Davidson, Donald - Inquiries into truth and interpretation

Philosophers are fond of making claims concerning the properties a language must have if it is to be, even in principle, learnable. The point of these claims has generally been to bolster or to undermine some philosophical doctrine, epistemological, metaphysical, ontological, or ethical. But if the arguments are good they must have implications for the empirical science of concept formation, if only by way of saying what the limits of the empirical are.