Daimon. Revista Internacional de Filosofía, nº 84, 2021 pp. 31-46

ISSN: 1130-0507 (papel) y 1989-4651 (electrónico)

http://dx.doi.org/10.6018/daimon.481381

Licencia Creative Commons Reconocimiento-NoComercial-SinObraDerivada 3.0 España (texto legal). Se pueden copiar, usar, difundir, transmitir y exponer públicamente, siempre que: i) se cite la autoría y la fuente original de su publicación (revista, editorial y URL de la obra); ii) no se usen para fines comerciales; iii) se mencione la existencia y especificaciones de esta licencia de uso.

 

The resistant effect of slurs:
A nonpropositional, presuppositional account*

 

El efecto resistente de los slurs:

Una propuesta presuposicional no proposicional

 

ALBA MORENO ZURITA**

EDUARDO PÉREZ-NAVARRO ***, ****

 

 

Abstract: The aim of this paper is to account for the resistance to cancelation, rejection, and retraction exhibited by slurs. The kind of explanation we offer is a presuppositional one. Like the most recent presuppositional accounts, moreover, ours is a nonpropositional presuppositional proposal. Our view is that, to be felicitous, utterances of sentences featuring slurs require certain components to be part of the common ground, but these components are not propositions, but world-orderings.

Keywords: slurs, resistance, cancelation, rejection, retraction, presupposition

 

Resumen: El propósito de este artículo es dar cuenta de la resistencia a la cancelación, el rechazo y la retractación exhibida por los slurs. La explicación que ofrecemos es presuposicional. Como las teorías presuposicionales más recientes, además, la nuestra es una propuesta presuposicional no proposiciona. Nuestra posición es que, para ser exitosas, las proferencias de oraciones que contienen slurs requieren que el common ground incluya ciertos componentes, pero estos componentes no son proposiciones, sino ordenamientos de mundos.

Palabras clave: slurs, resistencia, cancelación, rechazo, retractación, presuposición

 

---

Recibido: 28/05/2021. Aceptado: 01/07/2021.

* This paper has been funded by the Spanish Ministry of Science and Innovation under the research project “Disagreement in Attitudes: Normativity, Affective Polarization and Disagreement” (PID2019-109764RB-I00), by the Regional Government of Andalusia under the research projects “Public Disagreements, Affective Polarization and Immigration in Andalusia” (B-HUM-459-UGR18) and “The Inferential Identification of Propositions: A Reconsideration of Classical Dichotomies in Metaphysics, Semantics and Pragmatics” (P18-FR-2907), and by the University of Granada under a “Contrato Puente” fellowship and the excellence unit FiloLab-UGR (UCE.PPP2017.04). The authors would also like to thank Alex Davies, María José Frápolli, Claudia Picazo, Andrés Soria, and two anonymous reviewers for Daimon for their helpful comments and suggestions.

** PhD candidate at the Department of Philosophy I, University of Granada. E-mail: almorenozurita@gmail.com. Her interest areas are philosophy of language and social epistemology. A recent publication of hers is Almagro, M. & Moreno, A. (forthcoming), “Affective polarization and testimonial and discursive injustice”, in D. Bordonaba, V. Fernández Castro and J. R. Torices (eds.), The Political Turn in Analytic Philosophy: Reflections on Social Injustice and Oppressions, Berlin: De Gruyter.

*** Postdoctoral researcher at the Department of Philosophy I, University of Granada. E-mail: edperez@ugr.es. His interest areas are philosophy of language and philosophy of logic. Two recent publications of his are Pérez-Navarro, E. (2021), “The way things go: Moral relativism and suspension of judgment”, Philosophical Studies, DOI: 10.1007/s11098-021-01650-z, and Pérez-Navarro, E. (٢٠٢١), “No matter who: What makes one a relativist?”, Theoria: An International Journal for Theory, History and Foundations of Science, 36(2), 231–242.

**** The alphabetical order of the authors is meant to reflect their equal contribution to this paper.

 

 

1. Introduction

 

Sentences featuring slurs exhibit peculiar behavior. While there are a number of ways in which the effect of other speech acts can be blocked, exporting these strategies to avoid the harmful consequences of using a slur tends to be pointless. In particular, while one can usually cancel, reject, or retract the effect of other speech acts, these moves seem quite more difficult to make when the uttered sentence includes a slur.1 Compare the following sentences, in which “S” replaces a slur:

 

(1) The cat is on the mat.

(2) A is an S.

(3) If the cat is on the mat, we should vacuum it before my parents arrive.

(4) If A is an S, she shouldn’t go to that restaurant.

(5) No, it’s not.

(6) No, she’s not.

(7) I take that back. (Targeting one’s previous utterance of (1).)

(8) I take that back. (Targeting one’s previous utterance of (2).)

 

As we will see in due time, while (3) does not commit the speaker to the proposition that the cat is on the mat like (1) would, (4) is as derogatory as (2). This is what it means for the effect of slurs to be difficult to cancel. While (5) serves to reject the proposition that the cat is on the mat, one cannot successfully reject the derogatory effect of (2) by uttering (6). This is what we mean when we say that the effect of slurs is difficult to reject. Finally, while the speaker can successfully remove from the common ground the previously introduced proposition that the cat is on the mat by uttering (7), (8) does not seem as successful when the aim is to go back to a conversational state prior to the utterance of (2). That is, the effect of slurs is difficult to retract.

The aim of this paper is to account, building upon Marques and García-Carpintero (2020), for a feature of slurs that is meant to encompass this behavior—we say that their effect is resistant, either to cancelation, rejection, or retraction. This seems like a project worth embarking on, since the resistant character of slurs is part of what makes these words so harmful. Resistance to cancelation has been widely discussed (Jeshion, 2013a; Sennett and Copp, 2015; Cepollaro and Stojanovic, 2016; Camp, 2018; Marques and García-Carpintero, 2020), and resistance to rejection has been addressed to a certain extent in the literature too (Camp, 2013; Jeshion, 2013a, 2013b; Cepollaro and Stojanovic, 2016). Our explanation will be similar to previous ones given for resistance to cancelation, but we will also get inspiration from these strategies to account for resistance to rejection in a novel way. This paper also breaks new ground insofar as it focuses on resistance to retraction, a largely underexplored aspect of the behavior of slurs (see, however, McGowan, 2009; Simpson, 2013). Still, we take the work done here to be a programmatic effort, in need of further development in future contributions.

The kind of explanation we offer is a presuppositional one. Like the most recent presuppositional accounts, moreover, ours is a nonpropositional presuppositional proposal. To be felicitous, utterances of sentences featuring slurs require certain components to be part of the common ground, but these components are not propositions. As advanced above, this is inspired by Marques and García-Carpintero’s (2020) work, according to which it is reactive attitudes taken to be appropriate that the common ground needs to be enriched with. We follow Marques and García-Carpintero in using a nonpropositional account to explain why the effect of slurs persists in a range of constructions, as presuppositions do, but also in environments in which presuppositions are not expected to survive. We also use it to account for slurs’ resistance to rejection and retraction. However, our explanation only depends on one aspect of Marques and García-Carpintero’s proposal—the idea that the presuppositions involved are nonpropositional. We do not need them to concern reactive attitudes, and in fact will offer some reasons to understand them in terms of world-orderings instead. By doing so, we expect to provide an account of nonpropositionality that has independent interest.

The structure of the paper is as follows. In section 2, we explore what it means to say that the effect of slurs is resistant. We contrast the effect of sentences featuring slurs with that of other sentences and find that it is much more difficult to cancel, reject, and retract. In section 3, we survey a number of proposals that have relied on presuppositions to account for the meaning of slurs and how they have evolved to explain what makes the presuppositions triggered by slurs special. In our view, however, no account has satisfactorily done this until Marques and García-Carpintero’s, which we discuss in section 4 together with our proposed modifications and our reasons to incorporate them. Sections 5–7 are devoted to showing how the proposal resulting from section 4 can account for the resistant effect of slurs. In particular, section 5 explain slurs’ resistance to cancelation along Marques and García-Carpintero’s terms, while section 6 extends this explanation to flesh out Cepollaro and Stojanovic’s (2016) account of slurs’ resistance to rejection and section 7 applies it to slurs’ resistance to retraction.

 

2. The resistant effect of slurs

 

In this section, we describe the peculiar behavior of slurs, for which it is the aim of this paper to account. In particular, we characterize the derogatory effect of slurs as difficult to cancel, reject, and retract. We will summarize these features of slurs by saying that their effect is resistant.

There are a number of places where communication can go awry. Suppose I say

 

(1) The cat is on the mat.

 

The effect of this assertion is to add to the common ground the proposition that the cat is on the mat. This happens in two steps. In the first place, by uttering (1), I propose to add to the common ground the proposition that the cat is on the mat. In the second place, the proposition is added if my audience accepts my proposal.

Correspondingly, there are three ways in which I could utter the words in (1) without them having their default effect. First, something could go wrong before even the first step takes place. If the words are uttered in a linguistic environment that cancels their effect, the proposal to add to the common ground the proposition that the cat is on the mat will not even be made. This is what happens, for instance, when the words appear in the antecedent of a conditional, as they do in

 

(3) If the cat is on the mat, we should vacuum it before my parents arrive.

 

We summarize this by saying that the effect of an assertion is cancelable. Second, something could go wrong between the first and the second step. In particular, it could be that my audience does not accept my proposal. This is what happens when they reply to my assertion with something like (5) “No, it’s not”. We summarize this by saying that the effect of an assertion is rejectable. Lastly, something could go wrong after the second step. This is what happens when, after the proposition that the cat is on the mat has gone into the common ground, I say something like (7) “I take that back”. If this later assertion is accepted by my audience, the proposition will be taken out of the common ground. We summarize this by saying that the effect of an assertion is retractable.

It is worth noting that, in spite of the order in which we have introduced cancelation, rejection, and retraction, the first and the last of these phenomena can be seen as closer to each other than to rejection. After all, they are things that the speaker does, while rejection is done by the hearer. But note too that rejection and retraction target a certain content, while cancelation aims precisely at preventing such content from being conveyed. It is also worth mentioning that Caponetto (2020) advances a different taxonomy of ways of blocking the effect of a speech act. She discusses retraction like we do (Caponetto, 2020, 2407–10) and her amendment (Caponetto, 2020, 2410–12) bears a (passing) resemblance with our cancelation, but instead of rejection she considers annulment (Caponetto, 2020, 2404–6), which seems a wholly different phenomenon.

The picture above seems to work just fine with ordinary assertions. There are cases, however, in which cancelation, rejection, and retraction seem a bit harder. In particular, the derogatory effect of slurs, however it takes place, seems harder to cancel, reject and retract than the effect of an assertion. It is present even when they occur in the antecedent of a conditional, as in

 

(4) If A is an S, she shouldn’t go to that restaurant.

 

And also when the sentence in which they appear is negated or posed in the form of a question:

 

(9) A is not an S.

(10) Is A an S?

 

Slurs are also difficult to reject—it is not obvious that replying to a predicative use of a slur with (6) “No, she’s not” directly addresses its derogatory effect, and trying to do so is commonly felt to interrupt the flow of the conversation (Chilton, 2004, 64; Stanley, 2015, 170). And they are difficult to retract—derogation does not seem able to be undone by saying (8) “I take that back”.2

We summarize this by saying that the derogatory effect of slurs is resistant. The first two ways in which it is so—especially its resistance to cancelation—have been considerably explored in the literature (see Jeshion, 2013a; Sennett and Copp, 2015; Cepollaro and Stojanovic, 2016; Camp, 2018; Marques and García-Carpintero, 2020 for cancelation and Camp, 2013; Jeshion, 2013a, 2013b; Cepollaro and Stojanovic, 2016 for rejection). Slurs’ resistance to retraction, however, has gone relatively unacknowledged (see, however, McGowan, 2009; Simpson, 2013). In this paper, we advance an explanation of it, but our proposal is intended to account for slurs’ resistance to cancelation and rejection as well. The proposal is a presuppositional one. In the next section, accordingly, we survey this family of views.

 

3. Presuppositional accounts of slurs

 

In the previous section, we saw that the derogatory effect of slurs seems to be resistant to cancelation, rejection, and retraction. In the next section, we propose an account of the meaning of slurs that, as we will devote the rest of the paper to argue, explains this set of features. This is a presuppositional account of the meaning of slurs. There have been a number of proposals that have aimed at accounting for the meaning of slurs in presuppositional terms. Only the most recent of them, however, have tried to combine the presuppositional treatment with an explanation of what makes the derogatory content of slurs evaluative. We will follow the latter path, characterizing derogatory content as nonpropositional, presupposed content. Before doing this, however, we will devote this section to surveying how presuppositional accounts have evolved to this point.

The presuppositional account of the meaning of slurs can be traced back to Macià (2002). According to Macià, sentences featuring expressives trigger presuppositions; thus, they determine partial functions from worlds to truth-values, defined only for those worlds in which the sentence’s presuppositions are true. For instance,

 

(11) Toni Morrison is a Nobel laureate in spite of being African American.

 

determines a partial function that assigns truth to all worlds in which Toni Morrison is a Nobel laureate and is an African American, but only among the worlds in which African Americans are unlikely to win a Nobel prize. Making the sentence presuppose this is what “in spite of” does. As expressives, slurs will behave in a similar way.

It should be noted that utterances of sentences whose presuppositions are false in the context are infelicitous, but the presupposition can be made true in the context as the utterance is made. In cases like this, a process of accommodation is said to take place (Lewis, 1979). This means that the context is modified in any way needed for the utterance to be felicitous. In this case, in particular, the presupposition at issue would automatically be added to the context. The fact that this addition is automatic will play a fundamental role when we use presuppositions to account for the resistant character of the derogatory effect of slurs.

It is in Schlenker’s (2007) work that, for the first time, we find an account specifically targeting slurs (along with other expressives). Schlenker takes sentences featuring slurs to trigger presuppositions that have the following features (2007, 237). First, they are semantic presuppositions, as opposed to pragmatic ones, insofar as it is the meaning of the slur that triggers them. Second, they are indexical presuppositions, for they are bound to the context of utterance. Third, they have an attitudinal component, i.e., they make reference to the agent’s mental state. Finally, the agent whose mental state the proposition refers to does not necessarily have to be the speaker. Unlike in Macià’s view, these four features together make the presuppositions triggered by slurs expressive. Thus, if someone utters (2) “A is an S” and it is presupposed in the context that the agent believes that Ns are bad, where “N” is the neutral counterpart for “S”, the sentence will be true if A is an N and false otherwise. If there is no such presupposition in the context, however, (2) will lack truth-value.

Williamson (2009) and Nunberg (2018) find Schlenker’s proposal defective for the following reason. If, previously to the utterance of (2), it is not presupposed in the context that the agent believes that Ns are bad, the presupposition will be accommodated and become part of the common ground. However, it is problematic to accept an utterance of (2), while there should be no problem with accepting that a certain agent has a certain belief (Williamson, 2009, 151–2; Nunberg, 2018, 284; see also Marques and García-Carpintero, 2020, 142–3).

Objections like this have driven some authors to propose to specify the content of the presupposition triggered by the utterance of a slur in ways other than Schlenker’s. Cepollaro and Stojanovic (2016), first, and later Cepollaro (2020) in a more developed form, coincide with Schlenker in deeming these presuppositions different from descriptive ones,3 but render them as the mere presupposition that Ns are bad because of being Ns (Cepollaro and Stojanovic, 2016, 459). When we utter (2), therefore, we express two kinds of content. On the one hand, we express a descriptive content, which determines the truth-conditions of the sentence. This descriptive content is just like that of “A is an N”. On the other hand, we express an expressive content that is presupposed and can be rendered as “Ns are bad because of being Ns”. Similarly to Schlenker’s proposal, the sentence will be true or false depending on whether A is an N only if it is presupposed that Ns are bad because of being Ns. If it is not, it will lack truth-value.

 

4. A nonpropositional, presuppositional account of slurs

 

Cepollaro and Stojanovic avoid the problem faced by Schlenker’s presuppositions. However, they do so at the cost of leaving the attitudinal component out of the picture. In this way, they leave quite unexplained what makes their presuppositions expressive, other than being able to be made explicit through sentences that include evaluative terms like “bad”. Marques and García-Carpintero (2020) propose a presuppositional account that fleshes out what it means for a presupposition to be expressive while avoiding Schlenker’s problem. To do this, they need to extend the common ground to include not only possible worlds, but also other components. This goes in line with how the common ground has been proposed to incorporate partitions to account for the effect of questions (Stalnaker, 2014) or to-do lists to account for the effect of directives (Portner, 2007).

What Marques and García-Carpintero propose to add to the common ground, in particular, is a set of reactive attitudes (Strawson, 1962) that are taken to be appropriate by the participants in the conversation. Reactive attitudes are emotions that, as such, consist of two aspects—the intentional object that is the target of the attitude and the formal object or attitude that is directed toward that object (Kenny, 1963). We could thus have as part of the common ground a set that includes the reactive attitude of contempt toward a given group. Such an attitude will then be taken to be accepted by participants in the conversation. For an utterance of (2) “A is an S” to be felicitous, the set of reactive attitudes in the common ground needs to include contempt toward the Ns. If it does, whether the sentence is true or false will depend on whether A is an N. If it does not, the sentence will lack truth-value.

Marques and García-Carpintero’s view manages to put together two features that we find worth taking into consideration, and which might seem difficult to combine. On the one hand, by situating the derogatory content of slurs at the presuppositional level, they account for its elusive character, which is accentuated by the nonpropositional character of this content. As we will see in the next three sections, this makes it possible to explain the resistance that slurs exhibit to cancelation, rejection, and retraction. On the other hand, the fact that they rely on semantic presuppositions allows them not to lower the degree of responsibility that a speaker takes in using a slur. The problematic presuppositions are triggered by the meaning of the word; they are not an unexpected consequence at the end of the causal chain. Thus, it is in using a slur that the speaker is harming the target group, and she can be said to be responsible for this. Additionally, Marques and García-Carpintero contribute to the discussion by fleshing out what it means for content to be expressive. By doing so in a nonpropositional way, they account for slurs’ special connection with action—in particular, for their derogatory character, which is what makes them dangerous.

We think, however, that modeling the effect of accommodating a slur in the common ground in the way in which Marques and García-Carpintero do can lead to a psychologistic interpretation. A way of understanding what it means for reactive attitudes to be part of the common ground is to take speakers to share them, at least for the purposes of the conversation. This is what drives us to characterize this proposal as psychologistic, since it makes the derogatory effect of slurs depend on the particular attitudes of speakers. Once the derogatory effect of slurs is characterized in this way, we might be inclined to think that situations of injustice stem from implicit attitudes and, in prioritizing the latter as an explanatory element, make intervention measures depend on them too.

An alternative way of modeling the effect of accommodating a slur in a way that is more akin to structural explanations of what makes them dangerous (Haslanger, 2016) is through world-orderings. This move is inspired in Stanley’s (2015) account of political propaganda, which relies on the distinction between at-issue content and not-at-issue content. At-issue content is that understood as proposed to be added to the common ground (Potts, 2007, 666), while not-at-issue content is directly added to the common ground, without any need for the audience to accept it (Stanley, 2015, 135). In fact, the latter is added to the common ground without us having the chance to decide whether we accept it or not. According to Stanley, propagandistic discourse features not-at-issue content whose effect can be modeled along the lines of Starr’s (2020) account of imperatives. Starr models the effect that imperatives have on the common ground as that of ordering the possible worlds in it in accordance with their preferability. For instance, if Víctor commands his child Pablo not to eat any more cookies, the worlds in the common ground are ordered so that those in which Pablo stops eating cookies are ranked as preferable to those in which he does not.

Stanley takes the relevant order to have to do with socialization, and says that, when accommodated, slurs rank those worlds in which the speaker socializes less with the target group above those in which she socializes more. This account suffices for the purposes of this paper. Hence, our view is that sentences featuring slurs presuppose that the worlds in the common ground are ranked so that they are more preferable the less contact the speaker has with the target group. If no such ordering is in place, the worlds will be rearranged according to it to accommodate the use of the slur. Presuppositions are a variety of not-at-issue content: when they are accommodated, they are directly added to the common ground, with no proposal involved, so participants in the conversation do not have the chance to accept or reject them. Thus, we do not depart too much from Stanley’s account; if anything, we flesh it out a little.

An additional argument for understanding the effect of slurs in terms of world-orderings rather than reactive attitudes is that it results in a more parsimonious ontology. World-orderings also serve to model the effect of directives, while we know of no speech act, other than utterances of sentences featuring slurs, whose effect we can model in terms of reactive attitudes.4 By modeling the effect of slurs in terms of world-orderings, we subsume these speech acts under the more general class of directives, even if the associated command is made in a back-handed way. Moreover, enriching the set of operations that utterances can perform on the common ground instead of enriching the common ground itself may be seen as a generally preferable move.

 

5. Resistance to cancelation

 

As we saw in section 2, the derogatory effect of slurs seems resistant to cancelation, rejection, and retraction. An account such as the one resulting from the previous section will allow us to explain why they are so. In this section, we begin by using the account to make sense of slurs’ resistance to cancelation.

The derogatory effect of slurs seems to project along a variety of contexts, which has driven many authors to defend that it is the result of a presupposition. The phenomenon of projection occurs whenever the sentence resulting from embedding a sentence in a complex context triggers the same presuppositions as the original sentence. Consider, for instance, the following sentences:

 

(12) Nerea has stopped going to jiu-jitsu classes.

(13) If Nerea has stopped going to jiu-jitsu classes, I’ve bought her this gi for nothing.

(14) Nerea hasn’t stopped going to jiu-jitsu classes.

(15) Has Nerea stopped going to jiu-jitsu classes?

 

(12) presupposes that Nerea used to go to jiu-jitsu classes, and so do (13–15). We say that the presupposition that Nerea used to go to jiu-jitsu classes projects when the sentence that triggers it is embedded as the antecedent of a conditional, negated, or posed in the form of a question. The fact that presuppositions behave in this way may lead us to think that a presuppositional account of slurs should be given, as their derogatory character seems to project in similar cases:

 

(2) A is an S.

(4) If A is an S, she shouldn’t go to that restaurant.

(9) A is not an S.

(10) Is A an S?

 

(4), (9), and (10) are just as derogatory as (2) is, just like (13–15) presupposed that Nerea used to go to jiu-jitsu classes just like (12) did. We could thus say that this happens because (2) triggers a certain presupposition that projects in (4), (9), and (10). In particular, we could say that (2) presupposes that Ns are bad because of being Ns (Cepollaro and Stojanovic, 2016, 459).

However, the projective behavior of slurs does not exactly coincide with that of presuppositions.5 Consider these two sentences:

 

(16) If Nerea used to go to jiu-jitsu classes, she has stopped doing so.

(17) If Ns are bad because of being Ns, A is an S.

 

These cases are structurally identical: in both of them, we have placed the sentence that is supposed to trigger a certain presupposition—(12) in (16), (2) in (17)—in the consequent of a conditional whose antecedent is a sentence expressing the presupposition allegedly triggered by the former sentence.6 However, they behave differently. (16) does not presuppose that Nerea used to go to jiu-jitsu classes, while (17) is still derogatory. If it is so in virtue of its triggering the presupposition that Ns are bad because of being Ns, it presupposes that Ns are bad because of being Ns just like (2) does. This feature has been called the hyperprojectivity of slurs (Jeshion, 2013a; Sennett and Copp, 2015; Cepollaro and Stojanovic, 2016; Camp, 2018; Marques and García-Carpintero, 2020). Thus, the derogatory content of slurs does not exactly behave like a presupposition, which speaks against explaining it in this way.

Since presuppositions do not project in contexts like (16), we can embed sentences that in some sense imply a certain content in contexts with that structure to know whether the content is presupposed or not. If the content does not project, it will be a presupposition; if it does, it will not. This is what has been called the binding test (van der Sandt, 1992).

Proponents of presuppositional accounts of slurs have offered different explanations that try to make these accounts compatible with the hyperprojectivity of slurs. Cepollaro and Stojanovic (2016, 480) rely on the distinction, made among others by Abusch (2002) and Abbott (2006), between soft and hard triggers, where the former give rise to presuppositions that can be canceled in some contexts, while the presuppositions that result from the latter project in contexts such as (17). For instance, gendered pronouns are hard triggers (Heim, 2008):

 

(18) If Nerea is a woman, she can’t do jiu-jitsu.

 

(18) presupposes that Nerea is a woman even if it is a conditional whose antecedent expresses exactly that presupposition. The fact that this content projects in (18) seems to constitute no obstacle to understanding the implication that Nerea is a woman as a presupposition. In the same way, Cepollaro and Stojanovic say, the fact that (17) is as derogatory as (2) should not drive us to reject a presuppositional account of slurs.

Cepollaro and Stojanovic’s reply to the hyperprojectivity objection, however, seems stipulatory to us. They just introduce the distinction between soft and hard triggers, but they are not clear as to why slurs and gendered pronouns belong to the latter category.7 We find Marques and García-Carpintero (2020, 149–50), by contrast, closer to offering a really explanatory account. Their point is that, when we place (2) in a context such as (17), we are not really applying the binding test to it. For the binding test to be correctly applied, the antecedent of (17) should express the presupposition in (2) (see McCready, 2010, 9). But, as we said in the previous section, this presupposition is not propositional, while the antecedent of (17) will necessarily have propositional form8—otherwise, (17) will be ungrammatical. Thus, no matter what we place in the antecedent of (17), it will not express the presupposition in (2). It should not worry us, therefore, if no sentence with the form of (17) is free of derogatory character. No such sentence will be the result of applying the binding test. Thus, we cannot say that slurs fail the binding test just because the presupposition that Ns are bad because of being Ns projects in (17).

Despite appearances, though, Cepollaro and Stojanovic’s and Marques and García-Carpintero’s explanations can be said to be complementary rather than incompatible. We could say that slurs are hard triggers because the presuppositions they trigger are nonpropositional. Moreover, Marques and García-Carpintero’s account can also give us a hint as to why gendered pronouns are hard triggers too. Rather than propositional presuppositions, gendered pronouns can be taken to trigger presuppositions concerning the assignment of values to the variables—they pose certain restrictions on this assignment for utterances of sentences containing them to be felicitous. Since we can only place sentences with propositional form in the antecedent of a conditional like (17), no conditional will allow us to apply the binding test to a sentence featuring a gendered pronoun.

 

6. Resistance to rejection

 

The derogatory effect of slurs is not only hard to cancel. It is also hard to reject. In this section, we flesh out what resistance to rejection amounts to and show how a nonpropositional, presuppositional account of slurs like the one we have advanced can make sense of this feature.

Remember (2):

 

(2) A is an S.

 

An utterance of (2) could be replied by saying

 

(6) No, she isn’t.

 

However, we cannot be sure that, by doing so, it is the derogatory effect of the utterance that we are rejecting. A natural interpretation of (6) is just that A is not an N, but this is compatible with our sharing whatever negative attitude toward Ns the speaker of (2) holds (see Stanley, 2015, 135–6). Something else is needed if we want to make sure that we are understood as rejecting this attitude.

This fact is smoothly accommodated by a presuppositional account like ours. When we say something like (6), we are targeting the content asserted by the speaker of (2). If the derogatory effect of this sentence is explained through its presuppositions, it is then natural that replies like (6) do not manage to block that effect, as it is not through replies like this that we target presuppositions.

Instead, presuppositions are usually assumed to be identifiable through the “Hey, wait a minute!” test (von Fintel, 2004). Remember (12):

 

(12) Nerea has stopped going to jiu-jitsu classes.

 

If the hearer does not share the assumption that Nerea used to go to jiu-jitsu classes, it is natural for her to reply with:

 

(19) Hey, wait a minute! I didn’t know Nerea used to go to jiu-jitsu classes.

 

This is a standard way in which presuppositions can be rejected. Cepollaro and Stojanovic (2016, 467), however, note that this kind of reply is not natural when it is an utterance of a slur that we are replying to. It would be weird, for instance, to reply to (2) with

 

(20) ?? Hey, wait a minute! I didn’t know Ns are bad because of being Ns.

 

or whatever formulation we have chosen for the presupposition triggered by (2). Instead, they say, a rejection is likely to go along the following lines:

 

(21) Hey, wait a minute! You shouldn’t talk about Ns like that.9

 

Cepollaro and Stojanovic take this to be “a crucial difference between rejecting descriptive and evaluative presuppositions” (Cepollaro and Stojanovic, 2016, 467). It seems to us, however, that, by saying this, Cepollaro and Stojanovic are again describing the behavior of evaluative presuppositions rather than explaining why they behave in this way. Fortunately, we think we can find an explanation in Marques and García-Carpintero’s (2020) view just like we found one for the projective behavior of slurs. Note, first, that “I didn’t know that” requires a propositional prejacent. Thus, no matter what we place after “I didn’t know that”, it will not express the presupposition in (2), which is not propositional. Replies like (20) will therefore always fail to target this presupposition.

But also, propositional presuppositions are meant to model the information shared by participants in a conversation as to how the world is. They serve to describe the stage at which participants are in a process of inquiry, which is aimed at knowing which of the candidate worlds we inhabit. Nonpropositional presuppositions are nothing like this. Suppose we take these presuppositions to concern the participants’ reactive attitudes, as Marques and García-Carpintero do. In what kind of process do they play a role? Coordination is a natural candidate: we could say that part of the aim of a conversation is for speakers to align their reactive attitudes (cf. Gibbard, 1990, 110), and expressive presuppositions allow us to characterize the point they are at. However, it is far from settled that this is the role that reactive attitudes play in conversations (Pérez Carballo and Santorio, 2016). At any rate, we can be quite certain that, whatever process we should be talking about, it is not a process of inquiry. Speakers do not aim at discovering what the appropriate reactive attitudes are. This explains that utterances of sentences like (2) cannot be challenged with replies like (20). If, instead of reactive attitudes, we use world-orderings to model the effect of slurs, we will reach a similar conclusion.

Replying to (2) with (21) is better, but it also comes at a cost. Insofar as (21) concerns how we should talk instead of what the speaker of (2) actually said, its utterance is likely to be interpreted as an attempt at changing the subject of the conversation and, as such, a sign of lack of cooperativity (Chilton, 2004, 64; Stanley, 2015, 170).

We take these considerations to explain why it is so hard to reject the derogatory effect of a slur. Direct negation, as in (6), does not target the presupposition, but standard ways of rejecting presuppositions, as in (20), do not seem to work either. The fact that the derogatory effect of slurs is explained through presuppositions and that these presuppositions are nonpropositional in character is what is needed to account for slurs’ resistance to rejection.

 

7. Resistance to retraction

 

The last and least discussed feature of slurs for which we should want our proposal to account is their resistance to retraction. Retraction is the phenomenon whereby a speaker says something like “I retract that” or “I take that back” (MacFarlane, 2014, 108). By doing so, according to MacFarlane, the speaker cancels the normative effects of the speech act to which “that” refers, which is the target of the speech act of retraction. For example, by retracting an offer we withdraw a permission we had previously extended. In the case of utterances featuring slurs, however, the normative effects associated with such words seem hard to cancel in this way. In other words, you cannot “unring a bell” once it has been rung (see McGowan, 2009, 403; Simpson, 2013, 570).

A presuppositional account like the one we have advanced in this paper can explain this fact along the lines drawn by our explanation of slurs’ resistance to rejection. When I say “I take that back”, what I am taking back is the asserted content, not the presupposed content.10 The expression “I take that back” is naturally read as elliptical for “I take what I said back”, not for “I take what I presupposed back”—while we can find actual occurrences of “I take what I said back” in ordinary conversations, “I take what I presupposed back” is a philosophical construct. If we say “I take that back” trying to retract a presupposition, all we will do is retract the assertion that triggered that proposition. Now, if the derogatory content of slurs is presupposed, this would explain why we cannot undo the damage associated with the utterance of a slur.

Note, however, that a sufficiently rich context might allow an utterance of “I take that back” to target the presupposition triggered by a previous utterance instead of the content asserted through it. Consider the following dialogue:

 

(12) Nerea has stopped going to jiu-jitsu classes.

(22) In fact, she signed up for them but didn’t get to attend a single class.

(23) Oh, is that so? I take that back then.

 

Given (22), it is natural to think that (23) does not target the content asserted by (12), but the presupposition that Nerea used to go to jiu-jitsu classes. By uttering (23), the speaker is not introducing back into the common ground worlds in which Nerea has kept going to jiu-jitsu classes, but worlds in which she has never even gone to them. These worlds had been removed from the common ground by the speaker’s utterance of (12) followed by a process of accommodation.

It is not clear, though, that this mechanism can be used to retract a nonpropositional presupposition like the ones we are interested in. Consider the following dialogue, built to be structurally parallel to the one presented above:

 

(2) A is an S.

(24) In fact, Ns are valuable members of our society.

(25) ?? Oh, is that so? I take that back then.

 

(25) feels weird if intended to target the content asserted by (2)—why would the fact that Ns are valuable members of our society speak against the claim that A is an N? But, if (25) is intended to target the presupposition triggered by (2), it is hard to see how it could achieve its purpose. Ns will have been derogated regardless of one’s efforts to “go back in time”.

Up to this point, we could be accused of what, in the previous two sections, we accused Cepollaro and Stojanovic of. We have described how slurs behave with respect to retraction, but we have not explained why they do so. As with rejection, however, we think that the fact that the presuppositions triggered by slurs are nonpropositional allows us to account for the way in which they differ from other presuppositions when it comes to retraction. What makes the context of the first dialogue allow (23) to target the presupposition of (12) is that, in (22), the interlocutor provides the speaker of (12) with certain information. Cooperative principles then obligate the speaker to reintroduce into the common ground the worlds excluded through presupposition accommodation. There is no information that the interlocutor of the second dialogue could offer, though, that automatically forced the speaker of (2) to cancel the nonpropositional presupposition triggered by her utterance. Anything the interlocutor says is strictly compatible with the speaker of (2)’s negative evaluation of Ns. Of course, knowing certain facts may drive the speaker to reconsider her position. But, unlike in the first dialogue, refusing to do so does not make her uncooperative.

What if, instead of (24), the interlocutor replies denying precisely the presupposition triggered by (2)? This would require this presupposition to be made explicit in propositional form. But, as we saw when discussing slurs’ resistance to cancelation, this is not possible, as the presupposition triggered by (2) is not propositional. Thus, presupposition and nonpropositionality together allow us to account for the difficulty of retracting a slur.

 

8. Conclusion

 

Following Cepollaro and Stojanovic (2016) and Marques and García-Carpintero (2020), we have offered an account of the meaning of slurs that, by relying on presuppositions, explains much of their behavior, and explains the behavior that ordinary presuppositions leave unexplained by understanding the presuppositions triggered by slurs as nonpropositional. As we said in section 4, we believe that a nonpropositional, presuppositional account of slurs allows us to make sense of two features of this kind of term that are in apparent tension. On the one hand, their effect is elusive—it resists being pointed out, a previous move that is needed whenever we want to block it. On the other hand, however, this effect is triggered by the meaning of the slur. It is what slurs do, so speakers can be said to be responsible for this effect. Inasmuch as nonpropositional presuppositions are triggered by what slurs mean but impossible to make explicit using declarative sentences, we think that they have all the features that whatever lies behind slurs’ derogatory effect should have. Of course, future work should offer a much more detailed account of how a proposal like ours would deal with each particular case in which the presuppositions triggered by slurs behave in nonstandard ways.

 

References

 

Abbott, B. (2006), “Where have some of the presuppositions gone”, in B. Birner & G. Ward (eds.), Drawing the Boundaries of Meaning: Neo-Gricean Studies in Pragmatics and Semantics in Honor of Laurence R. Horn, Philadelphia: John Benjamins, pp. 1–20.

Abusch, D. (2002), “Lexical alternatives as a source of pragmatic presuppositions”, in B. Jackson (ed.), Proceedings of SALT XII, Ithaca, NY: CLC Publications, pp. 1–20.

Camp, E. (2013), “Slurring perspectives”, Analytic Philosophy, 54(3), 330–349.

Camp, E. (2018), “A dual act analysis of slurs”, in D. Sosa (ed.), Bad Words: Philosophical Perspectives on Slurs, Oxford: Oxford University Press, pp. 29–59.

Caponetto, L. (2020), “Undoing things with words”, Synthese, 197(6), 2399–2414.

Cepollaro, B. (2020), Slurs and Thick Terms: When Language Encodes Values, Washington, DC: Rowman & Littlefield.

Cepollaro, B. & Stojanovic, I. (2016), “Hybrid evaluatives: In defense of a presuppositional account”, Grazer Philosophische Studien, 93(3), 458–488.

Chilton, P. (2004), Analysing Political Discourse: Theory and Practice, London: Routledge.

von Fintel, K. (2004), “Would you believe it? The King of France is back! Presuppositions and truth-value intuitions”, in M. Reimer & A. Bezuidenhout (eds.), Descriptions and Beyond: An Interdisciplinary Collection of Essays on Definite and Indefinite Descriptions and Other Related Phenomena, Oxford: Oxford University Press, pp. 315–341.

Gibbard, A. (1990), Wise Choices, Apt Feelings: A Theory of Normative Judgment, Cambridge, MA: Harvard University Press.

Haslanger, S. (2016), “What is a (social) structural explanation?”, Philosophical Studies, 173(1), 113–130.

Jeshion, R. (2013a), “Slurs and stereotypes”, Analytic Philosophy, 54(3), 314–329.

Jeshion, R. (2013b), “Expressivism and the offensiveness of slurs”, Philosophical Perspectives, 27(1), 231–259.

Kenny, A. (1963), Action, Emotion and Will, London: Routledge.

Lewis, D. (1979), “Scorekeeping in a language game”, Journal of Philosophical Logic, 8(1), 339–359.

MacFarlane, J. (2014), Assessment Sensitivity: Relative Truth and Its Applications, Oxford: Oxford University Press.

Macià, J. (2002), “Presuposición y significado expresivo”, Theoria: An International Journal for Theory, History and Foundations of Science, 3(45), 499–513.

Marques, T. & García-Carpintero, M. (2020), “Really expressive presuppositions and how to block them”, Grazer Philosophische Studien, 97(1), 138–158.

McCready, E. (2010), “Varieties of conventional implicature”, Semantics and Pragmatics, 3, article 8.

McGowan, M. K. (2009), “Oppressive speech”, Australasian Journal of Philosophy, 87(3), 389–407.

Nunberg, G. (2018), “Speech acts in discourse contexts”, in D. Fogal, D. W. Harris & M. Moss (eds.), New Work on Speech Acts, Oxford: Oxford University Press, pp. 237–295.

Pérez Carballo, A. & Santorio, P. (2016), “Communication for expressivists”, Ethics, 126(3), 607–635.

Portner, P. (2007), “Imperatives and modals”, Natural Language Semantics, 15(4), 351–383.

Potts, C. (2007), “The expressive dimension”, Theoretical Linguistics, 33(2), 165–198.

van der Sandt, R. A. (1992), “Presupposition projection as anaphora resolution”, Journal of Semantics, 9(4), 333–377.

Schlenker, P. (2007), “Expressive presuppositions”, Theoretical Linguistics, 33(2), 237–245.

Sennet, A. & Copp, D. (2015), “What kind of a mistake is it to use a slur?”, Philosophical Studies, 172(4), 1079–1104.

Simpson, R. M. (2013), “Un-ringing the bell: McGowan on oppressive speech and the asymmetric plyability of conversations”, Australasian Journal of Philosophy, 91(3), 555–575.

Stalnaker, R. (2014), Context, Oxford: Oxford University Press.

Stanley, J. (2015), How Propaganda Works, Princeton: Princeton University Press.

Starr, W. B. (2020), “A preference semantics for imperatives”, Semantics and Pragmatics, 13, article 6.

Strawson, P. F. (1962), “Freedom and resentment”, Proceedings of the British Academy, 48, 1–25.

Williamson, T. (2009), “Reference, inference, and the semantics of pejoratives”, in J. Almog & P. Leonardi (eds.), The Philosophy of David Kaplan, Oxford: Oxford University Press, pp. 137–158.

 

Notas