Do eye movements go with fictive motion

Do eye movements go with fictive motion? Teenie Matlock (tmatlock@psych.stanford.edu) Department of Psychology, Stanford University Stanford, CA 94305 USA Daniel C. Richardson (richardson@psych.stanford.edu) Department of Psychology, Stanford University Stanford, CA 94305 USA Language is also full of sentences such as (2a) and (2b). Abstract Cognitive scientists interested in the link between language and visual experience have shown that linguistic input influences eye movements. Research in this area, however, tends to focus on literal language alone. In the current study, we investigate whether figurative language influences eye movements. In our experiment, participants viewed two-dimensional depictions of static spatial scenes while they heard either fictive motion sentences, such as The palm trees run along the highway, or non-fictive motion sentences, such as The palm trees are next to the highway. Overall, sentence type influenced participants’ eye movements. Specifically, gaze duration on the figure (e.g., palm trees) was longer with fictive motion sentences than with non- fictive motion sentences. Our results demonstrate that figurative language influences visual experience. They provide further evidence that fictive motion processing includes mentally simulated motion. (2a) The road goes through the desert (2b) The fence follows the coastline These sentences are figurative because they contain a motion verb (e.g., goes, follows) but express no actual motion (Matlock, 2001). 1 They contrast with literal sentences with motion verbs, such as The bus goes through the desert, or The herd of sheep follows the coastline, which feature mobile agents that move from one point in space and time to another (Talmy, 1975; Miller & Johnson-Laird, Despite the absence of actual movement with sentences such as (2a) and (2b), they have been claimed to involve fictive motion, an implicit mental simulation of “movement” through a construed scene (Talmy, 1983, 1996, 2000). On this view, the conceptualizer subjectively “scans” from one part of the scene to another, most notably, along the figure (i.e., prominent entity, subject noun phrase referent). For (2a), this means “moving” along the road, and for (2b), it means “moving” along the fence. According to the argument, fictive motion is a way to impose motion on what is otherwise a static scene. It enables the language user to compute information about the layout of a scene, for instance, a road in a desert in (2a), or a fence aligned with a coastline in (2b). Importantly, Talmy (1996) and other cognitive linguists do not maintain that fictive motion involves vivid imagery whereby the conceptualizer “sees” himself or herself (or any other animate entity) moving point by point along the figure in the scene being described. Instead, they take the motion to be relatively fleeting and tacit. (See also Langacker’s abstract motion, 1986, 2000, and Matsumoto’s subjective motion, 1996). 2 At first, the claim that people simulate motion while processing descriptions of static scenes seems bizarre. Why would motion be processed, for instance, with sentences such as (2a) and (2b) when neither the road nor the fence is Introduction Imagine that you and a friend are sitting in a courtyard chatting. During the course of the conversation, you occasionally glance over at a long, thin stationary object on the ground. You assume the object is a tree branch or a walking stick until your friend says, “Oh! Look what slithered onto the courtyard.” At that point, your perceptions and conceptions of the object dramatically change. The object goes from a piece of wood to a snake. Situations like these––in which language influences the interpretation of objects and actions––are ubiquitous. The question addressed here is whether this influence is limited to literal language, or whether it also includes figurative language. We are especially interested in whether sentences such as The road goes through the desert or The fence follows the coastline (figurative because they include a motion verb but express no motion) affect eye movements. Our results suggest they do. What We Know about Fictive Motion Everyday language is replete with sentences such as (1a) and (1b). These are literal descriptions of static scenes. Like Rumelhart (1979) and Gibbs (1994), we do not maintain a hard and fast distinction between “literal” and “figurative”. We simply use these terms here to operationalize two types of motion verb constructions: those that express motion and those that do not. Our study looks at just one type of fictive motion, Talmy’s (2000) co-extension path fictive motion. There are many others. (1a) The road is in the desert (1b) The fence is on the coastline