We use logic to structure the system of mathematics. Lord Russell was described as bewildered upon learning that original premises must be accepted on some human's "say so". Since human knowledge is so fragile (it cannot have all conclusions backed up by premises), is the final justification "It works, based on axioms accepted on faith"? In short, where do you recommend that "evidence for evidence" might be found, if such exists in the anterior phases of syllogistic construction. Somewhere I have read (if I can rely upon what little recall I still have) Lord Russell, even to the end, did not desire to rely on inductive reasoning to advance knowledge, preferring to rely on deductive reasoning. Thanks. Your individual and panel contributions make our world better.

I was intrigued that you take human knowledge to be very fragile. The reason you gave was that there's no way for all conclusions to be backed by premises, which I take to be a way of saying that not all of the things we take ourselves to be know can be based on reasoning from other things we take ourselves to know- at least, not if we rule out infinite regresses and circles. But why should that fact of logic (for that's what it seems to be) amount to a reason to think that knowledge is fragile?

Most of us - including most philosophers and even most epistemologists - take it for granted that we know a great deal. I know that I just ate lunch; you know that there are people who write answers to questions on askphilosophers.org. More or less all of us know that there are trees and rocks and that 1+1 = 2 and that cheap wine can give you a headache. Some of the things we know call for complicated justifications; others don't call for anything other than what we see when we open our eyes or (as in the case of things like 1 =1) understanding what we've been told.

This sort of reply is likely to prompt someone to ask "But how do you know that you know all those things?" That question will make some people fret, but here's a perfectly good answer: I don't know how I know all those things. Coming up with a good theory of knowledge is hard work and tends to produce controversial answers. But knowing things doesn't call for a theory of how we know things. People knew things for centuries before anyone got around to asking what exactly knowledge is and how it works.

A few things do seem clear, however. One is that not everything we know comes from syllogistic or any other sort of reasoning. Another is that we can use parts of what we know to evaluate the usefulness of other possible ways of knowing things. For example: by careful investigation, we've learned a lot about the unreliability of eyewitness testimony and memory (though we haven't learned that they're never reliable.)

But the most important thing is that there's no good reason to follow Descartes and thinking that knowledge must be based on foundations that are beyond all possible doubt. That's a premise eminently worthy of doubting, not least because it does such a lousy job of accounting for something that seems much less open to doubt: that we really do know a great deal about a great many things.

Read another response by Allen Stairs
Read another response about Logic, Mathematics