Gernot Horstmann¹, Stefanie Becker²; ¹Bielefeld University, ²The University of Queensland

An important question in attention research is to what extent attention is biased by bottom-up factors such as saliency versus top-down factors such as intentions and expectancies. A particularly relevant test is the surprise presentation of a critical stimulus. A surprise presentation means that the presentation is not expected in advance, and it is thus reasonable to assume that no intentions are directed to the critical stimulus. Surprise presentations may entail several components, such as not informing the observer of the critical stimulus, or rendering the critical stimulus novel in the sense that its features have not been presented before, and therefore have no top-down utility or history. Previous research found that announcing the critical stimulus, rendering it expected in some sense, has almost the same strong attentional prioritization effects as when the critical stimulus is not announced and thus unexpected. Here we measured eye movement parameters as a means to disentangle the predicted pre-attentive and post-attentive prioritization of the surprise stimulus, that is, the capture and binding of attention, respectively. We compared two groups with the same novel color singleton, where the change in the critical trial was announced in one group but not in the other. The novel singleton was strongly prioritized in the critical trial. The announcement of the change had little effect on pre-attentive prioritization, but shortened the post-attentive prioritization. Overall, the results suggest that implicit expectations generated during on-task experience and the incidental learning of statistical regularities are more important for the attentional prioritization of a surprising stimulus than explicit expectations generated by verbal instructions.