Title: Human Agent Architecture Has a Valuationist Structure: Implications for the Normativity of Attention
Abstract: The belief/desire model remains the dominant approach in contemporary philosophy for explaining action and investigating agency. But the model, rooted in folk psychology, is superficial and incomplete, providing a rickety framework for conducting rigorous philosophical inquiry. Meanwhile, in computational cognitive science and artificial intelligence, an alternative valuationist picture of agent architecture is convergently coming into view. At the heart of valuationism is a simple and sweeping claim: every time an agent acts, they do so on the basis of value representations. These are, very roughly, scalar, non-conceptual representations of the goodness or badness of one’s response options. In this talk, I give a philosophically rich account of this emerging valuationist picture of agency. I also consider the theoretical benefits for philosophy from giving up the belief/desire model. I focus on how valuationism supports an intriguing “accuracy-based” vantage point on the normative criticism of agency. This has major implications for understanding the normativity of attention, including when attention is inappropriate and when an agent is morally responsible for attentional failures.
If you would like to attend this event virtually by Zoom, please contact Drew Johnson at Department of Philosophy, Classics, History of Art and Ideas.