CENTER FOR RESEARCH IN
COGNITION & NEUROSCIENCES/ConnexConnex

Selective attention is often conceptualized as a mental spotlight that can be directed at will to illuminate any item of interest, acting as a sensory filter that selectively gates goal-relevant information for higher-order processing and conscious representation. Yet, over the past decade or so, it has become clear that our brains continuously predict what sensory signals are likely informative for goal-directed behavior based on the statistics of past agent-environment interactions, and hence, that action-oriented probabilistic learning is a much more pervasive feature of selective attention than generally assumed. In this talk, I will first present findings from several behavioral and EEG studies that reveal how, at the neural level, such probabilistic learning rapidly structures attention: what we subsequently automatically attend to or ignore. I will then discuss more recent work suggesting that these attentional biases may be sensorimotor in nature, giving attention a habit-like character. Finally, I will discuss the idea that attention training as cultivated by meditation may provide a method to unlearn to attend or mentally behave in habitual ways. Altogether the work presented in this talk suggests a reconsideration of how attention is typically approached: from a mere sensory filter to a relational process that continuously optimizes agent-environment interactions