EPISTEMIC MISTRUST

In her essay, “Epistemic Trust and Social Location,” Nancy Daukas (Reference Daukas2006) sets the background for a discussion of epistemic trust:^{Footnote 7}

Daukas is interested, especially, in epistemic trustworthiness: what is it for an agent to be worthy of epistemic trust. Let’s consider first what trustworthiness requires; we can then consider whether there are systematic ways in which some who are trustworthy are denied trust, and some who are not trustworthy claim it.

The relationship between teacher and student is not one of epistemic equality. At least in most cases, it is an important feature of the relationship that the teacher has some kind of knowledge that the student lacks. It is compatible with this, however, that there are epistemic conditions for each party that must be satisfied in order to participate cooperatively in teaching and learning. The first, most basic, principle is that each should extend to the other a threshold degree of first-order epistemic credibility. As Davidson (Reference Davidson1984) and others have argued, we cannot even begin to communicate with someone unless we presume that the other is in command of a substantial number of truths and undertake to abide by the simple rules of conversation dictating that one should aim to, e.g., be relevant, be informative (but not too informative), be perspicuous, and speak the truth (Grice Reference Grice, Davidson and Harman1975).

Daukas argues, however, that there are further requirements placed on productive epistemic cooperation. Because fruitful epistemic exchange requires that we take each other seriously, we depend not only on the credibility of our partners, but also on their ability to accurately gauge their own reliability. Epistemic cooperation with someone who consistently misjudges her own epistemic capacities is of little value, for she will contribute to the exchange without indicating to those who trust her the proper degree of confidence they should have in her contributions. “If someone frequently and sincerely claims epistemic authority that she does not possess, she is not worthy of trust, in an epistemic sense; and we should not believe P, or even consider the possibility that P, on the strength of her testimony alone” (Daukas Reference Daukas2006, pp. 110–111). So, Daukas suggests, successful epistemic exchanges are those in which we rightly extend to our partners, and they to us, a principle of second-order epistemic credibility. In effect, in order to be epistemically trustworthy, and so a meaningful partner in exchange, one must be a good judge of both his or her own cognitive capacities and the cognitive capacities of others. Failures of such self-assessment involve being overly confident in one’s own epistemic capacities (closed-minded, dogmatic, or impatient), or overly diffident (deferential or gullible). Failures in judging our interlocutor include being ungenerous (suspicious or skeptical) or excessively generous (uncritical).

We develop our second-order epistemic competencies—our ability to rightly judge and be judged trustworthy—in social interaction. As Daukas says:

School is a crucial site for internalizing epistemic norms and learning to “read” others. Students learn how to judge their own competencies and the competencies of others based on the responses they receive, and they see others receive, to contributions in the classroom. As Daukas points out, under conditions of race and gender hierarchy, it is not surprising that attributions of trustworthiness will go awry, and racially dominant groups will become over-confident of their trustworthiness, and subordinate groups will become less confident than they should be. Because those who are less confident in a domain are less likely to be successful and highly visible as authorities in that domain, “unjust epistemic exclusion on the basis of social location is therefore self-perpetuating: its consequences perpetuate the inequalities that fulfill, and therefore seem to justify, the discriminatory expectations that, in turn, perpetuate unjust epistemic exclusion” (p. 116). Let’s call this the “untrustworthiness feedback loop.”

This sort of looping effect is familiar and comes in different forms. For example, Abigail Stewart (Reference Stewart2007) describes an “evaluation bias feedback loop.” If, due to social bias, my contributions are wrongly judged less worthy than others’, I will be less successful by ordinary measures; and the lack of success of members of my group will have the result that we are less visible in authoritative positions. The lack of a critical mass of members of my group in high-ranking positions will not only lead others to doubt my credibility, but will reinforce attitudes towards the group that are responsible for biased evaluations. In short, bias against a group G leads Gs to be less successful, which reinforces bias against Gs.

Although Daukas emphasizes the effects of trust violations on confidence (in her view, lack of confidence explains lack of success), it is important to note that those students who are confident and trustworthy in their self-evaluation also suffer harm when teachers are untrustworthy, or even just perceived as untrustworthy. Consider again the example of Jonathan and Mustafa in Spanish class. According to Jonathan’s report, Mustafa asks a reasonable question and yet the teacher’s reply evinces not only rudeness, but also first-order ignorance of Egypt and second-order epistemic untrustworthiness: she is unable or unwilling to acknowledge her own epistemic incompetence. In this context, Mustafa and Jonathan do not come to doubt their own knowledge; rather, they rightly lose epistemic trust in the teacher. They conclude that she is not a trustworthy judge of her own epistemic abilities. However, a breakdown of epistemic trust results in a breakdown of epistemic cooperation in the classroom. As the teacher’s authority is challenged, she removes Mustafa from the class. The students take this to be an enactment of racial privilege: from their perspective, she demonstrates both epistemic and moral arrogance (and ignorance) in response to ethnic (at least somewhat racialized) difference. Moreover, Mustafa’s refusal to grant her authority on the matter, no matter how justified he is, results in expulsion from “her” space.^{Footnote 8} The students perceive this not just as unfair, but racially significant.

It would seem that a careful evaluation of the teacher’s actions would warrant an epistemic reengagement with her as a teacher of the Spanish language (she does not, after all, display epistemic untrustworthiness in her command of Spanish, even according to the student’s report). However, as Daukas points out, fruitful epistemic exchange relies both on epistemic and moral trust: we each must be able to trust that the other will be, to some threshold degree, sincere, honest, and open. This is especially true when power relations in the exchange are unequal, for the powerful who are dishonest and insincere can do serious damage when they are trusted by the less powerful. In the case of Spanish Class, doubt has been raised about Ms. S.’s moral trustworthiness in her treatment of Mustafa (and others). So it is not surprising that, without an effort to repair the epistemic and moral damage, the classroom has ceased to be a learning environment for students of color.

It is crucial to note that the students who disengage in Spanish Class (and, I would argue, in History Class (see below)), are not ignorant, unmotivated, or self-doubting; rather, they are knowledgeable and have epistemic self-respect. Moreover, their withdrawal of epistemic trust is warranted (even if not practically wise). However, an unwillingness to cooperate in the epistemic exchange is harmful to them: they are less successful in mastering the material; their distrust is interpreted as disrespect or stupidity; and, if they try to raise the issue or complain, they are branded as trouble-makers, as resistant to authority, etc. I think what emerges is an “epistemic distrust feedback loop”: if, due to my race, teachers act towards me in a way that either signals that they are not granting me epistemic credibility, or that shows them to be epistemically untrustworthy, I will, reasonably, withhold epistemic trust; but if I withhold epistemic trust, I will show signs of distrust, will do less well on my work, and will appear to be stupid and unmotivated. When I appear to be stupid and unmotivated, teachers act towards me in epistemically distrusting and untrustworthy ways. And the cycle begins again.^{Footnote 9}

EGO DEPLETION

It is easy to imagine that, if a student is regularly in classroom situations in which trust has broken down over the course of their high school years (perhaps even as witness to the travails of others), their disposition to extend epistemic trust to any teacher is diminished. If we add this on top of all the other disadvantages, no wonder students of color are having trouble in school. Research in social psychology allows us to explore some of the broader effects of epistemic distrust and hierarchical race relations with more than just imagination. Two key concepts are ego depletion (Baumeister et al., Reference Baumeister, Bratslavsky, Muraven and Tice1998) and effort optimism (Perry Reference Perry, Perry, Steele and Hilliard2003). I will consider ego depletion in this section and effort optimism in the next.

Richard Holton (Reference Holton2009) has recently argued, drawing on work in social psychology, that “there is a faculty of willpower—something like a muscle. When desires and resolutions clash, we can succeed in sticking to our resolutions by employing this faculty. Moreover, employing the faculty is hard work: it requires effort on the part of the agent” (p. 130). In his view,

It is not part of my goal here to defend Holton. However, it is clear that something like willpower is important in education: we don’t always want to attend class, do homework, read boring material for a test, and yet those who have the willpower to forego immediate desire satisfaction to “do what has to be done” are more successful. Willpower is a central aspect of executive function more generally.

A notable feature of willpower in this model is that “willpower comes in limited amounts that can be used up” (Holton Reference Holton2009, p. 128). This phenomenon is called “ego depletion.” According to Baumeister et al. (Reference Baumeister, Bratslavsky, Muraven and Tice1998), ego depletion is “…a temporary reduction in the self's capacity or willingness to engage in volitional action (including controlling the environment, controlling the self, making choices, and initiating action) caused by prior exercise of volition” (p. 1253). Baumeister et al. have conducted a series of experiments of the following sort: hungry subjects were presented with chocolates and radishes. Some were told they could only eat the chocolates, others that they could only eat the radishes. Some time afterwards, subjects were asked to solve (impossible) puzzles, taking as many tries and as much time as they would like. Those who exerted self-control to not eat the chocolates and only eat the radishes (many later admitted it was difficult to do so) quit sooner on the puzzle task than those in both the chocolates group and the control group. Baumeister et al. concluded that:

Moreover, ego depletion seems to have quite broad effects on executive function, e.g., it lowers one’s ability to stick to resolutions, lowers IQ, and makes one susceptible to bad arguments.^{Footnote 10} Although there are different accounts of how ego depletion works, and how it affects cognition, it is clear that under ordinary circumstances, managing anger and other strong emotions is a cognitive drain (Muraven et al., Reference Muraven, Tice and Baumeister1998; Stucke et al., Reference Stucke and Baumeister2006).^{Footnote 11}

Suppose, then, that efforts to exert self-control do diminish one’s willpower and executive function more generally. Let’s return to the examples and the phenomenon of trustworthiness in classroom settings. Recall that according to Eli’s report of History class, Ms. H. talks eagerly with Maya without including Eli in the conversation, and when she turns to Eli, speaks to him in a way that conveys a lack of confidence in his abilities. Eli wants to say, “F–k you, Ms. H.” Instead he is polite, but not forthcoming about the subject matter of the assignment. Eli is angry, perhaps reasonably so, but exhibits good self-control. However, as a result of his anger, not only did he find himself unable to take advantage of the teacher’s attention, he was no doubt also distracted from the lesson. Moreover, he carried his anger all day and was still angry after school. Given that he behaved well all day, he must have used self-control all day. This would have taken a huge toll on his willpower, leaving him with less energy and attention for his remaining classes and homework. On the Baumeister et al. model, coping with what he experiences as insult fatigues his willpower muscle. Given the relationship between motivation and willpower, these effects occur even if he is motivated to do the work.

No doubt all students suffer frustrations during the course of a day and have to exercise self-control; some students, regardless of race or ethnicity, will have substantial challenges that require more effortful self-control. Everyone is vulnerable to ego depletion. The point, however, is that if students of color are being treated in ways that they regularly experience as denying them credibility, respect, and trust in comparison with White students, then they will face greater challenges of self-control than Whites. And, interestingly, those who are committed to doing well and participating in the community will be the ones to exert the greater self-control, and so suffer the more substantial ego depletion.

So the options look worrisome. Teachers are sometimes untrustworthy. I believe that the students reporting the incidents in question offer reasonable interpretations. There is bias against Black students. But even in cases in which teachers are trustworthy, they can plausibly be perceived as untrustworthy. In either case, based on their experience, students of color find themselves faced with only bad options:

1. 1. Accept the teacher’s mistaken judgment of their credibility and lose confidence;

2. 2. Challenge the teacher and face the stigma of being disruptive and the punishment for it;

3. 3. Withdraw from the epistemic exchange and lose valuable learning opportunities;

4. 4. Exert self-control over their emotional responses of insult and (temporarily) deplete their capacities for cognitive work and epistemic agency; or

5. 5. Develop a Teflon skin that renders one insensitive to (at least certain kinds of) input from teachers (and others).

None of these are ideal strategies for academic success and, depending on the temperament of the student, some may not be available (though in some circumstances, something as simple as ingesting glucose can strengthen willpower so that one is less vulnerable to ego depletion!). These are not the only options, of course. Individuals who have developed a variety of epistemic strengths and hold some power in the conversational context may respond in more productive and self-affirming ways. But remember, we are talking about adolescent students in a public high school, where they are just learning how to be epistemic agents, and where social divisions are highly salient.

EFFORT OPTIMISM

Faced with the breakdown of epistemic cooperation in classrooms (and with students’ families), African American students lose “effort optimism” (Perry Reference Perry, Perry, Steele and Hilliard2003, p. 61). Intuitively, effort optimism captures the general attitude that one’s efforts towards a goal will have a positive effect on achieving that goal. Even if one faces setbacks, the effort is worth it and will lead, if not to success, then at least further down the path towards the goal. Effort optimism provides resilience when dealing with life’s challenges.

In the context of education studies, John Ogbu (Reference Ogbu1983) and others have argued that the persistent job ceiling for Blacks has prevented “effort optimism” from being “culturally sanctioned.” Developing William Shack’s (1970–1971) idea, he says:

Ogbu makes two claims. One claim is that Black students are disadvantaged academically because they lack effort optimism. The other claim is that Black students lack effort optimism because it is not a value in the Black community due to persistent encounters with racism. One might accept the first claim without accepting the second.

Theresa Perry (Reference Perry, Perry, Steele and Hilliard2003) welcomes Ogbu’s attention to the sociopolitical background of school achievement and his emphasis on power and status-mobility, not just “culture” (p. 64). And she draws an important connection between effort optimism and the contemporary educational climate that views education as job training: “For members of a castelike racial minority, the lack of a predictable and comparable relationship between the education and rewards in the marketplace over generations does influence effort optimism, particularly if one’s view of schooling and education is primarily instrumental” (p. 105). However, she rejects Ogbu’s claim that the African American community does not support effort optimism. In her view, African Americans have developed a distinctive approach to education, namely, “education for freedom, racial uplift, citizenship, and leadership” (p. 63) that has developed and sustained effort optimism over time.

Let’s bracket Perry’s debate with Ogbu over whether the Black community fosters effort optimism, or not. A hypothesis remains that effort optimism is relevant to Black achievement in school and is undermined by persistent encounters with racism. In the context of this paper, the question, then, is whether the violations of trust and trustworthiness that students of color encounter undermine epistemic agency so that effort pessimism is a predictable and damaging result.

Optimism, resilience, and self-efficacy are major themes in social and cognitive psychology, and I cannot do justice to them here. However, there is substantial evidence that optimism is a major factor in predicting success. The classic formulation of the issue is Seligman’s idea of “learned helplessness”:

We all suffer misfortune. But there are different ways of explaining misfortune.^{Footnote 12} Suppose you fail a math test. It might be seen as permanent or momentary (I always do badly in math vs. I didn’t study hard enough for this exam); pervasive/universal or specific (Teachers are unfair vs. Mr. B. is unfair); and personal (about me) or not (I’m stupid vs. We didn’t cover all the material in class). Those who tend to explain misfortunes in a particular domain as permanent, pervasive/universal, and personal, lack optimism. They will be more likely to give up after a failure rather than try harder, and will doubt that effort will make a difference.

Consider the example of Math class. According to Jerome, he does his math homework and gets the answers right. But his effort does not pay off because his work counts as failing simply because he doesn’t draw a box around the answers. He gets help from his father and turns in his work, but the teacher doesn’t record it (even though it was graded). He is told he is lying when he objects. And when this comes to a head, he is witness to the teacher rejecting his father’s word (and his legitimate authority). The story represents the teacher as deeply untrustworthy: she does not grant Jerome or his father threshold credibility, and she is a poor judge of her own epistemic capacities and those of her interlocutors. Jerome might plausibly ask the questions: Why me? Why is she treating me so disrespectfully? There will surely be a temptation to pessimistic answers: racism (permanent and pervasive), and I’m Black (personal). One might even argue that this explanation is reasonable and supported by induction given life as an African American in the United States.

There are, of course, other options. One might explain the incident a bit more optimistically: Ms. M. is prejudiced (but not every White person is); Ms. M. is trying to hold me to high standards so I will do better (she isn’t really prejudiced); and Ms. M. is threatened by my father’s intelligence (it’s not me or my Dad, she’s just insecure). But these can be a bit of a stretch, depending on the circumstances. Much of the literature on learned helplessness and effort optimism suggests that the cure for pessimism is changing one’s “mindset” (Dweck 1978, p. 2006). But the problem is that this strategy seems to favor those who are either very sophisticated in their understanding of social dynamics and can gain a distance from the pain caused by the events in order to apply this understanding, or those who are willing to make up explanations to suit their needs rather than carefully consider the evidence and the circumstances. Neither option is very promising, in part because the damage has been done. (And remember, we are talking about adolescents who may be developmentally somewhat self-absorbed!) As Miranda Fricker (Reference Fricker2007) puts it, “[w]hen one is wrongfully mistrusted, regardless of whether it is one’s competence or one’s sincerity that is being impugned, one is dishonored” (p. 46). The experience of dishonor, I believe, is common in racially charged settings.

EPISTEMIC INJUSTICE

In the previous two sections, I have argued that in classroom contexts there are race-related forms of epistemic distrust and apparent failures of trustworthiness that can have substantial implications for student motivation and achievement. Daukas represents the problem of epistemic mistrust as primarily a question of virtue: individuals who are epistemically untrustworthy—individuals who are poor judges of their own epistemic authority and/or wrongly judge the epistemic authority of others—lack epistemic virtue. If these epistemic failings are linked to moral failings, e.g., racism, and sexism, etc., then the individual also lacks moral virtue.

In these circumstances, however, justice is also an issue because under conditions of social injustice there is a pattern of epistemic exclusion that systematically disadvantages certain groups in reinforcing loops:

Given the economic and social significance of knowledge (many suggest we now live in a “knowledge economy” (Mason Reference Mason2013)), unjust epistemic exclusion results in widespread disadvantage for members of racially subordinate groups.

In addition to issues of virtue and disadvantage, however, it is worth considering the idea that a failure to grant others epistemic credibility due to their race is a moral wrong in itself. Miranda Fricker considers this in her book Epistemic Injustice (2007). Her central concept is testimonial injustice. In her view, in the central cases of testimonial injustice, an individual speaker suffers a credibility deficit based on systematic identity prejudice. Fricker (Reference Fricker2007) says,

In Kevin’s description of Biology class, systematic identity prejudice is a plausible explanation of the substitute teacher’s claim that Kevin stole the computer equipment (and plausibly also the accusation of being disruptive). Kevin, being Black, is assumed to be prone to criminal behavior. His denial is not considered credible until the Dean intervenes.

Being falsely accused of wrongdoing has bad consequences for the individual accused, sometimes very bad consequences, particularly when the error is not uncovered. In Kevin’s interpretation of the incidents in Biology class, the teacher is not morally or epistemically trustworthy: she claims to know who the wrongdoer is although she doesn’t actually know, and refuses to consider what Kevin or the others in the class say. Kevin’s word counts for nothing even though his actions do not warrant disrespect. The teacher’s actions are morally wrong. But we can gain further insight by considering the broader significance of epistemic trust.

In each of the examples of epistemic injustice, there are bad consequences. But the wrong is not just a matter of bad consequences. In Fricker’s (Reference Fricker2007) view, testimonial injustice wrongs an individual “in her capacity as knower” (p. 20). In the simplest sense, this involves unjustly depriving the individual of knowledge. Individuals who suffer testimonial injustice are excluded from the epistemic community and so are not granted fair access to a collective good. But what exactly is this good? If we think of knowledge as a product of inquiry and/or training, then one might construe the wrong by analogy with an unfair distribution of stuff: some of us unfairly gain greater access to or a greater proportion of the knowledge pie than others. If knowledge is the sort of thing that universities create and distribute, then unjustly depriving some groups of access to university education might be viewed in terms of an unjust distribution of knowledge-stuff (mutatis mutandis for K-12 education). Epistemic injustice might even be considered akin to theft: mistrust can cause an individual to lose knowledge, for if I am subject to suspicion, exclusion, or even ridicule when I put forth my knowledge in cooperative inquiry, then I may give up the belief, or the justification for it, in spite of my evidence (Fricker Reference Fricker2007).^{Footnote 13}

However, wrongs of mistrust and epistemic exclusion are not aptly understood by relying on a commodified understanding of knowledge, “for credibility is not a good that belongs with the distributive model of justice” (Fricker Reference Fricker2007, p. 19). The cooperative participation in a knowledge community—in the process of giving and receiving knowledge—is essential to who and what we are, for it is the activity that both trains us for and enables us to enact our rational capacities.

Being excluded from the rational community, even subtly, and even only occasionally, matters. And if such exclusion is systematic and prejudicial, the wrong can be severe.

Drawing on Bernard Williams’ work Truth and Truthfulness (2002), Fricker argues further that it is in and through trustful conversation that we are able to form selves at all. The idea is not just that we need a public language in order to have a sense of self, or humankind concepts to become the kind of agent we are (Hacking Reference Hacking, Heller, Sosna and Wellbery1986); we need trustful conversation in which each of us can pull together our various thoughts, desires, and fantasies into a relatively stable whole. As Williams (Reference Williams2002) puts it:

He goes on to describe these two tasks, or problems, of steadying the self and situating it in society:

What solves these problems, together, is trustful conversation:

So epistemic trust and credibility is not just needed to gain access to knowledge goods, but it is essential to becoming a socially situated self.

Of course it is compatible with Williams’ points that one can become a self through trustful conversation with a few intimates. But to become a socially situated self, one must engage in more public conversations with others. These public conversations, at least in our culture, begin at school, and by the time one is an adolescent, they are a crucial part of one’s daily work. The self is formed as one’s internal feelings and thoughts are formulated and reformulated in response to friends, teachers, coaches, and informed by literature, history, art, and music, as well as the push and shove of life.

There are at least two sorts of injustice that can emerge in this project of self-formation. One is that one is unjustly limited to a narrow range of conversations that do not enable one to form a self that is fully integrated into public space (in the extreme case I have in mind cultures in which women do not even speak the language used in public spaces); another is that the conversations one enters into provide one with limited or distorted resources for forming a self. This latter case may be one in which one participates as if one’s conversation partners are trustworthy, but they are not. Their responses, then, may lead one to self-interpretations that are distorted or socially stigmatized. For example, homosexual desire tentatively expressed in a homophobic context may prompt seemingly authoritative responses that represent such desire as shameful, and the shame is internalized; or in a racist context, some might respond to intellectual aspirations expressed by a Black girl by treating such desire as ridiculous, thus quashing the dream. Such conversations do provide information about the social identities available, but the prejudice encoded in the replies is hidden, cloaked by an air of moral and epistemic authority.

The broader worry here is that in a social hierarchy, collective understandings are structured to favor those in power. The selves we become are significantly a product of the social relations we enter into, including teacher-student relations. Individuals are judged by norms and offered identities that make them suitable for particular social roles. Those marked as destined for subordinate roles will not be recognized as full participants in conversations that make them eligible for more privileged roles. Their experiments with selfhood that reach beyond the roles assigned to them will be registered as failures. When the teacher brings to these relations prejudice that is echoed broadly in the culture, it is not surprising that education produces individuals who are most successful at becoming what they are expected to be. Anything else would be radical, dangerous.

RACIAL SEMIOTICS

I have been suggesting that it is useful to see the issue of race in education as not simply a matter of how individuals or institutions within the educational system respond to race, but how the system creates race. Yet, one might argue, the examples I have offered above are excellent instances of precisely how individual teachers and students respond to race. Don’t the students and teachers have a race prior to their interaction? Isn’t race already a fact to be dealt with in the circumstances? And isn’t the failure to deal with race fairly and equitably what causes the particular forms of racial injustice just described?

There is certainly some truth in this way of representing the problem. And, to be clear, my claim is not that an individual teacher or student creates race de novo; rather, teachers and students are parts of a structure that racializes individuals and sustains racial hierarchy, and this is how race is constructed. As I see it, focusing on the failure of individuals to respond correctly to the racial facts tempts us to misdiagnose the problem.^{Footnote 14}

First, the belief that an individual’s race is “given,” that it is part of the world that we are simply responding to, often rests on a mistaken folk biology of race. Second, and more importantly for our purposes, the supposed “givenness” allows people to evade their own contribution to the social realities of race. The contributions I have in mind are not simply the biased actions that occur in interaction, but our tolerance of the racial status quo. The problem is not just that individuals are prejudiced, though they are, or have unconscious biases, though they do. The problem is that these prejudices and biases are systematically produced and sustained through the social institutions and culture we all participate in, and they persist in spite of thoughtful efforts to eradicate them and good intentions on the part of individual agents interacting across racial differences.^{Footnote 15}

One reason why individuals’ good intentions are not enough is that unfairness towards Blacks as a group has a long history in this country, and the unfairness has had material consequences that are hard to overcome. Big and small injustices accumulate and place burdens on the disadvantaged that cannot be easily escaped. Antiracist attitudes on the part of individuals do little to compensate for this material disadvantage. This is why redistribution of resources and material support to underfunded schools is crucial. However, focusing primarily on the history of African American disadvantage encourages (often skewed) historical comparisons with other groups and a temptation to locate the injustice as having occurred in the past rather than the present. This emphasis on the past elides race as a factor in addressing the current situation. In such a “postracial” account, racial groups that were discriminated against in the past have a legacy of disadvantage to make up, but the problems are now racially neutral, because the real problems are disadvantages with respect to wealth, healthcare, education, and job opportunities, etc., that members of any racial group might face. Racism, or racial injustice, is not the problem anymore, or so the story goes.^{Footnote 16}

However, further consideration of the case studies we have examined suggests that we should consider a third factor in addition to conscious and unconscious bias, and accumulated material disadvantage. Racial injustice occurs and historical disadvantage is exacerbated because individuals of color, and their actions, are interpreted through an ideological interface—in ways similar to how language is an interface—that systematically disadvantages them. Simply asking individuals to be more conscientious in responding to racial difference neither adequately targets the normative issue, nor is sufficient to make lasting change in a context where racial meanings provide a dominant interpretive frame.

To explore this third factor we need to develop and apply the notion of social meaning. The idea of social meaning is relatively familiar in the case of spoken language: the meaning of the term is not determined simply by what I have in mind when I use the term (I can use the term incorrectly), but by the way a term is used and understood collectively. Use of the particular phrase ‘social meaning’ usually implies an extension of the notion of meaning beyond spoken languages to action more broadly. Lawrence Lessig (Reference Lessig1995) has done important work to clarify the notion:

Lessig provides the example of wearing a seatbelt in a cab in Budapest (prior to the mid-1990s). Doing so, especially if the driver suggested you didn’t need one, would be insulting, whether you intended the insult or not. “To wear a seatbelt was to insult the driver. That insult is a social meaning” (p. 952). Importantly, Lessig points out,

Semiotics, at least on one understanding, is the study of such social meanings, what they are, how they are created, reproduced, and disrupted.^{Footnote 17}

Even at this early stage of clarification it is useful, I think, to add two points to Lessig’s characterization. First, not only actions/inactions and statuses have social meanings. Things such as traffic signals, money, jewelry, and colors can have social meaning too. As we all know, pink means girl and blue means boy, right? Of course the social meaning of things depends on attitudes, i.e., collective understandings in a context; however, the same is true of everything that has social meaning, including actions/inactions, statuses, and the like. The whole point of saying that an action has a social meaning is to understand it as having a significance by virtue of collective understandings, not just the personal meaning given to it by the agent (or patient).^{Footnote 18} This emphasis on the social collective rather than the individual, and on the background symbolic resources rather than speech events, also distinguishes a semiotic approach from a speech act approach to social meaning.^{Footnote 19}

Second, Lessig (Reference Lessig1995) argues that in order for something to have a social meaning in a context, the meaning must be accepted by those in that context and not appear to be socially contingent:

Although I agree with Lessig that one way (perhaps an especially important way) that social meanings have force is by virtue of seeming “natural” or “necessary,” by existing “invisibly” and being “unnoticed” (pp. 959–960), another way is for the dominant group to take them for granted and impose them, even if they are not endorsed by everyone in the context, e.g., by those they are imposed upon. In such cases, the dominant can justify their actions by reference to social meanings, and such justification appears (and functions as) secure even if the subordinate reject the meanings and justifications in question.

In fact, Lessig’s own example of education demonstrates, I submit, that something can be socially meaningful in a context, and the meaning have social power, even if in the same context it is contested and recognized by some to be arbitrary. Education, of course, not only conveys information and academic skills, but also socializes children into a culture. Culturally mandated responses become automatized through repetition, close monitoring, and correction. In other words, education teaches us—inculcates—social meanings. But by Lessig’s own lights, it (often) does so coercively, even in cases where the coercion is not represented or commonly viewed as such. Yet surely the inculcation is often experienced as coercive (“stupid,” “insulting,” and “painful”) to the student. Lessig argues that the primary way of inculcating social meanings is to adjust context so that one comes to absorb them because it is in one’s interest to do so. This may be true. But one can certainly recognize that something is in one’s interest, while also resisting it together with the background conditions that make it in one’s interest. Returning to an example considered above, it is in Mustafa’s interest to defer to the teacher when she challenges his knowledge of Egypt, but it does not follow that he accepts her imposed social meaning of his comment or the circumstances that make it wise for him to defer. (And likewise, she may not accept his interpretation of her action.) Nonetheless, by virtue of her authority, she coerces his deference and he experiences her as coercive.

The concept of social meanings is an important tool for understanding social interaction because what we actually do in a context is not simply a matter of intention, but is partly determined by the meanings that are in the background in that context. The same is true of language, of course (e.g., I may intend to apologize but fail to do so because my choice of words is inapt, or because, for any number of reasons, my interlocutor doesn’t pick up what I’m trying to do). Mustafa, as we saw, attempted to correct the teacher about women’s roles in Egypt by making an assertion of fact. He interpreted her as confrontational because he exposed her ignorance and challenged her authority. This may not be how she saw the interaction; it might even be a misrepresentation of the classroom dynamics. But the meanings Mustafa and Jonathan relied on are easily and publicly available: For a teenaged male of color to contradict a teacher based on solid evidence has a meaning that the teacher must manage in order to retain her authority. Because Ms. S. lost epistemic authority concerning the facts, she asserted social authority by expelling Mustafa from the classroom. This expulsion, in turn, had meaning, namely, that his knowledge and experience are not valued, and that she is the qualified knower whose word matters.

Notice, however, that the challenge to the teacher’s authority could have occurred, even if Mustafa had not intended it; and the expulsion of Mustafa from the classroom would devalue his knowledge even if the teacher recognizes that he has knowledge of Egyptian customs. In such cases, one might make efforts to cancel social meanings but, as mentioned above, it isn’t entirely up to the agent what meanings their actions have, just as it is not entirely up to a speaker what her words mean. Elizabeth Anderson (Reference Anderson2010) notes, social meaning “does not consist simply in private thoughts isolated in the minds of discrete individuals. It consists in the fact that these thoughts enjoy a certain public standing, coloring the meanings of interactions even among people who would prefer that these meanings not apply” (p. 53).^{Footnote 21}

The example is further complicated by the racial/ethnic differences of the participants. In the example, the teacher is White, Mustafa is Egyptian, and Jonathan, Mustafa’s friend, is Black. Jonathan “reads” the interaction as racially significant: the expulsion of Mustafa is the expulsion of a male student of color; he takes it to express a devaluation of not only Mustafa’s knowledge but also the knowledge of non-White males more generally. This may be an incorrect reading of the teacher’s attitudes, but it is a reasonable reading of the social meaning of the teacher’s action, especially if we are considering that it occurs in a large public high school where the behavior of non-White males is always under suspicion. Moreover, the dominant culture is one in which the disruptive potential of males of color is more salient than their academic potential; it is one in which their assertions of fact are read as confrontational rather than illuminating. So in an academic context, treating males of color as disruptive and confrontational when they assert the truth carries a meaning that both draws on and reinforces this broader devaluation of their intellect whether or not it reflects the teacher’s implicit or explicit attitudes at the time. And, in turn, the message that their intellect is not valuable is something the students then must process, often in the damaging ways I have already outlined.

The effects of social meaning are also present in the other cases. When Ms. H., after engaging in a lively conversation with a female student, turns to Eli and asks, slowly, “Did you do your homework?,” her intentions might have been pure. Perhaps Eli hadn’t done his homework consistently in recent weeks. Perhaps her voice slowed because she wanted to emphasize the importance of finishing homework. But the contrast between the “sparkling” intellectual exchange between the two White women and the tone and content of Ms. H.’s remark to Eli has social meaning beyond her intentions. And Eli read the meaning. In that context, the teacher’s actions say that Eli is not a legitimate or qualified party to her exchange with his friend, that his ideas were not valuable, and that he was first and foremost academically suspect rather than an intellectual contributor. Understandably, he experienced insult.

My point is not to criticize the teacher, or simply to highlight the potential for misunderstanding. Rather, it is to call attention to the fact that racial meanings are part of our social world, whether we like it or not, and that these meanings have profound effects, again whether we like it or not. Although in the cases discussed above the interactions between students and teachers might have occurred even if the students had been White, they wouldn’t have had the same meaning (though, depending on other factors, they might have been harmful in other ways). For this reason, cross-racial comparisons are often inapt: noting that White males are also sent to the dean when they are “disruptive” doesn’t take into account the semiotic differences between the two actions. (Such comparisons also typically fail to note differences in what is considered a sufficient basis for punitive action.) The question at issue is how the teacher-student relationships are constituted and how they evolve. In ordinary cases, expelling a Black male from the class will carry the message that males of color (or, more specifically, Black males) are not academic achievers, that their intellectual contributions are not valuable. This message affects everyone in the classroom and changes the relationships between students of color and other students, and between both groups of students and the teacher.^{Footnote 22} In the case of the White male it is much more likely, I submit, that he will be regarded as an individual bad actor, not a representative of his race. Moreover, hearing the message of epistemic disrespect for Blacks repeated over and over eventually leads to very damaging results, e.g., beliefs that Black males don’t have what it takes to succeed academically, or that the teacher is not trustworthy, along with emotional responses that are undermining, e.g., withdrawal, anger, and depression.

I suggested above that although individual attitudes, both conscious or unconscious, play a role in perpetuating social injustice, focusing on them is neither normatively apt nor the best strategy for promoting social justice. Although sometimes individuals do wrong and are blameworthy, individual agents are not usually responsible (causally or morally) for the fact that there exist social meanings that disadvantage some racial groups and privilege others. Individuals may be responsible for knowing and managing social meanings where they can lead to injustice, but this is not always feasible. And although change on the part of some individuals may make a difference, social meanings are sustained by much broader structures that attitudes alone cannot change, (e.g., segregation, wealth imbalances, cultural and linguistic differences, and the media). Although both attitudes and the unjust distribution of resources are both part of the story, a further element —racial semiotics—is a crucial part of the mechanism of racial injustice, and as Trayvon Martin’s black hoodie testifies, it can be deadly.^{Footnote 23}

By adding a third dimension to our understanding of racial injustice—the social/cultural dimension beyond the psychological and material—we broaden responsibility for the wrongs that occur. Only some of us are responsible for providing working computers (and such) to schools, and only some of us are responsible for incidents of bias or misinterpretation that occur in a high school classroom. But we are all, collectively, responsible for the social meanings that are dominant in our culture and are all responsible for the ways that social meanings are circulated and perpetuated. The emphasis on psychology offloads the moral responsibility to individuals who are engaged in the immediate interaction, but the challenge may be too much for them, given that, in many cases, there will be almost nothing they can do to avoid the meanings they elicit (everything they do will be interpreted in ways that continue to be problematic). The real solution has to be bigger and broader, and engage different levers for social change.

CREATING RACE

Let us return briefly to the claim that education (among other institutions and structures) creates race. I have been arguing that, at least in some contexts, it is valuable to shift from diagnosing the problem of educational inequity in terms of individual responses to race, and instead to analyze it as a process by which we collectively create or maintain race. To do so is to call attention to how our practices are implicated in producing the racial reality we take for granted. The shift also pushes us past the individual agent as the site of problematic “responses” to explore the social mechanisms—the practices, the relations, the meanings, and the structures—that entrench racial hierarchy. But what does it mean to create (or maintain) race?

I have mentioned before that in earlier work, I have argued that gender and race, in the core senses, should be defined in terms of social relations (Haslanger Reference Haslanger2000). Considered simply, a relation constitutes two (or more) groups whose members serve as relata for the relation. Consider the social categories of landlord and tenant. These are constituted by a social relationship—renting to/from—that holds between the members of each group, i.e., membership in each of the categories is determined by virtue of standing in the relation. This relationship is possible only against a larger backdrop of economic, political, and legal conditions. Members of each category vary in their identification with the category and in their experiences as members of the category; the norms associated with the categories vary from context to context; and individuals can be members of neither or both categories and can move from one to the other or out of both. Although there are obvious and important differences between the categories of landlord/tenant and race, I claim that racial categories are likewise constituted by social relations.

Since the 1970s, feminists have argued for the social construction of gender using the slogan: gender is the social meaning of sex. Although some feminists now reject this slogan, the point is to distinguish, on one hand, the anatomical features by virtue of which one counts as male or female, e.g., having XX or XY chromosomes and a variety of primary and secondary sex characteristics and, on the other hand, the relations that one comes to stand in by virtue of the social meaning of the sexed body.^{Footnote 24} In effect, having a body considered to be of a particular sex entails, socially, that one will or should participate in some practices and not others, clothe oneself in some ways and not others, and enjoy some activities and not others, etc. Of course, we don’t all follow the “rules” that define us in relation to each other by sex, but that doesn’t mean they don’t exist. Gender results from the imposition of sex-differentiating roles, norms, and identities (i.e., from the social meaning given to sex).

As I see it, there are important parallels between gender and race. Transposing the feminist slogan yields, I argue: race is the social meaning of “color.” “Color” in the intended sense (indicated by the use of scare quotes) includes not only skin color, but also those features in a particular context that mark the body as having presumed ancestral origins in a particular region of the world. So eye, nose, and lip shape, hair color and texture, height and physique can all count as elements of “color” in the contemporary American context. And just as the interpretation of someone’s gender is separable from the interpretation of their sex, it is not uncommon to find someone’s race distinguished from their “color,” as illustrated by the descriptions “oreo,” “coconut,” or “banana,” and without the same inner/outer metaphor, “white chocolate.” (In a more positive vein, it is not incoherent for a person “colored” white to refuse to be White.)

According to my account (see Haslanger Reference Haslanger2000), races are constituted not by just any social meaning of “color” but, in particular, meanings that situate individuals within a relatively stable social hierarchy. It follows from this view that to eliminate “color” hierarchy is to eliminate races, and this is something we should aim for. However, this is not to recommend genocide! Cultural and non-hierarchical ethnic groups may remain even where there are no races, and terms that now pick out racial groups such as ‘White,’ ‘Asian,’ ‘Black’ may come to pick out non-hierarchical social groups with the same members. (In short, I think it is indeterminate at this point whether names for races rigidly designate racial groups or social groups that may, at some point, cease to be races.) Although I have argued that the ordinary English term ‘race’ and particular names of races such as ‘White,’ ‘Black,’ ‘Asian,’ ‘Hispanic,’ ‘Native/Indigenous,’ and, increasingly, ‘Arab,’ refer to racial categories, this is less important to me than a recognition of the social categories, whatever we call them.

Putting together the claim that race is the social meaning of “color” with the (sketch of an) account of social meaning in the previous section, we can better understand how race persists even though many of us now recognize that it is not a biological or genetic category, do not want to be racist, and actively work to embrace racial diversity. Very roughly, all bodies are marked and interpreted in light of cultural schemas that group them in accordance with the local ideology (Sewell Reference Sewell1992). ^{Footnote 25} In the case of race, individuals are seen as having ancestral origins in a place and at a time that is taken to be significant, and there are cultural narratives explaining features of the group and its history, e.g., what members of the group have in common and why they are here. In effect, “color” has social meaning. The social relations defining both gender and race consist in patterns of treatment towards bodies as they are perceived (or imagined) through collective frameworks of salience. Just as pink means girl and blue means boy in the contemporary United States, the bodily markings presumed to be evidence of (relatively recent) ancestry in sub-Saharan Africa “mean” (among other things) intellectual inferiority, unless canceled or inflected by other markers, e.g., of class, gender, and ethnicity.^{Footnote 26}

“Colored” and “sexed” bodies are situated socially in light of the collective meanings. Race occurs when the ideology in question supports and gives rise to hierarchy. Such narratives will typically imply that a given group is “fit” for some activities and not others, and this, in turn, motivates and “justifies” their role in divisions of labor, social domains, and other forms of activity. For these roles, then, there are norms: criteria for what it takes to do that labor or activity well. For example, consider a context in which those interpreted as having a “black” body are regarded as suited for menial labor. Such labor entails a different relation to education and other social goods than professional work. Correspondingly, there are different norms for those who perform such labor.^{Footnote 27} Even allowing for cultural differences, what makes a good professional is not the same as what makes a good ditch digger, janitor, or maid. For example, a good menial laborer is strong, but not too smart or too creative, doesn’t think for him or herself, but shows up on time and follows the rules.

If we suppose that a society needs some individuals to perform menial labor and to perform it well, then (in the modern context) rather than coerce individuals to fulfill this role, it is better for them to be motivated to perform it.^{Footnote 28} So in this picture, ideally, we should designate a group for these jobs and have them internalize the norms, i.e., the norms must be incorporated as part of one’s identity and represent what would count as at least part of one’s success as an individual (Foucault Reference Foucault and Sheridan1995). This process usually involves developing unconscious patterns of behavior that reinforce the role in oneself and others and enable one to judge and be judged by its associated norms. In order for large groups of people to internalize similar or complementary norms, there must be a cultural vocabulary—concepts, narratives, images, scripts, and cautionary tales—that provide the framework for action.^{Footnote 29} The cultural vocabulary, of course, will be very complex: the scripts and images for any particular individual will have elements that pertain to the multiple social roles and norms she or he is expected to satisfy.^{Footnote 30}

If we consider the pattern of experiences that males of color encounter through primary and secondary education, a pattern reflected in the examples discussed in earlier sections, it is very tempting to conclude that the educational system is preparing them for positions that are outside, or at best at the margins of, the knowledge economy. The relations that are maintained (hierarchical, mistrustful, untrustworthy), responses that are encouraged (frustration, anger, passivity, deference), the participation expected (marginal, resistant, diffident), are not the makings of a knowledge-worker. Knowledge is a valuable resource, however, and systematic exclusion has serious long-term effects:

Any group that is denied full participation in the knowledge economy will be vulnerable to exploitation and ongoing subordination. Moreover, as we saw above, to be denied access to trusting epistemic relationships is wrong in itself, for it deprives one of the resources to become self-integrated into a social network that one can wholeheartedly endorse. So unless the educational system can address not only the problematic attitudes of individual teachers and the disparity in resources students of color face, but also the racial semiotics that provide the backdrop for all of the relationships in academic settings—relationships between teachers, students, parents, administrators, and employers, etc.—the result of education will remain a racialized economy that systematically situates persons of “color” in disadvantaged positions.^{Footnote 31} This is what it is to create race.