

    II 304, line 1. Add to the end of the translation of EA 252:13-15, "before the king my lord."

    II 306, line 4. For mi-ih-is read mi-hi-is (also, sev- eral other s's on this page are typos for intended s). The qitil/qetil pattern found in this form and the other passive suffix-conjugation examples of mahdsu and rahasu cited here by Rainey may be compared with Ugaritic forms in syllabic transcription that also exhibit vowel harmony around gutturals (Huehner- gard 1987a: 271-73).

    II 310-12. It is extremely unlikely that the various quttil forms in these texts "might reflect the true D stem theme in contemporary Canaanite." A u vowel in the first syllable of active D forms would pro- vide the only instances in Semitic of active suffix- conjugation forms with the vowel pattern u ... i, and would be virtually impossible to explain from a his- torical-comparative Semitic prospective. (The fact that Ugaritic exhibits a D infinitive of the pattern qut- tal is not relevant, since infinitives lack voice.) Thus it is much simpler and more economical to assume that the u-vowel of quttil reflects the Babylonian "suffix-conjugation" (i.e., verbal adjective) pattern base, i.e., purrus. Note that the Amarna causative suffix-conjugation pattern is suqtil, and no one would argue that the initial s of such forms reflects the underlying Canaanite. Again, the form is simply the Babylonian suqtul, altered with a minimum of mor- phological information, viz., the i of the second syllable, to mark that it is intended as a NWS suffix-

    conjugation form. Indeed, the true Canaanite caus- ative suffix-conjugation pattern is hiqtil(a), as attested in hi-ih-bi-e "he has hidden" (EA 256:7; see Rainey, II 315). Several instances of similar phenomena may be noted: 1) in Ics suffix-conjugation forms such as saprdti, only the final -ti, not the ending -dti, is con- sidered NWS, the intervening -a- being part of the Akkadian paradigm (pars-i-ku; see Rainey's com- ments, II 285); 2) In II 288-92, Rainey notes that a few 3fs and 3mp suffix-conjugation forms with a medial vowel, qatVlat and qatVlu, suggest that in the underlying Canaanite "the thematic vowel did not elide" with the addition of the endings -at and -u, un- like in Akkadian, so that in the more common qatlat and qatlu forms, "the usual formation... is by the addition of the suffix ... to the appropriate Akka- dian form after the elision of the thematic vowel"; 3) an unusual passive form such as ta-sap-pdr-ta "you were sent" (EA 102:10) is not taken to reflect an underlying Canaanite pattern; 4) in II 379-80, Rainey notes that a few qattil forms must reflect the true infinitive of the underlying Canaanite dialect, so that quttil infinitives attested in the same letters (puhhir, ussir) must be understood as scribal crea- tions, blends of Babylonian purrus and Canaanite qattil; mu-se-ra in a Jerusalem letter (EA 287:57-58 [sic; not 27-29]) also "conforms to the qattil pat- tern" according to Rainey. Note also, e.g., Rainey's comment about the "occasional use of i vowel pre- fixes instead of prefixes with u in the D and ? stems," viz., that "they are colloquial features, analogic developments from the G stem prefixes, and not Canaanisms" (II 49), and his remark that hybrid forms such as ibassati "can be seen. . . as a testi- mony to [the scribes'] productivity in the generation of analogous forms." See also, finally, Huehnergard 1992: esp. 221 n. 53.

    II 315-16. In II 193, Rainey suggests that the form u-sa-ka in EA 82-29 might be understood as a passive causative (Hp) suffix-conjugation form, i.e., huqtal(a) (hophal; thus, "will be issued for you"); while another interpretation seems more likely to me (see above, ad II 193), the proposed huqtal(a) form should nevertheless be registered here.

    II 317-46. This long chapter, "Suffix Conjugation- Special Hybrids," is particularly important, and Rainey does a very good job of explaining the mor- phology and usage of such unusual creations as i-ba-

    BASOR 310 72

  • A GRAMMAR OF AMARNA CANAANITE

    sa-tu-nu "you are," iz-zi-iz-ti "I stood," in-ne-ep-sa- ta "you became," etc. Although these forms have Akkadian durative or preterite bases, the pronominal subjects suffixed to them mark them as Canaanite suffix-conjugation verbs. Once this is understood, it becomes possible to see why a form such as 3ms izziz or 3mp innepsu, which lack a prefix y- or t- to mark them as prefix-conjugation forms, should be interpreted as the 3ms and 3mp counterparts to izzizti and innepsdta, i.e., also as reflecting Canaan- ite suffix-conjugation forms, or why Rainey is able to make the following statement (p. 321): "Syntac- tically, all the inflected forms of basu [most of which have durative bases, like the example just cited] in these texts . . . function as qtl's."

    II 321, last line. For "When I was you" read "When I was young."

    II 338. It is difficult to imagine what sort of "reflex- ive nuance" the scribes would have associated with the t-forms of riahu "to remain, be left over"; Rainey's alternative explanation that the scribes pre- ferred to use the t-perfect as the bases of weak roots to which to add their Canaanite morphological mark- ers seems much more likely.

    II 365, 7th line from bottom. The phrase "indicative imperative yaqtulu" is presumably a typographical error for "indicative yaqtulu," as we find three lines later.

    II 366, line 10. For "main suffix forms" read "main prefix forms."

    II 367. To the examples of infinitives of II-weak verbs, add ma-ti-ma "to die" (EA 89:38), cited in II 386.

    II 370. The appearance of accusative infinitives of (w)asu in construct "written like genitives" is called "surprising." This is regular, however, in normative Babylonian; see GAG ?64i and Huehnergard 1987b: 188-91.

    II 376. One wonders whether the unusual construct infinitive forms of satu, viz., si-ti and si-te-su, reflect a mixing of III-i and III-e verbs, as seems to have happened in Ugarit Akkadian; see Huehnergard 1989: 54-56.

    II 389. The use of the infinitive to emphasize a finite verb, {pardsum(-ma) iprus/iparras}, while attested in core Old Babylonian, is not really "widely at- tested" except in Mari texts; see Aro 1961: 112-13.

    III 5. It is unlikely that the conjunction indum "be- long[s] to the Babylonian stock," since it is undoubt- edly to be connected with Hurrian undu; these forms occur in Akkadian for the most part in texts exhib- iting Hurrian influence, viz., from Nuzi, Bogazkoy, and Ras Shamra. The initial vowel of indu(m) may be the result of contamination with Akkadian inuma.

    III 7, also III 146-47. It might be mentioned that magal is well attested in Old Babylonian, a further indication of the OB nature of the Akkadian substra- tum that underlies the Canaanite of these letters. It is tempting to compare magal, the etymology of which is obscure (Rainey suggests a Sumerian loanword), with Mehri mekan "much, many."

    III 9-10. The section on the "status absolutus" is rather unsatisfying. It is unclear to me how we might ascertain that the grammatical category "sta- tus absolutus" is "practical[l]y unknown in WS," since apart from Arabic we lack sufficient evidence for the (non-) appearance of case-vowels in early WS dialects.

    III 11, lines 7-8. Transpose "latter" and "former."

    III 21. Cases of the apparent interchangeability of ana and ina are also attested in several peripheral di- alects, and occasionally in Middle Assyrian (Hueh- nergard 1989: 184 with, n. 335).

    Although the preposition arki occurs indepen- dently in a few instances, it is not listed among the simple prepositions here, but rather among the com- pound expressions, pp. 40-41.

    III 24, line 2. For "pp. 236, 295" read "pp. 236, 245."

    III 37. The translation of EA 283:9 given by Rainey to support his contention that KI there is not a logo- gram for itti does not inspire confidence: "Enter in [to], the king my (sic!) lord, behold, take!."

    III 39. The translation of EA 138:65-69 is unclear; see that of Moran 1992: 222.

    73 1998

  • JOHN HUEHNERGARD

    III 42. An example of biri used independently (i.e., without preceding ina/ana/sa) is also attested at Ugarit (Ugaritica 5 no. 81:21-23).

    III 43. On the frequent loss of the preposition in phrases such as ina libbi when they follow sa (thus, sa libbi) in core Akkadian, see GAG ? 115c.

    III 68. In EA 239:11, for it-ta-si read it-ta-si.

    III 70. The last example under adi (EA 364:24-28) should be deleted; Rainey suggests that "the addi- tion of an enclitic -mi to the conjunction may indi- cate that an adverb, 'still, yet,' is intended." Indeed, the following verb is a precative (li-de-mi), which cannot be part of a subordinate clause; thus, adi here must be an adverb.

    III 73. Another possible example of eli as a con- junction is perhaps found in a recently published Byblos letter from Kamid el-Loz (ZA 86 100:13; see pp. 107-8 in the same volume).

    On indum/undu, see above, ad III 5.

    III 74-89. The presentation of the conjunction inima/enima, which is ubiquitous and polysemous in these texts, is exemplary, and the concluding sum- mary (pp. 88-89) insightful and illuminating.

    In III 74, inuma, enuma, line 3, for "1990?" read "1992a." On the adverbial use of i/enuma at Ugarit, see now Huehnergard 1989: 196-97.

    III 92. In the transliteration of EA 254:38-46, the rendering of KI-e as ki-i15 in the first two instances and as ke-e in the third is presumably an oversight.

    III 93. The rarity of kima in the EA letters is all the more surprising given that it is especially a feature of Old Babylonian, which Rainey has shown to be the basic underlying Akkadian matrix of these texts.

    III 94-97. I am not convinced by Rainey's argu- ments that the nonconditional instances of summa mean "since, because" rather than that summa func- tions as a presentation particle (like Hebrew hinne) in such examples, as suggested by Moran.

    III 102. See Huehnergard 1989: 242-43 for a sum- mary of the occurrence of the "u of apodosis" in Akkadian; it is, of course, common in peripheral di- alects but is also attested in Old Assyrian, as Rainey

    notes; note also the following Old Akkadian exam- ple: sum!(us+bi)-ma MAS ANkE-mi gu-ti-u it-ru-u u a-na-ku8 mim-ma u-la a-qd-bi "If (you say), 'The Guti took away the livestock,' I will not (be able to) say anything" JRAS 1932 296:24. For wa introducing a protasis in Eblaite, see Edzard 1984: 111-12, 115.

    It seems to me that to ascribe the widespread use of the "u of apodosis" to Amorite influence is to explain ignotum per ignotius.

    III 116. The reference for agdmi/mdgammi is EA 362:13.

    III 116-119. While the discussions of the adverbs anumma and inanna are helpful, the mention (p. 118) of an instance in which the latter seems to replace the former prompts one to wish that a comparison of the usages of the two had been undertaken.

    III 122. In EA 95:37, for ut ERIN.MES read ut le-q[a] ERIN.MES.

    III 126. It seems unlikely that pa-na-nu in EA 137:75 and 144:34 is used as an adjective, since the form is not declined in either instance (where it would allegedly modify a genitive plural noun); note Moran's translations (1992 ad loc.) of the word as an adverb, "in the past," "before."

    III 126-127. The discussion of sa-an-TI-qd-di-ma in EA 149:27, which Rainey reads sa-an-di-qd-di- ma, is curious: Rainey suggests that the first dental might reflect "partial assimilation of the -t suffix" of NWS sant- "year" or, more likely, that we have here an instance of "MB/Peripheral dissimilation of -tt- > -nd-"; these suggestions are somewhat circular, how- ever, since the reading of TI as di is based on them.

    III 145. In the sentence before the citation of EA 317:10-12, for "It comes" read "The other comes."

    III 155-174. Rainey's analysis and discussion of the presentation particles are very useful. It is unfortu- nate, however, that he has chosen to gloss them with "behold," an obsolete word in English the use of which obscures the true function of such particles and precludes more nuanced translations of the passages in which they occur. One also misses a comparison with Hebrew hinne, with which these particles share many similarities in their syntax and semantics.

    74 BASOR 310

  • A GRAMMAR OF AMARNA CANAANITE

    III 179. Rainey is undoubtedly correct to suggest that the writings SAHAR-ra in EA 141 are to be read SAHAR.RA, with the second sign merely function- ing as an extension of the logogram; for similar examples at Ugarit, see Huehnergard 1989: 74.

    III 190. In the transliteration of EA 254:38-42, add sar-ru at the end of line 39.

    III 191. In EA 85:66, for IHa-za-nu read 1 ha-za-nu.

    III 196. In EA 60:31, for yi-ip[-q]i-[n]i read yi- ip[-q][-id-[n]i.

    III 210. Rainey claims that "all Assyrian dialects have only la, never ul" as negatives; see, however, Hecker 1968: 183-84 ?105, for Old Assyrian ula alongside Id.

    III 214. The distribution of negative adverbs in questions, with ld co-occurring with interrogative adverbs and ul otherwise, is true not only of Mari texts, but of Old Babylonian texts in general; see GAG ? 153c,f.

    III 223. It might be noted that the Babylonian dis- tinction between {ul + durative} for statement (ul illak "he will not go") and {Id + durative} for

    injunction (la illak "he may not go") is redundant in the grammar of Canaanite Amarna texts because the verb itself is morphologically marked as indicative (yaqtulu) or injunctive (yaqtul/yaqtula).

    III 224. Rainey claims that "the use of ul to negate the volitive is unique in all of cuneiform literature"; there are also examples, however, at Ugarit and Emar (see Huehnergard 1989: 239; Ikeda 1995: 111).

    III 227. Note that -ma is also absent as a conjunc- tion in Ugarit Akkadian; further, it is rare in the core dialect of Middle Assyrian.

    III 229. It is suggested here that -ma on an infinitive functioning as a finite verb adds "special emphasis"; in II 384, however, it is noted that with one excep- tion all such infinitives are followed by -ma or -mi, and claimed, with greater plausibility, that the en- clitics mark the infinitive in those instances as the logical predicate (see also III 266-67).

    III 251-280. In this last chapter, "Topic and Com- ment," Rainey provides an insightful summary of Gianto's monograph on word order in the Byblos letters (Gianto 1990), and includes many examples from letters from other sites as well.

    REFERENCES

    Note: Abbreviations in citations of Akkadian texts follow CAD. AbB 3 = Frankena, R.

    1968 Altbabylonische Briefe. Vol. 3: Briefe aus der Leidener Sammlung (TLB IV). Leiden: Brill.

    AHw = von Soden, W. 1956- Akkadisches Handworterbuch. 3 vols. Wies- 1981 baden: Harrassowitz.

    Aro, J. 1955 Studien zur mittelbabylonischen Grammatik.

    Studia Orientalia 20. Helsinki: Societas Orien- talis Fennica.

    1957 Glossar zu den mittelbabylonischen Briefen. Studia Orientalia 22. Helsinki: Societas Orien- talis Fennica.

    1961 Die akkadischen Infinitivskonstruktionen. Stu- dia Orientalia 26. Helsinki: Societas Orientalis Fennica.

    AuOr 5 = Arnaud, D. 1987 La Syrie du moyen-Euphrate sous le protec-

    torat Hittite: contrats de droit prive. Aula Ori- entalis 5: 21-41.

    Buccellati, G. 1976 The Case against the Alleged Akkadian Plural

    Morpheme -dnu. Afroasiatic Linguistics 3, no. 2: 28-30.

    CAD 1956- The Assyrian Dictionary of the Oriental In-

    stitute of the University of Chicago. Chicago: Oriental Institute.

    CH = Bergmann, E. 1953 Codex Hammurabi, textus primigenius. 3rd ed.

    Rome: Pontificium Institutum Biblicum. Driver, S. R.

    1892 A Treatise on the Use of the Tenses in Hebrew. 3rd ed. Oxford: Clarendon.

    Durham, J. W. 1976 Studies in Bogazkoy Akkadian. Ph.D. Disserta-

    tion, Harvard University. EA = Knudtzon, J. A.

    1907- Die El-Amarna-Tafeln mit Einleitung und 1915 Erlduterungen. Anmerkungen und Register

    bearbeitet von 0. Weber und E. Ebeling.

    75 1998

  • JOHN HUEHNERGARD

    2 volumes. Leipzig: Hinrichs; reprint Aalen: Zeller, 1964.

    Edzard, D. 0. 1984 Zur Syntax der Ebla-Texte. Pp. 101-16 in Stud-

    ies on the Language of Ebla, ed. P. Fronzaroli. Quaderni di Semitistica 13. Florence: Istituto di Linguistica e di Lingue Orientali.

    1985 Die 3. Person m. P1. TIPRUSU im Altakkadi- schen von Mari. Pp. 85-86 in Miscellanea Babylonica, Melanges offerts a Maurice Birot, eds. J.-M. Durand and J. R. Kupper. Paris: Edi- tions Recherche sur les Civilisations.

    Emar = Arnaud, D. 1985- Recherches au pays dAsftata. Emar VI. 4 vols. 1987 Paris: Editions Recherche sur les Civilisations.

    GAG = von Soden, W. 1995 Grundriss der akkadischen Grammatik. 3., er-

    ganzte Auflage unter Mitarbeit von W. R. Mayer. Analecta Orientalia 33. Rome: Pon- tificio Istituto Biblico.

    Gianto, A. 1990 Word Order Variation in the Akkadian of Byb-

    los. Studia Pohl 15. Rome: Pontificio Istituto Biblico.

    Gordon, C. H. 1965 Ugaritic Textbook. 3 vols. Analecta Orientalia

    38. Rome: Pontificium Institutum Biblicum. Hecker, K.

    1968

    Hetzron, R

    Grammatik der Kiiltepe-Texte. Analecta Ori- entalia 44. Rome: -Pontificium Institutum Biblicum.

    [. 1973- The Vocalization of Prefixes in Semitic Active 1974 and Passive Verbs. Melanges de l'Universite

    Saint-Joseph 48: 35-48. Huehnergard, J.

    1987a Ugaritic Vocabulary in Syllabic Transcription. Harvard Semitic Studies 32. Atlanta: Scholars.

    1987b Three Notes on Akkadian Morphology. Pp. 181-93 in Working With No Data: Semitic and Egyptian Studies Presented to Thomas 0. Lamb- din, ed. D. M. Golomb. Winona Lake, IN: Eisenbrauns.

    1989 The Akkadian of Ugarit. Harvard Semitic Stud- ies 34. Atlanta: Scholars.

    1992 Historical Phonology and the Hebrew Piel. Pp. 209-29 in Linguistics and Biblical He- brew, ed. Walter R. Bodine. Winona Lake, IN: Eisenbrauns.

    Idrimi = Smith, S. 1949 The Statue of Idri-mi. London: British Institute

    of Archaeology in Ankara. Ikeda, J.

    1995 A Linguistic Analysis of the Akkadian Texts from Emar: Administrative Texts. Ph.D. Dis- sertation, Tel Aviv University.

    Izre'el, S. 1987 The Complementary Distribution of the Vow-

    els e and i in the Peripheral Akkadian Dialect of Amurru-A Further Step towards our Un- derstanding of the Development of the Amama Jargon. Pp. 525-41 in Proceedings of the Fourth International Hamito-Semitic Congress, eds. H. Jungraithmayr and W. W. Miiller. Cur- rent Issues in Linguistic Theory 44; Philadel- phia: Benjamins.

    1990 A New Translation of the Amama Letters. (Re- view of Moran 1987.) Bibliotheca Orientalis 47: 577-604.

    1991 Amurru Akkadian: A Linguistic Study. 2 vols. Harvard Semitic Studies 40, 41. Atlanta: Scholars.

    JRAS 1932 = Smith S. 1932 Notes on the Gutian Period. Journal of the

    Royal Asiatic Society 1932: 295-308. KBo 1 = Figulla, H. H., and Weidner E. F.

    1916 Keilschrifttexte aus Boghazkoi, Vol. 1: Wissen- schaftliche Veroffentlichungen der Deutschen Orient-Gesellschaft 30/1. Leipzig: Deutsche Orient-Gesellschaft.

    KL 72:600 = Wilhelm, G. 1973 Ein Brief der Amarna-Zeit aus Kamid el-Loz.

    Zeitschrift fur Assyriologie 63: 69-75. Larsen, M. T.

    1988 Old Assyrian Texts, Nos. 71-98. Pp. 92-143 in Cuneiform Texts in the Metropolitan Museum of Art. Vol. 1: Tablets, Cones, and Bricks of the Third and Second Millennia B.C., ed. I. Spar. New York: Metropolitan Museum of Art.

    Mayer, W. 1971 Untersuchungen zur Grammatik des Mittel-

    assyrischen. Alter Orient und Altes Testament 2. Kevelaer: Butzon & Bercker.

    Moran, W. L. 1950 A Syntactical Study of the Dialect of Byblos as

    Reflected in the Amama Tablets. Ph.D. Disser- tation, Johns Hopkins University.

    1960 Early Canaanite yaqtula. Orientalia 29: 1-19. 1975 The Syrian Scribe of the Jerusalem Amarna

    Letters. Pp. 146-66 in Unity and Diversity, eds. H. Goedicke and J. J. M. Roberts. Baltimore: Johns Hopkins University.

    1987 Les lettres d'El Amarna. Correspondance dip- lomatique du pharaon. Litteratures Anciennes du Proche-Orient 13. Paris: ltditions du Cerf.

    1992 The Amarna Letters. Baltimore: Johns Hopkins University.

    PBS 13 = Legrain, L. 1922 Publications of the Babylonian Section of the

    Museum of the University of Pennsylvania, Vol. 13: Historical Fragments. Philadelphia: University Museum.

    76 BASOR 310

  • A GRAMMAR OF AMARNA CANAANITE

    RA 77 = Huehnergard, J. 1983 Five Tablets from the Vicinity of Emar. Revue

    d'Assyriologie 77: 11-43. Rainey, A. F.

    1989- A New Translation of the Amarna Letters- 1990 after 100 Years. Archiv fur Orientforschung

    36-37: 56-75. 1991- Is There Really a yaqtula Conjugation Pattern 1993 in the Canaanite Amarna Tablets? Journal of

    Cuneiform Studies 43-45: 107-18. Schub, M. B.

    1974 A Note on the Dialect of Tamim and Barth's Law. Zeitschrift der Deutschen Morgenldndi- schen Gesellschaft 124: 13-14.

    Soden, W. von 1973 Der akkadische Subordinativ-Subjunktiv. Zeit-

    schrift fir Assyriologie 63: 56-58. Testen, D.

    1992 A Trace of Barth's Preradical *i in Akkadian. Journal of Near Eastern Studies 51: 131-33.

    TT = Hrozny, F 1904 Keilschrifttexte aus TaCannek. Pp. 113-22 in

    E. Sellin, Tell TaCannek. Denkschriften der kai-

    serlichen Akademie der Wissenschaften in Wien, Phil. -Hist. Klasse 50/4. Vienna: Akademie der Wissenschaften.

    Ugaritica 5 = Nougayrol, J., et al. 1968 Ugaritica 5: Nouveaux textes accadiens, hour-

    rites et ugaritiques des Archives et Biblio- theques prives d'Ugarit, commentaires des textes historiques. Mission de Ras Shamra 16. Paris: Imprimerie Nationale.

    VAS 8 = Ungnad, A. 1909 Vorderasiatische Schriftdenkmadler, Vol. 8: Alt-

    babylonische Urkunden. Leipzig: Hinrichs. Westenholz, A.

    1978 Some Notes on the Orthography and Grammar of the Recently Published Texts from Mari. Bibliotheca Orientalis 35: 160-69.

    Whiting, R. M., Jr. 1987 Old Babylonian Letters from Tell Asmar. Assyr-

    iological Studies 22. Chicago: Oriental Institute. ZA 86 = Huehnergard, J.

    1996 A Byblos Letter, Probably from Kamid el-Loz. Zeitschrift fiir Assyriologie 86: 97-113.

    1998 77

    Article Contentsp.59p.60p.61p.62p.63p.64p.65p.66p.67p.68p.69p.70p.71p.72p.73p.74p.75p.76p.77

    Issue Table of ContentsBulletin of the American Schools of Oriental Research, No. 310, May, 1998Front Matter [pp.34-78]Ain Ghazal "Monumental" Figures [pp.1-17]Greater Canaan: The Implications of a Correct Reading of EA 151:49-67 [pp.19-24]The Use of Ivories as Interpreters of Political History [pp.25-33]Achish-Ikausu in the Light of the Ekron Dedication [pp.35-37]Early Islamic Settlement in the Southern Negev [pp.39-57]Review ArticleA Grammar of Amarna Canaanite [pp.59-77]

    Book Reviewsuntitled [pp.79-81]untitled [pp.81-82]untitled [pp.82-85]untitled [pp.85-89]untitled [pp.89-91]untitled [pp.91-92]untitled [pp.92-94]untitled [pp.94-96]

    Back Matter