Bartosova, J. , Chapman, C. , Kucerova, I. & Service, E.
McMaster University
English events can be described as telic, or having an endpoint, or as atelic. Stockall & Husband (2014) found greater processing difficulty (longer reading times) in self-paced reading for atelic compared to telic events. However, these events are also sensitive to the type of temporal adverbial with which they combine (Giorgi & Pianesi, 2001), e.g., Peter ate the cookie in/*for ten minutes (telic) versus Peter ran *in/for ten minutes (atelic). Effects of the adverbial on processing have not been previously investigated. Since some English predicates are compatible with both adverbials and interpretations (e.g., to iron the clothes), they can be used to test processing differences between telic and atelic events. Our results from a complex span task suggest that there is a processing cost after in-X-time adverbials (telic) but no processing cost after for-X-time (atelic) when the adverbial appears before the VP. Instead, atelic events show a processing cost after the whole sentence. Our results support the immediate partial interpretation hypothesis (Frazier & Rayner, 1990): while the parser commits to a telic interpretation as soon as it encounters in-X-time, it postpones making a semantic commitment until the end of the sentence for atelic events.