Eligibility and inscrutability

The philosophy of intentionality asks questions such as: in virtue of what does a sentence, picture, or mental state represent that the world is a certain way? The subquestion I focus upon here concerns the semantic properties of language: in virtue of what does a name such as ‘London’ refer to something or a predicate such as ‘is large’ apply to some object? This essay examines one kind of answer to this “metasemantic”1 question: interpretationism, instances of which have been proposed by Donald Davidson, David Lewis, and others. I characterize the “twostep” form common to such approaches and briefl y say how two versions described by David Lewis fi t this pattern. Then I describe a fundamental challenge to this approach: a “permutation argument” that contends, by interpretationist lights, there can be no fact of the matter about lexical

content (e.g., what individual words refer to). Such a thesis cannot be sustained, so the argument threatens a reductio of interpretationism. In the second part of the article, I will give what I take to be the best interpretationist response to the inscrutability paradox: David Lewis’s appeal to the differential “eligibility” of semantic theories. I contend that, given an independently plausible formulation of interpretationism, the eligibility response is an immediate consequence of Lewis’s general analysis of the theoretical virtue of simplicity. In the fi nal sections of the article, I examine the limitations of Lewis’s response. By focusing on an alternative argument for the inscrutability of reference, I am able to describe conditions under which the eligibility result will deliver the wrong results. In particular, if the world is complex enough and our language suffi ciently simple, then reference may be determinately secured to the wrong things.