The Syntax and Pragmatics of Anaphora
A Study with Special Reference to Chinese
Cambridge Studies in Linguistics, 70. Cambridge: University Press, 1994
(Paperback re-issue 2007) 331 pages. ISBN 0-521-03960-6. £38.
Reviewed by Laure Gardelle
École Normale Supérieure de Lyon
This study sets itself against the Government Binding analysis of anaphora as a purely syntactic phenomenon and proposes instead a pragmatic approach, based on the neo-Gricean model of conversational implicatures. Its basic tenet is that pragmatics, far from being what Chomsky termed an “epiphenomenon at best”, plays a major role in the production and interpretation of an anaphoric expression, even in intrasentential anaphora. Huang does not deny the influence of syntactic structure and semantic representations in the process, but argues that pragmatic inferences provide indispensable complementary and explanatory principles which constrain the interpretation and production of an anaphoric expression. The study focuses on Chinese, which has been termed a “pragmatic” language (along with Japanese and Korean, for instance), in the sense that constraints on anaphora are primarily due to principles of language use, as opposed to French, English or German, for example, which allegedly show more grammatical constraints. The book does not address all cases of anaphora, but prototypical ones (called “core anaphora” by the author), namely NPs with a co-referring textual antecedent; for those, it studies zero anaphora, pronouns, reflexives and lexical NPs.
Analyses are organised in three parts. Chapter 1 surveys the latest developments of neo-Gricean theory, starting from Grice’s approach, then examining several models, such as Relevance Theory and the models subsequently developed by Horn or Levinson, and finally outlining Huang’s own pragmatic approach. As the author notes, this chapter can be read independently. The argument of the book is developed in the next seven chapters. The author begins by showing the inadequacies of syntactic theories when dealing with Chinese, whether they be Chomsky’s typology of Empty Categories (chapter 2), Control (chapter 3) or Binding theories on long-distance reflexivisation (chapter 4). He goes on to substantiate his neo-Gricean approach, applying it first to intrasentential anaphora (chapters 5 and 6) and then extending it to discourse anaphora, based on a corpus of authentic conversations; he focuses alternatively on anaphora production (chapter 7) and resolution (chapter 8). The book closes on a final chapter which summarises the major findings and draws theoretical implications on the relationship between pragmatics and syntax (chapter 9).
In his analysis of zero anaphora, which is so frequent in Chinese that it has been termed the “norm”, the author sharply criticises Chomsky’s typology of Empty Categories, questioning the existence of PRO, pro and empty topics. He also shows that control cannot be adequately accommodated under a syntactic approach – but notes that the answer cannot lie in a purely thematic or semantic approach, such as those proposed by Fillmore and Jackendoff. Finally, a syntactic approach cannot deal with long-distance reflexivisation, in that it cannot specify the domain or the set of possible antecedents. A pragmatic, neo-Gricean theory, on the other hand, provides suitable answers. The author posits that on the one hand, anaphora is determined largely by the systematic interaction of the neo-Gricean interpretation principles set by Levinson: the I[nformativeness]- and M[anner]-principles, in that order of priority, which work in relation to the Q[uantity]-principle. Each of those principles affects the speaker and the recipient differently:
Speaker’s maxim: the maxim of minimisation (“Say as little as necessary”, i.e. produce the minimal linguistic information sufficient to achieve your communicational ends, bearing in mind the Q[uantity]-principle: do not provide a statement that is informationally weaker than your knowledge of the world allows, unless providing a stronger statement would contravene the I-principle)
Recipient’s corollary: the rule of enrichment (amplify the informational content of the speaker’s utterance by finding the most specific interpretation, up to what you judge to be the speaker’s m-intended point)
Speaker’s maxim: do not use a prolix, obscure or marked expression without reason.
Recipient’s corollary: if the speaker used a prolix or marked expression M, he did not mean the same as he would have had he used the unmarked expression U – specifically he was trying to avoid the stereotypical associations and I-implications of U.
In addition to those interpretation principles, there are “consistency constraints”, namely DRP (Disjoint Reference Principle, established by Farmer and Hamish, according to which the arguments of a predicate are intended to be disjoint, unless marked otherwise), information saliency and general implicature constraints – such as background assumptions. The author stresses that, to him, DRP is of a pragmatic, and not of a syntactic, nature; in other words, it is a matter of usage preference, based on world knowledge. This tenet has the advantage of automatically preventing any inconsistent pragmatic implicature from arising, or at least of cancelling it without violation of the Q > M > I hierarchy.
Building on those principles, the author establishes a hierarchy among the anaphoric types under study: zero (as the minimal type) < pronoun < reflexive < lexical NP. Use of one form implicates a local co-referential interpretation, unless a form to the left could have been used, in which case the recipient has to look for a reason why the “default” form was not selected. About discourse anaphora, Huang notes that in the literature on anaphora, it is widely acknowledged that two principles operate in determining reference: economy and clarity. To him, this is evidence that pragmatics are at play. Focusing on the supposedly “prototypical” use of anaphora – third-person reference to human entities – he concludes that production is dominated by the Q- and I-principles, while in anaphora resolution, the I- and M-principles prevail. Those analyses lead the author to consider the relationship between syntax and pragmatics. He shows them to be definitely interconnected and, following Levinson, suggests that grammatical rules, although they do belong to the realm of grammar, could be in fact “frozen pragmatics”, in other words, that they came to structure the language because they corresponded to pragmatic use. This is a final challenge to Chomsky’s syntactic theory, as it questions the very existence of a biologically determined human linguistic faculty which would justify a Universal Grammar.
The book makes a major theoretical contribution to research on anaphora. In addition to being extremely clear, even in its most technical arguments, it presents a unified theory of both intrasentential and discourse anaphora. The focus on Chinese, which obviously shows specificities, also sheds new light on some major theoretical issues. Because of the neo-Gricean framework, it sometimes takes a little flicking back and forth to understand ideas that one feels might have been expressed in a simpler way. For instance, the author writes: “the use of a pronoun will I-implicate a local coreferential interpretation, unless the pronoun is used where a zero anaphor could occur, in which case, the use of the pronoun will M-implicate the complement of the I-implicature associated with the use of the zero anaphor”. What is meant here is basically that the use of a pronoun when it is in competition with a zero anaphor is motivated, and that the recipient should therefore specifically take the additional information into account. This principle, which is the author’s thesis and is extended to each expression type under study, is very powerful and meets other studies outside the neo-Gricean framework, such as analyses of the choice of lexical information in NPs. It also calls for further studies on anaphora; in particular, it is necessary to re-examine the relationship between pragmatics, cognition and grammar. While not denying the role of syntactic rules, Huang posits that pragmatics govern grammar, at least in Chinese. This tenet recalls cognitive approaches, in which the choice of an anaphoric expression type is not depicted in terms of the linguistic coding of the expression type, either, but in terms of the status of the referent in the mental representation.
One question that remains to be addressed, therefore, is whether there is a relationship between mental status and conversational implicatures, and whether a theoretical model could unify those approaches. Besides, as Huang suggests, criticising purely syntactic analyses does not mean doing away with the grammatical aspects of anaphora; in particular, such approaches pose the question of the link between the pragmatic (or the cognitive) use of a form and the linguistic (procedural and / or qualitative) information it encodes. Huang had set himself such a broad scope already that he could not have addressed such questions, but his book is an invaluable stepping stone in the understanding of those issues and, ultimately, of anaphora.
Cercles © 2010