[PS-3.28] EVENT STRUCTURE AND PASSIVE FREQUENCY IN GARDEN PATH PROCESSING

King, R. & Townsend, D.

Montclair State University

We measured eye fixation times while reading sentences such as The prisoner halted / taught / abducted / taunted by the agent tried to escape (Malaia et al., 2009, 2012). The event structure hypothesis maintains that because the subject of optionally transitive telic verbs (halted) is an underlying object, these verbs produce a smaller garden path (O?Bryan et al., 2005). The passive frequency hypothesis maintains that verbs that appear more often as passives produce a smaller garden path (Hare et al., 2007).

For optionally transitive verbs (halted / taught), the garden path effect for first pass time on the by-phrase was smaller for telic verbs, and for total time on the main verb it was smaller for atelic verbs, suggesting revision of an intransitive interpretation (O?Bryan, 2003), all ps < .05. Both effects persisted when we used passive frequency as a covariate of verb type, both ps < .05. For obligatorily transitive verbs (abducted / taunted), event structure had no effect on the size of the garden path, all ps > .05.

These results support the event structure hypothesis with optionally transitive verbs and demonstrate that it is independent of passive frequency.