Christopher Hom

Racial epithets are derogatory expressions, understood to convey contempt toward their targets. But what do they actually mean, if anything? While the prevailing view is that epithets are to be explained pragmatically, I argue that a careful consideration of the data strongly supports a particular semantic theory. I call this view Combinatorial Externalism (CE). CE holds that epithets express complex properties that are determined by the discriminatory practices and stereotypes of their corresponding racist institutions. Depending on the character of the institution, the complex semantic value can be composed of a variety of components. The account has significant implications on theoretical, as well as, practical dimensions, providing new arguments against radical contextualism, and for the exclusion of certain epithets from First Amendment speech protection.

The norms surrounding pejorative language, such as racial slurs and swear words, are deeply prohibitive. Pejoratives are typically a means for speakers to express their derogatory attitudes. Because these attitudes vary along many dimensions and magnitudes, they initially appear to be resistant to a truth-conditional, semantic analysis. The goal of the paper is to clarify the essential linguistic phenomena surrounding pejoratives, survey the logical space of explanatory theories, evaluate each with respect to the phenomena, and provide a preliminary assessment of the initial resistance to a truth-conditional analysis

Pejoratives are the class of expressions that are meant to insult or disparage. They include swear words and slurs. These words allow speakers to convey emotional states beyond the truth-conditional contents that they are normally taken to encode. The puzzle arises because, although pejoratives seem to be a semantically unified class, some of their occurrences are best accounted for truth-conditionally, while others are best accounted for non-truth-conditionally. Where current, non-truth-conditional, views in the literature fail to provide a unified solution for the puzzle, this paper motivates a novel, semantic, analysis of pejorative language. The significance of the proposed solution is not only linguistic in nature, but also philosophical, as it both provides a new argument for, and sheds further light on, the nature of semantic externalism.

The problem of the unity of the proposition asks what binds together the constituents of a proposition into a fully formed proposition that provides truth conditions for the assertoric sentence that expresses it, rather than merely a set of objects. Hanks' solution is to reject the traditional distinction between content and force. If his theory is successful, then there is a plausible extension that readily solves the Frege-Geach problem for normative propositions. Unfortunately Hanks' theory isn't successful, but it does point to significant connections between expressivism, unity and embedding.

The following facts are straightforwardly unproblematic from a moral point of view: that no Chinese are c*'s', that there are no c*'s', but that there are Chinese people. The goal of the paper is to provide a general, explanatory, account of the logical form of sentences expressing such facts. We contend that pejorative terms like the C-word have empty extensions, and that this follows from a deeper analysis of pejorative terms as the phonological realization of a more complex syntactic construction. The semantic result is what we call the null-extension hypothesis for pejorative terms. In contrast, the prevailing view is that the referents of pejorative terms are identical to those of their corresponding correlates (e.g. that the C-word and 'Chinese' each refer to the same class of individuals). Call this the identity thesis. We argue that there is no coherent defense of the identity thesis that remains morally innocent.

In theorizing about racial pejoratives, an initially attractive view is that pejoratives have the same reference as their "neutral counterparts". Call this the identity thesis. According to this thesis, the terms "Jew" and the K-word, for instance, pick out the same set of people. To be a Jew just is to be a k*, and so to make claims about Jews just is to make claims about k*'s'. In this way, the two words are synonymous, and so make the same contribution to the truth-conditions of sentences containing them. While the fundamental claim for the identity thesis that Jews are k*'s' sounds anti-semitic, it need not be actually anti-semitic. The identity thesis is usually bolstered with the further claim that the pejorative aspect of the K-word and other such terms is located elsewhere than in truth-conditional content, so what makes the K-word a bad word is a non-truth-conditional association with anti-semitism that is not shared with the word "Jew". The exact nature and location of the negative moral content of pejoratives is a matter of some dispute among identity theorists. But whatever the intuitive appeal of the identity theory for those persuaded by such views, it is nevertheless inconsistent.

The Negation Problem states that expressivism has insufficient structure to account for the various ways in which a moral sentence can be negated. We argue that the Negation Problem does not arise for expressivist accounts of all normative language but arises only for the specific examples on which expressivists usually focus. In support of this claim, we argue for the following three theses: 1) A problem that is structurally identical to the Negation Problem arises in non-normative cases, and this problem is solved once the hidden quantificational structure involved in such cases is uncovered; 2) the terms 'required', 'permissible', and 'forbidden' can also be analyzed in terms of hidden quantificational structure, and the Negation Problem disappears once this hidden structure is uncovered; 3) the Negation Problem does not arise for normative language that has no hidden quantificational structure. We conclude that the Negation Problem is not really a problem about expressivism at all but is rather a feature of the quantificational structure of the required, permitted, and forbidden.

Fictional terms are terms that have null extensions, and in this regard pejorative terms are a species of fictional terms: although there are Jews, there are no k*'s'. That pejoratives are fictions is the central consequence of the Moral and Semantic Innocence (MSI) view of Hom et al. (2013). There it is shown that for pejoratives, null extensionality is the semantic realization of the moral fact that no one ought to be the target of negative moral evaluation solely in virtue of their group membership. In having null extensions, pejorative terms are much like mythological terms like 'unicorn horn' that express concepts with empty extensions, even though it was thought otherwise: people who falsely believed the mythology were mislead into thinking that ordinary objects (i.e. whale tusks) were magical objects, and pejoratives terms work likewise. For example, the K-word is supported by the ideology of anti-Semitism, and speakers who fall prey to its influence (perniciously or not) are mislead into thinking that ordinary people (i.e. Jews) are inherently worthy of contempt. In this paper, we explore the consequences of this parallelism, with an eye to criticisms of MSI. In particular, we will re-visit identity expressivist views - those that hold that there are k*'s' and that they are Jews, and hence deny null extensionality - arguing that this embeds a mistake of fiction for fact. Among the issues to be discussed are the role of fictional truth in understanding pejorative sentences and the relation of the semantics of pejoratives to offensive use of language. We conclude with meta-semantic reflections on the origins of word meanings.

A multidimensional account of the meanings of slurs holds that a slur has both literal, truth-conditional content (which is neutral) and conventional implicature (which is derogatory). This paper offers a careful examination of the motivations and commitments for a multidimensional account and argues that the theoretic costs for such a view are prohibitive. One of the primary motivations for a multidimensional account over a purely truth-conditional account is the apparent wide-scoping phenomenon of slurs (e.g. that derogatory content does not seem cancellable under negation). The paper argues that carefully distinguishing between predication and assertion not only dispels the misconception that the phenomena in question is centrally about scope but also vindicates the purely truth-conditional account as a more general and unified explanation.

Hanks (2015) develops a theory of propositions as speech-act types. Because speech acts play a role in the contents themselves, the view overturns Frege's force/content distinction, and as such, faces the challenge of explaining how propositions embed under logical operators like negation. The attempt to solve this problem has lead Hanks and his recent commentators to adopt theoretically exotic resources, none of which, we argue, is ultimately successful. The problem is that although there are three different ways of negating the sentence "Mary's card is an ace", current speech-act theories of propositions can only accommodate two of them. We distinguish between (1) "It is false that Mary's card is an ace" (sentence negation), (2) "Mary's card is a non-ace" (predicate negation), and (3) "Mary's card is not an ace" (content negation) and show that Hanks and his commentators cannot explain content negation. We call this Hanks' Negation Problem. The problem is significant because content negation is the negation required for logic. Fortunately, we think there is a natural way for Hanks to accommodate content negation (and all the other logical operators) as successive acts of predication. The view solves Hanks' Negation Problem with only resources internal to Hanks' own view.

Speech-act theories of meaning have enjoyed a rise in prominence with Hanks (2015) and Soames (2015). Both focus the vast majority of their attention on providing an analysis of the semantic contents of assertions, and more or less assume the extension of their accounts to questions and commands. This paper examines two foundational issues that arise when attempting to extend the speech-act analysis to imperatives. The first issue is that because imperatives successfully embed under logical operators like disjunction, a version of the Frege-Geach problem arises. The second issue is that embedding under negation forces any such theory to make transparent how contradictory commands are to be represented. We argue that both issues can be adequately addressed by an extension of the account in Hanks (2015), and therefore Hanks' view should be considered a viable overall theory for the semantics of imperatives.

There is a sharp, historical, theoretical divide in the recent literature on racial epithets that has been underappreciated. On one side is cognitivism; roughly the view that racial epithets should be explained via semantic, truth-conditional content. On the other side is non-cognitivism; roughly the view that racial epithets should be explained via content that is somehow pragmatic or expressive. The philosophical literature has overwhelmingly sided with the second. We hope to counteract this trend by pointing out that much of the non-cognitivist criticism of cognitivism commits what we call The Framework Fallacy (FF) which is basically to assume one's framework in order to criticize the other. The two major, cross-theoretical arguments are The Projection Argument (against cognitivism), and The Racist Frege Puzzle (against non-cognitivism). In offering a detailed assessment of each, we argue that the first commits the FF, while the second does not. This shows that from the appropriately abstracted point of view, the final theoretical tally favors a cognitivist explanation of racial epithets.

In their Inflammatory Language, Stojnic and Lepore present four major criticisms of content-based views of pejorative language: 1) the Projection Argument; 2) the Hyperprojection Argument; 3) the Specificity Argument; and 4) the Reclamation Argument. This paper argues that a content-based view can adequately respond to each of these criticisms. The paper goes on to consider their positive view, the Articulation Account, and argues that it suffers from being both under-specified and overly ambitious. Even when the view is plausibly precisified as a functional role theory of articulations, a serious dilemma arises: focusing on the sound or shape of the articulation is problematic when considering counterexamples like the racist use of 'Monday' as code for the N-word, and focusing on the functional role of the articulation is problematic when considering how little functional similarity there is between the Mandarin demonstrative term that sounds like the N-word and the N-word. The paper also presents the applications of two external criticisms deriving from the Identity Thesis and the Framework Fallacy, and concludes that the Articulation Account falls short of being a leading contender in the analytic space of views for pejorative language.