Abstract
Studies of transsaccadic object perception in which participants have to detect saccade-contingent changes in objects, have revealed that certain object characteristics are not maintained accurately across a saccade, while other types of extrafoveal object information are routinely integrated with the subsequent foveal view of the object. This suggests that transsaccadic object perception is selective. The purpose of the present paper is to examine what determines which object information is selected to be incorporated in the transsaccadic object representation. Our general hypothesis is that the degree to which the information is diagnostic for the object’s identity is an important determinant. We make a distinction between an object’s identity in sensu stricto (i.e., an object’s conceptual label that is activated in a stored object lexicon) and an object’s identity in sensu latiore (i.e., the role of the object in an evolving situation model). An overview is given of research recently carried out at our laboratory that supports this hypothesis.