Reference

9. Further issues

Terms in opaque or intensional contexts cannot be seen as having their usual referential roles. For, in these contexts, particularly those of propositional attitude ascriptions (see Propositional attitude statements), the replacement of a term by a coreferential term may not preserve truth.

There are a range of what might be called ‘negative’ views of reference. (a) Some philosophers have a ‘deflationary’ view according to which there is nothing more to referential notions than is captured by all instances of a schema like ‘e designates a‘, where what is substituted for ‘a’ ‘translates’ what is named by the term substituted for ‘e‘; ‘“Socrates” designates Socrates’ is a typical instance. This view accompanies a similarly deflationary view of truth (see Truth, deflationary theories of). (b) W.V. Quine (1960) argues that even once the translation of a sentence has been fixed the reference of any part of the sentence is inscrutable; thus there is no fact of the matter whether an alien’s ‘Gavagai’ in response to an environment of rabbits refers to rabbits, undetached rabbit parts, time-slices of rabbits, and so on (see Radical translation and radical interpretation §§1–4). Related to this, Donald Davidson (1984) takes an instrumentalist attitude to reference, denying both the need for, and the possibility of, a theory of reference. Putnam (1983) gives a model-theoretic argument that reference is indeterminate because any theory has unintended models. (c) Kripke (1982) presents an argument (which he finds in Ludwig Wittgenstein’s discussion of rule-following) that the meanings and references of terms are not determinate (see Meaning and rule-following). (d) Less sweepingly, Hartry Field (1973) has argued that in some cases there is no determinate matter of fact whether a term refers to one thing or another and we should see it as ‘partially referring’ to both: for example, ‘mass’ as used by Newtonians does not determinately refer to either proper mass or relativistic mass but partially refers to both. (e) Finally, those in the ‘structuralist’ tradition reject reference, and hence its role in meaning, altogether. They apparently think that the only possible theory of reference is one according to which a word resembles what it refers to. But this theory is refuted by the fact that language is arbitrary, by the fact that anything could be used to mean anything. Reference is thus left as simply ‘God-given’, which is unacceptable (see Structuralism in linguistics).

Finally, views on reference can bear on realist notions about the external world (see Realism and antirealism). Putnam, for example, draws antirealist conclusions from his model-theoretic argument. And consider the consequences of a holistic description theory for scientific terms. When, in time, we come to replace one scientific theory with another, it is natural to think that part of the reason we do so is that the theory does not accurately describe reality. Combine this thought with the holistic view that the reference of each term in the theory is determined by its associations with all other terms in the theory, and we get the consequence that all terms in the theory fail to refer. So, it was a mistake to believe in the entities apparently referred to by that theory. Worse, it is probably a mistake to believe in the entities of our present theory, for that theory will surely be replaced in time too. So we should not be scientific realists. Indeed, these considerations lead Thomas S. Kuhn (1970) and others to constructivism, a radically relativistic antirealism: rather than saying that the replaced theory does not describe reality, they say that it describes its reality, a reality that only exists relative to that theory (see Constructivism). Each theory has its own reality and no sense can be made of scientific entities existing ‘absolutely’. This line of thought can be resisted by rejecting the holistic description theory of reference in favour of a localist theory, perhaps one explaining reference in terms of causal relations to reality (see Holism: mental and semantic).