Week 8: Irrationality

Weakness of Will

GAGE neurocomputational model of decision making.

From P. Thagard, The Moral Psychology of Conflicts of Interest:

If the decision whether to eat or exercise were made simply by a rational calculation that maximizes expected outcomes, it would primarily be made in the prefrontal cortex, wherein resides the bulk of our linguistic and mathematical abilities. But evidence from functional magnetic resonance imaging suggests that decisions involving immediately available rewards involve parts of the limbic system associated with the midbrain dopamine system. In particular, the nucleus accumbens is well known to be involved in rewards involving food, alcohol, drugs, and sex. Hence the reason that people suffer from weakness of will is that their decisions are not driven by rational calculation about long term results, but rather by activity in emotionally powerful brain areas such as the nucleus accumbens that respond to immediate stimuli. Without conscious awareness of what is going on in our brains, most people have been in situations where the prefrontal cortex says no, but the nucleus accumbens says yes, yes, yes. Weakness of will is a natural consequence of how our brains are organized to sometimes give priority to motivational rather than cognitive influences.

Motivated Irrationality

The RIM Park scandal.

From P. Thagard, The Moral Psychology of Conflicts of Interest:

Self-deception is puzzling from the perspective of a unitary self, which would seem incapable of both believing something and of potentially realizing that the belief is unwarranted. From the perspective of the GAGE model, however, the self is far from unitary. Consciousness may appear to be unified, but it gives a very misleading picture of the workings of the mind/brain. Ideally, belief acquisition would be a purely rational process carried out by prefrontal cortex, with acceptance and rejection determined solely by considerations of evidence and overall coherence. But the brain does not have the processing and storage capacity to acquire beliefs of no importance, so we tend to focus belief acquisition on topics that matter to us. Whether a topic matters is intimately tied in with our emotional reactions to various situations, ranging from interest, curiosity, and enthusiasm to anxiety and fear. Hence it is not surprising that there are rich neuronal pathways connecting the prefrontal cortex with intensely emotional brain areas such as the amygdala and nucleus accumbens. There is abundant psychological evidence that people’s beliefs are determined in part by their motivations as well as by the available evidence. And there is no central homunculus to check the consistency of one set of beliefs that may be acquired based on evidence and warranted inferences against another set that may have arisen primarily because it fits with one’s personal goals. GAGE models decision making rather than belief acquisition, but it makes it easy to see how self-deception could arise by means of emotion-driven motivated inferences to beliefs that could potentially be recognized by purely neocortical processes as unwarranted.

Discussion Questions for Week 9

  1. Do psychological studies of inference show that people are often irrational?
  2. What are the main differences between the Tversky/Kahneman and Gigerenzer schools?
  3. Does an evolutionary perspective help to understand rationality?
  4. What are the normative implications of experimental studies of inference?

PHIL 680

Computational Epistemology Laboratory.

Paul Thagard

This page updated Oct. 31, 2005.