There is a large and growing literature on communal interpretive resources: the concepts, theories, narratives, and so on that a community draws on in interpreting its members and their world, also called “hermeneutical resources” or “epistemic resources.” In this article, I propose that we can advance this literature by considering the rules that govern social uses of interpretive resources.Footnote ¹ I develop the idea that there are rules governing uses of interpretive resources in section I. Section II and section III draw from the recent literature on dominant and resistant interpretive resources to suggest that some such rules in our community are shaped by the oppressive systems under which we live.Footnote ² Section II considers how the rankings of persons in oppression enable dominant group members to determine rules. Section III appeals to Charles Mills's The Racial Contract to derive rules that make it impermissible under oppression to use resistant resources. The conclusion enumerates several further points about the rules specific to dominant and resistant resources.

I. Resources and Rules

I follow Mary Kate McGowan in taking it that an activity is rule‐governed if “at least some behaviors (as contributions to the activity in question) would count as out of bounds or otherwise inappropriate (as contributions to the activity in question)” (Reference McGowanMcGowan 2009, 395). If we're playing simultaneous solitaire and you put a nine of clubs on a four of clubs in a foundation pile, you've made a contribution to the activity that is out of bounds. Why is some contribution to the activity “out of bounds”? Because it breaks the rules—it's against the rules of solitaire to put anything other than a five of clubs on a four of clubs in a foundation pile. If there are moves in the activity that are out of bounds, then there are rules governing the activity.

Interpretive resources are nearly ubiquitous for us, and we may be said to use them in many ways. I'm concerned here with those interpretive resources that we express with language; I'm concerned only with our linguistic expressions of them in the presence of others; and, I'm concerned with such expressions that serve to interpret the world. Paradigmatically, we use these resources this way in conversation. In conversation with Lin, Carmita uses the term sexual harassment to describe what she experienced at work. The term expresses an interpretive resource—presumably, the concept sexual harassment. I'm concerned here only with the resources that can be expressed with terms or strings of terms. Carmita's expression of the resource communicates to Lin a particular interpretation of a particular part of the world, namely, it interprets what Carmita experienced at work as an instance of sexual harassment. I'm concerned here only with such sincere and literal (rather than ironic or metaphorical) uses of terms and strings of terms that communicate to others a particular interpretation of a particular part of the world.

I'm concerned with individual concepts like sexual harassment as well as more complex interpretive resources like theories or narratives. Take, for example, Jennifer J. Freyd's theoretical construct expressed by the acronym DARVO—Deny, Accuse, Reverse Victim and Offender. The resource “refers to a reaction perpetrators of wrong doing, particularly sexual offenders, may display in response to being held accountable for their behavior.”Footnote ³ In an opinion published in the Boston Globe, Freyd and Louise F. Fitzgerald use language that expresses the interpretive resource to communicate an interpretation of the behaviors of Donald Trump (Reference Freyd and FitzgeraldFreyd and Fitzgerald 2017). Since the theoretical construct is used to interpret parts of the world, it is, prima facie, an interpretive resource. Since the resource is expressed in language to communicate an interpretation of the world, it's the sort of use of a resource that is the focus of this article.

In using resources this way, some moves are out of bounds by virtue of rules governing uses of terms. It's out of bounds to use the term fruit to refer to a graham cracker. It's out of bounds to use sexual harassment to refer to a graham cracker. It's out of bounds to use DARVO to interpret the behaviors of the moon. If McGowan is right, then, and an activity is rule‐governed if some moves in the activity are out of bounds, then using terms to communicate to others an interpretation of part of the world is a rule‐governed activity.Footnote ^4, Footnote ⁵

Where do these rules come from? I don't hope to give a complete answer here. I'll suggest, much more modestly, that in an oppressive system, these rules can be influenced by the system's rules. In order to see how, we'll need several clarifications about rules, rule‐governed activities, and the rule‐governed interpretation of the world.

(1) Not every behavior one performs while engaged in a rule‐governed activity is a move in that activity. If you utter a malapropism while we're playing simultaneous solitaire, your behavior may be impermissible, but it isn't impermissible according to the rules of solitaire. It isn't a move in solitaire at all. Many of the behaviors we perform while socially interpreting the world aren't moves in that activity: sneezing, sipping tea, walking, and so on.

(2) When interpreting the world with others, one is often also engaged in many other rule‐governed activities: spelling, using grammar, having a conversation, and so on. It doesn't follow that the rules governing these coincident activities also govern the application of terms (any more than the rules of English govern play in solitaire).Footnote ⁶ Nonetheless, it may happen that when rule‐governed activities overlap, their rules influence one another. Rules of etiquette at a posh dinner party, for instance, may make it impermissible to apply slang terms or graphic descriptions to parts of the world to which they would otherwise be appropriately applied. The dinner party rules influence the rules governing the application of certain terms.

(3) What is permissible in many (if not all) rule‐governed activities is a matter of the activity's rules and the state of play When the foundation pile was showing a four of clubs, it was impermissible to put a nine on top; but, eventually, when the eight of clubs was played, that made it permissible for me to play my nine. As the activity evolves, what was permissible may become impermissible and vice versa by virtue of moves that change the state of play—or, as David Lewis calls it, the activity's “score” (Reference LewisLewis 1979, 342–46). As McGowan puts it, “moves in rule‐governed activities enact changes in what is subsequently permissible in that activity” (Reference McGowanMcGowan 2009, 396).Footnote ⁷ For our purposes, a loose and intuitive understanding of “score” in general should suffice. In baseball, the components of score include not only runs for each team but also the inning, the half of the inning, balls, strikes, and outs. Different moves are permissible when there are four balls (a runner can walk to first) than when there are three. In conversation, components of score include “sets of presupposed propositions, boundaries between permissible and impermissible courses of action,” and so on (Reference LewisLewis 1979, 345). What's important for us is that moves in a rule‐governed activity change what is subsequently permissible. In socially interpreting the world with language, moves that change what is subsequently permissible include coining terms, refining definitions of extant terms, appropriating extant terms for different uses, and so on.

(4) There are rules of accommodation that govern conversation and, I propose, social interpretation. Lewis proposes that “conversational score does tend to evolve in such a way as is required in order to make whatever occurs count as correct play” (Reference LewisLewis 1979, 347). If I say, “let's talk quietly, my kids are asleep in the next room,” I've entered into the conversation the presupposition that I have children. The conversation hadn't presupposed it before, but it enters the score to accommodate my utterance. “Presupposition,” Lewis says, “evolves according to a rule of accommodation specifying that any presuppositions that are required by what is said straightaway come into existence, provided that nobody objects” (347). After my utterance, it is permissible to speak as though I have children, and it is impermissible to speak as though I don't or as though it is unknown whether I do. It is impermissible, for instance, to ask whether I have any kids.

When socially interpreting the world, the score also tends to evolve in such a way as is required to make linguistic usage count as correct play, provided nobody objects. My friend Amir once presented two boxes of pasta, one farfalle and one linguine, to a group and asked, “which kind of spaghetti do you prefer?” In context, in order for Amir's move to count as correct play in socially interpreting the world, it had to be either that the boxes had spaghetti in them (which all knew to be false) or that “spaghetti” has in its extension both farfalle and linguine, contrary to the usual rules for applying the term. Nobody objected or even remarked on it, and in the ensuing conversation, “spaghetti” was applied to pasta of all kinds. Prima facie, Amir's move triggered rules of accommodation, and once his use was accommodated, it became permissible in the conversation to apply “spaghetti” to pasta generally.Footnote ⁸ Such accommodations of nonstandard usage are common, suggesting that there are rules of accommodation governing social interpretation using language.

When a move in a rule‐governed activity changes what is subsequently permissible in that specific instance of the activity but not in instances of the activity generally, McGowan says that the move enacts an s‐rule. Rules that govern instances of the activity generally, by contrast, are “g‐rules” (Reference McGowanMcGowan 2009, 396). Accommodating Amir's use of “spaghetti” enacted an s‐rule, a rule that governed that one instance of social interpretation among friends. It did not enact a g‐rule making it permissible in English generally to use “spaghetti” as Amir did.

Of course, however, nonstandard uses do sometimes become standard, as with literally, disinterested, Madagascar, and so on. As McGowan notes, the distinction between s‐rules and g‐rules is “probably not sharp” and there is probably “a complex feedback system between the two” (396). We don't need to articulate the workings of this feedback system here, but it will be helpful to have a criterion to appeal to in claiming that there is some g‐rule corresponding to an observed s‐rule. I submit that if some s‐rule is enacted systematically, then there is a corresponding g‐rule. (This is meant to be a sufficient but not necessary condition.) If some move in social interpretation is systematically permitted, then there is a g‐rule making the move permissible.Footnote ⁹

(5) To charge that a move is out of bounds is itself a move in social interpretation. If I object to Amir's use of spaghetti, then his move doesn't straightaway enact the s‐rule. Moreover, in the usual case, I make it impermissible for Amir to ignore or dismiss my challenge without comment: He must either withdraw his move or give reasons to think it is correct play. These are both moves as well. If he withdraws his use, then my move—my objection—enacts an s‐rule making such use of “spaghetti” impermissible. If he adduces reasons to accept his usage, then, in the usual case, I must either withdraw my initial objection or give reasons for rejecting Amir's defense. And so on.

(6) In socially interpreting the world, participants sometimes have relative rankings, such that what is permissible for a participant with one rank may be impermissible for a participant with a different rank. Suppose Amir is in his first year studying English and I'm a fluent speaker tutoring him. In this case, it is permissible for me to correct Amir's usage, not just object to it. When I merely object, it is Amir's prerogative to withdraw his move or defend it; when I correct his use, the presumption is that his move is thereby withdrawn, though he may query my correction, protest, and, with enough evidence, have his move reinstated. On the other hand, it is impermissible for Amir to correct my usage. He may request explanation, express confusion, and so on, and, with effort, he may get me to withdraw a move I've made, but it is my prerogative to withdraw the move or not. This difference in our permissions is due to our difference in rank (at the time) when it comes to social interpretation using English. Differences in rank can confer a wide variety of differences in permissions; I explore only a few possibilities below.

There's much about these ranks that could fruitfully be clarified, but given space constraints, it's better for our purposes here to make just a few suggestive examples and notes. (i) Rankings are sensitive to the vocabulary in use and the parts of the world being interpreted in a context. If Amir is an expert chemist, and I'm not, then, ceteris paribus, it is permissible for him to correct my use of “alkaline” and not vice versa; in this context, he outranks me. Prima facie, if X ought to give semantic deference to Y for term T, then Y outranks X with respect to term T; and, if X ought to give semantic deference to Y when it comes to interpreting parts of the world p1–pn, then Y outranks X with respect to interpreting p1–pn. (ii) Rankings can be indeterminate. When it comes to interpreting our friendship, some considerations might suggest that you outrank me and others might suggest that I outrank you. There are some clear cases of rankings in social interpretation, but in many cases, things will be more complicated, and perhaps there will be no fact of the matter about interpreters’ relative rankings. I don't see that this undermines the usefulness of appealing to rankings in cases where they are relatively clear. (iii) Rankings may be negotiated in context. I may try to correct Amir and then have my attempt corrected by him. I thought I outranked him and was thus permitted to correct him, but he adduced reasons showing that I wasn't so permitted, and I withdrew my correction, revealing to us that he outranks me in this context. The rankings tend to evolve in such a way as is required in order to make whatever occurs count as correct play.

Finally, note that (iv) although the examples given so far suggest that rankings track epistemic positions, so that one with a better epistemic position will be ranked higher, this isn't always the case (see Reference FrancesFrances 2014 for discussion of what determines one's epistemic position). Institutional structures, for instance, can determine relative rankings as well. Literal military rankings make it impermissible in many contexts for a private to make a move in socially interpreting the world without first requesting permission from a sergeant, even if the private is in an epistemically better position. In the next section, I consider how the rankings of oppressive systems influence the rules governing social interpretation under those systems.

II. Resources, Rules, and Ranks

In this section and the next, I want to show that appealing to rules of social interpretation can be useful for articulating unjust social dynamics. Here, I'll articulate a previously unrecognized realization of hermeneutical injustice by pursuing a bottom‐up strategy—starting from an s‐rule and giving reasons to believe there is a corresponding g‐rule. In section III, I pursue a top‐down strategy, proceeding from g‐rules to their influence on individual instances of social interpretation.

First, a few disclaimers. I'll follow McGowan in saying that people are ranked under oppression; I'll follow Gaile Pohlhaus, Jr. and Rebecca Mason in broadly referring to dominant and nondominant group members; following Charles Mills and Robin DiAngelo, I'll talk about a racial system that ranks persons according to race; I'll also refer to a gender system that ranks persons by gender. The actual rankings of our oppressive system are much more complicated. The references to independent oppressive systems and to dominant and nondominant persons are abstractions that are useful at this early stage of analysis. White supremacy and patriarchy support and reinforce each other, and both interact further with other systems: our system is, as bell hooks says, “imperialist white‐supremacist capitalist patriarchy” (see, for example, hooks Reference hooks2004, 17). And persons aren't just dominant or nondominant; not all relatively nondominant ranks have the same prohibitions, not all relatively dominant ranks have the same permissions, and each of us occupies many social positions that trigger different permissions and prohibitions in different contexts. What I'll say here is in these ways too general and abstract to be accurate, but it is nonetheless helpful, I propose, insofar as it gives us a broad and sketchy “map of the terrain” of social‐interpretive rules under oppression. From here, we can add further refinements, fill in gaps, correct oversights, and so on.

Hermeneutical injustice is “the injustice of having some significant area of one's social experience obscured from collective understanding owing to persistent and wide‐ranging hermeneutical marginalization” (Reference FrickerFricker 2007, 154). Members of a group are hermeneutically marginalized when they are disadvantaged in making contributions to the resources available for interpreting some significant area of social experience (153). Fricker's well‐known primary example is a case of sexual harassment in the US in the 1970s—before the term sexual harassment had been coined. The US work force then, as now, was dominated by men. One aspect of this dominance was that the resources widely available for interpreting social interactions and agents at work were determined mostly by men and for the purposes of interpreting experiences typical of men. That is, women were hermeneutically marginalized with respect to the resources for interpreting workplace social interactions. As a result of this marginalization, women at the time suffered the injustice of having their experiences of sexual harassment obscured from collective understanding.

Fricker seems to take it that a group is hermeneutically marginalized if and only if they are disadvantaged in contributing to the available interpretive resources. The lacuna in hermeneutical resources where sexual harassment ought to have been is seemingly necessary for the hermeneutical injustice and marginalization in Fricker's lead example. On this conception of hermeneutical marginalization, social interpretive rules may realize hermeneutical marginalization only if they disable some group from contributing to hermeneutical resources. If there were a g‐rule making it impermissible for women in the workplace to introduce new terms or concepts, for instance, then this would realize women's hermeneutical marginalization in the workplace. If, however, there were g‐rules making it so that men are permitted to ignore any new terms or resources that women introduce, this wouldn't constitute women's hermeneutical marginalization on Fricker's conception because such a rule would not disable women from contributing to hermeneutical resources. Such g‐rules would allow women to develop hermeneutical resources, but it would disable them in their attempts to use those resources in communication.

It may be uncharitable to interpret Fricker's conception of hermeneutical marginalization so narrowly. In Epistemic Injustice, she most often gives a fairly broad and merely sufficient condition for hermeneutical marginalization. For example, “when there is unequal hermeneutical participation with respect to some significant area(s) of social experience, members of the disadvantaged group are hermeneutically marginalized” (Reference FrickerFricker 2007, 153). Prima facie, group members’ hermeneutical participation may be unequal even if they aren't disabled in contributing resources: they may be disabled in other aspects of hermeneutical participation. When we have in hand resources for talking about social interpretation as a rule‐governed activity, it's clear that there are many sets of rules that can disable a group's hermeneutical participation.Footnote ¹⁰ Social interpretive rules might prevent a group from developing hermeneutical resources, yes, but they might also prevent a group from communicating to others the resources they do develop; they might give dominant groups permission to ignore, change, or discount a marginalized group's resources; they might permit dominant groups to punish those who use a marginalized group's resources; and so on.

On the other hand, many of Fricker's remarks do suggest that an inability to contribute to shared hermeneutical resources is a necessary condition for hermeneutical marginalization. For example, “a hermeneutically marginalized subject is prevented from generating meanings pertaining to some areas of the social world …” (153–54). She focuses exclusively on how hermeneutical marginalization affects shared hermeneutical resources, resulting in lacunae (for example, 158) and prejudices (for example, 155) in the set of resources. Given this, it's helpful to broaden our understanding of hermeneutical marginalization and the ways it can, when persistent and wide‐ranging, obscure some significant area of a group's social experience from collective understanding. That is, it's helpful to explore other realizations of hermeneutical injustice.

Pohlhaus describes a social dynamic that, if systematic, would realize persistent and widespread hermeneutical marginalization. In Pohlhaus's case, interpretive resources developed by nondominant knowers—call these “nondominant resources”—are ignored or dismissed by dominant knowers. I say that these dismissals are social interpretive moves, and they enact s‐rules making it permissible for dominant knowers to dismiss or ignore nondominant interpretive resources. Pohlhaus claims that the dynamic she describes is near ubiquitous; if these situations systematically enact the relevant s‐rule, then by the criterion I suggested in section I, there is a corresponding g‐rule. Since this rule would make it permissible for dominant knowers to dismiss nondominant resources, it would present an obstacle to hermeneutical participation for those who would use nondominant resources that is not faced by dominant knowers using dominant resources. That is, wherever this g‐rule is followed, there would plausibly be systematic and widespread hermeneutical marginalization for nondominant knowers—not because of how the shared resources are, but because of how the rules of social interpretation are.

Many authors have criticized Fricker for seeming to conflate (a) a resource's being entirely unavailable in a community and (b) a resource's being unavailable in a community's dominant discourses or ideologies (Reference MasonMason 2011; Reference DotsonDotson 2011). Rebecca Mason charges that Fricker “glosses over … distinctions between dominant and nondominant hermeneutical resources” (Reference MasonMason 2011, 300). As a consequence, Mason alleges, Fricker takes it that if there is no resource adequate for interpreting some experience in a community's dominant discourses, then there is no resource for interpreting those experiences in the community at all. Similarly, on Kristie Dotson's reading, “Fricker seems to assume that there is but one set of collective hermeneutical resources” (Reference DotsonDotson 2011, 31). This fails to recognize the possibility (and often, the actuality) that marginalized communities do indeed develop resources that are adequate to their experiences, but dominant discourses fail to recognize these resources.

Pohlhaus suggests that this conflation obscures dominant knowers’ culpability for their ignorance of nondominant resources. In making the case that dominant knowers are so culpable, Pohlhaus describes the dynamic that interests us here. Pohlhaus says members of dominant groups may be presented with interpretive resources developed in and for nondominant communities and preemptively dismiss them. She offers the resources expressed by “white privilege,” “date rape,” and “heteronormativity” as examples.

Pohlhaus suggests that interactions like this are common. The hermeneutical injustice they instantiate is “much more ubiquitous in a world where marginally situated knowers do in fact know a great deal about their experienced world and have more venues for transmitting that knowledge to others” (732). She points to public receptions of then‐President Barack Obama's remarks on the arrest of Professor Henry Louis Gates, Jr., Justice Sonia Sotomayor's claim that “she would hope a wise Latina would make better judgments than a white male without her life experiences,” and quotations from Reverend Jeremiah Wright's sermons during the 2008 presidential campaign. “In all three cases, large numbers of epistemic agents drew the same distorted inferences from what was said by persons who were in some aspect of their social position marginally situated” (732–33). If interactions like this occur systematically in our community at present, and if systematically no one objects, then there is presently a g‐rule governing social interpretation that makes it permissible to dismiss or ignore interpretive resources developed by and for nondominant communities. Let me elaborate.

First, I take it that in a system without differences in rank, it would be impermissible in social interpretation for some participants to preemptively dismiss the interpretive moves of another. It's against the rules for you to just dismiss my play in solitaire, my chess move, my baseball pitch, and so on. Baseball games would go quite differently if one team could preemptively dismiss the other's home runs. In social interpretation, if we're colleagues from different departments and you deploy in discussion a term unfamiliar to me and developed by and for your field of study, I'm in violation of the rules governing social interpretation if I just dismiss the resource you used as though as it has no permissible applications (it applies to nothing at all) or no applications worthy of discussion (it makes something out of nothing). It's impermissible for me to carry on as though you've made either a trivial move or no interpretive move at all; rather, I ought to take it that I don't know which move you've made, and I should pursue avenues for finding it out—by asking you to explain the resource to me, for instance.

But in rule‐governed activities that have rules of accommodation, if a move is made and no one objects, then the move is subsequently permissible (unless or until someone objects). The move triggers the rules of accommodation and, if no one objects, it enacts an s‐rule making such moves permissible. Since social interpretation has rules of accommodation, then when a dominantly situated knower dismisses a nondominant resource,Footnote ¹¹ if no one objects, then the dismissal move enacts an s‐rule that makes such dismissals permissible.

If such s‐rules are enacted systematically, then by the criterion offered in section I, point 4, there is a corresponding g‐rule. If dominant knowers are systematically permitted to dismiss resources developed by marginalized knowers at present, then there is presently some general rule governing social interpretation that makes it permissible for dominant knowers to preemptively dismiss nondominant resources, contrary to what the rules would be in a system without oppression. The rule seems to target the resources for dismissal by dominant knowers, no matter the position of the knower who deployed the resource.

Moreover, the reverse—that nondominant knowers are generally permitted to dismiss dominant resources—is dubious, at least in contexts where the rules of oppression are in force. For one thing, the default rule for rule‐governed activities is, as suggested above, that participants can't just ignore one another's moves. In order to establish an exception to the default, those who want the exception or benefit from it have to wield considerable social power over those who don't want the exception established or who are disempowered by it. They must be able to do such degrading things as dismiss another person's conversational moves without suffering social, economic, or institutional sanctions. It's dubious that nondominant knowers have such power.

Is it right to suppose that dominantly situated knowers in oppressive systems—or, at least, in some of our oppressive systems—have this power? Here's one way we might make the case that it is. As McGowan points out, oppression is a rule‐governed activity that assigns ranks:

If this is right, then just as military and epistemic ranks confer differential social‐interpretive permissions, we can expect dominant knowers to have permissions that nondominant knowers do not. It may be that one of these is permission to dismiss or ignore social interpretive moves that use nondominant resources. Note, after all, that in other contexts, the higher ranks have a similar permission. Imagine that an undergraduate in his first chemistry lab applies to some chemical phenomenon a term that wasn't developed by chemists, that his professor knows wasn't developed by chemists, and that seems to the professor to refer either to something trivial or to nothing at all. Does she have permission to dismiss his move? It seems she does. Indeed, it would be appropriate for her to make a move signaling to him and other students that there is no such phenomenon as the one he's trying to refer to and that the term he used is inappropriate in the context. Prima facie, this move is appropriate because there's a g‐rule giving her permission to perform it. Similarly for military ranks: If a private refers to a part of a military base using a term that isn't used by the higher military ranks, and if it seems to the higher‐ranking person that his term refers to nothing at all, then she has permission to correct him and dismiss the resource he used. These are defeasible reasons to suppose that in oppressive rankings, the higher‐ranked dominant knowers also have this permission. If dominant knowers do have this permission, then this explains why they can dismiss others’ conversational moves without sanction. And if they have this permission thanks to their rank, this is why lower‐ranked, marginalized knowers aren't permitted to dismiss dominant resources. (In the next section, we'll see an additional reason for this asymmetry: it's partially constitutive of some kinds of oppression that dominant resources be accepted and resistant resources be rejected.)

Is this asymmetrical disabling of interpretive moves a kind of silencing? It would certainly be silencing if each case were like this: a language user attempts a move in social interpretation but fails, and she fails because her success is prevented by rules governing social interpretation under oppression (cf. Reference Langton and PostLangton 1998, 274). Compare one of Rae Langton's examples of silencing: “An actor has a role in which he shouts, “Fire! I mean it! Look at the smoke!”; but then a real fire breaks out in the theater, and he shouts, “Fire! I mean it! Look at the smoke!” (274). In our terms, the actor's second utterance is an attempted move in social interpretation, but his attempt fails because the audience dismisses his utterance as not a move in social interpretation at all. In the actor's case, the rules governing his speech at the time prevent his utterances from being taken as interpreting the actual world rather than the world of the play. But with the cases I've given, it's not that attempts at social interpretation with marginalized resources are prevented from success by the rules; it's that the rules make it so that their failure is permitted. What raises the possibility that the rule makes for systematic silencing is that it puts it in the hands of dominant knowers to determine whether such moves are or are not in fact moves of social interpretation. The rule makes it so that dominant knowers are systematically empowered to silence interpretive moves that use resources developed in nondominant communities. Surely this makes it the case that any who would use nondominant resources suffer persistent and wide‐ranging hermeneutical marginalization. Where this marginalization obscures some significant area of social experience from collective understanding, there is a hermeneutical injustice.

Three comments before moving on. First, in the foregoing development from individual encounters like Pohlhaus describes to the g‐rule, I seem to assume without discussion that, systematically, when a dominant knower dismisses a nondominant resource, no one will object. I give no reasons to believe it. A number of considerations in a variety of contexts may, of course, make it unsafe or imprudent to object to a dominant knower's interpretive moves, but this doesn't suffice to show that systematically, no one objects to such dismissals. Presumably, some dismissals of some resources are called out, and if we want to know whether they're systematically permitted to pass for a certain term, we need to do the empirical legwork. I'm not going to attempt it here. I take it to be plausible that for many nondominant resources, objections to dismissals are systematically absent, so the g‐rule is in force. Perhaps a smaller point here is that objecting to such dismissals in specific cases is important for the sake of preventing the g‐rule from taking hold.

A second related point is that I don't give sufficient reason for thinking that encounters like Pohlhaus describes occur systematically. I concede this point, and I won't try here to establish that they do. Again, whether they do or not is an empirical matter, and deciding it is beyond this article's purview. In the next section, I'll consider rules that more plausibly derive from systematic features of oppression.

Third, note that what I've considered in this section is just one way that the rules governing social interpretation might evolve to accommodate moves commonly made by dominant knowers. Collectively, dominantly situated participants in social interpretation have available to them a vast array of social, economic, and state resources for ensuring that the moves they want to be made are permissible and moves they don't want to be made are impermissible. If their collective will is to make a social interpretive move permissible or impermissible, it is likely to be done.

This can affect rules of interpretation, as above, and rules governing interpretive resources themselves, as Emmalon Davis shows in her discussion of epistemic appropriation. In epistemic appropriation, nondominant resources are both “detached from the marginalized knowers responsible for their production” and “utilized in dominant discourses in ways that disproportionately benefit the powerful” (Reference DavisDavis 2018, 702). Dominant knowers’ various privileges put them in a position to detach the rules governing nondominant resources from nondominant knowers who developed them and to change those rules to the benefit of the dominant.

III. Rules, Resources, White Fragility, and DARVO

In this section, I draw on Charles Mills's work on White ignorance to suggest that in addition to making it permissible for dominant knowers to dismiss nondominant resources, oppression makes it impermissible to use resistant resources because their use indicates a speaker's impermissible rejection of dominant resources. It is thus permissible to punish those who use resistant resources, with differential punishments for dominant and nondominant knowers. In addition to penalizing the use of resistant resources, this state of affairs incentivizes the use of dominant resources.

There are several conceptions of dominant resources to be found in the literature on hermeneutical resources. Pohlhaus, for instance, seems to conceive of dominant resources as those that are attuned to “the experienced world of the dominantly situated knower” (Reference PohlhausPohlhaus 2012, 722). Mason draws on Mills's conception of White ignorance, according to which dominant White knowers tacitly agree to accept epistemic resources (among other things) that misrepresent the world in ways that obscure the viciousness of their domination and/or make it appear justified (Mason 2012, 302). Dominant resources on this conception are, in the first place, those that conceal, justify, or otherwise help maintain an oppressive social system. In responding to Mason's and Dotson's charges against Fricker, Trystan S. Goetze proposes that the resources relevant to hermeneutical injustice are collective resources, and these are “interpretive tools that are shared by all” (Reference GoetzeGoetze 2018, 74).

Each of these conceptions can be useful depending on what one is concerned to articulate; for our focus on rules governing social interpretation in an oppressive system, I prefer the Mason/Mills conception. It situates dominant resources with respect to the rules of oppression directly: as we'll see shortly, it's essential to certain kinds of oppressive system that dominant resources be accepted. Something like this may also be true on Pohlhaus's or Goetze's conception, but they don't make it explicit.

Mills's remarks in The Racial Contract (1997) and “White Ignorance” (2007) suggest two rules governing the use of interpretive resources in a system of racial oppression; I'll sketch these rules and give reasons to think there are similar rules governing the use of interpretive resources in a system of gender‐based oppression. I suggest—but do not otherwise defend—the broader hypothesis that there are similar rules governing the application of resources in other systems of oppression. I don't assume that the rules for every oppressive system—racial, sexual, colonial, and so on—are the same, and I don't assume that the rules of each system influence the rules governing the application of interpretive resources the same way. And I don't want to ignore that oppressive systems are intertwined and mutually reinforcing. All I want to suggest about oppressive systems (other than those I discuss explicitly) is that it's worth asking of each whether it has rules similar to those discussed here.

First, on Mills's view, the epistemic resources that conceal or justify White domination aren't just available in the system, they're required for Whiteness.

The rules of White supremacy, according to Mills, make it so that one can't be White unless one adopts the system of dominant epistemic resources.Footnote ¹³

Second, it is necessary for political systems established by the Racial Contract to “bring about conformity to the terms of the Racial Contract among the subperson population” (83). I take this conformity to include acceptance of the dominant resources, at least when engaged in social interpretation with dominant group members. The Contract makes Whites into persons, but it relegates non‐Whites to a subordinate status Mills calls “subpersonhood”; since subpersons have no reason to accept the Racial Contract that subordinates them or the epistemic resources that justify and conceal their domination, their conformity to the Racial Contract must be achieved by “the two traditional weapons of coercion: physical violence and ideological conditioning” (83).

How might rules like these affect rules governing the application of terms? There are, of course, many ways; there will be different effects under different incarnations of White dominance (colonial, Jim Crow, color blind, and so on); the rules are plausibly sensitive to other social positions like gender and class. But, speaking generally, we should expect that there are rules in place making it impermissible for knowers to reject the system of dominant resources. When knowers racialized as White make interpretive moves that seem to reject the dominant resources in a context where the rules of White supremacy are in force, then, we should expect those knowers to be treated as though their moves were out of bounds. The rules of White supremacy call for such interpretive moves to be punished, to be treated as violations of the rules. When a knower racialized as a subperson makes a move that violates the rules, we should expect harsher punishment.

Which sorts of interpretive moves might seem to reject the dominant resources? The answer will depend a great deal on context and the extent to which dominantly positioned knowers are vigilant about detecting and punishing a speaker's violations of White‐supremacist rules. But I suggest that we can see such interpretive moves being punished when (i) knowers apply the interpretive resource expressed by “racism” in certain cases and (ii) this is met with exhibitions of “White fragility.”

(i) In White Fragility, Robin DiAngelo points out that “the dominant conceptualization of racism” is of “individual acts of cruelty,” so that “only terrible people who consciously don't like people of color can enact racism” (Reference DiAngeloDiAngelo 2018, 124). In our terms, say that in the system of dominant resources, the rules make it permissible to apply “racism” only to individual acts of cruelty by terrible people who consciously don't like people of color. Following this rule makes it “nearly impossible to engage in the necessary dialogue and self‐reflection that can lead to change [in the White‐supremacist system]” (124). I take this resource to be among the dominant resources because it obscures structural racism, institutional racism, and individual acts that reinforce them but don't always seem cruel.

I propose that a knower can seem to reject the system of dominant resources, then, by breaking the rule for applying “racism” in accordance with the dominant resource; one can apply it to persons who aren't obviously terrible or who don't consciously dislike people of color, for example, or to behaviors or attitudes that aren't (obviously) cruel. I want us to focus on cases in which knowers use “racism” this way.

(ii) DiAngelo points out that in these cases, one is often met with responses—from those, if any, accused of racism but also other Whites in the audience—characteristic of White fragility. DiAngelo spells out a variety of emotions and (verbal and nonverbal) behaviors associated with White fragility (122–23); she characterizes it as a form of bullying and as doling out punishment:

The punishments and bullying of White fragility are responses to moves that are out of bounds in White supremacy. When these moves are interpretive moves, White fragility is punishment for a violation of White‐supremacist rules governing the interpretation of the world. This suggests that there is a rule in the oppressive system making it impermissible—out of bounds and deserving of punishment—to reject the system of dominant resources. Indeed, DiAngelo tells us that these responses aren't just common defensiveness; they result from social forces that hold racial hierarchy in place (8, 114).

DiAngelo also gives support to the supposition that Whites and persons of color will receive differential punishments for rejecting dominant resources. She claims that when a knower racialized as White is punished with White fragility, it is, in part, punishment for violating “white solidarity.” One fails to do what Whites require of one another in the system, namely, to obscure or justify the system's injustices. That is, White fragility is also punishment for violating the necessary condition on Whiteness of which Mills spoke: accepting the dominant resources. Alternatively, when a person of color is punished with White fragility, the punishment is even harsher.

Call resources that challenge a dominant system of resources resistant resources.Footnote ¹⁴ If there are rules governing the social interpretation of the world under patriarchy that resemble the rules under White supremacy, then we should expect that (i) there are dominant and resistant resources in the gender system, (ii) some interpretive moves using the resistant resources seem to indicate a rejection of the system of dominant resources, (iii) there are practices of punishing those who make such interpretive moves, and (iv) there are differential punishments for dominant and marginalized gender groups. I'll make points i and ii together, and then turn to iii and iv.

(i, ii) Fricker suggests that sexual harassment expresses a resource that challenges the dominant gender system; Pohlhaus offers date rape.Footnote ¹⁵ Where sexual harassment applies, the dominant resources provide “flirting” to obscure how power is leveraged in sexual harassment and what is demeaning, disempowering, intrusive, and violating about it (Reference FrickerFricker 2007, 153). Similarly, the dominant resources supply “seduction” as the appropriate term for interpreting date rape (Reference MillsMills 1998, 28). In addition, just as the dominant resources supply a conception of racism that permissibly applies only to terrible, hateful people, and this obscures much other racism, it may be that the dominant resources supply a conception of rape that is permissibly applied only to forcible stranger rape, and this obscures date rape and much sexual harassment as well as other sexual abuses. In most applications, the resistant terms challenge the overarching dominant interpretation according to which men are entitled to access women's bodies for their own purposes.

(iii) Are speakers punished for using these terms in social interpretation? The costs of interpreting parts of the world with the resources these terms express should be well‐known. As noted above, applications of such resistant resources are often met with what Freyd describes as DARVO: deny, accuse, reverse victim and offender. This response, like White fragility, makes it miserable—often emotionally, economically, and physically threatening—to make an accusation of sexual offense, and it punishes those who do. It treats applications of “sexual assault” and “date rape” as interpretive moves that break the rules of social interpretation and deserve punishment.

Freyd's description of DARVO limits it to perpetrators of sexual offenses, but it's not only those accused of sexual offenses who punish those who use resistant resources. In the practices of victim‐blaming and shaming, those who follow the rules of gender‐based oppression mobilize to treat those who apply resistant resources as rule‐breakers.

And it's not only accusations using resistant resources that are out of bounds in the patriarchal system; it's any move that interprets the world in resistant ways. It was out of bounds, for instance, for Mary Koss to count date rape achieved via intoxication (rather than force) as rape in her 1987 study showing that 27.5% of college‐age women had experienced rape or attempted rape since age 14 (Reference Koss, Gidycz and WisniewskiKoss 1987). Although Koss's research has a respectable academic reputation, the nonacademic reception of her work exhibits similarities to White fragility. Two accusations found in Jody Raphael's survey of twenty‐five years of this nonacademic reception are relevant here: that Koss's work uses “an inflated definition of rape” (Reference RaphaelRaphael 2013, 30), and that it is part of “an ideological war against men” (22). The first accusation suggests that Koss's interpretation of date/acquaintance rape as rape is regarded as out of bounds, just as the moves that trigger White fragility are out of bounds.Footnote ¹⁶ The second accusation suggests that the hostile reception is in part a response to the appearance that Koss's research threatens patriarchal gendered dominance, just as White fragility issues from “social forces that hold racial hierarchy in place.” The continued hostility to Koss's research over twenty‐five years later is reminiscent of the bullying of White fragility: As DiAngelo put it, “I am going to make it so miserable for you to confront me … that you will simply back off, give up, and never raise the issue again (Reference DiAngeloDiAngelo 2018, 112; see Reference YancyYancy 2018).

Notice that treating Koss's research as part of a “war against men” is not dissimilar from the DARVO reversal of victim to offender. By the interpretive rules of oppression we're considering, this is no reversal at all when the offender is a man and the victim is a woman. In the dominant system, applying the resistant resource to such a case is an offense, and the victim of this offense is the actual sexual offender. The resistant interpretive move accuses him of doing something he wasn't entitled to do, but according to the dominant system, a man is entitled to a woman's body (unless she is another man's “property,” for example, his daughter or wife). Those who live under patriarchy are rule‐bound to treat the accuser as a rule‐breaker and the accused as a victim of cheating.

One might push an analogy with White fragility and the dominant rules for “racism”: the dominant conception of racism is to White fragility as the dominant conception of sexual offenses is to DARVO. Applications of “racism” to structures or institutions or to individuals who don't consciously hate people of color violate the dominant rules for applying the term, and they are punished with White fragility. Applications of “date rape” and “sexual harassment” violate the dominant rules governing the application of terms referring to sexual offenses—terms taken in the dominant system to refer to violent stranger rape or to property crimes against men—and they are punished with DARVO. I think this analogy is helpful for seeing the parallel between the social interpretive rules of White supremacy and patriarchy, but there isn't space here to fully defend or clarify it. I provide it, however, as a tidy way to express the point at hand: in both oppressive systems, there are practices of punishing social interpretive moves that seem to reject the system of dominant resources.

(iv) Are there differential punishments for dominant and nondominant group members in the oppressive gender system? As is painfully evident, and as the examples above suggest, those read as women receive harsh punishment for using the resistant resources. The same is plausibly true for members of other nondominant gender groups—trans persons, gender‐nonconforming, and nonbinary persons—but in the gender system, the very existence of these positions challenges the dominant resources, and, accordingly, persons who occupy these positions are punished as rule‐breakers for their mere existence (see, for example, Grant et al. Reference Grant, Mottet, Tanis, Harrison, Herman and Keisling2011). When a member of one of these marginalized groups is punished after applying resistant resources, then, it isn't obvious what the punishment is for.

It's not clear, however, that men face punishment at all for applying resistant resources. The situation is complicated, but I think they do. I'll discuss some of the complications below, but note, first, that this very nonobviousness supports at least the claim that dominant and nondominant gender group members receive differential treatment when they use resistant resources. It remains to be shown that they both receive differential punishments, though.

In the straightforward case, a man who resists patriarchal interpretations in a context where the rules of gendered domination are in force has his dominant position in the gender system questioned. In Anglophone communities over the past few decades, this questioning commonly takes the form of an accusation that the resistant man is gay. As George Yancy notes in his New York Times opinion “#IAmSexist,” when a man refuses—or even hesitates—to participate in the objectification of women, his peers often charge that he is gay (Reference YancyYancy 2018).Footnote ¹⁷ Under the rules of the gender system in force in such situations, male homosexuality is incompatible with hegemonic masculinity, the masculinity that grants a man entitlement to dominance (see, for example, Reference DonaldsonDonaldson 1993, 648). Resisting the dominant interpretation of women's bodies, then, immediately calls a man's masculinity into question, imperiling his position as a member of the dominant group. Applications of “sexual harassment” and “date rape” to cases in which a woman is the victim, then, should be straightforward challenges to dominant interpretive resources, and when men challenge the resources of gender dominance, they're “punished” with the threat of losing their dominant position. As with White supremacy, accepting the dominant resources of patriarchy seems to be prerequisite for membership in the dominant gender group.

But things are seldom straightforward when it comes to men's applications of terms like sexual harassment and date rape. I suggest that this isn't because it's permissible for men to challenge the dominant resources under the rules of gender oppression; rather, it's because the systems of patriarchy and White supremacy provide ways for men to apply these terms without challenging the oppressive systems. First, note that if the terms are applied to cases in which a man is the victim, it's not obvious that such applications challenge the patriarchal system of dominant resources; or, if such applications are challenging dominant resources, they don't seem to be challenging the resources that interpret women's bodies as objects for men's uses. In order to challenge these resources, rather, one must apply the resistant resources to cases in which a woman is the victim.

In many contexts, however, when men apply terms referring to sexual offenses to such cases, these applications reinforce dominance rather than challenge it. Some applications enact benevolent sexism, reinforcing the rules of the gender system that position women as to‐be‐saved by men. And some reinforce patriarchy and White supremacy, as when White men use such interpretive moves to justify racial violence (see, for example, Lindquist‐Dorr Reference Lindquist‐Dorr2004).

But again, I don't think we should take these cases as showing that men under patriarchy aren't punished for their uses of resistant resources. We should take them, rather, as showing that patriarchy supplies men with a number of ways to apply resources referring to sexual offenses without rejecting the system of dominant patriarchal resources. When women are interpreted as property and when it's needful to justify violence against Black men, there are patriarchy‐ and White‐supremacy‐approved applications of such resources. When a man does (manage to) seem to reject patriarchy's dominant resources, though, he is regarded as a rule‐breaker and his membership in the dominant gender group is thrown into question by other dominant group members.

If it's right that dominant knowers punish language‐users for seeming to reject dominant resources in social interpretation, then although Pohlhaus might be right that dominant knowers sometimes dismiss resources developed by marginalized communities because these resources seem to the dominant knower to refer to nothing at all or nothing important, there is an additional reason why a dominant knower might dismiss such resources: the dominant knower is following the rules that govern social interpretation in a system of oppression. Under the Racial Contract, a dominant White knower must be careful not to seem like she rejects the dominant resources or else she may forfeit her Whiteness and all (or at least some of) the privileges and safeties that come with it. If the point holds for systems of dominance generally, then the rules make it so that dominant knowers mustn't accept resources that challenge the dominant resources or else they might forfeit their dominance and (some of) its attendant privileges.

IV. Further Developments

Interpreting the world is often (if not always) a social activity. Using terms to express interpretive resources in this social activity is rule‐governed. The rules make some social uses of interpretive resources permissible, and they rule other uses out of bounds; they determine which interpretations are and aren't socially acceptable. In this article, I've given reasons to think that when interpreting the world under systems of oppression, these rules are influenced by the rules of the oppressive systems. In particular, I proposed in section II that it's permissible in oppressive systems for dominantly situated knowers to dismiss interpretive resources developed by and for marginalized communities, but not vice versa; and I proposed in section III that when oppressive systems produce interpretive resources that obscure or justify the injustices of the system, the rules of social interpretation in that system make it impermissible to reject these resources. I suggest that these resources (but perhaps not only these) are aptly called “dominant resources.” The rules of the oppressive system make it impermissible to reject the entire system of the dominant resources. There are presumably many ways that one might seem to reject the system of dominant resources. Here, I've suggested one way that is also a move in social interpretation, namely, applying a resource that seems to challenge the dominant system. These (but perhaps not only these) are aptly called “resistant resources.”

Following Charles Mills, I suggest that those who are dominant in an oppressive system will enforce these rules when they participate in socially interpreting the world; they will make moves to punish and eventually revoke dominant‐group membership from dominant‐group members who apply resistant resources; and, they will mete out even harsher punishments to subordinate group members who use resistant resources. I've given evidence that under patriarchy and White supremacy, such rules do govern our use of terms in socially interpreting the world. In conclusion, let me briefly suggest several further developments and applications of these points.

(1) These rules and the costs of violating them suggest one way that dominant groups can impose their interpretations of the world on subordinate groups. Iris Marion Young describes the injustice of cultural imperialism as follows: the dominant group “imposes on the oppressed group its experiences and interpretation of social life” (Reference YoungYoung 1990, 60). If a member of an oppressed group resists the dominant interpretations of social life, the rules of the system call for the weapons of coercion. The system punishes uses of resistant resources, and it can thus make the activity of interpreting the world so inhospitable to a subordinate group member that it can be in her best interest to apply dominant resources, even if she knows they're inaccurate, even if those resources obscure experiences important to her, even if they do damage to her sense of bodily autonomy, demean her, and so on. As Mills notes, the line in the Black American folk poem, then, is plausibly a matter of survival: “Got one mind for white folks to see/Another for what I know is me” (Reference Mills, Sullivan and TuanaMills 2007, 18).

(2) I've focused on cases in which speakers seem to resist the system of dominant resources by applying resistant resources. But note that there may be—and plausibly are—contexts in which one will face punishment or threats of exclusion just for failing to apply a dominant resource. In these contexts, one will seem to reject the system of dominant resources just because, for instance, one didn't join in using sexist slurs in the locker room, because one didn't use racist or classist terms at the country club, because one self‐identifies as genderqueer, and so on. In these contexts, dominant knowers are ever more vigilant about whether the dominant resources are being rejected.

(3) Given the costs of seeming to reject dominant resources, there are incentives for nondominant speakers to use dominant resources, namely, to escape the punishments for doing otherwise. But there may also be modest “rewards” for using dominant resources and/or rejecting resistant resources. A woman might receive a higher social interpretive rank in patriarchal contexts for criticizing feminists. A Black man might achieve a higher rank in White‐supremacist contexts for endorsing utterances of “all lives matter.”Footnote ¹⁸ Prima facie, nondominant knowers can achieve higher ranks in oppressive systems by following the rules of the system, but it remains nearly impossible for them to achieve equality with dominant knowers so long as they bear the visible marks of their nondominant race or gender.

(4) The costs of using resistant resources or even of failing to use dominant resources presumably make it so that dominant resources are more often used and less often challenged. This can lend to them a purport of legitimacy and accuracy, especially for those who engage in socially interpreting the world almost exclusively with dominant group members. It may thus come to seem as though everyone agrees that the world is as the dominant resources interpret it to be. The world as interpreted by dominant resources will seem to be the world. If this is how it seems to a dominant knower, then it's no surprise that they will respond to resistant resources as Pohlhaus says: as if such resources “attend to nothing at all, or to make something out of nothing” (Reference PohlhausPohlhaus 2012, 722).

But we should keep in mind that when this is a knower's position with respect to dominant resources, he is also likely to carry on as though those presenting him with resistant resources are breaking the rules and ought to be punished. For such a knower, the dominant resources will determine what various parts of the world are, just as the rules of solitaire make it that certain cards are nines of clubs, given how they look. If you treat a nine of clubs like a six, you're ignorant of the rules or dissembling.Footnote ¹⁹ From the perspective of such a knower, if you apply resistant resources, then if you've made a move at all, you've misinterpreted the world. Perhaps out of ignorance, perhaps because you're dissembling, perhaps because you're hysterical. At minimum, you're to be regarded as having broken the rules governing the social interpretation of the world, but you may also be in for the consequences of seeming ignorant, dissembling, unhinged, or in the extension of any of the dominant resources reserved for subordinate group members who break the rules of oppression.

(5) The social‐interpretive rules of oppression will sometimes be in force when communicating resistant resources. This suggests the epistemic importance for resistance movements of (i) cultivating and maintaining contexts in which the dominant rules governing social interpretation aren't in force and (ii) protecting resistant resources deployed in contexts where the dominant rules are in force. Having contexts where the dominant rules aren't in force, resistant contexts will facilitate the development of resistant resources. Protecting resistant resources—that is, objecting to their misuse, refusing to let them be preemptively dismissed, calling out those who punish their applications—will help prevent epistemic appropriation.

(6) Just as there are different rules governing dominant and resistant applications of “racism” as described by DiAngelo, there are plausibly different rules governing dominant and resistant applications of some or all race and gender terms. For instance, Talia Mae Bettcher describes different rules for applying “woman” in dominant communities as compared to trans subcommunities, concluding that “woman” expresses different concepts in the two communities (Reference Bettcher, Power, Halwani and SobleBettcher 2012, 238–44). Mostly, however, philosophical discussions of race and gender terms and concepts pay little attention to the possibility that one word form may play a role in both dominance and resistance, expressing a dominant resource here and a resistant resource there. But given that one of the goals of investigating race and gender terms/concepts is to understand the social world and how such terms/concepts might shape it, and given that dominant and resistant resources shape the social world in different (plausibly opposing) ways, the literature must take this possibility seriously. Moreover, if it's seldom recognized (even by theorists) that a word form sometimes expresses a dominant resource and sometimes a resistant resource, then this confusion itself plausibly plays a role in shaping our social world, and it too deserves further study.