Hybrid metaethical theories attempt to incorporate essential elements of expressivism and cognitivism, and thereby to accrue the benefits of both. Hybrid theories are often defended in part by appeals to slurs and other pejoratives, which have both expressive and cognitivist features. This paper takes far more seriously the analogy between pejoratives and moral predicates. It explains how pejoratives work, identifies the features that allow pejoratives to do that work, and models a theory of moral predicates on those features. The result is an expressivist theory that, among other advantages, is immune to embedding difficulties and avoids an overlooked difficulty concerning attitude ascriptions that is lethal to most other hybrid theories.

Well, a review of Hilary Putnam's Ethics Without Ontology.

Nondeclarative sentences, including mixed-mood sentences, pose a problem for standard truth-conditional approaches to providing a compositional semantics for natural languages, for utterances of them are prima facie not truth-evaluable.  We advocate instead a fulfillment-condition approach and show how such an approach, which nevertheless relies on truth-theoretic semantic machinery, can be applied to imperatives, interrogatives, molecular sentences containing them, and quantification into mood markers.  We then show how to integrate exclamatives and optatives into a framework similar to the fulfillment approach. 

Though it has been claimed that Frege's commitment to expressions in indirect contexts not having their customary senses commits him to an infinite number of semantic primitives, Terence Parsons has argued that Frege's explicit commitments are compatible with a two-level theory of senses.  In this paper, we argue that Frege is committed to some principles Parsons has overlooked, and, from these and other principles to which Frege is committed, give a proof that he is indeed committed to an infinite number of semantic primitives--an intolerable result.

Jonathan Harrison attacks Tarski's theory of truth, and similar theories, on the grounds that (a) truth cannot be a property of sentences; (b) if truth could be property of sentences, T-sentences would have to be necessary truths, which they are not; and (c) not only are T-sentences not necessarily true, they can be false.  In this paper, I show that Harrison's attack fails because he fails to understand important elements of a 'Tarski-style' truth theory.  The first section is a brief review Tarski's theory of truth and of extensions of his techniques to natural languages.  The second is a defense of Tarski-style truth theories against Harrison's three objections, in three corresponding parts.

Metaethical expressivism holds that ethical sentences, like imperative and exclamative sentences, fail to have truth conditions in any sense sufficient to adopt a truth-conditional semantic account of their meanings.   Expressivists have thus been forced to adopt a different kind of semantic theory, one that assigns to sentences something other than truth conditions.  The resulting theories, which have assigned to ethical sentences mental states or speech acts, have faced enormous--and, by now, familiar--difficulties.  In this paper, I recommend that expressivists return to the analogy with imperatives and interrogatives, and investigate the extent to which expressivists can adopt Success Condition Semantics, the most promising semantic account of nondeclaratives.  I argue that hybrid expressivists can indeed adopt a Success Condition Semantic account of ethical sentences, though, pure expressivists cannot.  The result is another strong argument to the effect that if one is going to be an expressivist at all, one should be a hybrid, rather than a pure, expressivist. 

Exclamative sentences and their utterances are surprisingly perplexing. What could appear simpler to explain than an utterance of 'Yes!' or 'Congratulations!'? The appearances, however, are deceiving. Exclamatives are quite varied—many have propositional content ('What great teeth you have!'), but some have only propositional objects ('What a car!'), and some don't even have these ('Ouch!', 'Arggg!'). Can such varied kinds of exclamatives be given a unified treatment? Some exclamatives stand in what appear to be various logical relations ('Congratulations on winning the race!', 'Damn you for winning the race!'). Are these really logical relations? If so, what grounds such relations? Exclamatives are sometimes embeddable as consequents ('If you cleaned up the room, thanks!') and as first disjuncts ('Thank you for cleaning up the room or should I be thanking somebody else?), but never as antecedents ('If thank you, I am pleased') or as second disjuncts ('Should I be thanking someone else for cleaning up the room or thank you!'). Why are they embeddable within some linguistic contexts, but not others? Are these explanations semantic or pragmatic? This paper seeks to unravel some of these perplexities concerning exclamative sentences.

Michael Ridge advocates a metaethical theory he calls "Ecumenical Expressivism."  The theory is supposed to be ecumenical in the sense that it holds that moral utterances express both beliefs and desires.  It is expressivist, in Ridge's revisionary sense of 'expressivism', in that "a moral sentence M is not conventionally used to express a belief such that M is true if and only if the belief is true."  I argue that Ecumenical Expressivism is unstable.  The source of the difficulty is a quite generic use of 'express'; it is simply unclear in just what sense—indeed, senses—Ridge thinks that moral utterances "express" both beliefs and desires.  After distinguishing four different kinds of expression relation, I show that there are two plausible interpretations of what Ridge could mean when he says that moral utterances express both beliefs and desires.  On the first interpretation, Ridge's theory is indeed expressivist (in his revisionary sense), but not interestingly ecumenical; on the second interpretation, the theory is interestingly ecumenical, but not expressivist (in his revisionary sense).  The moral should be unsurprising: it matters in which sense moral utterances are held to express various kinds of mental states, so expressivist theories must take a firm stand on the issue.  Unfortunately, several recent expressivist theories have failed to do just that.

Complete Version (This is the MA Thesis from which the shorter paper was culled. Most of the good stuff is contained in the second and fourth chapters.)