Encoding of text during reading differs fundamentally from the encoding of speech in that reading requires the active selection for processing of to-be-recognized words from a spatially ordered set of visual symbols. Two forms of visual selection are distinguished in the literature. Overt selection, which is accomplished by directing the eyes to a selected word of text and by fixating it for a brief interval, and covert selection, which controls the processing of words during a fixation. Models of eye movement control during reading seek to explain how these two types of selection are accomplished and coordinated. To advance theoretical developments in the field, Dr. Inhoff examines the selection-for-fixation and the selection-for-processing of words that are spatially adjacent to a fixated word and of previously read nonadjacent words. Two recently developed models of oculomotor control, that are predicated on substantially different selection assumptions, are used to formulate discriminating predictions for the selection of spatially adjacent (parafoveally visible) words; one of the models also provides working hypotheses for the selection of previously read words. Extensions of an established eye-movement-contingent display change paradigm are used to manipulate the temporal and spatial availability of linguistic information to the right and left of a selected target word upon its fixation and to control the allocation of attention during a fixation. A novel paradigm, involving the eyemovement-contingent presentation of spoken words, is used to elicit long-range eye movements to previously read words. Effects of these manipulations on the temporal and spatial pattern of ensuing eye movements are then used to delineate the nature and coordination of the two types of selection. Results of the proposed work should continue to advance extant theoretical conceptions of eye movement control during skilled reading, and it may advance our understanding of some reading disabilities that are related to poor overt and covert selection. It should also have ramifications beyond the domain of reading, as coordination of covert and overt selection may be part and parcel of effective visual selection in general.
Showing the most recent 10 out of 16 publications
