Learning what matters How to pay attention in a volatile world

This thesis explores how the human brain learns to focus attention in a constantly changing environment. It investigates the mechanisms of selective attention—how the brain prioritizes relevant sensory information while ignoring irrelevant or distracting stimuli. The research is divided into two parts: external attention, which deals with visual input from the environment, and internal attention, concerning information held in working memory.
The first part focuses on external attention, examining how the brain uses statistical learning to adapt to task demands in volatile environments. It finds that contextual stability significantly enhances the ability to ignore distractions, showing that reduced uncertainty in search tasks improves attentional control. These findings suggest that learning in less variable contexts generalizes to new but similar environments.
The second part investigates internal attention and its relationship to statistical learning within working memory. Unlike external attention, the study finds no evidence that statistical learning biases the prioritization of information held in memory. This suggests a fundamental difference between external and internal attentional mechanisms, potentially tied to their distinct functions—external attention optimizes sensory processing, while internal attention adapts stored representations for future use.
Overall, the thesis advances understanding of attentional systems by highlighting the critical role of environmental volatility in shaping attention. It also challenges assumptions about the similarity between external and internal attention, emphasizing the unique constraints and purposes of each system.