We are happy to announce a talk by Jan Köpping (GU Frankfurt) at the Semantics Colloquium.
Please register beforehand (s.walter@em.uni-frankfurt.de) to receive the access data to zoom on Thursday shortly before the talk starts.
Title: Transparent Negation in Dynamic Semantics
Date: April 29
Time: 4 pm – 6 pm ct
Abstract:
Dynamic semantic systems are designed to capture the truth-conditions of complete sentences as well as certain facts about anaphoricity. The second aspect is described in terms of discourse referents and accessibility: possible antecedents introduce discourse referents which may or may not be accessed by 3rd person pronouns, resulting in an interpretation that lets pronouns covary with their antecedent’s contribution and hence, allowing for bound readings even with indefinite antecedents behind the clause boundary. Within these systems, negation usually plays the role of a plug: it blocks the projection of discourse referents within its scope and thus reders them inaccessible to subsequent anaphoric elements. But, as is well known, this leads to some predictions that are not bourne out: firstly, doubly negated sentences seem to allow for covarying interpretations of external anaphoric elements again, which is unexpected in orthodox systems, since the plugging function of negation is not reversible; and secondly, discourse referents introduced by negated indefinites in the left clause of a disjunction are accessible to pronouns in the respective right part, which is also unexpected, since negated elements are usually not taken to introduce discourse referents at all. These two problems are tackled in several classical papers already, and they are also discussed in some more recent takes on three-valued systems that allow for semi-transparent negations. This talk argues that the range of constructions considered to be relevant for the semantics of negation is not wide enough, and that taking on board several other constructions shows that the assessment that negation acts as plug for discourse referents can only be maintained in a very specfic sense. Furthermore, the broadened empirical landscape suggests a new way of defining negation in dynamic systems that ultimately leads to a way more classical shape of the theory.