1 Introduction

An argumentum ad verecundiam is an appeal to authority or expertise that can constitute a reasonable argument, but is often applied fallaciously. Douglas Walton and Hans V. Hansen develop similar well-known accounts of this concept. Walton elucidates ad verecundiam in several critical argumentation volumes (1997, 2006, 2008), which shed light on how to distinguish a proper authority from an illegitimate one. He describes a legitimate authority as someone who has expertise within a domain of knowledge relative to someone else, whereas an illegitimate authority merely commands administrative office or influence. He then lays out more specific criteria for evaluating whether an appeal to authority is fallacious, including a list of six critical questions, followed by several ancillary sub-questions and additional guidelines. Hansen (2006) contributes to this discourse by noting that there are two additional sources of bias which can undermine the credibility of a putative authority: self-interest and vanity. However, he believes that these biases can be covered by Walton’s six critical questions, which inquire into the subject’s reliability. Thus, he endorses Walton’s theory without modification. I mention these theorists because they are representative of the standard view in informal logic.

It is notable that neither scholar explicitly addresses another source of bias that is known to have a significant and pervasive impact on judgments of expertise: namely, identity prejudice, such as sexism, racism, ablism, agism, and sanism. These sources can precipitate a false assumption of ad verecundiam due to distrust of a putative authority. For the purposes of clarity and concision, I will restrict my focus to gender bias in this paper; however, my arguments are generally applicable to other forms of discrimination, though they may not be exhaustive. That is, while they highlight core features of identity prejudice which have relevance for other discourses, they are not meant to exhaust the possibilities for critical analysis of prejudice for every oppressed group. Nonetheless, they should be instructive across several domains of social epistemology and identity politics.

Within feminist theory, Miranda Fricker (2007), Helen Longino (2002), Kathy Mack (1993), and Dominic A. Infante (1985) have written at length on how gender prejudice can undermine the amount of credibility accorded to speakers. Fricker identifies two sources of deflated credibility attribution—testimonial injustice and hermeneutical injustice—which harm and objectify vulnerable groups; Longino explains how knowledge-producing communities can be epistemically irresponsible, thereby silencing alternative points of view; and Mack and Infante illustrate how gender-based exclusions operate to silence women in modern society, underscoring the urgent need for the inclusion of gender-based criteria in the standard model for assessing expertise. Drawing on this scholarship, I argue that in order for Walton’s view to be theoretically adequate and epistemologically responsible, we must add at least three explicitly gendered criteria to his six critical questions. These criteria differ from Walton’s in that they move beyond analysis of the speaker as an autonomous agent, to question the epistemic justice of the process of judging of the speaker, and the epistemic responsibility of the speaker’s environment. Thus, I refer to these criteria as procedural and contextual. The critical questions that I intent to implement, drawing from Fricker and Longino, are: (1) Is one’s judgment of the speaker’s expertise affected by testimonial injustice? (2) Is one’s judgment of the speaker’s expertise affected by hermeneutical injustice? (3) Does the knowledge community under consideration satisfy the following four criteria of epistemic adequacy: (i) there are publicly recognized forums for criticism and dissent, (ii) there is uptake of criticism, (iii) there are public standards, and (iv) there is tempered equality. The first two questions are extrapolated from Fricker, and the third from Longino. Finally, to illustrate this account, I apply these criteria to the examples of Hillary Clinton and Sarah Palin, who were beset by dismissive and deflationary portrayals during the 2008 US presidential campaign, due not only to their political mistakes, but often on the basis of gender alone. This will show how we can use these criteria to identify false dismissals of ad verecundiam, stemming from tacit prejudice and entrenched stereotypes.

Before proceeding, some preliminary clarifications are in order. In Walton’s section on “appeals to authority” (2008), he discusses argumentum ad verecundiam and appeal to authority [or appeal to “expert opinion” (p. 208) or “expert authority” (p. 219)] simultaneously, and seemingly interchangeably. This is because he conceives of argumentum ad verecundiam as a species of appeal to authority/expert opinion/expert authority, which can be legitimate or fallacious depending on a number of interrelated factors. This type of appeal is contrasted against “the appeal to external or objective knowledge, which comes from scientific evidence such as experimental observations” or “nature,” in that it comes from “a personal source” who is amenable to critical scrutiny and dispute (p. 209). Walton says that appeals to authority are erroneous when they are “misinterpreted, taken too seriously, or not taken uncritically” (2008, p. 211); but I am concerned with instances where they are not taken seriously enough, and an errouneous dismissal is applied to the appeal. This is an important constraint, because if we are concerned with erroneous appeals to authority, then we should also be concerned with erroneous dismissals of such appeals (and such authorities), if we are to achieve theoretical precision and the practical ambitions that will be discussed in the next section (i.e. making informal logic relevant to daily life). Now, because Walton essentially conflates ad verecundiam with appeal to authority/expertise/expert authority without problematic results, the same identification will be replicated in this paper. There may be some confusion, however, about the scope of this appeal, specifically pertaining to whether it is meant to apply to (a) the authority, (b) the authority’s speech, or (c) the authority’s standpoint. (This is an issue that has been raised to me in professional correspondence). I can only answer this question indirectly, by saying that the appeal to authority and the legitimacy of the authority in question are necessarily imbricated, insofar as the appeal must be judged by reference to the authority’s status and the authority’s claim. Secondly, while the authority, as opposed to the authority’s claim, is the primary target of ad verecundiam, in keeping with Walton’s stipulation that ad verecundiam applies to a “personal source,” yet these two conditions are also intimately connected, since the content of a person’s claim can either confirm or impugn her expertise, depending on whether or not it is believable. (If someone makes a ludicrous claim, this calls into question her authority). Thus, these factors are impossible to completely disentangle. To see this more clearly, consider an example offered by Walton (2008):

Let us say that White backs up his argument by claiming that an expert, Green, has vouched for the proposition that White is maintaining. This move in the persuasion dialogue (between Black and White) has been advanced by White with the objective of persuading Black. (p. 216).

Here, the ad verecudiam applies to White’s invocation of Green’s authority, but in order to evaluate this appeal, we need to consider a range of personal and impersonal factors, such as the content of Green’s claim, whether “it is consistent with what other experts assert” (Walton, p. 218), whether it is “based on evidence” (p. 218), as well as Green’s “professional qualifications,” “experience in the field” (p. 219), “reliab[ility] as a source” (p. 219), and so on and so forth. Thus, we find that various factors are relevant to assessing an appeal to authority. My specific concern is erroneous dismissals of authority, which, like erroneous assumptions of authority, bear on the assessment of argumentum ad verecundiam, and depend on the same sorts of evaluative factors. To see this, consider the Black–White–Green example again, but imagine that Black erroneously discounts White’s appeal to authority because Green is a woman, and Black distrusts women’s testimony, either implicitly or explicitly. (This bias may be epistemically opaque to Black, in which case it is implicit). While the ad verecundiam in this scenario describes White’s appeal to Green’s authority, evaluating the legitimacy of this appeal requires consideration of numerous factors, from Green’s professional status, to the content of Green’s claim, to its relation to other claims, to the relevant domain of knowledge, and so on. The focal point of this analysis, as with Walton’s treatment, is the judgment of the legitimacy of an argumentum ad verecundiam/appeal to authority/expert opinion/expert authority. This judgment depends upon the legitimacy of (a) the authority, (b) the authority’s speech, and (c) the authority’s standpoint, and numerous other factors besides. This multifactorial approach coheres with Walton’s evaluative method, and thus it should not hinder the present analysis, any more than it did the original. Moreover, as informal logicians, we must not enact the mistake highlighted by Johnson (which will be discussed shortly), of getting bogged down in technical details at the expense of making practical gains and addressing issues of concern to real people.

2 Motivating Factors

Before explaining Walton’s theory, I should say something about why gendered criteria for evaluating false interpretations of argumentum ad verecundiam are eminently needed. Accordingly, in this section I highlight the general need for pragmatic criteria for the purpose of applying informal logic to the concerns of daily life, and I address the specific adverse effects of gender prejudice on women, in terms of dismissals of their credibility, trustworthiness, leadership ability, argumentation skills, and professional competence.

Advocating for a more pragmatic approach to logic, Ralph H. Johnson (1987) argues that the “standard treatment” of fallacies, which defines a fallacy as an argument that appears valid but is not, “uncritically assumes that vantage point of formal, deductive logic” (p. 242). Among other problems, this account is too divorced from “real-life settings” (p. 250) and the business of daily living to be of much use.Footnote 1 This point is driven home by Howard Kahane (1971), who recounts a striking exchange with his former student:

In class a few years back, while I was going over the (to me) fascinating intricacies of the predicate logic quantifier rules, a student asked in disgust how anything he’d learned all semester long had any bearing whatever on President Johnson’s decision to escalate again in Vietnam. I mumbled something about bad logic on Johnson’s part, and then stated that Introduction to Logic was not that kind of course. His reply was to ask what courses did take up such matters, and I had to admit that so far as I knew none did. He wanted what most students today want, a course relevant to everyday reasoning, a course relevant to the arguments they hear and read about race, pollution, poverty, sex, atomic warfare, the population explosion, and all the other problems faced by the human race in the second half of the twentieth century. (p. 7)

If Kahane is right and logical theory should be eminently practical and geared toward the concerns of ordinary people, it follows that it should incorporate criteria for evaluating when a judgment of expertise reflects a lack of appropriate sensitivity to a speaker’s credibility. Thus, it should incorporate conditions for evaluating when such a judgment is affected by testimonial or hermeneutical injustice, and when a knowledge-producing community lacks epistemic responsibility resulting in the marginalization of dissenting voices.

If it not immediately obvious that these types of exclusions occur frequently enough to warrant modification to the standard ad verecundiam formula, one does not have to look far for evidence of such prejudice. Because this evidence is staggering, I will concentrate on two particularly apposite sources. The first is Mack, who illustrates how deflated credibility ascriptions prejudicially affect women who testify in court. This is a salient focal point because one of the main driving forces in scholarship on appeals to authority is the question of how to evaluate expert testimony in trials. This is reflected in Walton’s emphasis on the role of expert testimony in courtroom disputes, where, he says, an alleged expert could be unreliable because he is paid to testify, selected in a partisan manner, or belongs to a “stable of house experts” who are recruited as witnesses because they are persuasive speakers (2008, p. 231). However, he does not discuss the opposite side of the coin: the tendency for the court to dismiss female witnesses due to sexism. The second set of theorists that I will highlight are Infante (1985) and Jordan-Jackson et al. (2008), who provide a more general account of how deflated credibility judgments undermine women’s ability to achieve their argumentative aims, based on a broad body of research.

Mack argues that women who testify about rape in court face deflated credibility judgments. When women testify in rape trials, courts in most American jurisdictions require third-party corroboration, or issue a warning to juries that uncorroborated testimony is not admissible (p. 344). This reflects a general presumption that women are not credible witnesses. Another sign of this bias is that people generally ascribe less authority to witnesses who use locutions common to women’s speech, including “superlatives, intensifiers (‘so’ or ‘such’), fillers (‘um’ or ‘you know’), empty adjectives, tag questions with rising intonation (even with an accurate assertion), hedges (‘sort of’), and politeness markers” (Mack, p. 330, citing research by Morrill and Facciola 1992). The same studies show that “women are more likely to speak hesitantly even if they are certain, while men are more likely to speak with assurance even if unsure or wrong,” and since confident speakers are seen as more credible, women are disadvantaged (p. 331). Further evidence shows that women face demeaning treatment in court, such as “patronizing language (‘little lady’) or other disrespectful forms of address (‘honey’); inappropriate comments on dress, marital status, or parental role; and sexual harassment” (Mack, p. 332, citing Md. report, Mich. report, Minn report, and Nev. report). These practices result in a general attitude of distrustfulness toward women testifying in rape trials, as well as low conviction rates and “lack of sympathy on the part of jurors” (p. 337).

Addressing the more general effects of gender bias, Infante conducts a study to explore whether women are perceived as more credible when they are induced to adopt a more argumentative style of communication, where argumentativeness is defined as a tendency “to recognize controversial issues in communication situations, to advocate positions on the issues, and to attempt refutation of the positions other people take” (pp. 33–34). He finds that more argumentative women are perceived as more credible, confirming his hypothesis. This is significant because women tend to be low in trait argumentativeness, implying that they are commensurately likely to be seen as less credible. Infante says that these results may be taken to suggest that “a concentrated effort should be made to motivate females to be more argumentative”; however, he cautions that “although greater argumentativeness may enhance credibility, that is not to say that it always is in the best interests of the individual to argue” (p. 43). It may, for example, disrupt interpersonal relationships which are predicated on traditional dynamics, or disadvantage women who have internalized these habits at the level of personal identity.

This research was more recently corroborated and expanded upon by Jordan-Jackson et al., in a comprehensive survey of current research on sex differences, perceived argumentativeness and verbal aggression, and how these factors influence relationship outcomes. Very briefly, the researchers found that women who deviate from their perceived gender roles are viewed negatively and often penalized (p. 242). Jordan-Jackson et al. contribute to existing research by using videotape analysis involving male and female dyads, to confirm that sex-role stereotyping often influences perceptions of argumentativeness and verbal aggression in ways that negatively affect women’s perceived credibility, expertise, trustworthiness, and professional competence. In particular, they find that “the female dyad was seen as less argumentative than the male dyad” (p. 252). This is significant because “a large body of research suggests that being argumentative is a constructive trait with numerous desirable characteristics such as greater credibility, leadership, critical thinking skills, and academic achievement (see for example, Infante and Rancer 1996; Rancer and Avtgis 2006)” (p. 252). Thus, because women tend be seen as less argumentative, they will also be seen as low in the corresponding positive traits.

In light of these findings, it could be argued that women should be encouraged to adopt traditionally male modes of argumentation. However, there is reason to reject this solution as both impractical and unacceptably chauvinistic, on grounds that, in Michael Gilbert’s words (1997), even if we deny gender essentialism, “most authors seem willing to accept that, for whatever reasons, there are identifiable gender differences in mode of communicating that are generalizable” (p. 50, emphasis in original). Thus, trying to manipulate women’s vernacular habits is not likely to be very productive; and insofar as this decision is morally arbitrary, it unfairly depreciates the value of women’s speech. As Gilbert puts it, given entrenched vernacular differences, trying to eliminate “female modes of reasoning… involves the unfair limitation of the power of one group by another” (p. 52). Therefore, rather than trying to modify women’s speech habits, we should, on ethical and pragmatic grounds, attempt to change the way that we ascribe credibility to them. One part of this project, I believe, is incorporating gendered criteria into the formula for evaluating appeals to authority, to draw our attention to our tacit biases.

3 Argumentum ad verecundiam and Walton’s Six Critical Questions

An argumentum ad verecundiam, as we saw earlier, is an appeal to authority which is often applied fallaciously. It was first defined by Locke in An Essay Concerning Human Understanding (1690), as an argument meant to “prevail on the understanding of another person,” to gain his assent and “silence his opposition” (Locke, pp. 159–160). Three centuries later, Walton defined it in his canonical Appeal to Expert Opinion (1997) as an appeal to an inappropriate sort of authority. Specifically, he characterized it as an appeal to “administrative authority” under the guise of an appeal to “cognitive authority,” where an administrative authority is someone with “a right to exercise command or influence… based on an invested office,” while a cognitive authority is someone capable of making pronouncements in his field “that carry a special weight of presumption… that is greater than the say-so of a lay-person in that field” (p. 321). In his 2008 text, Informal Logic: A Pragmatic Approach, Walton recapitulates this definition, describing argumentum ad verecundiam, following Locke, as “the misuse of an appeal to an authoritative source to try to prevail unfairly, or ‘silence the opposition’ in an argument” (p. 210), as well as the confounding of administrative and cognitive authority. To illustrate the fallacious usage, he gives the example of a famous comedian who recommends a brand of soft drink because it does not contain harmful additives and is good for one’s health. This spokesperson does not have expert credentials in health and nutrition, and is not even named, which renders the appeal flawed on two counts (2008, p. 225). There are also other conditions, which may bear on the evaluation of the person’s proper authority.

The argumentum ad verecundiam tends to be misused because it can be difficult to distinguish a proper authority from an improper one. This is why the argument is cited as a species of informal fallacy in logic textbooks: because it is frequently mislaid. Walton gives us instructions on how to identify a reasonable ad verecundiam as opposed to a faulty one. First, he says that a reasonable appeal should conform to the following structure:

E is an expert in domain D

E asserts that A is known to be true

A is within D

Therefore, A may plausibly be taken to be true (2008, p. 217)

If an appeal to expert opinion takes this form, then it “warrant[s] the drawing of a presumptive inference to the conclusion indicated” (Walton 2008, p. 217). That is, there is presumptive acceptance of A. However, as per Locke’s definition, a reasonable ad verecundiam is subject to scrutiny and refutation through critical questioning. An opponent can dispute the appeal by raising objections to the status of the expert in question. Thus, Walton offers a list of six critical questions for evaluating an appeal to authority, which must be asked to shift the burden of presumption to the other side. These are:

  1. 1.

    Expertise Question: How credible is E as an expert source?

  2. 2.

    Field Question: Is E an expert in the field that A is in?

  3. 3.

    Opinion Question: What did E assert that implies A?

  4. 4.

    Trustworthiness Question: Is E personally reliable as a source?

  5. 5.

    Consistency Question: Is A consistent with what other experts assert?

  6. 6.

    Backup Evidence Question: Is E’s assertion based on evidence? (2008, p. 218)

Walton then gives more specific directions on how to interpret these questions. For example, in deciding whether someone satisfies (1), we are urged to consider what degrees, professional qualifications, or certifications the person has, whether the person’s colleagues can attest to her expertise, whether she has a record of experience in the field, and whether she has a record of accomplishments and peer-reviewed publications (p. 219). There are additional guidelines, but the question of greatest concern to us—that of bias—surprisingly does not arise until the fourth critical question, where Walton offers three “critical sub-questions” that can be used to determine whether someone is a reliable source:

  1. (i)

    Is E biased?

  2. (ii)

    Is E honest?

  3. (iii)

    Is E conscientious? (p. 220)

Walton defines “bias” as “failure to represent both sides of an issue in a balanced way” (p. 220). The topic arises once more in a section on “insufficient and biased statistics” (p. 254), which deals with selecting representative samples. Strikingly, although one might expect to find identity prejudice mentioned in a discussion of expert opinion, it does not arise in this text, or in any of Walton’s other works. Although bias is discussed, it is not related to gender, identity politics, or epistemic responsibility. Nor does this subject arise in Hansen’s treatment, although he addresses two additional species of bias, self-interest and vanity. Although this is an improvement, is still misses an important source of bias.

Thus, it appears that reference to identity prejudice is altogether lacking in standard accounts of ad verecundiam. From a feminist perspective, this is an alarming oversight, given the well-documented effects of gender prejudice on judgments of credibility, trustworthiness, and expertise in women. In addition to the sources mentioned in the last section, Longino and Fricker explicate how gender prejudice can undermine the amount of credibility ascribed to a speaker based on gender stereotyping. If these theorists are right and gender is a relevant factor in the evaluation of speaker credibility, then Walton’s six critical questions cannot possibly exhaust the criteria required to evaluate whether an argumentum ad verecundiam is fallacious, since the expert’s credibility may be depreciated due to tacit prejudice. This suggests that we need gender-based criteria to help us identify when a person is devaluing a speaker’s expertise, and thus illegitimately ascribing a fallacious interpretation to an appeal to that person’s authority. This need is particularly pressing given that identity prejudice tends to be subconscious and epistemologically opaque, even to the evaluator him or herself. Because these gendered criteria apply to one’s assessment of a speaker’s authority, they are procedural, meaning that they apply to the process of judgment; and because they encompass the speaker’s hermeneutical context, they are contextual, meaning that they comprise factors beyond the individual speaker, within the broader community. These criteria will be explicated in the next section, and added to Walton’s questions as supplements.

4 Adding Three Gender-based Questions

To develop a set of critical questions that answers to the need for gender-based evaluation of argumentum ad verecundiam, I turn to Fricker’s account of epistemic justice and Longino’s theory of epistemic responsibility. This will help us identify the standard ways in which identity prejudice affects credibility judgments, which will in turn help us develop precise criteria for identifying prejudiced dismissals and faulty interpretations of ad verecundiam which rest on deflated credibility ratings.

In her pioneering work in virtue epistemology, Fricker explains that there are distinctively epistemic forms of injustice that can undermine the amount of credibility that we attribute to others. She identifies two types of epistemic injustice. The first is testimonial injustice, in which the hearer gives the speaker less credibility than she deserves due to prejudice. The paradigm case of testimonial injustice is identity prejudice, such as when a police officer discounts someone’s assertion because he is African American (Fricker, p. 4). A secondary type of epistemic injustice is hermeneutical injustice, in which a speaker’s credibility is undermined because she is unable to comprehend or communicate her experience to others due a lack of interpretive (hermeneutical) resources within her community. The central case of hermeneutical injustice is when a woman’s experience of sexual harassment is not acknowledged (by others, or perhaps herself) because the community lacks the resources required to understand sexual harassment as a kind of harm. As a result, the victim may not fully comprehend her own experience of objectification, much less be able to communicate it effectively to others (Fricker, p. 6). Nonetheless, she may have a strong intuitive impression that she has been harmed, and she may suffer substantive social disadvantages. Notably, these criteria direct attention away from the agent toward the agent’s circumstances, which may prevent or prohibit uptake and appreciation of her testimony.

Another relevant consideration, which also diverges from the agent-focused model of evaluation, is whether the speaker’s domain of expertise is an epistemically responsible one. In her pathbreaking work in social epistemology, Longino notes that domains of knowledge are not hermetically sealed contexts of rational inquiry. Rather, they are social communities which include and operate on implicit normative assumptions. Since these communities are inherently social and interpersonal, they should be subjected to evaluation in normative terms. Longino proposes four criteria for determining whether a knowledge-producing community is epistemically responsible. They are:

  1. 1.

    There must be publicly recognized forums for criticism and dissent,

  2. 2.

    There must be uptake of criticism,

  3. 3.

    There must be public standards, and

  4. 4.

    There must be tempered equality, meaning that minority members of the community must be treated as equal participants.

These criteria are crucial to the evaluation of expert authority because Walton’s critical questions directly cite “domains of knowledge,” but they do not include criteria for evaluating these domains’ degree of epistemic responsibility. In Informal Logic: A Pragmatic Approach, Walton cites “the domain of knowledge” (D) three times: He says that we must determine if “E is an expert in domain D,” if “A is within D,” and if “E is an expert in the field [or domain] that A is in” (pp. 217–218). He also indirectly invokes the domain of knowledge when he asks if “A is consistent with what other experts think” (2p. 218), since the consensus of other experts is necessarily relative to the domain in question. Walton’s questions are undoubtedly necessary for determining whether an authority is credible, but they are not sufficient, since the authority’s domain of knowledge may be infected by gender prejudice. This is the case, according to Longino, when someone is an expert in a male-dominated field, but there are no forums for criticism from women’s voices within that field, or these voices do not receive proper uptake. Adding Longino’s criteria will allow us to perform a more thorough investigation of a person’s expertise in light of her field or domain of knowledge. It will thus allow us to evaluate whether an appeal to her authority is illegitimately dismissed due to systemic prejudice.

These considerations prompt the addition of three new critical questions to Walton’s original list of six:

  1. 1.

    Is the judgment of expertise affected by testimonial injustice?

  2. 2.

    Is the judgment of expertise affected by hermeneutical injustice?

  3. 3.

    Does the knowledge community in which the appeal to authority is made satisfy the following four criteria of epistemic adequacy: (i) there are publicly recognized forums for criticism and dissent, (ii) there is uptake of criticism, (iii) there are public standards, and (iv) there is tempered equality.

I propose these criteria as an addendum to Walton’s six questions. They compel us to inquire into whether a judgment of ad verecundiam is epistemically just and responsible. Returning to the Black-White-Green example, if Black discounts White’s appeal to Green’s authority because he distrusts women’s testimony, he is failing to satisfy (1) the condition of testimonial justice. However, even if Black is being as epistemically responsible as possible in the circumstances, yet there is a gap in the hermeneutical resources required to identify and appreciate Green’s authority, there is still a failure of (2) the condition of hermeneutical justice. Moreover, if (2) is satisfied, but there are no forums for criticism or dissent in Green’s knowledge-producing community, then there is a failure of (3.2). From an epistemically enlightened perspective, we can recognize and cite these failures as grounds for shifting the weight of presumption onto Black. This brief sketch illustrates how the new criteria allow us to identify misjudgments of ad verecundiam, and shift the burden of proof onto the epistemically ignorant party or parties. In the next section, I provide a more in-depth analysis of how these criteria can be put into use in real life.

5 Applying the Criteria in the World

To illustrate the application of the new criteria, I will use the examples of Hillary Clinton and Sarah Palin, who were subjected to sexist depictions during the 2008 presidential election, and arguably continue to suffer gender-based discrimination in present reporting. I will also mention Native Canadian author Emma LaRoque—who is cited in Sue Campbell’s (1994) account of how sentimental reactions often fail to accrue social uptake in mainstream society—as a means of introducing the implementation of these criteria in a more parochial case with which many can relate, as opposed to high political office. To begin, I will explicate Campbell’s theory and LaRoque’s experience of discrimination, and how the new criteria can illuminate and block false dismissals of authority.

In her work on the politics of emotional expression, Campbell highlights the ways in which bitterness, sentimentality, and emotionality can be dismissed and silenced, especially when expressed by members of historically oppressed groups. She writes that “the association of the feminine with a feeling has been a long-standing historical ground on which to dismiss women” (p. 49), and thus sentimental reactions often result in flippant attitudes and a failure of social uptake. The term “social uptake,” first coined by Marylin Frye in A note on anger (1983), refers to the propensity of a listener to receive a speaker’s testimony. When social uptake fails, not only does the speaker lose public credibility, but she may even be prevented from understanding her own experience of oppression due to a dearth or unequal distribution of interpretive resources. According to Campbell, “if someone consistently fails to secure uptake for the feelings that get formed only through acts of expression, it cannot be clear even to that person what she or he is feeling, and many people’s emotional lives are, in fact, dominated by a confusion that is an inevitable consequence of persistent lack of uptake” (p. 55, emphasis mine). This, of course, mirrors Fricker’s observation that hermeneuical injustice can render a speaker’s experience reflexively opaque and incommunicable. This analysis shifts the focus from the individual subject onto the subject’s audience and social context. When a person’s legitimate expression of resentment is stifled, an appeal to her authority may be unfairly discounted.

This phenomenon is exemplified in Campbell’s description of Native author Emma LaRoque, who articulates her experience of systemic marginalization by a racist audience:

The interplay between audience reception and publishing cannot be minimized. As one of those earlier Native writers, I experienced and studied what might be called the Native- voice/white-audience dynamic. The interactions were often poignant. On another level, we were again rendered voiceless no matter how articulate we were. Apparently, unable to understand or accept the truth of our experience and perceptions, many white audiences, journalists, and critics resorted to racist techniques of psychologically labelling and blaming us. We were psychologized as “bitter,” which was equated with emotional incapacitation, and once thus dismissed we did not have to be taken seriously. (LaRoque 1990, xvi–xvii)

LaRoque’s audience presumed that her expression of anger was merely an emotional effluence, and failed to acknowledge her legitimate articulation of political dissent through reasonable anger. Now, let us see how the expanded criteria can be employed to parse this case of epistemic injustice. First, it is clear that LaRoque was afflicted by (1) testimonial injustice, since her audience dismissed her testimony on racist grounds. Secondly, she was somewhat affected by (2) hermeneutical injustice, insofar as “she began writing,” in Campbell’s words, “before the days when what Natives had to say about their own lives secured any uptake” (p. 54). This implies that the hermeneuical resources needed to comprehend Native oppression were partially lacking or inequitable distributed, rending her testimony partially unintelligible at the time. Third, although there were (3.1) publicly recognized forums for criticism and dissent, which explains how LaRoque was able to publish her anthology of Native Canadian women’s writings and critique of pervasive racism, there was not (3.4) sufficient uptake of criticism. Therefore, LaRoque’s audience fails to satisfy, at least, conditions (1), (2), and (3.4). Now let us imagine that someone has appealed to Laroque’s authority on Native oppression—as very well may have happened—and this person’s appeal was summarily dismissed due to distrust of LaRoque. On the present account, there would be grounds for denying this dismissal, and affirming the legitimacy of the appeal, since the dismissal failed on three counts.

Next, consider the more publicized examples of Clinton and Palin, who garnered substantial media attention and scholarly interest during the 2008 US presidential campaign. In their insightful critique of biased media coverage, Carlin and Winfrey (2009) note that female professionals often experience “what Kathleen Hall Jamieson (1995, p. 16) described as a double bind: ‘Women who are considered feminine will be judged incompetent, and women who are competent, unfeminine… those who succeed in politics and public life will be scrutinized under a different lens from that applied to successful men’” (p. 327). This applies to Clinton and Palin, who were depicted in the press as, respectively, unfeminine and incompetent. It should be noted at the outset that while both candidates deserve legitimate criticism on various grounds—particularly Palin, who lacked political insight and made egregious errors, such as ratifying the famous “bridge to nowhere”—both candidates also faced unfair discrimination on the basis of gender alone. These instances of epistemic injustice will be the focus of the present analysis. Performing this analysis is important because even if Palin’s testimony sometimes lacked rational warrant, she faced a type of discrimination that could, and often does, function to undermine women’s credibility even when their arguments are sound.

Following Rosabeth M. Kanter (1997), Carlin and Winfrey identify four standard stereotypes of professional women: seductress or sex object, mother, pet/child, and iron maiden (p. 327). They illustrate how these stereotypes were applied to Clinton and Palin during the 2008 campaign in ways that reduced their perceived credibility or “cause[d] them to seem less human” (p. 328). For our purposes, we can confine our attention to the second category, pet/child, which suffices to elucidate the discrediting effects of gender stereotyping. Carlin and Winfrey note that Senator John McCain used Palin to attract media attention, which fits Kanter’s (1997) “definition of a pet as a ‘cheerleader for shows of prowess’” (p. 336). This treatment presents Palin as “needing protection from the press” (p. 336). Objecting to McCain’s patronizing attitude, CNN’s Campbel Brown protests,

Tonight, I call on the McCain campaign to stop treating Sarah Palin like she is a delicate flower that will wilt at any moment. This woman is from Alaska, for crying out loud. She is strong. She is tough. She is confident. And you claim she’s ready to be one heartbeat away from the presidency. If that is the case, then end this chauvinistic treatment of her now. Allow her to show her stuff. Allow her to face down those pesky reporters, just like Barack Obama did today, just like John McCain did today, just like Joe Biden has done on numerous occasions. Let her have a real news conference with real questions. (Brown 2008, para. 12, cited in Carlin and Winfrey, p. 336)

While Clinton was accorded more respect than Palin, she was still subjected to pet/child stereotyping which depicted her as weaker than, and dependent upon, her husband, Bill Clinton. Fox News (2007) credited the former president with being willing to take “on the role of a spokesperson who is better able to explain [Hillary’s] positions on hot issues like Iraq”; and Patrick Healy (2007) notes that Bill was often depicted in the press as “the master strategist behind the scenes; the consigliore to the head of ‘the family’” (Carlin and Winfrey, pp. 336-37). Meanwhile, Bill’s attempts to help Hillary often reinforced sexist stereotypes, such as when he lamented to the press that he could not “make her younger, taller or change her gender’’ (Dowd 2008, cited in Carlin and Winfrey, p. 337). Finally, it is notable that when Hillary showed emotion in a debate with John Edwards and Barack Obama, “something male candidates (think McCain) do every debate,” she “was promptly accused of having a ‘meltdown’” (Watson 2008, cited in Carlin and Winfrey, p. 337). In a similar vein, Edwards and McDonald (2010) report that on the eve of the New Hampshire primary, Clinton’s voice “cracked with emotion,” but there was no evidence of actual tears (p. 324). However, the media harped on this emotional display, with some commentators construing it as a political machination. For instance, a cartoon by Mike Ramirez depicted Clinton with a crocodile’s head, saying, “I don’t cry for me, I cry for the country. Uh, are they still taping?” This harkens back to the double-bind that prevents women from being taken as seriously as men in political discourse.

Now let us assess these media depictions of Palin and Clinton using our new criteria for judging the legitimacy of a judgment of argumentum ad verecundiam. First, we should note that Walton’s criteria may fail to identify these politicians’ expertise, if questions (1) and (4), which gauge the speaker’s credibility and trustworthiness respectively, are applied uncritically. This was certainly done when journalists depicted Palin and Clinton as childishly dependent on men, and emotionally effusive and/or manipulative. If Edwards and McDonald are right, for instance, Ramirez may have failed to credit Clinton’s authentic expression of sensitivity due to a sexist perception of her as insufficiently feminine. This indicates, using our new criteria, that there was a failure of (1) testimonial justice in public discourse, insofar as commentators did not give Clinton the credence she deserved. The same can be said for Palin, inasmuch as she was portrayed as dependent on McCain. Although Palin generally lacked political acumen, when she did make reasonable arguments—such as when she promoted feminism and equal rights in an interview with Katie Couric—her assertions may not have received the uptake that they deserved. In terms of (2) hermeneutical injustice, there is evidence that interpretive resources were lacking in public discourse for interpreting women’s testimony. This is evinced both in the media’s frequent deployment of the four common stereotypes to both women, which indicates a disparity in hermeneutical possibilities for women compared to men, and in the politicians’ own denial of the importance of sexism. Significantly, both women explicitly denied that sexism was an issue in their political campaigns. Palin, for instance, stated that

When I hear a statement… coming from a woman candidate, with any kind of perceived whine about that excess criticism or maybe a sharper microscope put on her—I think that doesn’t do us any good, women in politics, or women in general wanting to progress this country…. [F]air or unfair, it is there. I think that’s reality, and I think it’s a given…work harder, prove yourself to an even greater degree that you’re capable, that you’re going to be the best candidate. (“Palin on How,” 2008). (Carlin and Winfrey, p. 327).

In similar fashion, Clinton insisted on being considered a “candidate who happened to be a woman” (Carlin and Winfrey, p. 328), as opposed to a woman candidate, thereby diminishing the role of sexism on her perceived credibility. While this denial may be, in part, a political subterfuge to finesse anti-feminine voters, it is fair to think that Clinton and Palin were unaware of the extent to which sexism distorted public opinion of their authority and political suitability. This lack of comprehension reflects the kind of hermeneutical injustice that Fricker and Frey identify as undermining women’s ability to grasp and articulate their experiences of oppression and demand fair recognition of their talents.

On the third question inspired by Longino, we find that there was a lack of (3.2) social uptake, insofar as the media indulged in the four stereotypes outlined above, and a lack of (3.4) tempered equality, insofar as the diversity of perspectives required to identify and appreciate this stereotypical treatment was absent or obfuscated. Thus, there was a failure of at least (1), (2), (3.2), and (3.4) in reporting on Palin’s and Clinton’s political runs.

To see how these failures affect the evaluation of argumentum ad verecundiam, imagine that I am appealing to Palin as an expert on Alaska’s taxation system (which was the lynchpin of her successful run for mayor of Wassilla, Alaska in 1996), or to Clinton on US foreign policy (on which she is an expert as America’s Secretary of State), and that my appeal is dismissed by an interlocutor due to child/pet stereotyping, false presumptions about women’s emotional lability, or some other biased attitude. Using the new criteria, we can identify this false dismissal of expertise as illegitimate, and successfully shift the weight of presumption onto the person making it. The new criteria give us explicit language for identifying the nature of the dissenter's mistake. This helps us to evaluate reasonable appeals to authority, and properly gauge which side bears the burden of proof.

6 Concluding Remarks

Walton offers six critical questions for evaluating when an appeal to authority should be accepted as legitimate. I have offered three additional questions for evaluating when an appeal to authority has been illegitimately dismissed due to the operation of epistemic injustice or epistemic irresponsibility on the part of a judge or community. These questions place an important constraint on the evaluation of argumentum ad verecundiam because they draw our attention to procedural and contextual epistemic factors that can distort judgments of expert authority, and thus judgments of appeals to expert authority. They can help shift the burden of presumption onto the epistemically flawed party or parties. These criteria are at least as useful in highlighting systemic prejudice (the goal of feminist epistemology) as in evaluating legitimate appeals to expertise (the goal of informal logic). By combining these two aims, my analysis helps to fulfill Johnson’s stated goal of bringing informal logic into line with the ethical concerns of ordinary people.