ABSTRACT
In this paper I want to examine some of the problems that arise from taking a cognitive-behavioral approach to location theory. The argument for this approach may be summarized as follows. Locational patterns in human geography are the physical expression of individual human decisions. Locational analysis must therefore incorporate some notions regarding human decision making. The simplest course is to set up idealizations or to develop some descriptive device to summarize aggregate human behavior. The idealization of rational economic man leads us to the normative location models such as those of Weber, von Thünen, Lösch, and their academic descendants. Empirical evidence suggests that it is possible to conceptualize behavior as a stochastic decision process and use probability distributions to discuss spatial behavior provided we are considering basically similar choices in a fairly homogeneous population. Descriptive mathematical functions may then be used as the foundation for a stochastic location theory -- a theory which has yet to be written and whose characteristics remain largely unknown in spite of recent general formulations. 1 But there are many situations in geography in 36which these descriptive devices or idealizations are clearly inappropriate and there is no alternative but to incorporate very specific statements about the cognitive processes involved in the act of decision.
