In lieu of an abstract, here is a brief excerpt of the content:

Reviewed by:
  • Anaphora: A cross-linguistic study by Yan Huang
  • Peter Siemund
Anaphora: A cross-linguistic study. By Yan Huang. Oxford: Oxford University Press, 2000. Pp. 396. ISBN: 0-19-823528-3. $35.00.

Huang’s book is extremely rich in coverage. Topics include NP-anaphora, VP-anaphora, discourse anaphora, null subjects and objects, reflexivity, long-distance reflexives, lo gophoricity, [End Page 634] logophoric pronouns, and switch reference. He gives an account of binding theory and control theory as well as various other syntactic, semantic, and pragmatic approaches to anaphora. According to H, the data discussed in this book are drawn from 550 languages.

The Ch. 1 introduction to the different types of anaphoric devices is followed by an evaluation of syntactic, semantic, and pragmatic approaches to the analysis of anaphora in Chs. 2–4. Ch. 5 takes up switch reference and discourse anaphora for discussion.

To evaluate H’s book in a single statement is not possible. The book certainly is, in the words of its author, ‘an extensive overview of the major contemporary issues surrounding anaphora’ (1), and in this respect it is an excellent and recommendable reference work. But such a breadth of coverage has its trade-offs. Readers who are looking for a coherent treatment of anaphora may expect something slightly different. Although the neo-Gricean pragmatic theory is developed inter alia by the author himself, the discussion of it is brief and contains some loose ends (see below). The same applies to readers who expect a typological study. True, there is extensive empirical coverage, and data from many languages are used for illustration as well as to point to weaknesses in current theorizing. But H’s book is not a systematic exploration of the patterns and limits of variation found across the world’s languages. Nevertheless, the reader who approaches Anaphora with these caveats in mind will be rewarded with a well-written and insightful survey of a highly complex grammatical domain. In view of the extensive scope of H’s book, I concentrate here on two aspects only.

A neo-gricean pragmatic theory of anaphora

From any perspective that includes more than a handful of languages, anaphoric linking, which in the following is delimited to reflexive marking, is a complex and confusing business. Reflexive anaphors vary with respect to the domain they are bound in, the syntactic positions they occur in, the kinds of antecedents they permit, the kinds of predicates they occur with, their morphological complexity, the degree of their morphological independence, the number and kind of additional expressions they stand in opposition to, and the additional functions they have besides reflexive marking. Syntactic theories of reflexive marking that work reasonably well for single languages usually fail once confronted with crosslinguistic variation. It is against this background that one has to see H’s proposal to shift the explanatory burden from syntax to pragmatics, and it certainly makes a lot of sense. Like Levinson (1987, 1991), H assumes a neo-Gricean set of pragmatic principles consisting of the Q-principle (roughly ‘Say as much as you can!’), the I-principle (‘Say as little as necessary!’), and the M-principle (‘Be brief, avoid prolixity!’). These principles are assumed to operate on or against stereotypical assumptions about the way the world, or rather language, is and yield meaning in the form of implicatures above and beyond what is literally stated. The most important background assumptions or generalizations in the reflexive domain can be summarized as follows (cf. 214ff.):

  1. 1. The general pattern of anaphora. Reduced, semantically general anaphoric expressions tend to favor locally coreferential interpretations; full, semantically specific anaphoric expressions tend to favor locally noncoreferential interpretations.

  2. 2. The semantic content hierarchy. Lexical NPs > pronouns > zero anaphors. The inherent semantic content of a lexical NP tends to be semantically more specific than that of a pronoun and the inherent semantic content of a pronoun more so than that of a zero anaphor.

  3. 3. Disjoint reference presumption (DRP). The co-arguments of a predicate are intended to be disjoint unless one of them is reflexive-marked.

  4. 4. Predicate meaning/reflexivization strategy correlation. The more ‘marked’ a reflexivizing situation (e.g. other-directed) is, the more ‘marked’ (i...

pdf

Share