Abstract
Scott Soames has recently argued that the fact that lawmakers and other legal practitioners regard vagueness as having a valuable power-delegating function in the law, evidenced by actual legislative practice, gives us good reason to favor one theory of vagueness – the partialdefinition/ context-sensitive theory – over another – the epistemic theory. The reason, Soames says, is that for a significant set of cases, the former helps explain this function, whereas the latter does not. If Soames is right, then facts about legal practice can in an important sense adjudicate between rival theories of vagueness, which is an exciting conclusion, both from the point of view of philosophy of law and philosophy of language. The argument is also likely to generate considerable optimism about what else we might expect to learn about language by looking at the law. The purpose of this paper is to significantly temper any such expectations, by arguing that we have to give up the one premise of Soames’s argument that he seems to take to be uncontroversial: that the legal content of a statute or constitutional clause is identical with, or constituted by, its communicative content. Following Mark Greenberg, we can call this a version of the communicative-content theory of law. Recently, the communicative-content theory has come under serious pressure from several prominent philosophers of law and legal scholars – including Greenberg, Lawrence Solum, and Dale Smith – who point out that legal textbooks are full of examples in which there appears to be some clear difference between the communicative content of a statute or constitutional clause and its legal content. I argue that the problem raised by these examples gives us good reason to reject identity- and constitution-based version of the theory, like Soames’s, but go on to provide a preliminary sketch of my own account of legal content – the Pro Tanto view, as I call it – and show how it avoids the problem by allowing us to explain away the apparent “gaps” in a principled and unified way. Despite being a version of the communicative-content theory of law, however, the Pro Tanto view does not suffice to vindicate Soames’s argument for the partialdefinition/ context-sensitive theory of vagueness, which I examine in some detail in the latter half of the paper. Due to the need for a fairly complex account of the relationship between communicative content and legal content, facts about legal practice do not, after all, seem to be able to adjudicate between rival theories of vagueness – at least not in the way envisioned by Soames. Discussing, briefly, the cases of Maurice v. Judd and Bronston v. United States, I conclude by arguing that my point about Soames’s argument is generalizable: since legal content is neither identical with nor constituted by communicative content, we should be quite cautious about drawing general conclusions about language on the basis of facts about legal practice.