Zusammenfassung
According to a common reading (and criticism), the doctrine of operationism purports to define the subject matter under investigation in terms of the methods of investigation. While such a procedure would be circular, and as such untenable, I argue that we have to distinguish between two questions that are frequently conflated in discussions of operationism, i.e., (a) the issue of how to operationalize a scientific question (i.e., how to pose it in a way that makes it amenable to scientific testing), and (b) the issue of what constitute conditions of application of a scientific term. I claim that it was the former question that was of foremost interest to both early and contemporary operationists in psychology. In my dissertation I argue that this aspect of operationism highlights an interesting problem, i.e., what assumptions about the subject matter have to be presupposed in order to be able to operationalize a question about it? This question is analyzed both, by means of historical case studies (which give accounts of the operationist positions of the psychologists S. S. Stevens, E. C. Tolman, and C. Hull), and by means of a contemporary case study. In my contemporary case study, I discuss the emergence of the concept of “implicit memory” in cognitive psychology and cognitive neuropsychology within the last 20 years. I show that a large part of the research on implicit memory is concerned with the design of experiments that are aimed at giving an adequate empirical description of the phenomenon. These experiments, in turn, have to rely on what I call “guiding assumptions” about the phenomenon. Operationalizing questions about the phenomenon of implicit memory importantly involves providing interpretations for such guiding assumptions. I argue that my construal of operationism raises interesting questions about the relationship between theory and observations during the process of scientific concept formation. While it has become a commonplace within philosophy that all observations are “theory-laden”, very little has been written about the question of what this actually means for specific scientific contexts. Using my contemporary case study as a point of departure, I develop a taxonomy of the ways in which observations may be said to be theory-laden. I use this taxonomy to re-evaluate problems that have traditionally been assumed to follow from theory-ladenness, such as the problem of underdetermination of theory by evidence. This philosophical concern, I claim, corresponds to the scientific concern with avoiding experimental artifacts. I provide an analysis of the notion of experimental artifacts in terms of my analysis of theory-ladenness.
