THEORY OF ANIMAL TIMING--MEMORY AND CHOICE

前几年在这个项目上已经看到了定量的发展

Previous years on this project have seen the development of a quantitative theory of animal timing, Scalar Expectancy Theory (SET), applied to a wide range of timing effects in animal learning. Past work has analyzed the way in which time delays are remembered and discriminated in a variety of animal psychophysics preparations. The preceding grant period has also studied temporal control in association leaning, and resolved some of the tensions between SBT and alternative stimulus competition theories of associative learning, prototypically the Rescorla-Wagner theory. The major focus of the competing continuation is the study of the role of memory sampling rate in continuous choice situations. Some recent challenges to our previous account of concurrent choice (the matching law, Herrnstein, 1961) are analyzed in depth. Matching is a well-worked phenomenon with a very large literature, but one which only recently has been amenable to differentiating between different molecular mechanisms. The major contribution in the present proposal is the analysis of the free-running rates at which memories are accessed in concurrent choice situations. This results in a system which is formally a Markov chain in space and time. We show that paradoxical preference results are explained by a combination of the rate at which memory is sampled and the probability with which choice for a given alternative is made. Application of these ideas to 2-way and 3-way choice is pursued. One result of this analysis is a characterization of a foraging analogue of choices between prey types and delays to prey, the classical diet choice problem. A later focus in the project is study of concurrent choice when the alternatives are averaged rates of payoff, analogous to patch choice problems in behavioral ecology.