To: Charles R.
From: Geoffrey Klempner
Subject: Xenophanes on human knowledge and its limits
Date: 13th May 2010 12:56

Dear Charles,

Thank you for your email of 7 May, with your second essay for the Ancient Philosophy program, entitled 'Reflections on the Question of Knowledge'.

Your essay is a useful and sensitive exploration of the limits of knowledge within a broadly scientific outlook.

We talk of what we 'know' in science, yet all the scientist can do is put forward the explanation which, at any given time, appears to be 'better' than any of the alternatives. As time goes on, explanations which were the 'best' available at the time are superseded as our capacity to put questions to nature develops and expands. 150 years ago, it could hardly have been predicted that what we term 'matter' would turn out to be mostly 'empty space'.

A point that needs to be stressed is that what I said in the unit about knowledge is an extrapolation from what Xenophanes actually said, with a strong emphasis (which you've noted) on the practical aspects of knowledge. This is what Plato does, e.g. in the Theaetetus and Meno when he discusses the question what it is that needs to be 'added' to belief to make it 'knowledge'. You go to an expert because you want reliable information, not guesses.

One thing that needs to be 'added' to make a belief knowledge is truth. Of course, whenever we make statements about what this or that person 'knows' we are assuming the truth of the knowledge claim in question. No-one has direct, infallible access to the truth. So a claim that 'A knows that P' can be defeated, either by evidence that A's belief that P does not meet the standard for knowledge, or by evidence that P is false.

This is no big deal. The fact that we don't have direct, infallible access to the truth doesn't make knowledge 'relative' or 'contextual'. Nor is it difficult to state what we mean by 'truth'. Aristotle's formula is still fully acceptable today: If I say X when X obtains then what I say is true. If I say X when X does not obtain then what I say is false.

Suppose I say, 'The last thought Aristotle had before he died was how much he loved his mother.' No human being can ever know this. But if I do make that statement, and IN FACT (unknown to us) the last thought Aristotle had before he died WAS how much he loved his mother, then my statement is true, and false otherwise.

In short, on this plausible view of truth, there are lots of truths which we do not and never will know, truths far removed from human perception or capacity for investigation, or evidence for which has been obliterated by the sands of time.

There are, of course, statements which we don't know how to evaluate because they are not simply 'factual' but depend on interpretation. If you are a Christian, does that mean you believe in the Virgin birth? Not according to the Bishop of Durham (who raised a storm of controversy a few years ago when he stated that the 'belief' in question should not be interpreted literally).

From the evidence of what he said, it seems that Xenophanes held a 'realist' view of truth, such as I have described. There is a truth about the earth and heavens and the ordering of the cosmos, but whatever that truth is, human beings cannot know. We can only theorise. In order to stress his point, Xenophanes resorted to a device which is still relied on today: God knows. We will never know Aristotle's last thought before he died but God does. We can never obtain a God's-eye view.

One philosopher (Michael Dummett) has gone so far as to suggest that the necessity for holding a realist view of truth is the best argument for the existence of God (he acknowledges his debt to Berkeley).

On a realist view, then, the truth is not as such 'contextual'. Our views may change over time, but the truth is whatever it is, regardless of what we believe.

On the other hand, it is quite plausible to argue that the very notion of 'knowledge' is a contextual notion. In order to understand the concept of knowledge, it is necessary to explore how this concept is used, what utility it has for us.

When do we say that someone knows something? One requirement, as explained, is truth. You have to agree with what they say. The other requirement concerns their right to say it. To state that someone 'knows' is to imply that, in some sense, that person may be regarded as a reliable source of information about the matter in question. We have a concept of knowledge because of our interest in testimony, as a source of information. The person in question has to be, in some sense, an 'authority' or 'expert' and not a mere guesser.

A humdrum example would be if I ask the other people standing in the bus queue when the next bus is coming. I'm not interested in guesses or estimates. I can do that. What I want to know is whether someone has, e.g. seen bus timetable, or has phoned the bus company information line on their mobile.

For the same reason, to be an eyewitness to an event gives one a special status. Much of what I 'know' about the world depends on the eyes and ears of other people.

The problem is that 'authority' is a difficult thing to pin down. There are many cases where we would normally say we 'know' something, but all it takes is a question and we realize that we don't know. 'Yes, I know. The next bus is at 9.25 am.' 'Do you know whether the bus drivers have not called a strike today?' 'I hadn't considered that.' 'Then you don't KNOW when the next bus is coming!'

If you take that to its logical conclusion, then no-one knows anything, because you can always find a defeating question. This looks like a paradox, or an argument for scepticism. But I think it is better seen as emphasizing the contextuality of knowledge claims.

All the best,

Geoffrey