Reviewed by Tommi Leung, United Arab Emirates University
Glyn Hicks’ The derivation of anaphoric relations examines how anaphoric relations in binding theory (BT) receive a novel analysis from the perspective of the minimalist program (MP).
Ch. 1 ‘Introduction’ recapitulates the basic history and tenets of BT and introduces the central agenda, i.e. how the current minimalist approach can account for the fruitful results given by BT. This is a major challenge to MP, given that the interest of BT has diminished in recent syntactic works that view BT as external to the narrow syntax. Ch. 2, ‘Binding theory and the minimalist programme’, summarizes the historical development of BT. The second part introduces several important notions in MP, e.g. minimal link condition and phase impenetrability condition. H argues that the theoretical challenge is to account for the myriad observations provided by BT while maintaining a simplistic theory of grammar.
Ch. 3, ‘The binding theory does not apply at LF’, points out the theoretical and empirical problems that emerge if BT applies at the logical form (LF). The evidence comes from the observation that principle B can be influenced by locality, phonological factors, morphological case, and verbal inflections that cannot be accessed at LF. In contrast, principle A potentially applies everywhere, including LF. By listing facts from reconstruction, quantifier scope reversal, and idiom chunk interpretations, H suggests that a narrow-syntactic approach to BT is more favored than the LF-approach.
Ch. 4 ‘Eliminating condition A’ attempts to eliminate condition A as a representational condition of grammar, and claims that anaphoric binding is an instance of Agree defined by MP. An Agree relation is established between the antecedent and the anaphoric with respect to ‘semanticosyntactic features’, e.g. referential features, operator features, and variable features. Since Agree is locality-constrained by means of (LF-)phase, the locality condition on BT can also be properly described.
Ch. 5, ‘Eliminating condition B’, claims that the binding domain for condition B is reduced to the phonetic form(PF)-phase. Condition B differs from condition A in that the binding condition of principle B cannot be reduced to Agree as argued in condition A. Alternatively, for a sentence to be grammatical, corresponding features of the pronoun and its locally c-commanding DP should not bear the same values. H claims that condition B is a violation of economy condition, since establishing dependencies (e.g. through Agree) is to ‘maximize featural economy’, whereas condition B is not.
Ch. 6, ‘Extensions to other Germanic languages’, looks at Dutch, Norwegian, and Icelandic pronominal and anaphoric systems. Crosslinguistic variation is attributed to the lexical feature specifications the languages employ, instead of defining different binding domains. Ch. 7, ‘Conclusions’, concludes that there is no binding theory as far as the particular mechanisms responsible for the binding facts are concerned. H’s attempt to completely eliminate BT from universal grammar, while meriting further empirical tests, is a significant step forward in the minimalist approach to grammar.