Indeterminacy of interpretation and inscrutability of reference: does the absence of a unique correct interpretation collapse into nihilism about meaning?

Donald Davidson, answered from the texts and cited to the page.

The position in brief

Among empirically adequate interpretive truth theories, there is no uniquely correct one. The Celsius-Fahrenheit analogy is the right frame: two temperature scales assign different numbers to the same physical states and are both adequate, relative to their calibration.¹ Neither scale is the uniquely correct one, but neither is arbitrary; both are fully grounded in the facts about temperature.

Similarly, two interpretive truth theories for a language may assign different truth conditions to the same sentences while both being confirmed by the same behavioral evidence.² The indeterminacy is at the level of theoretical concepts — which particular truth conditions and referents are assigned — not at the level of communication, which proceeds correctly on either adequate interpretation.³ This is not nihilism about meaning; it is the correct account of what meaning-determination comes to.⁴

Sources

1. Just as we must indicate whether the numbers we are using to measure temperature place the temperature on the Fahrenheit or the centigrade scale, so we must indicate which method of interpretation we are using (s or s', for example). One of the conveniences of sticking to the same method is that we can leave the indication of the method tacit.

   Subjective, Intersubjective, Objective, pp. 98–99

2. He goes to lengths to convince us that there may be two theories such that they imply all and only true observation sentences, are equally simple, and yet logically incompatible. Thus truth is apparently relative to a theory. But Quine settles, at the end, for a 'frank dualism' in which distinctive signs are used to state the theories. The theories turn out to be irreducible one to the other, but unconflicting.

   Inquiries into Truth and Interpretation, pp. 234–235

3. Correct interpretation keeps track of a complex pattern, and locates particular sentences and attitudes within it. But the counters we use to represent this pattern, namely our sentences, can represent it in more than one way. If the two interpretations of someone's utterance of 'Rome is a city in Italy' are equally correct, why is it that we feel that one is the right one?

   Subjective, Intersubjective, Objective, p. 98

4. How can a theory of absolute truth give an account of communication, or be considered a theory of meaning? It doesn't provide us with the materials for defining or analysing such phrases as 'means', 'means the same as', 'is a translation of', etc.

   Inquiries into Truth and Interpretation, pp. 224–225

Where the argument is made

• 'Radical Interpretation' (1973, in Inquiries into Truth and Interpretation, OUP 2001)
• 'The Inscrutability of Reference' (1979, in Inquiries)
• 'A Nice Derangement of Epitaphs' (1986, in Truth, Language, and History, OUP 2005) — on malapropism, passing theory, and the reality of language

How the argument was received

Searle's charge that inscrutability collapses into nihilism mistakes the level at which the indeterminacy operates: it operates at the level of theoretical assignment, not at the level of communication. Two interpreters using different adequate theories both correctly predict what the speaker holds true and what the speaker will do given those beliefs. Field's causal theory of reference as an alternative faces its own inscrutability: which causal chain is the reference- determining one? The specification of the relevant causal relation reintroduces the underdetermination the causal theory was meant to eliminate. Putnam's model- theoretic argument exploits the same underdetermination point and is answered by the same constraints: radical interpretation imposes behavioral and rationality constraints that formal model-theoretic satisfaction does not.