Ideal and Non-Ideal Theory and the Problem of Knowledge

This article analyses a hitherto neglected problem at the transition from ideal to non-ideal theory: the problem of knowledge. Ideal theories often make idealising assumptions about the availability of knowledge, for example knowledge of social scientific facts. This can lead to problems when this knowledge turns out not to be available at the non-ideal level. Knowledge can be unavailable in a number of ways: in principle, for practical reasons, or because there are normative reasons not to use it. This can make it necessary to revise ideal theories, because the principle of ‘ought implies can' rules out certain theories, at least insofar as they are understood as action-guiding. I discuss a number of examples and argue that there are two tendencies that will increase the relevance of this problem in the future: the availability of large amounts of sensitive data whose use is problematic from a normative point of view, and the increasing complexity of an interrelated world that makes it harder to predict the effects of institutional changes. To address these issues, philosophers need to cooperate with social scientists and philosophers of the social sciences. Normative theorising can then be understood as one step in a long process that includes thinkers from different disciplines. Ideal theory can respond to many of the charges raised against it if it is understood along these lines and if it takes the problem of knowledge and its implications seriously.