Tuesday, March 11, 2003


Just one more breather from moral philosophy: the celebrated 'problem of induction'.

What would count as evidence that inductive reasoning was unreliable? Nothing would count. Any evidence put forth against the reliability of induction would be either deductive or inductive inferences. Deductive inferences won’t get you anywhere. In that department, there is only Hume telling us that induction is not deductively airtight. Hume famously pointed out that the evidence afforded by our experience of the past does not give us certainty of the future, where “certainty” refers to the logical impossibility of error. All one can distill off from this argument is that induction is not deduction, which is trivial.

As for inductive evidence that induction is unreliable, i.e., not conducive to reaching the truth, well, this is an incoherent idea.

So, the question, Are inductive inferences rational? or Is induction truth-conducive? is a non-question. There is no problem of induction. Inductive reasoning is by definition rational, reliable, truth-conducive. "Reasoning" means "using deduction or induction".