Abstract
Abrupt onsets of objects involuntarily draw attention (Jonides & Yantis, 1988) and the eyes (Theeuwes et al, 1998) even when they are irrelevant to the task at hand. Luminance changes or other low-level transients do not fully explain why onsets capture attention. Instead, the new-object hypothesis proposes that it is the appearance of new structural or semantic information in the display that captures attention rather than the perceptual transient (Yantis & Hillstrom, 1994). In the present task, we investigated whether structural and semantic changes are sufficient to capture attention and the eyes. Observers moved their eyes to an object defined as the target by the shape of a frame around it, and concurrently another object in the search display smoothly and quickly morphed, thereby changing its structural and semantic information. Morphing an object also had the added advantage of minimizing the number of confounding low-level transients. If a strong version of the new-object hypothesis is correct, this “new” object should capture the eyes. A control condition used abrupt onsets rather than morphs in the periphery to compare capture rates. The morphed objects did not capture the eyes; however, the abrupt onsets did. These findings indicate that new structural or semantic information is not sufficient to capture the eyes. Instead, capture appears to require the onset of a new object to be accompanied by a sudden transient (Boot, Kramer, & Peterson, 2005).