Donald Davidson was an influential philosopher of language. His views are similar to those of W.V.O. Quine, in that he argues for a holistic approach to linguistic meaning that recognizes the indeterminacy of interpretation. But whereas Quine emphasizes the indeterminacy, Davidson believes the true indeterminacy is less extreme because we must adopt a "principle of charity," in which we assume as much as possible that another person shares the same beliefs we do. This shared background serves as the basis for interpreting their speech and their behavior. Assuming this shared background of basic beliefs is the only way to interpret their speech: if their beliefs are completely incommensurate with ours, we wouldn't even recognize the speaker as rational.

What makes interpretation possible...is the fact that we can dismiss a priori the chance of massive error. A theory of interpretation cannot be correct that makes a man assent to very many false sentences: it must generally be the case that a sentence is true when a speaker holds it to be. ("Thought and Talk," page 168-9)

Davidson's most provocative claim is that conceptual relativism — the view that different people might have different conceptual schemes that organize experience in different ways — is unintelligible as a doctrine. The title of one of the papers in this book is "On the Very Idea of a Conceptual Scheme."

The methodological problem of interpretation is to see how, given the sentences a man accepts as true under given circumstances, to work out what his beliefs are and what his words mean. ("Thought and Talk," page 162)

The method is intended to solve the problem of the interdependence of belief and meaning by holding belief constant as far as possible while solving for meaning. ... What justifies the procedure is the fact that disagreement and agreement alike are intelligible only against a background of massive agreement. ... If we cannot find a way to interpret the utterances and other behaviour of a creature as revealing a set of beliefs largely consistent and true by our own standards, we have no reason to count that creature as rational, as having beliefs, or as saying anything. ("Radical Interpretation," page 137)

Applying Davidson's principle of charity, we may arrive at multiple theories about the language the 'creature' speaks, each of which accounts for the empirical facts. The meaning of words and sentences derives from the overall theory, much as the meaning of a scientific term depends on its theory.

There are two approaches to the theory of meaning, the building-block method, which starts with the simple and builds up, and the holistic method, which starts with the complex (sentences, at any rate) and abstracts out the parts. The fist method would be fine if we could give a non-linguistic characterization of reference, but of this there seems no chance. The second begins at the point (sentences) where we can hope to connect language with behaviour described in non-linguistic terms. But it seems incapable of giving a complete account of the semantic features of the parts of sentences, and without such an account we are apparently unable to explain truth. ...

I suggest that what is invariant as between different acceptable theories of truth is meaning... In a nutshell: we compensate for the paucity of evidence concerning the meaning of individual sentences not by trying to produce evidence for the meanings of words but by considering the evidence for a theory of the language to which the sentence belongs. Words and one or another way of connecting them with objects are constructs we need to implement the theory. ("Reality Without Reference," pages 221 and 225)

Davidson is read less widely than Quine. The reason, I think, is that his prose style and his choice of topics are more academic than Quine's were. Especially early in his career (in the 1960s and early 1970s), he focused on specific technical issues in the theory of truth — a subject whose appeal is limited to professors of analytic philosophy. Even in his later years, when he applies his theories to 'live' philosophical issues, he uses fewer and less vivid examples than others such as Quine and Hilary Putnam.

A theory of intrepretation, like a theory of action, allows us to redescribe certain events in a revealing way. Just as a theory of action can answer the question of what an agent is doing when he has raised his arm by redescribing the act as one of trying to catch his friend's attention, so a method of intrepretation can lead to redescribing the utterance of certain sounds as an act of saying that snow is white. ("Thought and Talk," page 161)

One of the final papers in the book is called "What Metaphors Mean." In it, he argues that metaphors and metaphorical speech do not have any extra or figurative meaning. "Metaphors mean what the words, in their most literal interpretation, mean, and nothing more." The particular power of metaphors comes not from their meaning, but from their use and illocutionary force.

Just because a simile wears a declaration of similitude on its sleeve, it is, I think, far less plausible than in the case of metaphor to maintain that there is a hidden second meaning. In the case of simile, we note what it literally says, that two things resemble one another; we then regard the objects and consider what similarity would, in the context, be to the point. Having decided, we might then say that the author of the simile intended us — that is, meant us — to notice that similarity. But having appreciated the difference between what the words meant and what the author accomplished by using those words, we should feel little temptation to explain what has happened by endowing the words themselves with a second, or figurative, meaning. ("What Metaphors Mean," page 255)

I found his argument surprising, and also exciting because I think his approach could be used to support the view of 'normal' language usage I meant to propose in my dissertation; namely, the idea that the relevant identity relation for a word can vary with the context.