Whilst I still lack a resolution to the earlier worry, this one was merely a momentary post-pub anxiety. A couple of people have given me a good slapping (thanks Daniel, for one) and the fact is that, whilst my attention was on the non-individualist account of epistemology, the necessary addition of a plausible innativist account of content (rather than syntax) is not anything I can begin to believe in. I’ll leave it to mark a moment of madness whilst I sit in the corner with a dunce’s hat on.
One might as well be hanged for a sheep as a lamb. So whilst I’m having sudden doubts about Davidson, let me add another. I am not sure about the underlying argument for the constitutive principle of rationality. (I put this up with the hope that I will be able, in all conscience, to take it down again in a couple of days.)
I have always assumed it went something like this. Let’s start thinking about the metaphysics of meaning from a mundane third person point of view. (The justification for this will be partly by results and partly because none of the other starting points has been so very successful.) What would be the constraints on the radical interpretation of speech and action?
Well an interpreter could use evidence of what prompts utterance but what a speaker says will depend on what he or she believes. But to determine that depends on knowing what they mean. Hence we have two interconnected unknowns but only one sort of evidence. Radical interpretation can, therefore, only get off the ground if the interpreter adopts the principle of charity. She takes beliefs and utterances to be largely true and rational by her standards. But given that there are – by assumption – no facts about meaning unavailable to this stance this suggests that truth and rationality are constitutive principles of beliefs and meanings. They govern ontology as well as epistemology.
Two further moves:
1) Although the example suggests desert islands and grass skirts, radical interpretation begins at home. My only grasp of my native English depends on my original radical interpretation of parents and carers.
2) Against a student who objects that learning his native language didn’t seem much like this, I press the point that this is a rational reconstruction. His only justification for his knowledge of the meaning of words in English would be a description of such radical interpretation.
This second move has always seemed both crucial and plausible to me. Radical interpretation isn’t simply a description but a kind of idealised justification. Thus it isn’t closely tied to actual empirical details. There is, however, an echo of the epistemology of Descartes’ Meditations about this. Whilst for most people, most of the time, the actual groundings of interpretation can be ignored, a Descartes in his study can undertake a particular project. Further, the justifications implicit in radical interpretation are those that an interpreter could offer him- or herself.
Against this background my nagging worry is a quite different response to the question of how one might break into the interdependent circle of belief and meaning. Suppose that the Chomskians were right and that the poverty of empirical evidence especially refuting evidence for hypotheses about one’s surrounding linguistic culture is best explained by an innate universal grammar. We get language almost for free. If so, the background for radical interpretation is quite different. We can break into the circle of beliefs and meanings because aspects at least of the meanings are not so holistic. ((Of course I am simplifying grossly: UG is merely a grammar of syntax not semantics so much more will in fact be needed to pin down actual meanings. But let me press on for the moment with the fantasy on the assumption that we might think of a more sitable innativism about the grammar of contents.))
Now a first thought will be that this is merely a comment on the context of discovery not justification. To justify a claim for knowledge of meaning would require radical interpretation again. That’s the point of point 2 above.
But note how individualistic that picture of justification is. How about the idea that, relative to our human predicament, the interpreter need do much less by way of justification than such a full blown reconstruction? She can say that such and utterance probably means that so and so and, given UG, that amounts to knowledge.
My worry is that the full argument for the principle of charity and hence for the constitutive principle of rationality turns on an appeal to what would be necessary to constrain interpretation in radical interpretation but construed in such a way that the interpreter does her work as free as she can be from mere worldly favours. But if one adds together some appropriate innativism about contents and a less individualistic conception of earning the right to knowledge, won’t that undercut the rational reconstruction and hence the need for rationality?