Grammaticalization of the complex sentence: A case study in Chadic By Zygmunt Frajzyngier (review)

BOOK NOTICES 403 (128-48) investigates elliptical clauses containing zero argument forms. T demonstrates that these clause types are unrelated to anaphoric processes in the discourse and fulfill a variety of discourse functions . Ch. 8 (149-73) considers full clauses (those containing a verb and overt core arguments) and submits that overt coding of arguments occurs only for specific functional reasons, such as to report quoted speech or to create a recipient-oriented design. Drawing on this analysis, Ch. 9 (174-94) recognizes three major speech units in Mandarin: nominal and verbal clauses and the XV clause shape. T suggests that these units taken together support the operation of a phrase-based grammar in Mandarin spoken discourse as opposed to one built on traditional clause structure. Ch 10 (194-97) concludes with a brief summary. As T notes, the data used in this analysis are genre specific, and these findings should not be overgeneralized pending further examination of additional types of data. An accessible read for students as well as researchers, T's book provides a solid basis for such investigation. [Lucy Pickering Elliott, University of Florida J Grammaticalization of the complex sentence: A case study in Chadic. By Zygmunt Frajzyngier. (Studies in language companion series 32.) Amsterdam & Philadelphia: John Benjamins, 1996. Pp. 501. In this book, Frajzyngier displays his broad knowledge of Chadic languages and applies it to a very complicated issue: various types of complex sentences and their interrelations F offers many hypotheses , some speculative (as F himself admits), others solidly grounded. These ideas, together with the wealth of data from a relatively unknown language family, make this book an absolute must for everyone interested in the subject of complex sentences (from any theoretical perspective), modality, grammaticalization processes, and historical reconstruction . The book is too complex to review chapter by chapter, so I will limit myself to some of the highlights. Ch. 2 is devoted to a survey of different clause combining mechanisms and their semantic interpretation . As F shows, the Chadic possibilities only partially overlap with those of English and other IndoEuropean languages. The main argument of the book is that there is a distinction between the de dicto domain (the domain of speech) and the de re domain (the domain of reality ), and that Chadic languages have grammaticalized this distinction in a number of ways. This distinction plays a very important role in all areas of Chadic grammar A big chapter (4) is devoted to complements of verbs of saying, and F does a very good job of describing the various issues involved. F shows that the difference de dicto/de re is grammaticalized in a number of ways, such as difference in the order of matrix and embedded clause (for those Chadic languages that allow both). F also deals with the circumstances in which the verb of saying may be omitted (in those cases in which there is no problem understanding the grammatical relations) and how verbs of saying may be grammaticalized into de dicto complementizers Chs. 6-8 are devoted to complements ofvolitional verbs, perception verbs, and verbs of knowing, respectively . AU share a number of features with the de dicto domain, and are therefore coded similarly to complements of verbs of saying. Distinctions can be coded by different complementizers (if there is more than one), whereby the de dicto COMP codes the domain, and the other (called 'deontic' COMP) codes the actual wish, or in the case of perception verbs, the difference in COMPs can encode the difference between direct and indirect perception Another coding mechanism is presence vs absence of the de dicto COMP. Chs 9-10 deal with temporal and conditional clauses, respectively. The protasis can be marked, the apodosis can be marked, both, or neither. F finds that temporal markers can develop into conditional markers or the other way around, showing that gram- ' maticalization can be bidirectional. Ch. 1 1 deals with relative clauses and their structures . The syntactic shape is again dependent upon a de dicto or de re interpretation of the sentence, only this time the entire discourse must be taken into account for a proper interpretation. [Ferdinand de Haan, University of New Mexico.] Norsk grammatikk: Debatt i...