To: Julian P.
From: Geoffrey Klempner
Subject: Plato on knowledge and false belief
Date: 20 January 2005 12:54

Dear Julian,

Thank you for your email of 7 January with your University of London essay, 'Why does Plato struggle with the impossibility of false belief? How does he resolve this?', and your email of 18 January, with your essay, 'What things does Plato say we can know? How can we know them?'.

I have just read the first essay and will comment on before reading the second essay - which is what would have happened in the normal course of events.

Plato and false belief

This is a superb piece of work. From a stylistic point of view, your essay does have a slightly breathless feel - you were clearly conscious of the need to avoid excessive length (for which I am grateful!) but this is not an important issue.

Were you aware of the famous piece by the sophist Gorgias, 'On What Is Not'? The piece must have made an impression on Plato.

If any criticism can be made, it is that your treatment of the texts is so thorough that it doesn't leave much room for a broader view of the problem of false belief. It is important to distinguish two issues which come under the heading 'false belief' which are really different problems altogether. You show awareness of this, in your treatment of Plato's arguments in the Theaetetus and Sophist. But examiners like to have things spelled out.

The first problem is the one that most deserves association with the sophist issue. This is the problem of false judgement. The possibility of objective (non-relative) truth can be understood as the problem of giving an account of how false judgement is possible.

The second problem arises more directly from Parmenides, the need to explain logically how there can be such a thing as a false proposition.

The solution to the first problem requires the establishment of an appropriate distance between the knower and the objects of knowledge, a distance which is yet capable of being overcome (otherwise we would end up knowing nothing). The main criticism of naive idealism is that the idealist, in the desire to overcome scepticism, attempts to abolish the distance between the knower and the known (cf. Kant's Refutation of Idealism, Critique of Pure Reason 2nd edition, also Wittgenstein's private language argument).

Contemporary anti-realism (Michael Dummett, Crispin Wright) presents a different challenge from the idealist, but one which has a similar structure - the oscillation between scepticism and the impossibility of truth - and which in many ways fits the sophist case better.

Plato confusingly gives two solutions to the second problem. The first is readily recognizable as the recognition that language is structured in a certain way. It is not too anachronistic to compare the account of the blending of forms with Wittgenstein's theory of language in the Tractatus. The crucial question for Wittgenstein, as for Plato, is how a proposition differs from a name.

That ought to be sufficient. But Plato introduces another idea: that whenever we state a false proposition, we are saying something which is 'different from the truth'. He doesn't just mean the truistic point that there are two truth values, each of which is necessarily different from the other. The suggestion is that there is a more fundamental notion than negation, in terms of which negation can be analysed. (There is an excellent article on this by David Wiggins with a ridiculously long name - sorry I don't have the exact reference to hand, but it is in an Anchor paperback of articles on Plato edited by Gregory Vlastos).

This is mind-boggling stuff, and you will certainly gain extra marks for showing a keener awareness of the issue. (It is certainly worth an afternoon's thought.) You do say in your penultimate paragraph, 'Nor does he give an account of true negation: 'Theaetetus is not flying' which would is puzzling given that the subject-predicate theory is all that is required from a logical point of view, but less puzzling from the point of view of the negation-as-difference theory. This needs to be spelled out a lot more. It is not inconceivable that this could be the main focus of an exam question on the Sophist.

Plato and the objects of knowledge

This essay reads more like a summary of notes, and to that extent is less successful than the previous one. I didn't get a strong enough sense of the problem of knowledge as Plato sees it.

I think you will be in a better position to answer this question after you have done more work on epistemology. Russell's point, that a true belief is not knowledge if deduced from a false proposition, seems to have been overlooked in discussions of the nature of knowledge until Gettier came along ('Is Knowledge Justified True Belief?'). Gettier give some compelling examples of true belief plus an 'account' which turn out not to be knowledge. (It was a student paper which made his career as a philosopher.)

In response to Gettier, Gilbert Harman, in his book 'Thought' (Princeton 1973) defines knowledge as true belief which 'doesn't depend on anything false', which seems to heark back to Russell. An impossibly exacting requirement, unless we restrict what is to count as 'dependence'. Every attempt to deal with this challenge which doesn't throw in the towel and embrace scepticism has to compromise at some point, as you will discover. Obviously, one is not saying that Plato saw all this. But it was a wise decision to stay out of the mud and concentrate on higher things!

The road to Larissa example shows the deep problem for any account of knowledge as 'true belief plus an account'. Any account that you can give is just another belief or set of beliefs, so the question arises what makes *that* knowledge. Suppose you walk the road to Larissa. It is not clear how that helps, in the face of the challenge how you know that your memory is reliable, or that there haven't been road works in the meantime, or etc.

The slave boy example shows knowledge arrived at by a method of proof, a priori. Plato's solution to Meno's paradox is to show how you can 'know and not know something' in terms of the native capacity to follow a proof and the exercise of that capacity. This still doesn't help when we are concerned with empirical knowledge.

The Phaedo gives an important clue, in Plato's account of Formal causation (which became one of Aristotle's 'four causes', the others being efficient, material and teleological). 'Theaetetus is clever because that's just what cleverness is', is an answer you would give to someone who persisted with the question, 'Is Theaetetus clever?' after you had shown them his breadth of knowledge, his amazing feats of deduction etc. etc. The form of Clever is the formal cause of Theaetetuses being clever. So what? As you recognize in the latter part of your essay, there are many cases where our primary concern is this kind of question. What is virtue, knowledge, beauty etc.?

The Russell/Gettier problem doesn't arise for 'formal' knowledge - or knowledge gained through the process of dialectic - in the same way that it arises for empirical knowledge. The search for a complete 'account' has an ultimate terminus (even if no-one has actually reached it yet). (It is worth making the point that science hadn't happened yet, so it wasn't difficult for Socrates in the Phaedo to pour scorn on his former interest in the speculations of the Presocratics.)

Plato was, as you acknowledge, is hampered by his failure to give a clear account of the difference between knowledge by acquaintance and knowledge by description. But this is not a story about how Plato failed to give an intelligible account of knowledge and belief because he didn't have this distinction readily at his disposal. The problems of empirical knowledge and the knowledge gained through the methods of philosophical inquiry are still with us today.

There is a lot of good work here. My criticism is that things just get too complex, with the result that the reader fails to see the wood for the trees.

All the best,

Geoffrey