Reconciling Internalism with External Content through McDowell's Normativity — Epoche C2
What the Twin Earth argument leaves standing Set the year at 1750, before anyone on either planet could have said what the clear liquid in the rivers is made of, and consider two speakers alike down to the molecule: Oscar on Earth, where that liquid is H 2 O, and his counterpart on Twin Earth, where it is some other compound with the same look, taste and capacity to quench thirst. Hilary Putnam's argument of 1975 is that the word 'water' in Oscar's mouth applies to H 2 O and in his counterpart's mouth to the other stuff, so that the two men mean different things by the word while being, considered from the inside, indistinguishable. This essay asks what survives of internalism — the thesis that what a person is thinking is settled by how that person is, independently of her surroundings — once Putnam's argument is granted. The answer defended here is that what survives is not a smaller quantity of inner content but a different demand altogether: that a thinker's experiences and thoughts be available to her as reasons. John McDowell's Mind and World (1994) is the book that shows how to meet that demand without narrow content, and the normativity in this essay's title is that demand itself. Two corrections have to be made at the outset, because the earlier and shorter version of this essay got them wrong, and the argument cannot be repaired while they stand. The first concerns McDowell. That version credited him with a non-conceptual dimension of experience — a direct, unmediated openness to the world, not propositional but nonetheless normative — and made that dimension the internal contribution to content. McDowell holds no such view. The denial that experience has non-conceptual content is one of the central commitments of Mind and World , argued for at length in its third lecture, and 'openness to the world' is his phrase for experience that is conceptually structured through and through, not for a layer lying beneath the conceptual. The second concerns the Buddhist term vijñāna , which the earlier version enlisted as an ally of internalism. In the canonical analysis each variety of consciousness is named for the faculty and object on which it depends — eye-consciousness, ear-consciousness, and so on — so it is a doctrine of dependence, not of self-sufficient inner states, and cannot support the premise it was asked to support. Both corrections cost the earlier argument its mechanism. What follows supplies a better one. The argument, stated as an argument Putnam's paper does not merely report an intuition about the two Oscars; it isolates two assumptions ordinarily held together and shows that they cannot both be kept. The first is that knowing the meaning of a term is a matter of being in a certain psychological state, where 'psychological state' is understood narrowly — a state whose ascription does not presuppose the existence of anything other than the subject to whom it is ascribed. The second is that meaning determines extension: sameness of meaning implies sameness of the set of things the term applies to. The date of 1750 is not decoration; it is what makes the first assumption's antecedent hold undeniably. Neither Oscar has any beliefs about chemical composition, true or false, so whatever differentiates the two words cannot be a description that one associates with it and the other does not — there is no such description available to either. The two men are therefore in the same narrow psychological state with respect to 'water'. If meaning were fixed by that state the two words would have the same meaning, and by the second assumption the same extension; but they do not have the same extension, since one applies to H 2 O and the other does not. One assumption must be given up, and Putnam's verdict is that the narrow one goes. Two further components of the same paper prevent the conclusion from being quarantined as a curiosity about chemistry. The first is the division of linguistic labour. Putnam observes that he cannot tell an elm from a beech, that whatever he associates with the two words is for practical purposes identical, and that nonetheless 'elm' in his mouth means elm, because the labour of fixing extensions is distributed across a community in which some can tell the difference and the rest defer. Extension is a collective achievement before it is an individual one. The second is the indexical character of natural-kind terms: 'water' applies to whatever bears the same-liquid relation to the samples encountered here, so which stuff it applies to depends on which stuff is here. Tyler Burge's paper of 1979 removes the remaining exit. One might hope that the externality is confined to natural kinds, where a hidden physical essence does the fixing, leaving the rest of thought safely internal. Burge's patient believes he has developed arthritis in his thigh — a belief that is false, since the word in his community applies only to inflammations of the joints, but a belief he has, and one his doctor can correct. Now hold everything about the man fixed and vary only his linguistic community, so that in the counterfactual community the word covers rheumatoid ailments of muscles and joints alike. In that community his utterance expresses no false belief, and indeed no belief about arthritis at all. Since he is the same on the inside in both cases and his thoughts differ, thought content depends on the social environment. Nothing in the case turns on hidden essences; it turns only on the fact that words are learned from other speakers and remain answerable to them. Since nearly every concept anyone has is acquired that way, the dependence Burge exhibits is the normal case rather than a special one. Why the conclusion is felt as a loss If the conclusion were merely that reference depends on the environment, few would be troubled by it. What troubles people is a further thought about self-knowledge, and it is worth stating precisely, because the shape of the repair depends on it