ABSTRACT
Neuroscientists aim at higher things than leeches, of course? Learning, preferably in human beings, is a major focus. But learning, because of the very long time delays involved (what is learned today may not show up in behavior until next year), undoubtedly affects very many brain areas. Consequently a full account of the neural basis3 for associative learning may have to wait until
242 STADDON AND ZANUTTO
FEEDING AND REINFORCEMENT LEARNING
Animal learning was once the dominant field in experimental psychology. Even now, it is pretty lively-and theorists continue to mine its decades of research reports for data to model (e.g., Commons, Grossberg, & Staddon, 1991; Grossberg, 1982; Klopf, 1988; Schmajuk & DiCarlo, 1991; Sutton & Barto, 1981) . The vast majority of animal-learning experiments are on what is now termed reinforcement learning. (Animals cannot follow instructions and with-
out the inducement of explicit rewards or punishments, little useful behavior can be got from most of them.) Most animal-learning experiments involve hungry animals and food reinforcement. But feeding behavior is also involved in basic mechanisms of homeostasis, the maintenance of constancy in the milieu interieur made famous by Claude Bernard (1865/1927) . Feeding is usually considered as the prototypical motivational system. Yet, motivation is largely ignored in most animal-learning experiments and also in most of the theoretical schemes that have been developed to explain them. The animals are kept hungry in these experiments so (it is assumed) their motivational level may be assumed to be high-and constant.
