Abstract
When a new object onsets abruptly, it captures attention. But is this capture due to the onset transients, or to the onset object's status as a new entity? Franconeri, Hollingworth, & Simons (2005) examined this question using a procedure that introduced a new object without simultaneously introducing onset transients. Observers had to search for one of two target letters. At the start, an opaque annulus surrounded several static placeholder objects. The annulus shrank progressively, obscuring the placeholders at some point, whence the placeholders were turned into letters. On some trials, an additional letter was introduced here. As this new letter appeared whilst obscured by the annulus, there were no accompanying onset transients. The annulus then shrank further and the letters, including the new object, were revealed. At issue was whether attention engaged this new object first. Search set size was varied. The additional object may be a target, or a distractor, letter. If it captured attention, and it was the target, search times should be independent of set size. Franconeri et al.'s results showed no prioritization of the additional object, suggesting that a new object failed to capture attention. Yet crucially, for the additional object to be coded as new, it must be discriminable from the old objects which, in this task, meant that observers had to encode implicitly the old objects' locations. We showed in 5 experiments that, when encoding of locations was facilitated, the additional object succeeded in capturing attention. But when location encoding was disrupted, the new object failed to capture attention.
supported by NUS Grant R-581-000-042-112