Marketing practitioners and applied researchers often try to avoid academic speculation, preferring to ‘let the facts speak for themselves’. Their ap­ parent quest is for the unadorned description of marketing phenomena through direct observation of the facts and the first-hand experience of managerial realities. Some marketing academics, like certain of their counterparts in other disciplines, seek scientific explanation, if at all, only in the form of empirical generalizations built up from multiple observations (Hempel 1958; see also Dubin 1983; Vyas and Woodside 1984, for recent

examples). Theory and metatheory are still widely considered to be ir­ relevant to or even obstructive of useful empirical research and the effective practice of marketing. However, observation/practice on the one hand and theory/metatheory

on the other are inextricably linked. One of the most far-reaching conclu­ sions to emerge from the philosophy of science is the inevitably theory-laden nature of even the simplest observation (Kuhn 1970a; Popper 1972, 1980). Observation cannot be other than selective, reflecting a point of view; and descriptive statements are inescapably interpretations, ‘interpretations in the light o f theories' (Popper 1980: 107n; 1972: 46; emphasis in original throughout). The language in which observation is described is itself a model of reality, not the thing described; and the ‘facts’ are not logically prior to theories but are generated by them. Popular notions notwith­ standing, scientific advance does not consist in the production of successively more accurate descriptions of a subject matter independent of the con­ ceptions of scientists (Feyerabend 1970,1975; Kuhn 1970b; Lakatos 1970). The theory-dependency of observation has recently been acknowledged

in the marketing literature as part of a renaissance of interest in theoretical issues (e.g., Anderson 1983; Peter and Olson 1983). Consumer theory has also received attention (Howard 1983; Kassarjian 1982; Olson 1982). This interest is welcome in view of the atheoretical tendencies of many re­ searchers and the failure of even some of those who have seriously approached the task of theory construction to formulate testable propositions or to expose their explanations to critical evaluation (Bagozzi 1984; Foxall 1980a, 1980b; Jacoby 1978; Tuck 1976). This chapter is intended to contri­ bute to the progress of psychological explanation in consumer research by considering: (i) an analysis of choice which accords explanatory power exclusively to the environmental consequences of behavior, denying caus­ ative significance to intrapersonal events; and (ii) the relationship between that analysis and the prevailing paradigm for consumer research which derives principally from the cognitive psychology of human information processing. The focal approach to explanation, radical behaviorism, draws upon the

experimental analysis of operant conditioning phenomena and extrapola­ tions from that analysis to human social affairs in general. Both analysis and extrapolation owe much to the work of B. F. Skinner (1938, 1950, 1953, 1957, 1969, 1972, 1978), to which frequent reference will be made.1 Whilst operant conditioning has been mentioned in the marketing literature (Nord and Peter 1980; Rothschild and Gaidis 1981; Peter and Nord 1982), its discussants have been preoccupied with the search for managerial prescriptions, particularly in the manipulation of promotional stimuli.