To: Francis M.
From: Geoffrey Klempner
Subject: Is knowledge justified true belief?
Date: 29 September 2005

Dear Frank,

Thank you for your email of 18 September, with your University of London essay in response to the question, 'Is knowledge justified true belief?'

This is a very good piece of work which would be pushing towards a first (a mark in the 70's) in the exam. I would evaluate it as borderline.

The question about justified true belief is very well trodden ground, and to make an impression you have to show that you have at least attempted to gain a wider perspective on the moves and counter-moves.

Lehrer effectively does this in staking a claim about the 'goal of defining knowledge'. Asking about the goal or point of a definition is always a good thing to do when you are attempting a philosophical analysis. One question one might raise, however, is what is the assumption behind the idea of asking for the goal or point of a definition?

Naively, one might thing that we just have this concept of knowledge, and the aim of seeking a definition or analysis is to learn more about what that concept actually is. Whereas Lehrer's approach implies that everything is up for grabs, the aim is not so much to explain 'our' concept of knowledge - whatever that might be - but rather to create one, more or less from scratch. 'What would be a good concept of knowledge? Whatever that is, I'll take it, thank you very much.'

Along similar lines, I have seen one author approach the problem of free will by posing the question of 'Varieties of free will worth wanting.'

This implies that accepted linguistic usage is not adequate, on its own, as the ultimate court of appeal in questions of philosophical analysis. Philosophers are responsible for moulding and refining concepts, replacing vague or partially incoherent concepts with concepts which are sharper, clearer, ultimately more useful.

One might be forgiven for thinking that Gettier could have saved himself the trouble of writing his article if only he had read Russell. In deciding whether A knows that P, we need to look at each the beliefs involved in the justification which A would give for his belief that P. Provided that there are no false beliefs somewhere along the way - and of course provided P is true - then P counts as knowledge.

The problem with this as a recipe for 'obtaining truth and avoiding error' is that this is a recipe for perpetual suspension of judgement concerning whether someone *really* knows that something is the case or not. It is impossible in practice to investigate all the implicit assumptions behind the judgement that P is the case. In Gettier terms, you have to consider all the extravagant and improbable ways in which it might turn out that the connection between the truth of what is believed and the believer's justification is merely accidental.

Arguably, this is a problem with any of the refinements proposed to the 'justified true belief' definition. As soon as you try to plug the gaps created by Gettier, the spectre of scepticism looms.

By contrast, the 'justified true belief' definition is perfectly workable in practice. Is P true? Yes. Can A give a plausible justification for his belief? Yes. That's good enough for most purposes. Most of the time, following that formula, we will indeed obtain knowledge and avoid error with respect to claims about what this or that person knows.

As soon as one tries to plug the gaps, however, everything is thrown into confusion.

In other words, in responding to the genuine concerns raised by Gettier to the naive cut-and-dried definition of knowledge as justified true belief, we find ourselves pushed in the direction of a rather more questionable definition of knowledge - questionable because of the way that it resurrects the spectre of scepticism.

Obviously, what you want to do in these circumstances is not simply retreat to the naive definition of knowledge, but rather combine the 'best' response to Gettier with a robust defence against scepticism. I have my own views about this, but that's another story.

I think that, if you were looking for ways in which your essay could be further improved, then the scepticism issue would certainly be one valid issue to explore.

Another, related point concerns the 'third person' nature of the definitions of knowledge, whether we are talking of the original naive version, or the various Gettier-inspired versions. The assumption is that 'we' have knowledge, we are capable of deciding with confidence whether a given proposition is true, and the only question is how to decide whether or when to attribute knowledge to others. This relates to the debate between 'reliabilism' and 'foundationalism', another of the issues which you will be exploring for the Epistemology paper.

All the best,

Geoffrey