To account for the integration of extrafoveal and foveal views across eye movements, two representational systems have been proposed. The first is a permanent lexicon of object detectors, which fire whenever evidence for particular objects is apprehended, regardless of their location. The second is a set of temporary, episodic object files, which each are constructed for a specific object occupying a particular location in space. In the experiments from which these two representational systems were inferred, observers typically were required to make a saccade to a location in between two extrafoveal target objects. The target objects were or were not shifted during the saccade and subsequent ease of target identification was measured. In the present study, participants were required to make a saccade to a specific target object as opposed to a location in between two objects. This situation, in which a single object to which the eyes are to be sent is selected, is not only less artificial but also creates different attentional demands. As such, the goal of this study is to test the theoretical propositions of the dual representational system under more natural circumstances for saccadic and attentional processes.