Tuesday, 11 August 2009

‘Diverse logics’, capacity and normativity

This morning I’ve been invited by the Centre for Philosophy and Mental Health at the University of Plovdiv, Bulgaria to contribute to a seminar on ‘Psychiatry and the Logics of Practice’ at the INPP meeting in Lisbon, 2009. They say:

The idea of the symposium is to announce the meeting between two groups who have been ‘digging the same tunnel from opposite sides’: psychiatrists interested in using philosophical and logical research, and philosophical logicians willing to reform their discipline so as to make it adequate for the study of the human world. On our part, we recognize the difference between the (explicit and mathematized) logic of textbooks and the (essentially implicit and heterogeneous) diverse logics in which we think and act, and we believe that the development of reliable methods for the study of these logics can be of great help for approaching the problems of human mentality.

Since Lisbon isn’t that far off, a rational thought might be: I don’t actually have anything relevant on my hard disk and so it’s too late. But, in another life as an administrator at the LSE, my first line manager, Michael Arthur, convinced me that one should never turn down work.

On that basis, this is my first hunch (I will try and generate another couple in the next few days): there’s something to be said to address an issue raised by Natalie Banner in discussions of her PhD on assessments of decision making capacity. Start with the distinction between procedural and epistemic rationality drawn by Jose Bermudez.

Procedural rationality consists in reasoning in ways that conform to the formal principles of the logic of consistency, while one is epistemically rational to the extent that one reasons in accordance to the norms of good reasoning. Whereas procedural rationality is a matter of inference, of the conclusions that it is appropriate to draw from a given belief or set of beliefs, epistemic rationality is principally a matter of the dynamical relations of how beliefs relate to evidence and how they should be changed in response to changes in the structure of evidence. [Bermudez: 2001a]

Now the niceties of debates about internal and external reasons aside, there seem to be Wittgensteinian and Rylean reasons to think that procedural rationality has to be approached with some care. There are two main reasons for this. First, there is the Wittgensteinian critique of platonic foundations for logical compulsion. Contrary to Achilles’ wishes, logic never takes one by the throat. Second, grasping a principle cannot be explained through psychologistic relations to mental elements or talismans.

These considerations suggest, at first, a revisionary possibility. If they lack external support, perhaps the canons of procedural rationality are merely one set among many possibilities. If so this might help underpin a notion of ‘diverse logics’ in the brief. But I accept Lear’s forceful rejection of this idea (if not his take on its significance) and a too close assimilation to Quinean pragmatism and the revisable web of belief.

They also suggest that the distinction itself between procedural and epistemic rationality lacks one characteristic that might have been attractive: that the former could be sketched out other than from within. That in turn suggests that the distinction is akin to Mary Hesse’s construal of the distinction between theory and observation. Against the backlash against a complete separation of theory and observation – a backlash that held there was no significant distinction – she argued that one could draw one by picking out the observational judgements but only relative to a background theory. Theory underpinned the purely observational status of some claims. But that meant the distinction could not be the neutral arbiter of theory. In this case, against a background of reactions to rules, one can pick out the deductive operations of merely procedural rationality. But the contrast is drawn only against a shared horizon (sorry!) of epistemic rationality.

Against a background of these sorts of considerations, Natalie pressed the point: if procedural rationality is only a part of the overall picture of a rationally structured capacity to make decisions, and if even it depends on a background of shared epistemic rationality, then surely that shows that there’s no further relevance of Wittgenstein’s discussion of rules? Rules are only relevant for procedural rationality.

But that, I think, misses one of the morals of the rule following considerations. Since general principles have a limited role in the explanation of judgement (that is not to say they cannot serve their ordinary justificatory role), we have instead to invoke a form of judgement particularism, of seeing a move to be the right move in the context. Wittgenstein’s discussion of rules highlights both the central role of normativity and the balance between generality and particular judgement. It’s not only of relevance where the generality takes the form of a context-free prescription but anywhere where one can ask the question in virtue of what is this judgement normatively constrained?