To: Mark S.
From: Geoffrey Klempner
Subject: Coherentist theory of epistemic justification
Date: 9 November 2007 14:14

Dear Mark,

Thank you for your email of 2 November, with your University of London Epistemology essay in response to the question, ''The fact that your beliefs can form a coherent set and yet be false shows that coherence cannot suffice for justification.' - Discuss.'

This is an excellent piece of work, which demonstrates considerable knowledge of this subject area. However, it is not so commendable as an answer to the question.

I accept that you need to look at all aspects of a topic in order to be sufficiently prepared for an exam. However, part of the reason for writing me essays is to focus on what is required to give the best answer to a question in the examination. The examiner is not just testing your knowledge: you are also being tested on your ability to respond to the challenge presented by a specific question in the most relevant and persuasive way.

I apologize if I seem to be labouring this point. But it is one that comes up regularly with my UoL mentees. The best solution, as other students seem to have discovered, is to write (for your own benefit) an essay or report which surveys the problem area, then write a second essay (for me) which answers a specific question. This is the best way I can help you.

Considering the question you have been asked, the first thing one needs to consider is whether, given that we accept that a belief can be 'justified' but false, we do in fact accept, as a matter of principle, that a justification which is false -- or contains something false -- cannot be considered a 'justification'.

This seems pretty difficult to deny. It is hardly necessary to go into specific examples to show that if you believe something on false assumptions, your belief is not justified. But what if the assumptions hang together and only some of them (the minority) are false? (as is indeed often the case). Then your justification is flawed but still adequate.

So the fact that the belief P that we are seeking to justify might be false is not an objection. We are only considering whether we are justified in believing it, not whether it is true or whether we know it to be true. Nor is the fact that some of the beliefs in the justifying set are false sufficient to undermine the claim that the set as a whole justifies the belief that P, provided that sufficiently many beliefs in the justifying set are true.

So the next thing to consider is whether, indeed, it is possible for a person's entire corpus of beliefs to be false, as the question seems to imply. This sounds incredibly unlikely. What does seem possible is that sufficiently many might be false to undermine the justification for the given belief P.

A practical example of this would be paranoid delusions, where a large number of beliefs hang together, exhibit strong coherence and mutual explanatory power, and yet the result is that they 'justify' a belief which is completely insane. Or perhaps a society where everyone has extremely weird, irrational (from our point of view) beliefs about the world. Like the Azande tribe who believe (allegedly) that they carry their souls in a stick.

Prima facie, this doesn't look like a very persuasive criticism of a sufficiency coherentist theory of justification. Obviously, the coherentist will say, the person suffering paranoid delusions has a deep incoherence in his total belief system. In order to maintain the tottering structure of his paranoid delusion, it is necessary to embrace patent absurdities (which human beings are, admittedly, very good at doing).

The second case is harder, because it raises general issues about rationality. But even here, there seems no good argument against the common sense view that a large group of people can reinforce one another's irrational beliefs, and that this is shown by a crucial failure of coherence at some point.

The key weapon in the coherentist's defence is the role of experience. However, it is also a serious potential weakness (as you point out) because it seems to remain a pure coherentist while accepting that failure to cohere with what one sees and feels and hears is a pretty devastating indictment of a system of beliefs, however coherent it might be in other respects.

Nevertheless, the coherentist will say, this observation does not force us to embrace the 'myth of the given' or become semi-foundationalists. It is an uncontested truism that all our information about the world ultimately comes our senses. However, as good coherentists, we accept that no report is immune from challenge.

...And so on. I'm not trying to write a model answer, but just give an example of the kind of answer that I believe the examiner is looking for: one which takes the question, if necessary challenges the assumptions behind the question, and seeks to construct the most economical case which either defends the assertion in question, or refutes it without adding any extraneous information.

All the best,

Geoffrey