Corrections and Higher-Order Unification

  Request Code

  API Access   Call/Text an Expert

We propose an analysis of corrections which models some of the requirements corrections place on context. We then show that this analysis naturally extends to the interaction of corrections with pronominal anaphora on the one hand, and (in)definiteness on the other. The analysis builds on previous unification--based approaches to NL semantics and relies on Higher--Order Unification with Equivalences, a form of unification which takes into account not only syntactic beta-eta-identity but also denotational equivalence.