Leda Cosmides, Professor of Psychology at the University of California, Santa Barbara, recently came to Dartmouth College to discuss her department’s work in formulating a computational and evolutionary-functional approach to understand the underlying mechanisms of motivation. She began with a picture of the Korean War: two U.S. soldiers are grieving over a comrade’s death while a third soldier fills out casualty tags. Traditional science is most concerned with the computations occurring in the third soldier’s head, but Cosmides hopes to better understand the emotions of the first two soldiers. She wondered why humans anger, grieve, and love, and whether a computational approach could elucidate these human responses. Ultimately, Cosmides developed a scientific answer that explains motivation in terms of internal regulatory variables, computing devices, and decision rules.
Current theories in psychology such as conditioning and goal-seeking models account for only a small subset of motivational phenomena. Scientists need new theoretical concepts and tools to frame the psychological issues of motivation. By describing the brain’s operations as a series of computational programs made to process information from the internal and external environment, Cosmides hopes to evoke a deeper understanding of human behavior. Psychologists often attribute human behaviors to natural selection. However, it is difficult to define what specific purpose motivational systems were designed to fulfill. By defining specific adaptive problems that have arisen in society, evolutionary biology and behavioral ecology help to answer some of these questions.
Internal regulatory variables (IRVs) are the mechanisms that govern human behavior. IRVs hold magnitudes that can be expressed with a value. They serve as computational inputs for mathematically determining how an individual responds to external stimuli. Cosmides focused specifically on the IRVs of welfare tradeoff ratios and kinship indices. If IRVs are indeed evolved specializations that govern human behavior, then a particular IRV value should give consistent results among a wide range of behaviors. For example, a high kinship index should reflect both a low desire to mate with close genetic relatives and a high inclination to care for close genetic relatives. In simpler terms, IRVs should give converging evidence on the behaviors the IRV regulates.
Maternal Perinatal Association (MPA) has been the most successful indicator of an IRV’s magnitude. MPA measures the extent of genetic relationship between siblings. MPA is determined by how long an individual has seen its biological mother nurture its neonate sibling. However, MPA only applies to the older sibling. For younger siblings, psychologists use the duration of co-residence to quantify the kinship IRV. Cosmides found that MPA reveals a 0.88 correlation between kinship index and the corresponding behavior toward relatives, while duration of co-residence reveals a 0.71 correlation.
To determine the predictive power of the welfare-tradeoff ratio (WTR), subjects were given surveys of ten questions. Each question gave an option to judge how much money the subject would sacrifice for a stranger. For example, would you rather take $66 or give a stranger $23? From the ten questions, psychologists could determine the switch point, the point at which a subject would no longer be willing to sacrifice any more for another person. From the surveys, Cosmides concluded that WTR is person-specific and that WTR is a surprisingly precise and consistent predictor of an individual’s tradeoff ratios.
Recalibrational emotions like anger and gratitude may distort IRV values. An experiment in which subjects made tradeoffs with computers has shown that if other’s WTR towards you is low, you, in response, will be less willing to bargain. Subjects recalibrate their own WTR based on the actions of others.
Through controlled experiments and statistical analyses, Cosmides suggested that our qualitative and subjective emotions and motivations may actually be the result of carefully computed processes. After all, data indicates that people generally have consistent IRVs. Ultimately, Cosmides’ work suggests that the psychology of motivation may be, at its most basic level, just a series of calculations.