Allocation of Visuo-Spatial Attention and Saccade Programming During Reading

Abstract Current conceptions of visuo-spatial and lexical processing during reading (e.g., Reichle, Pollatsek, Fisher and Rayner, 1998) assume that attention confines recognition processes to a single word at a time and that interword saccades are triggered by the completion of an initial stage of lexical processing while attention is exclusively allocated to the fixated word. Three experiments examined these assumptions by determining whether readers obtained useful lexical information to the right and left of fixation prior to the initiation of interword saccades. The results of Experiment 1 reveal that lexical information from a parafoveal word to the right of fixation can be obtained before a saccade to this word is committed to action. The results of Experiments 2 and 3 show that readers obtain useful lexical information from a target word to the left of fixation, irrespective of whether the target is subsequently reread. These findings indicate that more than one word is attended at a time during a fixation, that successive areas of attention overlap, and that programming of a saccade to a parafoveally visible word is not completed before attention is allocated to it.