I’m reworking, today, a paper (submitted to a journal which did not want a draft of the paper on this blog) which compares and contrasts Wittgenstein’s and Campbell’s views on levels of explanation and rational explanation. In it, I argue that despite the similarity between their rejections of a necessary connection between rational explanation and neurological explanation, there is a significant difference.
Campbell’s rejection of a priori connections is put to the service of a denial of coherence to the very idea of levels of explanation. Such an idea presupposes, he argues, the idea that causality presupposes intelligibility in the world and Hume refuted that. That is, Campbell ties the idea of a level of explanation to a causal structure which makes sense to us. He then argues that because causation is in fact merely brute (following Hume), thinking that causal explanation tracks an intelligible structure in the world mistakenly projects a synthetic a priori structure onto the world. Philosophers are tempted by such synthetic a priori claims but should reject them.
Wittgenstein’s discussion of rational explanation suggests that there is, contra Campbell, reason to hold that there is such a synthetic a priori. It is the price of the assumption that a state is intentional or content laden.
As part of the context of this, my paper describes Putnam and Oppenheim’s ontological view of levels of explanation and Marr’s epistemic view. It then sets out Dominic Murphy’s discussion of the fact that psychiatry fits neither of these approaches because it invokes explanations that cross all levels. One reviewer of the draft paper asks: if psychiatry doesn’t fit any picture of levels of explanation, why would readers be interested in Campbell’s criticism of the very idea of levels of explanation? Murphy’s claims undermine the motivation to consider Campbell’s.
But it seems to me that this is wrong. Murphy’s claims address forms of explanation currently in vogue. They leave as possibilities that either ontological or epistemological housekeeping might return the scene to Marr’s or Putnam and Oppenheim’s views of what must be. Campbell goes further: if he is right, there cannot be any levels with which psychiatric data can be reconciled. No house-keeping is possible.
Against that, Wittgenstein argues for something distinct. There need not be any harmony between levels of explanation, but there is a reason to hold to some distinct forms of intelligibility. The rationale for a synthetic a priori is that that is the way intentionality comes into sight.
Is that clear?