Cognitive Feelings in the Predictive Mind: Emotion, Meta-cognition and Predictive Processing

Abstract

Cognitive feelings are affective states concerning the subject’s own mental processes and capacities. They include the feeling of knowing, the feeling of error, the feeling of confidence, and the feeling of forgetting, to name but a few. A question that remains open concerns their underlying mechanism. We know that cognitive feelings correlate with certain process properties, most notably process fluency, but there is no encompassing theory of how they emerge. We know that cognitive feelings influence behaviour, but there is no definite notion of the way in which they do so. Predictive Processing, a theoretical framework that conceives the brain as a hierarchical prediction machine, provides a promising way to tackle these issues. In return, a good theory of cognitive feelings will further our understanding of the interconnections between mental activity and affective states within the Predictive Processing paradigm. In this paper, we offer an account of cognitive feelings within the framework of Predictive Processing in an effort to shed light on how cognitive feelings emerge and how they guide mental action.

Forgetting Ourselves in Flow: An Active Inference Account of Flow States and How We Experience Ourselves Within Them

“…For example, building off the work of Van de Cruys (2017), Kiverstein et al (2019) suggest that "error dynamics" -the rate at which the embodied cognitive system is reducing prediction error -is at the heart of valence, such that "when an agent succeeds in reducing error at a faster than expected rate (or recognises the opportunity to do so) this feels good" (p. 2860), and vice-versa with respect to negative valence (see also Andersen et al, 2023;, Van de Cruys et al, 2020. As recent work by Fernández Velasco and Loev (2024) makes clear, Hesp et al (2021)'s notion of "deeply felt affect" and that of "error dynamics" are convergent: with respect to flow, according to both accounts, the positive valence associated with flow states is rooted in the organism's ability to reduce free energy at a better rate than they expected.…”

A computational perspective to affect generation in perceptual decisions

“…A core idea expressed explicitly (or implicitly) in different accounts of affect generation is the idea that affect is a regulatory signal (Carver, 2015;Proust, 2014;Velasco & Loev, 2022). Similarly to a signal inside a thermostat regulating the heating in a room, affect functions as monitoring signal, detecting discrepancies (e.g., conflicts, errors), and indicating the need to invest control to overcome them.…”

“…Many of these decisions are accompanied by affective feelings, such that some decisions feel good whereas other decisions feel bad. It has been argued that these affective feelings are not merely epiphenomenal but instead serve a crucial functional role of regulatory signals for cognitive control (Carver, 2015;Dignath et al, 2020;Proust, 2014;Velasco & Loev, 2022). Although affective impacts on decision-making can originate from decision-unrelated processes (i.e., incidental affect; Västfjäll et al, 2016), in the current work we focus on affect that arises as a function of the decision-making process itself (i.e., integral affect).…”

When the Predictive Brain Meets the Hard Problem of Consciousness

“…It is therefore possible to speak about hierarchical generative models of precision (Koelsch et al, 2019). This inference about the causes of precision could be equated to what psychologists called the fluency heuristic (Servajean et al, 2023; see also Hesp et al [2021], Van de Cruys [2017], Van de Cruys et al [2017], Van de Cruys et al [2021], Van de Cruys and Wagemans [2011] and Velasco and Loev [2022]). Indeed, according to some psychologists, our brain infers the causes of the fluency (i.e., the level of ease) with which our own cognitive processes occur.…”