Abstract
In the past my colleagues and I have presented evidence for the existence of a primitive memory for attention deployments. This memory system orients attention to features and locations of recent behavioral importance. A large part of the proposal is that the memory system facilitates attentional selection of features in the environment such as color, orientation, etc. I will review the findings behind these proposals. The bulk of this evidence comes from studies of history effects in vision, such as priming of visual search, and implicit learning in exogenously cued attention shifts. In addition, recent challenges to this view will be discussed (in particular, arguments that the facilitation from repetition in visual search tasks occurs at a relatively late stage of perceptual processing). Our recent investigations of these proposals suggest that there is indeed evidence for higher-level priming effects in visual search under certain conditions, but those findings do not challenge the basic proposals regarding the primitive memory system, since the evidence for feature-specific facilitation is robust.