1TRANSCENDENTAL ARGUMENTS AND REALISM

It almost amounts to a platitude that realism and transcendental arguments do not go well together. To see why, let us briefly rehearse the characteristics of transcendental arguments1: (i) their premises have to be justified independently of any experience about the external world, i.e. they must be justified introspectively or a priori. (ii) The premises cannot be rejected by the skeptic. (iii) Transcendental arguments uncover necessary conditions of thought and experience. (iv) The conclusion of the argument entails truths about the external world.2 But as Barry Stroud has emphasized since the late sixties,3 the premises of transcendental arguments seem to have only psychological status. They, therefore, do not warrant any inferences about the external world. In this manner, Stroud forces the proponent of transcendental arguments to answer the following question: “How can we proceed ’deductively’, or by necessity, from premises only about our thinking in certain ways to conclusions about the independent world which we thereby think about?”4 In order to meet this challenge, a defender of transcendental arguments has in principle three strategies to choose from5: First, one could opt for idealism. This is, basically, the Kantian strategy. It validates the inference from “inner ” evidence to the external world, b