Task-irrelevant perceptual learning without reinforcement

• Applicant: Dr. Patrick Bruns
• Subject Area: General, Cognitive and Mathematical Psychology
• Term: from 2015 to 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 284165930

Task-irrelevant perceptual learning refers to improvements in sensory abilities for a feature that is irrelevant to a given task. Theories of perceptual learning generally assume that such passive learning is mediated by concurrent internal or external rewards. According to them, diffusive neuromodulatory signals from the reward system interact with bottom-up processing of the task-irrelevant feature, thereby helping to exceed an assumed learning threshold. However, despite a huge amount of research on perceptual learning, it is currently unclear whether and how task-irrelevant learning occurs without the involvement of reinforcement signals. The global goal of this proposal is to better understand the mechanisms of task-irrelevant perceptual learning. Three related hypotheses about the necessity of reinforcement for task-irrelevant learning will be tested using psychophysical learning experiments: First, from classical conditioning, it is well-known that learning specificity for a rewarded stimulus feature strongly depends on whether or not similar stimuli are exposed without reward. It will be tested whether this principle is valid for task-irrelevant perceptual learning, which would suggest that the specificity of task-irrelevant learning does not directly involve reinforcement signals. Second, it is well-known that emotional salience and reward history of a stimulus significantly enhance perceptual processing and memory, while the effects of these factors on task-irrelevant learning remain unclear. Considering the tight relationship between the emotional system and the reward system in signalling behaviourally relevant events, emotional salience and reward history of a stimulus are likely candidates for mechanisms that should drive task-irrelevant perceptual learning without involvement of a reinforcement signal. The role of these factors will be tested in two experiments. Third, it will be tested whether stimulus salience and behavioural relevance are necessary for task-irrelevant learning to occur. Because statistically structured sources of information reflexively receive prioritized processing even without conscious awareness of the regularity, the paradigm of statistical learning will be used to test this prediction. Successful results of these experiments would better clarify the effects of attention, reward and emotion and thereby substantially advance our knowledge about the mechanisms of task-irrelevant perceptual learning.

• DFG Programme: Research Fellowships
• International Connection: USA
• Host: Professor Dr. Takeo Watanabe