This post is based on their co-authored papers 'Intuition and Reason: Re-assessing Dual-Process Theories with Representational Sub-Activation', forthcoming in Teorema, and 'The Fear Factor: Reconsidering the Roles of Emotion in Reasoning', currently under review.
Nonetheless, despite their significant contribution to our cognitive landscape, the role of emotions in reasoning is often overlooked, and is often understood only negatively. One reason for this may be due to an understanding of reasoning in terms of two separate, and often antagonistic reasoning systems. One system (Type-1) is fast, automatic, emotional and/or subconscious, whereas the other (Type-2) is rule-based, analytical, deliberative and/or explicit (e.g. Stanovich, 1999).
Recent evidence from work in fields as diverse as psychology, philosophy and neuroscience has challenged ‘dual-process’ theories. In line with this evidence, we favour a unitary reasoning system (forthcoming), showing that reasoning is the outcome of a compound process consisting in cognitive and affective aspects and crucially that there is a clear modulatory relationship between the two.
Focusing on this modulatory relationship, it is natural to think of emotions as analogous to perceptual states (cf. Prinz 2004), and, as such, are under environmental control, e.g. I feel scared in the presence of a predator. Crucially, emotions are also represented in the mind like perceptual states and thus form connections with each other, as well as other representations, and are influenced by the strength of the connections obtaining between them. Specifically, the stronger the connection between two representations, the higher the probability that activation of one will trigger activation of the other and in turn will influence which representations will (ultimately) feature in conscious thinking.
Given these characteristics, emotions can both facilitate and hinder reasoning. Generally, emotions contribute positively to reasoning by enhancing the formation of associations between representations of things to which we attend. For instance, we normally only have to go through a painful experience once in order to form the association between, say a knife, and the feeling of being in pain (cf. Tillas, 2010; forthcoming).
Thus, emotions contribute significantly to learning as they drive attention to things in the world that are important to the subject. Indeed, there are good arguments suggesting that emotions are a necessary requirement for reasoned inference, since there may be many more inferences and associations that cohere with some goal or other than can be calculated within the limitations of the human brain. Nonetheless, we clearly are capable of dealing with incredibly complex sets of associations without computational collapse, in the main due to an ability to identify those associations and inferences, which have immediate valence.
On the other hand, emotions could also hinder reasoning by dampening existing associations between representations or concepts that are routinely deployed in reasoning. This process most often occurs outside of the subject’s conscious control and leads to alien inferences or inferences that the subject would have never made in the absence of specific emotional states and responses. For instance, I might consider that jumping from the fifth floor is a better option than using the stairs when panicking in the event of a fire.
Crucially, this warping effect of emotional states upon our conceptual networks, and in turn thinking, can be counterbalanced, in the longer run, by voluntarily activating conceptualised versions of alternative emotional states, and the concepts associated with them. Simply put, despite the fact that emotions are normally environmentally controlled, there are also versions of emotional states (conceptualised) that we can activate on demand.
Cognitive manipulation of conceptualised versions of emotions allows us to recondition the strength of the connections within our representational networks. For instance, by voluntarily activating positive emotional states, such as safety or comfort, whilst thinking about flying, one could counterbalance a generally negative pre-disposition against it. This is a further positive contribution of emotions to reasoning. The suggested view stands in direct contrast to the ‘two systems’ theories and shows how conceptualised versions of emotions could play an additional bridging or modulatory role between the two.