Political Theory And Epistemology: Quine, Habermas, And Cohen

Yesterday, I ended by claiming that a particular kind of challenge to Rawlsian theory, the one which claimed that Rawlsian theory fixes the boundaries between the political and the moral too firmly, too decisively, also has implications for the nature of validity claims in political theory, amongst other things. This is because if the boundaries aren't fixed, then the proper content of the political - and indeed the proper content of the moral - is not fixed either, and on a lot of understandings of the meaning of the predicate 'truth', claims whose validity varies like that just aren't true: something doesn't become not true just because we change our minds about it, yet the implication of the idea that the proper content of the political is not fixed is just that if we change our minds about it, then it changes. Rawls himself begins to acknowledge something like this when he says that reasonableness rather than truth is the proper validity claim to be applied to his theory in 'Political Liberalism', something for which he got, and continues to get, no little stick. How fair that stick is, more or less, is what I'll be writing about today: it's going to get epistemological, for which I apologise.

I'll begin by talking about probably the most impressive attack on the Rawlsian abjuring from truth, G. A. Cohen's 'Facts and Principles' (I'm taught by him, alright: I've been exposed to his ideas more than anyone else's apart from Rawls's, so he will inevitably figure in a lot of debates I'm interested in). 'Facts and Principles' attempts to show that any fact which we think supports any normative principle does so in light of another normative principle, the implication being that, if normative principles are grounded, they are grounded by some ultimate normative principle, which is self-evident, a priori, something like that. This is an attack on Rawls's abjuring from truth because it is an attack on his decision procedure, which utilises some relatively general facts, and its claim that it produces principles of justice, and indeed an attack on all decision procedures as creating, rather than revealing, principles of justice: principles of justice exist prior to such procedures, and are true if they are valid. I think that Cohen is wrong, partly because I find the quasi-Arendtian conception of politics in Rawls persuasive, and partly for prior epistemological reasons. Now, I've already talked about the quasi-Arendtian conception of politics, so to avoid repeating myself, I'm going to talk about the prior epistemological reasons.

Since it is where my resistance to Cohen begins, it's probably best to begin with Quine's 'Two Dogmas Of Empiricism'. In this paper from the fifties, Quine argues that the distinction between the analytic and the synthetic - the true by definition and the true by virtue of facts in the world, roughly, which are the two dogmas in question - doesn't hold, basically because it is impossible, without vicious circularity, to specify what it is that is meant by true by definition. Everything is both a matter of definition and the empirical evidence we have, even things like the law of the excluded middle, that things must either be true or false, and so there are situations - perhaps not situations we can imagine - in which an apparently unavoidable logical truth like the law of the excluded middle isn't true, because it is partly dependent on the interpretation of the empirical evidence we find most persuasive, gven our other beliefs. Truth then becomes, effectively, rather than a property of statements considered in isolation, a property of statements as part of system of belief: in the case of the law of excluded middle, it is because, in light of everything else we believe, it makes sense that is is true.

Now, it might seem totally ridiculous to think that something as fundamental as the idea that statements are either true or false is dependent on it being compatible with the best interpretation of the rest of our beliefs, and so Quine must be wrong. But the thought that it is totally ridiculous to doubt the truth of the law of the excluded middle is perfectly compatible with Quinean epistemology, because it is such a key part of our system of belief that calling it into to doubt would call more or less every other belief we have into question, and since the truth of our beliefs is supposed to depend on their fit with our other beliefs, given the excellent fit of that particular belief with all our other beliefs, it would be ridiculous to doubt it, given those other beliefs. This is not, though, to say that it would not be ridiculous to doubt it given some other set of beliefs, and because the wisdom of doubting one particular belief is relative to the set of beliefs as a whole, we can't make the claim, on the basis of our current beliefs, that it would never be ridiculous to doubt the law of the excluded middle.

I think there's another way of putting Quine's point, which Habermas has used. The two dogmas which Quine attacks - the idea of the synthetic and of the analytic - are supposed to represent empirical and a priori knowledge, knowledge gained through our experiences and knowledge gained through pure reason. We could see these forms of knowledge as knowledge typical of humans as embedded in the world and of humans as external to the world respectively, as an in-itself and a for-itself, as object and subject. Part of Habermas's critique of much contemporary European is based on a critique of the adoption of these two standpoints as mutually undermining, and leading to contradiction and confusion. The idea is that when we look at the world as a for-itself, it becomes very difficult to understand how we gain access to it, how it is that we can verify that this is actually how things are in the world, since we are separated, distant, from it. Conversely, though, when we look at the world as an in-itself, because we are so thoroughly attached to it, although this is undoubtedly how the world appears to us, we have no external point of view from which to verify that this is how it is, no point from which the contingencies of the world might not be distorting our perceptions of it. Either we have no grip on the world, or we have no Archimedean point from which to be sure that we are not victims of total mistake.

Not only this, though, but the two perspectives also undermine each other, for when we look at ourselves from the point of view of the for-itself, we seem to be objects in the world, a kind of in-itself, and when we look at ourselves from the point of view of the in-itself, we seem not to be objects in the world, but subjects, different from simple objects in the world, a kind of for-itself. Both points of view are intrinsically unstable, for they both require the other, which then contradicts them (one can generate the kinds of paradoxes which Hobbes's account of freedom is subject to when considering the radical subject and its relation to the world, for example, as appear, I think, in Sartrean existentialism). Habermas's proposed solution is to give up both, and to move to a kind of pragmatism, rather like Quine's: everything has to have presuppositions, but when we considered something else, we can call into question the presuppositions of what was initially considered, a freedom which can only exist in constraint.

This escapes the dualisms of subject and object because the presuppositions are given by intersubjective agreement, agreement which gives itself its own grounds, much as Quine's set of beliefs sets the conditions by which any individual belief is to be assessed: everything can both be considered as given and as contingent, just not simultaneously. Habermas claims that this also undermines as whole set of associated dualisms - the free-will/determinism one, most obviously - as well as showing the falsity of an awful lot of contemporary European thought, because they rely on shifting from one of the two points of view to the other without acknowledging it - they involve themselves in what he calls a performative contradiction (for example, if Foucault's claim that all discourses are totalising instruments of power is true, then not only is his own discourse a totalising instrument of power, it is hard to see what he is opposing this claim to, from where he is getting the normative weight that this claim bears, because for one thing, the idea that being subject to a totalising instrument of power is bad or wrong is surely part of a discourse as well) but this is by the by here.

The relationship of these ideas to Cohen's claims about the fact-dependence or otherwise of ultimate normative principle is what is important now. The point is, if Cohen is right, and ultimate normative principles are totally fact-independent, neither Quine nor Habermas can be right, because, relatively simply in Quine's case and perhaps slightly more complicatedly in Habermas's, if any individual belief has its validity tested against its fit with all our other beliefs, then surely, at least in principle, some change in our factual beliefs could alter the truth or otherwise of a given principle. Cohen doesn't actually offer an argument against either (he discusses holism very briefly, in what I consider to be a totally inadequate way), merely offering a kind of phenomenological account of our processes of moral reasoning, which is, I admit, very persuasive in its simplicity, and since he doesn't offer an argument against the views he is arguing against, I think it's fair to dismiss his challenge until he does. So Rawls is vindicated, at least in this sense, against Cohen, if he can consistently adopt these kinds of epistemological positions.

I think he can, and that in fact he does, more or less. As I've already stressed, the Original Position is supposed to model the debate of principles of social regulation between people conceived of as free and equal, yet that surely implies some kind of indeterminacy in their conclusions in exactly the same way as Quine and Habermas's epistemologies do. In the face of that kind of indeterminacy, it would be bizarre to claim truth in the neo-Platonic sense that Cohen wants him to. This is not to dismiss Cohen's non-meta disagreements with Rawls - as my piece yesterday showed, I think he's probably right about those - but at a meta-level, I think Rawls is right: not only in his conception of the political, but also in his more broad epistemological commitments. Epistemology seems to tell us how to categorise normative claims, but not their whole content, and in this particular case, it seems to be telling us that Cohen's kinds of truth-claims just don't exist.

Unfortunately - or fortunately - there's nowhere obvious to go from this kind of conclusion, so I'm not sure what I might write about tomorrow, if indeed I will write anything at all. If anyone has any suggestions, I'm doing papers in Contemporary Political Philosophy - otherwise known as Cohen and his disagreements with Rawls - Political Theory from Machiavelli to Burke, and Contemporary European Social and Political Thought, as well as a core Political Theory paper, so I'd in principle be happy to write on any of those things. That said, I might come up with something all by myself.