Banbury 2005 — Abstracts
Deconstructing the Law of Effect
C. R. Gallistel
Rutgers Center for Cognitive Science
Matching is a robust widespread foraging strategy, whereby animals (including humans) proportion their investment (time spent) in foraging locations to the incomes they have derived from them. In three experiments with experimentally naive mice, we varied the coupling between investment and income--from no coupling (income independent of investment) to complete coupling (income proportional to investment). Under all conditions, the mice matched from the outset. In the third condition (income proportional to investment), the matching took the form of unstable all-or-nothing preferences for one location or the other. The results imply that matching does not depend on the return that subjects get from their investment. It is not the result of behaviors being strengthened by their consequences. It is a purely feed-forward process, which takes no account of the impact of the subject's past behavior on what it has observed.
