Composition of multiple dimension representations in context during real-time comprehension

We investigate the role of context in the comprehension of competing semantic representations of sentences with aspectual verbs (AspVs). On the Structured Individual Hypothesis, AspVs select for structured individuals as their complement, construed as a directed axis along various dimensions. During comprehension, the verb’s lexical functions are exhaustively retrieved and the AspV+complement composition yields multiple mutually exclusive dimension representations, which are later constrained by context. Results from this eye-movement study show that AspV sentences engender additional processing cost independent of context. That is, while processing multiple dimension representations is costly, the exhaustive lexical retrieval and dimension composition are initially encapsulated from context.