Wedgwood sets his sight on the semantic project that is expressivism, the view that the fundamental explanations of the meaning of normative statements are the types of mental states that those statements express. This view is in contrast to the factualist approach, which holds that the fundamental explanation of the meaning of normative statements are in propositions that are made true or false by facts of the world, such that the meaning of a normative statement is due to the correspondence of the statement’s proposition with the truth. Wedgwood notes that the most skilled developments of expressivism come from the hands of Simon Blackburn and Allan Gibbard, and though they both strive to account for how expressivism can accommodate many of the theses thought to be possible for the moral realist, because both of these philosophers think that the fundamental meaning of normative statements are the mental states they express, if moral discourse requires the truth of factualism then the expressivist’s psychologistic semantic project fails.
The common argument for expressivism, Wedgwood notes, is the argument from the truth of internalism in moral motivation to non-cognitivism, and from there, that expressivism is the best semantic thesis for non-cognitivism. Another argument for expressivism is that the ontological commitments of a truth-conditional moral semantics are too strange to take seriously, which leaves us with expressivism as the most ontologically economical view. Later Wedgwood will argue that the ontological commitments of a factualist semantics are not mysterious, but for now his intent in the chapter is to show that expressivist semantics cannot accommodate a basic function of moral discourse, such that it is a mistake to favor expressivism.
Wedgwood then turns to the objections to expressivism, looking first to Geach’s Fregean objection that moral terms can be used in sentences that contain sentential operators where that operator has the largest scope and not the moral term, and these sentences make sense, so it seems mistaken to claim, as the expressivist does, that normative terms lose meaning when they do not have the largest scope in the sentence. Wedgwood makes a nice addition to this objection by noting that the objection shows that an expressivist theory must account for how logical operators have the same function whether or not there are normative terms in the statement, such that such a view must be able to explain how “the meaning of normative terms and of the logical operators allows all the customary forms of inference to be valid” (The Nature of Normativity, 44).
While this argument defeated early versions of expressivism, Wedgwood admits that Gibbard’s expressivist theory avoids Geach’s objection, and his alteration of it. For Gibbard, a normative statement expresses the mental state that is a “normative judgment”, where a normative judgment is a mental state that maps out a plan of what to do in the situation. More specifically, this judgment consists in negative attitudes towards a complete set of actions possible in the situation, save for one action, which is met with the attitude of disagreeing with disagreeing with that action, such that it is permissible. Going further, Gibbard posits that a hyperplan would be a complete and consistent plan about what to do for every possibility. A normative statement, on Gibbard’s view, expresses the mental state of an attitude of disagreeing with all options except one, such that when a logically complex statement is uttered, with a logical connective that operates on two moral terms, the two moral terms have meaning if they are allowed by the fully complete hyperstate about what to do, such that the meaning of the logical operator is confirmed by whether those terms stand in that relation to each other in their consistency in the hyperstate. Wedgwood seems right to conclude that Gibbard’s expressivist account can explain how moral terms can have meaning even when they are not of the largest scope of the sentence, and how normative terms can function in logical inferences in the same way as non-normative terms.
Wedgwood then turns to his main objection to expressivism, which is, neatly summarized, that normative discourse is a discourse guided by truth, where, “in making normative statements, speakers aim to comply with, and are assessed or evaluated according to, certain standards of justification or warrantedness” (47). Statements can be criticized in a wide variety of ways, such as when they do not properly express the mental state of the speaker, when their contradictory nature makes them unjustified, and when they are stated to be conclusions from statements from which they do not logically follow. The conclusion of such points is that an explanation of the meaning of normative statements must offer an account for such standards of justification, and why it is that we should want to meet these standards of justification, or in other words, why being justified is important to us in normative discourse. It is the latter result that Wedgwood thinks is particularly troubling for the expressivist, as from the thesis that the meaning of a normative statement is the mental state it expresses, it in no way follows that there is any profound reason that we should want our normative statements to be justified in an epistemically satisfactory way if normative statements merely express our attitudes. If normative statements just express our attitudes about our plans, then how do we enter into normative discourse to deliberate what we ought to do?
In that moral deliberation aims at something, it seems the only hope for the expressivist, according to Wedgwood, is to posit that there is a property that a statement can have that is the point of such moral deliberation, such that the deliberator wants their statement to have this property. Wedgwood refers to this property as “winningness” such that a statement that has this property is a winning statement in the sense that the statement has the property they want it to have. But suppose we fill in the spaces, such that we use this “winning” property to explain the drive of moral deliberation? We might then say that normative statements avoid the criticisms of being non-contradictory and useful for valid inferences when they are “winning” statements. But, when fully fleshed out in such a way that it can properly succeed in these functions, the property of “winningness” just is “truth”. The result is that expressivism, as a semantic theory for normative statements, cannot in fact account for a central feature of normativity, moral deliberation, because moral deliberation requires that the truth in some way guides are discourse such that “truth” can be a property of moral statements, and as expressivism is a deflationist theory of truth that holds that ‘truth’ is not a property, it cannot adequately explain moral discourse.
Yet if the expressivist project fails, then our alternative is the factualist position that Wedgwood outlined in the beginning of the chapter, being roughly that normative statements are essentially propositions that are made true or false by facts about the world, and in what distinguishes it from the error theory/fictionalism, that some of the propositions are true.