Adapting to Motion With and Without Attention

The motion aftereffect, sometimes referred to as the "waterfall illusion," is one of the oldest visual illusions ever discovered, dating as far back as Aristotle. When we are exposed to a continuously moving stimulus, such as a waterfall, it causes selective adaptation of motion-sensitive neurons that code for that direction. For example, observing a waterfall for 30 seconds causes our "downward motion" neurons to adapt and reduce their firing rate. Upon looking at a static (non-moving) scene, it appears to be slowly drifting in the opposite direction (in this case, upward). You can experience the motion aftereffect by watching this short video:

The Ambiguous Role of Attention

An important question for cognitive scientists is whether and how attention plays a role in motion aftereffects. Since motion perception starts at the earliest stages of neural processing, it is possible that motion aftereffects don't require active, focused attention: as long as the motion stimulus is hitting our retinas, it should trigger the relevant motion detectors in the early visual cortex to adapt, and lead to motion aftereffects. However, scientific research into this question has produced mixed results, with some studies showing little to no effect of attention and others showing considerable involvement of attention.

To resolve this discrepancy, a new study by Daphne Roumani and Konstantinos Moutoussis published in this month's issue of Perception, adapted observers to a peripheral motion stimulus while they completed a task of varying difficulty in the center of the screen. The task (Rapid Serial Visual Presentation, or RSVP) required participants to identify target letters that flashed briefly in the center of the screen.

The difficulty of the task was manipulated by changing the frequency at which the targets appeared (the more frequent the targets, the more difficult and attention-demanding the task). After several seconds of adapting to peripheral motion while completing the central RSVP task (or no task), participants judged the direction of an ambiguous set of dots moving in the periphery (the test stimulus).

By examining participants' bias when responding to the test stimulus, the researchers were able to measure the strength of the motion aftereffect across different attentional loads. The researchers also manipulated the overall strength of the adapting motion stimuli to determine whether attentional effects depend on the strength of the adapting motion itself.

Their findings showed a moderate effect of attention: when the central RSVP was more attention-demanding, participants exhibited a weaker motion aftereffect. However, this effect was only found for the strong motion adaptation conditions. When the motion stimuli in the periphery were weak (i.e., less coherent), the motion aftereffects were weaker and did not vary based on attentional load.

Overall, the authors conclude that attention does have an effect on motion aftereffects, but these depend critically on the strength of the motion stimuli. Moreover, the modulatory effects of attention are limited; that is, in the high attentional-load condition, motion aftereffects were only about 10.5 percent weaker than in the low attentional-load condition, and only about 23.6 percent weaker than a baseline condition with no RSVP task at all. Therefore, although attention can modulate the strength motion aftereffects, motion aftereffects can still manifest robustly in the absence of attention.