Do Apes Read Minds?: Toward a New Folk Psychology

Chapter 3. The Asymmetry of Folk Psychological Prediction and Explanation

3 The Asymmetry of Folk Psychological Prediction and Explanation The fact is that the average person is able to explain, and even predict, the behavior of other persons with a facility and success that is remarkable. —Paul Churchland Standard Folk Psychology Emphasizes Prediction (and Assumes Explanation Follows) There is no doubt that we attribute beliefs and desires to others to a great extent, and for many reasons. We gossip about what others desire or believe, we condemn people simply by citing their beliefs (“He rejects evolution because he thinks that dinosaurs still exist and they’re really good at hiding!”), and we ascribe propositions to manipulate our audience or as an act of celebration. However, the ubiquity of our belief and desire talk in our explanations of behavior should not be taken as evidence that we consider beliefs and desires when predicting behavior. Despite this fact, it is a widespread assumption that psychological prediction and explanation are symmetrical. The general account of symmetry in folk psychology goes like this: When I predict what a person will do by appealing to initial belief-desire conditions C and the relevant information F, I can infer with some degree of probability that the person will engage in the behavior B. And when I explain behavior B, I look for appropriate initial conditions C and information F that implies B. For prediction, beliefs and desires are the inputs, and the behavior is derived. For explanation , the behavior is the input, and the beliefs and desires are derived. The slogan for the symmetry thesis is “No prediction without explanation, and no explanation without prediction.” Because in folk psychology the symmetry of prediction and explanation is thought to be cognitive rather than logical, the psychological symmetry amounts to the view that prediction and explanation are accomplished by the same cognitive processes. 38 Chapter 3 That is, the two folk psychological practices are subsumed by the same mechanisms. If the psychological symmetry thesis is true, then any account of the structures underlying folk psychology must accommodate both our predictive practices and our explanatory ones. This view has been held either implicitly or explicitly by a number of philosophers of various stripes (Fodor 1987, 1991; Churchland 1989; Gordon 1995a, 2000; Stich and Ravenscroft 1996; Nichols and Stich 2003; Sehon 2005). This assumption appears also to be widespread in the research on theory of mind in apes and children, as reflected in the writing of developmental psychologists Elizabeth Robinson and Peter Mitchell, who describe explanations as nothing more than “backward predictions” (Robinson and Mitchell 1995). Despite its popularity, I argue that we should reject the psychological symmetry thesis. If no symmetry exists between psychological prediction and explanation, then it cannot be the case that a single model of folk psychology works for both practices. A rejection of symmetry does not provide evidence that different cognitive mechanisms are at work within our acts of predictions or within our acts of explanation. It only implies pluralism between prediction and explanation. Rejecting the psychological symmetry thesis has two implications. For one, it suggests that the current attempts to model folk psychology are flawed insofar as they presume a set of mechanisms that are implicated in both our predictive and our explanatory practices. In addition, rejecting symmetry has methodological consequences for practicing psychologists and animal cognition researchers. The experiments in these disciplines are designed to find evidence of folk psychological abilities by having participants predict behavior, and from their findings, researchers make inferences about participants’ ability to explain behavior. Given the asymmetry of folk psychological prediction and explanation, researchers should develop experiments on belief attribution that examine the ability to explain behavior. The experimental paradigms designed to test for belief attribution in human children, especially in young infants, as well as those used to test for belief attribution in other apes, should be reevaluated. Before we begin to examine the symmetry thesis, we need some understanding of what is meant by prediction and explanation. A prediction can be described as an expectation that some event will occur in the future. Understanding what is meant by explanation requires a little more work. An explanation may be of an event that will happen or of an event that has already happened. Explanations are not limited to events but may be given for states of affairs, capacities, functions, intentions, and so on. Asymmetry of Folk Psychological Prediction and Explanation 39 Following Bas van Fraassen...