To: Stuart B.
From: Geoffrey Klempner
Subject: Plato's Republic and the tripartite theory of the soul
Date: 14th October 2008

Dear Mark,

Thank you for your email of 6 October, with your essay in response to the University of London Greek Philosophy: Pre-Socratics and Plato question, 'Does Plato, in the Republic, offer a good logical ground for thinking that the soul has three parts?'

Right at the start, you declare what you understand by a 'good logical ground' for making a conceptual distinction. This is coloured to a considerable extent by the model of Linnaean classification which you cite. On this view, classifying a phenomenon is a matter of 'carving at the natural joints' (a phrase from Plato's 'Phaedrus') which is what the dialectician aims to do.

However, your example from Kant's 'Jasche Logic', that 'to be logically grounded an argument must (a) have grounds and (b) not have false consequences is not altogether helpful. It is true that if one can ever find a counterexample (true premisses, false conclusion) to an *argument*, then we know without examining the structure of the argument that it must be invalid. A valid argument always leads from true premisses to a true conclusion. We also know that a valid argument is sound (i.e. a reason for believing the truth of the conclusion) if its premisses are true.

The difficulty I have is that there seems to be a gap between the idea of a logically consistent classification which can be used for this or that purpose, and an attempt to dialectically determine the ultimate reality of the phenomenon or entity in question. I take it that Plato in Republic is seeking to determine the 'deep logical structure' of the psyche (leaving on one side, as you note, the question discussed in the Phaedo concerning the relation of the psyche to the physical world). His use of the 'Principle of Contraries' is intended to reveal the conceptually necessary cuts.

In terms of this project, the attempt to reduce the number of parts to two, or increase the parts to more than three, is not just less useful or illuminating but incoherent. Each of the three parts can be further divided as in Linnaean classification (as you observe) but any increase in parts is necessarily subordinate logically and conceptually to the primary cut.

The question, therefore, is whether Plato has proved his case for the necessity of a tripartite classification. If he hasn't then his scheme is not 'logically grounded'.

Consider the contemporary arguments over mind-body dualism. A 'dualist' is not someone who merely observes that it is sometimes useful and valid to talk about a distinction between an individual's mental and physical qualities, but rather one who insists that a material account of the mental is ultimately not possible -- for example, on the grounds that there cannot be an explanation couched in purely materialist terms of the phenomenon of 'qualia'.

Thus, the disagreement between Plato and a proponent of a dual view of psyche as governed by 'reason and passion', or between Plato and a Humean who claims that the passions are the only thing with the power to move an agent to action, is one that needs to be resolved. Only one can be right. Ultimately, only one can be 'logically consistent'.

Much of what you say is consistent with this approach: for example, your criticism of Hume's argument, 'Reason alone can never produce any action...'. An additional point to make is that we can't always expect these issues to be resolved. Rival conceptual analyses can remain in play so long as we haven't found a knockdown argument which reveals one or more of the analyses as embodying inconsistency. But it is one thing to say that a conceptual distinction appears consistent, or that one cannot find any logical objection, and another to claim that it is in fact *correct and true*.

I think that the question equates the notion of being 'logically grounded' with correctness or truth, rather than merely with being defensible or consistent so far as one is able to tell. In that case, one possible answer to the question would be, 'None of the arguments I have considered succeed in showing that Plato's tripartite classification is not logically grounded.'

For what it's worth, my answer to the question would be, No. G.E.M. Anscombe in her seminal monograph 'Intention' (Blackwell 2nd ed. 1963) gives a powerful argument against the idea that we can be motivated or moved to action by appetites, defined merely as mental pushes or pulls. A desire is necessarily a reason for action. As such it already embodies a rational component. (So reason/ passion dualism and Hume are both wrong.)

Anscombe gives an example of someone who desires a saucer of mud. What would that mean? If desires are just mental pushes and pulls, then there is no reason in principle why a person can't desire a saucer of mud. You give them the saucer and they don't do anything with it. They have no explanation to give of why they want it or what it does for them. What we expect is an answer along the lines, e.g. 'I wanted to smell the rich river smell' (Anscombe's example), or 'I wanted to cover my face with mud as a camouflage'.

Human beings have natural appetites, which we recognize in language as things worthy of being desired -- thirst, hunger, etc. However, there are reasons for other things too -- like taking the desires of other persons into consideration, and conflicts naturally arise. But these conflicts are within the arena of reason, not a conflict between a separate faculty named 'reason' and something else.

A complicating factor here is that we know that we are members of the animal kingdom and that a dog or a bird can be 'hungry' or 'thirsty' just as we are. For Plato, this is evidence of our 'animal nature' and the idea that hunger qua hunger is the same, whether it is in a man, or a dog or a bird. The alternative view would be that in man, the 'rational animal' these phenomena are transformed acquiring an additional layer of meaning, precisely because we are able to cite them as 'reasons for action'.

All the best,

Geoffrey