Seeing what isn't there; The costs of perceptual learning

Perceptual learning is defined as plasticity in one's sensory brain-areas and helps us to better perceive the visual environment. A natural question to ask when evaluating studies of perceptual learning is what is the cost of perceptual learning? If there is no cost, why wouldn't sensory processing already be at its maximal level of performance? To date studies of perceptual learning have concentrated only on its benefits, and the possible costs of such learning are ignored. Here we show that perceptual learning can also lead to misperceptions, such that subjects actually perceive stimuli when none are physically presented. We demonstrate this counterintuitive result by using a reinforcement procedure in which motion stimuli, which were too dim for subjects to detect, were temporally paired with the targets of a letter task. After learning, subjects not only showed enhanced sensitivity to the motion direction of the trained stimulus but often reported seeing dots moving in the trained direction when no stimulus was displayed. We further show that these misperceptions result from a perceptual bias and are not attributable to a response bias. These results show that there are costs as well as benefits to perceptual learning and that sensitivity enhancements for a specific feature can also be accompanied by misperceptions of the visual environment.