Just How Testimonial , Epistemic , Or Correctable Is Testimonial Injustice ?

In her book Epistemic Injustice: Power & the Ethics of Knowing, Miranda Frickerargues that there is a distinctly epistemic kind of injustice, which she calls testimonial injustice, resulting from identity-prejudicial credibility deficit – identity prejudic causing a hearer to give a deflated level of credibility to a speaker’s word. Sheargues that testimonial injustice is correctable through hearers developing a capacity for self-correcting for it within a virtue-epistemological framework. In this essay, I examine Fricker’s argument, and conclude that Fricker’s exposition of identityprejudicial credibility deficit is sound, but that she has not demonstrated either that testimonial injustice is a distinctly testimonial phenomenon or that identityprejudicial credibility deficit is a distinctly epistemic phenomenon, or how a virtueepistemic framework might be successfully employed as a corrective methodology. I further suggest that her conception of testimony as distinctly evidential is too narrow to be applied to everyday instances of identity-prejudicial credibility deficit, and should be reconfigured within a broader framework.


In her book Epistemic Injustice
, Miranda Fricker argues that there is a specifically epistemic kind of injustice, which consists in wronging somebody in their capacity as a knower, and that one form of epistemic injustice is testimonial injustice, which occurs 'when prejudice causes a hearer to give a deflated level of credibility to a speaker's word' (Fricker 2007, 1). 1 She refers to this prejudice as 'identity prejudice' because it is based on the identity of the speaker as a 'social type' -someone socially stereotyped by the hearer, for example by virtue of their race, gender or ethnicity. Fricker thus defines such instances of testimonial injustice as instances of 'identity-prejudicial credibility deficit'. She distinguishes them from instances of what she calls 'epistemic bad luck' -cases in which the hearer reasonably misjudges the speaker's words or body language, such as by interpreting his avoiding eye contact with her as a sign of his insincerity, when it is actually a sign of the speaker's shyness. Both cases, she says, result in the credibility of the speaker being deflated by the hearer, but in the case of testimonial injustice the speaker is also epistemically 'wronged' by the hearer (41)(42).
It is possible, says Fricker, for a speaker to be given excess, as well as deflated epistemic credit as a result of identity prejudice which carries an epistemic 'social charge'. An example might be a speaker having an accent which implies she has a 'creditworthy' educational or social background (though the 'wrong' kind of accent might equally cause the speaker's credibility to be reduced in the mind of the hearer). Fricker states that she is concerned only with cases of identity-prejudicial credibility deficit, but I will argue that this is not in fact the case.
Testimonial injustice, says Fricker, is especially harmful where the identity prejudice that informs it amounts to social stereotyping, which she defines as: a widely held disparaging association between a social group and one or more attributes [. . .] [which] displays some (typically epistemically culpable) resistance to counter-evidence owing to an ethically bad affective investment (35).
Such widely accepted social prejudices, Fricker declares, are antithetical to the dissemination of truths, freedom of speech and the spread of knowledge and also constitute ethical wrongs against speakers whose credibility is deflated, because they are degraded not only qua knowers, but also qua human beings. I would argue, though, that it is more likely that it is socially stereotyping speakers as humans that leads to their credibility as knowers being deflated.
Fricker notes that a secondary effect of negative identity-prejudicial stereotyping may be that speakers whose credibility is deflated by others may suffer a loss of confidence in themselves as knowers, or in their authority as speakers, and become too quick to give up their own beliefs when challenged, or accept too uncritically the opinions of others. More radically, they may begin to see themselves as others see them, or seek to conform to the stereotypical image imposed on them by others. One example of such pernicious self-stereotyping is that recounted by the Canadian philosopher Karyn L. Freedman, who was violently raped as a young woman and subsequently wrote an account of her experience and its aftermath in her book, One Hour In Paris. Freedman recalled how she did not relate her experience, other than to her closest family and friends, until almost a decade after the event, largely because she had assimilated the 'dominant world view' that the world was only unsafe for women who were not 'sufficiently careful or intelligent'. Thus, she feared disclosure might lead to her being judged to have contributed to, or even collaborated in, the assault on her. (Freedman 2014) Probably the most notorious historical example of collective 'self-fulfilling' racial stereotyping is that of German Jews who were bombarded with viciously degrading and dehumanizing images of themselves by the Nazi media during the 1930s. The 'surrender' of Jews in Germany and elsewhere to this stereotypical image of themselves, said the Austrian philosopher, essayist, and Holocaust survivor Jean Améry, was no moral failure, but simply the acknowledgment of a social reality. 'This, so they must have told themselves' he said, 'is how the world sees us, as lazy, ugly, useless and evil; in view of such universal agreement what sense does it still make to object and say that we are not that way!' (Améry 1999, 87). 2 Thus, I am in agreement with Fricker's contention that social stereotyping exists, or even that it is widespread, but this does not of course imply that it necessarily results in credibility deflation.
Correcting for testimonial injustice, says Fricker, means an epistemic agent training herself to be a 'responsible' hearer -to 'match the level of credibility she attributes to her interlocutor to the evidence that he is offering the truth' (Fricker 2007, 19, my emphasis).
Fricker's illustration of testimonial injustice in practice revolves around two fictional cases. The first is contained in Harper Lee's much acclaimed 1960 novel To Kill a Mockingbird, set in a courtroom in Maycomb County, Alabama, in 1935, which Fricker describes as a work of 'historically truthful fiction'. A black man, Tom Robinson, is on trial charged with raping and beating a white girl, Mayella Ewell. 'It is obvious to the reader', says Fricker, 'and to any relatively unprejudiced person in the courtroom, that Tom Robinson is entirely innocent' (23). There is compelling evidence, for instance, that Tom would have been physically incapable of beating the girl. However, any rational evaluation of the evidence is rendered impossible by the all-white jury's prejudice against Tom as a 'Negro'. They cannot believe his assertion, for example, that it was Mayella who had tried to kiss him, because the possibility that a white girl may have tried to kiss a black man -as was in fact the case -was simply unthinkable. Faced with the epistemic reality, within the larger social reality of racial prejudice, that whether or not he tells the truth he will not be believed and will be found guilty, Tom runs away, thus 'proving' his guilt in the eyes of his white accusers. The jury's prejudiced perception of Tom, notes Fricker, pervades many aspects of their reaction to his evidence. For example, his truthful assertion that he had helped out Mayella to do odd jobs around the house because he felt sorry for her transgresses their innate belief in white superiority, since the notion that a black person might feel sympathy for a white person, and thus by implication feel a sense of superiority towards her, appears to them incredible, and confirms them in their view that he is a liar. The jury's duty, on the basis of the evidence, is clearly to acquit Tom, but their racial prejudice overpowers any sense of duty, and thus Tom becomes the victim of testimonial injustice.
Fricker's second illustration is also fictional -Anthony Minghella'a screenplay for The Talented Mr Ripley, based on Patricia Highsmith's novel and set in Venice in the 1950s. Herbert Greenleaf, a rich American industrialist, is visiting the city in order to investigate the whereabouts of his son, Dickie, who is engaged to Marge Sherwood, but has mysteriously disappeared after spending time travelling with his 'friend' Tom Ripley. Ripley encourages Greenleaf to believe Dickie has run away and possibly committed suicide, but Marge comes to believe, correctly, that Ripley has murdered him. Though when Marge expresses her distrust of Ripley and her disbelief of Ripley's story, Greenleaf, to whom 'Ripley has done a successful job of sucking up' silences her, mistakenly 'pathologizing' her assertion that Dickie would not kill himself to be 'a sweetheart's wishful thinking' (Fricker 2007, 87). He even ignores the key piece of evidence, that Marge has found Dickie's rings at Ripley's place, including one from her he had sworn never to remove, responding with a 'familiar put-down': 'Marge, there's female intuition and then there are the facts' (88). Her truthful assertions simply convince Greenleaf, as a result of his social stereotyping of her, of the rationality of his belief that Marge's account is untrustworthy. The viewer, however, knows that Marge is right. Ripley has successfully exploited Greenleaf's adherence to contemporary prejudicial attitudes towards women among men of his age and social background to make him disbelieve Marge, a fact to which Greenleaf himself is blind. This is therefore another case of testimonial injustice, but Fricker ultimately deems Greenleaf non-culpable in his prejudice, because his prejudicial stereotyping of Marge is so ingrained as to be unconscious. Tom's jurors, on the other hand, she finds culpable for not believing Tom due to their refusal to take into account any of the compelling evidence presented by his white legal counsel, Atticus Finch, that Tom was telling the truth. In fact, they deliver their verdict in defiance of, and perhaps even partly as a consequence of being provoked by, the pleas of Finch to set aside their prejudices about Negroes and 'do their duty' to judge Tom solely based on the evidence. Fricker next considers the possibility of correcting for identity-prejudicial credibility deficit. Correctability, she finds, requires the hearer to have both a sensibility towards the presence of identity-prejudicial influences affecting her evaluation of the speaker, and the will and ability to engage corrective epistemic virtues in order to try to neutralize them. One way to develop such virtues, Fricker suggests, is through familiarization. Epistemic agents who are prejudiced against a particular accent, for example, may be able, with regular exposure to hearing the testimony of speakers with that accent, to come to associate it with a speaker's trustworthiness (though of course such exposure could equally confirm their prejudicial view). The ideal outcome, she suggests, would be somehow to condition the hearer's sensibility, through experience, to be able to recognize the possible danger of his forming an identity-prejudicial evaluation of the speaker's credibility and to correct for it spontaneously, so that the hearer's evaluation emerges 'ready-corrected'. However, since such an ideal is unlikely to be realized in practice across a range of identity prejudices, Fricker concedes that corrective action may require a certain amount of reflective application, in tandem with a degree of spontaneity. She suggests that to the extent one thereby succeeds in correcting for the relevant identity-prejudicial credibility deficit, one has acquired the virtue of testimonial justice.

How Can One Account and Correct for Testimonial Injustice within an Epistemic Framework?
Propositional telling -someone telling someone else that p -says Fricker, 'is the parent case of testimonial injustice, since the basic wrong of testimonial injustice is the undermining of the speaker qua knower', and conveying information is the 'most basic and immediate point of telling' (Fricker 2007, 60). 3 Therefore, 'the epistemology of testimony is the relevant framework within which to place all cases of prejudicial credibility deficit' (61). Consequently, we need to be able to account and correct for testimonial injustice within an epistemological evaluatory framework.
Fricker first examines the two most historically important candidate accounts of belief justification and knowledge acquisition: the reductionist account and the non-reductionist account, also sometimes called the antireductionist account. The reductionist account is the classic inferentialist account, which requires testimonial utterances to be inductively verified by the hearer, whilst the non-reductionist account is the classic noninferentialist account, which grants the hearer a presumptive right, or a priori entitlement (depending on which version is followed) to accept the speaker's word as a default position, in the absence of contrary evidence. She rejects the reductionist account, primarily on the widely accepted ground that its requirement that all testimony must be verified inductively is impracticable, and also not reflective of how we evaluate the vast majority of everyday testimonial encounters in practice. Fricker's rejection of reductionism is conceptually logical, since the notion that all testimony should be inductively verified is clearly contrary to the basic thrust of her argument that belief in testimony is situated in a hearer's personal evaluation of the testifier rather than the subject matter of her testimony. It is, though, an objection typically registered against global reductionism, which is the traditionally espoused Humean form of reductionism, rather than its more widely accepted contemporary form, known as local reductionism. The latter requires only that, in the case of each particular testimonial utterance, the relevant speaker's word should be verified through the hearer's face-to face evaluation of the that speaker, supported by any background evidence available to the hearer. One has to assume, though, that Fricker also rejects local reductionism and I can see two grounds for her doing so. The first is that local reductionism has no starting point of trust in the word of the speaker and so is not specifically anti-prejudicial. The second is that it does not envisage hearers having any epistemically specific training for their task of monitoring the speaker. As Elizabeth Fricker (1994Fricker ( , 1995 makes clear in her widely quoted account of local reductionism, hearers are not required to possess or cultivate particular epistemic virtues. They are simply encouraged to evaluate the speaker's sincerity and competence, as evidenced by her 'beliefs, desires, and other mental states and character traits', by employing 'commonsense psychology or person-theory', though this might, in the case of the 'right' hearer, include virtues necessary or beneficial for successful epistemic evaluation, such as open-mindedness, attentiveness, rationality and so forth. (Fricker 1994, 148-149) On the other hand, Miranda Fricker raises no conceptual objection to either leading version of modern non-reductionism: that formulated by C.A. J. Coady 4 and that formulated by Tyler Burge. 5 It is not wholly clear that she endorses them, but arguably this can be assumed from the fact that their default starting position of a presumption of trust in (Coady) or an entitlement to accept the speaker's word, as long as what she says is intelligible (Burge) is wholly consistent with her conception of a testimonially just hearer. Fricker claims, however, that non-reductionism is wrong in perceiving that the critical faculties of speakers are in 'snooze' (critically unreflective) mode unless and until alerted by some 'cue for doubt', when a mental 'shift of gear' sends them into 'critical' mode. In reality, she contends, our critical antennae are never wholly switched off, but are in a constant mode of 'critical openness' to the word of others and are always alert for signs or cues which might alter our view, in one direction or another, of how far to trust the speaker. (Fricker 2007, 66) I am not sure of her grounds for making this claim, since I perceive non-reductionism as simply creating normative entitlements to trust speakers, but I can see why she makes it, namely in support of an argument that correcting for anti-prejudicial credibility deficit requires critical openness. A state of critical openness, in turn, I assume the argument goes, requires the adoption of virtue-epistemological methodology in order to 'wake up' snoozing hearers.
Fricker's chosen epistemological methodology for cultivating responsible hearers is thus to pursue a 'virtue epistemological account of testimony'. Virtue epistemology, unlike both reductionism and non-reductionism, presents as a specifically epistemic agent-centred account, the basic idea underpinning which is that 'knowledge is true belief that is gained as a result of the reliable operation of [the agent's] epistemic virtues or cognitive faculties' (Pritchard 2014, 58). Cognitive faculties in this context are natural, or innate faculties, like perception, intuition, and memory, whilst epistemic virtues are 'cultivated' character traits like conscientiousness, intellectual and moral courage, and open-mindedness. The idea of knowledge being acquired through learned cognitive virtues is an old one. Linda Zagzebski observes that it can be traced back to Plato's notion of episteme ('knowledge' or 'understanding') which he associated with techne -practical human arts or skills acquired in fields as diverse as medicine, shipbuilding and cookingand also Aristotle's notion of epistemic virtues which must be learned over time, like those of builders, or musicians (Zagzebski 2001, 240). Virtue theory in effect builds on this idea by giving the notion of these learned practical skills a moral dimension, as intellectual virtues. In the context of virtue epistemology, Fricker sees this as her notion of a 'responsible hearer'one who is suitably trained or socially educated to 'see the world in moral colour', and to exercise 'rational sensibility, without inference, so as to be critically open to the word of others' (Fricker 2007, 71). It is not possible, within the length or scope of this essay to do full justice to Fricker's carefully worked through portrayal of a responsible, or virtuous, hearer. I would briefly summarize it in a testimonial context, though, as that of a hearer who possesses the capacity to recognize cues relating to trustworthiness, which enable her to make correct, non-prejudicial judgements concerning the credibility of testifiers -for example about their sincerity. The concept of the virtuous hearer, adds Fricker, embraces attitudes which are both intellectual and emotional. A perception of a speaker as trustworthy, for example, contains a feeling of trust in her. Her virtuous hearer would thus aim, over time, to develop both a critical reflexive awareness of identity-prejudicial stereotyping, and the means to neutralize it. She acknowledges that the ability to issue assessments of speakers 'ready-corrected' for identityprejudicial credibility deficit would in practice be hard to achieve due to 'the psychologically stealthy and historically dynamic nature of prejudice' and its 'powerful visceral force' (98). Indeed, she accepts that some identity prejudices, such as those relating to colour or gender, may be so historically and socio-culturally embedded that any resulting testimonial injustice should be considered, in effect, non-culpable (as in the case of Herbert Greenleaf), due to there being no collective social consciousness of the relevant identity-prejudice as an epistemic shortcoming. On the other hand, she adds, if a collective consciousness of a previously unconscious identity prejudice subsequently emerges, individuals who are guilty of continuing to make identity-prejudicial credibility judgements may then become considered to be culpably prejudiced.
Virtue epistemology is not without its critics; some commentators, for example, hold that it cannot account for cases of 'epistemic luck', 6 in which justified or true beliefs are acquired by chance. However, two specific issues arise in the context of its utility as a framework for correcting for identity-prejudicial credibility deficit. The first is how far it is possible in the real world to aim for correcting irrational and sometimes deeply embedded prejudices in oneself through some kind of rationally acquired 'virtuous' process. The second is why, even if this can be done, an epistemic agent would need a virtue-epistemological framework to achieve it, rather than applying general moral, intellectual, and social virtues, such as reasonableness, fairness and intellectual honesty. In this light, it seems to me that there are two key capacities, which virtue epistemologists need to demonstrate that virtue epistemology possesses, which other epistemological accounts do not. The first is a particular capacity to train epistemic agents to develop the requisite heightened levels of awareness of the operation of identity prejudice within themselves and others, and the second is a particular capacity to equip epistemic agents with some specific methodological tools to correct for such prejudices.

The Issues Concerning Testimonial Injustice
I will assume for the purposes of the remainder of this essay that it is not disputed that identity-prejudicial credibility deficit exists as a phenomenon, which is present within many, if not most, societies. On the basis of that assumption, it seems to me that the field of possible objections, which have been raised to the notion of identifying and correcting for epistemic injustice and specifically testimonial injustice elaborated by Fricker, can be narrowed to the following: (1) There is nothing specifically testimonial about Fricker's depiction of testimonial injustice, save for the context in which it occurs -that it is, in other words, an instance rather than a kind of identity-prejudicial judgment.
(2) There is nothing uniquely epistemic about the operation of identityprejudicial credibility deficit in the context of a testimonial exchange; it is rather an application, in an epistemic context, of a priori social and cultural prejudices or biases, such as racial or gender prejudice, held against others not only as knowers, but more generally as humans.
(3) Identity-prejudicial credibility deficit, like identity prejudice itself, is not a monolithic phenomenon, which can be accounted for or corrected through normative rules or practices. Moreover, even if the underlying prejudice is widespread, particular instances of its occurrence, such as in a testimonial encounter, may be differently contextualized according to the individual characters and background experiences of relevant hearers and speakers, and their personal interactions.
(4) There is no real evidence that testimonial injustices, or epistemic injustices more generally, are self-correctable, or, even if they were, that epistemic virtues specifically accessed through the application of virtue-epistemological principles are required to correct for them.
Adam Piovarchy (2021) specifically challenges Fricker's moral distinction between what she labels culpable and non-culpable instances of testimonial injustice. He suggests that it is not only the judgements of hearers like Herbert Greenleaf -those whose social backgrounds deprive them of the understanding that they are perpetrating a testimonial injustice -that should be deemed non-culpable, but the judgements of all those who lack the social capacity to correct, or indeed perceive the need to correct, for their prejudices, such as the jurors in Tom Robinson's trial. His argument is that neither the testimonial evaluations of Greenleaf, nor the Maycomb County jurors are epistemically blameworthy, because they are not making epistemic evaluative judgements. In fact, the whole point is that their minds are so socially and culturally distorted that they are unable to make such evaluative judgements, however compelling the evidence might have been to an unprejudiced hearer. As Fricker herself observed in the case of the Maycomb County jurors, for example, 'they cannot, in the courtroom context, perceive Tom Robinson as anything other than a lying Negro' (Fricker 2007, 36). The idea that his version of what happened should be preferred to that of a white girl 'is virtually a psychological impossibility' and thus bound to make them reject Finch's call to them to do their moral and legal duty. (25) An alternative view might be that both Greenleaf and the Maycomb County jurors are epistemically culpable for discounting evidence purely on the basis of identity-prejudicial stereotyping, but the jurors are also morally culpable of a dereliction of their legal duty to consider evidence put before the court on its merits, and to deliver justice. Greenleaf's culpability might thus be more excusable than that of the jurors. In either case, though, I would agree that the notion of any distinction between culpability and nonculpability as epistemic in nature seems questionable. José Medina's earlier (2011) critique of Fricker's conception of testimonial injustice is much broader than, though to some extent consistent with, that of Piovarchy. He argues that Fricker has given too little attention to the role of identity-prejudicial credibility excess as a phenomenon and its role in deflating the credibility of others. In essence, he appears to see most kinds of identity-prejudicial actions as a refraction of social and historical imbalances between a society's 'haves' and 'have nots'. Consequently, he argues: The proper analysis of a testimonial exchange requires looking into what happens before and after the exchange, looking into what happens in other exchanges and in society as a whole. (17) Medina argues further that identity-prejudicial credibility excess, contra Fricker, is a regular cause of testimonial injustice, making speakers dogmatic and bullying, and hearers intimidated and inhibited to the extent that they deflate their own credibility and thus wrong themselves. 7 Moreover, he contends, according members of some social groups disproportionally excess credit typically results in members of other groups getting disproportionally less credit. Whilst this is plausible, I would not follow Medina so far as to say that the operation of these social forces is universal, or automatic. I would suggest rather that credibility assessments of speakers, especially in the context of face-to-face testimonial exchanges, will typically also involve reactions to them as individuals as well as members of groups, though it is surely right to say that reactions to speakers as individuals may frequently be overwhelmed by reactions to them arising from a priori stereotypical prejudices.
Medina therefore looks at the case of Tom Robinson as one which is driven by the identity-prejudicial credibility excess attributed to Mayella Ewell and her father as members of the privileged white community, rather than the identity-prejudicial credibility deficit attributed to Tom Robinson. Moreover, like Piovarchy, he regards the comparative credibility differential between the two as being already firmly in place before the start of the trial, and thus the injustice directed towards Tom being not directed towards him specifically as a knower, but rather as a Negro, or more specifically as a Negro man charged with raping a white woman. For example, Medina notes, Tom's evidence concerning Mayella's advances towards him is not so much rejected as 'rendered incredible (in fact almost unintelligible) in that culture' (Medina 2011, 25). That the credibility of Mayella and her father, though she is a poor witness, must be inflated to the point of unconditional belief, and that of Tom deflated to the point of disbelief, is a direct consequence of the 'social imaginary', as Medina calls it, which comes between the jurors and their reason, and make them blind and deaf to the evidence. In fact, Medina declares, the jury in the Robinson case were affected not just by blindness to the evidence but by 'meta-blindness' -blindness to their blindness -in contrast to others in the courtroom, most notably the black spectators, who were able to see the evidence for what it was. The same, it seems to me from Fricker's account of him, can be said of Herbert Greenleaf. The fact that there can be no epistemic middle ground available between those who could see and those who were meta-blind, Medina claims, rules out the possibility that any epistemic injustice the defendant suffered could have been corrected. Corrective action such as Fricker envisages would have required there to be some 'unbiased space' in which hearers could have engaged with their own prejudices on epistemic grounds, such as the weight of evidence, but metablindness makes this impossible: 'The jurors' ignorance of their own prejudices -more than the absence of any particular virtue -is what got in the way of epistemic justice' (30). In other words Medina, like Piovarchy, rejects the notion of the injustice in the Robinson case as either testimonially related, or epistemological in nature. This seems to me to necessarily imply a rejection of Fricker's thesis of correctability as something which can be aimed for or achieved within an epistemological framework.
Medina's analysis seems plausible to me within the racial context in which he puts it, though the conception of identity-prejudicial credibility deficit as being typically comparative in nature in all situations is perhaps one that would require more detailed analysis. Medina refers to his account as a contextualist extension of Fricker's thesis, which he generally endorses, but it seems to me that it is a refutation of at least some of the ideas that are central to Fricker's notion of the problem of testimonial injustice as fundamentally epistemological, or epistemologically correctable. Benjamin Sherman's critique of Fricker's thesis of testimonial injustice, on the other hand, is psychologically and behaviourally centred, and challenges her contention that epistemic justice is a virtue one 'can and should aim for in practice' (Fricker 2007, 98-99). Virtue theory, he says, may appear to some to attractively combine moral and epistemic goodness, but it has its disadvantages. Firstly, echoing to some extent the arguments of Piovarchy and Medina, Sherman argues that in focusing on individual vices and virtues as the centre of moral life, virtue theory ignores the constraints imposed by immutable historical situations and social structures on the exercise of personal virtues. Secondly, he contends that virtue theory must claim that there are virtues which are stable -for example that an epistemic agent who possesses the virtue of honesty is consistently honest across most situations -which isn't necessarily something one can 'aim for'. Virtue epistemology, he notes, is attacked by those who argue that situational factors irrelevant to moral and epistemic reasoning 'have much greater impact on human thought than virtue theory can accommodate' (Sherman 2016, 234). One cannot therefore assume that a universally stable state of virtue is possible -that someone who is courageous or openminded in some situations, for instance, will be courageous or openminded in others. (235) This, in turn, casts doubt on the notion that epistemic virtues are ideal states of mind that can be aimed for. Thirdly, Sherman asks, to the extent that testimonial injustices are not committed knowingly, how can the psychological ability to recognize and to correct for them be developed, given that there is no evidence of which he is aware that errors of epistemic judgment are reliably neutralizable by purely introspective 'critical reflection'. In fact, he suggests, though reflecting one's identity prejudices 'might bring to mind a few errors' it is more likely, on the vast majority of occasions, to reinforce one's initial value judgements, and the more psychologically deeply-rooted the judgment is, the more likely it is to be thought fair and accurate on reflection. (238) Moreover, he says, one's idea of what is virtuous might be erroneous; you might for example think it is virtuous to bully your fellow jurors into agreeing with what you sincerely but mistakenly believe is the correct interpretation of someone's testimony. In reality, suggests Sherman, most people consider themselves just and epistemically responsible, and don't normally defer to the judgements of others if they disagree with them. Consequently, though he has no objection to a conception of testimonial injustice as mistaken belief, and thus an epistemic vice one might strive to avoid, he cannot agree that there must be some virtue-theoretic ideal state of being testimonially just, let alone one that can be aimed for.
Sherman concludes that the underlying psychological problem of Fricker's thesis of correcting for testimonial justice is that its prerequisite is that the hearer see himself as prejudiced, which is not feasible in the case of the deepest-rooted, and often unconsciously held, stereotypical prejudices, such as racial and gender prejudice. Unsurprisingly, Sherman lines up with the other commentators and indeed Fricker herself, in concluding that one cannot correct for prejudices that are so deeply embedded as to be unconsciously held, but raises a further objection. He argues that Fricker has provided no evidence that there is a corrective meta-virtue for identity prejudice -'some skill or disposition as avoiding prejudices in general' (244). The mere fact that testimonial injustice might be a vice, he declares, does not of itself imply that there must be a virtue of testimonial justice which can correct for it. (Ibid.) We can, then, perhaps do no more than be vigilant about being vulnerable to prejudice, and rely on the best available social and scientific information to identify instances of it. Moreover, he argues, since we can't know that there aren't other, yet unidentified, prejudices with which we may have to deal, we can't ever know that we have attained a general, stable disposition of being testimonially just. On the contrary, thinking we have achieved it might merely make us complacent and less vigilant about new prejudices. Better then, perhaps, to accept that we will sometimes be unknowingly testimonially unjust, and try to be vigilant in looking out for, and if possible correcting, individual failings on our part. In fact, he suggests, the best solution might be simply to abandon virtue-theoretical thinking altogether. Fricker's insights, he argues don't need a virtue-theoretical framework, because all moral theories worry about the harms and injustices she shows as resulting from testimonial injustice, and all theories of rationality and epistemic responsibility are on guard against the errors of judgment involved. If abandoning virtue theory altogether is unacceptable, Sherman suggests, we could retain the notion of testimonial injustice as a vice, but abandon the notion of testimonial justice as a generally attainable virtue.
In fairness to her, Fricker does at least partially acknowledge the problems Sherman highlights. 'It may be', she says, 'that the ever changing and selfrenewing nature of prejudice in society means that the best we can really hope for is to achieve the required social reflexivity of judgment through repeated efforts of critical reflection, in which case, in so far as we may achieve the virtue [of testimonial justice], our possession of it will not be full but partial' (Fricker 2007, 97). The difference may be one of emphasis: that whereas Fricker seems to see the issue in terms of social realities and the limits of individual resistance to social pressure, Sherman seems to see it as more fundamentally one of human psychology and the limits of introspection.
Either way, it seems to be common ground that the barriers to achieving testimonial justice are predominantly social and cultural, especially in the case of identity-prejudicial judgements that are deeply and widely socially and culturally embedded, as in the case of racial prejudice and gender prejudice. Correcting for them is consequently not merely a matter of moral self-improvement, but of socially and culturally counter-intuitive thinking and action -correcting, that is, for one's own social and cultural upbringing and thus in that sense acting against oneself as a 'social being'. I would not say such corrective action cannot be aimed for, but it does seem to me to that achieving it represents a considerable challenge.
The above commentaries represent a necessarily limited critical appraisal of Fricker's thesis, but nevertheless raise a range of issues. Bearing them in mind, but also following my own reading of Fricker's account of testimonial and epistemic injustice, I now suggest the following answers to the questions raised at the beginning of this section: (1) Testimonial injustice is not a distinct phenomenon, but rather a context in which the phenomenon of identity-prejudicial credibility deficit manifests itself.
(2) Identity-prejudicial credibility deficit is credibility deflation based on an a priori prejudice against the speaker as a person rather than an a posteriori prejudice based on an evaluation of the content of the speaker's utterance, and thus not a specifically epistemic phenomenon, though the harm it inflicts may be epistemic. (3) Identity-prejudicial credibility deficit is one form of a more widely experienced social and cultural phenomenon of identity-prejudicial judgements, which is typically, but not necessarily, derived from the social stereotyping of speakers as members of a racial, ethnic, or social community or group. However, particular instances of it may also be affected by their context, the relative social situations and characters of the individual parties, and, in face-to face encounters, their personal interaction.
(4) Fricker has not shown how a specifically virtue-epistemological framework is conceptually or methodologically necessary or desirable for the task of self-monitoring and self-correcting for testimonial injustice, or that there is a specifically epistemic virtuous state of being 'testimonially just'. More broadly, since stereotypical identity prejudice is a social phenomenon, and often held unreflectingly, the extent to which it is individually correctable through the application of what Fricker calls 'a distinctly reflexive critical social awareness' is debateable.
Having largely dealt with the issues raised by other commentators which seem to me particularly plausible, I will finish by adding some comments of my own regarding Fricker's conception of testimonial injustice.

Looking Back and Moving On
The term 'testimonial injustice', as employed by Fricker, embodies three elements. The first of these is what she describes as negative identityprejudicial stereotyping, such as racial or gender stereotyping. The second element is where negative identity-prejudicial stereotyping causes one person to give the word of another person less credibility than she otherwise would give it. If we pause there, we have the notion of identity-prejudicial credibility deficit -deflating a speaker's credibility as a result of whatever identity prejudice we hold against them. This notion is basically descriptive in nature -it describes the state of someone who holds a particular prejudice applying it in a particular way. The third element though introduces the notion of epistemic wronging -of wronging a speaker 'specifically in her capacity as a knower' (Fricker 2007, 20). In common with most other commentators, I have no issue with the first two elements, and fully recognize the importance and value of Fricker's extremely well-worked exploration of the issues and her conclusions. I do, however, in company with Piovorchy, Medina and Sherman, have issues with the third element and these are highlighted by Fricker's use of the cases of Tom Robinson and Marge Sherwood. These cases are undoubtedly powerful but extreme examples of identity prejudice at work, and the testimonial encounters they portray are specific in two crucial aspects. Firstly, they are cases of competing testimonies, in which two different versions of the same event are presented by different speakers in circumstances in which one version of the event must be wholly accepted, and the other wholly rejected. Moreover, the judgemental dice are loaded, since the personal and social consequences for the hearers of accepting one of the versions are far more damaging to them than those of accepting the other. In the case of Greenleaf, believing Marge means accepting that his son has been murdered and also that he has been cruelly duped by the murderer. In the case of the Maycomb County jurors, acquitting Tom Robinson means publicly undermining the social structure around which the relationship of white people to a black people is founded. Consequently, they are both cases in which existing deeply embedded identity prejudices are actually reinforced by the circumstances surrounding the testimonial event. Secondly, as already noted, in neither case is any serious attempt to evaluate the evidence as evidence in an epistemic manner remotely possible, since such an evaluation is pre-empted by socially conditioned indefeasible assumptions: in the case of Tom, that all negroes are liars, and in the case of Marge, that women are emotional rather than rational by nature, and that female intuition is by its nature epistemically unreliable. If, therefore, testimonial injustice requires the speaker to be wronged specifically in his or her capacity as a knower -by way of an epistemically wrongful rejection of his or her testimonial utterance -neither Marge Sherwood nor Tom Robinson appear to be victims of testimonial injustice, though they are clearly victims of an identity-prejudicial disbelieving of their assertions. In other words, both scenarios appear to me to be antithetical not only to Fricker's notion of correctability, but more fundamentally to her definition of testimonial injustice, and even her conception of identity-prejudicial credibility deficit.
My other issue concerns how Fricker's conception of testimonial injustice might work when applied to testimony more generally. As she acknowledges, there are many kinds of testimonial encounter, even within the limited category of spoken face-to-face testimony. Much of it -almost certainly most of it -occurs in everyday requests for information or advice, such as requests for travel information or advice regarding the selection of some item to purchase, for example, or routine workplace exchanges, in which hearers may well be sometimes unreflectively evaluating the information they receive in an identity-prejudiced manner. In fact, hearers may already have engaged their identity prejudices in the selection of their informant, and thus never actually reached the point at which their identity-prejudiced response could become epistemic in nature. As Sherman points out, such instances of unreflective identity prejudice are also the most difficult to recognize within oneself, let alone correct for, and reflecting upon them might not necessarily result in them being seen as epistemic vices requiring correction. But I would suggest that the difficulty is in reality much greater and more indeterminable than even Sherman suggests. Consider, for example, the following scenario: I visit a car showroom in search of a new car, and, upon entry see the model I have in mind, and notice two sales people, one of whom is male and the other female. I am keen to know more about the car's specification and performance and instinctively approach the male salesperson, being socially conditioned to believe a man will possess more technical knowledge of this kind than a woman. However, when I start to interrogate the salesman, he informs me that his colleague 'knows more about the technical stuff' than he does, and calls her over to talk to me. I accept his word, and after an informative conversation, follow his colleague's recommendation to purchase the vehicle I originally had in mind.
Trying to deconstruct this scenario, it appears that I had initially been guilty of identity-prejudicial credibility deficit per Fricker, but identityprejudicial credibility excess per Medina. I had then corrected for it, but not through a process of virtuous self-correction. Initially, it was on the basis of what the male salesperson I had prejudicially selected told me, so I deserve no epistemic credit for my decision to engage with his colleague beyond accepting the truth of his utterance, thus arguably reinforcing my identity-prejudicial reliance on him. On the other hand, I had thereafter accepted my female informant's advice, so maybe in the end I deserve credit for correcting for my originally prejudiced credibility judgment. But how far the virtue I have thus acquired is specifically epistemic is still unclear. And there again, to what extent was my acceptance of the female salesperson's advice at least in part a result of confirmation bias, since she had confirmed my original preference? Moreover, assuming that I do decide that I have been guilty of identity-prejudicial credibility deficit and must correct for it,andthus consciously choose to approach a female salesperson on the next similar occasion, am I thereby exercising the virtue of testimonial justice, or irrationally according the unknown female salesperson excess credibility?
The point is that I have no hard and fast answers to these questions. Fricker's thesis, much as I accept its conceptual plausibility, seems to give me no practical help in cognitively deciphering the scenario I describe, or determining my future actions. That scenario may seem trivial, compared to an allegation of rape or murder, but it is surely closer to that in which identity prejudice is likely to be played out in everyday life, and thus far more relevant to the vast majority of those who are the victims of identityprejudiced judgements. My contention, in other words, is that the price that Fricker seems to have paid for building her thesis around the Robinson and Greenleaf cases is to have significantly narrowed its scope. Consequently, I am inclined to read her book as a valuable and important, though to some extent flawed, conceptual analysis, of two egregious illustrations of identity-prejudicial credibility deficit, but as far as the wider application of her thesis is concerned, to see it more as a valuable introduction to a yet to be fully explored, and much more widely experienced, phenomenon of identity-prejudicial social judgements in societies generally.

Concluding Remarks
There is little doubt that Fricker's examination of the pernicious influence of identity prejudice on testimonial evaluation is of great significance, and notable both in its scholarship and originality. She has valiantly struggled with what seems to me to be the dilemma which is at the heart of her characterization of the debate, which is whether and how one can tackle the irrational behavioural phenomenon of identity-prejudicial credibility deficit by applying the kind of rational solutions offered by epistemic theories and methodologies. Ultimately, I believe, that task is in reality not only beyond Fricker, but also beyond the exclusive reach of epistemology, and indeed philosophy. It probably requires the development of the public will and ability, separately in the case of each instance of identity prejudice, to invoke changes in collective psychology and social and moral perceptions towards its victims, especially when it comes to the most fundamental cases, such as racial and gender prejudice. To the extent that such a project is feasible, it will probably require the combined attention of a whole range of professional disciplines, as well as considerable public goodwill, and political and legislative support, over an indeterminable period of time. Consequently, I would myself see more benefit in treating the problem of identity-prejudicial credibility deficit as being one which is not specifically testimonial or epistemic, but rather one of how identity prejudice reveals itself more generally in an irrational inability to believe the word of others. Notes 1. I shall mostly refer herein to testifiers as 'speakers' and recipients of their testimony as 'hearers', since a speech act in which a single speaker addresses one or more hearers is the form of testimony which is the subject of Fricker's study. 2. Jean Améry spent a year in Berlin, in 1935, where he witnessed the phenomenon he describes at first hand, before fleeing Vienna into exile in Belgium, in 1938. His adopted surname Améry was a francophone anagram of Mayer, his original surname. 3. This is not an assessment with which I agree, as will become apparent later in this essay, but in the interests of presenting her case fairly, I will follow her for now. 4. The version espoused in Coady (1992) that hearers have a general presumptive epistemic right, as a default position, and absent any clear evidence to the contrary, to believe the testimony of others. 5. The version espoused in Burge (1993) and restated with modifications in Burge (1997) that 'a person is entitled to accept as true something that is presented as true and that is intelligible to him, unless there are stronger reasons not to do so' (Burge 1993, 467). 6. See, for example, Lackey (2007), and Kallestrup and Pritchard (2013). 7. Medina regards identity-prejudicial credibility deficit as necessarily a collective, rather than an individual phenomenon.

Disclosure Statement
No potential conflict of interest was reported by the author(s).