What are the rules that govern the redeployment of attention to targets presented sequentially across space? In providing an initial answer to this question, I will draw from the conceptual and practical work that has been done in the fields of inhibition of return (IOR; the finding that response times are slower to a target presented at a previously cued location) and the attentional blink. (AB; the finding that identification accuracy for the second of two targets is impaired when it is presented shortly after the first).

In the predominant view, IOR has been attributed to some form of inhibitory mechanism. On this view, previously cued locations are marked with inhibitory tags which must be overcome before redeploying attention to that location. I will present evidence consistent with the view that IOR is not a unitary phenomenon based on inhibition. Rather, the magnitude of IOR is influenced by expectations that the observer has developed from everyday interactions with the physical world. These expectations include directional momentum and directional reading biases.

Attentional redeployment also governs the AB deficit which, for this reason, is thought to be a phenomenon of high-level vision. Recent evidence has questioned this view by showing that the AB depends critically on masking events in low-level vision. This may lead one to conclude that the AB is a phenomenon of low-level vision. In my talk, I plan to show how this conclusion would be unwarranted, and why the AB must continue to be regarded as a phenomenon that is modulated by attentional processes in high-level vision.

Returning to the topic of this meeting, I will conclude that attentional processes as indexed by IOR are modulated not only by inhibition but also by expectations. In the same vein, I will conclude that the attentional processes that mediate the AB are not limited to low-level vision.