Two dynamic strategies

Abstract

The dynamics of semantic interpretation may be located in two natural components of an interpretive system: either in the processes that construct the syntactic representation, or in those that perform the evaluation in the models. Both strategies have been implemented in current dynamic systems of interpretation; the first for instance in DRT, the second in DPL. This paper compares the two strategies, using the real life example of temporal reasoning in natural language. The focus is on the logical issues, in particular the compatibility of the two forms of dynamics, the structure of their action repertoires, and the interaction of interpretation and inference.