but there’s nothing bad in this.B3.Madonna’s show wasn’t lewd, because there is no such thing.
A and B disagree about the application of the predicate ‘lewd’ to Madonna’s show: B1 grants that some things are sexually explicit beyond conventional boundaries and bad because of that, but thinks that Madonna’s show is not among such things. In contrast, B2 and B3 reject the very idea that being sexually explicit beyond conventional boundaries can be bad per se: B2 retrieves, articulates and rejects what A’s utterance presupposes, while B3 employs a metalinguistic negation. This metalinguistic strategy recalls Oscar Wilde’s famous reply in his trial (Foldy 1997: 8):
1 26.Carson: Do you consider that10 blasphemous?Oscar Wilde: (. . .) the word ‘blasphemous’ is not a word of mine.
The same happens with slurs:
1 27.A.Madonna is a wop.B.No, she isn’t.
B’s denial is not particularly effective in targeting the evaluative content of A’s utterance, or is at least ambiguous between saying that Madonna is not Italian and refusing to accept A’s utterance on evaluative grounds.11 It can be very tricky to respond to slurring utterances in a way that does not take their derogatory content on board. It is even more complicated than it is for thick terms, mainly because of the interplay between multiple factors. Consider these exchanges (from the COCA):
1 28.A.I understand Black culture. I grew up around black people all my life. If the truth be told, I probably know niggers better than you, Monsieur Delacroix.B.I would prefer you not use that word in my presence.
2 29.A.You a fag?B.(. . .) I’m a homosexual.
3 30.A.I heard he was a fag (. . .)B.Historians don’t generally put it quite that way, but yes, da Vinci was a homosexual.
The strategies employed in (28)–(30) do not contemplate the articulation of the implicit content: the speakers either prohibit the use of the slur altogether or correct it by substituting it with a non-derogatory counterpart. In pedagogical contexts, on the other hand, where people explain how to react to slurring utterances, we can find the proper articulations of the derogatory content of slurs (which are harder to find in standard exchanges):
1 31.What you just said was really inappropriate because you are implying that there is something wrong with being gay or lesbian when there isn’t.12
Why do speakers avoid articulating the presupposed content of slurs in typical conversations? For one thing, there is a crucial difference between rejecting descriptive and evaluative presuppositions. Consider (32) and (34) and their presuppositions π32 and π34. From now on, I use ‘?’ before an utterance to signal that it is odd or infelicitous, ‘??’ to signal that it is strongly infelicitous, and ‘*’ when it is redundant.
1 32.Lila stopped doing yoga.
2 π32.Lila used to do yoga.
3 33.Hey, wait a minute, I didn’t know Lila used to do yoga!
4 34.You finally realize you were dating a loser.
5 π34.You were dating a loser.
6 35.? Hey, wait a minute, I didn’t know that I was dating a loser!
7 36.Hey, you shouldn’t talk like that about my ex.
The presupposition conveyed by (32) describes a state of affairs (Lila used to do yoga), whereas the one that (34) triggers expresses a value judgment (the interlocutor’s ex is a loser). Imagine that in both of the conversations where (32) and (34) occur, the interlocutors do not share the triggered presupposition, nor are they willing to accept it. While it would be appropriate to respond to (32) along the lines of ‘Hey, wait a minute, I didn’t know that . . .’13 (as in 33), such answers are not optimal for rejecting non-shared evaluative content (as we see in 35). If the speaker of (34) takes for granted that his interlocutor’s ex is a loser (even though this presupposition is not shared), the addressee cannot reply with something along the lines of ‘Hey, wait a minute, I didn’t know that my ex is a loser.’14 Something like (36) would be more appropriate.
The take-home message is that when HEs introduce an evaluative presupposition, the question is not whether the conversation’s participants already believed it, but rather whether they are willing to accept it or not. In general, the ‘Hey, wait a minute, I didn’t know that . . .’ schema does not work well with evaluative contents, and this could make the strategies of articulation more complicated.
There are further factors that make the articulation of evaluative presuppositions particularly hard, especially where slurs are concerned. For one thing, it is both cognitively and socially costly to retrieve, explicitly articulate and reject a presupposition in a conversation (see Chilton 2004, 64); it is more convenient to just reject the use of a term rather than articulating the associated presupposition. Furthermore, speakers are not willing to bear the costs of articulating and rejecting the presupposed content, unless they deem it necessary (for instance, when the question under discussion is precisely why it is bad to use a slur, as in pedagogical contexts). Lastly, slurs are censored in most public contexts and their use can be prosecuted.15 It is only to be expected that when speakers want to reject a slurring term, they opt for a meta-linguistic solution, thus avoiding any involvement with that kind of language: they’d rather forbid the use of the term, or substitute it with a non-loaded counterpart, than explicitly articulate the presupposed content.
1.2.3 An Alternative Explanation of Projection
In the previous sections, I proposed to account for the projective behavior of slurs and thick terms by relying on well-known features of presuppositions. This explanation of the projection of the evaluative content does not rely on the nature of the evaluation itself, but on the way in which it is encoded, i.e. presuppositionally. In other words, the negative HE-evaluations of ‘wop’ and ‘lewd’ do not project because they are evaluative; they project because they are presupposed. This is a crucial difference between my account and the hybrid expressivist account of slurs developed by Jeshion (2013a, 2013b, 2014, 2016a, 2016b, 2017, 2018, 2020). According to the hybrid expressivist strategy, the derogatory content of slurs projects because it is expressive rather than descriptive. Jeshion analyzes slurs as truth-conditionally equivalent to their neutral counterparts. In her framework, ‘wop’ has the same truth- conditional meaning as ‘Italians’: utterances like ‘Leonardo da Vinci is a wop’ are strictly speaking true, utterances like ‘Leonardo da Vinci is a boche’ are strictly speaking false.16 Note that Jeshion’s account is not committed to the existence of a neutral counterpart for every epithet: all she needs to postulate is the existence of a group-defining property, something which turns out to be crucial for a uniform account of slurs and thick terms (see section 4.1). For Jeshion, what distinguishes non-loaded expressions (neutral counterparts or more complex non-loaded paraphrases) from epithets is that the semantics of slurs also includes—at a non -truth-conditional level—an additional element, i.e. an expressive component. In particular, slurs encode contempt.17 What ‘wop’ means is something like ‘Italian,’ pronounced with a contemptuous intonation, or accompanied by pejorative modifiers (such as ‘dirty,’ ‘rotten,’ etc.). Contempt is to be understood as the attitude that one holds towards those that one regards as inferior as persons. In Mason’s words, contempt should be understood as “presenting
