Oppressive Speech

Abstract I here present two different models of oppressive speech. My interest is not in how speech can cause oppression, but in how speech can actually be an act of oppression. As we shall see, a particular type of speech act, the exercitive, enacts permissibility facts. Since oppressive speech enacts permissibility facts that oppress, speech must be exercitive in order for it to be an act of oppression. In what follows, I distinguish between two sorts of exercitive speech acts (the standard exercitive and the covert exercitive) and I argue that each such exercitive affords a distinct model of oppressive speech. Notes 1I thank Sally Haslanger, Rae Langton, Ishani Maitra, Marion Smiley, Lynne Tirrell and two anonymous reviewers for helpful comments on earlier drafts. I also thank audiences at Brandeis University and the Workshop on Gender and Philosophy at MIT. Finally, I thank David Lewis. 2To say that such speech is highly protected is to say that any justification for its regulation must meet the heightened standard of strict scrutiny. This requires establishing that the state has a compelling interest in regulating what it proposes to regulate and that the proposed regulation is narrowly tailored to serve that interest. Strict scrutiny is the highest level of protection afforded by the First Amendment. For a discussion of various categories of speech and levels of scrutiny, see Schauer Citation1982. 3This distinction (between causing and constituting) is already familiar to those acquainted with, for example, MacKinnon's work on pornography. In that work, she argues that while pornography causes a variety of harms to women, the proper legal grounds for its regulation resides in the harm it constitutes1987; 1993. Langton Citation1993[ offers a speech act account of how pornography could constitute such harm (i.e. subordination) and the first model of oppressive speech offered here relies heavily on Langton's work. This parallel raises a question about the relation between subordination, on the one hand, and oppression, on the other. As we shall see, oppression is a systematic disadvantage suffered in virtue of one's membership in a socially marked group. Thus, if subordination is systematic and if it obtains because of that membership, then such subordination is also oppression. According to MacKinnon, pornography systematically subordinates women because they are women. As a result, pornography oppresses women, on MacKinnon's view. Thus, the speech act work on the alleged subordinating force of pornography also affords a model of oppressive speech. See, e.g., Langton [Citation1993, McGowan Citation2003. 4I thank Lynne Tirrell for helping to make this point explicit. 5Similarly, suppose that a husband routinely abuses his wife but that he does so in a gender-just society. In such a case, his actions constitute abuse but they are not acts of gender oppression per se. In order for his abuse to be oppressive, it must occur in a broader context where women are systematically disadvantaged in virtue of being women. 6Langton Citation1993 was the first to use speech acts to make sense of MacKinnon's claim that pornography (understood here as speech) constitutes harm. Our first model of oppressive speech (the standard exercitive) is modelled after Langton's use of Austin Citation1973[. 7In addition to Langton, several other theorists have since applied Austin's theory of speech acts to feminist claims about pornography [Hornsby Citation1995; McGowan Citation2005; West Citation2003 but few have applied it to racist hate speech. One noted exception is Altman Citation1993. 8What may well be confusing is that although the permissibility facts enacted by the employer's utterance are neither legal nor moral ones, the act of enacting of these permissibility facts is both illegal and immoral. 9While familiar examples of standard exercitives involve the explicit stating of the content of the permissibility fact enacted, I leave open the possibility that such exercitives might express that content in some other manner (e.g. via presupposition or implicature) Grice Citation1989. This definition also leaves open the possibility that standard exercitives are indirect speech acts. See Searle Citation1979. 10For the origin of this model, see fn. 6. 11Elsewhere I have argued that some racist hate speech may fit this model of racially oppressive speech. See McGowan ms.. 12Rules of accommodation adjust (within certain limits) so that what actually transpires counts as correct. Thus, any utterance that triggers such a rule, thereby enacts a change to the conversational 'score' and thus enacts a change to what is subsequently permissible in that particular conversation. Rules of accommodation and this notion of 'score' are introduced in Lewis Citation1983. 13There are further differences between standard exercitives and conversational exercitives. Here is but one example. With standard exercitives, one expresses (even if only via implicature) the content of the permissibility fact enacted, but this is not so with conversational exercitives. In the above example, I did not express the content of the permissibility fact (i.e. The proper referent of 'the house' is my house and it is hereby inappropriate to try to refer to any other house with the expression 'the house') I enacted. For a discussion of further differences, see McGowan Citation2004. 14See fn. 12. 15The case of presupposition is a bit tricky. One successfully presupposes so long as one's interlocutor does not immediately question that presupposition. For a discussion of this complication, see Lewis Citation1983: 234. 16Since we are here using 'rule' in a rather loose sense, any permissibility fact is a rule in this sense. Obviously, the converse is also true: any such rule is a permissibility fact. 17The distinction between g-rules and s-rules is probably not sharp and there is likely to be a complex feedback system between the two sorts of rules. (G-rules clearly affect which s-rules are enacted by any particular move. Moreover, s-rules collectively may well affect what the g-rules are.) Such details, although interesting, do not matter for present purposes. 18Plenty of norms are implicit. Consider, for instance, cultural norms regarding how much physical space to give a conversational partner. (Such norms differ, for example, between the United States and France.) Although such norms are operative, they are not made explicit. 19Certain aspects of this second model seem to be implicit in Langton and West 1999. There they argue (among other things) that pornography communicates messages covertly (e.g., via presupposition and implicature) and that the introduction of such messages thereby legitimates, in complex ways, the mistreatment of women. Since legitimation is a form of permissibility, Langton and West here hint at the covert exercitive force of (pornography as) a conversational contribution. Both that paper and my work rely heavily on Lewis Citation1983. 20Of course, this is not sufficient. The s-rules also depend on the particular moves made and they depend on what has transpired thus far in the rule-governed activity in question. 21Although only whites can make the move in question, the s-rule is enacted, not via an exercise of the authority of the speaker, but in virtue of triggering the g-rules of the system. For this reason, the uttering of 'splet' is a covert (rather than a standard) exercitive. 22If, however, the speakers are aware of this correlation and intentionally exploit it (so that the speaker intends to prohibit women from speaking by uttering 'blop'), then arguably the speaker is an oppressor. 23I do not mean to suggest, by this example, that sexist speech is limited to (or more prevalent among) the working class. Although the élite tend to sanitize their sexism, it is no less oppressive. Consider the following exchange that takes place in an élite men's club. Winston: And, how was last night? Charles: Success! In a sing-song voice. This is clearly an allusion to Dangerous Liaisons and hence clearly a signal of sexual conquest. Winston: I am so glad that your parents had the foresight to send you to the Chenter School. Charles: curious look Winston: Your sharing skills are superior as a result. smiling My turn! 24This depends, in large part, on John's accepting Steve's utterance as (conversationally) appropriate. Cf. fn. 15. 25It is an act of gender oppression, in part, because women are systematically disadvantaged in virtue of being women. It is not a legally actionable act of oppression, in part, because the state does not legitimately regulate such aspects of 'private' life. 26Of course, such signs are prohibited only when they enact (illegal) segregatory policy. Posting such signs by a public washing machine or exhibiting them in a museum would certainly not be illegal. 27Finally, suppose that the oppressive s-rule allegedly enacted by Steve's utterance was already operative in the conversation (and that employee lounge) before Steve's utterance. (Suppose, for example, that a similarly sexist comment had been made earlier in that particular conversation.) In such a case, it may seem that Steve's utterance cannot enact that s-rule exactly because that s-rule was already operative (and thus had already been enacted). To conclude this, though, is to overlook something important about the ontological status of permissibility facts, in particular, and social constructions, in general. Because social constructions ontologically depend on (a complex form of) collective recognition, all social constructions (not just permissibility facts) are re-enacted at each and every moment at which they obtain. If, for example, we all ceased to regard a certain permissibility fact as obtaining, it would cease to obtain. The collective recognition of that fact (at time t) makes it the case that it obtains (at time t). Thus, whatever constitutes the collective recognition of a permissibility fact at a time enacts the permissibility fact at that time. Arguably, Steve's utterance does just that (even in a case where the s-rule in question is already operative). Since his utterance is part of what makes it the case that the permissibility fact in question is regarded as obtaining at that time, his utterance enacts that fact at that time. For a helpful discussion of social construction, see Searle Citation1995. This perpetual re-enactment of social constructions may well be what Butler Citation1997 means by reiteration. 28Of course, there is a trivial sense in which such speakers have such authority. In virtue of being a participant in the rule-governed activity in question, each participant has authority over how that rule-governed activity evolves in the case at hand. The baseball player may not have authority over the rules of baseball as such but he does have authority over how this particular baseball game is played. He has such authority in virtue of being a participant in the game in question. 29As discussed in fn. 27, s-rules and systems of oppression are social constructions and since such constructions are ontologically dependent on collective recognition, there is an important sense in which they are enacted at each and every moment at which they exist. In this way, perpetuation is, in fact, a form of enactment. 30The phenomenon identified here could have consequences for the law. It is possible that, under some circumstances, sexist remarks might constitute an otherwise illegal act. Suppose, for example, that Steve's utterance enacts an s-rule such that women are prohibited from the employee lounge. In such a case, Steve's utterance enacts an otherwise illegal discriminatory policy.