question of what effects they produce are deeply related; the experimental research in psychology provides interesting insights in this respect. In chapter 4, I defend the two central claims of my view against potential objections: the points raised here may be a bit technical, but they serve to address or anticipate potential criticisms. First, I focus on the uniformity claim, according to which slurs and thick terms should be analyzed along similar lines. My strategy is to distinguish superficial differences from authentic ones, and show that the latter do not challenge a uniform account, as certain features of the semantics of slurs and thick terms can explain the phenomenal divergences one observes. Then, I defend the presuppositionality claim, according to which HEs encode evaluative presuppositions. I address the main objections raised against presuppositional accounts of slurs and thick terms, and show that these objections do not constitute an obstacle to the presuppositional approach. Finally, chapter 5 is dedicated to nonstandard uses of HEs, including the reclamation of slurs and the variability of thick terms. The gist of my proposal is that both phenomena—typically treated as mutually independent—stem from a fairly common mechanism in language use, namely, echo. I devote special attention to reclamation and defend three main claims about (i) how reclamation often involves multiple phases, (ii) the role of in-groupness, and (iii) the empowering effects of irony.
In a nutshell, my analysis of HEs, as presented in part I, is the following: Both slurs and thick terms pick out a certain descriptive property and at the same time they trigger an evaluation of that content (let us call it HE-evaluation). For example, ‘wop’ means ‘Italian,’ but at the same time it triggers the presupposition that Italians are bad because of being Italian; ‘lewd’ means ‘sexually explicit beyond conventional boundaries’ and triggers the presupposition that things or individuals that are sexually explicit beyond conventional boundaries are bad because of being so, etc. Slurs have a stronger projective power than thick terms; this fact can be explained by appealing to certain features of their descriptive content, together with extra-linguistic factors.
Part II is a critical review of alternative theories put forward to account for slurs and thick terms respectively. This overview of the competing approaches is not only meant to survey the existing accounts and argue for the superiority of the presuppositional theory; it is also an occasion to spell out in greater detail my proposal, by comparing it to alternative views. Because slurs and thick terms have been mainly studied separately, the theories I consider often deal with only one of these classes and neglect the other. The overview I provide in this part is far from a complete survey of all the theories on the market (see Popa-Wyatt 2020). Rather, I select three families of accounts for their relevance to specific aspects of my own approach, while leaving aside others— including expressivist accounts of HEs, for instance—for the sake of brevity. The first two chapters of part II are dedicated to “single-source” approaches, as Jeshion (2016b) calls them: truth-conditional approaches and deflationary approaches. In chapter 6, I first discuss the truth-conditional analysis of slurs, focusing on Hom (2008, 2010, 2012) and Hom and May (2013, 2014, 2018), and then consider the truth-conditional theory of thick terms, formulated by Kyle (2013). The main problem with truth-conditional approaches is to account for the projection of the evaluative content. I show that the strategies employed to explain projection without appealing to non-truth-conditional meaning ultimately fail. In particular, I show that (i) Hom and May’s data in support of a non-pejorative interpretation of embedded slurs actually stem from meta-linguistic readings, and that (ii) pragmatic mechanisms (both Hom and May’s offense and Kyle’s conversational implicatures) cannot explain projection data. Therefore, the truth-conditional analysis of HEs fails to explain the basic data and should be abandoned. The intuition that I share with these approaches is that slurs and thick terms linguistically encode evaluative contents.
Chapter 7 is dedicated to deflationary accounts of slurs and thick terms, which in contrast deny this intuition and defend the opposite view. First I focus on derogatory epithets, by discussing the so-called deflationary accounts, according to which slurs have a fairly unexceptional semantics. I include in this category all those accounts according to which the derogatory content associated with slurs is not part of their encoded meaning (again, broadly understood: not just truth conditions but also conventional implicatures, presuppositions and the like). I present four approaches that employ different tools: taboo and prohibitions (Anderson and Lepore 2013a, 2013b), contrastive preferences (Bolinger 2017), manner implicatures (Nunberg 2018) and subordinating speech acts (Langton, Haslanger and Anderson 2012). The first three accounts have difficulties in explaining the genesis of derogatory epithets, the complicity phenomenon (i.e. the fact that slurs’ pejorative content does not seem to be ascribed to the speaker only ), and the contrast between slurs and other ‘affiliatory’ terms. The speech act account, however, has trouble dealing with the so-called authority problem, as well as difficulties in showing that the derogatory content of slurs is not lexically encoded. In the second part of the chapter, I discuss the deflationary account of thick terms, developed by Pekka Väyrynen, which holds that the evaluation associated with thick terms (T-evaluation in Väyrynen’s terms) conversationally arises as a pragmatic implication. These implications are not part of the asserted content, nor are they the main point of the utterance in literal uses of thick terms in normal contexts: they are typically backgrounded. The pragmatic account has the advantage of being more parsimonious than any other account of thick terms, but it fails to recognize some instances of projection of the evaluative content (the alleged defeasibility data). In sum, in the first two chapters of part II, I show that single-source approaches fail to properly account for the behavior of HEs and I argue that the difficulties that these proposals have speak in favor of a hybrid approach.
In chapter 8, I discuss an alternative hybrid approach: the conventional implicature (CI) view. It agrees with truth-conditional accounts that the derogatory content of epithets is linguistically encoded, but it holds with pragmatic theories that the pejorative content does not belong to the truth-conditional dimension of meaning (Potts 2005, 2007c, Williamson 2009, McCready 2010, Gutzmann 2011, Williamson 2009, Whiting 2013). While the previous approaches I present have been independently developed for both slurs and thick terms, the CI view has been proposed only for slurs, 1 since the alleged defeasibility of the evaluative content of thick terms made scholars discard a conventional implicature approach. After presenting the account and showing how similar it is to the presuppositional approach (presupposition and conventional implicature being “close neighbor[s] in the linguistic literature,” Abbott 2006: 2), I point out the aspects for which the presuppositional view ought to be preferred, also looking at some experimental literature.
The picture I sketch in this work is that HEs represent a device through which language implicitly conveys linguistically encoded evaluations. On my account, HEs rely on presuppositions, which are—in Chilton’s (2004) words—“at least one micro-mechanism in language use which contributes to the building of a consensual reality.” By employing these terms, we implicitly take for granted a certain moral perspective, a certain set of beliefs concerning what is good and what is bad (an “ethos,” as Gibbard 2003b calls it). We implicitly apply a certain lens to the world and expect everyone else to do the same. Because the presupposed content is presented as not open to discussion, if it is not challenged, it has the potential to shape contexts. In this sense, using HEs is a powerful tool through which language not only encodes evaluation but also is able to impose it. Talking about the stereotypes evoked by slurs, Nunberg (2018) talks about cognitive shortcuts that we employ to make sense of the world; I argue that this is true not just for slurs and stereotypes but also for HEs in general, as they are devices through which language can convey evaluations in a way that is both linguistically encoded and implicit
