Printer Friendly

Semitic linguistics in the new millennium.


It is always a fitting tribute to a scholarly field when a number of its most distinguished practitioners gather to reflect on its growth and historical development. The work under review contains eighteen state-of-the-art essays by well-known specialists on various aspects of Semitic linguistics. These invited articles were originally presented at an international symposium held at Tel Aviv University in January 1999, sponsored by the University and many external sources. The basic questions on the table were: where had Semitic linguistics been in the course of the twentieth century and before, and where would it or should it be headed in the twenty-first? The tasks for the participants necessarily involved a thorough digestion of the numerous leading publications, and all the papers suggest intimate familiarity with the enormous relevant bibliography in each subfield. It is my purpose here to present commentary on each of the essays intended to benefit this growing and dynamic area by offering an examination of the big picture as well as the microscopic details which science always involves.

The volume is conveniently subdivided into six parts: methodologies, overviews, the wider Afroasiatic perspective, languages in contact, dialectology of the modern languages, and an all-encompassing section entitled "Broadening Our Horizons." The final segment contains four book reviews on recent publications, two by the editor and one each by David Testen and Aaron Dolgopolsky. We shall examine the book reviews as well, which are more along the lines of review articles.

The editor's introduction describes, among other interesting topics, the Chomskyan and Russian Schools of Semitic Linguistics (pp. 13-20). The former is emphasized in a quoted "call for papers" for a conference held in Fez, Morocco in March 1999. It is hyperbolic to claim, as that "call for papers" does, that "today most Chamito-semitic (sic) studies are inspired by the generative trend" (p. 14). There are, of course, many outstanding Semitists working in nongenerative frameworks, and good work knows no theoretical model--only solid argumentation and accuracy. Consider in this regard the 129 articles published in Semitic Studies in Honor of Wolf Leslau on the Occasion of his 85th Birthday, ed. Alan S. Kaye (Wiesbaden: Harrassowitz, 1991), very few of which are penned in the framework of generative-transformational theory.

Highlights of the Russian School discuss the work of the late Igor M. Diakonoff (d. 1999) (who contributed an important paper to the aforementioned Leslau Festschrift), Qonstantin I. Marogulov, Nicholas Marr, and N. V. Yushmanov. Diakonoff was undoubtedly the most outstanding linguist in the Afroasiatic (2) and Semitic fields. The fact is that Marr is little read today for numerous reasons, not the least of which is that his work was thoroughly discredited, and few Semitists have read Marogulov or even know the name, with the possible exception of a handful of NA specialists. Since Diakonoff produced a number of pupils who became active scholars, one can legitimately speak of a Diakonoff School, (3) and names such as Alexandra Y. Aikhenvald, Viktor Porkhomovsky, Leonid Kogan, and Alexander Militarev readily come to mind.


1. Gideon Goldenberg's "Semitic Linguistics and the General Study of Language" (pp. 21-41) is, in many ways, a sequel to his "The Contribution of Semitic Languages to Linguistic Thinking" (Jaarbericht van het Vooraziatisch-Egyptisch Genootschap Ex Oriente Lux 30 [1987-88]: 107-15). This essay makes some profound statements which are right on target, such as: "Linguistic Studies will be doomed to be banal without vast and profound knowledge of each language which one investigates ..." (p. 22), and "many studies, however, relating to Semitic languages refrain from original linguistic thought, either reposing on achievements of the past, or--which is by no means better--embracing servilely whatever theory is in vogue" (p. 23). It is refreshing to read down-to-earth critical statements like these about an entire academic enterprise, and we are grateful for his straightforwardness and candor, not to mention his astuteness.

This essay also contains erudite observations on the gerund(ive) = gerondif = converb "any adverbial derived from a verb" (pp. 28-30), and Suffixaufnahme or "double case," in which a possessor in the genitive agrees with the nominal head (pp. 32-37). According to Goldenberg, "double case" is attested in Sumerian, Hurrian, Urartian, Basque, Kartvelian, Cushitic, and other languages, and it appears likely that linguists will become increasingly interested in this fascinating topic (p. 33).

2. Joseph L. Malone was faced with the arduous task of commenting on "The Chomskian School of Semitic Linguistics" (pp. 43-55). In fact, he writes that "the very first spade work ... convinced me of the impossibility of what I had at first hoped to develop" (p. 43). First of all, one may legitimately ask if there is such a School, or whether it would be more accurate to speak of Chomsky-inspired work, or perhaps even better, work done within the Chomskyan Paradigm? Secondly, there is most definitely a Chomskyan School of Linguistics (and perhaps one of Political Science as well), but how does the former differ from the designated School of Semitic Linguistics? Are these not one and the same School, except that the one in Malone's title deals exclusively with the Semitic languages? Thus, I think it appropriate to reword his title to the latter of the aforementioned choices.

The author decided, in the final analysis, to obtain most of the material for discussion by examining the articles on Semitic languages which were published in Linguistic Inquiry (from its inception in 1970 to the end of the millennium), the linguistics journal of MIT, Chomsky's home base. This was a reasonable approach to the task at hand.

While discussing research dealing with modern Arabic dialects, I am surprised to see Maltese included among them. Although it is certainly an Arabic dialect from the diachronic point of view, for many reasons it is not accurate to consider it an Arabic dialect synchronically (p. 45). I have discussed the situation of Maltese (see, e.g., my "Arabic," in The World's Major Languages, ed. Bernard Comrie [Oxford: Oxford Univ. Press, 1987], 664-85), and this perspective has, I believe, gained general acceptance.

Malone's article singles out for discussion a number of influential authors who have focused their research on Heb., including a young Noam Chomsky in his Master's thesis The Morphophonemics of Modern Hebrew (Univ. of Pennsylvania, 1951), published 28 years later (New York: Garland, 1979). One can make a solid case that transformational-generative grammar, and generative phonology in particular, begins in 1951, six years before the publication of Syntactic Structures (The Hague: Mouton, 1957), the customarily cited starting date.

3. Jo Ann Hackett's "The Study of Partially Documented Languages" (pp. 57-75) is a comprehensive survey of epigraphic Northwest Semitic languages, including Phoenician, Punic, Moabite, Ammonite, Edomite, Aramaic, and Ugaritic. In a commentary about Amarna Akkadian, Hackett affirms that she agrees with Anson Rainey's view that this variety must be understood "as a mixed or 'fused language'" (p. 62). However, it must be noted that many (dare I say most?) languages are mixed in the sense that they show influence as a result of language contact. Is there a special methodology for dealing with these languages, as opposed to nonmixed ones? I think not. Does she mean that this particular dialect is so fraught with contact phenomena that we must be especially careful by paying particular attention to them?

In the discussion of the classification of Ugaritic (p. 64), Hackett seems unaware of my essay rejecting its positioning as a Canaanite language ("Does Ugaritic Go with Arabic in Semitic Genealogical Sub-Classification?" Folia Orientalia 28 [1991]: 115-28). Thus, I am in the same camp as the others cited by the author: Robert Hetzron, Rainer Voigt, John Huehnergard (who also discusses the classification of Ugaritic in this volume, p. 130), Alice Faber, Anson Rainey, and the pioneer of this theory, Albrecht Goetze, in a 1941 article in Language.

Hackett, as was true with Goldenberg above, has some very strong opinions about the field, with which I am in firm agreement (p. 68):
 I would go so far as to say, in fact, that the majority of what is
 written about epigraphic Northwest Semitic languages and dialects is
 the work of amateurs. It gets published because many editors know no
 more than these authors, and because the publish-or-perish ethic has
 produced so many new venues. The result of these combined
 factors--desperation to find something new to say, to find something
 to publish; inadequate knowledge of biblical Hebrew and of Aramaic;
 lack of linguistic training; venues and editors who have no training
 in judging the quality of this kind of work--can be more clutter
 than valuable ideas.

Although statements like these may seem out of place or inappropriate to some, Hackett calls it as she sees it, and harsh words are sometimes necessary for the betterment of a field. One sometimes wonders why so many scholars are reluctant to call a spade a spade, or even loath to let others do so.

Corrigenda: On p. 73 of the author's bibliography, it should be noted that Sabatino Moscati was not the sole author, but rather the editor and one of four coauthors of An Introduction to the Comparative Grammar of the Semitic Languages (Wiesbaden: Harrassowitz, 1964).

The introduction to the section on "Overviews" by the editor is a fine summary of the five essays it contains (pp. 79-84). However, the first name of the famous University of Chicago grammatologist and Assyriologist, Ignace J. Gelb, is misspelled (p. 79).

4. Peter T. Daniels' "The Study of Writing in the Twentieth Century: Semitic Studies Interacting with Non-Semitic" (pp. 85-111) is a masterful treatment by a student of Gelb's and one of today's most outstanding grammatologists. It is astounding to ponder Daniels' point made at the beginning of his article--that, had Americanists known of the cuneiform decipherment techniques employed by Edward Hincks between 1846-52 (pp. 85-86), the Mayan inscriptions "could have been read much sooner than the full century later ..." (p. 86).

This article, like others in the collection, is valuable in that its evaluation of numerous publications sets the record straight. Calling Florian Coulmas's Blackwell Encyclopedia of Writing Systems (1996) "a great disappointment" is probably an understatement, while noting that Peter T. Daniels' and William Bright's edited The World's Writing Systems (1996) "is meant to provide information on the ways writing systems represent language, and there is certainly no intent or claim of historical thoroughness" (p. 92) does not, in my opinion, go far enough in its praise. Suffice it to mention that the latter handbook remains the most comprehensive treatise on this subject up to the present time.

Daniels comments on Gelb's achievements by offering an enlightened perspective on the latter's Principle of Uniform Development, which the author calls his "most important hypothesis" (p. 93). According to this hypothesis, West Semitic writing is syllabic--not alphabetic--about which Daniels has this to say: "It is this principle that leads him [Gelb] to gloss over the scripts of Ethiopia and India and to misrepresent Korean" (p. 98). My own viewpoint has always been that West Semitic (Phoenician, Hebrew, Aramaic, etc.) writing systems are alphabetic for the following basic reason. A West Semitic scribe who wanted to write a CVC syllable would have used two graphemes, not one. Consider the pausal form for "hand": yad, written <yd>. Why would the mere addition of a vowel diacritic (e.g., a games in Heb.) cause the system to switch from syllabic to alphabetic? Daniels is, of course, correct to maintain that Gelb considered vocalized Hebrew and Aramaic alphabetic (p. 96).

Daniels is also fair in criticizing his teacher's view that Mayan writing was not phonetically based (expressed in A Study of Writing [Chicago: Univ. of Chicago Press, 1952], 56). It turns out that Gelb was wrong on this point. (4)

Let me turn to one area of disagreement with the author. Daniels asserts that it is "now uncontroversial" that the Greek invention of vowel letters was "an accident of the phonemic structure of Phoenician versus Greek" (p. 103). One must surely recognize the creative leap required to take the grapheme aleph for the glottal stop and adapt it for the vowel alpha. In my view, that is quite an innovation, which even the author refers to as "the Greek innovation of vowels" (p. 101). I believe that the Greeks should be properly credited for this remarkable achievement.

5. John Huehnergard's "Comparative Semitic Linguistics" (pp. 119-50) is a thorough discussion of the research of a plethora of authors (with 14 pages of bibliography). It is interesting to reflect, now years later, on the two schools of Assyriology mentioned: (1) the Benno Landsberger School, which "explicitly rejected the study of Akkadian within the framework of Semitic linguistics" (p. 122), and (2) the Comparative School, as practiced by, e.g., Wolfram von Soden (Landsberger's student), Albrecht Goetze, Ignace J. Gelb, and others. Both approaches have produced significant works, to be sure. Incidentally, we are never told why Landsberger, who knew many languages, rejected the Comparative Semitics approach. Perhaps, if I may hazard a guess, he was of the opinion that it would be too easy to misuse the Arabic and Hebrew lexicons and project back to Akkadian lexical items and phrases which, in fact, never existed. Indeed, one can cite as a parallel within the Semitic field that a few Ugaritologists and Hebraists have been accused of overusing (read: misusing) the vastness of the Arabic lexicon. In this regard, see, e.g., John Kaltner, The Uses of Arabic in Biblical Hebrew Lexicography (Washington, D.C.: Catholic Univ., 1996), and my review in Hebrew Studies 39 (1998): 196-99.

Turning to the important matter of linguistic classification and subclassification, I agree with Huehnergard's rejection of Edward Ullendorff's (1961) position: "Classification is harmless, unobjectionable, and at times even useful if limited to describing present-day habitat and the prevailing geographical circumstances, but it becomes positively dangerous, i.e., obscuring rather than illuminating, if meant to explain genetic connexions" (p. 130). What would have prompted Ullendorff, such a well-known comparativist, to launch this scathing attack on genetic subgrouping, one of the most fascinating and fruitful aspects of comparative-historical linguistics? Huehnergard is certainly right when he proclaims: "Classification should be the first factor in our reconstruction of Proto-Semitic and thus of early Semitic linguistic history" (p. 132). However, even though classification and subclassification are necessary ingredients in genetic linguistics, I do not believe, contra the author's opinion, that it is necessarily erroneous to invoke the hypothesis "that the choice of isoglosses on which subgroupings are based is subjective or even arbitrary" (ibid.). Using the geminate verb, as the author does, as evidence of a shared ancestry, and including Sabaean in Central Semitic for this reason, is not necessarily incontrovertible (p. 133), as we are led to believe, since, to use the Sapirian term, languages have a drift of their own, and parallel development is a powerful force with which to reckon. Let me explain precisely what I mean by way of illustration. The fact that Akk. (a)ninu "we" parallels Chadian and Nigerian Ar. anina "we" does not mean that these two languages should be lumped together. All it means it that compensatory lengthening is at work in each case because both languages independently lost the pharyngeals. These historical sound changes reflect typological tendencies and parallel developments at work. They do not constitute proof of genetic inheritance from an earlier prototype language.

Concerning Huehnergard's pronouncement that PS has glottalized stops and the impact thereof on IE Studies (p. 136), I see no evidence that there was any such impact when Paul J. Hopper first formulated the glottalized stop (ejective) hypothesis for PIE ("Glottalized and Murmured Occlusives in Indo-European," Glossa [1973] 7:2: 141-66). There are no references in Hopper's bibliography to the writings of any Semitist, nor is there any mention of any Semitic language. His discovery was a reinterpretation of the traditional view of the PIE stop system, and rests upon the foundations of advances in phonological typology and language universals.

Let us examine the author's point concerning Akk. kbt "heavy," which is cognate with kbd in all the other languages. He states: "... which of these is the original must be decided, if it can be decided, not by the number of languages on either side of the ledger but rather by considerations of the typology of sound change and of the tendencies and constraints in the formation of Semitic roots" (p. 132). First of all, it can certainly be decided. Secondly, the PS root should be reconstructed with the final voiced stop, in my opinion. There are some definite parallels to this phenomenon in Arabic dialectology. Consider CAr. sahhao "beggar" > sahhat in ECA, which is paralleled by CAr. nabio "wine" > ECA nibit. Of course, the voiced interdental fricative first shifted to a voiced stop (theoretically speaking), and then devoiced in final position--an environment often triggering stop (plosive) devoicing (witness the comparable situations in Germanic and Slavic).

I would like to suggest, contra the author's aforementioned pronouncement, that numbers sometimes do count. Let us consider the PS word for "dog": Akk. kalbu(m), Aram. kalba, Ar. kalbun, Old South Arabian-Ethiopic kalb, and Heb. k[epsilon]l[epsilon]b = [k[epsilon]l[epsilon]v]. It is obvious that the Heb. should be explained as coming from the PS *kalb-, since all of the other languages have this form. It would be extremely difficult to argue the reverse position that all five other languages diverged from the original Heb. bisyllabic prototype.

Corrigenda: p. 119 states that "the lamented Robert Hetzron's final offering was his edited work, The Semitic Languages (London: Routledge, 1997)." This is not true. His last work, published posthumously, is coauthored with Berhanu Chamora: Inor (Munich: Lincom Europa, 2000). P. 132 has a word missing: "A trivial phonological example is [the] root meaning 'heavy' ..." On p. 135, we note the use of "the modern colloquial Arabic speech forms." Why not refer to "Arabic dialects," as is customarily done?

6. Geoffrey Khan's "The Study of Semitic Syntax" is a review of the recent as well as older syntactic literature written by one of the few experts working in this field today (see my review of his Studies in Semitic Syntax, JAOS 111 [1989]: 135-37). After consideration of the Jewish and Arab grammatical traditions (pp. 152-53) and such names as Joseph Ben Noah and Sibawayhi, the following scholars are discussed, among others: Ewald, Gesenius, Noldeke, Dillmann, Praetorius, Brockelmann, Wright, and Reckendorf. The modern period includes mention of the work of Joseph H. Greenberg, Joshua Blau, Gideon Goldenberg, Olga Kapeliuk, M.M. Bravmann, Irene Garbell, Francis I. Andersen, A. F. L. Beeston, and R. J. Williams. Quite interestingly, Khan demonstrates that "in some respects the study of Semitic syntax has come full circle and taken on some features of the earliest treatments of Semitic syntax in the Middle Ages" (p. 164).

Corrigenda: It should be noted that Talmy Givon's last name is consistently misspelled without the accent mark throughout.


7. Michael O'Connor's "Semitic Lexicography: European Dictionaries of Biblical Hebrew in the Twentieth Century" (pp. 173-212) begins with the following proclamation, which is difficult to prove: "Semitic lexicography is a slow-moving field, but it moves more rapidly than other branches of Semitic linguistics" (p. 173). Although aware of the difference between lexicography and lexicology, the author asserts that he does not make use of this distinction in his essay (p. 175), being more concerned, I suppose, with the actual dictionaries which scholars have at their disposal for research in the classical Semitic languages. All Hebraists and most Semitists should, I think, be familiar with the standard Biblical Hebrew dictionaries (pp. 187-91)--from the pioneering Francis Brown, S. R. Driver, and Charles A. Briggs, A Hebrew and English Lexicon of the Old Testament with an Appendix Containing the Biblical Aramaic (1907) (based on Wilhelm Gesenius' nineteenth-century lexicon). Wilhelm Gesenius and Frants Buhl, Hebraisches und Aramaisches Handworterbuch uber das Alte Testament (1915), and the first edition of the Lexicon in Veteris Testamenti Libros by Ludwig Koehler and Walter Baumgartner (1953), written in German and, as O'Connor so aptly puts it, "a sort of English" (p. 188), as well as the third edition of this work, coauthored by Johann Jakob Stamm et al., Hebrew and Aramaic Lexicon of the Old Testament, translated by M. E. J. Richardson (Leiden: E. J. Brill, 1994-2000).

O'Connor is enthusiastic about one of the two newer lexicons: the Madrid Diccionario biblico hebreo-espanol, edited by Luis Alonso Schokel (1994), while he is less flattering towards the Sheffield Dictionary of Classical Hebrew, edited by David J. A. Clines, a project now about half completed (1993-). He puts it as follows (pp. 203-4):
 The 1990s lexica have, I think, gone too far in searching for a
 stripped-down approach that may be linguistic in impulse, though not
 necessarily in method or result. The Madrid dictionary has the
 advantage that the product is stripped down as well. The Sheffield
 dictionary, with its pinched sense of the linguistic enterprise, is
 bloated in contrast to any other dictionary of Biblical Hebrew.

Let me take up one lexicographical detail. O'Connor, in discussing the Piel of dbr "to destroy," mentions the Sheffield-listed cognate in Akk. dabaru "to overthrow" (pp. 202-3). Koehler and Baumgartner (p. 199) cite the Akk. as meaning "push back" and also Akk. dabru "outrage." Further, they cite Ar. dabarahu "he followed behind his back." All of these aforementioned meanings are far removed from "destroy." Is it merely coincidental that Ar. has dammara (with the interchange of labials via dissimilation of the initial voiced stop) meaning "destroy"?

8. Baruch Podolsky's "A Selected List of Dictionaries of Semitic Languages" (pp. 213-21) mentions under the general and multilingual category (p. 213) V.E. Orel and O. V. Stolbova, Hamito-Semitic Etymological Dictionary: Materials for a Reconstruction (Leiden: E. J. Brill, 1995) with the following comment: "Hardly useful for a Semitist" (ibid.). See my review in BSOAS 69 (1997): 365-67. Along these lines, why does the author not mention Christopher Ehret, Reconstructing Proto-Afroasiatic (Proto-Afrasian)? See my review in the Canadian Journal of Linguistics 41 (1996): 278-82.

For Amharic, the author cites Wolf Leslau's Concise Amharic Dictionary (Wiesbaden: Harrassowitz, 1976), but he should also have included that author's larger 1,500-page English-Amharic Context Dictionary (Wiesbaden: Harrassowitz, 1973).

Under "Other Ethiosemitic" dictionaries, the author should have mentioned Wolf Leslau's Zway: Ethiopic Documents: Grammar and Dictionary (2001). See my review in Word 53 (2002): 234-38.

While I am delighted to see my two dictionaries of Nigerian Arabic mentioned (1982 and 1986, p. 215), I am disappointed not to see some others. Chief among these are: Hamdi A. Qafisheh, NTC's Gulf Arabic-English Dictionary (Chicago: NTC Publishing Group, 1997) (see my review in JNES 60 [2001]: 313-14); Hamdi A. Qafisheh in consultation with Alan S. Kaye, NTC's Yemeni Arabic-English Dictionary (Chicago: NTC Publishing Group, 2000); Jeffrey Deboo, Jemenitisches Worterbuch (Wiesbaden: Harrassowitz, 1989); Richard S. Harrell, ed., A Dictionary of Moroccan Arabic: Arabic-English (Washington, D.C.: Georgetown Univ. Press, 1996); and Harvey Sobelman and Richard S. Harrell, eds., A Dictionary of Moroccan Arabic: English-Moroccan (Washington, D.C.: Georgetown Univ. Press, 1963).

Corrigenda: p. 214: J. Milton Cowan, the editor of the English version of Hans Wehr's A Dictionary of Modern Written Arabic should be J Milton Cowan; p. 219: Haifa is correct for Hafia. The list of Hebrew dictionaries duplicates information contained in O'Connor's paper.


9. Two contributors have chosen to focus on the wider connections of Semitic within the AA phylum. First, Helmut Satzinger's "The Egyptian Connection: Egyptian and the Semitic Languages" (pp. 227-61), using an 11-page bibliography (pp. 250-61), surveys the evidence--in phonology, morphology, morphosyntax, and lexicon--which demonstrates the close affinity of Egyptian to Semitic. While I can agree with the author's assessment that Egyptian cognates "are not very conspicuous" in comparative AA lexicons, I must disagree with his pronouncement that "... Egyptian is a single language whereas the other branches are--with the exception of Berber--groups of numerous languages" (p. 233). Although many have written from the viewpoint that Berber is one language, the truth of the matter is that one should speak of the Berber language family, which consists of several Berber languages and dialects.

The lexical comparisons in the 100-word Swadesh list are striking: e.g., PS and Egyptian *mwt "die." However, I must confess to being skeptical of PS *sm' "hear" and Egyptian sdm via metathesis (p. 233), as well as the results of the "new phonetics" (ibid.) proposed by Otto Rossler, where PS oub(v)b "fly" (noun) = Egyptian 'ff. I would think that PS *'wf "to fly" (verb) is cognate, as is noted by the author: "... comparatists continue to claim that a regular correspondence between Semitic '(and [TEXT NOT REPRODUCIBLE IN ASCII]) and Egyptian ' is beyond any doubt" (p. 234).

Satzinger has presented the evidence that traces of case endings in Egyptian point to the old Semitic system of nominative in -u and genitive in -i, but absolute (not accusative) in -a (p. 237). His arguments are convincing, but opposed to the opinions of two other noted Egyptologists: John Callender and Antonio Loprieno, who have claimed that Proto-Egyptian had the Semitic accusative in -a (ibid.).

It is odd that the names of three prolific comparativists in Egypto-Semitic Studies are conspicuously absent: Cyrus H. Gordon, William A. Ward, and Carleton T. Hodge, yet we find mention of a little-known (1936) article on Egypto-Semitics by Shmuel Yeivin.

10. Rainer Voigt's "The Hamitic Connection: Semitic and Semito-Hamitic" uses the misleading term "Hamitic" for the non-Semitic branches of AA (pp. 265-96). Even though he is not the first person to do so, this usage improperly encourages this outmoded and erroneous linguistic designation, which is apt to confuse general linguists into thinking that PAA (= Proto-Hamito-Semitic) neatly split into PS and Proto-Hamitic. This we know was not the case and was one of the prime reasons Joseph H. Greenberg decided to substitute the term AA as a replacement for Hamito-Semitic (at first, he used the hyphenated term, "Afro-Asiatic"; however, this designation goes back to 1914--see my "Comparative Afroasiatic and General Genetic Linguistics" (with Peter T. Daniels), Word 43 (1992): 429-58, at p. 431.

Voigt has gone to great lengths to explain various features of Proto-AA phonology, basing reconstructions on the classic work by Marcel Cohen, Essai comparatif sur le vocabulaire et la phonetique du chamito-semitique (Paris, 1947), Igor M. Diakonoff, and Otto Rossler, among others. However, he also devotes considerable space to his theory that Akkadian, Twareg, and Bedauye (Beja) "represent the most archaic languages of the respective language groups," positing one suffix and three prefix conjugations for Proto-AA (p. 282). This theory remains highly controversial and not all specialists would accept it.

Let me take up the intriguing matter of the shwa semitohamiticum (pp. 274-75). The */[TEXT NOT REPRODUCIBLE IN ASCII]/ is reconstructed by some for PS *l[TEXT NOT REPRODUCIBLE IN ASCII]bb- "heart," which yields libbu(m) in Akk., but lubbun in CAr. The entire idea is borrowed from PIE shwa indogermanicum, as in *p[TEXT NOT REPRODUCIBLE IN ASCII]ter "father." Thus, the reconstruction of PAA *s[TEXT NOT REPRODUCIBLE IN ASCII]m- "name" accounts for the variation of Heb. sem, but Akk. sumu(m). I do not think we have enough evidence to reconstruct any vowel here, and so *sVm- is a more accurate reconstruction. Robert Hetzron reconstructed *sim (in his "Afroasiatic Languages," in The World's Major Languages, 649). A similar situation with a consonant presents itself when one tries to reconstruct the root for "kill"--PS *qtl or *qtl. Should one give more weight to Heb. emphatic /t/ or Ar. non-emphatic /t/? The truth of the matter is that there is no way to justify either one; Heb. could have resulted via assimilation to the uvular stop, and Ar. /t/ could have resulted via dissimilation. These are the inherent limitations of the comparative method. The best we can do is to reconstruct qt/tl.


11. Stephen A. Kaufman's "Languages in Contact: The Ancient Near East" (pp. 297-306) begins by maintaining that PAA is related to PIE, and that the Urheimat for PAA is the Nile Delta (p. 298). Needless to say, these two positions are not mainstream, and I believe only a handful of Afroasiaticists would accept either. I do not accept the IE-AA connection as proven, and scholars such as Robert Hetzron have also cast skepticism on these wide connections. (5) Furthermore, Hetzron states that the Sahara desert is the Urheimat (see "AA Languages," in The World's Major Languages, 649).

Let me comment on the first language mentioned by Kaufman, which also happens to be the first written language (pp. 298-99). The author maintains: "Isolated attempts to relate Sumerian to other language isolates (such as Ural-Altaic, Japanese, or Korean) cannot be said to have born any edible fruit ... " (p. 298). Japanese and Korean may be language isolates, although some linguists believe they are related, but Ural-Altaic is a well-established, although conjectural, language family--in any case, not a language isolate.

After comments dealing with the influences of Sumerian on Akkadian, Greek on Aramaic and Hebrew, Persian on Aramaic, the author focuses on inner-Semitic: Ugarito-Akkadian, Amarna Akkadian, Aramaic on Hebrew, Aramaic on Arabic, and offers a strong endorsement of the view that the Deir 'Alla inscription is between Hebrew, Aramaic, and Ammonite. This is followed by a section on diglossia (p. 303), in which it is hypothesized that the time lag between formal and colloquial language in the ancient Near East is approximately two centuries (ibid.).

Corrigenda: p. 297: type-cast is incorrect for typecast.

12. Olga Kapeliuk's "Languages in Contact: The Contemporary Semitic World" contains an extensive twelve-page bibliography on the four linguistic areas treated: Ethio-Semitic, Neo-Aramaic, Israeli Hebrew, and colloquial Arabic dialects (pp. 307-40). The author is correct to emphasize that the Cushitic impact on the Ethio-Semitic languages has been extensively investigated during the twentieth century. This influence can be seen both in grammar and vocabulary. Insofar as the latter domain is concerned, Kapeliuk reports that Wolf Leslau identified 1,300 Cushitic loanwords out of a total of 6,500 roots in his Etymological Dictionary of Gurage (Wiesbaden: Harrassowitz, 1979). Cushitic has also greatly influenced phonology, although not all scholars agree that the glottalized consonants in Ethio-Semitic result from the Cushitic substratum. The mainstream view today is that Ethio-Semitic (and Akkadian) inherited glottalization from PS.

One area which has long fascinated me has been the extent to which Arabic vocabulary has penetrated the various Ethio-Semitic languages. In this particular domain the contributions of Wolf Leslau have been considerable. Kapeliuk explains how this penetration occurred (pp. 312-13). In Harari, e.g., the language of Muslims in the walled city of Harar, 27.5% of the lexemes are borrowed from Arabic, which, as the language of the Koran, is a sublime catalyst, as is Arab commerce (p. 313).

Neo-Aramaic contact with Iranian languages has a long history of study. The author mentions the pioneering work of the great Semitist, Theodor Noldeke (Grammatik der neusyrischen Sprache am Urmia-See und in Kurdistan, Leipzig, 1868 [p. 308 n. 2]). The author, in speaking of the Iranian and Turkish influence on Neo-Aramaic mentions that fact that 68.9% of the nouns in the Jewish Neo-Aramaic dialect of Persian Azerbaijan (described by Irene Garbell in 1965) are Kurdish or Turkish loanwords (p. 316). Even in morphology, the entire Neo-Aramaic verbal system has been reshaped in accordance with the structure of Persian and Kurdish (p. 317).

Many scholars have discussed the contact situation with Israeli Hebrew. The pioneering work on this subject is the unpublished 1930 Berlin doctoral dissertation by Irene Garbell, Foreign Influences on Israeli Hebrew (in German) (p. 320). The author convincingly shows that Israeli Hebrew is just a normal living Semitic language and, quoting Gideon Goldenberg, she affirms that its morphology "is by far the most conservative of all the living Semitic languages" (p. 321). As for the thesis developed by Paul Wexler's The Schizoid Nature of Modern Hebrew: A Slavic Language in Search of a Semitic Past (Wiesbaden: Harrassowitz, 1990), I agree with her strong refutation: "... it seems to be no more than a provocative prank 'pour epater le bourgeois' personified for the occasion by more traditional grammarians" (ibid.).

Kapeliuk's discussion of the contact phenomena with Maltese, Cypriot Maronite Arabic, Uzbekistani and Afghan Arabic, and Arabic-based pidgins and creoles of Africa, although brief, is insightful. On the latter topic, she affirms that "it is possible that the existence of a pidginized form of Arabic goes back as far as the Middle Ages" (p. 324). There can be no doubt that Pidgin Arabic is the world's first attested pidgin language--in the eleventh century (see for details my "Peripheral Arabic Dialectology and Arabic Pidgins and Creoles," in Languages of the World 2 [1991], 4-16; also in Actas del Congreso Internacional sobre Interferencias Linguisticas Arabo-Romances y Paralelos Extra-Iberos, ed. J. Aguade, F. Corriente, and M. Marugan [Zaragoza, Spain, 1994], 125-40).

Corrigenda: p. 331: One word has been omitted from the title of the Georg Krotkoff fest-schrift, Humanism, Culture and Language in the [Near] East, ed. A Ashrafuddin and A. H. M. Zahniser (Winona Lake, Ind.: Eisenbrauns, 1997).


13. Otto Jastrow is one of those rare scholars who has contributed to Arabic as well as Neo-Aramaic dialectology by accomplishing fieldwork and writing up his meticulous analyses thereof (pp. 347-63). His "Arabic Dialectology: The State of the Art" argues, and rightly so, that Arabic dialectology is "the perfect introduction for anybody intending to work in Semitic linguistics or philology" (p. 347). Early on in his general discussion of Arabic Sprachinseln, we read that "Malta has remained Arabic-speaking to this day" (p. 348). I cannot agree with this statement for reasons expressed in my commentary earlier on the article by Joseph L. Malone (see above).

Considering the development of the modern dialects from the more original types spoken in the Arabic peninsula, Jastrow affirms that he does not accept the pidginization hypothesis developed by Cornelis (Kees) H. M. Versteegh in Pidginization and Creolization: The Case of Arabic (Amsterdam: John Benjamins, 1984) (see my "On the Importance of Pidgins and Creoles for Historical Linguistics," in Diachronica 2 [1985]: 201-30, and "Correction Note," in Diachronica 3 [1986]: 130). He notes (and most Arabists would agree): "This theory is a blatant overstatement that does not capture the essence of the linguistic development of Arabic" (p. 349).

After a thorough survey of the major literature on the Arabic dialects of Egypt, the Levant, Iraq--where it is certainly true that we know more about how the minorities speak (or spoke, when we are referring to Jews) than about the dialects of the Muslim majorities--North Africa, and the Arabian peninsula, the author turns his attention to the Sprachinseln: Uzbekistani and Afghan Arabic, which the author calls a "continuum of language islands" with "some affinity ... [to] the Anatolian q[TEXT NOT REPRODUCIBLE IN ASCII]ltu dialects" (p. 352), Cypriot Maronite Arabic, Chadian and Nigerian Arabic, and Bedouin dialects in general.

Jastrow is right to suggest, due to a number of factors which have to do with the prestige and beauty of Classical Arabic, that "Arabic dialectology as a discipline exists mainly outside the Arab world" (p. 353). I think that the preceding statement, although ironic, is nevertheless true. However, he is not totally correct that "the North American continent has more or less given up fieldwork on Arabic dialects," and "nowadays, however, the United States contributes very little to the field" (p. 355). Let me state, in an attempt to refute the author's critical remark, that I, for one, spent seven months in 2000 doing fieldwork in Saudi Arabia and, previous to the trip, I have conducted Arabic dialectological fieldwork in Egypt, Lebanon, Syria, Jordan, the Sudan, Chad, Kuwait, Qatar, Yemen, and Israel and the West Bank.

Corrigenda: p. 350: Tlemcen is correct for Tlemcen; p. 352: terra incognita is correct for terra inognita; p. 356: indebted is correct for endebted.

14. Jastrow's second essay, "Neo-Aramaic Dialectology: The State of the Art" (pp. 365-77) surveys the scholarly literature of the four groups of Neo-Aramaic languages and dialects. I say languages and dialects because some of the varieties are not mutually intelligible. The groupings are: (1) Western NA, spoken only in three Syrian villages about an hour's drive from Damascus; (2) Turoyo, spoken in Mardin Province, Turkey; (3) Northeastern NA, spoken by Christians and the few remaining Jews in Turkey, Iraq, and Iran, and (4) Neo-Mandaic, spoken in Iraq and Iran. To show how important the factor of mutual intelligibility is, Jastrow uses Urmia (Azerbaijan) NA as illustrative. The Christians and Jews of Urmia could not understand one another's NA "dialect" and thus, had to communicate with each other in either Azeri Turkish or Persian (p. 368).

Although Jastrow expresses his firm belief that "Neo-Aramaic linguistics is ... bedeviled by the contempt of scholars of classical Aramaic varieties, who for the most part do not care at all about the modern languages," and recommends a course in one NA language for all of these scholars (p. 374), I am skeptical that this should be an absolute desideratum. While I am somewhat sympathetic in principle to Jastrow's suggestion, I am reluctant to endorse his position because, to cite a parallel, I cannot see that the study of modern American English dialects, for example, would contribute much to the research efforts of Indo-Europeanists working on an ancient IE language, or even, for that matter, Anglicists working on Old English. Do Hittitologists have to become au courant with African American Vernacular English or Brazilian Portuguese?

Corrigenda: p. 375: In two references Gustav Bergstrasser should be replaced by Gotthelf Bergstrasser.

15. Marie-Claude Simeone-Senelle's "Les langues sudarabiques modernes a l'aube de l'an 2000: Evaluation des connaissances" (pp. 379-400) is a thorough survey of the discovery, early research on, and continuing study of the six languages belonging to this subgroup of Semitic languages called Modern South Arabian. The six languages are: Mehri, Hobyot, Harsusi, Bathari, Jibbali, and Soqotri (her ordering, p. 379).

The author estimates that about 100,000 Yemenis and 140,000 Omanis have one of these six languages as their mother tongue. In addition, in all likelihood, the overwhelming majority of the 240,000 are Arabic-speaking as well. It is interesting to note that the existence of South Arabian spoken languages was already mentioned in the tenth century A.D.; however, the world had to wait until the earlier part of the nineteenth century to learn that these languages, in fact, existed. One can follow the history of the state of South Arabian Studies through 1945 by examining Wolf Leslau's Modern South Arabian Languages: A Bibliography, Bulletin of the New York Public Library 50.8 (1946), 607-33.

Two names are most prominent in the modern era of progress in this field. They are Bertram Thomas, who in 1929 discovered two new languages spoken in Oman--Harsusi and Bathari--and T. M. Johnstone, who discovered what he then called a dialect of Jibbali spoken on the border between Yemen and Dhofar. Hobyot, as it was called, turned out to be a new language, the sixth. The author and Antoine Lonnet are the two leading contemporary scholars researching this field.

One of the most important points Simeone-Senelle successfully brings out is that there is no direct link between the Old South Arabian and the Modern South Arabian languages. It is very likely that the modern ones go back to old languages which were never written, well before the penetration of Arabic, which came from the northern part of the peninsula. She does, however, classify the Old and Modern South Arabian languages together along with the Ethio-Semitic ones. She stresses the importance of the Modern South Arabian subgroup, affirming that they have retained old features lost in all of the modern Semitic tongues (p. 394).

16. David Appleyard's "New Finds in the 20th Century: The South Semitic Languages" (pp. 401-30) surveys the past and potential future contributions that the study of the 30-odd modern South Semitic languages can make to Comparative Semitics. Let me first take up one of the characteristics of South Semitic--the geminated imperfect. South Semitic is known for its preservation of PS and PAA *yVqattVl, the imperfect indicative. Ethio-Semitic languages and Akkadian, the representatives of the two geographical ends of the Semitic-speaking world, retain the geminated imperfect indicative, as in Akkadian iparras "divides/will divide" and Ge'ez y[TEXT NOT REPRODUCIBLE IN ASCII]qatt[TEXT NOT REPRODUCIBLE IN ASCII]l "he kills" = Proto-Modern South Arabian nongeminated *y[TEXT NOT REPRODUCIBLE IN ASCII]qat[TEXT NOT REPRODUCIBLE IN ASCII]l. Appleyard reports, quoting research published by Norbert Nebes (1994), that Sabaean and the other Epigraphic South Arabian languages do not contribute any evidence favoring a geminated imperfect, and consequently, it has been suggested that they may belong to Central Semitic rather than South Semitic (p. 405) (see also Huehnergard, p. 129, who remarks that Hadrimitic may be different from the other Epigraphic South Arabian languages).

Secondly, let us consider the definite article (pp. 411-12). As is well known, PS did not have a definite article. Appleyard rightly concludes that the definite article developed independently in Epigraphic and Modern South Arabian and in the various Ethio-Semitic languages. I agree with the author's conclusion that the definite article in the Modern South Arabian languages is "due to the influence of some form of Arabic, or better North Arabian diffusion" (p. 412). Consider Mehri g[epsilon]:d "skin," but ag[epsilon]:d "the skin" (cf. Israeli Heb. y[epsilon]l[epsilon]d "boy," ay[epsilon]l[epsilon]d "the boy" (hay[epsilon]l[epsilon]d in more acrolectic speech = Biblical Hebrew [hayy[epsilon]l[epsilon][TEXT NOT REPRODUCIBLE IN ASCII]).

Thirdly, let us delve into Ethio-Semitic dialectology (pp. 419-24). The author is correct to stress that mutual intelligibility is the customary means used to decide whether we are speaking of a language or a dialect, but, as he further annotates, must we insist on complete mutual intelligibility to talk of dialects (p. 420)? Tackling the question of Amharic and Argobba, the author presents a sentence in Argobba which is identical to is Amharic equivalent, and another in which only one word is identical (ibid.). There is no easy answer to the nomenclature problems of language vs. dialect--in Semitic, AA, or indeed any other language family. In fact, the old adage attributed to Uriel Weinreich and his father Max, that "a language is a dialect with an army and a navy," might be the best we can do to demarcate these terms. I suggested that Argobba developed as the Muslim dialect of Amharic (see my review of Wolf Leslau, Ethiopic Documents: Argobba, Canadian Journal of Linguistics 44 [1999]: 75-78).

We know relatively little about dialectal and sociolectal variation within Ethio-Semitic. For a language like Tigrinya, the author points out some impressionistic comments by Edward Ullendorff and Wolf Leslau (p. 422); however, he quite rightly calls for the first scientific investigation of the situation. Further, Appleyard notes: "As with dialect geography, the sociolinguistic aspects of Ethiopian dialectology have barely been explored, or indeed hardly begun" (ibid.).

The editor writes in his introduction about what we may term "hyphenated" Semitic linguistics (cf. his title, "Introduction with Some Notes on Psycho- and Neuro-linguistics (sic) and with Some Comments on the Potential Use of Computers in Linguistic Studies," pp. 433-39). I take up only one point:
 ... the so-often redefined and rejected apparent truism that every
 beginner in Semitics learns, namely that Semitic languages treat
 consonants differently from vowels, and that consonants carry the
 lexico-semantic component while vowels serve to form the grammatical
 component of a word. While I know that I am putting myself into the
 cross fire by raising the issue, let me risk my neck by asking
 whether research into slips of the tongue can teach us anything about
 Semitic languages, or perhaps from a different angle: can Semitic
 linguistics add anything to the study of human cognition and language
 generation from insights gained in research of slips of the tongue?
 (p. 435).

The editor seems not to be aware of the multifaceted research going on in this field. The articles published by Jean-Francois Prunet, Renee Beland, and Ali Idrissi ("The Mental Representation of Semitic Words," Linguistic Inquiry 31 [2000]: 609-48), and by Sabah Safi-Stagni ("Agrammatism in Arabic," in Perspectives on Arabic Linguistics III, ed. Bernard Comrie and Mushira Eid [Amsterdam: John Benjamins, 1991], 251-70, and "Morphological Structure and Lexical Processing: Evidence from Arabic," in Perspectives on Arabic Linguistics VII, ed. Mushira Eid [Amsterdam: John Benjamins, 1995], 93-106) conclusively demonstrate that Semitic consonantal roots are psychologically real. Based on evidence from language games and speech errors, roots have been proven to be independent morphological units, and these facts are best accounted for by the traditional root and pattern (= template) morphology. Many other relevant publications are cited in the article by Prunet et al., op. cit., 645-48. Moreover, David L. Appleyard in a commentary on Wolf Leslau's Ethiopian Argots (The Hague: Mouton, 1964) examines some examples of the distorted speech of Gojjami minstrels: warrada "descend" [right arrow] warannada and raggama "curse" [right arrow] ranaggama, concluding that native speakers are certainly linguistically aware of a triconsonantal root (p. 423). The author explains: "... because the rules of morphology operate with the new consonantal string in the normal way, making an originally triliteral root like warrada into a quadriliteral root warannada, which inflects just like an original quadriliteral such as marammara" (p. 424).

17. Edward L. Greenstein's "Some Developments in the Study of Language and Some Implications for Interpreting Ancient Texts and Cultures" (pp. 441-79) reviews (with a 14-page bibliography, pp. 465-79) some of the vast literature on discourse analysis (pp. 444-46) and poetics (pp. 446-49). An interesting aspect of the former is discussed when the author notes, e.g., that Biblical Heb. [wayyel[epsilon]x] "and he went" may sometimes merely indicate the "beginning of a new episode" (pp. 445-46). As an example of the latter (based on an article by Victor Avigdor Hurowitz in 1997), the author notes the metonymic development of Heb. n[epsilon]f[epsilon]s "soul" to "appetite; desire" (p. 449). The semantic transition to "food" is not hard to imagine and, I think, is the correct translation of n[epsilon]f[epsilon]s in Isaiah 58:10 (ibid.). I would like to add evidence from ECA as extra proof because nifs (the cognate to n[epsilon]f[epsilon]s) has come to mean "appetite; desire," as in nifsi masduuda "I have no appetite/desire" (see Martin Hinds and El-Said Badawi, A Dictionary of Egyptian Arabic ([Beirut: Librairie du Liban, 1986], 876).

One of the most interesting sections of this article deals with anthropological interconnections (pp. 449-57). Not wishing to write a long excursus on the merits or lack thereof of the Sapir-Whorf Hypothesis since many of these are already available, let me just say that not all linguists would accept the view expressed by John J. Gumperz, who is quoted as saying: "With sufficient effort, it is possible to say anything in any language" (p. 450). I, for one, have never been able to fathom how one can render the mathematical notion of [pi] (pi) into a language which does not have any numerals, or one in which after the words "one" and "two," the words "three," "four," "five," etc. are one and the same.

The author delves into the realm of semantics by discussing the notion of semantic primitives. More Semitists should certainly be cognizant of this area. Anna Wierzbicka, whose 1972 and 1991 books are cited, has written a newer volume on the topic: What Did Jesus Mean? Explaining the Sermon on the Mount and the Parables in Simple and Universal Human Concepts (Oxford: Oxford Univ. Press, 2001); see also my review in Language 79 (2003): 226-27.

In a commentary on some of the semantic literature, the author is very critical of John Myhill's 1997 contention that Biblical Heb. yare' means "fear" in the sense of "revere," whereas pahad means "fear" in a negative sense (pp. 452-53). He contends: "It is in fact only Myhill's presupposition that Hebrew expresses the basic emotions such as fear and compassion in an abstract fashion" (p. 452). However, Greenstein is going too far when he proclaims: "My own sense has always been that Hebrew, like other ancient Semitic languages, does not express the abstract lexically but rather expresses abstract reasoning implicitly ..." (ibid.). But then the author seems to modify this position when he talks of "tendencies": "Hebrew and the other ancient Semitic languages tend to concrete rather than abstract expression ..." (p. 456). I would have thought that all languages have some abstract nouns or other grammatical units.

Corrigenda: p. 469: One word is left out in the four volumes edited by Jack M. Sasson: Civilizations of the Ancient Near [East].

18. Bruce Zuckerman's "Working with a Little More Data: New Finds in the Twentieth Century: The Semitic Languages of the Ancient World" discusses the discoveries since the end of the nineteenth century, especially in Northwest Semitic, which have added to or altered the conceptions of previous data available (pp. 481-97). His examples include texts from Elephantine, Ugarit, Lachish, Arad, Deir 'Alla, Qumran, Ebla, etc. Also important are the newer tools and monographs available for the study of such new texts: e.g., the Chicago Assyrian Dictionary and the Complete Aramaic Lexicon.

Most of the essay deals with a return to the original photographs (if available) of the inscriptions, seals, stone stelae, and cuneiform tablets, and some new techniques, such as computer imaging. I sympathize with the author's conclusion: "Sometimes I feel as though we are drowning in the data, that all our technological techniques really do not cut down our subjectivity in interpreting the data but only increase it. And when that happens, I begin to wonder, is it the data we are reading or simply our imaginings of the data?" (p. 496).

Corrigenda: Zellig Harris's name is misspelled as Zelig on pp. 483 and 496.


1. Shlomo Izre'el's review of Robert Hetzron, ed., The Semitic Languages (London: Routledge, 1997; pp. 501-10).

Let me correct the reviewer's statements concerning my article (coauthored with Judith Rosenhouse) in this volume, "Arabic Dialects and Maltese" (pp. 263-311). He states: "Kaye and Rosenhouse, after some deliberation, regard all Arabic languages as dialects (except for Maltese)" (p. 503). We use the term "Arabic dialects," since this is the customary designation by both specialists and nonspecialists alike. However, we note that if two dialects are not mutually intelligible (on the whole), we would prefer to call them separate languages. Thus, Uzbekistani Arabic, Nigerian Arabic, Saudi Arabian (Najdi) Bedouin Arabic, and Moroccan Arabic are really best considered separate languages, much as Mandarin Chinese, Wu Chinese, and Cantonese Chinese are distinct languages. It should be noted that linguists quite often refer to different languages by using the term "dialects." Consider the Canaanite dialects, the Gurage Dialects, the Berber dialects--even early Semitic/IE dialects. See further my remarks above in the discussion of David Appleyard's essay.

The reviewer also finds fault in that we say we use "IPA-based" symbols (p. 263). The phonetic symbols we employ are either IPA or IPA-based, which means that we use /[theta]/ /[TEXT NOT REPROPULIBLE IN ASCII]/, and /h/ for /<h->/, /s/ for /[integral]/, etc., for ease of recognition by Semitists, and lack of typographic availability of certain symbols.

2. David Testen's review of Edward Lipinski, Semitic Languages: Outline of a Comparative Grammar (Leuven: Peeters, 1997; pp. 511-20).

Since Lipinski makes references in this book to other AA languages, Testen is justified to remark: "Using Afro-Asiatic material to resolve Semitic problems can be a very hazardous enterprise, particularly if all the comparanda are adduced principally on the basis of superficial resemblance" (p. 513). My sentiments precisely! In this connection, see for further details my review of V. Orel and O. Stolbova, op. cit., and my review of this volume in World 51 (2000): 77-82.

Moreover, Testen is right to point out that Lipinski's views on Semitic language classification are different from the mainstream views (pp. 513-14). From my perspective, the classification posited into North, South, East, and West is merely geographical, not genetic.

3. Shlomo Izre'el's review of Patrick R. Bennett, Comparative Semitic Linguistics: A Manual (Winona Lake, Ind.: Eisenbrauns, 1998; pp. 521-25).

I have two remarks on this review of an introductory textbook on Semitic linguistics. The first concerns lexicostatistics, which Bennett uses. I agree with the reviewer, who notes that "most of us will find this unacceptable" (p. 523). Let me reiterate what John Huehnergard wrote in his essay: "It is disappointing that a number of scholars continue to use glottochronology and lexicostatistics as the basis of their internal classifications of Semitic" (p. 124 n. 2).

My second comment has to do with an astute observation on a premise of Bennett's methodology, which the reviewer explains: "His insistence on exploitation of data (over using empty theories) is to be endorsed without hesitation, as well as his repeated warning against adhering to theories as truth" (p. 524). I believe I can paraphrase this philosophy by the following: Theories come and go, but data will always be around, and there is more than one way (I mean equally valid ways) to explain data.

Corrigenda: p. 521: The infelicitous term "course book" should be changed to textbook.

4. Aaron Dolgopolsky's review of Gabor Takacs, Etymological Dictionary of Egyptian, Vol. I: A Phonological Introduction (Leiden: Brill, 1999).

Most of Dolgopolsky's remarks center around various AA etymologies proposed by Takacs. I think Dolgopolsky is right in rejecting the author's PS *na/ihna "we" in favor of only nihna (p. 532). Akk. ninu "we" and Ge'ez n[TEXT NOT REPRODUCIBLE IN ASCII]hna are important data; however, the Ge'ez form might go back to a form in a, since there are cases where *a > [TEXT NOT REPRODUCIBLE IN ASCII] (see August Dillmann, Ethiopic Grammar, second ed., ed. Carl Bezold, tr. by J. Crichton ([London: Williams and Norgate, 1907], 35-36. Besides, CAr. nahnu "we" is a secondary development even within this language, since we can compare various dialectal forms which point to /i/, such as Egyptian 'ihna and Moroccan hna(ya). Nigerian and Chadian Ar. anina "we" must go back to a *anihna--not *anahna.

Dolgopolsky rejects the equation of Egyptian jdn "ear" and Jegu?udu[eta]e and Birgit udu[eta]i (both Chadic languages), saying they borrowed this word from Chado-Sudanese Arabic uden (p. 533). I find this highly questionable. First of all, why would these Chadic languages borrow the word for "ear"--a basic body part? In looking at Ngizim, a Chadic language spoken in Potiskum, Nigeria, close to Maiduguri (where Nigerian Arabic is spoken), the word for "ear" has not been borrowed: agud (see Russell G. Schuh, A Dictionary of Ngizim [Berkeley and Los Angeles: Univ. of California Press, 1981], 197). In Kanuri (a Nilo-Saharan language in contact with Nigerian Arabic), the word for "ear" is s[TEXT NOT REPRODUCIBLE IN ASCII]mo (see Norbert Cyffer, English-Kanuri Dictionary [Cologne: Rudiger Kopper Verlag, 1994], 58). Why is the Arabic loanword uden or the like not used? More likely, either Takacs is right or these words are just coincidentally similar to Arabic.


Many of the authors in this volume speak of the desiderata necessary in their respective subfields. Huehnergard, e.g., talks about his "long wish list" (p. 133). He mentions that the field needs an etymological Arabic dictionary and an etymological Semitic dictionary (ibid.). In my view, these are far more crucial than some of the other items on his list, such as descriptions of the Modern South Arabian languages or a comparative Aramaic Grammar. Khan talks of a discourse approach to syntax for the spoken languages (p. 166). Jastrow's number one need in Arabic dialectology is extensive fieldwork in Saudi Arabia followed by fieldwork on the dialects of the Sprachinseln and religious minorities (p. 355). Appleyard would like to see descriptions of Amharic and Tigrinya dialects (p. 424). Unquestionably, although much has been accomplished in the twentieth century by many outstanding Semitists, the field of Semitic linguistics still lags far behind the IE field. This is so for several reasons: IE linguistics is older, more linguists have chosen this specialization, and IE-speaking countries tend to be wealthier than other nations, and thus tend to sponsor research more in IE linguistics.

I remain pessimistic about the future, however. I feel that the next generation of researchers will be fewer in number than at present. As Jastrow writes concerning Arabic dialectology: "I may be too pessimistic, but sometimes I have the feeling that Arabic dialectology as a discipline is on the decline" (p. 355). We can only trust that this tome, with so many high-quality contributions, will serve as the inspiration to attract new specialists to Semitics. We can also only hope that our pessimism is judged with the passage of time to be totally unwarranted, and that when this century comes to a close, many of today's Semitists' wish-lists have items on them which will have seen the light of day.

An index of authors, topics, and Semitic words discussed would have been most useful and would have also contributed to the shelf-life of this remarkable achievement.

This is a review article of Semitic Linguistics: The State of the Art at the Turn of the Twenty-First Century. Israel Oriental Studies, vol. 20. Edited by SHLOMO IZRE'EL. Winona Lake, Ind.: EISENBRAUNS, 2002. Pp. 535. $59.50.

1. The following abbreviations are used here: AA = Afroasiatic, Akk. = Akkadian, Ar. = Arabic, Aram. = Aramaic, CAr. = Classical Arabic, ECA = Egyptian Colloquial Arabic, Heb. = Hebrew, IE = Indo-European, NA = Neo-Aramaic, PAA = Proto-Afroasiatic, PIE = Proto-Indo-European, PS = Proto-Semitic, Sem. = Semitic.

2. Diakonoff preferred the term Afrasian for what is also known as Hamito-Semitic or Semito-Hamitic, Erythraic (coined by Archibald N. Tucker), or Lisramic (coined by Carleton T. Hodge).

3. See the Diakonoff Memorial Volume, ed. M. Lionel Bender et al. (Munich: Lincom Europa, 2003).

4. Daniels once confided to me that Gelb never changed his mind once an idea of his was committed to print.

5. Many long-range hypotheses have been postulated. See my "The Current State of Nostratic Linguistics," in Nostratic: Analysis of a Linguistic Macrofamily, ed. C. Renfrew and D. Nettle (Cambridge: McDonald Institute for Archaeological Research, 1999), 327-58.


COPYRIGHT 2003 American Oriental Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kaye, Alan S.
Publication:The Journal of the American Oriental Society
Date:Oct 1, 2003
Previous Article:The many voices of the Mahabharata.
Next Article:Mikha'il Mishaqa: virtual founder of the twenty-four equal quartertone scale.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |