To: Brent S.
From: Geoffrey Klempner
Subject: Knowledge, belief, rationality and truth
Date: 16th December 2009 12:05

Dear Brent,

Thank you for your email of 8 December, with your essay for the University of London Epistemology module, entitled, 'A brief summary of the nature of knowledge'.

The task you have set yourself -- summarising what you have learned about the issues around belief, truth, rationality and knowledge -- is very modest. Even so, there are errors/ unclarities in what you have said.

What is true belief? You say, 'In epistemology the notion of TB is simply recognized as any cognitive content an agent may hold of which it is believed by that agent to be true.' But earlier, you said, 'It is not necessary for a belief to be true. One can believe that the moon is made from cheese, but this doesn't make it true.'

Let's say Tim believes that the moon is made of cheese. Tim believes that his belief is true. But the belief is not true. Tim is wrong. But suppose Tim says, 'I believe that the moon is made of cheese. Of course, it is not true that the moon is made of cheese, but I believe it anyway.' Does that make sense?

In what sense is it axiomatic that to believe that P is to believe that P is true? I can make myself cry when the circumstances are not appropriate. Why can't I make myself believe that P when the circumstances are not appropriate (because it is not true that P)?

You say, 'The significance of JTB being of more value than mere TB, is that we are in a better position to rule out instances of luck'. Is that the only reason? Plato gives a different explanation in Theaetetus: beliefs which lack 'an account' have a tendency to 'run away'. In other words, the problem Plato seems to be talking about is subjective, relating to the mind of the knower, rather than objective, i.e. concerning the degree of reliance that anyone might place on the belief in question. In other words, it is good to have 'an account' because you are less likely to change your mind later (=the belief 'running away'). Plato is talking about the sense of certainty rather than the probability of truth.

You mention Gettier, but there is nothing about responses to Gettier, so I have nothing to say about this.

You also look at epistemic rationality, which doesn't strictly come under the heading of nature or definition of knowledge. You are actually the second student recently to raise the 'problem' of trivial beliefs. I assume you were both reading the same epistemology text. I am at a loss to understand how this could even look like a problem, as you have posed it.

Of course we want to ensure that as many of our beliefs as possible are true rather than false. But it would be absurd to equate this with 'having as many true beliefs as possible'. What is rationality? There are competing definitions. The question is especially problematic when we consider moral rationality by contrast with prudential rationality. But let's stick with prudential rationality. It is prudentially rational to apply a cost-benefit analysis to proposed courses of action. Included in this analysis, of course, are the potential costs and benefits of prolonging the amount of time you give for deliberation or shortening it. Some situations demand immediate action. Among the courses of action that one may apply a cost-benefit to, are investigating the world or looking for justification for our beliefs.

True beliefs are useful. Why? Why is it useful for me to know that Sirius is approximately four light years away? Well, one day I might be lucky and the question will come up on University Challenge (say, I keep a score of the number of questions I am able to answer each week). Or, one day, I may need to leave Earth and travel to the nearest star, and it would be useful to know how much rocket fuel to put in my space ship.

Well, no. A cost-benefit analysis says that we ought to gain as many true beliefs as possible for the purposes of possible action. If I don't want to spend too much time waiting at a bus stop in the cold wind and rain, it's a good idea to know when my bus is due to come. On the other hand, it would be potentially useful to know the entire bus timetable, but the cost in terms of effort of memorisation would outweigh the benefit.

My other student also explained 'deontic' rationality. But he came up with the objection that accepted norms of justification which are internalised by the subject might not be valid/ justified from the point of view of a more knowledgeable observer, which seemed plausible to me.

Is knowledge valuable? We've established that true belief is valuable (for the purposes of action). We've also established that if you have 'an account', then you are likely to be more confident in your belief (Plato) and that, as a rule, justified beliefs have a greater probability of being true than unjustified beliefs. But we haven't said yet what knowledge IS. That was the point of responding to Gettier.

It makes sense to say, 'Is knowledge valuable?' or 'Why is knowledge valuable?' if you have a given definition of knowledge (=response to Gettier). Then you can ask, of the definition in question (call it K) 'Is K valuable?' or 'Why is K valuable?'. In fact, one can go a step further and say that it is a condition on the adequacy of a definition of knowledge that it has the consequence that knowledge is valuable. (This bears a very superficial similarity to Tarski's convention T: It is a condition on the adequacy of a definition of truth that it has the consequence that 'P' is T if and only if P for every sentence P of the language.)

I would really like you to do more. I would like you to pick exam questions rather than write on a general topic as you have done here. You will learn more about how to write a good essay (i.e. one that will score good marks in an exam) and also give me more material to critique (rather than leave me to spin a lot out of relatively little as I've done here).

Also, I would like you to aim for a word count of 2000-2500 words. The aim of that is to make you 'dig deeper'.

All the best,

Geoffrey