Friday, June 7, 2013

Fairy tales and the coherence theory of knowledge

To: Ruy R.
From: Geoffrey Klempner
Subject: Fairy tales and the coherence theory of knowledge
Date: 10th March 2010 14:22

Dear Ruy,

Thank you for your email of 3 March, with your essay for the University of London Epistemology BA module, in response to the question, 'The fact that there are coherent fairy tales shows that one might have a coherent system of beliefs without those beliefs amounting to knowledge.' Discuss.

As a rule, when answering a question you should aim to give as few 'hostages to fortune' as possible. By that I mean that you should not set out to defend *more* than the question requires you to, or make more assumptions than are needed in order to construct a convincing case.

In the question, the statement, 'there are coherent fairy tales' could be taken either as a challenge to a coherence theory of justification or as a challenge to a coherence theory of truth. However, the question goes on to say that what this shows is 'one might have a coherent system of beliefs without those beliefs amounting to knowledge'. In other words, you are being asked to defend a coherence theory of justification. You are not being asked to attempt the harder task of defending a coherence theory of truth!

You might think that you would gain extra credit from the examiner by (successfully, as you hope) giving more than the question asks for. However, the examiner will very likely conclude that you don't sufficiently understand the distinction between a coherence theory of justification and a coherence theory of truth. Also, you only have one hour in an examination, so you should not attempt to write two essays when you should be writing only one.

What is justification and why do we prize it? Justification isn't the same as truth. A belief can be justified but false. However, the fact that a belief 'happens to be true not false' is no good to us if we have no confidence in that belief. Justification provides the basis for that confidence. As Plato states in the Meno, beliefs which lack an adequate 'account' have a tendency to 'run away'.

You make some good points about the advantage that a coherence view of justification has over the traditional foundationalist view. In particular, I liked your statement, 'knowledge seems to be a process that grows dialectically between evidence and our intellectual construction', and especially the addendum, 'as I get acquainted with new evidence or new propositions that increase the coherence of the whole, the system debugs.'

Although I am not a programmer, I find the metaphor of 'debugging' very fruitful, because it implies a process of testing which approaches asymptotically towards a goal. You test the program in every way you can, and wait for bugs to show themselves. Sometimes fixing bugs produces bugs elsewhere and the whole process has to be gone through again. In a similar way, my coherent set of beliefs comes under strain from various sources and adjustments have to be made. Confidence increases as the system overcomes successive challenges.

But now the question arises: could such a system in reality be on a road to nowhere? Is it possible that because of some fatal failure to 'reality check' or 'debug', I am merely constructing an ever-more elaborate fairy tale? The analogy would be with programs whose basic design is fatally flawed, so that they can never be successfully debugged. In the real world, persons who are diagnosed as suffering paranoid delusions would be an example where apparent coherence of beliefs exists side by side with loss of connection to reality.

Having criticized you for attempting to defend a coherence theory of truth, it does occur to me at this point that a coherence theory of justification might gain support from a coherence theory of truth. I suppose that was your idea. The problem is that a pure, or genuine coherence theory of truth is very hard to defend. All that you say about the difficulties in establishing whether a proposition is 'really' true doesn't advance the case for a coherence theory of truth. On any theory, you have to allow that a belief which you thought was true, turns out to be false.

You admit, at the end of your essay, that defining truth in terms of coherence has the consequence of 'downgrading truth to a human construction and, who knows? Perhaps it is.' How far do you envisage this downgrading will go? It was once widely believed that the earth was flat. Was it true that the earth was flat? Did the flat earthers inhabit a 'flat earth' of their construction? Or did they inhabit a (roughly) spherical earth, believing falsely that it was flat?

My own view, for what it is worth, would be that there isn't much to say about truth, not because there is no deep metaphysical problem about how thought connects to reality, but rather because we need a simple, 'minimalist' account of truth in order to state coherently what that problem is. To use your phrase, there is a 'dialectic' to be worked through, the dialectic of 'realism' versus 'anti-realism' about truth.

All the best,

Geoffrey