McCollough Effect

All of us have experienced visual adaptation in some form or another. Stare at a bright red light for half [Page 479]a minute, look away and you will see a green spot on your retina before it fades away quickly. You just adapted to the (red) light and, as a result, saw a (green) color that is opposite or complementary to the original. In general, persistent exposure to a stimulus causes the neural circuitry responsive to the stimulus to adapt. The adaptation results in afterimages or aftereffects, namely, percepts that are dissimilar from the actual stimulus. Aftereffects are generally negative insofar as the feature value perceived following the adaptation is opposite that of the adapter (the stimulus that causes the adaptation), as in the preceding example.

Celeste McCollough's discovery of an aftereffect that was contingent on a feature of the adapting pattern radically altered the study of adaptation. The aftereffect, known as the McCollough effect, has since been extensively replicated in other laboratories. It arises from the alternating presentation of stripes (gratings) of perpendicular (orthogonal) orientations of complementary colors for a modest time period (5–10 minutes). Figure 1 illustrates the adapting stimuli, which consist of red, horizontal and green, vertical gratings. Subsequent to the adaptation, achromatic (i.e., black and white) horizontal gratings appear greenish and achromatic vertical gratings pinkish, opposite to the colors shown during adaptation. Thus, similar to the overwhelming majority of aftereffects, the McCollough effect is a negative aftereffect.

Figure 1 McCollough effect

The McCollough effect is a dramatic departure from other negative aftereffects, however. In contrast to simple aftereffects that do not require a test stimulus for the misperception to be evident, the McCollough effect is a contingent aftereffect that is created by relatively brief experimentally induced correlations between stimuli that are usually uncorrelated in the real world. The aftereffect is different depending on whether the test stimulus is horizontal or vertical in the previous example. Also, in contrast to simple aftereffects that last a few seconds, the McCollough effect is stable for weeks, or even months.

There are at least two other fascinating aspects of the McCollough effect. The observer does not have to maintain fixation on a point on the adapter; in fact, the observer can let the eyes wander over the pattern, and the resulting aftereffect will be largely unaffected. Second, the effect does not transfer interocularly: If adaptation is limited to one eye, no discernible negative aftereffect is observed in the non-adapted eye. Given that the McCollough effect is contingent on adapter orientation, that orientation selectivity does not emerge before primary visual cortex (or V1), and that orientation-tuned cells in V1 typically have binocular responses, one would expect the McCollough effect to transfer to the non-adapted eye. That it does not is surprising, and suggests a subcortical locus, where information from the two eyes is segregated.

Stemming from the McCollough effect, research has expanded into investigations into mechanisms or theories to explain the McCollough effect, explorations of aftereffects on other visual dimensions, and discoveries of positive contingent aftereffects.