Causation With a Human Face

The experiments described seem to show, that as an empirical matter, features of causal relationships having to do with invariance influence ordinary subject's judgments (as expressed in responses to various causal strength probes) about how strong" or"good" those relationships are or about the extent to which the subjects find it appropriate to describe those relationships as causal. In fact, several distinguishable invariance-linked features seem to influence causal judgment, including those incorporated into Cheng's causal power model and the considerations having to do with background invariance explored by Vasilyeva, Lombrozo and colleagues. (I expect other invariance based features of the sort described in Chapter 5 also to be influential.) Of course, as contended throughout this book, showing these empirical influences on causal judgment does not establish that they are normatively justified or appropriate. However, as argued in Chapter 5, there are strong normative reasons why various sorts ofinvariance considerations should matter in causal judgment. Thus to the extent that we find invariance considerations influencing people's judgments and behavior, we have reason to think that they are judging and behaving in normatively appropriate ways-that they are in these respects "rational" in their causal cognition, thus providing support for the rational model approach advocated in earlier chapters. Returning to the role of intuition/judgment about cases in establishing conclusions about either causal cognition or "causation itself", I note that many accounts of causation developed by philosophers do not assign an explicit role to invariance-based considerations (at least under that description) either in how we think or how we ought to think about causation. Suppose that a traditional philosopher consulting her intuitions/judgments about various cases reports either that (i) invariance-based considerations do not influence her judgments or, alternatively, (ii) simply fails to report that they do (perhaps because the possible relevance of invariance does not occur to her). If the philosopher is like the subjects in the experiments described in this chapter, we have reasons to suppose that in reporting (i) she is mistaken and that in reporting (ii)she has missed something important that influences her judgments. If the philosopher claims then that she is not like ordinary subjects and that in virtue of her philosophical training her different, invariance-neglecting judgments are more reliable or more informative than those of ordinary subjects, we can remind her that there is a normative theory supporting the role of invariance in the ordinary judgments, thus casting doubt on her claims of special expertise. In this way a combination of empirical results and normative theory can play a role in circumscribing and critically assessing claims deriving from intuition"--