Causatives and causation: A universal-typological perspective by Jae Jung Song (review)

670 LANGUAGE, VOLUME 74, NUMBER 3 (1998) (Nina Elenbaas, 61-72); (4) Phonology: 'Metrical complexity' (Jan Kooij, 147-58), 'Dutch orthography : A near-optimal phonological transcription?' (Anneke Nunn & Anneke Neijt, 195-206), 'Emphasis spread in two dialects of Palestinian' (Ruben van de Vijver, 245-56); (5) Phonetics: 'Lexical stress and spoken word recognition: Dutch vs. English ' (Klaske van Leyden & Vincent J. vanHeuven , 159-70), 'On the rise and the fall of Spanish diphthongs' (Elisabeth Mauder & Vincent J. van Heuven, 171-82); (6) Syntax and Semantics: ? dynamic binding approach to intervention effects on negative polarity item licensing' (Martin Honcoop, 85-96), 'How "greedy" is the French imperative?' (Aafke Hulk, 97-108); (7) Syntax: 'Inalienable possession m locational construction: An apparent problem' (Hans Broekhuis, Leonie Cornips, and Maarten de Wind, 37-48), 'On heads and the linear correspondence axiom' (Hans Broekhuis, 25-36), 'Associative DPs' (Hans den Besten, 13-24), 'Object drop in Dutch imperatives' (Janneke Visser, 257-268), 'The spread of the reflexive adjunct middle in the Limburg dialects: 1885-1994' (Leonie Cornips, 49-60), 'Constructions of inalienable possession : The role of inflectional morphology' (Sergio Baauw, 1-12), 'Constraint interaction in binding and the feature specification of anaphoric forms' (Sergio Menuzzi, 183-94); and (8) Semantics: 'The nature of quantification of high-degree: 'very', 'many', and the exclamative' (Gertjan Postma, 207-20), 'Boundary tones and the semantics of the Dutch final particles hè, hoor, zeg and./'oft' (Robert S. Kirsner & Vincent J. van Heuven, 133-44). Most of the papers are written within the latest theories of their corresponding fields. As is often the case in collections of papers from a linguistics association meeting, papers on syntax have the lion's share (7 out of 22). Most of them are within one or another of the (too?) many different proposed analyses of the minimalist program. The second in importance (well after syntax) is phonology with 3 papers (of which 2 are written within the optimality framework ). In conclusion, this book offers a good overview of what is happening in linguistics in general and in Dutch linguistics in particular. [Alain ThéRiAULT , Université de Montréal.] Causatives and causation: A universaltypological perspective. By Jae Jung Song. (Longman linguistics library.) London & New York: Addison Wesley Longman Limited, 1996. Pp. xvi, 295. Song purports to provide here a novel, universally valid typology of causative expression based on an apparently cursory examination of causative constructions in 613 languages. Nevertheless, in the space of his 187-page text (188-295 being taken up by a 'database' index, a bibliography, and a subject index), S actually cites causative sentences that he has culled from the grammars of only 92, or 15%, of these languages. An original feature of S's typology is his concentration on causative constructions with factitive meaning to the almost complete exclusion of any with permissive and assistive meaning. Not only does S not concern himself with this basic semantic trichotomy, but occasionally he also conflates factitive causative constructions with mere factitive ones (cf., for example, the noncausative, factitive Houailou examples na waa vc-bavara 'he flattens it' and na waa ?e-a 'he improves it' [53]). Reproducing isolated examples of causative sentences from the grammars of the 92 languages referred to above, S asserts that causative relations universally find expression in one of three types: the compact type of causative construction, the and type, and the purp (purpose) type. S's compact type includes both lexical causatives (e.g. English kill 'cause to die', feed 'cause to eat', etc.) and morphological ones (e.g. Turkish Ali Hasani öldürdü 'Ali killed Hasan', which exhibits the causative suffix -DIR- as opposed to noncausative Hasan oldii 'Hasan died'). S views lexical and morphological causatives as 'compact' because they are characterized by a formal 'fusion' of the two elements (or phases) that constitute any causative situation, viz. '[Vcause]' and '[Veffect]'. The and type of causative construction consists of two more or less independent clauses, one of which expresses [Vcause]; the other—typically a coordinate ('and so') clause—expresses [Veffect] as a result or consequence of [Vcause], According to S, a number of African and South Asian languages exhibit the and type of causative by virtue of the...