Preview benefit in speaking occurs regardless of preview timing

摘要

Speakers access information from objects they will name but have not looked at yet, indexed by preview benefit—faster processing of the target when a preview object previously occupying its location was related compared to unrelated. This suggests that speakers distribute attention over multiple objects, but does not reveal the time course of processing of a current and to-be-named object. Is preview benefit a consequence of attention shifting to the next-to-be-named object shortly before the eyes move to that location, or does preview benefit reflect a more unconstrained deployment of attention to upcoming objects? Using the multiple object-naming paradigm with a gaze-contingent display change manipulation, we addressed this issue by manipulating the latency of the onset of the preview (SOA) and whether the preview represented the same concept as (but a different visual token of) the target or an unrelated concept. Results revealed that preview benefit was robust, regardless of the latency of the preview onset (SOA) or the latency of the saccade to the target (the lag between preview offset and fixation on the target). Together, these data suggest that preview benefit is not restricted to the time during an attention shift preceding an eye movement, and that speakers are able to take advantage of information from non-foveal objects whenever they are visually available.