Dictionnaire Songhay-Anglais-Français. Vol. 1: Koyra Chiini (264 pp.); Vol. 2: Djenné Chiini (208 pp.); Vol. 3: Koroboro Senni (352 pp.). By Jeffrey Heath. (review)

496 LANGUAGE, VOLUME 76, NUMBER 2 (2000) syncope, aphesis, and vowel shortening—each succinctly explored with particularly well-chosen examples . He has whet this reader's appetite for further knowledge with his impressive section on historical phonology (49-54). One diachronic phenomenon, *b> w in word-final position, is a good example of a sound change in progress, since one notes assabab — assabow 'cause' (49). He has only scratched the surface on the topic of Arabic loanwords (54): since most KCh speakers are Muslim, the Arabic influence is considerable. However , I agree with H when he remarks that 'Arabic loanwords are tricky' (54). By this he means presumably that some have come from Classical Arabic (e.g. laabudda 'definitely'), since Timbuktu was and still is a center of Islamic learning, while others derive from Hassaniya or Maghrébine Arabic, and yet others have filtered down through other West African languages, such as Fulfulde. In some cases, it is most difficult to pinpoint Classical or colloquial Arabic as the etymon (e.g. kul 'each, all' which in fact, might have been borrowed from both). The most detailed sections of the grammar deal with syntax, and many intricate constructions are cogently described. For example, H discusses a 'whatchamacalht ?' formation involving hajje (or haya-je) and haywane, which is said to be probably related to haya 'thing'. Interestingly enough, the lexeme 'thing' shows up in an English synonym for 'whatchamacallit ', viz., 'thingamajig' or 'thingamabob'. Mention is made of a common Semitic construction , the cognate object (133), in which, e.g. koy 'go(ing)' is repeated = 'the going I went'. Since this is a well-known construction in Arabic, one wonders if this could possibly be attributed to language contact . In conclusion, H's grammar is a most welcome addition to the ever-growing arsenal of accurate information presently available to researchers in linguistic typology and language universals among other various subfields oflinguistics. [Alan S. Kaye, California State University, Fullerton.] Dictionnaire Songhay-Anglais-Français. Vol. 1: Koyra Chiini (264 pp.); Vol. 2: Djenné Chiini (208 pp.); Vol. 3: Koroboro Senni (352 pp.). By Jeffrey Heath. Paris: L'Harmattan, 1999. These three comprehensive trilingual dictionaries ofone ofthe major language families ofMali (members of which are also spoken in other West African countries) are a solid first for Songhay lexicography. Volume 1 covers the Songhay of Timbuktu (called Koyra Chiini [KCh]), which extends further along the Niger River to Goundam and Niafunké. Volume 2 deals with the Songhay of Djenné, well known for its mosque and weekly market. It is different from Koyra Chiini in that it has a seven-vowel system to the latter's five, and it also contrasts somewhat in syntax and vocabulary. Volume 3 covers Koroboro Senni (properly Senni), the Songhay of Gao, the ancient capital of the Songhay Empire. This variety is spoken along the Niger River from Gourma Rharous to the border with the Niger Republic. There is, not unexpectedly, an overlap among the three tomes, e.g. daaga is glossed as 'a caste ofTuaregs ' in Vols. 1 (65) and 3 (76) although it is translated only as 'Tuareg' in Vol. 2, but with the following annotation: 'originally denoted a Tuareg caste, the Imrad' (47). Because of this situation, the following remarks deal mainly with the first volume which, incidentally, has a fine photograph of the author in traditional boubou, a ganduuma (99), on its cover page. As I perused the three volumes, my Arabist interests were stimulated. It was not surprising to see the tremendous number of Arabic (A) loanwords in all the varieties. The etymology of most ofthese is cited according to the well-known root system; however, for a few items which differ, the exact A etymon is cited, such as KCh daa?iman 'always' < A daaPiman 'always' (65). Let me turn to the many fascinating items beginning with al-, the A definite article (21-31), although a few of these are not A in origin, such as almeetu 'matches', a French loanword (29). It should be noted that a few A loanwords occur as doublets: alwaajib 'duty, obligation, necessity' points to Classical A as donor, whereas alwaaiib points to Maghrébine A. Evidence for a lateralized pronunciation of the daad (emphatic = pharyngealized -velarized /d/) can be cited in KCh alwalaa 'Muslim ablutions (before prayer)' (30) < A wadaaia 'purity', and alkaali 'Islamic judge' < A alqaadi 'judge' (27). The assimilation of the A definite article before apicals can be noted in words such as annuura 'brilliance, elegance' (31) < A annuur, arrahma 'grace, mercy' (32) < A arrahma, or assama (also samaa without the article) 'minaret' (33) < Maghrébine A samîa 'minaret'. Also quite noteworthy is the free variation among many of the lexemes. Heath is to be commended for the decision to record these variants since they facilitate research in dialectology. Consider: masar — masar 'smallpox' (192), or citaab — kitaaw — kitaw ~ kitaabu 'Koranic volume' (63). Several interesting semantic changes in the A loanwords parallel the situation ofother languages which have borrowed heavily from this language, such as Persian, Malay, or Swahili. For example, alhamdila 'congratulations!' < A alhamdu lillaah 'praise be to God', alhikima 'wonder, miracle' < A alhikma 'wisdom' (25), and dawa 'ink' < A dawaah 'inkwell ' (67) (cf. A hibr 'ink'). I must differ with the author on three A etymolo- BOOK NOTICES 497 gies: (1) the word 'God' allaahu < A allaahu with the preservation of the nominative case, as in the Muslim call to prayer, allaahu lakbar 'God is great'—not from allaah (28); (2) wallaahi 'by God' < A wallaahi and not wa allaahi (249); and (3) kalaa 'no'! is said to derive 'perhaps from Tamashek' (154), whereas I believe it is clearly from A kallaa 'no' ! with degemination (cf. KCh kul 'all, every' < A kull [176], and alhaku 'reward for a service' < A alhaqqu 'right' [24]). H's dictionaries are well-executed and easily serve as models for other linguists who plan to engage in lexicographic work on other relatively unstudied African languages. [Alan S. Kaye, California State University, Fullerton.] Pathways of the brain: The neurocognitive basis of language. By Sydney M. Lamb. Amsterdam & Philadelphia: John Benjamins, 1999. Pp. xxii, 416. This has proven to be a very difficult book to evaluate. On the one hand, it reads as though it were an introductory linguistics textbook although it was certainly not conceived as such. Consider my reasons for this pronouncement: rudimentary statements such as: 'speech and writing are markedly different from one another' (16), 'linguists can represent texts in phonological transcription' (19), 'the relationship between sound and meaning is in principle arbitrary' (35), or 'In initial position the phonetic distinction between the stops which are usually called ''voiced'' (/bdg/) and those called "voiceless" (/ptk/) is really the absence or presence, respectively, of aspiration' (279) (a statement with which I disagree). On the other hand, in an attempt to deal with more involved linguistic issues such as the relationship between language , thought, and intelligence (291-92), Lamb raises important but by no means original issues, e.g. whether there can be thought without language, concluding , disappointingly, that the term thought 'can't really be taken too seriously' (292). But it is most difficult to reconcile this aforementioned perspective with the following scintillating sentences penned by L: 'What is thinking? What are we doing when we are thinking?', and 'Can we think without words?' (11). Tackling the last question head-on, surely every American English native speaker knows, more or less, how much $14 buys today (say, dinner in a restaurant but not a diamond ring or car) and that it is more than $13 but less than $15. Clearly, the concept of $14 would not mean very much if there were no words representing numeration per se. The answer to L's aforementioned question is not completely satisfying since the question is not really answered. However, L inconsistently further expatiates with scattered remarks throughout such as: 'Tust as the forms oflanguage can influence thought ... so notation can influence scientific thinking' (273). So much for 'thought' and 'thinking' not being taken so seriously . Some chapters appear to be textbookish but germane in fields other than linguistics. For example, Ch. 16, 'Introducing the brain', has all the makings of an introduction to a physiological psychology text (293-319). Why should a linguist know that the hypothalamus lies under the thalamus (297), or 'the hippocampus is actually a paleocortical structure' (298)? Ch. 17, 'Neurons and nections', is in parts even more complicated for linguists than the preceding chapter, with exhausting statements such as: 'And myelinated axons conduct action potentials up to one hundred times faster than unmyelinated ones' (323). The reader of this tome will surely be aware of the author's relationship to stratificational linguistics and will therefore be curious about its current states. Mention of strata is still a part of L's views. But one does not necessarily have to be a stratificational linguist to come to the conclusion, as L and other linguists do, that language is an intricately woven system of systems ("The stratification of language' [34-97] and 'Applying the unity fallacy to the linguistic system' [284-89]). L is still using the terms phonon and morphon, both of which never became popular. The phonons 'apical', 'front', 'back', 'labial ', 'nasal', and 'unmarked' (all of which every linguist knows) do not add much to our knowledge, I believe (i.e. referring to them as phonons), nor do morphonic representations tell us any more about the workings of language than its older designation morphophonemic did. Indeed, this tome covers much familiar ground as L asks, 'What is a morpheme?' elucidating with the alternation wife-wives and the suppletive good-better (51 ). He also continues to use the term sememe. Finally, let me comment on two absorbing topics. Firstly, L opines that knowing a language is somehow tantamount to knowing its lexemes (285). But clearly, this is a distorted view since we can have two fluent native speakers, one ofwhom knows twice as many words (or more) as the other. These persons, thus, share the linguistic systems of the language but not, in equal measure, its vocabulary. Secondly, in the section on the origin of language (285-89), mo· nogenesis is, L writes, 'actually not even a very sensible question' (287). I agree, since I think we will never be in a position to prove monogenesis or polygenesis . The 'sophisticated' form of this question, according to L, is 'do all the languages now spoken in the world have a common source?' (287). L answers in the affirmative and continues: 'We may also entertain the possibility that Proto-World might have been somewhat less highly developed than the languages we find on this planet today, a truly primitive language in that it was simpler in some ofits features ...