1. False Confessions

There are many factors that contribute to people falsely confessing to crimes that they didn't commit. First, there are situational factors that can significantly impact the likelihood of false confessions, including the length of the interrogation, sleep deprivation, the presentation of false evidence, and maximization and minimization tactics.Footnote ⁴ Let's examine these briefly in turn.

According to guidelines outlined by Inbau et al. (Reference Inbau, Reid, Buckley and Jayne2001), it is advised that single interrogation sessions not exceed 4 hours. Yet Drizin and Leo (2004) found that in cases in which interrogation time was recorded, 34% lasted 6–12 hours, 39% lasted 12–24 hours, and the mean was 16.3 hours. Moreover, lengthy interrogations are often accompanied by other factors that can increase the likelihood of false confessions, such as isolation from significant others, which ‘constitutes a form of deprivation that can heighten a suspect's distress and incentive to remove himself or herself from the situation,’Footnote ⁵ and sleep deprivation, which ‘strongly impairs human functioning.’Footnote ⁶

Regarding false evidence, it is permissible in the United States for police to outright lie to suspects, and so when involvement in criminal activity is denied, purportedly decisive evidence of guilt can be offered in response. Consider, for instance, the case of Marty Tankleff who, in 1989, was accused at the age of 17 of murdering his parents ‘despite the complete absence of evidence against him. Tankleff vehemently denied the charges for several hours – until his interrogator told him that his hair was found within his mother's grasp, that a ‘humidity test’ indicated he had showered (hence, the presence of only one spot of blood on his shoulder), and that his hospitalized father had emerged from his coma to say that Marty was his assailant – all of which were untrue (the father never regained consciousness and died shortly thereafter).’ Following these lies, Tankleff became disoriented and confessed, but then immediately recanted. ‘Solely on the basis of that confession, Tankleff was convicted, only to have his conviction vacated and the charges dismissed 19 years later’ (Kassin Reference Kassin, Drizin, Grisso, Gudjonsson, Leo and Redlich2010, pp. 17-18). That the presentation of false evidence contributes to such confessions is reinforced by self-report studies, where suspects say that the reason they confessed is that they took themselves to be trapped by the weight of the evidence against them.Footnote ⁷

There are also maximization and minimization tactics, which research has shown can lead to false confessions.Footnote ⁸ Maximization is a ‘hard-sell’ approach that involves the interrogator trying to scare or intimidate the witness, offering false claims about the evidence, and exaggerating the seriousness of not cooperating. Minimization is a ‘soft-sell’ approach in which the interrogator ‘tries to lull’ the witness into a ‘false sense of security by offering sympathy, tolerance, face saving excuses, and even moral justification’ (Kassin and McNall Reference Kassin and McNall1991, p. 235). Such techniques come in three different forms: ‘those that minimize the moral consequences of confessing, those that minimize the psychological consequences of confessing, and those that minimize the legal consequences of confessing’ (Kassim et al., Reference Kassin, Drizin, Grisso, Gudjonsson, Leo and Redlich2010, p. 12). For instance, the interrogator may offer sympathy and understanding to normalize the crime, saying, for instance, ‘I would have done the same thing;’ the interrogator might offer minimizing explanations of the crime, such as that the murder was spontaneous or accidental; and the interrogator might communicate promises through pragmatic implication that the suspect will be punished less severely if he or she confesses. All three forms can put pressure of varying degrees on a suspect to confess to a crime that he or she did not commit, especially when used in combination with some of the other techniques, such as the presentation of false evidence.Footnote ⁹

In addition to situational features, there are dispositional factors that increase the likelihood of false confessions, and the two most commonly cited concerns are juvenile status and mental impairment, including developmental disabilities and mental illness. This is supported by the fact that these groups are wildly overrepresented in the population of proven false confessions. For example, of the DNA exonerations in the U.S. involving false confessions, 31% of the false confessors were 18 years or younger and 9% had mental health or mental capacity issues known at trial.Footnote ¹⁰ In their sample of wrongful convictions, Gross, Jacoby, Matheson, Montgomery, and Patel (Reference Gross, Jacoby, Matheson and Montgomery2005) found that 44% of the exonerated juveniles and 69% of exonerated persons with mental disabilities were wrongly convicted because of false confessions. There are a number of factors at work here. In both groups, for instance, there can be impairments in adjudicative competence, such as the ability to assist in one's own defense. There can also be a diminished capacity to grasp legal terms, such as Miranda rights.

Finally, false confessions are often facilitated by the very innocence of the suspect. Awareness of one’s own innocence leads people not only to waive their Miranda rights to silence and to counsel,Footnote ¹¹ but also to be more open and forthcoming in their interactions with police.Footnote ¹² If you have nothing to hide, you might wonder why you should remain silent and get an attorney. Yet it is not uncommon for the testimony of those who are innocent to be used against them, such as by calling into question their reliability or sincerity on the basis of minor inaccuracies. In addition, when a suspect confesses, this often leads the police to regard the case as solved, thereby closing the investigation and increasing the likelihood of overlooking exculpatory evidence.Footnote ¹³

2. Testimonial Injustice

With these points in mind, let’s now take a closer look at the phenomenon of testimonial injustice. The standard view is that a speaker is a victim of testimonial injustice when she is afforded a credibility deficit in virtue of a prejudice on the part of a hearer that targets her social identity.Footnote ¹⁴

A speaker suffers a credibility deficit when the credibility that she is afforded by a hearer is less than the evidence that she is offering the truth, and a hearer has the relevant kind of identity prejudice when she has a prejudice against the speaker in virtue of the latter’s membership in a social group. Prejudice here is being understood in terms of not being properly responsive to evidence. A prejudicial stereotype, for instance, is a generalization about a social group that fails to be sufficiently sensitive to relevant evidence. Where this prejudice ‘tracks’ the subject through different dimensions of social activity – economic, educational, professional, and so on – it is systematic, and the type that tracks people in this way is related to social identity, such as racial and gender identity. For instance, if a police officer rejects a woman’s report of sexual assault merely because his sexism leads to him discrediting her, and despite evidence that supports her credibility, this would be a paradigmatic instance of testimonial injustice. In particular, the police officer’s sexist beliefs manifest as a prejudice that targets the victim’s gender identity in a way that results in her testimony being regarded as less credible than the evidence supports.Footnote ¹⁵

When a hearer gives a speaker a credibility deficit in virtue of her social identity, Miranda Fricker argues that the speaker is wronged ‘in her capacity as a knower,’ and is thereby the victim of testimonial injustice. For the purposes of this paper, I will focus entirely on the claim that only credibility deficits are relevant to this phenomenon. Here, Fricker explicitly considers the question of whether credibility excesses can result in testimonial injustice, and denies that they can be, at least in the paradigmatic sense that is of interest to her. She writes:

Fricker grounds her denial that credibility excesses can lead to testimonial injustice in her rejection of a distributive model of credibility.Footnote ¹⁶ In particular, she argues that credibility is not finite in a way that lends itself to a distributive treatment.Footnote ¹⁷ Consider, for instance, finite goods, such as wealth, land, or food. Not everyone can own 20 acres because there is only a limited amount of land to go around. For some to have a lot of it necessitates that others have a little, or none at all. But other goods don’t limit one another in this way. To give moral praise to one person need not be to deny it to another. We can say equally of Abraham Lincoln, Ida B. Wells, and Martin Luther King Jr. that they are exceptional moral agents. In this way, there is an important sense in which moral praise is an infinite good: there is often enough of it to go around.

Credibility, according to Fricker, is like moral praise rather than like land: it is an infinite good. If two friends tell me about their vacations this summer, believing that one of them snorkeled in Thailand need not impact my trusting that the other went hiking in Peru. I can give them both as much credibility as I like since not only is there plenty to go around, but giving some of it to one need not take any away from the other. Because of this, it seems it isn’t unjust to give someone more credibility than is owed since this doesn’t deprive someone else of a good that is deserved.

3. Credibility Excesses and Testimonial Injustice

I began this paper by highlighting that the U.S. Supreme Court regards confession evidence as possibly the most powerful evidence of guilt admissible in court. At the same time, confessions are often acquired through coercion, manipulation, and deception – such as by using maximization tactics and presenting false evidence – as well as by targeting vulnerable suspects – such as those who have been sleep deprived, juveniles, and the mentally impaired. What I now want to show is that when the testimony of a confessing self is privileged over a recanting self because of prejudice, whether racial or otherwise, this results in a unique kind of testimonial injustice that is due to a credibility excess.Footnote ¹⁸

This can be seen most clearly by focusing on several features of false confessions. First, false confessions are highly resistant to counterevidence, which, it might be recalled, is a key feature of the prejudicial stereotypes often at work in instances of testimonial injustice. Despite awareness of the reality and prevalence of false confessions, as well as their causes and effects, false confessions are frequently taken to be sufficient for grounding convictions. This occurs even when there is powerful evidence on behalf of defendants’ innocence in particular cases, and the stakes simply couldn't be higher. In a recent article showing how false confessions trump exculpatory DNA evidence, for instance, Appleby and Kassin discuss the case of Juan Rivera, who was convicted of the rape and murder of an 11-year-old girl in Waukegan, IL on the basis of his confession, even after DNA testing of semen at the scene excluded him. ‘The state's theory of why DNA belonging to someone other than the defendant was found in the victim was that the young girl had prior consensual sex with an unknown male, after which time Rivera raped her, failed to ejaculate, and then killed her’ (Appleby and Kassin, Reference Appleby and Kassin2016, p. 127). The fact that Rivera was convicted of the child's murder shows that the state's outrageous theory was regarded as more credible than the possibility that he confessed to a crime he didn't commit – in other words, a single confession trumped evidence that would otherwise be taken to be decisively exculpatory. The most plausible explanation for this is that the false confession received a massive, unwarranted excess of credibility.

This is even more vivid when it is noticed that the totality of the evidence against confessions is often substantial, while the evidence in their favor is remarkably thin. Returning to the case of Juan Rivera, the evidence in favor of his innocence wasn't only the DNA that excluded him, and the fact that the state needed to construct an incredible theory to explain this. He was also a 19-year-old former special education student, who had been questioned by detectives for four days, during which he steadfastly denied any knowledge of the crime. Around midnight on the fourth day, after the interrogators became accusatory:

Because of trial errors and post-conviction DNA testing, Rivera had three separate jury trials and was found guilty and sentenced to life in prison on all three occasions. It wasn't until the Center on Wrongful Convictions became involved that the Illinois Appellate Court ruled in 2012 that Rivera's conviction was ‘unjustified and cannot stand,’ and thus that the state would dismiss all charges. Rivera had served 20 years in prison.

If we look at the case of Rivera, the extent to which his false confession needs to be given an excess of credibility is not only undeniable, it is shocking. To see this, notice that knowledge is taken to be incompatible with the presence of defeaters, with can be either doxastic or normative, and either rebutting or undercutting.Footnote ²⁰ A doxastic defeater is a doubt or belief that you have that indicates that one of your beliefs is either false (i.e., rebutting) or unreliably formed or sustained (i.e., undercutting). A normative defeater is similar, except it concerns doubts or beliefs that you should have, given the evidence available to you. So, for example, if I believe that the animal in my backyard is a bobcat by seeing one there, I might get powerful evidence that such a belief is false by you telling me that bobcats have never lived in my state, or that my basis is a poor one by my optometrist reporting to me how much my vision has deteriorated. If I accept both instances of testimony, then I have doxastic defeaters, rebutting in the first case, undercutting in the second. But even if I reject the testimony in question, I am still on the hook for this counterevidence if I do so for no good reason at all. Why? Because it is evidence that I should have.Footnote ²¹ The justification that my bobcat-belief might have initially enjoyed, then, has been normatively defeated.

These tools can help us see the extent to which Rivera's false confession was given, over the course of decades and by people at every stage of the process – including police officers, prosecutors, and jurors – a massive excess of credibility that resulted in a distinct form of testimonial injustice. Given all of the research discussed above, Rivera was, first and foremost, a prime candidate for providing a false confession: he was a special education student, who had endured multiple lengthy interrogations, was sleep-deprived, and was shown to be in the middle of a psychotic episode. Moreover, Rivera's original confession was riddled with inaccuracies and implausible information. All of this, by itself, should challenge the reliability of Rivera as a source of information about his guilt. In other words, those accepting Rivera's confession had undercutting defeaters (whether doxastic or normative), since they had evidence that clearly showed that their beliefs that Rivera raped and murdered the 11-year-old child were unreliably formed or sustained. That is, they had evidence that the source of their beliefs about Rivera's guilt – namely, Rivera himself – was not reliable under the interrogation conditions in question. But they also had doxastic rebutting defeaters, since the DNA evidence excluded him as a source of the semen at the scene of the crime, thereby calling into question the truth of their beliefs that he was guilty of the crimes.

When one has a defeater of any kind, the only way in which the target belief can be rationally retained is if one has a defeater-defeater – that is, a further belief or evidence that defeats the original belief or evidence. So, for instance, the rebutting defeater for my bobcat belief might itself be defeated if I come to learn that a bobcat recently escaped from the local zoo. Or the undercutting defeater might be defeated if I discover that my optometrist consulted the wrong chart when concluding that my vision is unreliable. But notice: there is simply no way in which the state's incredible theory in which the 11-year-old was sexually active with some unknown male, and Rivera didn't ejaculate despite raping her, successfully works as a defeater-defeater here. In other words, there is no interpretation of the available evidence that makes this theory more plausible than the alternative one: namely, that Rivera falsely confessed under duress to a crime he didn't commit.

That false confessions are resistant to counterevidence is further supported by looking at the sheer number of instances of testimony that often need to be discounted in order to retain belief in the correctness of a corresponding conviction. Consider, for analogy, how the testimony of victims of sexual assault is often rejected or discounted because of bias or prejudice,Footnote ²² but how numbers can sometimes add up to tip the balance.Footnote ²³ So, for instance, a handful of girls and women accusing Larry Nassar of sexual harassment or assault wasn't enough for many to believe them, but when over 300 women came forward, the public started to side with their word over his denials.Footnote ²⁴ We saw something similar in the case of Bill Cosby, where 60 women reported being victimized at the hands of the once respected comedian and actor.Footnote ²⁵ Now while this involves the addition of new testifiers, we can see a structurally similar problem at work in false confessions. A confessing self often reports guilt only once – under conditions of coercion, manipulation, deception, sleep deprivation, stress, and so on – while a recanting self reports innocence hundreds, even thousands of times, often over a period of years. And yet despite this, the one report of guilt utterly swamps the thousands of reports of innocence, with no justification for this radical asymmetry in treatment of confession versus recantations. This provides another lens through which we can see that false confessions are highly resistant to counterevidence.

Second, false confessions reveal how credibility can be finite, and thus how its proper distribution is crucial for assessing whether a speaker is the victim of testimonial injustice. Typically, when we talk about distributing credibility, we have in mind doing so across different people. If a woman says she was assaulted and the accused assailant denies this, then the question is: which person do we believe? But in cases of false confessions, we are talking about distributing credibility across different times in the life of the same person. There is the earlier, confessing self and the later, recanting self. The question then becomes, which self do we believe: the earlier or the later one?

Of course, the mere fact that two people disagree, even about matters of fact, does not by itself require that credibility be finite between them. I may tell you that a local restaurant is open while someone else tells you it's not. That we offer competing reports here does not require that only one of us be deemed worthy of trust or belief: you can be credible, even if wrong on a particular occasion, and I can lack credibility, even if right in a one-off case. Many disagreements are the product of innocent mistakes or lack of information, and so there can still be enough credibility to go around.

But not all disagreements are like this. It's precisely when someone's credibility itself is on the line that its finitude rears its head. False confessions provide the clearest case here: when someone confesses to murder and then recants shortly thereafter, there are no errors or gaps in evidence to explain the disagreement away. To give credibility to the confessing self is ipso facto to deny it to the recanting self. Credibility becomes scarce.Footnote ²⁶ What this shows is that false confessions uniquely pit one against oneself, and reveal how an excess in credibility can lead to an egregious kind of testimonial injustice.Footnote ²⁷

Third, by virtue of the state saying that the reality described by the confessor in cases of false confessions – one that is reported only through coercion, manipulation, deception, sleep deprivation, and so on – represents her truest states, the confessor's status as a knower is reduced to what she reports only under conditions devoid of, or with diminished, epistemic agency. This is especially problematic since the question of whether one is a murderer can literally be a matter of life and death. So, while it is true that the recanter – the later self who accurately, consistently, and steadfastly describes a different reality that is not extracted through coercion, manipulation, or deception – is wronged in being afforded a massive credibility deficit, there is a unique and powerful epistemic wrong done to the earlier self who receives a credibility excess. Indeed, the excess given in false confessions quite literally amounts to the state saying that confessors are knowers with respect to the testimony in question only insofar as they are not epistemic agents.Footnote ²⁸

There is an instructive parallel here: in ancient Athens, the testimony of enslaved persons, who were the property of their masters or the state, was typically inadmissible in judicial proceedings except under torture. As Michael Gagarin writes, ‘One of the most criticized features of classical Athenian law is the bizarre institution of … ‘interrogation under torture’. A well-known rule held that in most cases the testimony of slaves was only admissible in court if it had been taken under torture, and in the surviving forensic speeches the orators frequently…praise the practice as most effective’ (Gagarin, Reference Gagarin1996, p. 1). Just as Athenian courts regarded the testimony of enslaved persons as reliable only when obtained via torture – and thus offered under conditions devoid of epistemic agency – so, too, do our courts privilege the testimony of confessing selves, even when confessions are extracted through interrogation techniques that undermine or compromise our epistemic agency. This is especially problematic when the confessing selves are vulnerable members of society, such as juveniles and those with developmental impairments, and when the credibility excesses are fueled by prejudice, such as racism and sexism.

We might take a step back here, then, and identify an interestingly different sort of testimonial injustice. More precisely, what I call agential testimonial injustice occurs when testimony is extracted from a speaker in a way that bypasses or subverts the speaker's epistemic agency and is then given an unwarranted excess of credibility. Agential testimonial injustice is extraordinarily vivid in cases of false confessions that have been extracted in various ways, but testimony obtained in ways that deny epistemic agency is not limited to such cases. Many abusive relationships, for instance, involve coercion of various degrees, including in testimonial contexts, and when what is reported under such conditions is unjustifiably privileged, one is the victim of this kind of testimonial injustice. Imagine, for instance, a woman testifying that her partner has never been abusive while he is sitting next to her in an interrogation room, but she then retracts this once she is able to extricate herself from his control. If the former testimony is weighed far more heavily than the latter for no good reason, particularly when one is aware of the broader context of the abuse, this would be an instance of what I'm calling agential testimonial injustice.

It is worth developing in a bit more detail the precise nature of two different epistemic wrongs involved in this kind of testimonial injustice. The first kind of wrong is the one highlighted above, which specifically involves the excess of credibility given to the extracted testimony. Here, one is epistemically wronged by virtue of being regarded as a testifier – a giver of knowledge – only when one's testimony is extracted and is thus the product of a process that subverts one's epistemic agency. This can be seen both by the weight that the confession is given relative to other evidence and, specifically, the privileging of the confessing testimony over the recanting testimony.

The second kind of epistemic wrong involved in agential testimonial injustice results from the very act of extracting testimony from a speaker in a way that subverts her epistemic agency. This can happen in different ways. The clearest and most extreme case is where the extraction, such as the interrogation tactics in question, leads subjects to believe in the truth of their own reports, either wholly or partially. For instance, following Kassin and Wrightsman (Reference Kassin, Wrightsman, Kassin and Wrightsman1985), Kassin et al. (Reference Kassin, Drizin, Grisso, Gudjonsson, Leo and Redlich2010) characterize ‘coerced-internalized’ false confessions as those in which ‘innocent but malleable suspects, told that there is incontrovertible evidence of their involvement, come not only to capitulate in their behavior but also to believe that they may have committed the crime in question, sometimes confabulating false memories in the process’ (2010, p. 15). In cases of coerced-internalized false confessions, then, both one's testimony and one's doxastic states have come under the will of another. By virtue of employing techniques that are coercive, manipulative, and deceptive, interrogators here are able to alienate a suspect from her own epistemic resources and bring about beliefs that are disconnected from her epistemic agency.Footnote ²⁹

These cases of coerced-internalized false confessions quite straightforwardly involve this second epistemic wrong. To see this, notice that epistemic agency is commonly understood as requiring a subject's responsiveness to reasons or evidence.Footnote ³⁰ On a strong reading of this, I exercise my epistemic agency with respect to my belief that p when my belief that p is responsive to reasons. When interrogators are able to manipulate not only the testimony of suspects, but their doxastic states as well, they are quite clearly interfering with the reasons-responsiveness of the suspects’ beliefs. This results in a clear instance of the second epistemic wrong involved in agential testimonial injustice since a subject's epistemic agency is being subverted in the obtaining of her confession.

But even when subjects don't internalize their own guilt, and thus continue to believe in their innocence despite saying otherwise, there is an important sense in which their epistemic agency is compromised in the extraction of their testimony. To make this clear, let's look at a couple of cases where a speaker reports what she herself does not believe but in a way that does not at all interfere with her epistemic agency.

Consider, first, lying – where a speaker states that p, believes that p is false, and states that p with the intention to be deceptive with respect to whether p.Footnote ³¹ Even though a liar aims to be deceptive in her reports, this does not at all interfere with her ability to be responsive to reasons in her beliefs or her testimony. Indeed, a liar might even be a responsible epistemic agent regarding her doxastic states, despite the fact that her statements aim away from the truth. Consider, next, cases of selfless assertion, Footnote ³² where there are three components to this phenomenon: first, a subject, for purely non-epistemic reasons, does not believe that p; second, despite this lack of belief, the subject is aware that p is very well supported by all of the available evidence; and, third, because of this, the subject asserts that p without believing that p. A classic cases of a selfless assertion is where a Creationist teacher correctly reports that Homo Sapiens evolved from Homo erectus to her students, even though she doesn't believe this herself. Here, the reported belief in question is resistant to counterevidence, and so the belief itself is at least not properly responsive to reasons. Nevertheless, the reporting of the selfless assertion does not in any way violate the speaker's epistemic agency and, in fact, is grounded in it. In particular, the speaker in such cases fails to report what she herself believes, but she does so for straightforwardly epistemic reasons. In this way, she is appropriately sensitive to reasons, not with respect to her own beliefs, but with respect to her testimony. On my view, then, selfless assertions straightforwardly reflect epistemic agency.

Let's now turn to the case of extracted testimony. For instance, Kassin et al. (Reference Kassin, Drizin, Grisso, Gudjonsson, Leo and Redlich2010) discuss ‘compliant false confessions,’ which are ‘those in which suspects are induced through interrogation to confess to a crime they did not commit. In those cases, the suspect acquiesces to the demand for a confession to escape a stressful situation, avoid punishment, or gain a promised or implied reward’ (2010, p. 14). Here, even if the beliefs of the suspects are responsive to reasons, this is utterly disconnected from the obtaining of their testimony. Unlike in the case of selfless assertions, for instance, where the offering of the reports is precisely what is grounded in the responsiveness to reasons, the tactics used to extract the confessions – coercion, manipulation, deception – subvert the epistemic agency of the suspects. In so doing, such speakers are the victim of this second kind of epistemic wrong involved in agential testimonial injustice.

Finally, recall that Fricker's conception of testimonial injustice focuses specifically on credibility deficits that result from prejudices targeting a speaker's social identity, such as race or gender. We might ask, then, what social identity is relevant in cases of false confessions, especially since it is not only members of underrepresented groups that are victims of the sort of credibility excess at issue in agential testimonial injustice. By way of response here, notice that Fricker's conception of social identity is unnecessarily narrow. While race and gender are certainly highly relevant to the credibility that speakers are given across many different contexts, there are other aspects of social identity that are important here as well, such as socioeconomic status, occupation, and so on. For instance, many people in the workplace find themselves on a hierarchy where those at the top are given more credibility than those lower down, even in areas that are entirely disconnected from their areas of expertise. In a hospital, physicians might be believed over nurses about questions that do not pertain to medicine, and if this happens regularly, and simply in virtue of professional status, it seems correct to say that the nurses are victims of testimonial injustice. Similarly, members of groups associated with delinquency, deviance, or moral deficiency, such as ‘criminals’ and ‘prisoners,’ are frequently the targets of systematic prejudice. A criminal record, for instance, presents a major barrier to employment,Footnote ³³ the label of ‘prisoner’ or ‘ex-con’ is highly stigmatized,Footnote ³⁴ and ‘offenders’ tend to be demonized as dangerous, dishonest, and disreputable.Footnote ³⁵ Some have even extended this prejudice to ‘suspects,’ as Edwin Meese III famously said in 1985, while he was Attorney General of the United States, ‘you don't have many suspects who are innocent of a crime. That's contradictory. If a person is innocent of a crime, then he is not a suspect’. Footnote ³⁶ Thus, a plausible explanation is that at least one of the social identities targeted in cases of false confessions is ‘criminal’ or ‘suspect,’ triggering in others the belief that the confessor is guilty. Of course, this can combine with other prejudices, such as those involving race, ethnicity, and socioeconomic status. But, as we have seen, confessions alone can lead to egregious acts of testimonial injustice.

Agential testimonial injustice thus involves a testifier suffering two epistemic wrongs, both through the process by which the testimony is extracted and by virtue of the credibility excess it receives. I now want to briefly turn to why confessing selves might be given a credibility excess in the first place. And notice just how crucial it is to address this question. For convictions based largely on false confessions can't be explained simply by pointing to the fact that recanting selves receive a credibility deficit. In many cases, if you subtract the confession, you lose the conviction, too.Footnote ³⁷ So, for instance, even if a defendant's testimony of innocence at a later time is rejected, what is often also needed to convince a jury of his guilt is the veracity of the original confession. Put bluntly, calling the recanter a liar isn't enough for a conviction – the confessor also needs to be regarded as a truthteller.

4. Why?

The first, and perhaps most obvious, reason why the testimony of confessors is privileged is that most of us find it very difficult to imagine ourselves confessing to something we didn't do, and so we conclude that the suspect must be guilty. This is especially compelling when a violent crime is at issue, such as murder. The problem with this is that there is ample psychological research showing otherwise. For instance, in a well-known experiment by Kassin and Kiechel (Reference Kassin and Kiechel1996), 69% of college students who were falsely accused of causing a computer to crash by pressing a key that they were told to avoid signed a confession. When false evidence is presented of guilt, this percentage is even higher, and it is not uncommon for suspects to even come to believe in their own guilt, either fully or partially. A study by Nash and Wade (Reference Nash and Wade2009) used digital editing software to fabricate evidence of participants ‘stealing’ money from a ‘bank’ during a computerized gambling experiment. When presented with this evidence, all of the subjects signed the confession form, with 63% fully internalizing the act and 20% partially internalizing the act. The authors conclude, ‘a combination of social demand, phoney evidence and false suggestion from a credible source can lead a substantial number of people to falsely confess and believe they committed an act they never did’ (Nash and Wade, Reference Nash and Wade2009, p. 629). This can be seen in the case of Michael Crowe, whose sister was stabbed to death in her bedroom. ‘After a series of interrogation sessions, during which time police presented Crowe with compelling false physical evidence of his guilt, he concluded that he was a killer, saying: ‘I'm not sure how I did it. All I know is I did it’. Eventually, he was convinced that he has a split personality – that ‘bad Michael’ acted out of a jealous rage while ‘good Michael’ blocked the incident from memory. The charges against Crowe were later dropped when a drifter in the neighborhood that night was found with [his sister's] blood on his clothing’ (Kassin et al., Reference Kassin, Drizin, Grisso, Gudjonsson, Leo and Redlich2010, p. 15).

Another reason we favor the confessor over the recanter is that false confessions affect the perceptions of others, including eyewitnesses, alibi witnesses, and forensic experts. In one study, 61% of those who had witnessed a staged theft changed their identifications after learning that certain lineup members had confessed.Footnote ³⁸ In another study, only 45% of participants maintained their support of an alibi for a suspect after being told that she confessed to stealing money, a number that dropped to 20% when the experimenter suggested that their support might imply their complicity with the alibi.Footnote ³⁹ What this data shows is that false confessions not only mislead in the first instance, they also beget additional misleading evidence downstream. When this is combined with how counterintuitive false confessions seem to many, including judges and jurors, conditions become optimal for wrongful convictions.

Of particular relevance for our purposes here is that there is reason to believe that racial prejudice or bias is at work in convictions based on false confessions. Andrew Taslitz explains what happens when the interrogation techniques we discussed earlier are used in conjunction with racial discrimination:

Given that 85% of juvenile exonerees who falsely confessed are African American,Footnote ⁴⁰ there is further reason to conclude that racism is a significant factor when looking at why confessing selves are given a credibility excess.

Finally, the practical interests of those most responsible for securing justice often lead them, intentionally or unintentionally, to weigh confessions far too heavily, to disregard exculpatory evidence, and to rely on incredible theories to support their conclusions. This is often seen in the case of prosecutors, who can be blindly driven by a desire to ‘win’. For instance, in a widely ridiculed interview on an episode of 60 Minutes, ‘Chicago: The False Confessions Capital,’ then-State's Attorney Anita Alvarez discussed the case of the ‘Dixmoor Five’ in which DNA evidence ruled out five defendants who had falsely confessed to the rape and murder of a 14-year-old girl. After serving a total of 95 years behind bars, all five were exonerated in 2011, and the Illinois State Police settled in 2013 a civil rights case brought on their behalf for a record $40 million. Moreover, the semen found inside the 14-year-old matched Willie Randolph, who was a convicted rapist with 39 arrests. Despite all of this, when asked about this case in a 2015 interview, Alvarez still said that it was possible that the five defendants raped and murdered the girl, and that Randolph wandered past the field where her body was and committed an act of necrophilia.Footnote ⁴¹ Since the total evidence overwhelmingly tells against this outrageous theory, the most plausible explanation is that Alvarez was motivated by her practical interests, which she thinks will be served by refusing to admit mistakes by her office. Of course, we see this same phenomenon outside the courtroom as well. Climate change deniers will massively privilege one scientist's testimony over that of thousands of others because it suits their purposes, as do voters with the testimony of their preferred candidates for office. When such credibility excesses are driven by prejudices, they clearly result in acts of testimonial injustice.