Abstract
We examined grouping under inattention using the method developed by Driver et al. (2001). On each trial, observers were presented with two successive displays that appeared briefly. Each display consisted of a central square target (a small square made up of random black and white pixels) surrounded by background elements. The observer's task was to judge as rapidly and as accurately as possible whether the two successive target squares were the same or different (only a single pixel changed). The organization of the background elements stayed the same or changed across successive displays, independently of whether the successive target squares were same or different. The colors of the elements always changed between the two successive displays to disentangle a change in color per se from a change in organization. The different background organizations were: Grouping elements into a vertical/horizontal pattern by color similarity, grouping elements into a shape (triangle/arrow or square/cross) by color similarity, grouping elements (or line segments) into a shape with no segregation from other elements, and a connected shape that served as a control. We measured the influence of the background organization on the speed and accuracy of the same/different judgments. After the last trial in each background condition subjects were asked surprise questions about the immediately preceding background. The RT and error pattern in the same/different judgments implies that background elements were grouped into a vertical or horizontal pattern by color similarity, and into a shape when no segregation from other elements was involved, suggesting that quite an elaborated form of grouping can take place without attention. However, grouping into a shape that requires resolving figure-ground relations for segregated units appears to require attention. Regardless of the attentional demands of grouping, the products of grouping are not available to awareness without attention.