ABSTRACT
This chapter reviews a model of Gordon and Trainor et al. of the harvester ant species Pogonomyrmex barbatus based on observations taken in the field over a period of years by Gordon. Social insects organized in colonies present a particular fascination for the observer and a particular challenge for the experimenter in trying to account for the group behavior in contradistinction to the behavior of individuals in the group. The more traditional approach to understanding colony behavior has been to assume a caste structure for the colony. In the Oster-Wilson model, the caste organization should provide the key to the ecology of social insects, in so far as the insects differ from their solitary counterparts. In the Hopfield model, the interactions between units are fixed at the outset and remain fixed throughout the dynamics. This feature allows Hopfield to construct a quadratic "energy" function, which decreases monotonically as the dynamics proceeds.
