Cohesion in English By Michael A. K. Halliday and Ruqaiya Hasan (review)

676LANGUAGE, VOLUME 54, NUMBER 3 (1978) into an event' (194). The argument strikes me as faulty: unless we make the unusual assumption (and there is no indication that Menzel makes it) that non-underlying representations must conform to semantic well-formedness constraints, one cannot motivate the application of a syntactic rule such as Equi-NP Deletion in terms of semantic well-formedness. If a verb requires an embedded action, then the embedded action must be well-formed, i.e. subjectless, in the underlying representation. Besides unconvincing arguments, the other drawback of the book is inexplicitness and general infelicity of organization and presentation. No precise delimitation of subject matter, of the range of facts to be accounted for, of basic assumptions, or of proposed explanatory principles is provided anywhere. We are given no definition of such key concepts as 'complement', 'surface structure', or 'deep structure'. The semantic classification of sentences is given in at least two different versions: on p. 40, 'proposition', 'event', 'process', 'action', 'state', 'property', and 'relationship' are mentioned; on p. 200, 'proposition' and 'relationship' are omitted, but 'act' and 'activity are added. Again, on p. 60 it seems that actions, acts, and processes are subtypes of events, while states and properties are subtypes of 'states' (in a generic sense)—an idea raised here but never picked up again (cf. 200). The status of 'propositions' is unclear: if they are coincident with declarative sentences (41), then what exactly is the difference between the labels 'declarative sentence' and 'proposition'? As already mentioned, the place of the semantic distinction between fact and non-fact also remains unclear in the proposed classifications . In addition to such unclear and imprecise passages, the structure of the book places an undue burden on the reader. The actual argument starts in Chapter 3 (p. 63); the preceding two chapters provide a survey of the transformational literature on complementation and of the semantic classification of sentences. Although Menzel repeatedly emphasizes (13 ff., 21) that in these discussions he only talks about things that will be relevant for his own ensuing argument, these two chapters seem somewhat pointless : since the readers do not yet know what Menzel's argument will be, they do not know what to look for in these surveys. A brief introductory section to the book, summarizing its contentions, would have provided guidelines for plowing through these two survey chapters. In spite of its deficiencies of argumentation and presentation, I found Menzel's book useful. Arguments for the main thesis—that at least some syntactic and semantic categories define coextensive classes—are at times unconvincing, but at other times sound. Thus 1 find Menzel's explanations for collocational restrictions among main-clause verbs, head nouns, and complement-clause types, in terms of semantic well-formedness constraints, to be convincing. All in all, the monograph is a somewhat inexplicit and disorderly discussion of the relationship of semantic and syntactic categories within the domain of English complement constructions. As such, it provides difficult but thought-provoking reading for anybody interested in English grammar, in complementation, or in the general relation between meaning and form in language. [Received 25 July 1977.] Cohesion in English. By Michael A. K. Halliday and Ruqaiya Hasan. (English language series, 9.) London: Longman, 1976. Pp. xv, 374. $20.00. Reviewed by Wolfgang U. Dressler, University of Vienna This is a very significant contribution to the study of intersentential coherence and of its interaction with sentence structure. It represents the fruit of about ten years of joint research undertaken by the founder of Systemic (or Scale-and- REVIEWS677 Category) Grammar and his wife, who has specialized in stylistics and discourse analysis. In fact, the first three chapters are a revised version of Hasan 1968; chapters 4 through 6 were written by both authors; and the last two chapters were added by Halliday. Cohesion is seen as 'a necessary though not a sufficient condition for the creation of text' (298). A text is 'a unit of situational-semantic organization' (25), either of written or oral speech, and may be of any length. It 'is not a structural unit'; nor is cohesion 'a structural relation' (6), but a semantic concept. The basic types of cohesive ties...