Safir Ken. The Syntax of Anaphora

Oxford University Press, 2004. — 329 p. — (Oxford Studies in Comparative Syntax).This book establishes the need for a competitive approach to the distribution and interpretation of anaphoric relations in natural language, and makes a particular proposal about the sort of competitive theory of anaphora that might be on the right track. Linguists are especially interested in anaphoric relations because they provide evidence for the existence of an innate human language faculty in that they are conditioned by both structural relationships in natural language syntax and choices of morphological form. This work is another attempt to develop an explanatory theory of anaphoric relations general enough to capture the deep similarities across languages, while at the same time one flexible enough to account for the variety of patterns that have been observed to be possible. Rejecting binding theory and predicate-based approaches to Principle B and Principle C, which are both derived in this approach from the Form-to-Interpretation Principle, the competitive approach does however exploit c-command and dependent identity relations to determine which morphological forms that represent dependent identity are in competition for a given anaphoric interpretation. The morphology of anaphors, pronouns and referring expressions then plays a role in how they compete. For example, where a reflexive anaphor is possible, as in “John praised himself”, less dependent forms, like “him” and “John” are excluded for the interpretation where John is the x such that x praised x. This simple approach (with its antecedents in the literature that are explored in some detail) is then defended across a wide range of languages and constructions where it provides unique explanatory force, and in so doing, apparent counterexamples are addressed, many in detail. Ancillary proposals about the morphology of anaphors and the locality of anaphoric relations are then defended, and the variety of anaphoric interpretations are explored with respect to their role in determining the nature of dependent identity and the conditions on dependency competitions regulated by the Form-to-Interpretation Principle. The role of competitive algorithms in the architecture of generative grammar is then explored, and other minimalist approaches to anaphora are compared with this one.