geoff

On an "Existence of God(s)" topic, I posed a question related to a classic epistemological question that I now realize is more suitable for this forum. Before I pose the question, the following preamble will be helpful in guiding the discussion.

Since at least Socrates' Meno it has been common to think of knowledge as justified true belief. It may be formulated as follows:

A person S knows some proposition p just if

(1) P is true.

(2) S believes p.

(3) S is justified in believing in p.

(1), (2) and (3) are individually necessary and jointly sufficient for something to be an instance of propositional knowledge. (3) has proven to be the most difficult condition to meet. Edmund Gettier wrote a seminal essay (http://www.ditext.com/gettier/gettier.html) in 1963 which attempted to argue that one could come up with many instances in which justified true belief did not in fact constitute knowledge. Henceforth, this dilemma has been called the Gettier problem.

So ends the preamble. Now the question: how would you propose to justify your ostensibly true beiefs?

(I would note that this question has been vigorously pursued by epistemologists of all stripes and, in various degrees, in all ages. See, e.g., John Pollock's Contemporary Theories of Knowledge".)

Wyrdsmyth

Hi Geoff,

Looks like it's Old Chestnut Time again.
What has always struck me as funny about the definition for knowledge being "true, justified belief" is how redundant it is. If I could fulfill condition (1) by demonstrating how P is true, then it will naturally follow that (2) I'll believe in it and (3) I'll be justified in my belief.

Unlike Gettier, I think condition (1) is the difficult and most crucial item. We all believe things, and we all justify our beliefs... People can and do believe things that are false, and justify that belief. Hell, I've done it myself. That's easy. But how do you show something is true?

Throughout history, we have often called something 'knowledge' and it turned out not to be true. We're probably doing that about many things even today. We probably always will. At one time we humans in Western society 'knew' Newton's laws were true, and that they were proven. Newton's laws were the laws of nature. We believed it and could justify it.

But, it turned out they weren't true. Newton's laws were models that worked eminently well within certain parameters, but they were not exactly flawless. Newton did not discover these laws 'written on the walls of the universe,' as it were.

So, if we have to show that P is true before we can claim to know anything, it may turn out we don't know a whole heck of a lot.

P.S. Are you a teacher?

[ May 21, 2002: Message edited by: Wyrdsmyth ]</p>

Vorkosigan

My objection is also to (1). We can't know P is true, because that would mean we had some source of knowledge that did not involve (2) and (3). In my view we only have (2) and (3), we can never have (1). If philosophers have a problem with (3), they must be even more screwed up than I thought.

Vorkosigan

[ May 21, 2002: Message edited by: Vorkosigan ]</p>

jordan_tar

Why "justified" - could you elaborate?

geoff

Wyrdsmyth and Vorkosigan:

('1) affirms only that some proposition p is true (e.g., p could be "2+2=4", "the sun is shining", "my car black"). ('3), in my mind, is precisely the problem in question. We know p (e.g., take any of the propositions listed above), is true, but how we are personally justified in believing that p is the rub.

Wyrdsmyth:

No, I'm not a teacher.

Jordan_Tar:

"Justification" means, variously, to provide a rational account or to provide sufficient warrant for a particular belief. "Justification" is that which distinguishes true belief from opinion. Justification typically comes in two varieties: (1) evidential, in which one is justified in believing that p if and only if sufficient evidence grants such belief or (2) deontological, in which one is justified in believing that p if and only if she is within her epistemic rights.

For a good online dictionary for philosophical and other terms, see <a href="http://www.xrefer.com/search.jsp." target="_blank">http://www.xrefer.com/search.jsp.</a>

Vorkosigan

Sorry, but we're still not past (1) here. (2) and (3) are interesting problems here only if you accept (1). You can't make a proposition true by "affirming" it. Have you written "affirm" where you meant "assume?"

Your dictionary link doesn't work.

BTW, Gettier's "seminal" paper is really...silly (you should read it Wordsmyth, it's short and a good belly laugh). In his example of Smith, Jones with the Ford and Brown, how did they confirm Brown was really in Brest-Litovsk? By the same kinds of processes that Smith had originally used to know whether Jones did or did not possess a car. No one possesses the godlike perspective it requires to know for certain that Brown is indeed in Brest-Litovsk. The problem with this is not (3), which is quite an interesting debate in the philosophy of science, but (1). There is no way ever to get to (1). We can only pile up evidence and argument until (1) seems irrefutable, although at that point we will only have a really strong (2).

Vorkosigan

[ May 21, 2002: Message edited by: Vorkosigan ]</p>

tronvillain

You know, I seem to recall that we did this once before. Anyway, as I said, we virtually never have justified true beliefs, we simply attempt to approximate them probablistically.

The three conditions don't really make sense, since if someone is truly justified in believing in p, then p will be true.

[ May 21, 2002: Message edited by: tronvillain ]</p>

geoff

Some of you evoked the notion of probability to explain the ability of a person S to know some proposition P to be true. In other words, P is true because P is more probable than not-P (I'm using a very weak sense of probability. A stronger sense could be used, but the weaker sense proves the point just the same). It seems you are saying, conversely, that we cannot know that P is true (in the strong sense of logical certainty) because, at best, our ability to know any proposition P is restricted to probability.

Admittedly, many of our beliefs are only probable. Similarly, many of our beliefs (beliefs which seem intuitively to be true) are unable to pass the stringent justification criterion. But isn't the distinction not one between knowing P to be true with certainty and knowing P to be true probabalistically? (I would think that all of us agree that 2+2=4 is a true proposition. We don't know that fact probablistically; we know it with certainty.) Rather, isn't the distinction really between what beliefs we can know with certainty (else we would lapse into complete skepticism) and how those same beliefs can only be explained in terms of probability? The distinction is basically between what we can know and what we can provide an account for. (Vorkosigan, I thought I had read one of your posts from a previous thread that affirmed this idea.) Isn't, e.g., 2+2=4 de facto true? Do we really need "godlike" omniscience to know 2+2=4? Isn't that proposition true regardless of the fact that we are finite?

If I'm misrepresenting the intent of your thoughts, please correct me.

[ May 21, 2002: Message edited by: geoff ]</p>

Sicarius

Well, using that formula, one can permit any sort of specious reasoning to become knowledge.

(1.) We take the belief, in which for this case I will use something completely rediculious. Or better yet, I'll use the example of God, and the belief that he keeps the world from destruction, etc...
(2.)Subject S believes the above.
(3.) And finally, using specious reasoning, and observing that the world is, (within reason) not instantaniously breaking apart and killing us all, (1.) has become basic knowledge.

To put it simple, the formula is quite faulty, due to the fact that its utterly openended, and can be put into intepretation for any statement.

Vorkosigan

riginally posted by geoff:
(I would think that all of us agree that 2+2=4 is a true proposition. We don't know that fact probablistically; we know it with certainty.)

Well, yes, because we've defined a logical system with certain axioms, so certainty is possible, because perspective is limited. But we could never be certain in the sense of (1) above that Brown is really in Brest-Litovsk. We'd only be able to pile up evidence (I saw him there) and disprove alternatives (I called his house in New York and they said he wasn't there).

Rather, isn't the distinction really between what beliefs we can know with certainty (else we would lapse into complete skepticism) and how those same beliefs can only be explained in terms of probability? The distinction is basically between what we can know and what we can provide an account for.

There isn't any distinction between those. They are the same thing. We can't "know" anything we can't provide an account of. How could we? We can't ever know any beliefs with certainty, because certainty as it is normally used is a social convention, a conversational shorthand that allows us to avoid lengthy explanations. Just today I had the following conversation:

Vork: Was Michelle in class last week?
Glenda: Yes, I borrowed a pen from her.
Vork: Are you sure?
Glenda: You bet.

Glenda was not about to give me a long explanation that ruled out reasonable alternatives while simultaneously putting forward proof. Even if she is sincere, all she is saying is "Any reasonable person would agree with me that the preponderance of evidence and experience suggests there is a high probability of Michelle having been here last week."

To me what prevents us from lapsing into complete skepticism, Geoff, is that knowledge is social. We exist in networks of intersubjectivity. the "reasonableness" of an idea to others is an important check on our logic and evidence. Ultimately "certainty" is both social and provisional.

Vorkosigan