Moral Arguments and the Frege-Geach problem

This thesis deals with the Frege-Geach problem, which is arguably the main objection faced by the expressivist view on moral discourse. The key idea of the Frege-Geach objection is that the expressivist cannot give an account of the meaning of conditionals sentences involving moral predicates and the validity of arguments involving moral sentences. After all, as traditionally understood, validity requires truth-aptness. Philosophers such as Crispin Wright and G. F. Schueler challenged the very idea according to which desire-like attitudes can stand in logical relations. They held that if the components of the moral modus ponens are not truth-apt, the argument cannot be properly categorised as ‘valid’. After presenting and critically examining five different attempts to solve the Frege-Geach problem, I provide a new way solution to it. My main goal is to develop an expressivist framework within which it is possible to give an account of evaluative conditionals and validity without relying on the contentious assumption that desire-like attitudes can stand in logical relations. My position is influenced by Grice’s notion of conversational implicature and Vranas’ logic of prescriptions. I further argue that there is a relation between the evaluative and the prescriptive domains of discourse, and that a plausible way of understanding this relation is via the conversational implicature. I follow Peter Vranas with respect to the question whether prescriptions can have a logic, and show how to get from evaluations (‘x is bad) to prescriptions (‘don’t do x!’). Finally, I trace a distinction between two levels of validity – the psychological and the logical one – and show that the moral modus ponens is valid in both senses. My thesis therefore concludes that the Frege-Geach problem is not a knockdown objection against expressivism.