Discourse perspectives on syntax Edited by Flora Klein-Andreu (review)

REVIEWS163 reason for being there, and on seeing linguistics as (properly) an intellectual crossroads rather than an isolation chamber. At times, Givón is perhaps carried away with the logic of his form/function correlations (and his evolutionary sequences), and the particulars of individual language systems are not picked up; but the over-all design of his theory is sufficiently supple that these matters could be handled merely by a change in emphasis. [Received 12 June 1985.] Discourse perspectives on syntax. Edited by Flora Klein-Andreu. New York: Academic Press, 1983. Pp. xvii, 266. $34.50. Reviewed by Bernard Comrie, University of Southern California What is the correct relation between discourse and syntax? In principle, one might take a number of stands, of which the following are some of the main ones (though there are also many intermediate positions): (a)Syntax should be analysed in complete isolation from discourse, discourse properties being used neither in the statement of syntactic regularities nor as a methodological crutch in gaining insight into syntactic regularities. (b)Discourse may (or even must) be used methodologically, in enabling the analyst to gain insight into the correct characterization of syntactic phenomena; but it does not form part of the statements used in a syntactic description. (c)The very statement of syntactic regularities is incomplete without explicit reference to discourse. (In its extreme version, this possibility would involve the very denial of syntax as a separate phenomenon, all syntax being reduced to discourse.) Of course, one might also take different stands among these with respect to different syntactic phenomena. Needless to say, none of the contributions to this collection takes stand (a): all are concerned with ways in which the study of syntactic phenomena in context can throw light on those syntactic phenomena. Thus Ellen Contini-Morava, 'Relative tense in discourse: The inference of time orientation in Swahili' (3-21), argues that the meaning of a relative tense is simply temporal location relative to some point of orientation. It is up to the hearer to infer the point of orientation, which cannot be derived automatically from the syntax (e.g. from the time reference of time adverbials, or of nearby absolute tense forms). Barbara A. Fox, 'The discourse function ofthe participle in Ancient Greek' (23-41), claims that Ancient Greek participles (as opposed in particular to finite verb forms) are used to refer to backgrounded events, i.e. those that do not occur in temporally ordered sequences . Similarly, Sandra A. Thompson, 'Grammar and discourse: The English detached participial clause' (42-66), suggests that such clauses serve a local backgrounding function. Paul J. Hopper, 'Ergative, passive, and active in Malay narrative' (67-88), presents an interesting characterization of the ergative construction as non-predicational, with emphasis on the verb—and therefore used for sequenced events in a narrative—while the active and passive are both predicational : the active says something about the agent, and the passive about the patient. Benji Wald, 'Referents and topic within and across discourse units: Observations from current vernacular English ' (91-116), gives a detailed analysis of the discourse functions of 'new-rWi' (as in There was this guy) as a way of introducing a new participant into a story—a phenomenon that seems to be remarkably recent in English, with no attestations before the 1930's (94). 164LANGUAGE, VOLUME 62, NUMBER 1 (1986) Carmen Silva-CorvalAn, 'On the interaction of word order and intonation: Some OV constructions in Spanish' (117-40), argues that the preposed object in OVX constructions can have different discourse functions (discourse link, center of attention, focus of contrast, contradicted expectation) which are dependent on the intonation contour; perhaps my initial question on the relation between syntax and discourse should therefore be formulated as relating rather to the relation between grammar (which would include intonation) and discourse. Flora Klein-Andreu, 'Grammar in style: Spanish adjective placement' (143-79), presents a unitary characterization of the difference between pre- and postposition of adjectives: the latter is contrastive (i.e. establishes a difference), whereas the former merely characterizes without establishing a difference. This account subsumes the various lists of factors and exceptions of traditional analyses, which KA attacks for failing to consider data in...