Abstract

The purpose of this research was to examine how linguistic and visual processes interact, and to determine the locus of this interaction. Three studies were conducted to this aim. The first, a normative study, asked participants to list the objects they saw after viewing 17 naturalistic scenes for a short time in order to establish which objects were most salient. The second and third studies used an eye-tracking methodology to record participants' eye movements as they looked at static and dynamic scenes (the same as those in the Normative Study) and listened to related sentences. The sentences differed with respect to the class of their main verbs (causative vs. perception/psychological verbs), and the scenes differed according to the positioning and motion (or apparent motion in the static scenes) of the human agent (towards, away, or neutral) relative to the object referent of the verb's direct object (the target object). Results indicate that the direction of apparent motion of the agents in the scenes did affect how frequently the target object was listed by participants in the Normative Study. For the two experiments, the only consistent finding was that motion context had a significant main effect on how quickly eye movements were launched towards the target object after the verb was uttered. The effect of verb type was found to be less consistently significant, depending on the statistical procedure employed, which fails to support the notion that verb-specific information can guide eye movements towards the target object