Wikijunior:Philosophy/Epistemology

Epistemology

Epistemology is the study of how we know what we know.

What is Knowledge?

The fundamental question that epistemology wants to answer is what knowledge is. A definition of knowledge that many philosophers have used is justified true belief. That is, you know something if:

You believe it to be true. Obviously, you can't know the sky is blue if you think it isn't!
It actually is true. You can't know that the sky is actually purple, since it isn't.
That belief is justified.

This third point is the most complicated part. What justifies beliefs? Well, philosophers have suggested some ideas:

Coherentism: A belief is justified if it fits well with other beliefs you have.
Foundationalism: A belief is justified if it's based on basic beliefs; beliefs that are so obvious that they don't need further justification.
Reliabilism: A belief is justified if you form it through a reliable process.

So, how does our belief about the sky fall into this? Well, coherentists don't look at beliefs individually, they view them as part of a big web. Whether or not the belief that the sky is blue is justified depends on the other beliefs in your web. Let's say that you believe that your eyes are always wrong, and whatever you see is the opposite of what is true. This belief does not fit well (or cohere) with the belief that the sky is blue, because you came to that conclusion using your eyes. So, your belief that the sky is blue would not be justified. However, if you believe that your eyes are generally right, then your belief that the sky is blue would be justified, since those beliefs cohere well with each other.

To a foundationalist, this is really simple. Your eyes give you direct knowledge of what the sky looks like. It's non-inferential, meaning that it's not derived from other beliefs. You just instantly know that the sky is blue by looking at it, and this belief justifies itself.

Reliabilism is a bit different from the other two theories, because it's what's called an externalist theory, while the others are internalist theories. Internalist theories say that you must have access to the factors that justify your beliefs, whether they be immediate experience (foundationalism) or other beliefs (coherentism). Meanwhile, reliabilism is externalist, meaning that you don't need to have access to the justifying factors for your beliefs; in this case, you don't need to be aware about whether or not the process that caused you to form your belief is reliable, all that matters is that it actually is reliable. With our sky belief, it's justified as long as you aren't colorblind, hallucinating, or, for some other reason, not able to properly determine the color of the sky.

The New Evil Demon Problem

One thought experiment that has been brought up a lot in the debate about what theory of justification is correct is the "new evil demon problem." It goes like this:

Imagine that there are two people, John and Jimmy, who both look up at the sky at the same time, and both see that it's blue. However, Jimmy lives in a world where an evil demon has deceived him so that he sees the sky is blue when it, in fact, is orange. Meanwhile, John is in a world where the sky is blue. They both form this belief (and all of their other beliefs) through the exact same process, and otherwise have completely identical mental states, but all of Jimmy's beliefs are the opposite of the truth, due to the evil demon.

The question is, are Jimmy's beliefs justified?

Coherentists say yes, since he believes that he isn't being deceived, and that coheres well with his belief that, say, the sky is blue. Foundationalists also say yes, since the belief is formed through his ordinary sensory experience, which produces basic beliefs. Reliabilists, however, say no, since all of his beliefs are false, meaning that the process in question is not reliable. Which do you think?

The Gettier Problem

Recently, the "justified true belief" definition of knowledge has become less popular, due to what are called Gettier cases. Here is an example:

You see the clock to your right, that says that it's 2:00. This makes you believe that it is 2:00, and it is in fact 2:00 at that time. However, the clock is actually broken, and always displays 2:00 no matter what the time is. It's just by a sheer coincidence that you happened to look at it when it was exactly 2:00.

Here, you do believe that it's 2:00, it is true that it's 2:00, and this belief is justified, because you based it on what the clock said; this coheres well with your belief that clocks are generally reliable, so it's fine for coherentists, clocks actually are generally reliable, so it's fine for reliabilists, and it's based on your immediate experience of the clock, so it's non-inferentially justified for foundationalists. But, do you really know that it's 2:00? Your belief is true essentially by accident, and there was no connection between your justification and why the belief was actually true. So, this suggests that knowledge is more than just justified true belief.

Some people have suggested ways to get around this:

The No-False-Lemmas (NFL) Condition: This adds one more requirement to knowledge beyond justification, truth, and belief. This extra requirement is that the belief must not be based on any false beliefs. For instance, your clock belief is not knowledge, since it requires you to falsely believe that this clock is not broken.
The Tracking Theory: This theory, popularized by philosopher Robert Nozick, replaces the requirement for justification with sensitivity and adherence requirements. Sensitivity means that in the closest possible world where the belief were false, you wouldn't believe it, and adherence means that in the closest possible world where the belief were true, you would still believe it. The clock belief fails the sensitivity condition since, in the closest possible world where it is false that it's 2:00, the clock would still say that it's 2:00, so you would believe that that is the time, falsely.
The Causal Theory: This theory was proposed by Alvin Goldman and says that you can only know something if the factors that make the belief true in some way cause you to believe it. The clock belief fails this requirement because the fact that the clock said 2:00 was completely unrelated to the fact that it was 2:00 at the time, so your belief that it was 2:00, which arose due to the clock, is equally unrelated to that fact. However, if the clock were working, then the fact that it was 2:00 would have caused the clock to say that it was 2:00, and your belief would therefore be caused by the fact that it was indeed 2:00.
The Defeasibility Condition: This adds another requirement for knowledge, similar to the NFL condition, except the defeasibility condition is that there must be no true statements ("defeaters") that would undermine one's justification for that belief, regardless of if you are aware of them or not. For instance, the clock belief is not knowledge, because you are not aware of the true statement that the clock is not working.
Biting the Bullet: This is where you simply admit that the clock belief is actually knowledge. This is not a solution to the problem, but rather a "dissolution" of it. That means that this doesn't attempt to solve the problem, but instead says that it is not a problem at all.

But there is still one problem. That problem is, how do we decide which of these theories is correct? How do we decide what knowledge actually is?

The Problem of the Criterion

Another name for that problem is the problem of the criterion. More formally, it's based on the relationship between two questions:

What do we know (what beliefs count as knowledge)?
How do we know (how can we determine what beliefs count as knowledge)?

The answer to the first question seems to assume that we know the answer to the second question, since we can't figure out what counts as knowledge if we don't know how to do that. But the answer to the second question also seems to assume that we can answer the first one, since we would need some examples of knowledge to determine what the rule that we can use to figure out what counts as knowledge is. So, we're basically going around in circles, and can't get anywhere. Or can we?

There are three main solutions to this problem:

Methodism: Answer the second question first. We start with the conditions for what knowledge is, and we use that to find examples of it.
Particularism: Answer the first question first. Figure out examples of knowledge, and then use those to figure out what the rule is.
Skepticism: Don't answer either question, and instead come to the conclusion that we can't actually know anything.

There is no consensus as to which solution is right, but almost all philosophers agree that the last solution is wrong.