Karl Popper and the Social Sciences

THE LIMITED RANGE OF SITUATIONAL ANALYSIS

clearly not all beliefs are produced through rational processes alone. Just as with desires and norms, people often adopt beliefs via subconscious mechanisms .To the extent that they are generated through subconscious processes, the study of desires, norms, and beliefs clearly falls outside of the purview of situational analysis. Yet quite obviously such processes must be of great interest to social science. The second shortcoming of situational analysis concerns the subconscious psychological processes that occur within the scope of situational analysis rather than outside of it. Recall that Popper requires that we never abandon the rationality principle when building our situational models.That is, we must continue to assume that actors implicated in our models always respond “adequately” to their situation even if prima facie evidence strongly suggests that the action was irrational. However, drawing on the work of Jon Elster, I will argue that the rationality principle is not always adequate for explaining all phenomena to which we apply situational analysis. Situational models, I will try to show, often require the inclusion of psychological mechanisms to explain irrational action, as well as irrational belief formation.This is, I admit, a highly contentious claim, given that Popper’s expulsion of psychology from explanations of human action is a central feature of his situational analysis. Nonetheless, I maintain that psychology can be incorporated into situational analysis without seriously violating the spirit of Popper’s methodological vision. In the following sections I discuss these two general problems with situational analysis and then conclude with an analysis of Elster’s model of political revolutions. Elster’s explanation offers a fine example of situational analysis that incorporates psychological components. His approach also offers a schema for social science explanation that, like situational analysis, gives pride of place to instrumental rationality but is open to explanations invoking nonrational and irrational action, should explanations assuming rationality prove inadequate. Social scientists wishing to employ situational analysis would do well to follow Elster’s guidelines. T H E LIMI T ED RANGE OF SI T UAT IONAL ANALYSIS Recall that for Popper the desires and beliefs of individual actors in a situational model are assumed to be part of the model itself. Desires and beliefs are not to be conceptualized as psychological properties of the individuals who figure in the situational model. They are to be transformed into aims and information held by a typical individual implicated in the situational model. The model is thus intended to explain how an abstract “anybody” would act in the situation, given that information and those beliefs (MF, 168). Popper says that the actual psychological states of actual individuals implicated in the model are not relevant to what the model seeks to explain. Returning to his model of Richard the pedestrian, Popper tells us, 100 KARL POPPER AND THE SOCIAL SCIENCES I propose to treat Richard’s aims and Richard’s knowledge not as psychological facts, to be ascertained by psychological methods, but as elements of the objective social situation. And I propose to treat his actual psychological aim of catching the train as irrelevant for solving our particular problem, which only requires that his aim—his “situational aim”—was to cross the road as quickly as was compatible with safety. (MF, 167–168, Popper’s italics) We shall not be interested, Popper goes on to say, whether Richard was thinking about Verdi’s operas or Sanskrit texts as he pursues the goal of crossing the street, which might not be in his conscious mind at all as he does so. In short, Richard’s psychological state, including the goals in his conscious mind, are not strictly relevant; what is relevant is his objective goal, as opposed to his subjective thoughts. The goal built into the situational model—the goal that, for whatever reason, interests us—is the actor’s objective goal. As we saw in chapter 1, situational models can thus be understood as ideal types rather than as complete replicas of particular social situations. No attempt is made to reconstruct the actor and the situation in all its complexity; only certain salient features are reconstructed. The next section of this chapter will explore some problems with totally expelling psychology from situational models. But here I want to note that, even if we drain all psychology out of situational analysis and turn the subject’s desires and beliefs (including his or her normative beliefs) into objective properties of our models, we still must be concerned with questions regarding the formation of...