Responding to Sanist Microaggressions with Acts of Epistemic Resistance

Abstract People who have mental health diagnoses are often subject to sanist microaggressions in which pejorative terms to describe mental illness are used to represent that which is discreditable. Such microaggressions reflect and perpetrate stigma against severe mental illness, often held unconsciously as implicit bias. In this article, I examine the sanist attitudes that underlie sanist microaggressions, analyzing some of the cognitive biases that support mental illness stigma. Then I consider what responsibility we have with respect to microaggressions. I argue that all people share in a collective responsibility to engage in acts of epistemic resistance that challenge sanist attitudes so that it is easier for bystanders who witness microaggressions, and targets of microaggressions in particular, to identify microaggressions and to point out biased behavior. The act of pointing out bias is best understood as an act of epistemic resistance that is more effective and meaningful in the context of other acts of epistemic resistance. Ultimately, whether to point out bias is an individual decision that one must make after weighing the risks involved; engaging in a range of acts of epistemic resistance, on the other hand, is a moral responsibility everyone shares.

have mental illness, making mental illness represent the discreditable. I doubted she had any idea how stigmatizing her language was. To her, the words probably meant nothing; they were just exaggerated descriptions. To me, the words were full of meaning: I had been psychotic and suicidal when I was hospitalized. My colleague never would have used sexist or racist language in a conversation like this. Using language that discredits people with mental illness is still socially acceptable, however.
People who have mental health diagnoses are often subject to seemingly innocuous uses of language that are in fact pejorative insults that reinforce negative stereotypes of mental illness and consequently belittle the experience of people who have mental illness. Such microaggressions reflect and perpetrate stigma against severe mental illness, often held unconsciously as implicit bias. In this article, I examine the sanist attitudes that underlie sanist microaggressions, analyzing some of the cognitive biases that support mental illness stigma. Then I consider what responsibility we have with respect to microaggressions. I argue that all people share in a collective responsibility to engage in acts of epistemic resistance that challenge sanist attitudes. In addition, when we all share this responsibility, this makes it easier for bystanders who witness microaggressions to identify them and to point out when someone's behavior is biased.
Bystanders who point out bias risk many harms, especially bystanders who are also targets of microaggression. Bystanders thus face a difficult choice either to self-silence or to expend epistemic labor to help privileged people recognize their bias. For targets of microaggressions, both choices can result in contributing to one's own experience of epistemic harm, yet sometimes pointing out someone's bias is necessary for the perpetrator of a microaggression to recognize and change their behavior. The act of pointing out bias is best understood as an act of epistemic resistance that is more effective and meaningful in the context of other acts of epistemic resistance. Ultimately, whether to point out bias is an individual decision that one must make after weighing the risks involved; engaging in a range of acts of epistemic resistance, on the other hand, is a moral responsibility everyone shares.

Sanist Attitudes and Diminished Epistemic Credibility
Sanism consists of prejudicial attitudes toward people with mental illness, 2 reinforcing negative stereotypes that demean people with mental illness similar to how racism demeans people of color and sexism demeans women (Perlin and Dorfman 1993, 51; see also Perlin 1993;. Sanism produces "systematic subjugation" of people with mental health diagnoses (Poole et al. 2012, 20). People hold sanist attitudes when they treat people with mental illness as inferior, childlike, incompetent, violent, or dangerous. Sanist attitudes assume a psychocentric view of mental illness that conceives of mental illness as a medical or psychological problem of the individual rather than as a social problem with social causes (LeBlanc and Kinsella 2016). Because it is largely invisible, sanism is generally socially acceptable. Sanism is perpetuated by "ordinary common sense," a belief that because one sees the world a certain way, others must see it the same way and therefore one's perception of the world must be true (Perlin 2016, 117).
Sanism occurs on a continuum of privilege and is experienced differently by different people depending on how they are socially situated (Meerai, Abdillahi, and Poole 2016). People who are members of marginalized groups, such as those based on race, gender, or sexual orientation, commonly experience compounded prejudice, where they experience oppression within the mental health community as well as sanism within their marginalized community. 3 As a form of identity prejudice, sanism is a type of oppression like sexism and racism that feminist philosophical perspectives can help illuminate. This article draws upon feminist analyses of epistemic injustice to help us understand sanist microaggressions and epistemic resistance more deeply.
Sanism is based on stereotypes about mental illness. Stereotypes are generalizations about people based on a shared trait as a way to determine quickly what kind of person someone is (Fricker 2007, 30;Anderson 2010, 45; see also Manzo 2004;Rusch, Angermeyer, and Corrigan 2005;Livingston and Boyd 2010;and Courtwright 2013). Some of the common dominant stereotypes about people diagnosed with mental illness include the views that they are incompetent and unable to take care of themselves; that they are unpredictable, dangerous, and violent; that they have a character flaw such as deficient willpower; and that they are permanently defective (Eminson 2004, 123;Corrigan and Kleinlein 2005, 16;Rusch, Angermeyer, and Corrigan 2005, 530). 4 We also sometimes hold complimentary stereotypes, such as the belief that people with mental illness are especially creative, but these positive stereotypes are drastically outweighed by negative stereotypes (and are themselves problematic in tending to romanticize mental illness). Sanist attitudes reflect and reinforce the many negative stereotypes about mental illness.
Some of the most pernicious sanist attitudes, and the ones with which I am concerned here, involve attributing diminished credibility to people with mental illness on account of their illness. Epistemic credibility is the power of being trusted to provide reliable testimony and to be a competent participant in epistemic practices. We attribute credibility to people based on what we think we know about them. Credibility judgments can be warranted or unwarranted depending on whether our appraisal of the trustworthiness of the person and their testimony fits the evidence. Various factors influence our credibility judgments, including past reliability of the person we are assessing, social norms about complying with authority, emotional reactions, moral commitments, and other beliefs we hold (Origgi 2012, 227-33). Unjust credibility judgments are those that are based more on prejudice than on evidence. When we assume that people have less credibility than their actual situation warrants, based on negative stereotypes about them, this causes them to have an unjust credibility deficit based on their social identity, or what Miranda Fricker calls an "identity-prejudicial credibility deficit" (Fricker 2007, 28).
Credibility is granted within a "credibility economy" that governs our epistemic relations with one another (Marsh 2011;Origgi 2012). Access to knowledge-production and meaning-making is therefore intrinsically political (McKinnon 2016, 438). Although credibility is not a zero-sum game, in that one's credibility is possessed at the expense of another's diminishment, it is a comparative notion: some people have more credibility in relation to others (Medina 2011, 18-19). Credibility is a form of social power, involving the epistemic power to control one's social environment (Marsh 2011, 280-81). To be lacking in credibility is therefore to be lacking in social power and privilege. People who experience credibility deficits are epistemically marginalized; their diminished credibility prevents them from being able to participate fully in epistemic activities. Credibility is determined by those in power: those who have epistemic credibility have the power to grant credibility to others. People who have diminished credibility, on the other hand, are not taken seriously as epistemic agents, so their judgments of other people's credibility are discounted.
People who have severe mental illness, especially psychotic disorders like schizophrenia or bipolar disorder, are often judged as having less credibility because they are seen as incompetent, defective, unpredictable, and irresponsible-all of which makes them appear to be unreliable testifiers and incapable of participating meaningfully and appropriately in epistemic practices. 5 In this way, people with mental illness are often subjected to testimonial injustice, in which their testimony is discounted because of their social identity (having mental illness), as well as to hermeneutical injustice, in which they lack the hermeneutical resources to understand their situation because of their social identity (Lakeman 2010;Sanati and Kyratsous 2015;LeBlanc and Kinsella 2016;Scrutton 2017;Kurs and Grinshpoon 2018). People who have mental illness are often seen as lacking a sense of epistemic relevance and lacking the capacity to provide information that can make a meaningful epistemic contribution (Carel and Kidd 2017, 339-40;Kidd and Carel 2017, 180); in addition, they are often precluded from hermeneutical understanding through strategies of exclusion from epistemic practices and through strategies of expression, where they are not taken seriously because of their communicative style (Carel and Kidd 2017, 342;Kidd and Carel 2017, 184-85).
Sometimes people do have diminished epistemic capacities because of their illness, and in such cases diminished credibility may be warranted. For example, symptoms such as disorganized thinking, delusions, paranoia, poverty of thought, and uncontrollable or inappropriate emotions can all affect rational processes involved with various epistemic practices. Even when people do demonstrate incompetence in some area, however, this does not necessarily justify reduced credibility. We tend to overgeneralize and assume that incompetence in one area means incompetence in all areas. But just because a person has delusions or hallucinations does not mean they lack relevant capacities for all epistemic participation (Sanati and Kyratsous 2015;Kurs and Grinshpoon 2018). Assuming global incompetence in a person with severe mental illness, even when they have florid psychosis, is inaccurate, unfair, and unjust.
Certain cognitive biases reinforce sanist attitudes toward mental illness. With shared reality bias, people align their own views about mental illness with culturally dominant views, endorsing stereotypes such as those listed above in a way that leaves no room for alternative views (Anderson 2010, 46;2012, 170). With illusory correlation bias, people see correlations where there are none, developing stereotypes based on unusual occurrences that are nonetheless vivid in the popular imagination (Anderson 2010, 46;Balzan et al. 2013, 320). Stereotypes that people with mental illness are dangerous, unpredictable, and violent are based on newsworthy occurrences of violence that have little connection to people's day-to-day experience with mental illness.
In the case of bias toward in-group favoritism, people have a bias for members of their own group (Anderson 2010, 46). People identify those who have mental illness as unlike themselves in order to affirm their own mental health status in opposition: unlike those with mental illness, they are healthy, rational, and sane. With this affirmation, they solidify their own identity, defining themselves in relation to what they are not (Ottati, Baudenhausen, and Newman 2005, 107-8;Gosselin 2018, 82). Mental illness represents the "Other" because it is among those things we most fear. Because it threatens our rationality, volitional control, and subjective awareness, "losing our mind," as we say colloquially about mental illness, suggests the loss of our identity as human beings. In-group favoritism that exalts one's sanity and rationality by shunning those with mental illness as "Other" protects people from identifying with such loss.
In surveys, many people will admit to stigmatizing mental illness and directing discriminatory actions toward people with mental illness (for example, Repper and Perkins 2003, 35-36;Stuart 2006). 6 Stigma is not always so blatant, however. Much of the stigma directed at people who have mental illness is held unconsciously as implicit bias and is directed at them in the form of microaggressions.

Microaggressions and Implicit Bias
Microaggressions consist of speech or behaviors that subtly demean a person based on their marginalized identity. Microaggressions have been classified into three types: microassaults (conscious speech or behavior intended to hurt a person), microinsults (subtle communications that demean a person's marginalized identity), and microinvalidation (speech or behavior that excludes or negates a person's thoughts, feelings, and experience) (Sue et al. 2007, 274;Sue et al. 2008, 330-31). They can be verbal, nonverbal, or environmental (Sue et al. 2008, 331). Microaggressions typically fit several common themes, including in a racial context: being seen as a foreigner, being assumed to be less intelligent or educated, being assumed to be a criminal, being seen as a secondclass citizen, exoticization or presumed similarity, color blindness, denial of individual racism, the myth of meritocracy, pathologizing certain cultural values or communication styles, and environmental invalidation (Sue et al. 2007, 275;Nadal 2011, 477). People can experience multiple types of microaggressions at the same time, depending on how they are socially situated; Black people, for instance, can experience sanist and racist microaggressions simultaneously when they are subject to anti-Black sanism. The type of microaggression I am considering in this article consists of microinsults that regard people with mental illness as inferior (sometimes other than or less than human), incompetent, and lacking epistemic credibility; my colleague used this type of microaggression in our conversation.
Microaggressions lead to many harms, including feelings of powerlessness, feelings of invisibility, forced compliance and loss of integrity, and pressure to represent one's group (Sue, Capodilupo, and Holder 2008, 333-34). They can cause emotional, behavioral, and cognitive responses and reinforce larger structural inequities (Friedlaender 2018, 7-8) as well as contribute to "everyday suffering" (Fleras 2016, 9). The harm of microaggressions is cumulative, so any given microaggression can have greater impact than a single act might have. Perpetrators of microaggressions often feel that targets of microaggressions are overreacting, but this is because they are ignorant of the systemic effects of microaggressions. Perpetrators tend only to see the single act they committed, but targets see the act as yet one more in a long line of such acts.
The power of microaggressions lies in their invisibility (Sue et al. 2007, 275). Microaggressions, particularly microinsults and microinvalidation, are usually unconscious and unintended, committed by well-meaning people who see themselves as good and decent people and who do not recognize the harm they cause by their unconscious utterings and behavior (Sue et al. 2007, 275;Fleras 2016, 9-10). Not recognizing the microaggressions they commit, perpetrators tend to believe they are acting according to dominant moral norms. As a result, they may become defensive when their actions are identified as prejudicial, viewing themselves as being "unjustifiably reprimanded for morally acceptable behavior" (Friedlaender 2018, 13). In a sanist society where negative views toward the mentally ill are socially acceptable, people may not recognize when they are committing sanist microaggressions and may become defensive at being identified as sanist. Since they contribute to oppression unknowingly and despite good intentions, perpetrators of microaggressions commonly commit civilized oppression (Harvey 2007).
Microaggressions benefit perpetrators, as microaggressions reinforce existing power structures that privilege the perpetrators (Friedlaender 2018, 9). Not only do perpetrators typically not recognize when they are committing microaggressions, but also targets of microaggressions often second-guess themselves about whether what they thought happened really did happen (Sue et al. 2007, 279). Microaggressions prioritize the perspective of the seemingly innocent perpetrator who claims not to have done anything wrong at the expense of the perspective of the target of microaggressions who is made to second-guess their experience (Fleras 2016, 3). This perpetuates existing power relations by diminishing the reality of the target's experience while affirming the perception of the perpetrator's experience.
Microaggressions also reinforce the belief that existing power structures are just. Committing microaggressions toward others reinforces the belief in a meritorious world where people who have mental impairments are somehow at fault for their impairments and thus deserve to be discredited, demeaned, and excluded (Corrigan and Kleinlein 2005, 24-25). A perpetrator subjecting a person to epistemic injustice helps confirm the perpetrator's belief that the perpetrator deserves their social, career, and personal success (whereas the subjugated person does not), thus reinforcing their epistemic privilege (Alcoff 2010, 134;see Gosselin 2018, 85). Similarly, a perpetrator committing a microaggression helps confirm the perpetrator's belief that the perpetrator is superior to their target (people who have mental illness) and thus deserves their privilege.
Unconscious and unintentional, microaggressions are usually based on implicit bias. Implicit bias is a negative attitude held unconsciously or unknowingly toward people who share a certain trait. We develop implicit bias when we internalize culturally dominant, negative stereotypes about groups of people (Gendler 2011;Holroyd, Scaife, and Stafford 2017b). Implicit bias affects our comportment toward people, including our body posture and positioning; our interpretations of people's action and speech (see Peet 2017); the language that we use; and macro behaviors such as hiring decisions. Biased attitudes toward people are implicit in the sense that they are not held consciously; similarly, actions that result from implicit bias are done without intending to be negative and without even realizing that they are negative (Holroyd, Scaife, and Stafford 2017b, 4). People are often ignorant of their biases, not recognizing when their language and behavior is negative toward certain groups of people.
Biased attitudes toward people are also implicit in the sense that they may be beyond a person's control (5); because they are unconscious, a person may have no conscious means of correcting them even when the person is capable of recognizing them as biased. Implicit bias is thus automatic in the sense that it is not initiated through conscious effort and cannot be revised based on rational considerations (Holroyd 2012). People professing conscious views that go against their implicitly biased behavior results in cognitive dissonance that people may feel motivated to try to address. Implicit bias is thus sometimes bias that is unendorsed (Holroyd, Scaife, and Stafford 2017b, 5-6). Because biased behavior occurs without intending or even recognizing its bias, however, it can feel impossible to exert conscious control over it to change it. Implicit bias tends to be supported by beliefs or belief-like states that are negative toward a certain group of people based on a trait they share. Since they are held unconsciously, the beliefs that undergird implicit bias may not be fully formed or endorsed beliefs but rather a different kind of belief-like state. For example, Eric Schwitzgebel proposes they are "in-between beliefs" in that they do not track our dispositions consistently (Schwitzgebel 2010). Neil Levy suggests that they are "patchy endorsements" that have propositional structure but are not responsive to reasons the way regular beliefs are; because they do not stand in inferential relationships with other attitudes, they cannot be integrated into a person's other mental states and so in a sense cannot be attributed to the person (Levy 2015). Tamar Szabό Gendler understands them as "aliefs," states that are constituted by sets of three kinds of content that are co-activated by one's environment: representational content, affective states, and behavior (namely readying of motor responses) (Gendler 2008; see also Albahari 2014). In contrast to beliefs, aliefs are associative, automatic, arational, and activated by a person's environment.
These are all plausible interpretations of implicit bias, and it may be that there are different types of implicit bias that people hold that these accounts characterize. The account that seems to capture features of implicit bias most fully, in my view, especially with respect to the microaggressions I am considering here, is conceiving of them as aliefs: the implicit biases that underlie microaggressions are associative clusters of affective states and behavioral tendencies. They are automatic, arising without conscious awareness; they are arational, not responding easily to reasons or evidence that opposes them; and they are activated by a person's environment as behavioral tendencies that arise in response to a person's situation. There is some debate over whether a person has, or ought to have, control over these belief-like states, as I shall explain below.
Sanist implicit bias was likely the basis for my colleague's use of derogatory language toward people with mental illness, as she appeared unaware that her language had any meaning other than the force of exaggeration. This gap between my colleague's obliviousness about using language that stigmatizes mental illness and my profound discomfort in hearing it constitutes what José Medina calls "epistemic friction," a place where there are alternative epistemic viewpoints that should be evaluated in relation to each other (Medina 2013, 48-49). Epistemic friction encourages scrutiny and self-criticism, making people examine what they had previously taken for granted (176). It requires people to confront and interrogate their ideas in relation to alternative perspectives that become highlighted. Epistemic friction invites a shifting of epistemic viewpoints as epistemic equilibrium is sought. Both "object-level blindness," or ignorance of the nature of the situation, and "meta-blindness," or ignorance of one's ignorance, are examples of epistemic friction (148-49). Epistemic friction occurs especially in places where a person is ignorant to features of the world around them, in particular to features of the people they interact with, where such features are actually salient to the nature of interactions.
Only people who are critically reflective can recognize when epistemic friction occurs and feel motivated to try to understand and rectify the circumstances causing the friction. Since my colleague was oblivious to this friction, it was up to me to decide whether to bring her attention to it. This is a common situation: because perpetrators of microaggressions are frequently (and often willfully, as I explain below) ignorant of the effects of their behavior, the responsibility of responding to microaggressions often falls on witnesses of microaggressions, and in particular targets of microaggressions.
Because the intentionality behind microaggressions is often unclear, targets of microaggressions typically have difficulty figuring out how best to respond to them (Mercer et al. 2011, 457). Targets of microaggressions often have to do a sanity check, where they check with other members of their marginalized group to verify if what they perceived happen did in fact happen (Sue, Capodilupo, and Holder 2008, 332). This adds to the cumulative harm that targets of microaggressions experience, subjecting them to both testimonial injustice and hermeneutical injustice in making them feel as if they are perceiving something no one else perceives (Friedlaender 2018, 16). I second-guessed myself for a long time after my colleague's remarks, wondering if what I thought happened really happened. When I decided that it did, I felt very uncertain whether I should bring it to her attention because I was not sure what the effects of doing so might be.

Responding to Microaggressions
The fact that many microaggressions tend to be unconscious and unintentional makes it difficult to determine what kind of responsibility a perpetrator has. It may be helpful first to consider what kind of responsibility a person has for implicit bias. Philosophers debate whether we are responsible for holding implicit biases in a backward-looking way (in terms of blameworthiness) when we are often not even aware that we possess them. For example, Levy argues that we are not responsible for mental states of which we are unaware because without our awareness they cannot express who we are (Levy 2011;. Michael Brownstein, on the other hand, argues that we are responsible for implicit biases because they reflect who we are, emanating from our dispositions and from what we care about (Brownstein 2016). Jules Holroyd, Robin Scaife, and Tom Stafford add that we are responsible for implicit biases insofar as we ought to be aware of them (Holroyd, Scaife, and Stafford 2017a). In addition, philosophers also debate whether we are responsible for holding implicit biases when we do not have direct control over them. Jennifer Saul suggests that we are not responsible for our implicit beliefs in a backward-looking way because we do not have control over them: the circumstances that led to our developing them are out of our control and we do not have immediate and direct control over changing them (Saul 2013). Holroyd counters that although we cannot directly control our beliefs by changing them at will, we can have indirect control over them: we can control how we interpret and weigh the evidence on which our beliefs rest, and we can adopt commitments to reject prejudicial beliefs (Holroyd 2012; see also Holroyd, Scaife, and Stafford 2017a). Thus, we ought to take control over them to the extent that we can, and we are blameworthy for implicit biases if we do not work on changing them.
In light of these arguments, what seems most fair to say is that we are responsible in a forward-looking way to uncover our implicit biases and to confront them by actively working on changing them. To do otherwise would be to be negligent (Riggs 2012). This corresponds with Christina Friedlaender's argument that perpetrators of microaggressions may not be blameworthy for speech and behavior that they do not intend and are not even conscious of, but they are responsible for it in a forward-looking way through Iris Marion Young's social connection model in which shared responsibility is discharged through collective action (Friedlaender 2018, 17). Individuals have to work together to educate themselves about microaggressions, to learn to recognize when their thoughts and behavior might constitute microaggressions, and to try to reduce implicit bias against marginalized groups (18). Perpetrators of microaggressions may have some backward-looking duties to make amends for their actions when they are able to recognize the harm their actions cause. But the emphasis is on forward-looking duties that are aimed at individuals preventing themselves from committing microaggressions in the future. Only when individuals work together to promote critical self-reflection and self-education do they change the social environment in which microaggressions occur.

Difficulty Accepting Responsibility
Although it is easy to see what responsibility individuals have to prevent microaggressions from occurring, it is hard to motivate perpetrators of microaggressions to recognize the harm their actions cause and to want to change. It can be difficult to acknowledge that we have implicit bias or to recognize that we commit microaggressions because not only is the bias held unconsciously-without our awareness and without our direct control over it-but also we have self-affirming tendencies that prevent us from noticing our own bias. Sometimes we need someone else to point out our biases and identify our actions for what they are. Because of this need, people who witness microaggressions, whether or not they are their targets, may have a responsibility to identify them to perpetrators so as to invite perpetrators to recognize their actions for what they are and to be critically self-reflective in changing them. We might think of this as a bystander responsibility to identify microaggressions and implicit bias when they are observed. There are many barriers to carrying out this responsibility, however.
Identifying microaggressions and pointing out the ways in which someone might be biased can cause harms such as discomfort (sometimes extreme discomfort), embarrassment, hurt feelings, loss of credibility, and retaliation. Pointing out bias involves bringing attention to someone's moral flaw, making them seem like a wrongdoer. This may diminish the perpetrator's credibility, for they may no longer appear to be impartial and disinterested participants in epistemic activity; they may appear to have a motive that their bias serves, or they may appear morally indifferent to the effect of their bias. Pointing out bias can also cause a perpetrator to feel embarrassment and other negative emotions. Feeling like they need to protect perpetrators of microaggressions from embarrassment and hurt feelings, targets of microaggressions often feel like they have to "rescue" perpetrators by trying to take care of the perpetrator's feelings (such as by brushing over the offense or relieving the perpetrator of any responsibility for apology) before taking care of their own feelings (Sue, Capodilupo, and Holder 2008, 332). Moreover, as noted above, perpetrators of microaggressions tend to believe they are doing nothing wrong because they are acting according to socially acceptable norms (such as socially acceptable sanist attitudes). As a consequence, they are likely to react defensively, demonstrating ways that they believe they are not biased, and perhaps attacking the credibility of the bystander pointing out the bias. Bystanders who have considerably less power than those whose bias they are identifying, such as those who are members of marginalized groups, have significantly more to lose in pointing out bias. Being on the receiving end of a defensive reaction can be very uncomfortable; a person can also suffer greater harms, such as loss of credibility or retaliation, when the biased person has the authority to judge others' credibility or the power to cause harm through retaliation (such as when a bystander reports discrimination at their workplace).
Pointing out bias can be harmful to bystanders who witness biased behavior, but the potential for harm is especially acute for targets of microaggressions. Then the discrediting that they are subjected to as a member of the marginalized group that is the aim of the microaggression is applied to their act of pointing out bias: since they are seen as unreliable or incompetent epistemic agents, their act of pointing out bias is discounted. The person can be seen as "making a big deal" out of something that is seen to be trivial, for example, the use of language; or for what is seen as trivial reasons, for example, because the language feels personally offensive. This further downgrades the credibility of the target. As noted above, when perpetrators of microaggressions fail to see the systematic effect of microaggressions, and fail to recognize the cumulative harm they cause, perpetrators interpret targets of microaggressions as overreacting to what seem to perpetrators to be only a single, insignificant act. Interpreting targets as overreacting is a way of discrediting them.
On the other hand, if the perpetrator of a microaggression perceives the target of the microaggression in a way that does not fit the stereotype assumed by the bias, the target might be seen as an exception: "Those people with mental illness are psychotic, suicidal, maniacs, and so on-but you're not like that; you're not one of them." This exceptionalism bias treats the person who does not appear to fit the stereotype as the exception that proves the rule, reinforcing the social hierarchy that subjugates people in marginalized groups. From a sanist perspective, this exceptionalism bias reinforces the social hierarchy that privileges "sane" people and subjugates people with mental illness as inferior, incompetent, pathetic, and even different from or less than human. I suspect my colleague sees me in this way, and that is why it never occurred to her that her language could be hurtful to me. She does not see me as psychotic, suicidal, or crazy-even though I have been all of these. Seeing someone who is clearly a member of the marginalized group as an exception is a way of discrediting them by understanding their situation and perspective in a false and harmful way, alienating them from a core part of their identity.
Pointing out bias to my colleague could also be seen as confirming various stereotypes. I could be interpreted as personally confirming a stereotype of White women with mental illness who are "hysterical," high-strung, or over-emotional 7 if I am seen as someone who gets offended easily. Black people who point out bias can be seen as confirming a stereotype of angry, aggressive, or "uppity" Black people, or paranoid Black people who see themselves as victims of White people's action too readily. When Latinx people point out bias, they can be seen as confirming a stereotype of hotheaded or feisty Latinx people, or lazy Latinx people who, again, see themselves as victims of White people's actions too readily. People commonly hold expectancy-confirming biases in which they interpret evidence in ways that confirm what they already believe (Ottati, Baudenhausen, and Newman 2005, 109-10). When pointing out bias is seen to confirm stereotypes, this is another way of dismissing a person's concern and seeing the person as discreditable. Bystanders who witness microaggressions may be subject to many potential harms, but targets of microaggressions risk even more.
All of these negative reactions to having one's action be identified as a microaggression and to having one's bias pointed out are ways that perpetrators resist taking responsibility for their thoughts and behavior. The failure of a person with privilege to understand the perspective of people who are marginalized is a "contributory injustice" that creates epistemic harm (Dotson 2012). What we know and how we know is shaped by how we are socially positioned; what we can do as epistemic agents depends on how much and what kind of access we have to resources and to opportunities to develop and practice our capabilities (Pohlhaus 2012). People who have privilege usually adopt dominant epistemic perspectives and often do not have the means or motivation to change their views or to develop alternative understandings. People like my colleague who do not suffer from severe mental illness typically do not find themselves in situations where they are required to think about their own mental health status and whatever disability may arise from it, and consequently they often fail to pay attention to other people's mental health status, possible impairments, and whatever effects these have on the way they position such people in relation to others. Even if my colleague had some knowledge of some mental disorders, she is likely to have less understanding of the perspective, needs, capabilities, and resources of someone with a psychotic disorder than of someone who has other, less severe disorders. Thus, it might never occur to her that to use language that disparages people who experience psychosis or suicidality could be harmful to such people. Moreover, because her perspective corresponds with dominant negative views about psychosis and mental illness more generally, she has no need or even motivation to develop alternative perspectives unless a situation of epistemic friction demands it.
Not only do privileged people not often find themselves in circumstances that would require them to challenge their dominant perspectives, but they also typically are not motivated to do so. Privileged people sometimes lack the motivation to become aware of their biases, because this would threaten their self-concept as a "good" person and threaten existing power relations in which they possess superior epistemic power. For this reason, perpetrators of microaggressions often become defensive about their actions, as noted above. The desire to avoid learning about one's biases, such as those around mental illness, is a form of what Gaile Pohlhaus, Jr. calls "willful hermeneutical ignorance" (Pohlhaus 2012). Willful hermeneutical ignorance occurs when people who have privilege rely on dominant understandings that they develop through their privilege and refuse to learn alternative ways of understanding, particularly understandings that come from marginalized perspectives. People who adopt dominant sanist perspectives may be resistant to learn about perspectives of those who have mental illness because doing so threatens their self-conception as well as the existing power relations that benefit them as "sane" people. When perpetrators of microaggressions resist taking responsibility for their thoughts and behavior, it typically falls on bystanders witnessing microaggressions-and, in particular, targets of microaggressions-to decide whether and how to respond.

Acts of Epistemic Resistance
Bystanders who witness a microaggression typically have two options in deciding how best to respond. They can ignore the microaggression, consequently silencing themselves, or they can point it out to the perpetrator. Both responses can involve contributing to epistemic harm. When I laughed alongside my colleague because I was too uncomfortable to object to her use of language, I was contributing to epistemic injustice, not only letting her get away with causing epistemic harm by not pointing it out to her, but also leading her to believe, by my laughter, that I saw the world the way she did and that I endorsed her use of language that vilified people with severe mental illness. When targets of microaggressions silence themselves in this way, they contribute to harm against themselves. In appearing to endorse the idea that people with severe mental illness are inherently discreditable, I participated in creating epistemic harm to myself as one who is part of the group of people with mental illness. Self-silencing is a form of testimonial injustice that Dotson calls "smothering," in which a person selfcensors because they know they will not be believed or taken seriously (Dotson 2011). When other people seem to lack the hermeneutical resources to make sense of a marginalized person's perspective, it can seem pointless and even harmful to share that perspective. Remaining silent might appear to be the best response. Remaining silent is a way of being complicit in epistemic injustice, however, by seeming to support the actions that cause epistemic harm.
Instead of silencing themselves, bystanders who witness microaggressions can respond to willful hermeneutical ignorance by performing what Medina calls "acts of epistemic resistance" (Medina 2013, 48-50). These acts are not singular achievements of extraordinary individuals but rather chained actions that are echoed by others, creating a social network of people who through their collective acts of resistance are able to change social norms (225-26). They are chained actions in both a forward-looking and a backward-looking sense. In the backward-looking sense, acts of resistance are linked to previous acts that create the circumstances that enable a given act to be meaningful. Acts of resistance are only recognized as acts of resistance within a certain social context that enables people to understand their significance. In other words, any individual's action is only powerful enough to be echoable by others when it is performed in a context that allows for its recognizability, that is, when there is a social environment that can make sense of such action. In the forward-looking sense, acts of resistance are linked to the acts of other people who echo the action, repeating it in their own way and within their own context. These acts are thus powerful enough to inspire others to act similarly. The group of people who perform these echoable actions constitute a social network of people who share some of the same epistemic goals, to increase hermeneutical resources by making a situation better understood.
Hermeneutical heroes are individuals who perform these chained actions in ways that are recognized and imitated by others and who stand out not because their acts are significantly different from the acts of others but rather because there is something about their circumstance that makes their acts particularly inspiring to others (111 and 114). For example, after Ken Steele published his story about his experiences with schizophrenia in New York City Voices, other people with schizophrenia began to share their stories, and many families who had been silenced by stigma became comfortable talking about their family members and asking for help (Steele and Berman 2001, 219-39). Steele was a hermeneutical hero because he was a trailblazer in sharing his personal story in many New York state and in national venues; what he did was what many people with mental illness do in sharing their stories, but his act was particularly inspirational to many who later echoed his action. Hermeneutically heroic acts are thus emblematic of everyday acts of resistance. Engaging in hermeneutical heroism is not necessary for social change, but performing chained, echoable actions is.
Targets of microaggressions are specially situated to engage in certain acts of epistemic resistance because of their marginalized position. Although someone who is hermeneutically marginalized may have to overcome more barriers to be able to develop dominant perspectives and the resources and capacities that are associated with privilege, they have other epistemic capabilities and resources that are unique to their marginalized position (Alcoff 1999;Wylie 2003). For example, Medina argues that marginalized people are better situated to develop what he calls "kaleidoscopic consciousness," or the ability to perceive and even hold multiple perspectives rather than simply the dominant one or simply dual perspectives (Medina 2013, 73-74 and 200). Having kaleidoscopic consciousness enables a person to be both self-aware and aware of how others see them (190)(191)(192)(193)(194)(195)(196)(197)(198)(199)(200)(201). Hermeneutically marginalized people can develop kaleidoscopic consciousness because they are sometimes forced by existing power structures to understand and live according to dominant perspectives, yet they also develop marginalized perspectives through the ways in which they are situated in relation to these power structures. For example, psychiatrist Kay Redfield Jamison, who has manic depression, and law professor Elyn Saks, who has schizophrenia, both have kaleidoscopic consciousness because they are both professionals who help mental health clients, which enables them to adopt a dominant perspective through this role, and mental health clients themselves, through which they adopt a marginalized perspective (Jamison 1995;Saks 2007).
Whereas privileged people commonly develop only dominant perspectives and develop only the epistemic resources and capacities required to succeed in a context of privilege, marginalized people frequently develop both dominant and alternative perspectives and can draw upon a wider range of resources and capacities in determining how to act. This gives them greater flexibility in responding to situations of epistemic friction and engaging in acts of epistemic resistance. Possessing multiple perspectives, Jamison and Saks, for instance, are well positioned to respond when their colleagues make disparaging or ignorant remarks about mental illness because they can share their insider knowledge of what it is like to be mentally ill in a way that those in dominant positions can understand and appreciate.
When marginalized people such as the targets of microaggressions perform acts of epistemic resistance, they provide the hermeneutical resources for people with privilege to be able to recognize their privilege and to confront their biases. When targets of microaggressions perform acts of epistemic resistance, they must put aside emotional responses to experiencing stigma such as shame, anger, or indifference-all common responses to mental illness stigma (Watson and River 2005)-for the sake of the greater good of helping others increase their hermeneutical resources and capabilities. They are thus epistemically generous, sharing their own epistemic resources and capabilities in the service of developing these in others (Baehr 2011, 110-11).
Bystanders who witness microaggressions but who are not themselves the targets of them may also perform acts of epistemic resistance. Because they have privilege that targets of microaggressions lack (not being targets themselves), however, they have different epistemic resources to draw upon and thus different ways and degrees of sharing these. A bystander who is not the target of a microaggression can point out the perpetrator's bias from a relatively privileged position, which in some contexts gives them greater authority so they will be taken more seriously by the perpetrator, but which also reinforces existing power relations that privilege some (for example, "sane" people) and burden others (for example, people with mental illness).
When a bystander who is the target of a microaggression points out a perpetrator's bias, on the other hand, this challenges existing power relations by asserting performatively the importance of their marginalized perspective over the dominant perspective in which microaggressions occur. Although targets of microaggressions are vulnerable to many harms identified above, they also have a special power by virtue of their position that is recognizable by others and thus echoable by others, even if it is not always recognizable to people who commit microaggressions. When a target of a microaggression draws on their firsthand experience to dispute the perpetrator's assumption that the perpetrator has done nothing wrong, this affirms the reality of the target's experience and challenges the idea that existing power relations are just. Bystanders who are not targets of microaggressions can make an important epistemic contribution in pointing out a perpetrator's bias, and may in some cases be taken more seriously, but targets of microaggressions can challenge existing power relations in an especially powerful way because of their marginalized status. Even when targets of microaggressions are subject to epistemic harm and discrediting by people whose bias is being identified, they exhibit power that is recognizable, appreciated, and echoable by others when they challenge existing power relations.
An objection may be raised to my suggestion that we conceptualize the act of identifying microaggressions and pointing out bias as an act of epistemic resistance that some people are specially positioned to carry out, and thus have a special duty to perform: people who are marginalized have special hermeneutical resources that can help those who are made ignorant by their privilege, but it seems unfair that they have a responsibility to help the privileged as opposed to requiring the privileged to help themselves. Expecting marginalized people to educate privileged people about their privilege and the ways in which they perpetuate oppression is a form of epistemic exploitation (Berenstain 2016). When a marginalized person feels that it is their duty to help privileged people gain the hermeneutical resources they need, they may be internalizing this epistemic exploitation, contributing to their own experiences of injustice in a different way. Internalizing epistemic exploitation is thus another way of committing epistemic harm to oneself that commonly occurs to targets of microaggressions.
In assessing whether the epistemic labor involved in performing acts of epistemic resistance is worth the effort and the risk of potential harm, we must examine the power relation between the bystander who identifies the microaggression (who may or may not be its target) and the perpetrator of the microaggression. When the perpetrator has significantly greater social power, there is much greater potential harm to the bystander and greater epistemic exploitation of someone who is already marginalized. When the bystander has more equal social power relative to the perpetrator, on the other hand, this epistemic labor is not as exploitative or otherwise harmful, and it is easier, more justifiable, and perhaps even more obligatory for a person situated in this way to perform acts of epistemic resistance. People's intersectional identities are thus relevant in assessing harm. In my own case, although I am marginalized with respect to mental health status, my colleague and I are equally privileged in many other ways as middle-aged, White, female, heterosexual professors. Because of my relative privilege, I may have more of a responsibility to draw my colleague's attention to her bias than someone who is more marginalized because I have relatively less to lose in terms of epistemic harm.
Even when they have otherwise relatively equal social power, however, targets of microaggressions can be so marginalized due to the social identity that is targeted that other social identity features may be irrelevant. Targets of microaggressions are especially vulnerable to epistemic injustices that undermine their credibility, further marginalizing them. As noted above, people with mental illness are particularly vulnerable to testimonial injustice and hermeneutical injustice; the diminished credibility associated with these can be so marginalizing that not only may people with mental illness not be taken seriously when pointing out bias, but they may also suffer significant harm. This is why it is so important that pointing out bias occurs in a larger context of epistemic resistance where many different kinds of actions resisting sanist attitudes occur: the context helps confer credibility that is otherwise diminished.
In order to be taken seriously, therefore, the act of pointing out bias must be done in a context where other, similar actions are performed so that the act is not viewed in isolation but rather as a systemic response to a systemic injustice. It is in this way that any individual action contributes to social change. Responding to microaggressions by identifying them and pointing out the bias that underlies them should be conceptualized not simply as the duty of an individual bystander but rather as an act of epistemic resistance that is part of a set of chained, echoable actions. Another way to think about this is that it is one of several ways that individuals can contribute to the shared responsibility that individuals have with respect to microaggressions (Friedlaender 2018).
All individuals have a responsibility to contribute to this shared responsibility by engaging in whatever acts of epistemic resistance they can. By engaging in a range of such acts, individuals create the social environment in which any single act-such as identifying a microaggression and pointing out the bias that underlies it-is part of a systemic effort at change and is thus made more meaningful within that context. Since pointing out someone else's bias can lead to a range of potential harms, especially for targets of microaggressions (as noted earlier), a bystander who witnesses a microaggression must make an informed decision about whether the risk of harm is worth the effort. In a social environment that normalizes acknowledging and confronting bias, pointing out bias is more easily done. To create this social environment, however, all individuals must engage in a range of acts of epistemic resistance that involve confronting and changing dominant perspectives such as sanist attitudes.
Many kinds of acts of epistemic resistance can be employed to counteract sanist attitudes besides identifying microaggressions and pointing out bias. One important form of epistemic resistance that individuals with mental illness can engage in, and that may be more powerful than pointing out bias, at least initially, is self-disclosure. The primary way that people overcome stigma is through interpersonal contact with people with known mental illness (Couture and Penn 2003;Corrigan et al. 2012). Steele says, "A bold but necessary move, self-disclosure is a first step toward successfully addressing the stigma associated with being mentally ill" (Steele and Berman 2001, 221). When people know others with mental illness, their attitudes toward mental illness soften and shift.
Another important act of epistemic resistance, which all individuals should engage in, is educating themselves about mental illness experience with an openness to learning about alternative ways of viewing such experience. Targets of microaggressions can contribute to this effort by sharing their understandings of their mental illness experience through firsthand accounts as a way to increase education about mental illness experience. All individuals collectively can make the effort to seek out firsthand accounts in order to better educate themselves.
All firsthand accounts provide valuable perspectives, but individuals should be especially receptive to alternative narratives that defy the dominant paradigm of viewing mental illness as chiefly a disease or a psychological problem. Mohammed Abouelleil Rashed identifies three themes of what he calls "Mad narratives," narratives created not by scientists or therapists but by people with mental illness themselves as ways to explain their experience outside of these dominant paradigms. Some of the themes he identifies are viewing mental illness as a spiritual transformation (Rashed 2019b; see also Razali, Khan, and Hasanah 1996;Morrison and Thornton 1999;Kabir et al. 2004;Endrawes, O'Brien, and Wilkes 2007;Waldron 2010;Crabb et al. 2012), a dangerous gift (see also Dubrul 2014; International Bipolar Association), or "healing voices" that can aid personal growth (see also Romme et al. 2009;McCarthy-Jones 2012, 315-31;Longden, Corsten, and Dillon 2013;Romme and Morris 2013;Thomas and Longden 2015). Social justice narratives that explain mental illness and recovery in terms of social marginalization and collective struggle (Harper and Speed 2012;Morrow and Weisser 2012) are other alternative narratives that should be sought out. Narratives that analyze the social causes of mental illness, such as those that analyze the social causes of depression in women (for example, Ussher 2010; Gosselin 2014), and narratives that challenge the dominance of the pharmaceutical industry (for example, Moncrieff 2008;Busfield 2013), are further examples of narratives that provide valuable alternative perspectives that challenge dominant narratives and that need to become part of our collective knowledge.
Asserting the humanity of people with mental illness in a public setting is another way for individuals to perform epistemic resistance. When Pamela Spiro Wagner, who has schizophrenia, wrote a letter to the editor of a Hartford newspaper declaring that those with mental illness must be treated as fully human beings, she received accolades and support for articulating this view (Wagner and Spiro 2005, 242-43). When people are reminded of the humanity of those with mental illness, they can work consciously on reducing their implicit bias.
Another form of epistemic resistance involves reclaiming pejorative language by reconceptualizing mental illness experience as Mad identity rather than as inherently pathological and by taking a public stance of pride. Proponents of Mad identity reject the psychocentric view of mental illness that views it as necessarily disabling and focus on the way mental illness experience is integrated, sometimes in a positive way, into one's identity (Schrader, Jones, and Shattell 2013;Rashed 2019a;Rashed 2020). Targets of sanist microaggressions can reconceptualize how they view their own mental illness experience and reappropriate derogatory language such as "mad" or "crazy" to be prideful terms expressing identity. Many authors of memoirs on mental illness intentionally use language in this way (for example, Jamison 1995; Steele and Berman 2001;and Saks 2007). People who are not themselves targets of sanist microaggressions can contribute to epistemic resistance by learning more about mental illness experience, learning more about Mad identity, being an ally of people with mental illness, supporting the reclamation of derogatory language, and supporting the myriad expressions of Mad Pride.
Friedlaender is right to identify some of the responsibilities that all individuals have with respect to microaggressions as including self-education, critical self-reflection, and reducing implicit bias (Friedlaender 2018, 18). When individuals become more aware of their implicit bias through education and interpersonal contact with people with mental illness, they are better able to work intentionally on reducing their bias, as I see occurring in my Philosophy of Mental Health class every year. As students learn more about the experiences of people with mental illness and interact with people who have disclosed mental illness (including myself), they give up their negative stereotypes about mental illness and have more respect and appreciation for the struggles of people with mental illness.
All individuals share in the collective responsibility to resist sanist attitudes and behavior by engaging in acts of epistemic resistance whenever they can. Bystanders, however, have a special but supererogatory duty when witnessing microaggressions to help perpetrators of microaggressions recognize their behavior for what it is and feel motivated to change it, because perpetrators sometimes need their biases pointed out to them by someone else. Identifying microaggressions and pointing out bias is a special act of epistemic resistance that can only be performed by those witnessing microaggressions. Targets of microaggressions have a special responsibility to advocate for themselves in identifying microaggressions to perpetrators, but because of the risk of harm involved, they may not always be able to carry out this duty; they must assess the risk of harm for themselves and make a judgment about when it is appropriate to carry out this duty. Identifying microaggressions is more effective in a social environment where more people are aware of sanist attitudes and work intentionally to resist them through self-education about mental illness experience and critical self-reflection about one's own thoughts and behavior. Pointing out bias is thus an important responsibility that some people have when they find themselves witnessing biased behavior, but it is most effective and meaningful when it is accompanied by other acts of epistemic resistance that all people are capable of engaging in.

Confronting My Colleague
People who have mental illness are frequently subjected to microaggressions that draw upon negative stereotypes of mental illness to make mental illness represent that which is discreditable. Because sanist attitudes are so commonplace in American society, perpetrators of microaggressions rarely recognize when they commit microaggressions; various cognitive biases, moreover, make it difficult for perpetrators to acknowledge problematic behavior and ways in which they are biased. Therefore, it often falls to bystanders, and in particular targets of microaggressions, to bring attention to the biased behavior. Yet bystanders, and targets especially, risk experiencing a range of social harms in pointing out someone else's biased behavior. As marginalized individuals, targets of microaggressions not only risk being subjected to social and epistemic harms but also risk contributing to their own experience of epistemic harm when they self-silence or expend epistemic labor to help privileged people recognize their bias. For any bystander, but especially for targets of microaggressions, the decision about whether to respond to microaggressions should be based at least partly on a determination of risk of harm based on relevant social power, weighed against the need of privileged people to have their bias pointed out to them by someone else. Pointing out sanist bias should be seen as an act of epistemic resistance, one of many chained, echoable actions that collectively aim at countering sanism, and must be performed in the context of many people carrying out their collective responsibility to resist sanism with a variety of acts of epistemic resistance.
In my case, I deliberated for a long time before deciding to confront my colleague about her sanist behavior. I realized that because we were socially situated in many similar ways, the risk of harm was low; the greatest risk was discomfort. I also realized that because we are part of a campus culture that includes bias training and encourages antibiased behavior, my action would occur in a context where she would recognize sanist behavior for what it was if it was pointed out to her. Eventually, I had a conversation with my colleague and told her how her language struck me. Immediately upon my pointing it out to her, she saw how her language was biased against people with mental illness, and she told me that others had pointed out to her that she had committed microaggressions like this previously. My action was thus one of several acts of epistemic resistance that collectively affected her self-awareness. My colleague admitted that her language use was problematic, and she apologized. Had she not apologized or not committed to changing her behavior, I would have let it go as I calculated that it was not worth my expending more epistemic labor on this than I already had. With her apology, however, we were able to move on from this and take the first steps toward restorative epistemic justice (see Almassi 2018). We remain friends, and she is committed to being more consciously aware of the language she uses and more committed to self-education about areas where she is biased. Her commitment to change was just the result I hoped for. Association Meeting and the 2019 Feminist Ethics and Social Theory Association Conference. I thank audience members in these venues for their helpful commentary, as well as two anonymous reviewers whose insightful feedback was critical in improving this article.

Notes
1 For the sake of gender neutrality, in this article I use the pronouns "they, their, theirs" in referring to singular persons. I find these pronouns easier to read and less troublesome than other efforts at gender neutrality. Using gender neutral pronouns is important to avoid invoking gendered stereotypes about mental illness.
2 Sanism affects all people who have mental health diagnoses or who receive treatment for mental health conditions, but I use the term people with mental illness here to focus on people who have severe mental illnesses such as schizophrenia, bipolar disorder, or major depression that sometimes impede their mental functioning sufficiently to make them "ill." Our attitudes toward, stereotypes of, and credibility judgments of people who have different mental disorders vary considerably; in this article, I am focusing my discussion on bias against people who have severe mental illness.
3 Black people with mental health diagnoses, for example, commonly experience anti-Black sanism, where they are subject to prejudicial attitudes on account of both their mental health identity and their Black identity (Meerai, Abdillahi, and Poole 2016). They often feel "in the middle," not accepted by members of their Black community on account of their mental health problems and also not accepted by members of their mental health community on account of their racial identity. As a consequence, they often experience racism within the mental health community and sanism within the Black community. Moreover, because Black people may ( justifiably) respond to cultural oppression with behavior that can be mistaken for mental illness symptoms (such as paranoia and hostility), they are often subject both to overdiagnosis of certain disorders (like schizophrenia) (Kposowa, Tsunokai, and Butler 2001;Chien and Bell 2008;McGuire and Miranda 2008) and to having real mental illness symptoms not taken seriously (Keating and Robinson 2004;Whaley 2004;Hollar et al. 2007). 4 As the biological model of mental disorder has become more widely accepted, stereotypes around severe mental illness have tended to converge around the ideas that people with severe mental illness are incompetent, permanently defective, and dangerous, because it is perceived that people do not have control over their condition, a condition that is regarded as terrible and fearsome (Corrigan and Watson 2004;Thachuk 2011). 5 Note that people can be granted less credibility when they adhere to negative stereotypes, but they can also be granted less credibility when they counter negative stereotypes, almost as if they are being "punished" for defying expectations (Munroe 2016). Mentally ill people who do not fit our stereotypes of mental illness can be seen as having diminished first-personal credibility with respect to their illness because they are "not sick enough." 6 For example, according to some surveys, half of employers are hesitant to hire someone with a past history of mental illness or someone who is currently receiving treatment for depression, and 70% of employers express reluctance to hire someone with a history of substance abuse or who takes antipsychotic medicine. Also, nearly a quarter of employers surveyed would dismiss someone who had not disclosed a mental illness but who was later found to have one (Stuart 2006, 523). In addition, according to one survey, 62% of employers would consider hiring someone with a physical impairment, but only 37% stated they would consider hiring someone with a known mental health problem (survey cited in Repper and Perkins 2003, 35-36). 7 Women with mental illness in general are more likely to be dismissed as having psychological weakness and to not have their concerns taken seriously. White women in particular are typically associated with certain mental illnesses that are feminized in a way that suggests psychological weakness; these include "hysteria" in the late nineteenth century (Shorter 1992;Showalter 1997), depression and anxiety (Herzberg 2009;Dusenbery 2018), borderline personality disorder (Nehls 1998), and chronic fatigue syndrome (Blease, Carel, and Geraghty 2017).