Abstract
Much recent research has suggested that the units of attention are often discrete whole objects, and has further revealed some of the factors that mediate what can count as an ‘object’. Here we address a new question: How fast do these object representations form? This question is difficult to address with standard cueing or divided-attention paradigms, since these tasks themselves require strict timing parameters. Here we employed the ‘multiple object tracking’ task, in which subjects had to track a set of unpredictably and independently moving targets in a field of featurally-identical moving distractors: e.g., a subject might have to track 4 out of 8 moving line segments. ‘Target merging’ manipulations were then employed to frustrate this tracking: each target was merged with a distractor via various types of grouping cues (‘connectors’); e.g., two line segments might simply be joined into a single long line, of which subjects must continue to track one end. We have found in earlier studies that such manipulations cause the target-distractor pair to be treated as a single object (i.e. to ‘merge’), which prevents tracking of only the targets. In the current study these connectors appeared for only a short time during the tracking interval: they onset after the motion had begun and then disappeared before the motion ended. The critical variable was the duration for which the connector was visible. Observers were able to successfully track the targets through connector durations of up to about 300 ms; at longer durations, tracking was drastically impaired. The critical duration beyond which tracking suffered varied with the particular grouping cue used, and also with the method by which the connector appeared (suddenly vs. gradual fading-in vs. growing from each end). Such results reveal the speed with which the objects of object-based attention form in various conditions, at least in a situation with a high attentional load. (BJS was supported by NIMH 1-R03-MH63808-01. JF was supported by NSF SBR-9875175.)