page 93 note 1 On Stress and Arabic Phonology: A Generative Approach, by Abdo, Daud A. (Beirut: Khayat Book and Publishing Co., 1969), pp. 128, $4.50.Google Scholar The subjects covered in this review article follow, as closely as possible, the topics as they are presented by the author with his ordering. The reason for this is that I intend this article to be read in conjunction with the work itself (hopefully, in hand), so that any linguist, data-oriented or not, no matter what his theoretical persuasion or fields of specialization and interest, can follow my criticisms and suggestions exactly. I emphasize the analysis of the Arabic data in terms of generative/transformational grammar in the light of Arabic linguistics and dialectology. I think it always a useful exercise for a linguist to read a work with a review article in hand. It is, of course, sad to report that many journals are reluctant to publish old-fashioned review articles. RPh is an exception, due primarily to the factual approach of its editor, Yakov Malkiel. This, in its own right, is particularly noteworthy. Long vowels are marked either as VV or V^(V¯), as is done by the author (inconsistently, I might add). No difference is intended or implied between their interpretations.Google Scholar I would like to express my thanks to Professors Charles Wendell and Robert Hetzron, UCSB, Joseph L. Malone, Columbia University, and Carleton T. Hodge, Indiana University for useful suggestions on an earlier version of this review article. Needless to say, I alone am responsible for all statements, factual or otherwise.Google Scholar

page 94 note 1 Even Sir Hamilton Gibb, with reference to Modern Literary Arabic, i.e. MSA, states: ‘…in most respects the language is closer today to the language of al-Jāhiz than it was one hundred years ago’ (page 10 in Stetkevych, Jaroslav, The Modern Arabic Literary Language: Lexical and Stylistic Developments (= Publications of the Center for Middle Eastern Studies, no. VI; general editor, Polk, William R.; Chicago and London: University of Chicago Press, 1970). In my opinion, this is just not true.Google Scholar

page 94 note 2 See my ‘Remarks on Diglossia in Arabic: Well-Defined vs. Ill-Defined’, Linguistics, vol. 81 (1972), pp. 32–48, for a detailed discussion.Google Scholar

page 95 note 1 See the introduction of Arberry, A. J., Arabic Poetry: A Primer for Students (Cambridge University Press, 1965), pp. 1–27.Google Scholar See also Wright, William, A Grammar of the Arabic Language (trans. from the German of Caspari, 3rd ed. revised by Smith, W. Robertson and de Goeje, M. J.; Cambridge University Press, 1962), vol. 2 part fourth (first published 1862), pp. 350–90.Google Scholar

page 95 note 2 The /-u/is deleted in Moroccan Arabic (colloquial) virtually by all speakers. Chadian Arabic, on the other hand, based on my own field work in Fort-Lamy and Abéché, summer 1970, retains both the /-u/ marking nominative and the /-i/ marking genitive. I never heard a native speaker of Chadian Arabic (bilingual or multilingual or not) deviate from this rule.Google Scholar

page 95 note 3 Ariel A. Bloch, University of California at Berkeley, has been working on a monograph showing interesting syntactic distributions of the numerals in colloquial Arabic dialects (especially Cairo and Damascus Arabic). We anxiously await his results. See now his ‘Morphological Doublets in Arabic Dialects’, JSS, vol. 16 (1971), pp. 53–73.Google Scholar

page 95 note 4 It should be pointed out that most linguists do not accept Ferguson's koiné hypothesis. I, however, do accept the hypothesis (with some reservations) and hope to show elsewhere its validity utilizing data from Sudanese and Chadian Arabic dialects. It is true that many of Ferguson's features can better be explained through ‘drift’. See Sapir, Edward, Language: An Introduction to the Study of Speech (Harvest Books ed., 1921), ch. 7, pp. 147–70.Google Scholar

page 97 note 1 This is a well-known phenomenon in Semitic as a whole, and also one which J. Cantineau spent much time on, i.e. his parler différentiel. Cf. Mitchell (1956), pp. 113–4.Google Scholar

page 97 note 2 These are not to be found, say, in Wehr, Hans, A Dictionary of Modern Written Arabic, ed. Cowan, J. Milton (Cornell University Press, 1961), pp. 217 ff.Google Scholar Cf. Cl. Ar. /'istaâba/ and /'istajˇwaba/, synonyms for /'ajˇâba/ ‘to answer’. Cf. Beeston, A. F. L., BSOAS, vol. 35, 1 (1972), p. 140.Google Scholar

page 98 note 1 See Jakobson, Roman, ‘Mufaxxama: The “Emphatic” Phonemes in Arabic’, in Studies Presented to Joshua Whatmough on his Sixtieth Birthday (The Hague: Mouton, 1957), pp. 105–15.Google Scholar

page 98 note 2 The term was coined by Chafe, W. L., as far as I know. See his ‘Idiomaticity as an Anomaly in the Chomskyan Paradigm’, FL, vol. 4, 2 (1968), pp. 109–27,Google Scholar especially for his discussion of ‘paradigms’ utilizing Kuhn, Thomas, The Structure of Scientific Revolutions (Chicago and London: University of Chicago Press, 1962).Google Scholar

page 98 note 3 Underlying segmental representation is equated with morphophonemic (underlying phonological) representation. It is worthwhile noting that many generative phonologists use the term ‘phonemic’ sometimes meaning ‘autonomous phoneme’ and other times ‘systematic phoneme’. This may cause confusion.Google Scholar

page 99 note 1 The /li-/ underlying structure is a semantic one, and this kind of evidence is another indication of the power of a generative semantics (semanticism) over a generative syntax (syntacticism).Google Scholar Cf. Chafe, W. L., Meaning and the Structure of Language (Chicago and London: University of Chicago Press, 1970), esp. pp. 1–23. Chafe ends the book by saying (351–2): ‘If I have done no more than add weight to the assertion that linguistics will stagnate unless it thoroughly explores this territory, I have accomplished my principal aim.’ I hope to spend a whole paper on this topic in the near future.Google Scholar

page 100 note 1 I find the term ‘contemporary Arabs’ rather awkward. ‘Speakers of contemporary Arabic’ is far more appropriate.Google Scholar

page 101 note 1 Europeans, it seems sometimes, are better aware of the problem than Americans. See, for example, Kuipers, A. H., ‘Unique and Typological Universals’, in Pratidānam: Indian, Iranian and Indo-European Studies Presented to Franciscus Bernardus Jacobus Kuiper on his Sixtieth Birthday, ed. by Heesterman, J. C., Schokker, C. H., and Subramoniam, V. I. (The Hague: Mouton, 1968), pp. 68–88. Kuipers writes (p. 84): ‘It would have been possible to pass over in silence a reverse methodology, contempt for data, and hollow phraseology in one or two linguists or even a school of linguists. But the theory-oriented, fact-despising approach, though not yet universal, is assuming the proportions of a near-universal, and it constitutes a real cultural danger, particularly in the U.S.A. [emphasis mine].Google Scholar ‘No science can do without a theoretical foundation, least of all linguistics. But when theory is given priority over fact-gathering, when theory becomes a goal in itself and assumes the arrogance of authoritarian doctrine [emphasis mine], turning its back to reality, ignoring, dismissing or altering such facts as are brought to its attention, and when a leading theoretician can pass off the tritest inanities as new universal insights, then there is something wrong with the field of science. ‘And there is something wrong, indeed, with American linguistics [emphasis mine]. There is not a single science which has so shamefully neglected its material, its object of study. The flora, the fauna, the geology, everything which has been surveyed and described down to the last species and rock-formation. Only the linguists have left their field –one the richest in the world – to a very large extent an untapped terra incognita, about which they would rather speculate than gather information.’ Kuipers continues (pp. 85–6): ‘Why is linguistics no longer a successful science in the U.S.A.? It was in Europe in the 19th century, and in America in the first decades of the 20th (Boas, Sapir, Whorf, Bloomfield)… It is the fact that in linguistics there are few sanctions on bad work… In linguistics, as we have seen, anything goes [emphasis mine], and one can even be quite frank about it.’Google Scholar See also for a much too positive review of King 1969, Campbell, Lyle in Language, vol. 47 (1971), pp. 191–209.CrossRefGoogle Scholar

page 105 note 1 Cf. Cantineau's, J. parler différentiel and parler non-différentiel. For example, Cairene /mahammad/ ‘Mohammed’ but /yaa/ ‘Oh’ + /mahammad/ → /yamhammad/ (elision of /a/), or /ma'a/ ‘with’ but /ta'aalam'aaya/ ‘come with me!’ This is unknown, insofar as the literature is concerned, and I hope to make this subject into a whole paper in itself. These characteristics of /i/ and /u/ as opposed to /a/ are certainly related with the facts that both reduce to /u/ in Ethiopic and Damascus Arabic, and often in Hebrew and other Northwest Semitic languages too. Hebrew also reduces •/a/, occasionally. Cf. Hebrew /šbî1/ and Arabic /sabîl/ ‘path’. The crucial environment is unaccented open syllable, yet note that /a/ is a vowel whose norm is non-reduction.Google Scholar Cf. Malone, Joseph L., ‘Wave Theory, Rule Ordering, and Hebrew-Aramaic Segolation’, JAOS, vol. 91 (1971), p. 59, n. 54 and p. 62, n. 62. I hope to treat this at a later date also.Google Scholar

page 106 note 1 This problem is treated in detail in my ‘Arabic /žiim/: A Synchronic and Diachronic Study’, Linguistics, vol. 79 (1972), pp. 31–72.Google Scholar

page 107 note 1 Garbell, Irene, ‘Remarks on the Historical Phonology of An East Mediterranean Arabic Dialect’, Word, vol. 14 (1958), pp. 303–37.CrossRefGoogle Scholar Professor Malone comments (personal communication): ‘These anaptyctic data are highly interesting, and I've been taking plenteous notes especially for the verb, since I formulated the synchronic repercussions for Mandaic in my A Morphologic Grammar of the Classical Mandaic Verb (Univ. of California doctoral dissertation, Berkeley, 1967) (Ann Arbor, University Microfilms No. 68–10,366), pp. 181 ff. To take the example of the imperfect with V-suffix, it spreads beyond Arabic and is widespread at least throughout NW Semitic. E.g., Syrian Arabic /ykattbu/ and Iraqi Arabic /yikitbuun/ are at least typologically cognate to Mandaic <nikitbun>, where the process is quite systematic, and Hebrew ya′avδuˇ, where the process is limited to a preceding guttural (more widespread in the Babylonian Massora, where catalyzed also by preceding liquid). I have not found a NW language completely alien to the process, as long as the orthography is sufficiently rich to betray it. Seems to have got started in guttural/liquid environment (αconsonantal, αvocalic?), and then spread irregularly over the NW.’,+where+the+process+is+quite+systematic,+and+Hebrew+ya′avδuˇ,+where+the+process+is+limited+to+a+preceding+guttural+(more+widespread+in+the+Babylonian+Massora,+where+catalyzed+also+by+preceding+liquid).+I+have+not+found+a+NW+language+completely+alien+to+the+process,+as+long+as+the+orthography+is+sufficiently+rich+to+betray+it.+Seems+to+have+got+started+in+guttural/liquid+environment+(αconsonantal,+αvocalic?),+and+then+spread+irregularly+over+the+NW.’>Google Scholar

page 107 note 2 This whole problem is certainly related to such facts as Damascus Arabic /'ltllo/ ‘I told him’ and Cairene /′ultílu/ ‘I told him’. That is to say, the gemination of /1/ and the stressed anaptyctic vowel relate the two phenomena diachronically.Google Scholar

page 110 note 1 This transcription follows Schramm, Gene M., The Graphemes of Tiberian Hebrew (Berkeley and Los Angeles: University of California Press, 1964).Google Scholar A better transcription, for our purposes here, is /wayyāqom/ ‘and he got up’. On this point, see most recently, Levin, Saul, The Indo-European and Semitic Languages (Albany: State University of New York Press, 1971), p. xxxiv and passim.Google Scholar