Reviews

Reviews Links & Letters 1, 1994 125 Reviews CHRISTIAN MAIR. Injnitival Complement Clauses in English: A Study of Syntax in Discourse. Studies in English Language. Cambridge: Cambridge University Press, 1990. vii+264 pages. This book, the first in the new descriptively and empirically oriented CUP series Studies in English Language, is a corpusbased monogra~h on to-infinitival com- U I plement clauses in English. It is organised into four chapters: an introduction on methodological issues (chapter l), two core chapters on to-in- finitival clauses functioning as subjects and as complements to verbs respectively (chapters 2 and 3) and a conclusion (chapter 4). 1 will start with a discussion of Mair's methodological perspective and premises. The author sets these out by asserting: A belief that the syntactic structure of a language is nota well-defined autonomous formalism but closely interrelated with and partly determined by semantic and pragmatic factors (Mair 1990: 3). Hence the subtitle a study ofsyntax in discourse. The approach to syntax Mair espouses is known as functionalism and is associated with the linguist Talmy Givón (Givón 1979). There is nothing inherently wrong with such a perspective, except when it leads the author to observe that: [I]t is a corollary of the view of syntax adopted here that such widely unchallenged linguistic conventions as the distinction between competence and performance, or between grammaticality and acceptability ... cannot be taken for granted any longer (Mair 1990: 6). Not everyone would agree that the Chomskyan notions mentioned here have not been challenged (Hymes 1971, Labov 1972), and, furthermore, a substantial number of linguists would disagree with the conclusion Mair draws. Throughout the book often misguided sentiments such as these are reiterated and Mair feels an apparent urge to come out strongly against theoretical approaches to syntax, notably Chomsky's framework. However, from the fact that the latest Chomsky referente is 1981 it appears he reaily properly acquainted himselfwith recent work in Government-Bindine U Theory or Principles and Pa-rameters Theorv (Chomskv and Lasnik, forthcoming). he book thus suffers fiom an unfortunate intransigent orthodoxyvis-a-vis other approaches to syntax. Of course, this same intolerance is also found in manv theoreticallv oriented studies. and one can only regret the lack of open-mind- edness on both 'sides'. A further criticism regarding Mair's methodology is that, although it is 'ideologically guided by the functionalist approach to language, it can aptly be described as eclectic in that it does not really have a ~rincioled basis. This becomes ob- I I vious in the introductory chapter where 126 Linlcs & Letters 1, 1994 Mair discusses the various uses that different linguists have made of the term cornplement. He notes that it is not always obvious whether a particular constituent is a complernent or a peripheral unit of sorne sort, and because there is gradience in this area with fuzzy boundaries between the various types of infinitival clauses he decides not to cornrnit hirnself to a definition of the notion. Of course, cornplernents are volatile entities, even in current theoretical work, but nevertheless it seerns that defining one's terrns is an absolutel~ rninirnal requirement for any type of study, whatever the sphere of investigation and rnethodological outlook, and one cannot sirnply opt out of pinning down oneS terms of reference. The lack of a secure principled methodological foundation and the refusal to cornrnit hirnself to defining his terms has undesirable consequences in that it leads to irnprecise and ad hoc staternents. An example of this is to be found in Mair's discussion of (extraposed) for-infinitival subject clauses of the following type: It will be necessalyfor me to se11 my car (Mair 1990: 40). The author rightly rejects the possibility of takingfor me to be a Prepositional Phrase and says that: [I]t could be described as a complex subject noun phrase, consisting of the subject-in- troducing particle for and the subject proper ... This descriptive staternent is unhelpfül in that it leaves the reader wondering exactly what is the syntactic status of he so-called subject-introducing particles. The ~ossibility of anal~sing for as a complernentiser is not considered. Mair seerns to be aware of the fact that fünctionalisrn is perhaps not descriptively rigid enough: [1] t is weil to remind oneself that - what-ever its drawbacks - the formalist approach to syntax had [sic] one virtue. Its claims were explicit enough to allow criticism - both from inside and outside a given theoretical frarnework ... What are the criteria to decide whether a functional generalisation works or not? And what is the degree of rigour required even in the formulation of a functional generalisation? ... Clearly, the functional paradigm cannot succeed unless the utmost effort is spent on increasingly precise descriptions of the unformalisable that are richly documented by empirical evidence (Mair 1990: 11). One rnight wonder at this juncture whether beingprecise about the unformalisable is a contradiction in terms. This rnonograph is not principally concerned with grarnrnatical analysis (it broadly follows the descriptive frarnework of Quirk et al. 1985), but rather with explaining discourse syntactic processes. It is in this area that Mair has the most interesting things to say. A good example is his discussion of infinitival subject clauses (ISCs). There are three types of predicates that can take infinitival clauses as subjects, exernplified by the sentences below: (1) It is unhealthy to be in tbesunfor too long. (2) It is a crime to throw rubbish onto tbe sheets. (3) You see the rising sun as a god. You know he isn't a god, but to tbink of tbe rising sun as a god explains something about the human feeling, about the new day coming and the excitement of the recurrence of the seasons and al1 that (Mair 1990: 27). From these sentences it can be seen that adjectival, nominal and verbal predicates take ISCs. In (1) and (2) the subject clauses have been extra-posed and this is because long andlor cornplex phrases tend to occur sentence-finally (the principles of end weight and endfocw). In (3), a slightly arnended attested example transcribed frorn a radio broadcast, the subiect clause rernains in situ. Mair discusses the sernantic restrictions on the three types of predicates in (1) - (3) and makes a nurnber of observations regarding ISCs narnely that in general they are infrequent and in the vast rnajority of cases they are extraposed. The situation for ISCs depending on verbal predicates, as in (3), is exceptional. According to the author these Reviews Links & Letters 1, 1994 127 are less likely to extrapose and are even less QUIRK, R., S. GREENBAUM, G. LEECH frequent. There are good reasons for this. and J. SVAR~K. (1985). A Com- The resistance to extraposition can be prehensive Grammar of the EnglkhLanexplained by obsewing that verbal predigccage. London: Longman. cates tend to be quite long and informationally heavy. This is especially clear in Bas Aarts (3) above. The subject dause in this sen- Department of English Language and tence is inf~rmationall~ not prominent ~i~~~~~~~~ (the risingsun is mentioned in the directly university college London preceding context) and hence moving it to a sentence-final position would disturb the informational balance of the sentence. The reason why ISCs depending on ver- ANDREW SPENCER. ~~~h~l~+l bal predicates are extremely rare is that Theory Oxford: Basil Blacbell, verbal predicates are usually dynamic and require an animate subject and clearly 19g0. 512 Pages. ISCs come low on what has been called the anima9 hierarchy. Spencer's book is a nearly exhaustive in- The final evaluation of &is monotroductory to morphology, with exercises graph is that, on the whole, it is a meritofrom very &fferent languages, and recomrious piece ofwork, very well-written and mendations for further reading at the end well-researched, with interesting contri- of each chapter. The book also contains butions to our knowledge of discourse a subject index, a name index and a iansyntactic processes. It is, however, disguage index, which makes looking for appointing as far as analytical content and very specific information an easy and f'st the rather narrow-minded methodologitask. In addition, the most important cal out-look are concerned terms are written in boldface the first time they appear in the text. Spencer's book pressuposes little knowledge of mor- References phology, but discusses fairly complicated issues, which can however be easily understood because of his accessible style CHOMSKY, N. (1981). Lectures on Gov- and a gradual introduction to the most mmmta&Bi&ing Dordr&t: Foris. intricate aspects of morphological theory. CHOMSKY, N. and H. Lasnik (forthcom- Morphology is a difficult subject, ing). "Principies and Pararneters with a lot of complex relations with pho- Theory". J. Jacobs, A. von Stechow, nology, syntax and semantics; it interacts W. Sternefeld and T. Vennemann with al1 the components of the grammar. (eds.). Syntax: An InternationalHand- The borderline between phonology and book of Contemporary Research. Berlin morphology, for instance, is not clear at and New York: Walter de Gruyter. all, and a particular sequence might be GIVON, T. (1979). On Understanding considered an allophone or an allomorph Grammar. New York: Academic Press. depending only on the theoretical frame- HYMES, D. (1971). "Competente and work adopted. Spencer's book reflects al1 Preformance in Linguistic Theory'. the relations morphology has with other R. Huxley and E. Ingram (eds.) Lancomponents of the grammar (except for guage Acquisition: Models and Mesemantics, which receives little attention), thodr, pp. 3-24. New York: Academic and he reviews the most important pro- Press. posals that have been made with respect LABOV,W.(~~~~). SociolinguisticPatterns. to the subject from different points of Philadelphia: University of Pennsylview, most especially within the framevania Press. work of generative grammar.