Epistemic Virtue Signaling and the Double Bind of Testimonial Injustice

Catharine Saint-Croix; Catharine Saint-Croix

doi:10.3998/phimp.4049

1. Introduction

Virtue signaling is a familiar concept. According to Westra (2021),

Virtue signaling is the act of engaging in public moral discourse in order to enhance or preserve one’s moral reputation. […] What makes the act in question an act of virtue signaling is not the content of the moral expression itself, but rather the status-seeking desires of the person or corporate entity making it. (p. 156)

Examples abound: Corporations decorate their social media profiles with rainbows during Pride month, celebrities decry scandalous events in interviews, and so on. In each case, the expression aims not only to put forth a judgment, but also to influence the audience’s judgment of the actor’s moral character. But, what about profile pictures framed by “Vaccines work!”? Or memes posted to an anti-vaccine group echoing the group’s view that “Only sheep believe Big Pharma!”? These actions don’t express moral views—both claims are empirical (if imprecise).¹ Nevertheless, they serve a similar purpose: to influence their audience’s judgments of the speaker. Whereas rainbow profiles guide their audience to view the agent as morally good, however, these acts guide their audience to view the agent as epistemically or intellectually good. They are instances of epistemic virtue signaling (EVS). I argue that EVS is a common part of everyday conversational dynamics, facilitating the coordination and uptake of testimony.

However, it also poses risks. Because our understanding of such signals is informed by the socio-epistemic norms we absorb from those around us, EVS is entangled with the biases of our broader social context. This helps to explain the connection between testimonial injustice and EVS. In particular, EVS sheds light on a double bind confronting those marginalized by testimonial injustice: to remain silent or to engage in defensive signaling, each choice carrying its own epistemic and moral harms. Understanding this double bind reveals the structure of deeper injustices, including the erosion of medical autonomy experienced by pregnant women of color in the United States today.

The first goal of this paper is to offer an account of EVS.² Section 1 broadens the concept of virtue signaling, arguing that the motivations and structure of moral virtue signaling can be found across different discourses and drawing attention to the socio-epistemic norms that we often signal through speech and action. Section 2 focuses on the concept of epistemic virtue signaling itself. There, I argue that there are two forms of EVS, propositional (2.1) and demonstrative (2.2), and argue that it is a common part of our everyday interactions. Section 2.3 discusses the policing of EVS and the nuance of how epistemic virtue signals are received by different groups. Section 3 introduces the double bind of testimonial injustice. Subjects of this double bind face a choice between risking testimonial injustice or engaging in defensive EVS. On either side of this bind, they suffer both moral and epistemic harm. Furthermore, those who choose defensive EVS risk not only further loss of credibility, but also ingraining the very prejudice they seek to overcome. This section also discusses the impacts of deception on the harms of this double bind (3.1) and situates the double bind with respect to Hirji’s (2021) account of oppressive double binds and Lackey’s (2020) concept of agential testimonial injustice (3.2). Section 4 shows how prejudice and EVS interact in the context of medical autonomy, focusing on Altman et al’s (2019) research on the care gap faced by pregnant women of color, as compared with their white counterparts. Section 5 closes by reflecting on epistemic dependence (Hardwig 1985) and its relationship to EVS.

2. (Epistemic) Virtue Signaling

To understand EVS, it will be helpful to take a closer look at the analogous phenomenon, moral virtue signaling. Simply referred to as ‘virtue signaling’ or ‘moral grandstanding’ in the literature, moral virtue signaling has three important features. First, it is an act. Generally, virtue signaling will occur through speech acts employing moral talk, which Tosi and Warmke (2020, p. 3) define as “communication about moral matters”, though it need not.³ Either way, it must meet the second condition: it is an act that engages with our public moral discourse. So, for example, a social media post about one’s ire regarding a recent Supreme Court decision is a candidate for virtue signaling, but so is wearing a pin that says, “Go Vegan!” or discussing the latest moral panic over coffee with friends. These acts engage with public moral discourse because they concern topics and norms that are shaped by public moral discourse.

Whether an act actually is an instance of virtue signaling depends on the third and final feature: its motivational profile. Instances of virtue signaling are characterized by a motivational profile directed toward improving or preserving one’s moral reputation. This motivational profile need not be explicit or dominant, and it should not suggest insincerity. Consider a Catholic who sincerely holds the view that abortion is sinful under any circumstance. They might announce this view among friends with the primary aim of having a conversation about the ethics of abortion, but nevertheless have been covertly (even to them) motivated by a desire to be viewed as a Good Catholic by their like-minded friends. And, the “Go Vegan!” pin is a candidate for virtue signaling because wearing it constitutes engaging in public moral discourse, but whether it turns out to be virtue signaling depends on whether the wearer’s motivation includes influencing others’ views of their moral character. With this in mind, it is also worth distinguishing between virtue signaling and signals of virtue. Virtue signaling is a kind of act undertaken by an individual with a particular motivational profile, while a signal of virtue is just that: something that signals virtue. For example, working at the soup kitchen might be a signal of one’s virtue—as this is something virtuous people do—without being an instance of virtue signaling—you might do this without any thought toward others’ impressions of you.⁴ As we turn to EVS, we will find that each of these core features has a counterpart.

In order to understand the role that EVS plays in our lives, however, we first turn to the notion of public discourse itself, moral and otherwise.

2.1 Public Discourse and its Purpose

Public moral discourse is not our only form of public discourse. In fact, we have many others: intellectual, scientific, political, artistic, and so on. These discourses are, of course, intertwined and overlapping, but what’s important for our purposes is that they are similar to public moral discourse in many ways. Tosi and Warmke (2016) argue that public moral discourse has an important epistemic role: by exposing us to arguments for particular moral views, the public moral discourse allows us to challenge and improve our own views, thereby improving our moral decision-making individually and as a society (Tosi and Warmke, 2016, p. 206). This is just as true for other discourses: If an agent is deciding whether they should save their paycheck or splurge on a new laptop, they might want to know whether the country is heading into a recession. To find this out, they can turn to the public square—perhaps scrolling through social media, opening The New York Times, or tuning into a podcast—and survey the arguments presented there. They can then weigh these reasons and, now informed by sincere, epistemically responsible testimony (one hopes), proceed to make their economic decisions on the basis of (presumably) better beliefs. While this engagement with public economic discourse lacks the apparent gravitas of moral discourse, it is no less influential and no less important.

But, as Westra (2021) points out, there is more to public moral discourse than this deliberative conception suggests. In addition to potentially changing individual agents’ beliefs about what is true, it also changes their understanding of the operative social norms—the ways others expect them to behave. Westra, following Christina Bicchieri, adopts an expectation-based view of social norms:

A social norm is a rule of behavior such that individuals prefer to conform to it on condition that they believe that (a) most people in their reference network conform to it (empirical expectation), and (b) that most people in their reference network believe they ought to conform to it (normative expectation). (Bicchieri, 2017, p. 35)

Social norms, then, are constituted by two kinds of expectations, both of which involve reference networks. These are the groups of people who matter to our decision making—those whose opinions, behaviors, and attitudes we care about. And, these reference networks change depending on the decision itself and the context in which it is being made.

How are these expectations learned? One way, Westra suggests, is through virtue signaling (2021, Sec. 5). When individuals around us engage in public moral discourse, they not only express their views but, in virtue of having done so publicly, express views about what one ought to believe and do. In doing so, they are “publicly endorsing a normative standard about the kinds of behaviors that are acceptable or unacceptable,” (Westra, 2021, p. 11). As more and more members of one’s reference network do this, it becomes strong evidence for the pair of expectations involved in social norms. That is, it becomes strong evidence that within one’s reference network, people behave in this way and believe that one ought to.

Returning to the discussion of virtue signaling in broader public discourses, we can see that Westra’s norm communication account also applies to public discourse in general, not just public moral discourse. Consider the example of climate change denial within public scientific discourse, such as one’s reference network repeating the claim, “anthropogenic climate change is a hoax,” especially where such claims are presented without argument or evidence (as in, say, a meme or a comment thread). Just as in the case of public moral discourse, this provides evidence for both the empirical and normative expectations that make up a social norm. In this case, the norm would be something like everyone ought to believe that anthropogenic climate change is a hoax. The repetition of the claim among an individual’s reference network provides evidence for the empirical expectation by demonstrating that many members of the network do, in fact, have this belief. It also provides evidence for the normative expectation in two ways. First, where this claim is made without evidence, it suggests that the claim is so well known as not to need supporting evidence. Second, where these expressions are made angrily or with exasperation, as they often are, this emotional component conveys the normative content—members of the reference network not only believe it, but also believe that others ought to believe it. Related claims, for example “only idiots would believe a few Volvos can change the planet!”, convey the relevant normative expectation more directly: irrespective of their reasons, an individual holding the belief in question is subject to (epistemic) social sanctions. And there is nothing particularly remarkable about this example: in general, public discourse provides evidence about the relevant social norms—in this case, norms about what “only idiots” would believe.

We now have two ways that public moral discourse might influence individuals’ beliefs and, by extension, their behaviors:⁵ either through engaging participants’ rational faculties for moral deliberation or through teaching them what their communities expect of them, on pain of social sanction. And, we’ve seen that these functions apply not only in the case of public moral discourse, but public discourse in general. In fact, it is reasonable to think that wherever there are (a) social norms and (b) public discourse that would allow one to demonstrate adherence to those norms, virtue signaling will occur. The epistemic realm is no different. There are socio-epistemic norms—social norms, in Bicchieri’s (2017) sense, that govern epistemic practices—and they can differ across reference networks.⁶ Some socio-epistemic norms are quite widespread. For example, “don’t believe in wishful thinking” is a common socio-epistemic norm. But some are not so widespread. Claims like “I do my own research”, a hallmark creed of conspiratorial thinking (Levy, 2022), belittle deference to experts and thereby suggest a socio-epistemic norm that eschews expertise in favor of independent epistemic practices. An individual may or may not conform to such socio-epistemic norms, and they may signal to others that they do through their engagement in public discourse.⁷

3. Identifying Epistemic Virtue Signaling

We want to be (or at least be seen as) good people. But, we also want to be (or, again, at least be seen as) intelligent, well-read, thoughtful, scientifically literate, and so on. We want, in other words, to be seen as epistemically good in much the same way that we want to be seen as morally good. Whether it arises because being a good epistemic agent is intrinsically valuable, morally valuable, or something else entirely, this desire is a common one. And, in the same way that we can engage skillfully with public moral discourse in order to influence how others view our moral character, so, too, can we harness public discourses to influence how others view our epistemic character. This is the heart of EVS.

Before moving forward, note that while the “virtue” in “epistemic virtue signaling” can refer to the signaling of specific virtues of character—curiosity, intellectual courage, etc.—it need not.⁸ The term “virtue signaling” is not tied to a virtue theoretic account of either ethics or epistemology. In both cases, this term is meant to pick out perceived good-making features of the virtue signaler relative to the normative domain in question. Just as expressing specific beliefs with morally good content or implications vaguely signals some general moral goodness of the agent, so, too, does publicly expressing certain beliefs (such as scientific beliefs) vaguely signal some general epistemic goodness. And, because the virtue signals are tied to socio-epistemic virtues— social norms, in Bicchieri’s (2017) sense, that pick out what signaler’s reference network takes to be epistemically good—rather than any genuine epistemic virtues, the signaled traits may not reflect traits that are in fact epistemically good-making.⁹ As a result, the collection of traits that constitute “epistemic virtues” in this context are quite broad, and might include things like being intelligent, knowledgeable, well-read, an independent thinker, and so on, in addition to features traditionally recognized as epistemic virtues.

With this in mind, we can understand EVS as the act of employing relevant socio-epistemic norms in order to enhance or preserve one’s epistemic reputation, where the relevant socio-epistemic norms are those the signaler takes to be accepted by their reference network. EVS might occur through the direct content being conveyed (“I believe the science!”), through the manner in which the content is conveyed (adopting a posh accent), or both at the same time. It is helpful to consider two categories of EVS: Demonstrative EVS and Propositional EVS, which is an important subset of the former. Demonstrative EVS is EVS that occurs through acting in a way that comports with the behaviors expected of the epistemically virtuous, while propositional EVS uses the propositional content of the signaling act to do the same. We begin by looking at propositional EVS, which most closely resembles discussions of moral virtue signaling and most directly engages with public discourse.

3.1 Propositional EVS

Propositional EVS occurs when the propositional content conveyed by the act in question is the means by which the signaler’s epistemic reputation is to be enhanced or preserved. So, supposing that the wearer has the right kind of motivational profile, wearing a pin that says, “Vaccines work!” will be an act of propositional EVS because believing the proposition expressed by “Vaccines work!” demonstrates a way of being epistemically good: deference to scientific consensus.

In general, the content used in instances of propositional EVS will be propositions that are already widely accepted by the agent’s reference network. This is because EVS involves status-seeking from the agent’s own perspective. So, insofar as the agent is correct about the socio-epistemic norms in her reference network, the propositions she espouses in order to demonstrate adherence to those norms will also be widely accepted in that network. However, broad acceptance of the content isn’t a necessary condition: If one’s reference network has a different set of norms, such as prizing iconoclastic thought, then espousing a view that is not widely accepted might have the same salubrious effect. What matters is the background socio-epistemic norms and how the content relates to those norms.

Propositional EVS might occur through self-credentialing as well. Introducing oneself with a title—“Please call me Dr. Doolittle” or “Hello, I’m Bill S. Preston, Esquire”—signals a level of educational achievement or relevant skill. Here, the relevant content is a conversational implicature: “Please call me Dr. Doolittle” carries the conversational implicature that the speaker holds a PhD, MD, or similar degree, while “Hello, I’m Bill S. Preston, Esquire” carries the conversational implicature that the speaker holds a JD. Another example is the narrative device Avowal of Prior Skepticism, which is a common technique in many pseudo-scientific contexts (Stone, 2014). This occurs when an individual makes a claim like “I used to be a skeptic, but…,” after which, the speaker explains their, e.g., paranormal experience, then draws the listener’s attention to the fact that the strength of the evidence they just related is what caused them to change their mind. This tactic identifies the speaker as epistemically virtuous by underscoring their evidence-responsiveness. And, as Stone (2014) finds, it is a successful tactic.

Given the role of the reference network in determining which socio-epistemic norms a speaker is attempting to exploit, the propositional content used to signal epistemic virtue might not be content that would be accepted by any wider scientific or intellectual community. So, for example, someone nestled in an anti-vaccine reference network would use skepticism about or outright denial of the effectiveness of vaccines to promote her epistemic prestige within that network. This is because such communities tend to have epistemic norms that are antagonistic toward deference to scientific consensus.¹⁰ What matters for successful propositional EVS is how the propositional content in question relates to the audience’s socio-epistemic norms.

3.2 Demonstrative EVS

EVS does not always rely on the propositional content of the act, however. Instead, it may rely on the way the act is carried out. Demonstrative EVS occurs when individuals behave—for example, in their manner of speech, affect, presentation, etc.—in ways that they presume will be taken as demonstrating good epistemic character (rigor, trustworthiness, etc.) by the reference network. Because it exploits the way a proposition is asserted rather than its content, demonstrative EVS may be used in service of convincing others to accept a proposition they currently reject or doubt.

Propositional EVS may be understood as a subset of demonstrative EVS. Since speech acts are acts, propositional EVS is demonstrative EVS carried out through an act with propositional content. For example, asserting, “I used to be a skeptic, too, but then…” is a behavior that a speaker might expect their audience would understand as demonstrating their good epistemic character. But, demonstrative EVS covers a much wider range of behaviors, many of which are quite subtle.

The adoption of particular accents—such as the “Received Pronunciation” among British accents or General American English (as opposed to African American Vernacular English or many among the family of Southern Regional accents)—is a common example of demonstrative EVS. These forms of speech are often associated with high levels of educational achievement. Regardless of the veracity of that association, the consequence is that speakers who adopt them either improve or protect the level of credibility and overall intellectual regard others assign them.¹¹ Choice of parlance operates similarly because grammatical forms and vocabulary, too, often mark education and social class. As a result, one can adopt such parlance and thereby signal associated epistemic qualities, such as intelligence and diligence. For example, in a 1987 volume on the nature of language acquisition, Sternberg (1987, p. 90) writes, “vocabulary is probably the best single indicator of a person’s overall level of intelligence.” Though Sternberg’s more recent work has embraced a broader understanding of intelligence, as have many other academic researchers, this idea nevertheless persists culturally. Nor did this association begin in the 20th century. For example, in his 4th century Confessions, St. Augustine observes this effect in his description of others’ reactions to the Manichean Bishop Faustus:

My ears had already had their fill of such stuff, and now it did not seem any better because it was better expressed nor more true because it was dressed up in rhetoric; nor could I think the man’s soul necessarily wise because his face was comely and his language eloquent. But they who extolled him to me were not competent judges. They thought him able and wise because his eloquence delighted them. (Augustine’s Confessions, Book V, Ch. VI, Sec. 10)

Here, we see that it is not the content itself, but the way the content is expressed—eloquently and “dressed up in rhetoric”—that persuades interlocutors to view him as “able and wise”.

Sartorial decisions can also constitute EVS. Often, these choices use stereotypical features of intellectual professions in order to signal the epistemic virtues associated with that profession. For example, the use of lab coats in advertising, congressional hearings, and television shows to signal that the wearer is a medical expert falls neatly under the category of demonstrative EVS. Similarly, among many populations wearing glasses raises the wearer’s perceived intelligence (Harris, 1991; AlRyalat et al, 2022). Reese Witherspoon’s character in Legally Blonde, Elle Woods, tries to do exactly this on her first day of law school. Elle is well aware that, because she is young, attractive, and actively fashionable, she is often dismissed as unintelligent. Attempting to combat this perception, she chooses an outfit for her first day of law school that signals that she’s meant to be there, exclaiming “I totally look the part!” to herself as she gets dressed. But, because Elle is only familiar with stereotypical representations of law students and academia she chooses an outfit—complete with a brocade smoking jacket and horn-rimmed glasses—that is ridiculed by her schoolmates. Because she is not, in fact, familiar with the social norms of her reference network—actual law students—she lacks the knowledge and skill to successfully signal the relevant characteristics.

Before moving on, it is worth addressing a worry the previous two sections might inspire: perhaps the category of EVS is too disunified. If everything from using one’s title to wearing glasses can count as epistemic virtue signaling, perhaps this isn’t a well-delineated category after all.¹² While the tokens of EVS are, indeed, quite varied, the power of this concept lies in its explanatory value. EVS helps to explain behaviors—such as Elle donning her smoking jacket, a fumbling first date’s rattling off of endless factoids, or Freshman Comp essays full of ten-dollar words—that otherwise look irrational or strange. These are all tokens of the same type: they are attempts to influence the audience’s judgment of the actor’s epistemic status or character. EVS will certainly cross-cut domains governed by very different norms, which means that understanding the ramifications and reception of any particular act of EVS will require in situ analysis. Nevertheless, having the category of EVS helps to draw attention to its characteristic motivational profile and, therefore, to better understand agents’ decision-making. Moreover, EVS is an important category because it draws attention to the coordination problems it helps to solve. When we need to acquire or communicate information, what is the appropriate level of complexity or context to provide? How much trust should we accord someone? Effectively signaling one’s epistemic virtues is an important part of everyday conversation. When I go to the hardware store and make sure to use all of the jargon at my disposal to talk about my current woodworking project, my choice to convey that expertise (such as it is) to the clerk helps them appropriately calibrate the information they provide. If I decide to engage in EVS by mentioning how much I like Minwax’s 250 VOC products, for example, the clerk can infer that [MOU1]I am knowledgeable with respect to woodworking—they don’t need to tell me about how important it is that I get different grits of sandpaper. In this way, EVS can contribute a great deal to the fluidity and common ground of everyday conversations.¹³ This is an essential, valuable function of EVS, and part of why it is so common.

3.3 Policing Epistemic Virtue

Like moral virtue signaling, EVS is policed by its audience. Consider the cases discussed in the previous section. In both cases, the speaker uses the mode of presentation to elicit a judgment in their audience: a judgment that the speaker is epistemically virtuous in some way. And, in both cases, the audience rejects the signal. In Elle Woods’ case, this is because the signal is inept. In Augustine’s case, the signal is apt, but Augustine himself already has a stable evaluation of the content and speakers who would promote that content. In both cases, the audience rejects the signal and marks it as an intentional act of signaling, which causes them to lower their credibility judgments of the signaler. But the signaling in question need not be as obviously clumsy or ostentatious as these examples to induce sanction. Oppenheimer (2006), for example, shows that unnecessary use of ponderous vocabulary and unnecessary complexity has the effect making the speaker appear less intelligent, rather than more.¹⁴ Just as in the moral case, getting “caught” engaging in inapt EVS damages your epistemic reputation.

The aptness of a particular signal, however, depends on the context. The media scandal that erupted in 2020 concerning Jill Biden’s use of her title of “Dr.” offers a particularly high-profile example of epistemic virtue policing that demonstrates the complexity and nuance of both signals and signaling. In his December 11 editorial in the Wall Street Journal, Joseph Epstein personally addressed Dr. Biden, hectoring, “Any chance you might drop the ‘Dr.’ before your name? ‘Dr. Jill Biden’ sounds and feels fraudulent, not to say a touch comic,” (Epstein, 2020). This needling relies on a clash of reference networks with different norms about the significance of “Dr.” as a signal. For Biden, who holds a Doctor of Education (Ed.D.) degree, the title of “Dr.” marks a particular epistemic and practical achievement: obtaining a degree that requires sufficiently deep epistemic engagement with a topic. Because she has done so, the title “Dr.” is an appropriate signal within her reference network.¹⁵ Epstein, despite addressing his editorial to Biden herself, was speaking to a different reference network: the non-academic public, whose familiarity with the title is largely limited to medical doctors. Two years earlier, National Public Radio’s Public Editor Elizabeth Jensen responded in a similar way to listener complaints concerning NPR’s coverage of the confirmation hearings for Brett Kavanaugh’s nomination as a Supreme Court Justice. During the hearings, NPR’s coverage gave Kavanaugh the title of “Judge”, but gave no title to Christine Blasey Ford, who holds a doctorate (Ph.D.) in educational psychology. Listeners wrote in to NPR complaining that, “the disparity of hearing Kavanaugh called ‘Judge Kavanaugh,’ at times, and not hearing Ford referred to as ‘Doctor’ rankled. One called it ‘offensive,’ saying it showed how women are disrespected in relation to men. Another called it an ‘insidious bias.”’ In defense of NPR’s policy of avoiding ‘Dr.’ for Ph.D. holders, Jensen wrote, “for most listeners, a ‘Dr.’ practices medicine,” (Jensen, 2018). According to these reference networks, the title of “Dr.” conveys very different achievements.

But mere disagreement about the practical achievement of a particular degree is not enough to account for Epstein’s condescending choice of words. Epstein justifies his claim that the title is “a touch comic” by pointing out that Dr. Biden’s degree is “a Doctor of Education, earned at the University of Delaware through a dissertation with the unpromising title ‘Student Retention at the Community College Level: Meeting Students’ Needs”’. Here, Epstein clearly implies that the epistemic achievement indicated by Biden’s degree and her dissertation title is dwarfed by the epistemic achievement of someone who attains a medical degree. Thus, the title is seen not just as an indicator of one’s history, but also as a marker of epistemic status. As we’ll see, this pattern of policing plays a crucial role in the double-bind of testimonial injustice, to which we now turn.

4. The Double-Bind of Testimonial Injustice

Fricker (2007) identifies the central case of testimonial injustice—“prejudicial dysfunction in testimonial practice” (p.17)—as a credibility deficit. Credibility deficits occur when the recipient of a piece of testimony deflates the credibility of the testifier in a way that wrongs them in their capacity as a knower (ibid, p. 20). A grounding example of this kind of injustice is deflating one’s estimate of a speaker’s credibility in virtue of their race. Doing so insults and undermines the speaker’s capacity as a knower because it renders that capacity a function of an irrelevant, uncontrollable trait: their race.

For our purposes, what is important about this concept is that testifiers are not naive to it. Regardless of whether they possess the concept of testimonial injustice, many women, transgender people, and people of color, for example, are familiar with the fear that their testimony will not be taken seriously as a result of their demographic features. Often, speakers simply silence or smother their testimony, opting not to experience the potential injustice.¹⁶ But, consider the agent who chooses, instead, to refuse silencing. In this case, they face a double bind: speak up and suffer the injustice, or combat the prejudice through EVS and risk further injustice. We consider these binds in turn.

Perhaps the proper response to the fear of testimonial injustice is refusing to be silenced and speaking up. Here, we encounter the first bind: If one simply speaks, they risk suffering the testimonial injustice. So, simply speaking up is not a good option. Consider the second bind, then. If the fear is that one will not be taken seriously, then perhaps the proper response is to ward off the potential testimonial injustice by engaging in defensive EVS:

Defensive EVS. Epistemic virtue signaling carried out for the purpose of overcoming or thwarting potential deflation of one’s credibility in virtue of the audience’s assessment of the signaler’s epistemic character.

Examples of this are commonplace. Imagine a young woman who finds herself in an all-male university programming course:

Reina’s Recursion. Reina is a student in her University’s Algorithms and Data Structures course. By happenstance, she is the only woman in her class. One afternoon, Reina notices an error in her professor’s presentation—the professor has written down the algorithm they’re studying in a way that breaks the recursion. Reina considers pointing out the error but pauses out of fear that her classmates and professor might not listen to The Girl. Swallowing the fear, Reina raises her hand and, when called on, lowers her voice, uses a bit of technical jargon that emphasizes her familiarity with the topic, and peppers her comments with claims that she suspects her audience will regard as markers of the tech cognoscenti.

Reina labors under threat of the negative stereotype that women are not particularly good at programming and, in response, defensively deploys several epistemic virtue signals. The first two are examples of demonstrative EVS: by lowering her voice and adopting jargon, Reina presents her comments in a way that she suspects will bolster her credibility in the eyes of her reference network. The last is an example of propositional EVS: by espousing such views, Reina expects to signal epistemic virtue, and thereby earn credibility, in the eyes of her audience.

Here, it is worth pausing to note that Defensive EVS is closely related to code-switching. Code-switching is the act of adapting one’s behavior to meet the norms governing a particular context.¹⁷ There are many reasons one might engage in code-switching, but EVS draws our attention to an important motivation: securing credibility.¹⁸ The examples of shifting accents discussed in Section 2.2 are a kind of code-switching carried out for the purpose of EVS because modes of speech and cultural markers are closely associated with educational achievement. In focus group discussions, Scott (2013, p. 319) finds black women engaging in a mixture of code-switching and EVS in order to dispel negative stereotypes and perform competence:

“I find myself talking properly—articulating, enunciating, expanding my vocabulary while in class. (Sandra)

Oftentimes when I speak, I speak very intelligently. I study hard, get good grades, and I like to let people know. I’ll leave my paper out, like, “Yeah, I got the only A in the class—it was me.” To let people know we’re intelligent. (Oni)

I actually have conversations with people and let them know I’m not your typical Black woman you met somewhere else down the road. I’m not her, and I will let you know that easily. (Jessica)

In communicating with people, I work very hard at using code switching. So I talk proper English that I learned in school, especially in the classroom or around people I attend school with. And I’m learning not to do certain behaviors, such as resting my hand on my hip or roll my eyes, when in certain environments. (Linda)”

And, in at least some contexts, code-switching is successful at its aims. McCluney et al (2017), for example, finds that black employees describing their code-switching behaviors were regarded as displaying professionalism by their colleagues. Moreover, we might expect that marginalized individuals who, unlike Elle Woods, frequently feel the need to engage in code switching for the purpose of EVS are likely to be particularly adept at it (Britt & Weldon 2015). Given that EVS cross-cuts code-switching, these results may carry over as well, though empirical research remains to be done.¹⁹

Nevertheless, defensive EVS carries several risks, the most obvious of which is backlash. Recall that moral virtue signaling is often seen as self-undermining. Tosi and Warmke (2020, Ch. 6), for example, argue that the characteristic motivational profile of virtue signaling—improving or preserving one’s reputation—undermines the speaker’s moral character. They argue, “if someone uses public moral discourse to seek greater social status and make herself look impressive, for instance, she departs significantly from the motivations traditionally associated with virtue” (p. 125). Here, they have in mind motivations to do something because it helps others or because it is the right thing to do. Regardless of whether this is a sound judgment, a similar pattern arises in the case of EVS.²⁰ This is what happened to Elle Woods in the example above: Because her attempt at EVS is clumsy, it is noticed by her audience (her new academic peers). And, instead of simply being ignored, the clumsy attempt at EVS makes her the target of ridicule. Similarly, consider how you might react to someone who, in the middle of a conversation about something else, declares unprompted that their IQ is 155. This is clearly an instance of EVS, since the speaker is using this claim about their IQ to attempt to influence your judgment of their epistemic virtue. But this attempt at EVS is unlikely to succeed because, just as in the moral case, it appears self-undermining. After all, shouldn’t the quality of the speaker’s argument stand on its own? Furthermore, assuming the speaker’s reference network does not have socio-epistemic norms favoring outright declarations of formal measures of intelligence (or perhaps does have norms against such declarations),²¹ this failure to adhere to the relevant socio-epistemic norms has the effect of signaling the opposite of the speaker’s intention, tarnishing rather than burnishing their epistemic reputation. This effect can be found in the data concerning Avowal of Prior Skepticism, the narrative form of EVS discussed in 2.2, as well. Stone (2014) finds that this tactic is self-undermining when the audience is made aware of it as a tactic for enhancing the speaker’s credibility: “Avowal of Prior Skepticism serves to increase the believability of a paranormal explanation for an anomalous event as long as participants are not made suspicious of the motives behind the avowal” (p. 272). EVS, in other words, can backfire.

Returning to the double-bind, Reina’s defensive EVS is risky because if the speaker is “caught”, her credibility will be still undermined, and perhaps beyond the deflation she originally feared. To see this, remember that Reina finds herself in this double bind because of her awareness that she may face testimonial injustice. So, she feels the need to signal. If she is right, however, and her audience does, in fact, have epistemically prejudicial attitudes toward her, they may be more vigilant in looking for signs of EVS. Thus, the virtue signaling she must do in order to secure an appropriate judgment of her credibility puts her at high risk for even further deflation of her credibility because of the very prejudice that coerces her EVS in the first place.

How might this happen? On one hand, she might misstep. If, for example, she mistakes RAM for ROM—a mistake almost any computer science student would recognize—her comments will confirm her classmates’ prejudices. (Better, as they say, to stay silent and be thought a fool than to speak up and remove all doubt.) In this case, engaging in virtue signaling will leave her in worse regard than had she not bothered. Even without factual error, however, her EVS may still trigger backlash if her vigilant audience catches it. Consider, for example, the intense scrutiny paid to Elizabeth Holmes’ renowned contralto.²² While women with deeper voices are generally regarded as more trustworthy and competent (O’Connor & Barclay 2017, Krahé et al. 2021), suspicion that one is changing their voice may evoke the opposite reaction. This is consistent with empirical research on impression management: focusing on hiring contexts, Roulin et al (2015) show that detected impression management reduces employers’ evaluations of candidates. Moreover, they show that employers are more likely to notice honest, as opposed to deceptive, attempts at impression management. Insofar as these results carry over to the general dynamics of EVS, it suggests that honest efforts at EVS are more likely to be detected and therefore, more likely to be punished, than dishonest ones.²³ So, Reina faces a delicate balance even as she gains experience with EVS because attempting to combat prejudice against her faces compounding risks: not only is signaling risky, but because it is honest signaling to a vigilant, negatively prejudiced audience, that risk is greater than it would otherwise be.

There are more subtle harms of finding oneself in this double bind as well. Many people deeply value their status as members of a knowledge community, and it can be a significant part of one’s self-worth. As a result, this double bind can give rise to a deep sense of precariousness, wherein one’s continued membership in the community is fragile and dependent on continually monitoring others’ estimation and acceptance of your membership.²⁴

Moreover, in many contexts the behaviors one might employ in the course of EVS will also be associated with social identities, such as race, class, and gender. Members of marginalized or oppressed groups who engage in EVS to a dominant audience using those signals may, as a result of those associations, feel as if they are rejecting their own identities (Carter 2005, Durkee & Williams 2015). As Oni, a focus group participant for Scott (2013, p. 324), puts it, “I feel like I’m trying to perform for them, and that makes me feel bad as well because I feel like I don’t know how to be true to myself and dispel a lot of the stereotypes they have about us.” The effort of combating negative stereotypes inculcates feelings of inauthenticity and exhaustion.

4.1 Deception, Real or Imagined

The negative outcomes of this double-bind are deeper and further-reaching when the signaling involves deception, whether real or imagined.

Individuals who feel forced to engage in defensive EVS may do so by stretching the truth, omitting unflattering facts, lying, or otherwise deceiving their audience.²⁵ These cases might involve outright lies, such as Representative George Santos fabricating a college education (Penzenstadler and Quintana, 2023), or more subtle deceptions, such as our young programmer using jargon to suggest a deeper level of knowledge than she actually possesses. In either case, the exposed virtue signaler again loses credibility beyond the feared credibility deficit. But, a host of other consequences follow, which are both epistemically and morally significant.

Focusing on the virtue signaler themselves, coming to the conclusion that one must lie or deceive in order to ensure that they are taken seriously is harmful on several counts. One’s status as a knower is not merely precarious, as it is for anyone who faces this dilemma regularly. Because this requires viewing intentional, active deception as a live, perhaps necessary, option, it may also undermine the agent’s sense of themselves as a reliable epistemic agent. Moreover, it deepens their sense of being an outsider to the community. The agent who faces this dilemma fears that they are a member of their knowledge community only insofar as they are not too well-recognized by that community—no matter which horn of the dilemma they choose, they face malrecognition, either through others’ prejudice or their own deception. To the extent that this form of recognition is “a vital human need”, as Taylor (1994, p. 26) puts it, understanding that one’s position in their own community is laced with malrecognition is antithetical to that need.

This applies regardless of whether the speaker actually engages in any deception. If they do, further harms may arise. The speaker themselves may acclimate to deception, becoming more likely to lie to others in the future.²⁶ Additionally, Hewlin’s (2003, 2009) research into facades of conformity, which are “false representations created by employees to appear as if they embrace organizational values” (2003, p. 634), suggests that frequent, low-level deception in the service of impression management leads to emotional exhaustion. While there is little empirical research into EVS as such, insofar as deceptive EVS is similarly aimed at managing one’s impression, similar effects of exhaustion and burnout may occur, even where the signaling is successful.²⁷ Beyond this, if the virtue signaler is successful in their deception, they may engender false beliefs in their audience, making the speaker quite literally a less-reliable source of testimony and harming the audience by giving them falsehoods upon which they might erroneously rely.

Finally—and most worryingly—if the deception is uncovered or the hearer believes that they have uncovered deception, it may ingrain the very prejudice the speaker was attempting to combat. Where a hearer may have absorbed prejudicial attitudes from their culture or the attitudes of those around them without particular examples, the (even merely purported) deception provides a direct example. Moreover, since prejudicial attitudes are often held in the form of a nebulous generic construction, such as “addicts lie” (as opposed to a universally quantified form like “all addicts lie”), these attitudes are easily buttressed by individual observations. And, there is strong incentive for the hearer to seek deceptions and respond to them this way: by confirming a prejudicial attitude and thereby “justifying” it, they may reduce feelings of uncertainty or guilt associated with that attitude.

4.2 Situating the Double Bind

The double bind described in Section 3 is a clear example of Sukaina Hirji (2021)’s norm-based account of oppressive double binds. Drawing on Frye (1983), Hirji argues that double binds should be understood as “choice situations where an agent is forced to choose between cooperating with and resisting some oppressive norm, and in which whatever they do, they end up reinforcing to some degree the oppressive structures that constrain their options.” (p. 658). On this view, oppressive double binds are not merely situations in which one will suffer no matter which choice they make. Instead, their suffering is directly related to their goals; the very agency the individual employs in making their decision is undermined by that decision.

In Reina’s case, and cases like it, the oppressive norm in question is epistemic. Under this norm, members of dominant groups exert undue influence over whose testimony is taken up and whose is diminished or rejected. This norm engenders the core injustice that Fricker (2007, p.145) identifies in testimonial injustice: “the subject’s being wronged in their capacity as a giver of knowledge.” And, it constrains Reina’s choices to options that undermine her epistemic agency. If Reina chooses not to engage in EVS, she cooperates with the norm. She exercises her agency, but only in a way that acquiesces to others’ dominance. If, instead, she chooses to engage in EVS, she resists. In doing so however, she risks being further diminished as a participant in her epistemic community.²⁸

Moreover, as in the case of code switching (McCluney et al, 2021), the effectiveness of EVS will depend on how well it is received by its audience. Given that this reception will be mediated by the audience’s attitudes concerning the oppressive norm governing the double bind, this has further ramifications for the decider’s agency. In any choice situation, the best option is defined by the outcomes. But, oppressive double binds like this one are made worse by the fact that the oppressive norms and attitudes that construct them are all but invisible. Because this double bind arises in situations in which an agent fears (but perhaps does not know) that they may be subject to testimonial injustice, an agent like Reina must assess her options amidst the fog of stereotypes, microaggressions, uncertain allies, and so on.²⁹ Thus, her ability to effectively employ her epistemic agency is seriously diminished.

This suggests a connection with Jennifer Lackey’s (2020) notion of agential testimonial injustice.³⁰ Lackey focuses on cases of false confessions, in which agents’ testimony is extracted from them through coercion, manipulation, or torture. In such cases, she says, “the confessor’s status as a knower is reduced to what she reports only under conditions devoid of, or with diminished, epistemic agency” (p. 59). Though the double bind of testimonial justice described here does not involve coercion, since an individual like Reina can simply choose not to offer testimony at all, it nevertheless involves a distinct diminishment of epistemic agency. Her choices, if she wishes to testify, are choices that undermine her agency.³¹

5. Application: Medical Autonomy

This section examines how testimonial injustice and EVS interact in the erosion of medical autonomy, focusing on the care gap faced by pregnant women of color in the modern American healthcare system.

Between dismissal of elderly patients, mismanagement of pain among African Americans, the abuse of mentally ill or disabled patients, and denial of care to (or outright abuse of) transgender patients, prejudice in health care settings is well-documented (Alhusen et al, 2016; Alio et al, 2010; Bower et al, 2018; Hoffman et al, 2016; Jaffee et al, 2016; Knaak et al, 2017). Many of these cases involve testimonial injustice: healthcare providers deflating the credibility they assign to patients’ testimony as a result of identity prejudices.³² This case is particularly important because being believed by one’s healthcare provider is an essential part of medical autonomy. And, in situations in which healthcare is prohibitively expensive or difficult to navigate, the stakes are quite high: an individual who is denied medical autonomy may not be able to afford to find a second provider. As a result, this double bind will be felt acutely.

The gap in quality of experience and outcome that pregnant women of color face, as compared with their white counterparts, is well-documented (Hill et al, 2022). Altman et al (2019) studied this gap by interviewing patients who experienced it. As their study suggests, providers’ management of information is a key aspect of that gap:

[Patients] perceived providers to be intentionally or inadvertently using information sharing as a mechanism for controlling interactions with and influencing decision-making of patients. Acts of withholding information, providing partial information, or providing misleading information were often perceived by participants to reflect the provider’s assumptions around the patient’s ability to make “good” decisions on their own behalf and therefore influenced patients’ autonomy and self-determination. (Altman et al, 2019, p. 3)

Patients perceived their providers as using judgements about the epistemic capacities of their patients in determining how they packaged information for them—information crucial to their decision-making. And, therefore, crucial to their medical autonomy. Patients interviewed by Altman’s team described feeling as though they were neither cared about nor valued as knowers, specifically. And, this experience was directly tied, at least for some patients, to whether they engaged in EVS. Describing her experience, one patient recalled,

…I always received a different response once people realized that I went to UC Berkeley. Like somehow it comes up and then they treat me differently, which is really interesting because it feels unfair because what if I hadn’t, you know? (Ibid, p. 5)

Altman et al (2019) highlight this participant’s narrative as emblematic of a theme running through the experiences of many interviewees. Patients whose participation in higher education became known to providers experienced a difference in care—they felt that they were perceived differently because of their education. For this participant, the role of her EVS in eliciting better treatment from her providers only underscored the extent to which racial stereotypes like, as she put it, “crazy Black woman” were influencing her treatment and her providers’ willingness to enable her medical autonomy.

This patient’s case draws out the harm of this double bind. If she does not engage in defensive EVS to display her education to her healthcare providers, her medical autonomy is at risk. But, if she does engage in EVS, she experiences alienation, precarity, and guilt:

And I feel like a part of me [mentions my prestigious education] on purpose because I know that they’re going to treat me better after I say that, which makes me feel a little bit bad because I am accessing privilege. (Ibid, p. 5)

Medical autonomy should not be a privilege of those who flash adherence to a particular set of socio-epistemic norms, and patients ought not fear that their access to autonomy is limited by the extent to which they can convince healthcare providers of such adherence.

6. Why it matters

Moral virtue signaling is often thought, to put it bluntly, to be a bit gouache. In the epistemic realm, however, the story is quite different. We rely on signals of others’ epistemic virtue deeply, and, I think, unavoidably. This is because our moral reputations play a very different role from our epistemic reputations. As Hardwig (1985) discusses, we are deeply epistemically dependent on one another. Our everyday lives require us to rely on others’ epistemic capacities regularly, from the morning weather report to the doctor’s office to deciding whether to take a friend’s advice about fixing your bike, epistemic dependence is a fundamental part of life in a complex society. And—if we are not overly credulous—we are constantly, if unconsciously, assessing the credibility of those upon whom we rely. In doing so, we rely on what we take to be signals of epistemic virtue. We must, because we seldom have direct knowledge of others’ epistemic backgrounds.³³

In this article, I have focused on misleading, prejudicial, or otherwise flawed signals and signaling, but this is by no means the majority of the signals to which we respond. Responding to a speaker’s confidence or their familiarity with their subject matter is a reliable, if imperfect, way to modulate the credibility we assign. And, as we saw in Section 2.2, EVS is an important part of everyday conversation. This is because it helps to solve the epistemic coordination problem that plagues many of those conversations, as the hardware store jargon example illustrates. Moreover, my EVS in that situation is defensive: it successfully preempts my interlocutor acting on any assumption they might have that, because I am a woman, I don’t know anything about woodworking. While I can’t know whether they had such an assumption, not knowing is better than having to find out that they did. More importantly, when the women in Altman et al (2019)’s study were able to successfully employ EVS, they overcame the injustice they initially encountered and received information that was calibrated to their capacities. The combination of epistemic dependence and the need to coordinate makes epistemic virtue signaling an essential part of communication.

But, because our understanding of these signals is informed by not just our own experience with them, but also the socio-epistemic norms we have absorbed from those around us, these signals are bound up with the biases of our broader social context. This helps to explain the connection between EVS and testimonial injustice. Focusing on the double-bind that emerges from this link highlights the fact that those subject to testimonial injustice cannot be responsible for fixing it, at least not without significant risk to themselves. The mere threat of testimonial injustice requires the potential victim to strategize their communication at the very least, and potentially in a no-win situation.

This underscores the importance of Fricker’s (2007, Ch. 3) virtuous hearing, but also reveals the limitations of that solution. Fricker’s discussion of testimonial injustice is limited to prejudicial dysfunction, which relies on recalcitrant and emotional prejudice. But the cases we’ve been considering here, in which an agent’s socio-epistemic norms may arise, not only from internal animus but also—and alternatively—from absorbed expectations about credibility, call for a different kind of solution. This is a project I leave to future work.³⁴

Notes

That said, there are moral undertones to these cases, as will be true of many similar ones. Delving into this relationship is beyond the scope of this paper and will be left to a lengthier treatment. ⮭
Jack Warman independently offers a brief account of the nearby concept of ‘intellectual grandstanding’, which hews closely to Tosi and Warmke’s (2016; 2020) discussion of moral grandstanding. The concept I introduce here is broader in scope, most notably because it is not restricted to speech acts and the primary cases of interest will not be “abuse of intellectual talk”, as man (2021, p. 212) defines intellectual grandstanding. Readers interested in applications of Tosi and Warmke’s notions of piling on, ramping up, and trumping up should refer to Warman (2021). ⮭
While Tosi and Warmke (2020) focus on speech acts, a well-placed sigh, wearing a “Save the Whales” t-shirt, or blacking out one’s social media profile picture all seem to be equally potential acts of virtue signaling that function in the same way. So, making a distinction between speech-virtue signaling and non-speech-virtue signaling seems artificial. ⮭
The fineness of this distinction leads Tosi and Warmke (2020) to the conclusion that the term “virtue signaling” should be abandoned in favor of “moral grandstanding”. However, I take the distinction between virtue signaling and signals of virtue to be, like the distinction between yarn balling and balls of yarn, fairly straightforward. More importantly, “grandstanding”—whether moral or epistemic—conjures a very specific, rather obnoxious form of virtue signaling. Focusing on this terminology, then, threatens to obfuscate many subtler cases that deserve attention. Here, again, readers interested in this more ostentatious abuse of intellectual talk should refer to Warman (2021). ⮭
One need not find both models convincing to see the value in the concept of virtue signaling. Moreover, even if neither model accurately describes dynamics of individuals’ beliefs in relation to public discourse, it is sufficient if these models capture agents’ own understanding of how public discourse functions—if they do, then they explain such agents’ choice to engage in virtue signaling. ⮭
Epistemic bubbles—a social epistemic structure in which some relevant voices have been excluded through omission (Nguyen 2020, p. 142)—exacerbate this effect. Because individuals stuck in epistemic bubbles do not encounter pushback by definition, the socio-epistemic norms they develop can differ dramatically. ⮭
This is especially important in the context of echo chambers, which often use adherence to certain epistemic norms as a membership requirement (Nyugen 2020). ⮭
For discussions of intellectual virtues in virtue epistemology, see Zagzebski (1996), Baehr (2011), Morton (2012) and Montmarquet (1993) ⮭
Epistemic virtue signaling may be neither holy, nor Roman, nor an empire, as it were. ⮭
This underscores the point that the “correct” epistemic norms are irrelevant to EVS. The “Vaccines work!” pin will fail as an epistemic virtue signal in a context in which the observers of the signal are anti-science, regardless of the actual status of deference to scientific consensus as an epistemic norm. ⮭
See Bernstein (1960) for a general discussion of formal language as a class marker, Collins et al (1996) for discussion of the “Received Pronunciation”, Young et al (2013) for the specific case of AAVE, Dodsworth (2013) for discussion of Southern accents in the United States, and Labov (2012) for a general discussion of the politics of accents in the United States. ⮭
Thank you to an anonymous referee for suggesting this worry. ⮭
Focusing on the linguistic analysis, there is much more to be said about epistemic virtue signaling as a speech act (Austin, 1962), its contributions to the common ground (Stalnaker, 2002), and its relationship to Gricean conversational maxims (Grice, 1975), but this discussion is beyond the scope of the present work. ⮭
For example, comparing two professional translations of Descartes’ Meditations, Oppenheimer finds that participants consistently rated author of the the less complex translation more intelligent, regardless of whether they knew that the author was Descartes. ⮭
There is also much to be said here about the political decision Biden was making in this context, given that she—like any academic woman—is likely well aware of the dynamic that Epstein exploited to create the scandal. Her insistence on her title, even in the face of Epstein’s reference network, is an effort to change the relevant norms. ⮭
See Dotson (2011) for extended discussion of silencing and smothering. ⮭
Code switching is well-studied in linguistics, where the term is used to pick out bilingual speakers’ ability to switch between languages depending on their context. In sociology, Du Bois’ The Souls of Black Folks provides an early illustration of the cultural form of code switching in his discussion of ‘double-consciousness’. ⮭
Code switching is distinct from EVS in this way: While EVS is partially constituted by its motivational profile, code switching need not have any motivation at all. ⮭
It is also worth noting that, while there is significant research on the frequency of code switching, there is relatively little on the effectiveness of code switching on others’ views of. McCluney et al (2017) and Roberts et al (2014) are exceptions. Moreover, the extent to which research on professionalism and hiring carries over is unclear, in part because professionalism is generally regarded as a mutable trait, whereas epistemic virtues may be regarded as innate. ⮭
There is an important question here about whether EVS should be regarded as self-undermining in the same way that moral virtue signaling is. However, this discussion would take us too far afield for present purposes. ⮭
Such norms might be moral, but might also be matters of intellectual humility or good judgment. In fact, both epistemic and moral norms might legislate against this kind of declaration in many contexts. As in many cases, moral and epistemic norms concerning humility can be intricately intertwined. ⮭
In a May 2023 tell-all with the New York Times, the second paragraph reads, “In case you’re wondering, Ms. Holmes speaks in a soft, slightly low, but totally unremarkable voice, no hint of the throaty contralto she used while running her defunct blood-testing start-up Theranos,” (Cheung 2023). Prior to this, a 2019 article for New York Magazine’s The Cut dwells on the question “What Kind of Person Fakes Their Voice?”, with Holmes as its central character (Heaney, 2019). ⮭
As of this writing, however, there is no direct empirical research on epistemic virtue signaling in broader contexts. ⮭
This worry shares much with Hookway’s (2010) discussion of the role of the participatory perspective in creating forms of epistemic injustices. ⮭
Roberts et al (2008), for example, provides evidence that those who feel that their identities are devalued in relevant context are more likely to downplay those identities when crafting self-presentation, using behavior and speech to either suppress the devalued identity or bolster similarity with valued identities. Where these identities are devalued in virtue of negative epistemic stereotypes, this intentional management of one’s impression on others is a form of EVS. ⮭
Some empirical research suggests that repeated deception may even wear a biological rut into our brains. Garrett et al (2016) show that the brain’s stress response, via the amygdala, to engaging in self-serving deception decreases with repetition. This, they argue, helps to explain the observation that major frauds—Madoff, Enron, and others—began as small transgressions. ⮭
Hewlin (2009, p 735-6) also notes that the presence of this behavior is strongly related to individuals’ perception of their environments. Specifically, subjects who felt that their work environments were “non participative”, meaning that they are less receptive to diverse ideas and expressions, were significantly more likely to engage in facades of conformity. This suggests that, even if people are generally disposed toward honest behavior, prejudicial or hostile environments may encourage less honest behavior. ⮭
Thank you to an anonymous referee for suggesting drawing out this connection. ⮭
Studying the effects of “resume whitening”, altering the name, achievements, or both on a resume tso as to appear less racialized, Kang et al (2016) found that “minorities may be particularly likely to experience disadvantage when they apply to ostensibly pro-diversity employers” (p. 1). Their results suggest that apparent openness to diversity may not be reflected in individuals’ behaviors. ⮭
Thank you to an anonymous referee for suggesting this connection. ⮭
Lackey’s cases also involve a credibility excess given to the coerced testimony (2020, p. 61). In cases of agential testimonial injustice that do not involve coercion, it is no surprise that no such credibility excess would present itself. However, this should still be considered agential testimonial injustice because coercion is just one way to undermine an individual’s epistemic agency. The undermining that goes on in this double bind is one of many, much subtler ways to do so. This suggests that the category of agential testimonial injustice may warrant further investigation. ⮭
See Drożdżowicz (2021); Kidd and Carel (2014); Crichton et al (2017); Scully (2018) for discussion of epistemic injustice in healthcare contexts. ⮭
And, as Hardwig (1985) points out, even if we do have knowledge of their backgrounds, we cannot verify their expertise unless we are ourselves experts in the same subject matter. ⮭
Acknowledgements: I am grateful to T.Y. Branch, Robin Dembroff, Georgi Gardiner, and Ravi Thrakal for helpful feedback that improved this paper, as well the Values in Science Group at Université du Québec and audiences at the 2023 Midwest Epistemology Workshop, the 2023 Pacific Meeting of the American Philosophical Association, and the 3rd Political Epistemology Network Conference. ⮭

References

Alhusen, J. L., Bower, K. M., Epstein, E. & Sharps, P. (2016). Racial discrimination and adverse birth outcomes: An integrative review. Journal of Midwifery & Women’s Health, 61(6), 707–720. http://doi.org/10.1111/jmwh.12490

Alio, A. P., Richman, A. R., Clayton, H. B., Jeffers, D. F., Wathington, D. J., & Salihu, H. M. (2010). An ecological approach to understanding Black–White disparities in perinatal mortality. Maternal and Child Health Journal, 14, 557–566. http://doi.org/10.1007/s10995-009-0495-9

AlRyalat, S. A., Jumaah, M., Al Hajaj, S. W., Al-Noaaimi, F., Alawneh, Y., & Al-Rawashdeh, A. (2022). The Effect of Wearing Eyeglasses on the Perception of Attractiveness, Confidence, and Intelligence. Cureus, 14(3), 1–6. http://doi.org/10.7759/cureus.23542

Altman, M. R., Oseguera, T., McLemore, M. R., Kantrowitz-Gordon, I., Franck, L. S., & Lyndon, A. (2019). Information and power: women of color’s experiences interacting with health care providers in pregnancy and birth. Social Science & Medicine, 238, 1–9. http://doi.org/10.1016/j.socscimed.2019.112491

Austin, J. L. (1962). How to do things with words. J. O. Urmson (Ed.). Harvard University Press.

Baehr, J. (2011). The inquiring mind: On intellectual virtues and virtue epistemology. Oxford University Press.

Bernstein, B. (1960). Language and social class. The British Journal of Sociology, 11(3), 271–276. http://doi.org/10.2307/586750

Bicchieri, C. (2017). Norms in the wild: How to diagnose, measure, and change social norms. Oxford University Press.

Bower, K. M., Geller, R. J., Perrin, N. A., & Alhusen, J. (2018). Experiences of racism and preterm birth: findings from a pregnancy risk assessment monitoring system, 2004 through 2012. Women’s Health Issues, 28(6), 495–501. http://doi.org/10.1016/j.whi.2018.06.002

Britt, E., & Weldon, T. L. (2015). African American English in the middle class. In J. Bloomquist, L. J. Green, & S. L. Lanehart (eds), The Oxford Handbook of African American Language (pp. 800–816). Oxford University Press. http://doi.org/10.1093/oxfordhb/9780199795390.013.44

Carter, P. L. (2005). Keepin’ it real: School success beyond black and white. Oxford University Press.

Chozick, A. (2023, May 8). Liz Holmes wants you to forget about Elizabeth. New York Times. https://www.nytimes.com/2023/05/07/business/elizabeth-holmes-theranos-interview.html (accessed July 23, 2023).

Collins, B., & Mees, I. M. (1996). The phonetics of English and Dutch (3rd ed.). Brill Academic Publishers.

Crichton, P., Carel, H., & Kidd, I. J. (2017). Epistemic injustice in psychiatry. BJPsych Bulletin, 41(2), 65–70. http://doi.org/10.1192/pb.bp.115.050682

Dodsworth, R. (2013). Retreat from the southern vowel shift in Raleigh, NC: Social factors. University of Pennsylvania Working Papers in Linguistics, 19(2), 31–40.

Dotson, K. (2011). Tracking epistemic violence, tracking practices of silencing. Hypatia, 26(2), 236–257. http://doi.org/10.1111/j.1527-2001.2011.01177.x

Drożdżowicz, A. (2021). Epistemic injustice in psychiatric practice: epistemic duties and the phenomenological approach. Journal of Medical Ethics, 47(12), 1–5. http://doi.org/10.1136/medethics-2020-106679

Durkee, M. I. & Williams, J. L. (2015). Accusations of Acting White: Links to Black Students’ Racial Identity and Mental Health. Journal of Black Psychology, 41(1), 26–48. http://doi.org/10.1177/0095798413505323

Epstein, J. (2020, December 11). Is there a doctor in the white house? Not if you need an M.D.. The Wall Street Journal. https://www.wsj.com/opinion/is-there-a-doctor-in-the-white-house-not-if-you-need-an-m-d-11607727380 (accessed June 9, 2025).

Fricker, M. (2007). Epistemic injustice: Power and the ethics of knowing. Oxford University Press.

Frye, M. (2000). Oppression. In L. Code (Ed.), Encyclopedia of feminist theories (p. 370). Routledge.

Garrett, N., Lazzaro, S. C., Ariely, D., & Sharot, T. (2016). The brain adapts to dishonesty. Nature Neuroscience, 19(12), 1727–1732. http://doi.org/10.1038/nn.4426

Grice, H. P. (1975). Logic and conversation. In D. Davidson & G. Harman (Eds.), The logic of grammar (pp. 64–75). Dickenson Publishing.

Hardwig, J. (1985). Epistemic dependence. Journal of Philosophy, 82(7), 335–349. http://doi.org/10.2307/2026523

Harris, M. B. (1991). Sex differences in stereotypes of spectacles. Journal of Applied Social Psychology, 21(20), 1659–1680. http://doi.org/10.1111/j.1559-1816.1991.tb00497.x

Heaney, K. (2019, March 21). What kind of person fakes their voice? The Cut. https://www.thecut.com/2019/03/why-did-elizabeth-holmes-use-a-fake-deep-voice.html (accessed July 22, 2023).

Hewlin, P. F. (2003). And the award for best actor goes to…: Facades of conformity in organizational settings. The Academy of Management Review, 28(4), 633–642. http://doi.org/10.2307/30040752

Hewlin, P. F. (2009). Wearing the cloak: Antecedents and consequences of creating facades of conformity. Journal of Applied Psychology, 94(3), 727–741. http://doi.org/10.1037/a0015228

Hill, C., Corbett, C., & St. Rose, A. (2010). Why so few? Women in science, technology, engineering, and mathematics. American Association of University Women (AAUW). https://www.aauw.org/app/uploads/2020/03/why-so-few-research.pdf (accessed June 9, 2025).

Hill, L., Artiga, S., Ranji, U. (2022). Racial disparities in maternal and infant health: Current status and efforts to address them. Kaiser Family Foundation. https://www.kff.org/racial-equity-and-health-policy/issue-brief/racial-disparities-in-maternal-and-infant-health-current-status-and-efforts-to-address-them/ (accessed January 28, 2023).

Hirji, S. (2021). Oppressive double binds. Ethics, 131(4), 643–669. http://doi.org/10.1086/713943

Hoffman, K. M., Trawalter, S., Axt, J. R., & Oliver, M. N. (2016). Racial bias in pain assessment and treatment recommendations, and false beliefs about biological differences between blacks and whites. Proceedings of the National Academy of Sciences, 113(16), 4296–4301. http://doi.org/10.1073/pnas.1516047113

Hookway, C. (2010). Some varieties of epistemic injustice: Reflections on Fricker. Episteme, 7(2), 151–163. http://doi.org/10.3366/epi.2010.0005

Jaffee, K. D., Shires, D. A., & Stroumsa, D. (2016). Discrimination and delayed health care among transgender women and men: Implications for improving medical education and health care delivery. Medical Care, 54(11), 1010–1016. http://doi.org/10.1097/MLR.0000000000000583

Jensen, E. (2018, September 28). NPR doesn’t confer “Dr.” on Ph.Ds. Here’s why. National Public Radio. https://www.npr.org/sections/publiceditor/2018/09/28/652538222/npr-doesnt-confer-dr-on-ph-d-s-heres-why (accessed June 9, 2025).

Carel, H., & Kidd, I. J. (2014). Epistemic injustice in healthcare: A philosophical analysis. Medicine, Health Care and Philosophy, 17, 529–540. http://doi.org/10.1007/s11019-014-9560-2

Knaak, S., Mantler, E. D., & Szeto, A. (2017). Mental illness-related stigma in healthcare: Barriers to access and care and evidence-based solutions. Healthcare Management Forum, 30(2), 111–116. http://doi.org/10.1177/0840470416679413

Krahé, B., Uhlmann, A., & Herzberg, M. (2021). The voice gives it away. Social Psychology, 52(2), 101–113. http://doi.org/10.1027/1864-9335/a000441

Labov, W. (2012). Dialect diversity in America: The politics of language change. University of Virginia Press.

Lackey, J. (2021). False confessions and subverted agency. Royal Institute of Philosophy Supplement, 89, 11–35.

Levy, N. (2022). Do your own research! Synthese, 200(5), 1–19. http://doi.org/10.1007/s11229-022-03793-w

McCluney, C. L., Robotham, K., Lee, S., Smith, R., & Durkee, M. (2019, November 15). The costs of code-switching. Harvard Business Review. https://hbr.org/2019/11/the-costs-of-codeswitching (accessed June 9, 2025)

McCluney, C. L., Durkee, M. I., Smith II, R. E., Robotham, K. J., & Lee, S. S. L. (2021). To be, or not to be… Black: The effects of racial codeswitching on perceived professionalism in the workplace. Journal of Experimental Social Psychology, 97, 1–12. http://doi.org/10.1016/j.jesp.2021.104199.

Mills, C. W. (1997). The racial contract. Cornell University Press.

Montmarquet, J. A. (1993). Epistemic virtue and doxastic responsibility. Rowman & Littlefield.

Roberts, L. M., Settles, I. H., & Jellison, W. A. (2008). Predicting the strategic identity management of gender and race. Identity: An International Journal of Theory and Research, 8(4), 269–306. http://doi.org/10.1080/15283480802365270

Morton, A. (2012). Bounded thinking: Intellectual virtues for limited agents. Oxford University Press.

Morton, J. M. (2014). Cultural code-switching: Straddling the achievement gap. Journal of Political Philosophy, 22(3), 259–281. http://doi.org/10.1111/jopp.12019

Nguyen, C. T. (2020). Echo chambers and epistemic bubbles. Episteme, 17(2), 141–161. http://doi.org/10.1017/epi.2018.32

O’Connor, J. J., & Barclay, P. (2017). The influence of voice pitch on perceptions of trustworthiness across social contexts. Evolution and human behavior, 38(4), 506–512. http://doi.org/10.1016/j.evolhumbehav.2017.03.001

Oppenheimer, D. M. (2006). Consequences of erudite vernacular utilized irrespective of necessity: Problems with using long words needlessly. Applied Cognitive Psychology, 20(2), 139–156. http://doi.org/10.1002/acp.1178

Pateman, C. (1988). The sexual contract. Stanford University Press.

Penzenstadler, N., & Quintana, C. (2023, January 20). George Santos’ college education is a myth. Is he the only one lying? We checked. USA Today. https://www.usatoday.com/story/news/investigations/2023/01/20/george-santos-congress-members-degrees-education/11034419002/ (accessed June 9, 2025).

Roberts, L. M., Cha, S. E., & Kim, S. S. (2014). Strategies for managing impressions of racial identity in the workplace. Cultural Diversity and Ethnic Minority Psychology, 20(4), 529–540. http://doi.org/10.1037/a0037238

Roulin, N., Bangerter, A., & Levashina, J. (2015). Honest and deceptive impression management in the employment interview: Can it be detected and how does it impact evaluations? Personnel Psychology, 68(2), 395–444. http://doi.org/10.1111/peps.12079

Scully, J. L. (2018). From “She would say that, wouldn’t she?” to “Does she take sugar?” Epistemic injustice and disability. International Journal of Feminist Approaches to Bioethics, 11(1), 106–124. http://doi.org/10.3138/ijfab.11.1.106

Stalnaker, R. (2002). Common ground. Linguistics and Philosophy 25(5), 701–721.

Steele, C. (2011). Whistling Vivaldi: How stereotypes affect us and what we can do. Norton.

Sternberg, R. (1987). Most vocabulary is learned from context. In M. G. McKeown & M. E. Curtis (Eds.), The Nature of Vocabulary Acquisition (pp. 89–105). Lawrence Erlbaum Associates.

Stone, A. (2014). An avowal of prior scepticism enhances the credibility of an account of a paranormal event. Journal of Language and Social Psychology, 33(3), 260–281. http://doi.org/10.1177/0261927X13512115

Taylor, C. (1992). The politics of recognition. In C. Taylor, Multiculturalism and “The politics of recognition”. Princeton University Press.

Tosi, J., & Warmke, B. (2016). Moral grandstanding. Philosophy & Public Affairs, 44(3), 197–217. http://doi.org/10.1111/papa.12075

Tosi, J., & Warmke, B. (2020). Grandstanding: The use and abuse of moral talk. Oxford University Press.

Warman, J. (2021). Reflections on intellectual grandstanding. Southwest Philosophy Review, 37(1), 209–217. http://doi.org/10.5840/swphilreview202137122

Westra, E. (2021). Virtue signaling and moral progress. Philosophy & Public Affairs, 49(2), 156–178. http://doi.org/10.1111/papa.12187

Young, V. A., & Barrett, R. (2018). Other people’s English: Code-meshing, code-switching, and African American literacy. Parlor Press.

Zagzebski, L. T. (1996). Virtues of the mind: An inquiry into the nature of virtue and the ethical foundations of knowledge. Cambridge University Press.