Language

Warwick Danks's monograph concerns Arabic verb patterns with a long vowel after the first root consonant (traditionally numbered III and VI, respectively). Originally a doctoral thesis at the University of St. Andrews, The Arabic verb aims at making 'the complexities of the Arabic language accessible for specialists in linguistics and [presenting] linguistic theory comprehensibly to Arabists with no advanced linguistic training' (xiii). The variety of Arabic investigated by D is Modern Standard Arabic (MSA) (2).

After a preface and some technical remarks (xiii-xviii), Ch. 1, 'Introduction' (1-13), establishes the subject of the book, presents very basic linguistic concepts (i.e. Saussurean structuralism), and discusses methodological considerations. D draws on Christopher Beedham's method of lexical exception (Beedham 2005; see pp. 4ff.). The actual language data are obtained in three ways: by excerpting Hans Wehr's fundamental dictionary of MSA (Wehr 1994), by questioning native speaker informants, and by consulting arabiCorpus, a web-based resource accessible at http://arabicorpus.byu.edu (see pp. 13, 15). Overall, though, D rarely quotes examples of actual language use from texts. Most passages cited either stem from other sources like grammars and textbooks, or are presumably invented. The latter particularly applies to the many sentences quoted from the questionnaire for the informants.

Ch. 2, 'Verbal morphology and the lexicon' (15-37), presents basic data on the morphology of the Arabic verb by paying special attention to the different verbal patterns and their statistical frequency based on Wehr's dictionary (29ff.). Some of the morphological information may be considered rather superfluous in a book dealing primarily with patterns III and VI, but the statistical data are quite interesting: 465 pattern III verbs are listed, equivalent to 15.7% of all triliteral root verbs. It has to be borne in mind that these data apply only to TYPES; the TOKEN count in actual usage may differ considerably. Already in this chapter, D makes extensive use of the chi-square test, a statistical hypothesis test (32ff.). Although this methodology yields interesting results, its complexity may not be readily comprehensible to many readers unfamiliar with mathematical statistics. Considering the complexity of this methodology, D's statistical results seem overall a bit meager (see the summary, p. 102: for example, 'one-quarter of pattern III and one-third of pattern VI verbs do not conform to these dominant meanings' or 'The mutual-reciprocal relationship between the patterns has been validated statistically as a real phenomenon, but it is still inadequate to explain the meanings of over 30% of the pattern III-pattern VI verb pairs').

Ch. 3, 'Alternative morphologies' (39-62), discusses newer proposals that seek to supplant the traditional root-pattern analysis of Semitic morphology, that is, word- and stem-based approaches (40ff.), as adopted, for instance, by Robert R. Ratcliffe (1997). In addition, D introduces prosodic templatic morphology (55), which assumes the arrangement of three tiers in the formation of words (vowel melody, CV skeleton, root). But the use of CV skeletons such as CvCCvC for kuttib- 'was caused to write' (56) blurs important differences in the language (e.g. vowel quality, consonantal length, status of approximants). The same CV skeleton, for example, is applicable to words like mahǧar 'exile', ʾazraq 'blue', kattab- 'caused to write', or tarǧam- 'translated', so that its information value is somewhat limited. And what exactly determines the vowel melody of a given word? In the end, therefore, the traditional root-pattern analysis seems much more compelling; no one would dispute the fact that other mechanisms like analogy, for example, are at work in Semitic morphology as well (similarly D, 50ff.).

Ch. 4, 'Understanding Arabic verbal semantics' (63-81), tackles the intriguing problem of the lack of systematically predictable semantics in the derived verbal stems, based on the secondary [End Page 634] literature. D states rightfully that no approach so far has yielded satisfactory results. Since D's target language is MSA, it is methodologically strange that he quotes data from several grammars of Classical Arabic (66ff., 71). In Ch. 5, 'Evaluating the pattern III-pattern VI semantic relationship' (83-102), D examines different semantic notions such as mutuality, reciprocity, or simulative meaning for patterns III and VI, again presenting valuable statistical data. D justifiably rejects conativity as the basic meaning of pattern III; instead he proposes implied mutuality as the dominant function.

Ch. 6 is devoted to 'Transitivity and valency' (103-30). After defining transitivity in Arabic, which also involves prepositional objects (105ff.), D discusses valency, showing that, while the bulk of pattern III verbs are transitive (two arguments) and pattern VI verbs intransitive (one argument), 24% of pattern III verbs have valencies greater than two and 44% of pattern VI greater than one (112). The ta-prefix of pattern VI is presented as essentially detransitivizing in deriving pattern VI from III (113). The chapter ends with the discussion of verbs that are difficult to classify.

Ch. 7 examines 'The pattern III template: From form to meaning' (131-56) by evaluating formal characteristics like the vowel length in patterns III and VI, which, according to D, represents verbal plurality. D rightfully rejects proposals that draw a parallel between vowel lengthening in these patterns and in certain internal plural patterns (140). The rest of the chapter is devoted to 'the long ā in Arabic morphology' (143-56), for which D proposes a fairly vague 'consistent aspectual significance' in the C₁āC₂ (and C₂āC₃) sequence of verbs and nouns (156). This is continued in Ch. 8, 'An aspectual model for Modern Standard Arabic' (157-87), in which aspect is analyzed and defined by introducing Zeno Vendler's aspectual categories (162ff.) and Mari Olsen's scheme (165ff.).

Ch. 9, 'Aspectual categorisation of patterns III and VI' (189-211), applies the discussion of the previous chapter to D's data and introduces his hypothesis that the C₁āC₂ sequence in the verbal stems III and VI marks atelicity. This is certainly an important result, but seems somewhat vague, as many verbs in other verbal stems are atelic as well. In this chapter, D makes extensive use of data obtained via arabiCorpus, once more presenting valuable statistical data (93% of actual usage of patterns III and VI, for example, is attributable to verbs in categories unmarked for telicity, 211). In addition, D pays special attention to verbs of inception (201ff.).

Ch. 10 further elaborates on 'Inceptive aspect' (213-35) by introducing a new category of INCEPTIVE for verbs like Habila 'be/become pregnant' (213ff.) and subdividing it into INCEPTIVES OF STATE and INCEPTIVES OF ACTIVITY. The introduction of these new categories is doubtlessly problematic and probably even superfluous given that any of the basic aspectual categories established by Vendler may be induced instead (cf. Van Valin & LaPolla 1997:84). This, incidentally, easily explains the behavior of the verb Habila above. The short Ch. 11, 'The passive in patterns III and VI' (237-46), draws attention to the use of this diathesis in patterns III and VI, but cannot give much information since the passive is very rare (see pp. 238, 245), the exception being the participles (241ff.). 'Conclusions' (247-55), which summarizes the findings of the book, the bibliography (257-63), four short appendices (265-75), and a name index (277-78) and subject index (279-81) conclude the book.

D's approach, which aims at two different target audiences, inevitably leads in most chapters to many redundancies for one or both of them. This obviously cannot be helped and may be considered somewhat cumbersome, but in the end, it may also be advantageous for the clarity of D's account. Some aspects of his methodology are problematic. It is generally amiss to consult native speakers of Arabic on MSA since their mother tongue is a spoken variety, not the standard language. Therefore, the native speakers' very ambiguous reactions are not surprising (i.e. 'lack of agreement between informants', 234; 'the response from native speaker informants is confused', 244). Some passages do not seem entirely relevant for D's subject. See, for instance, the discussions of the verbal noun (masdar; 149ff.) or of the particle qad (161ff., 178). In his account of the 'aspectual model' (169), D quotes some examples that primarily show the aspectual differentiation between the simple past (faعala) and the complex tense kāna yafعalu, not necessarily the aspectual value of pattern III. In some sections, it would have been helpful to cite German-language publications, for example, Carl Brockelmann's Grundriss (1908-13) on comparative Semitic linguistics [End Page 635] (73ff.) or Ernst Jenni's fundamental study (1968) on the Hebrew Piʿel (75ff.). Finally, the Arabic data are not entirely reliable or correct: for example, ʾayyūb 'Job' for ʾayūb (107); maktab 'office' instead of aktab (133; also Arabic script); yabnūn 'they build' rather than yabniyūn (178; also Arabic script); Husayn 'Hussein' for Hussayn (206, also 230); عishrīn 'twenty' instead of عashrīn (207 twice); (a)llāh 'god' rather than llah (207 thrice); tisعat ʾashhur 'nine months' and tisع sinīn 'nine years' for tisعa ʾashhur and tisعat sinīn (219); and buTūla 'championship' instead of baTūla (229).

There are additional concerns as well. For example, the hypothesis (51) that verbs sharing two root consonants have closely related meanings, as quoted from Bohas 2006, is much older (see the discussion in Voigt 1988:80ff.). The 'p-stem' verb (yafعalu) and the participle are hardly interchangeable in circumstantial clauses (146). The claim that the equivalent of the agentive by-phrase in the passive is absent in Arabic (148, 239) is untrue, even for Classical Arabic (see Ullmann 1989:76ff.). The clause as cited in No. 133 (149) is ungrammatical: wa-عaynayhā maftūHa should be wa-عaynāhā maftūHa (incomplete quote?). The negation lam requires the so-called apocopate ~ jussive yafعal, not the subjunctive yafعala (159). The use of the 's-stem' (past faعala) in conditional clauses with ʾidhā 'if' is not 'somewhat archaic' (160) as it is still the norm in MSA. Concerning his corpus, D is certainly right that the language of newspapers is less susceptible to archaism, poeticism, or colloquialism than literature (193), but it is susceptible to interference from European stylistic models instead. The masdar of pattern VI is taC1āC2uC3 (tafāعul), not mutaC₁āC₂aC₃a (= masdar III; 246).

The book has a clear structure and stringent argumentation. Its approach is intriguing and provides some interesting ideas and results. But in the end, in view of the dearth of actual language data, some questionable claims, and other minor weaknesses, neither of the target audiences may be wholly satisfied.