Mood, Aspect, Modality Revisited, 2016
This paper reflects upon the relation between semantic/pragmatic theory and psycholinguistic expe... more This paper reflects upon the relation between semantic/pragmatic theory and psycholinguistic experiments within the realm of aspectual coercion. What predictions can be derived from standard and not so standard theories for incremental aspectual interpretation? We contrast operator-based and underspecification accounts of aspectual coercion with the Event Calculus of Hamm & van Lambalgen (2005), a pragmatic theory of tense and aspect, which proposes a classification of coercion into different types. We present the results of an ERP study on one particular kind of coercion (adding an eventuality = "additive coercion") showing that aspectual coercion is not necessarily triggered by an aspectual mismatch. We will then present a self-paced reading experiment on contextual influences in additive coercion which suggests that aspectual processing immediately takes into account the larger discourse context. Taken together, the first two experiments provide clear evidence against a semantic operator-based account but indicate that aspectual coercion heavily relies on pragmatic inference. Two further self-paced reading experiments compared different subtypes of aspectual coercion with each other (deletion of an eventuality = "subtractive coercion" and * I would like to thank the audiences of the TRAIT workshop How categorical are categories? in Wroclaw 2013 and an invited talk at the German department in Cologne 2013 for their comments on the work presented in this paper. I would also like to thank Johannes Dölling, Fritz Hamm, Janina Radó, Edith Scheifele, Matthias Schlesewsky and two anonymous reviewers for valuable comments on the research reported in this paper. The usual disclaimers apply, any remaining errors are mine. The research was funded by the German Science Foundation within projects C1 and B1 in the Collaborative Research Centers 441 and 833 at the University of Tübingen. two types of iteration of an eventuality = "abstract type shift"). The results indicate that different kinds of coercion do in fact lead to differences in how they are processed. Most strikingly, different coercion subtypes which should involve the same operator-an iterative operator-revealed different coercion costs. Taken together, the experiments demonstrate that a operator-based coercion theory or underspecification account cannot fully account for the data. The predictions of Event Calculus, however, were fully confirmed by the experiments. Event Calculus thus allows us to derive empirically valid predictions about the underlying cognitive processes while incrementally constructing a temporal model of the unfolding discourse.
Uploads
Papers by Oliver Bott