Bill Snedden

THAT'S where it was! Okay, this is what I was talking about, only I described it almost completely incorrectly.

They say that memory is the first thi...what was I saying?

Regards,

Bill Snedden

Francois Tremblay

Geesh. This is, as tronvillain demonstrated, pure semantics legerdemain. The biggest fallacy committed by these examples is the analytic/synthetic dichotomy. Once you dissociate the two, then of course word games would become problematic. But if one retains the full context of the entities being discussed, there is no room for problems.

bd-from-kg

Franc28:

Well, of course it’s semantics. We’re not talking about what the word “knowledge” “really” means (a meaningless question), but what’s the most useful way to define it. One criterion of usefulness is that the definition should agree pretty well with established usage, another is that it should sometimes be applicable in nontrivial cases. Not everyone finds such questions interesting. If you don’t, no problem. No great issue is at stake here.

tronvillain:

You said:

Are you serious?

On Sept.10 last year I believed that I’d be able to watch an NFL football game the following Sunday. The main justification for this belief was that a number of NFL football games were going to be broadcast that day. As it turned out, this was not actually the case for reasons we are all sadly familiar with. Was my belief that I would be able to watch football that Sunday therefore unjustified? This makes a mockery of the notion of “justified belief”.

In this context it’s a given that there is such a thing as false justified belief. Otherwise no one would bother to define “knowledge” as justified true belief, since all justified beliefs would be true by definition. Whether a belief is true and whether it is justified are completely different questions. The notion of “justification” in epistemology is grounded in what is known, not in what is true. If a belief is well-grounded in what you know, it is “justified” regardless of whether it’s true. Thus if I was justified on Sept. 3 in believing that I’d be able to watch football that Sunday (which I surely was) and I had the same grounds for believing the same thing the following Monday, this belief was also justified.

But if you insist that the justification for a belief must be true for the belief itself to be justified, some minor tinkering with the example will take care of this. I can say that my justification for believing that the track would be muddy was that it been muddy after the last few hundred heavy rains. The idea that it is justified to believe that D will follow the next occurrence of C if it has followed the last several hundred occurrences of C is called the principle of induction, and is considered by most sane people to be a valid justification of a belief.

Francois Tremblay

I'm not sure how debating contextless quandaries is going to help you define "knowledge", because any such epistemic definition would inevitably be based on a proper notion of identity.

I usually just say "knowledge is correlation to reality", which is perhaps simple but right away points to what must really be done, i.e. find what is real in this or that case.

[ January 29, 2002: Message edited by: Franc28 ]</p>

tronvillain

Actually, I do consider your belief that you would be able to watch football that Sunday unjustified. As you point out, justification is grounded in the known, but you did not know that a number of football games were going to be broadcast that day. You considered that highly probable, so you considered yourself justified in your belief.

It's quite difficult to achieve full justification, and in the case of the future it's virtually impossible. This makes actual knowledge somewhat difficult to obtain, and it is because of this that we resort to probabilistic justification. Unfortunately, this opens up the door for false justified belief.

bd-from-kg

tronvillain:

Actually it’s impossible to obtain “full justification”, in the sense of having reasons that justify absolute metaphysical certainty, about anything whatsoever. Thus if we define knowledge to be this kind of absolute certainty, it doesn’t exist. A word with no application is useless. So unless we are going to chuck the word out of the language and expunge it from the dictionaries, we need to define it as something less than absolute certainty.

We could, of course, define knowledge as “true belief”. (I assume that this is what Franc28 has in mind when he suggests defining it as “correlation to reality”.) But this has problems as well. For example, say Smith “sees” rats under the porch because he’s suffering from delirium tremens, but as it happens there really are rats under the porch. Smith has a true belief, but it can hardly be called “knowledge”.

So we are led to the idea of defining knowledge as “justified true belief” (JTB). But we have to cut a little slack. If we say that we must know with certainty that the justification itself is true and that it must imply the belief with certainty we’re back where we started: no belief can meet this criterion. So we have to relax either the requirement that we must be absolutely certain that the justification is true, or that the justification implies the belief with certainty. Obviously in either case it will be possible to have justified false beliefs.

This is where Gettier comes in. He points out that the JTB definition of knowledge doesn’t work, because (if we don’t take the extreme, self-defeating position that the justification must be certain and imply the belief with certainty) there are many cases in which a belief is justified and true, yet we would not ordinarily say that the person having the belief has “knowledge” because the belief happens to be true for reasons completely unrelated to the justification.

My point was that ManM’s proposed solution doesn’t work for most “Gettier” cases.

If you still want to define knowledge as “absolute certainty” or to insist that any “justification” must render that belief absolutely certain for it to qualify as knowledge (which comes to the same thing) , there is no reason for you to be interested in Gettier’s critique of the JTB definition of knowledge.

Francois Tremblay

This is not a very good argument. You are confusing truth with naive realism - a proposition that came from an error of interpretation cannot usually be rationally called "true". My error of interpretation makes my proposition discord with reality.

For example, in the case of the rat, Smith is omitting from his context the fact that he is suffering from delirium tremens (he may not be able to recognize that fact, but that is another matter). Therefore his proposition cannot be rationally labeled "true".

[ January 29, 2002: Message edited by: Franc28 ]</p>

tronvillain

bd-from-kg:

I define "knowledge" as "justified true belief" and I do not cut it any slack. Of course, as a result I think that knowledge is not something we ever really have, but rather is something we attempt to approximate.

tronvillain

Oh, and I suppose I agree that ManM's solution doesn't really work. It's just that it's close enough to mine that the difference didn't really seem that significant.

bd-from-kg

Franc28:

Ok, I’m a little slow sometimes, so let’s take this one step at a time.

P1: “X is the case” is true if and only if X is the case.

- See Tarski.

P2: A belief is true if and only if the thing believed is true.

- By definition.

C1: If Y believes “X is the case”, his belief is true if and only if X is the case.

- Follows immediately from P1 and P2.

C2: If Smith believes there are rats under the porch, his belief is true if and only if there are rats under the porch.

- Follows immediately from C1 by simple substitution.

P3: Smith believes there are rats under the porch.

- By stipulation

C3: Smith’s belief is true if and only if there are rats under the porch.

- Follows immediately from C2 and P3.

P4: There are rats under the porch.

- By stipulation

C4: Smith’s belief is true.

Where did I go wrong? Which premise is false? Which conclusion does not follow from the previous steps as indicated?

I’m baffled.