Davidson, Donald (1917–2003)

5. Radical interpretation

Davidson requires of the meaning-giving T-theory that it be empirically warranted under the practice of radical interpretation. This means that empirical considerations must be respected in choosing between different but true truth theories for, say, German, between a theory that issues in (S) and one that issues in the true but non-interpretive (S*). The favoured T-theory is to be selected on the basis of evidence plausibly available to a radical interpreter (RI), ‘someone who does not already know how to interpret utterances the theory is designed to cover’ (1984: 128). More particularly, an RI does not know the language they are trying to interpret and has no access to bilingual informants, prior dictionaries and the like. An RI can generally tell when an informant ‘holds-true’ a sentence even though they do not know the interpretation of the sentence (1984: 135). So, among the primary data for an RI are, for example:

Data like (E) are collected from a variety of speakers across a variety of times to confirm or support a generalization like:

Sentences like (GE) provide evidence that the speakers of the community take some form of words to express a certain truth. Davidson, remarkably, does not consider anything else as potential evidence for interpreting another (1984: 135).

What licenses an inference from data like (GE) to the corresponding T-sentences? Davidson’s answer is that a principle of charity is presupposed. According to this principle, the favoured truth theory for a language L must entail T-sentences according to which most of the sentences that speakers of L hold true are true. Under radical interpretation, sentences held true must usually be true because interpretation is partly constituted by this principle of charity. So, once sentences like (GE) are collected, we can infer corresponding T-sentences via a principle of charity (see Charity, principle of §4; Meaning and understanding §2; Radical translation and radical interpretation §§7–10).