Why naturalists must give up deduction, or return to Hume

This paper argues that there is a rational inconsistency in the naturalist project when rejecting analyticity and a priori knowledge but using deduction in philosophical enquiry. This is because deduction should be construed in the Humean sense – as demonstrative reasoning which allows necessary inference from conceptual truths in the premises to a conceptual truth in the conclusion. We look at a few different accounts of deduction, ones which might be able to assist the naturalist by offering an alternative reading of deduction, but in the end, reject all of them. Once it is accepted that the classical Humean account of deduction is correct, there is no plausible explanation for what it is naturalists do when they use deductive reasoning, given their other theoretical commitments. The naturalist is left with two options: accept that analyticity (logical truth) and a priori knowledge are possible and keep deduction as a legitimate method of enquiry, or forego both. (abstract 154 words)