Ben asked:

What is knowledge?

============

Justified true belief is a standard and reasonably widely (but not universally) accepted definition for
knowledge — there are some problems with it. Before I go into that, though, I need to clear up what is
meant by saying that knowledge is a particular type of belief — one which is both justified and true.
Remember, it is a definition of 'knowledge', not of 'truth'.

Belief: For me to be said to know something, I need to believe it. If I don't believe that snow is white, I
can't be said to know that snow is white.

True: It is not enough for something to be knowledge just because it is believed. It also has to be true,
independent of any belief I have. It has to actually be a fact, in the world. So if I believe that
Shakespeare wrote Hamlet, and (see next condition) I have very solid grounds for believing it (lots of
books say so etc), but it turns out that Bacon wrote it instead, then I never knewthat Shakespeare
wrote Hamlet. I just, mistakenly, thought that I knew it.

Justified: However, even if I believe something that is also true, if my justification for that belief is not
adequate (and we can question what counts as adequate justification), then I can't be said to know it.
Say I claim to know that you were eating a ham sandwich as you asked your question, and say that it
turns out that you were, in fact, eating a ham sandwich as you sent it. But if I am asked 'how did you
know?' and I say 'because I am eating a ham sandwich as I write the reply', then we would say that
this was not a justification for believing you were eating one as you wrote, and that my "knowledge"
was no more than a lucky guess, or a coincidence. Knowledge can't turn out to be true just by
accident — there must be a good reason for holding it.

One thing to emphasise (on this account) is the distinction between knowledge and truth. Knowledge
is something that depends on people and their beliefs, whereas truth seems to be something that
does not depend on people at all, just on what really is. This is often referred to as the distinction
between epistemology (what is known) and ontology (what is). Things can be true without us being
able to say they are true, and we can say they are true without them being true. This takes care of the
case where there is disagreement about truth: if people disagree about what is true, then there is a
matter of fact (independent of belief or justification) about what is true — and only the person whose
belief (with justification) aligns with truth has knowledge — the other has a justified but false belief
(unless both are wrong!).

Example: 'I believe Father Christmas exists' is a statement of belief, not of knowledge. On the
"justified true belief" account, it is also knowledge if and only if (a) it can be justified and (b) it is
(independently) also true.

Some people (me included) think that there are problems with the definition of knowledge as justified
true belief. One reason (the one I share) is the problem about what is true. We don't seem to have
any way of checking whether something is true, apart from the methods of forming justified beliefs.
So, if we can never independently determine what actually is true (as distinguished from what we
think is true), then we can never be sure we know anything (the problem of scepticism). Yet we do
claim to know lots of things — surely we do in fact know lots of things. So knowledge might better be
thought of as properly justified belief, and we can therefore know things that might, in theory, turn out
to have been wrong (a fallibilist view of knowledge). Of course, some people don't like this last
statement — they would say that if it turns out to have been wrong, we didn't know it at all.

Of course, there is disagreement about types of justification and how good each is — all knowledge
does seem to depend on lots of other things we know, ways we go about finding things out, our
beliefs about how the world works, what counts as justification and so on. So, many people think that
knowledge has to do with fitting in with all these other things — not just in ourselves, but in the
community we move in — these are coherence and consensus models of knowledge, and it is a
version of this that I would defend. But that is not to say that this is an easy account of knowledge to
defend — in fact, I am still working away at it.

Tim Sprod

The branch of philosophy dealing with this question is called epistemology. There are many theories
of knowledge, among them the coherence theory and the pragmatic theory. I'll try to explain one of
the "classical approaches" here, the justification theory of knowledge.

Knowledge must primarily be based on reason and evidence, rather than feeling or intuition.

Knowledge further requires:

1. Belief: I can only know that London is the capital of the UK if I (at least) believe that it is.

2. Truth: I also could believe something false: "Paris is the capital of the UK".

So, knowledge needs true belief based on evidence. Still this is not necessarily knowledge. An
example: in ancient Greece a few people were heliocentrists. They believed, that the earth revolves
around the sun (which turned out to be a true belief), they had reasons for their belief, but not enough
evidence to know that the earth went around the sun: at that time it seemed more evident, that the
sun revolved around the earth.

It was thought that justification, when added to true belief, yields a necessary and sufficient condition
for knowledge. Its sufficiency, however, was refuted by Edmund Gettier. He showed that having a
justified true belief still might be insufficient for knowledge.

An example: Suppose that Helen, one of my sisters, tells me that she is pregnant, on the grounds that
her pregnancy test at the clinic was positive. So I believe that one of my sisters is pregnant for a good
reason: my belief is justified. Further suppose that my belief is true, but not because Helen is
pregnant. There was a mix up at the clinic and not she is pregnant, but my other sister, Christine. My
belief was true and justified, but there was no knowledge.

Then, what more than justified true belief is required for knowledge? One answer is this. A belief
counts as knowledge only if it was acquired by a reliable method. A method for acquiring beliefs is
reliable just if it leads one to acquire beliefs which are true and does not lead one to acquire beliefs
which are false. Trusting hospital pregnancy tests is an example of what may seem, in most contexts,
a reliable method for acquiring beliefs. But in the above example, the context of the mix-up at the
hospital meant that it was not a reliable method. And this is why my true justified belief that one of my
sisters is pregnant does not count as knowledge. For a belief counts as knowledge only if it was
acquired by a method that was, in the context, reliable.

What, then, is knowledge? One answer is this: knowledge is true justified belief that was acquired by
a method that was, in the context, reliable. A subject's belief counts as knowledge when they have
good reason to have that belief, the belief is true, and it was acquired by a method that was, in the
context, reliable. That's a lot of conditions, isn't it?

That's why some people are dissatisfied with (these variations of) the justification theory of
knowledge, they say "If that's what knowledge is, then we have very little of it, if any!"

According to another approach to the question of knowledge, the causal theory of knowledge, we can
know something without personally having a proof or even justification of it. We have knowledge that
something is the case whenever our belief is caused in the right way: A subject's belief counts as
knowledge if and only if it is caused by that which makes it true. I know that it is raining if and only if
my belief that it is raining is caused by that which makes it true — its raining.

Simone Klein