Mental curve tracing with elementary stimuli.

It has been proposed that certain spatial relations are determined by an operation, or "visual routine," that can trace along a boundary (Ullman, 1984). This proposal was supported by Jolicoeur, Ullman, and Mackay's (1986) finding that the time required to determine if two Xs are on the same curve increased monotonically with the separation of the Xs along that curve. In the present study the generality of the curve tracing hypothesis was explored across four experiments by using elementary stimuli that eliminated interweaving curves, displaced the fixation point away from the curves and target Xs, and provided a simple alternative to curve tracing--namely, determining whether or not the Xs fell on the same side of the figure. Stimuli consisted of two curves (150 degrees arcs) and two Xs, and each stimulus was presented for 150 ms. In Experiments 1 and 2, subjects were instructed to decide as quickly as possible if the two Xs fell on the same curve or on different curves. Even for these elementary stimuli, mean reaction time (RT) for same trials increased monotonically with the distance separating the Xs. Mean RT for different trials, however, decreased with the distance separating the Xs. In Experiments 3 and 4 alternatives to curve tracing were tested. For same trials the evidence strongly favored curve tracing. However, different trials were apparently solved on the basis of judgmental processes presumably operating in parallel with curve tracing. Curve tracing rates fluctuated across experiments and seemed to be partially governed by the width of the "pathway" provided for the trace.