To: Anna H.
From: Geoffrey Klempner
Subject: Coherence theory of knowledge
Date: 16th February 2010 12:22

Dear Anna,

Thank you for your email of 5 February, with your essay for the University of London BA Epistemology module, in response to the question, 'Critically assess the claim that if you believe that p and p coheres with everything else that you believe, then you know that p.'

Note that the correct thing to say is, 'Tom believes that p,' and not, 'Tom believes that 'p'.' Quotes are used to indicate direct discourse, as in, 'Tom said, 'it is sunny today',' while indirect discourse dispenses with quotes, as in, 'Tom said that it is sunny today.'

This essay is a good length for an examination answer, so if you can write this much in an hour, then you would be doing well.

In your essay, you raise a number of valid points regarding the correspondence theory of knowledge. However, the most important point doesn't get a mention, so I will discuss this first.

You start off by alluding to the discussion of knowledge and belief by Plato in the Theaetetus. Although the passage you quote leaves the answer unresolved (210 a-b), by taking this dialogue together with the Meno, Plato is generally credited with the definition of knowledge as 'true belief with an account' or 'justified true belief'.

What is so noticeable about the coherence definition of knowledge quoted in the question, is that neither truth nor justification is mentioned. You go on to say that 'a belief is justified 'if and only if' it coheres with a system of beliefs, but you don't say anything about truth.

If all coherence does is provide justification for a belief, then surely the question will arise whether you can have a coherent set of beliefs, many or the majority of which are false. That is a characteristic of persons suffering from paranoid delusions. However, we can also travel back in time to when rational, knowledgeable people 'knew' that the earth was flat, or that the sun goes round the earth. How can you 'know' something that isn't true?

With this question in mind, it makes more sense to raise the issue that you raise, concerning the threat of infinite regress or circularity. Why do we seek justification for our beliefs? In order to have the highest possible assurance that they are indeed true. However, if A justifies B and B justifies C... and so on up to Z, and Z justifies A, then it would be perfectly possible for all the beliefs in the 'coherent set' to be false.

The objection that 'a coherent set can be false' is usually levelled against the coherence theory of truth, but it is still telling here.

So far as the coherence theory of knowledge is concerned, there is a response: some coherent sets are better than others, the goal being, as you say, a 'maximum of explanatory coherence'. As one approaches the maximum, one might suppose, the chance that a large number of beliefs are false is, one hopes, reduced.

However, as you also note, another factor to take into consideration is experience. It is much harder to hold a 'false' coherent set of beliefs in the face of contradictory experience. Here, you need to note that coherence theorists differ in how they deal with experience. To suggest that experience is an external check on a system of beliefs would be to fall into the illusion of 'the myth of the given', as Sellars calls it. We never encounter 'raw experience' as such, but rather make experiential judgements, which are sensitive to our prior beliefs. So experience is important, but it must be viewed as an intrinsic part of the coherent structure, not extrinsic.

You go on to quote the foundational account. That there is an alternative account is not, as such, a criticism of the coherence theory of knowledge. If I am criticizing your theory, whatever theory that may be, it is no good my saying that I have a different theory. I have to pick holes on your theory in order to justify my preference for my theory!

What you say towards the end of your essay, with regard to the question of seeing an apple, does leave me in some doubt as to whether you are arguing for or against the coherence theory of knowledge. As noted above, coherence theorists do not regard it as acceptable to allow the factor of experience to be extrinsic to the coherent set of beliefs -- the result would no longer be a coherence theory of knowledge but rather a foundational theory of knowledge. So how is experience special?

I give you an apple, and you are in no doubt that it is an apple. It looks and feels like an apple, it has an apple smell. However, I am a practical joker who loves to give people fake apples then see their surprise when they try to bite them. What are you going to say at this point? That at least you can see something round and green? Would you be prepared to redefine perception in terms of judgements made about sense data? The objections to that idea are every bit as strong as the objections made against the coherence theory.

All the best,

Geoffrey