First I’ll give a simple rebuttal, then i’ll provide a more thorough treatment.
I’ve got three keys on my keychain. One of them opens up the door in front of me. I try the first two; they don’t work. What kinds of assumptions do I have to make in order to know that the third key will open the door? If you think such assumptions need justification, well, then we haven’t got anything to talk about because you’re operating on the principle that nothing whatever can be proven and therefore language is meaningless.
A more rigorous treatment of the problem reveals that the question is more or less nonsense. The questioner asks us to justify the statement “logic is reliable”, or else capitulate that the statement is an unprovable assumption. Unfortunately, this doesn’t make any sense.
First, I’ll examine what it would mean if logic were unreliable, then show that if you can only ask whether logic is reliable if logic is reliable. According to the University of Hong Kong’s philosophy department, logic is “the study of the principles of correct reasoning”. Under this definition, if logic were unreliable, meaning human understanding of 'the principles of correct reasoning’ is unreliable, then we humans wouldn’t be able to distinguish good reasoning from bad reasoning (at least not reliably, which is essentially as bad as not at all since everything becomes suspect). Austin Cline from Brian’s link confirms this interpretation, “Logic is what allows us to distinguish correct reasoning from poor reasoning.”
(You might argue that we are in fact unable to reliably distinguish good arguments from bad, otherwise why is there so much disagreement? The short answer is that while logic is required for telling when an argument’s conclusion is sound, it is not sufficient. Logic may allow us to distinguish between good and bad reasoning, but an argument can be logically consistent and have wrong conclusions if it starts from false premises. Logic doesn’t preclude ignorance.)
Now, if you want to show that logic is reliable, you can’t use logical argument to prove it — that would be using the validity of X to prove that X is valid. You would have to start from the position that logic may or not be reliable. From this position, how would we distinguish a correct argument for reliable logic from a bad one? That is, if we start by assuming that maybe we can’t distinguish good arguments from bad, any argument for anything at all will have indeterminate truth-value. Let me emphasize what I mean by indeterminate: if you can’t tell whether or not an argument is sound or invalid, then there is effectively no difference. If you can’t see, then blue and red are indistinguishable.
Now let’s recap: in order to answer the question, “Is logic reliable?”, one must begin without assuming that logic is reliable. This question proposes that perhaps there is no observable difference between good and bad arguments, and then it asks for someone to argue that this is not so. The question assumes that questions may not be answerable, thereby invalidating itself.
As for the idea that God is the guarantor of all truths, well, the best argument against that is that God doesn’t exist and hence can’t guarantee a cell-phone warranty let alone all of everything true.
Posted: August 19th 2007
See all questions answered by George Locke