Abstract
Can one illusion trigger another illusion?
It has been reported that a stationary patch of random dot motion is perceptually displaced along the direction of motion (Ramachandran et al. 1990). This illusion has drawn new interest since a MRI study reported that cortical representation of the motion patch is shifted in the opposite direction of motion, indicating that neural activity of the primary visual cortex can be dissociated from perceived location (Whitney et al. 2003). So the second question is, from what level on along the visual pathway, does neural activity match the illusory percept?
The motion aftereffect is a nice tool to hack into the level of visual processing due to the existence of static and dynamic test, which is sensitive to retinotopic low-level MAE and non-retinotopic higher-order MAE, respectively. We observed MAEs at the boarders of the illusory shifted patch of motion to investigate the level of percept - neural activity accordance.
The adaptation stimulus was a patch of random dot motion (3 deg × 3 deg) placed at the center of a larger patch (9 deg × 9 deg). The two patches had opposite direction of motion and the inner patch appeared to be displaced. In blocks measuring the static test, we presented a stationary spot (0.5 deg, 500ms) around the physical boarders of the two patches after adaptation. For the dynamic test, we presented three discrete phases (250 ms each) of Gabor patch (0.5 deg, 3 cpd).
The experimental results indicated that static MAE follows the physical border of the patches, say, the direction of motion aftereffect switches at the physical boundary. In contrast, the changeover point of dynamic MAE is shifted to the direction of inner patch motion, indicating correspondence of percept and higher-order activity. To answer the two questions, yes, an illusion can trigger another illusion, as long as it involves the higher areas where neural activity is likely to be in accordance with the illusory percept.