List of Past Events
The architecture of human motivation: A computational and evolutionary-functional approach
Dr. Leda Cosmides
Tuesday, February 10, 2009, 01:00pm - 02:00pm
Center for Evolutionary Psychology, University of California, Santa Barbara
The cognitive revolution has advanced our understanding of many areas, but it has barely touched the study of motivation. We present a new approach to motivation that is both computational and grounded in models from evolutionary biology. The concept of internal regulatory variables is a key element in this computational approach to motivation. Intuition says that we anger, grieve, wish to help or hurt (etc) because we feel like it. This is of course true-but scientifically, the goal should be to go beyond vague intuitions and subjective descriptions of this kind. Our working hypothesis is that what generates these feelings is a hidden, previously unmapped layer of neurocomputational procedures & representations: (i) internal regulatory variables, (ii) devices that compute them, and (iii) decision rules these variables feed. These cause us to experience specific motivations, value specific outcomes, and express specific behaviors given certain inputs.
The motivational systems that regulate hunger, thirst, or breathing are interpenetrated by networks of internal regulatory variables: computational elements that evolved to track narrow, targeted properties of the world that were important for regulating metabolic and behavioral responses in ancestral environments. The same should be true of the motivational systems that regulate social behavior: these require computational variables that track properties of other people that make them valuable or dangerous as social interactants. We will illustrate the approach with data about the kinship index, a computational element that regulates altruism and sexual aversion toward kin. Time permitting, we will also illustrate with a new theory of anger.