There are expressions whose interpretation seems to bring in additional content which, in principle, is not part of the semantics of any of the units that make them up, as in Peter began the book, interpreted as ‘Peter began [reading / writing /…] the book’. This constitutes a challenge for compositional approaches, since the expected isomorphism between syntactic structure and semantic structure is absent.
At the core of this phenomenon is a mismatch in the semantic features of some of the units that are composed together. The clash, however, does not lead to ungrammaticality, but triggers an interpretive operation designed to restore compatibility, by adding the content required to adjust the semantic features that are at play.
The best-known proposals related to these phenomena have focussed on mismatches that affect lexical-conceptual content. Among the notions that have been applied to explaining how these processes are resolved are those of type-shift and type mismatch, particularly since Partee & Rooth (1983) and Partee (1987, 2003, 2007); fitting operations take place to resolve these conflicts, and allow semantic composition to follow the usual patterns (Winter 2007). From a somewhat different theoretical standpoint, the notion of coercion has been developed, especially after work such as de Swart (1998, 2003, 2011).
The goal of this workshop is to tackle those cases in which at least one of the conflicting features encodes grammatical or procedural meaning. Among these are the functional projections responsible for reference assignment in the nominal and verbal domains (classifiers, determination, quantification, tense, mood, aspect, evidentiality), for the manifestation of information structure (topics, foci), and for individualising propositional attitude and illocutionary force information. Most significantly, adjustment operations are, in many respects, systematic and predictable, and do not have the strong contextual dependency that characterises conversational pragmatic processes, which suggests that the interpretive processes are partially governed by grammatical features.
The core questions we intend to focus on include the following:
Under what conditions are non-matching feature combinations allowed?
Between which units, and levels of representation, are mismatches legitimated?
What are the status and properties of the operations for restoring mismatches?
What are the limits and limitations of these processes?
What are the implications and consequences of mismatch resolution processes for linguistic theory in general, and for the semantics/pragmatics interface in particular?
What are the consequences and implications of mismatch resolution for language acquisition and learning?
What do these processes tell us about the way in which human cognition represents and manages meaning?