To: Tony S.
From: Geoffrey Klempner
Subject: Pythagorean theory of reincarnation
Date: 10 April 2003 11:07

Dear Tony,

Thank you for your e-mail of 31 March, with your second essay for the Ancient Philosophy program, in response to the question, 'Critically discuss the Pythagorean theory of reincarnation. What issues does the idea of reincarnation raise concerning the nature of the self, and the definition of personal identity?'

There are two main issues to discuss here: whether reincarnation is inconsistent with the preservation of personal identity; and whether it is possible to have a form of reincarnation without personal identity.

It could be argued that what we are really concerned with is the identity of the subject, conceived as the ultimate nugget of individuality, whether the 'personality' is preserved in some way or not. This is how I understand the way 'Atman' is conceived in Buddhist philosophy. To illustrate this, in the philosophy of mind program there is an example of a drunk, 'Dirty Dick' whose memory of his former existence as a company executive Richard Bull has been obliterated by alcohol. There is still the physical continuity, but why should that interest us? The temptation is to think of individuality as detachable from all that the subject is consciously or subconsciously aware of, as something separate from the contents of consciousness in the way that a jar is separate from the jam inside. Deep inside, we imagine, there is the same *point of view*, the same mental eye looking out onto the world.

If this concept of a contentless point of view or spark of individuality could be sustained (I don't think it can) then we have all we need for a theory of reincarnation. It is not necessary for Pythagoras to remember who he was in previous lives. For even if there remains not a trace of memory of Euphorbus, the spark, the nugget of individuality which is now in Pythagoras was once in Euphorbus, whether anyone can ever know this or not.

In fact, clearly, no-one can ever know this. Because even if Pythagoras 'remembers' being Euphorbus, it is still possible for this memory to be false. In terms of the vital spark theory, nothing could count as verifying or falsifying a claim of identity in these circumstances. However, that does not show the theory to be false, unless you subscribe to a verificationist theory of meaning.

Interestingly, you raise as a problem for the memory criterion of personal identity, the example of 'Pythagoras with his whipped friend'. Suppose we reject the vital spark theory, which does not require memory, and go instead for a theory of personal identity which requires continuity of memory. There is, as I have already pointed out, a seemingly insurmountable problem of finding a criterion to distinguish between true and false memories. However, putting that aside, I don't see that it is a problem that the puppy cannot tell us what it 'remembers'. We still seem to understand what it is to have knowledge, yet be unable to communicate that knowledge. Imagine, for example, being paralysed by a stroke.

The problem of distinguishing true and false memory is the main objection to Pythagoras theory, as an account of reincarnation which preserves personal identity. But the objection is not fatal to all forms of reincarnation. This is where we enter interesting territory.

Arguably what makes the difference between a genuine memory and a false one is a causal link. If I seem to remember that I locked the back door last night, and the door was locked last night it doesn't follow that my memory is true. For example (as often happens) my wife locked the door and told me. It genuinely seems to me as if I was the one who locked the door, but in this case my memory has deceived me.

There is, of course, a causal link between the locking of the door and my apparent memory, namely, being given this information. But it is the wrong kind. Another example of the wrong kind of causal link is if someone found Bertrand Russell's brain miraculously preserved in a jar and injected some of the cells into my brain, so that I gained some of Russell's memories. That still would not make it true that I wrote 'The Problems of Philosophy' even if I seem to have the memory of having done so.

Is it necessary to have the same body in order to be the same person? Is that what it takes for the right kind of causal link between an action and one's memory of performing that action at a later time? If, as some AI theorists have argued, a human brain is ultimately nothing more than a computer running a program, then there could be such a thing as downloading the program onto disc and uploading it into another brain. Through that process, I could *be* Russell, or Hitler. Indeed (and this is an unavoidable consequence which all but wrecks the theory as a theory of identity) there could be hundreds of 'genuine' Russells and Hitlers.

As I point out in the unit, the relevance of these sci-fi speculations to Pythagoras is that he appears to have conceived of the soul as essentially a mathematical tuning or harmony, and this is just the image that AI theorists employ.

All the best,

Geoffrey