Much effort has been spent in the last few years on synthesizing classical location theories and rewriting them "in behavioral terms" [18, 24, 25, 26, 33, 35, 53,541. A common, though never very explicit, goal in this work appears to be a single general theory of human spatial choice, which will give "the real reasons why people behave . . . in the way that they do" in decisions about where to conduct many kinds of activity [Harvey, 35, p. 371 This paper leaves aside troublesome questions about the more precise definition and the merits of such aspirations (for discussions see [16, 25, 33, 42, 43, 51, 631 ). The paper considers ( a ) the identification of the properties of alternatives entering choice behavior and (b) the incorporation of these properties in a general cognitive behavioral theory of spatial choice processes.Discussion is confined to changes of choice behavior within spatial structures: with particular reference to consumer choice of shopping place from a given set of centres. The paper is constrained in these ways to allow two intractable problems to be avoided for the moment, namely, the problem of unravelling the complex interactions between changes in spatial structures and changes in learning and behavior within structures, and the problem of requiring the collection and aggregation of data on the choices of individuals over a long period. (These prob-
REVIEWDespite the need to revise dubious theoretical assumptions about the variables that condition individual choices, work to date has not confronted the question of attempting to identify both the properties of spatial alternatives as they are perceived by individuals and the ways these might vary with successive choices over time. For example, it has long been the practice simply to assume that the attributes of alternatives that influence selections of a shopping place embrace one or more of the following: distance from home, ease of parking, centre size, variety of goods (functions), quality of goods, prices of goods, and so on.2 However, such objective attributes need not be the real dimensions of centres that influence choice, even when they are endorsed by shoppers in field surveys [22, 46, 601. Consumers conceivably respond to subjective alloys, such as "value for my time and money," which may not be related in a simple way to objective centre properties. Moreover, consumers may vary the kinds of attributes used to select alternatives as their learning proceeds, and may be imperfectly aware of the properties they employ.Nonetheless, behavioralists in geography and regional science, in attempts to revise traditional approaches to consumer behavior, have in general left aside the problem of identifying the properties of given spatial alternatives as they enter choice processes. Rushton continues to assume that distance and town size are appropriate surrogates for whatever might be the true attributes of centres that condition choice of shopping place [54]. Golledge also assumes that the traditional objective attributes of prices of goo...