Abstract

A growing empirical literature on the cognitive effects of moods and affects points out to the fact that we seem to be more (epistemically) reliably, more (epistemically) cautious and more reflective and analytical when we cognitively perform under the grip of a negative mood/affect in comparison to a neutral or positive one. Concretely, I will argue, we form epistemically safer beliefs under the grip of negative affect.  I call this body of empirical data, the Empirical Datum. I want to explain the Empirical Datum, proposing that valence has epistemic significance on its own. This thesis has not been explored in the literature yet, as far as I know, since most philosophers tend to treat emotions, positive and negative, epistemically alike. More concretely, I want to defend the idea that valence allows us to track the epistemic standards, which, as contextualism teaches us, are partially constituted by practical stakes, at play in a given epistemic context. Thus, negative valence informs us that the stakes are higher than those represented by positive valence or neutral states. This way we can explain our explanandum: given negative valence’s representational content about high epistemic standards, it makes sense that negative valence’s functional role is associated with sophisticated cognitive programs that aim at satisficing those very high epistemic standards. The epistemic function of valence, independently motivated as an explanation of our empirical datum, promises to open novel theoretical possibilities: it promises to help the contextualist explain how we know about epistemic standards but also to open the path to naturalize our normative practices surrounding knowledge attribution, grounding them on non-lingüistic representations of standards of knowledge.