Page 259 - Encyclopedia of Philosophy of Language

But the intuitive idea is that the properties, mech- anisms, and processes posited in the special sciences must be explicable by more fundamental ones, and, while it is hard to come up with a general formula, particular cases are often clear enough. If cognitive psychologists posit, for example, a capacity to access a special pictorial memory, they owe an explanation of the mechanism of access, an explanation appealing to simpler psychological or neural mechanisms. So the ambitious program is committed to giving an account of the relation between words and the world, not just taking that relation as primitive. It is most plausible to suppose that the semantic can be explained by the psychological. So naturalist programs have reductive intentions towards semantics.
There is a lot of scepticism about the prospects for naturalizing semantics. Both Kripke and Putnam argue that the reductive intent is unnecessary and incapable of fulfillment. It is just a mistake to think of the human sciences in general, and semantics in particular, as protosciences. For one thing, that is to miss the normative element of a theory of meaning. No regularity of behavior, or psychological dis- position underlying that regularity, can give sense to the notion of a term being misapplied. Suppose that some benighted tourist points at a wallaroo and says Lo! A kangaroo. They have said something false. But that is so only if their tokens of 'kangaroo' apply to all and only kangaroos, not to kangaroos and wall- aroos. But they have just demonstrated that their 'kangaroo'-using disposition is the disposition to apply the term to both varieties of macropod. Since such dispositions and their ilk are, in his view, about the only facts the naturalist has to trade in, Kripke is decidedly sceptical about the prospects for an account of reference in naturalistic terms with sufficient resources to give a robust account of misrepresentation.
So defenders of the ambitious program propose to develop a theory of meaning with reference as its core semantic notion, deriving a theory of sentence meaning from that core. But they do not take ref- erential relations to be primitive; on the contrary, these are to be explained by appeal to more fun- damental facts.
3. Description Theories and Causal Theories
Traditional theories of reference, concerning either singular or general terms, were not naturalist theories. Consider, for example, the theories that have been offered for general terms, such as photon and tiger. These are both examples of natural kind terms, one theoretical, the other observational.
3.1 Theoretical and Observational Terms
Theoretical terms have been the focus of intense debate in philosophy of science. In the heyday of logical empiricism, the idea was to explain the sem-
antics of 'photon' and kindred terms by reducing their semantics to the semantics of the observation language; perhaps by direct definition, perhaps more subtly. Not just evidence but also meaning depended on observation; hence the 'logical' in 'logical empiri- cist.' But attempts to explain the meaning of theor- etical terms by appeal to observational ones simply have not succeeded. So it has been supposed that the semantics of theoretical terms is derived from the role they play in the theories that contain them. The term 'species,' for example, derives its semantic properties from its role in the theories of evolution, ecology, and population genetics, the theories that essentially appeal to that notion. So 'species' means something like 'evolving interbreeding ecologically coherent populations,' for that is how these theories charac- terize species. The idea was originally proposed as a theory of the meaning of theoretical terms, but unless a term's meaning is to be completely decoupled from its reference, we can take it to be a theory of reference as well: the term refers to all and only those biological entities that (perhaps only approximately) fit that characterization.
This conception of the semantics of theoretical terms faces serious problems, for it follows that the semantics of a term change not just when the theory undergoes revolutionary restructuring, but whenever the theory is significantly modified. Though a number of philosophers have embraced some version of the idea that theory-change entails meaning-change, it has implausible consequences. It becomes difficult to explain how even those who accept differing versions of the same basic theory routinely succeed in com- municating with each other, and still more difficult to explain the successful communication between those whose views are very far apart. But Darwin and Owen, for example, despite their very different views on the nature of species, understood each other perfectly well; understanding was a precondition of their deep disagreement. Moreover, this view of the semantics of theoretical terms makes it difficult to give an account of the cumulative aspects of scientific change.
So these theories of the semantics of theoretical terms may not work on their own lights, but even were they to work, they are inadequate by the atomist's lights. For term meaning is explained by appeal to other semantic facts. Logical empiricists attempted to reduce the meaning of 'photon' and its ilk to the semantics of observation terms. The defenders of 'theoretical role' semantics explained term meaning by appeal to the meaning of the sentences that jointly make up a theory. We have here no explanation of representation by a more fundamental class of fact.
3.2 Componential Analysis
Theoretical role' semantics at best explains some sem- antic facts by appeal to others; the same might well be said of many theories of the meaning of words
Reference: Philosophical Issues
237