Abstract
Most theories assume that attention imposes the primary constraint on the number of objects one can track, either through a limited number of foci, or a limited amount of attention. Yet tracking must also rely on updating one's knowledge of object locations, knowledge that is probably never perfect, even for one object. Could inherent imprecision in spatial knowledge be the only factor limiting tracking? We measured the imprecision in representations of location in displays with featurally identical targets and non-targets. After a tracking period, all items disappeared. In half of trials, all reappeared again, and targets were identified, affording a measure of tracking performance. In the other half of trials, all but a single target reappeared. Participants clicked in the location of the absent target, affording a measure of the imprecision in their spatial knowledge. (A separate experiment found no tracking deficit for items that disappeared momentarily compared to ones that did not). We tested loads of as many as eight targets among variable numbers of distractors, and we used a separate experiment with 16 targets to identify (and exclude) responses that likely came from a distribution of mostly guesses. Collectively, five experiments showed that imprecision was invariant with respect to the number of targets successfully tracked, their sizes, and even whether they moved or not. Moreover, imprecision was not affected by increasing numbers of non-targets, though tracking performance was. And imprecision did not increase for more crowded displays, though tracking performance suffered considerably. These results suggest that inherent imprecision in spatial knowledge imposes the primary constraint on tracking. High target loads, dense and crowded displays do not consume more attention or exhaust available foci. They simply increase the probability that inherently imprecise location knowledge will lead to tracking errors.