What We Talk About When We Talk About Knowledge

So we accept that knowledge (or at least the possibility of knowledge) exists. Now what? Now we try to establish what we know about knowledge, starting with its conditions. What are we talking about, precisely? Let's unpack a particular claim of knowledge.

I know the Eiffel Tower is in Paris.

What do I mean when I say this? In other words, what would have to happen for me to be wrong? Well, at the most basic level I appear to be saying that I personally think that the Eiffel Tower is in Paris. If I believed that it was in Berlin, say, then I would be lying when I said I knew it was in Paris. So I am making a claim about what I believe. We can generalize this and say that knowledge requires belief.

I'm not just talking about my belief, though - otherwise I would have said 'believe' instead of 'know'. I'm making a claim about the outside world, in particular about the architecture of Paris. If I believe that the Eiffel Tower is in Paris, but I happen to be wrong, then I don't really know it, do it? The concept of 'wrong knowledge' is incoherent. So we can say that knowledge requires both belief and truth.

Now we are getting somewhere. If I know something, that must mean that I believe it and that it is true. But is this sufficient? Suppose that I have no acquaintance with any facts about Paris, or the Eiffel Tower in general. In fact, I only heard about the Eiffel Tower (which I dimly imagine to be a rifle-thin copy of Saruman's lair from the Lord of the Rings) a few minutes ago, when it came up in conversation. Feeling intellectually outgunned, I jump into the conversation with a wild guess: The Eiffel Tower is in Paris! Do I believe it? Let's say yes, since I'm the kind of person who very quickly develops certainty for no reason. Is it true? Yes, by sheer luck I've picked the right city. But do I actually know that the Eiffel Tower is in Paris?

Of course not. So there must be a third component of knowledge; some reason to think that your true belief is true. I must have some kind of justification for believing that the Eiffel Tower is in Paris. Photos, for instance, or multiple references in books and films. Knowledge, then, requires (at least) three things: that you believe a certain thing, that that belief is correct, and that you have good reason to hold that belief. This is what philosophers call "justified true belief", or JTB, and it was considered a full definition of knowledge for considerably more than a thousand years.

Is it, though? Can you think of a case where the three conditions - justification, truth and belief - are all met, but the belief still does not count as knowledge? If you can deduce it yourself (I certainly couldn't), then you'll find my next post on epistemology trivial.