Dworkin's semantic sting argument is aimed against a certain attempt (which he considers Hart to have made) to apply philosophy of language in legal and political philosophy. But note that his objection to Hart's theory of law could be framed differently, without mentioning semantics or language at all: it is an argument that Hart was wrong to think that participants in a legal system share a rule of recognition, in the sense of an agreed way of deciding what rules are valid under the law of their community. Instead of opposing the misconception that the language of the law can be meaningful only if lawyers share criteria for the truth of statements of law, Dworkin could oppose the misconception that the members of a community can make legal judgments only if they share uncontroversial tests of legal validity that exhaust the grounds on which such judgments are legally justified.
Nicos Stavropoulos has restated the semantic sting argument as an argument against a ‘traditional semantic theory’ that he calls the ‘Criterial Model’ (Stavropoulos 1996). The Criterial Model holds that the content of some or all of the concepts used in legal practice, like ‘contract’ (and including the concept of law), is specified by a set of shared beliefs of users of the concept, which provide ‘criteria’ (in Dworkin's sense of the term) for the application of the concept. Stavropoulos offers an alternative understanding of the content of such concepts, borrowing eclectically from Dworkin and from Saul Kripke, Hilary Putnam, and Tyler Burge. He views the main concepts of legal discourse as referring to real properties, so that it is the task both of theorists and of lawyers to identify those properties, and there is room for ‘substantive disagreement’ over the nature of those properties. Just as ‘Aristotle’ refers to Aristotle, and ‘water’ refers to water, and ‘arthritis’ refers to arthritis (instead of referring to whatever has a set of properties that speakers believe Aristotle, water, or arthritis to have), a legal concept like ‘contract’ refers to contracts (instead of referring to whatever has a set of properties that speakers believe contracts to have). The content of those concepts is given by the real properties of the things that people use the words to designate (i.e. the properties of Aristotle, of water, of arthritis, and of contracts). The content of such concepts is not determined by the (more or less mistaken, and more or less shared) views about those properties that users of the words have in mind.
Like Dworkin, Stavropoulos uses the linguistic or ‘semantic’ mode,[3] treating a legal dispute, such as whether an agreement is a contract, as a dispute over whether the concept ‘contract’ applies to an agreement. Stavropoulos treats Hart (and others) as having a semantic theory for the concept of a contract which is defective because it takes the content of the concept to be determined by shared beliefs or ‘criteria’. But Hart doubtless would have viewed disputes over alleged contracts as disputes over the application of the rules at a time in a jurisdiction for contract formation, or concerning the legal effects of misrepresentation, mistake, or frustration, etc.; in the view of a theorist like Hart, a contract is simply an agreement made enforceable (or valid for purposes other than enforcement) by the rules of the legal system (and enforceable in ways determined by those rules). Such a theorist might claim that there is no disagreement over the content of the concept (the members of the community agree in understanding a contract as an agreement that is enforceable according to law). But this agreement concerning the concept may leave a great deal of disagreement over whether particular transactions count as enforceable agreements.
