Carving at the Joints: Distinguishing Epistemic Wrongs from Epistemic Harms in Epistemic Injustice Contexts

Abstract

This paper examines the relatively underexplored relationship between epistemic wrongs and epistemic harms in the context of epistemic injustice. Does the presence of one always imply the presence of the other? Or, is it possible to have one without the other? Here we aim to establish a prima facie case that epistemic wrongs do not always produce epistemic harms. We argue that the epistemic wrongness of an action should never be evaluated solely based on the action's consequences, viz. the epistemic and practical harms suffered by the wronged party. Instead – as we shall show – epistemic harms necessarily follow from epistemic wrongs. To conclude, we suggest ways in which extant accounts of epistemic wrongs and epistemic harms as they cash out in epistemic injustice contexts might be refined in light of our argument.

1. Introduction

In Epistemic Injustice: Power and the Ethics of Knowing, Miranda Fricker defends a distinctively epistemic kind of injustice in which an agent is “wronged in her capacity as a knower” (2007: 44). Fricker labels this “epistemic injustice.” Typically, the focus is on the ethical dimensions (including wronging or harming) that underpin two basic epistemic practices: (1) conveying knowledge to others, and (2) making sense of our social experiences. Subsequent scholarship builds on Fricker's work by ameliorating the concept of epistemic injustice and/or foregrounding some of its myriad analogs and iterations (especially those cashed out in real-world conditions) (see Coady 2010, 2017; Dotson 2011, 2014; Dunne and Kotsonis 2023; Lackey 2020a, 2023; Luzzi 2021; McGlynn 2019; Medina 2013, 2017, 2023).

Somewhat curiously, however, there remains a lacuna regarding the causal relationship between epistemic wrongs and (primary)¹ epistemic harms, despite growing interest in epistemic injustice scholarship. In this paper, we address this lacuna. Our goal is to forge a better understanding of the conceptual distinction between epistemic wrongs and epistemic harms in epistemic injustice settings so that scholars might avoid overhasty conflations or misattributions. To orientate the discussion better, we narrow our focus to the following fundamental questions:

1. 1. What is the difference between epistemic wrongs and epistemic harms in epistemic injustice settings?

2. 2. Do epistemic wrongs always produce epistemic harms in epistemic injustice contexts?

3. 3. Do epistemic harms, in epistemic injustice settings, necessarily follow from epistemic wrongs?

To our knowledge, questions 2 and 3 have not been thoroughly explored. One might argue that this is because answers to these questions are intuitively obvious. However, as should become apparent, this is far from true. Indeed, upon close inspection, epistemic wrongs and epistemic harms have a far more complex relationship than many seem to realize.

We begin our discussion by defining (1) non-epistemic wrongs and non-epistemic harms (at least in the way that Fricker conceives of them) and (2) the relationship between these two concepts.² A better understanding of both can aid us in considering ways in which epistemic wrongs and epistemic harms might relate to one another. We then define epistemic wrongs and epistemic harms. Roughly, following Fricker,

An epistemic wrong is an injustice suffered by an agent in their capacity as a knower. In other words, an epistemic wrong is a form of externally-imposed compromised epistemic agency. (2007: 44)³

An epistemic harm makes an agent qua knower worse off. When someone suffers a testimonial injustice, according to Fricker, “they are degraded qua knower, and symbolically degraded qua human” (2007: 44).

Although it may initially appear that the epistemic wrongs of epistemic injustices always produce epistemic harms, we will provide examples of epistemic injustice where this is not necessarily the case. We conclude that it is possible to be epistemically wronged without suffering an epistemic harm in epistemic injustice contexts. Based on this hypothesis, we also examine if the converse might also be true. Is it possible to suffer an epistemic harm due to an epistemic injustice without being epistemically wronged? We discuss cases in which it might seem that this could happen, but conclude that the epistemic harms of epistemic injustices necessarily follow from epistemic wrongs. To be epistemically harmed by an epistemic injustice requires that one is first epistemically wronged, though, as we maintain, one could be epistemically wronged by an epistemic injustice without being epistemically harmed.⁴ Lastly, we outline how these issues require further conceptual refinements and amendments, more than one finds in extant analyses of related phenomena.

2. Distinguishing wrongs from harms

Suffering harm is, unfortunately, part and parcel of everyday life. Crudely put, to suffer harm is to experience some kind of damage that makes one worse off (Unruh 2023).⁵ The amount of harm one can suffer ranges from minimal (losing a lock of hair) to maximal (losing one's life). One can also suffer different kinds of harms. These include physical harms (a black eye or a broken nose), mental health harms (PTSD or workplace depression), economic harms (pension depreciation or losing money), social harms (losing social capital or reputational loss), and, of course, epistemic harms (being harmed in one's capacity as a knower or being harmed in one's capacity as an epistemic agent).

One can also suffer harm as a result of others' actions (e.g. malicious gossip) or others' unwillingness to act (e.g. failing to call the police when witnessing a crime).⁶ This includes harms suffered via the actions of human institutions (e.g. the court or the government). For example, someone is harmed if the government decides to seize their property because of a war. However, the party inflicting the harm need not be different from the party suffering the harm (e.g. self-mutilation or suicide). There are also harms that are not a consequence of human actions (or inactions). Persons can suffer harm (e.g. damage to one's property) from a natural phenomenon (a tornado or an earthquake) or from non-human animals (a dog chewing at one's front door). We can classify harm in other ways as well. One can suffer permanent harm (having one's reputation irreparably damaged) or non-permanent harm (a bleeding nose or catching a cold). Harm can be systemic (experiencing ongoing prejudice) or once-off (a burnt-out vandalized car). Harm can be anticipated (stepping into a boxing ring) or unexpected (being assaulted while walking along a busy sidewalk).

Similarly, the infliction of harm can be distinguished in terms of the different kinds of harm inflicted (physical, emotional, economic, social, epistemic, etc.), the severity of the harm inflicted (ranging from minimal to life-threatening), the party inflicting the harm (i.e. self-harm, other-inflicting harm, harm inflicted by humans, harm inflicted by non-humans), the duration of the harm (ranging from momentary to permanent), whether the harm is systemic or singular, whether the harm takes time to manifest or not (water consumption from lead pipes), and whether the harm was anticipated by the victim or not.

Moreover, harm can be inflicted intentionally (punching someone purposefully) or unintentionally (accidentally bumping into someone's car). Harm can be inflicted knowingly (malicious gossip) or unknowingly (well-meaning but poor financial advice). Harm can be inflicted out of choice (hitting an unaware stranger) or out of perceived necessity (harming others when defending oneself from their attack).

What the above examples show is that the concept of harm is quite broad. It can be used to capture a multitude of different outcomes stemming from a range of diverse behaviors. That said, the common denominator allowing the above examples to be classified as instances of harm is that an agent suffers some kind of damage. It does not make sense to talk about harm if there is no damage suffered (or caused). An agent must be demonstrably worse off because of some state of affairs, event(s), or behavior(s).

Another concept that is used to describe a variety of different phenomena is that of a wrong. We take it that to wrong someone is to violate an all-things-considered duty owed to them (see Darwall 2010; Scanlon 2000, 2007; Thompson 2004). This can be called the breach of duty clause. Breach of duty is a failure to perform a duty (or a responsibility) that is required. It is a failure to behave with the level of respect and care that someone of ordinary prudence would have exercised under the same circumstances. For example, I am wronged if someone trespasses on my property (enters my property unlawfully, without good reason, or without my explicit consent). This is the case even if I am unaware of the trespassing and nothing has been stolen or damaged. The trespasser violates a duty owed to me: that I have a moral and legal right to my private property (Enoch 2002). Like harm, wrong can be committed intentionally or unintentionally, knowingly or unknowingly, out of choice or out of perceived necessity, systematically or singularly, and it can be anticipated or unexpected (see Bowen 2022). Wrong can also come in degrees of severity and manifests differently in myriad contexts.

Wrongs are often classified into two types: legal wrongs and moral wrongs (Thompson 2004). For an act to be legally wrong, it must violate some legal right. For instance, kidnapping someone is a legal wrong because it violates the legal right they have to be free from constraint. Kidnapping can also be categorized as a moral wrong. According to Stephen Darwall, the concept of moral right is:

conceptually related to that of a moral duty or obligation to the rightholder. Your right against me entails that I have a moral duty or obligation to you. An obligation to someone is a “directed” or “bipolar” obligation. If I violate such a duty to you, I wrong you. (2010: 152)

Thus, when we kidnap someone, we wrong them both legally and morally because we violate both a legal and a moral duty owed to them. We violate both their legal and moral right to be free from constraint. We also violate the authority they have as a person to demand that we do not act in a way that violates their rights (Darwall 2010; Scanlon 2000, 2007; Thompson 2004). That said, legal wrongs and moral wrongs are not the same. A legal wrong might be morally right, and a moral wrong might be legally acceptable. Lying to someone, for instance, is morally frowned upon (when the act of lying fails to meet certain moral criteria). But, in most cases, lying is legally permitted (assuming that one is not under oath in a court of law).

Most importantly for our purposes of carving epistemic wrongs and harms at the joints, wrongs and harms are distinct concepts. They should neither be conflated nor used interchangeably (Feinberg 1984; Kumar 2003). One can harm another person without wronging them. Imagine a restaurant that opens close to a competitor. The new restaurant's food is superior, the prices more reasonable, and, soon enough, the competitor closes down. Here, the owner of the competitor restaurant is not wronged, but he is harmed. To give another example, consider the following. A soldier defending her country against an illegitimate invasion injures an enemy soldier.⁷ This can be seen as an act where the soldier inflicts a harm (physical damage) but does not wrong her enemy. We do not commit an injustice to someone when we harm them to protect ourselves and our country.⁸

Likewise, we now argue, one can wrong another person without harming them. Trespassing is a good example. Take trespassing that does not result in any material or personal damage. The trespasser violates a duty owed to me, that I have a moral and legal right to my private property. The trespasser has wronged me, but they have not harmed me. As Joel Feinberg notes, the trespasser has “wronged the owner of the property, but they have not caused any harm to them” (1984: 35). We can say that the trespasser has violated a duty of care to the ownership rights of the rightsbearer. However, the former has not harmed the latter in any way.

3. Distinguishing epistemic wrongs from epistemic harms

What about epistemic harms and epistemic wrongs in the context of epistemic injustice? Is there any difference between the two? To answer this, we must first define the two concepts. Fricker, who was one of the first to argue that there is a distinctively epistemic kind of injustice, identifies two kinds of epistemic injustice: (1) testimonial injustice and (2) hermeneutical injustice.

1. 1. Testimonial injustice occurs when a prejudice causes “a hearer to give a deflated level of credibility to a speaker's word” (Fricker 2007: 1). One of Fricker's examples of testimonial injustice is that of a prejudiced white male police officer who does not believe the testimony of a witness because she is black.⁹

2. 2. Hermeneutical injustice occurs “when a gap in collective interpretive resources puts someone at an unfair disadvantage when it comes to making sense of their social experiences” (Fricker 2007: 1). One of the examples that Fricker uses to explain this kind of epistemic injustice is that of Carmita Wood: a victim of sexual harassment at Cornell University in the 1970s. Wood was unable to understand and communicate her experience owing to the hermeneutical marginalization of women in her society.

Fricker characterizes each kind of epistemic injustice as involving a specific kind of epistemic wrong. Testimonial injustice occurs when, “someone is wronged in their capacity as a giver of knowledge”; hermeneutical injustice occurs when “someone is wronged in their capacity as a subject of social understanding” (Fricker 2007: 7). Nonetheless, a common feature of both kinds of epistemic injustice – one that allows them to both count as instances of epistemic injustice – is that they are ultimately cases in which an agent is “wronged in her capacity as a knower” (Fricker 2007: 44).¹⁰

Fricker argues that epistemic injustices bring about primary (intrinsic) harms and secondary (extrinsic) harms. The primary harm of epistemic injustice is being “degraded qua knower” and hence being “symbolically degraded qua human” (Fricker 2007: 44).¹¹ Fricker links this to epistemic objectification. She argues that those who suffer epistemic injustices are “demoted from subject to object, relegated from the role of active epistemic agent, and confined to the role of passive state of affairs from which knowledge might be gleaned” (Fricker 2007: 132).¹² For example, some kind of prejudice might lead an audience to treat a speaker as a mere object through which they can glean information (much like when one infers the age of a tree from its growth rings). Being wronged in one's capacity as a knower thus involves a certain dehumanization.

For Fricker (2007), the secondary harms of epistemic injustice are extrinsic and can be distinguished into (1) practical secondary harms and (2) epistemic secondary harms:

1. 1. Practical secondary harms of epistemic injustice might involve suffering practical costs because of the injustice in question. An example of practical secondary harms in cases of testimonial injustice is a person being wrongly convicted because of a prejudiced jury. Practical secondary harms related to hermeneutical injustice include the case of Carmita Wood, who was denied compensation from Cornell owing, in part, to her inability to make sense of and state her reason for leaving the institution.

2. 2. Epistemic secondary harms can be derived from both testimonial injustices and hermeneutical injustices. These include cases in which a person mistakenly or wrongly loses confidence in their beliefs, in the evidential weight of their lived experience, or in their epistemic competencies owing to injustices suffered (Dunne 2020, 2023).

As mentioned, Fricker (2007) maintains that suffering an epistemic injustice necessarily entails been wronged qua knower. On her view, one cannot suffer an epistemic injustice and not be wronged as a knower. We do not wish to question this or try to come up with counterexamples (assuming there are any to be found). After all, Fricker has explained that she is looking for “a distinctively epistemic kind of injustice” (2007: 1), and has found it in cases where agents are wronged qua knowers. Another conclusion we can draw is that, for Fricker, there is a distinctive kind of testimonial injustice-induced primary harm occurring every time an agent is wronged qua knower (viz. degraded qua knower through a process of objectification that dehumanizes the agent). This is the point that we focus on here.

Fricker's account seems to imply that epistemic wrong always produces primary epistemic harm. Contra Fricker, we argue, she does not consider cases (assuming she believes that such cases exist) where one suffers an epistemic injustice and is not degraded qua knower. Hence, the question we seek to answer is the following: can one suffer the distinctive kind of primary harm that Fricker is describing without being wronged? That is, can someone be wronged as a knower without being degraded qua knower? If yes, then it seems that the epistemic wrong and epistemic harm are not as closely linked as they often seem to be. Describing such cases will then help us better understand the relationship between the two concepts. This should not be seen as an attempt to target Fricker's views. For all we know, she is simply interested in cases where epistemic wrongs bring about primary epistemic harm. Nonetheless, scholars working on epistemic injustice fail to discuss instances where epistemic wrong does not yield epistemic harm. Our aim is to show that such cases exist, and that they can aid in better understanding epistemic injustice.

Before proceeding to discuss cases of epistemic injustice that involve suffering an epistemic wrong without suffering an epistemic harm, it is important to note that not everyone agrees with Fricker's claim that objectification is the primary harm of testimonial injustice. Gaile Pohlhaus (2014), for instance, argues that the primary harm of testimonial injustice is derivation: relegating someone “to the role of epistemic other, being treated as though the range of one's subject capacities is merely derivative” (107). For Pohlhaus, the primary epistemic harm in cases of testimonial injustice is not that one is seen as an object. Rather, it is that one is seen as an unreliable subject, as one who is not “capable of contributing to epistemic practices uniquely, that is from [one's] own distinct lived experiences” (107). Though the conceptual divergences in accounts of primary harms are noteworthy, nothing salient hinges on such–viz–mapping connections between epistemic harms and epistemic wrongs in epistemic injustice contexts.

4. Epistemic wrongs without epistemic harms

Whether one agrees with (a) Fricker's view that the primary harm of epistemic injustice involves objectifying an agent, (b) Pohlhaus' view that epistemic othering is the primary harm of testimonial injustice, or (c) some other view about primary harms, the fact remains that instances of epistemic injustice are cashed out in terms of their adverse impact on an agent's epistemic agency.¹³ But, what if there is no adverse impact on an agent's epistemic agency? Can there be a wrong without a harm in such contexts?

Consider the following case: someone who belongs to a marginalized social group is in a car accident and then rushed to hospital. She is seen by one of the hospital's consultants who is a white, heterosexual, middle-aged male. She is told that she will need spinal fusion because the nature of her injuries has exacerbated her pre-existing scoliosis. Now, imagine that the patient believes that white, heterosexual, middle-aged males should never be trusted. Not only that, but, as someone who has lived with severe scoliosis, she knows that fusion might make her posture more upright, but the consequent level of pain is simply not worth enduring. Because of this, she assigns the doctor's expert opinion a low degree of credibility and elects not to have the surgery. Is the doctor epistemically wronged by the patient?

On Fricker's conception of epistemic injustice, the answer is yes. He is wronged in his capacity as a knower; he is wronged as a giver of knowledge. But is he epistemically harmed? The answer seems to be no. The doctor is not excluded “from participating fully in the most basic of social epistemic practices based on an unwarranted, but widely held, identity prejudice” (Pohlhaus 2014: 101).¹⁴ There is no widely held identity prejudice against white, heterosexual, middle-aged male doctors. The doctor is not a marginalized speaker who “is likely to be discredited and/or silenced when the information she contributes moves beyond the scope of the world experienced from dominant subject positions” (Pohlhaus 2014: 110). It is the doctor who is in the dominant social position, while the patient is a member of a marginalized social group. The patient is unable to influence the doctor's participation in social epistemic practices (in fact, it may actually be the patient who is generally excluded from such practices). As such, we can conclude that the doctor has not experienced any adverse impact on his epistemic agency as a knower or giver of knowledge via the wrong he suffered from the patient.

To give another example, consider the following case. George is a Canadian trained, board certified, and highly experienced counsellor. He is working with Helen who has been suffering from anxiety attacks related to a recent perimenopause diagnosis. To help her cope with anxiety, George advises Helen to use breathing techniques and offers to teach these techniques to her. However, Helen is prejudiced against male counsellors educated outside Europe (a widely held prejudice in her social circles¹⁵), and believes that they should rarely be trusted. Because of this, she assigns a low level of credibility to George's expert opinion, and decides not to learn how to use the breathing techniques. This is another case in which an agent is epistemically wronged without being epistemically harmed. George has been wronged in his capacity as a knower, as a giver of knowledge. Still, he has not been epistemically harmed because of this wrong. He has not suffered damage that has made him epistemically worse off. The wrong he suffered did not have a negative impact on his epistemic agency. It might even be that George benefits epistemically from the wrong he suffered. It might lead him to realize that certain people “resist” change, and that he can benefit from more training in this area (arguably an epistemic benefit of sorts).

One could object that the primary epistemic harm in the doctor/patient example discussed above consists in (a) the objectification of the doctor (if one agrees with Fricker) or (b) treating the doctor as an unreliable subject (if one agrees with Pohlhaus). Yet, both of these options seem to involve a wrong inflicted on the agent and not the primary harm they suffered. An intrinsic epistemic harm occurs when there is a negative impact on an agent's epistemic agency: when they experience some kind of damage that makes them all-things-considered worse off.

We can also adjust the example of the doctor so that there is a positive impact on the doctor's epistemic agency (he is better off) owing to the testimonial injustice he suffered. Suppose that the director of the hospital hears of the incident with the prejudiced patient and is impressed that the doctor did everything in his power to convince the patient that she needed surgery. Because of this, the director decides to promote the doctor to chief surgeon in the hospital's orthopedic department. Acquiring this prestigious title means that the doctor is taken seriously by patients and colleagues. People listen to what he has to say more than ever before. It turns out, after all, that the injustice he suffered has benefited him. Had the patient not treated him as she did, the doctor would not have had the chance to try to convince her to change her mind (and thereby incidentally impress the director).

Some might object that this example is unrealistic. They might insist that testimonial injustice brings about only epistemic harms and never epistemic benefits. We accept that our example is somewhat contrived, and that, in most cases, suffering epistemic injustice does not have a positive impact on epistemic agency. Still, no matter how far-fetched, this example demonstrates that there is a possibility (regardless of how small or unlikely it is) that one can epistemically benefit from epistemic wrongs (there is likewise a small and unlikely possibility that one can benefit from non-epistemic wrongs).

Some might likewise object that, while Fricker and Pohlhaus focus on cases where members of a social group are systematically treated unjustly by the dominant group,¹⁶ we are discussing a case in which a member of the dominant group suffers a non-systematic epistemic injustice. The case we are discussing might be considered irrelevant to understanding the process by which groups and group members are marginalized and excluded from participating in social epistemic practices. This may generally be so, but the objection does not undo our argument. It remains the case that epistemic wrongs do not always bring about primary epistemic harms. And the fact that we used an example of a non-systematic injustice to support this does not undermine our argument.

Some may question whether our argument is useful for understanding epistemic injustice. What is the point of highlighting limited cases where epistemic wrongs do not bring about epistemic harms? What exactly are we hoping to contribute to the discussion? Our answer is twofold.

First, forging a distinction between epistemic wrongs and epistemic harms contributes to better understanding the nature (context-specific salient features, myriad instantiations, etc.) of epistemic injustice. The concepts of epistemic wrong and epistemic harm are so pertinent to examinations of epistemic injustice that untangling the two is of significant value in and of itself. Delineating the two is crucial to deciphering the justification underpinning designations of wrongs or harms, and also in which cases to appropriately attribute blame. Arguably there is a real risk, on the current conception, that epistemic harms abound. We see the proliferation of epistemic harms to be indicative of the need to engage in ameliorative conceptual revision with regard to epistemic wrongs and harms, including forging deeper understandings around the precise relationship between the two.

Second, our argument suggests that to suffer an epistemic injustice does not always imply that one has suffered an epistemic harm. When evaluating the epistemic wrongness of an action, we should not only take into account the harms suffered by the agent and vice versa. The two are distinct concepts; harms do not necessarily follow from wrongs. If we ignore this, then we run the risk of evaluating the wrongness of an action solely on its consequences (see Congdon 2017). This can lead to the misclassification of certain wrong actions as neutral or even right actions.

5. Epistemic harms without epistemic wrongs

So far, we have argued that epistemic wrongs arising from epistemic injustices do not always produce epistemic harms, that one can be wronged without suffering epistemic harm. We now examine if the opposite is true: do epistemic harms always follow from epistemic wrongs? Can one suffer an epistemic harm without being wronged? In this section, we attempt to answer this question. We shall do so by discussing cases of credibility excess testimonial injustice. This kind of epistemic injustice involves cases where an agent suffers epistemic harm without been epistemically wronged.

Contra Fricker, Emmalon Davis (2016) argues that testimonial injustice is not always a matter of credibility deficit (as in the cases discussed in the previous section). It can also be due to credibility excess. Davis discusses what she calls identity-prejudicial credibility excess epistemic injustice. Here, one is harmed as a result of receiving a credibility excess from one's audience. According to Davis,

an identity-prejudicial credibility excess occurs when a speaker is assessed to be credible with respect to some bit of knowledge on the basis of prejudicial stereotypes associated with the speaker's social identity. (2016: 487)

Davis uses the following example. Treating a person from China as an expert on computers because of the prejudice that Chinese people are technologically savvy is a case of identity-prejudicial credibility excess. For Davis, this is an epistemic injustice because the audience does not perceive the agent as a person with unique experiences and viewpoints. Instead, the agent is perceived as if they are “fungible or interchangeable with others who share the same identity” (Davis 2016: 488). Davis characterizes the primary harm of epistemic injustice as “a form of epistemic othering, through which the capacities of the speaker are prejudicially assessed in such a way that bypasses or circumscribes the speaker's subjectivity” (2016: 490).

The example of the Chinese agent is a case in which a person is both epistemically wronged and epistemically harmed by their audience. They are wronged qua knower and this wrong also harms them qua knower. The wrong suffered by the Chinese agent is that a community treats them as interchangeable with other Chinese people (owing to the technologically savvy stereotype). The primary (or intrinsic) epistemic harm inflicted on the Chinese agent is that she is excluded from fully participating in social epistemic practices as an inimitable individual. Hence, this is a case where the primary epistemic harm suffered by the agent stems from the epistemic wrong inflicted to them. Now, the pertinent question is as follows. Can we tinker with cases of credibility excess testimonial injustice (or any other kind of epistemic injustice) to come up with a case where an agent is harmed qua knower without being wronged qua knower?

Consider the following example: Meera is proud of her nationality. Owing to this, she readily assumes the role of spokesperson for her social group (the role is not imposed upon her in any way). Having assumed this role, Meera's audience is either unable or unwilling to view her as an individual with unique experiences and contributions to make. They see her as interchangeable with others with whom she shares the same nationality. Following Davis (2016), we can conclude that Meera has suffered an epistemic harm qua knower. She is obstructed from participating fully in the most basic of social epistemic practices. But has she suffered an epistemic wrong?

Perhaps a case can be made that Meera has not been wronged, at least not by her audience, because she was the one who assumed the role of the spokesperson. She did not consent to be treated as an individual without unique experiences and contributions to make. Yet, there is still a prima facie argument that this role was not imposed upon her by the audience. This is an important way in which our example differs from Davis' Chinese agent case. Meera assumes the role of spokesperson voluntarily. In Davis' case, the Chinese agent is involuntarily assigned the role of spokesperson for her nationality. Nonetheless, despite this difference, there still appears to be an epistemic wrong in Meera's case. Meera has been treated as interchangeable with other members of her social group, based on an identity prejudice, and not as an individual with unique contributions to make. This is the case irrespective of whether one understands the epistemic wrong as being inflicted by the agent herself (when she readily assumed the role of spokesperson) or by the audience (who were unable or unwilling to see her as more than just a spokesperson for her group).

It seems that we have failed to come up with a case of someone being harmed qua knower without being wronged qua knower.

Some might object that we have not examined enough cases of epistemic injustice to justify this argument. Perhaps there are cases of epistemic injustice that we have not considered where an agent is harmed qua knower without been wronged. We do not think that such a case exists. This is because Fricker (2007) frames the concept of epistemic injustice in a manner that requires primary epistemic wrongs to follow from epistemic harms. First, she has excluded cases of distributive unfairness regarding epistemic goods from her conception of epistemic injustice. This is because they are not wrongs specifically done to someone in their capacity as a knower. In so doing, Fricker has removed many cases where one might be epistemically harmed (though not in one's capacity as a knower) without been wronged (one is harmed by not receiving epistemic goods, but not wronged because one was not entitled to those goods).¹⁷ This is not to imply that Fricker is unjustified in excluding such cases. She is interested in epistemic injustice and thus focuses on cases that are distinctively epistemic.

Most importantly, Fricker (2007) generally conceptualizes primary epistemic harms in a way that ties them to epistemic wrongs: to be objectified, more often than not, requires that one is wronged. There are, though, obvious counterexamples; indeed, even Fricker allows for certain exceptions (see McGlynn 2019, 2020).¹⁸ More often than not, this view of objectification holds true even if, for some non-epistemic reason, we believe that we are justified in treating others in an epistemically unjust way.

If we were to epistemically objectify the world's most notorious villain (e.g. Hitler), we would still be wronging him epistemically (qua knower). The same goes for possible cases of epistemic self-defense where an agent treats others in an epistemically unjust manner to protect themselves. Assuming that such cases exist, they will still be cases in which we are treating others wrongly in their capacity as knowers. The argument for this causal link between epistemic harms and epistemic wrongs is not unique to Fricker. It can also be found in the work of most scholars working on epistemic injustice. It does not matter if, like Pohlhaus, one disagrees with Fricker regarding the primary harm of testimonial injustice. What matters is that Pohlhaus follows Fricker in conceiving of epistemic injustice as distinctively epistemic, and, in so doing, ties primary epistemic harms to epistemic wrongs. This is the very assumption we take issue with and argue against, since, to our mind, epistemic wrongs do not always produce epistemic harms.

One might wonder what is problematic with conceiving of intrinsic epistemic harms as necessarily following from epistemic wrongs in contexts of epistemic injustice analyses. Our aim in this paper is to foreground the importance of thinking carefully about the assumed relationship between epistemic harms and epistemic wrongs to pave the way for better understandings of the nature of epistemic injustice. Moreover, if we wish to understand primary epistemic harms as not necessarily following from epistemic wrongs (as is the case of non-epistemic harms and non-epistemic wrongs), then we will have to deviate from Fricker's original understanding of epistemic injustice. In the interests of accuracy and clarity, scholars who agree will need to make this explicit in their writings. Still, we do not know what this new understanding of epistemic harms and wrongs under the umbrella of epistemic injustice should look like. Trying to develop such an understanding lies outside the scope of this paper. For our purposes, it suffices to note that, following Fricker's understanding of epistemic injustice, one cannot conceive of a primary epistemic harm that has not been produced by an epistemic wrong.

6. Conclusion

To sum up, our aim in this paper has been to critically re-examine the assumed connection between epistemic wrongs and epistemic harms in the context of epistemic injustice. We argued that epistemic wrongs produced or arising from epistemic injustices do not necessarily bring about epistemic harms. We likewise noted that this realization highlights the fact that the epistemic wrongness of an action should never be evaluated solely based on the action's consequences (harms inflicted). If we do so, we run the risk of not including under our conception of epistemic injustice cases where an agent is wronged but not harmed. We also argued that epistemic harms necessarily follow from epistemic wrongs when an epistemic injustice has been committed. Contra Fricker, to combat the proliferation of epistemic harms which arise based on her original conception, we suggest refinements must be made to narrow the parameters of epistemic harm. Finally, we contend, that this is a distinction with a difference, one that many important practical issues (ethical or otherwise) depend upon.