1. IGNORANCE OF RACIAL DISPARITIES IN POLICING PRACTICES

Racial disparities in policing practices have been thrust back into the spotlight since 2014, when several high-profile cases of black men being killed by police officers (including Michael Brown in Ferguson and Eric Garner in New York City) set off a series of protests across the nation. Since then, activist groups across the United States, such as Black Lives Matter, have worked hard to call attention to the deaths of black citizens at the hands of police officers. These activists often try to place such deaths into the context of a web of related problems faced by many black Americans, such as the over-policing of black communities, the overrepresentation of black children in the school-to-prison pipeline, and the disproportionate rates of incarceration of black men and women.

All of the aforementioned problems are well-documented, but I will focus on an example of the over-policing of black communities.Footnote ² The targeting of minority populations by the police in cities that have “stop and frisk” policies, such as New York City and Chicago, has been given a good deal of attention. In May 2015 the American Civil Liberties Union reported that in Chicago 72% of people stopped over a four month period in 2014 were black, even though black people only make up 32% of the population (ACLU 2015). The findings of the ACLU's report were covered by local and national news outlets, and have led to the passage of reforms to the policy. New York City's stop and frisk policy was found unconstitutional in 2013 for showing similar racial disparities, and the issue was not just widely covered but played a central role in the current Mayor Bill de Blasio's successful campaign for the position. The Department of Justice also released a scathing report about racial disparities in the Baltimore Police Department, including its higher rates of stopping black residents. The DOJ reports that black residents are stopped three times more than white residents of Baltimore, and that “BPD made roughly 44 percent of its stops in two small, predominantly African-American districts that contain only 11 percent of the City's population” (U.S. Department of Justice 2016: 6–7). My point here is that, while these trends are perhaps not common knowledge to all, there has been ample publicly accessible coverage of examples of racial disparities in policing practices.

Many white Americans, however, remain skeptical about the existence of these kinds of racial disparities. For example, a Washington Post–ABC News poll from December 2014 found that 60% of white respondents reported being confident that the police treat white and black people equally, compared with only 20% of black respondents (Balz and Clement Reference Balz and Clement2014).Footnote ³ The poll also found that “There is a 2-to-1 gap between whites and blacks on whether police are adequately trained to avoid the use of excessive force, and slightly larger differences on whether the police try hard to maintain good relations with varied community groups and whether they are held accountable for any misconduct” (Balz and Clement Reference Balz and Clement2014). Importantly, this poll was conducted after the high-profile killings of Brown and Garner by white police officers and the subsequent acquittals of the officers involved in each case. Both incidents were not only widely publicized, but were followed by highly visible protests (especially in the wake of Brown's shooting) led by black citizens. Even after such incidents and the subsequent outpouring of black protest and testimony about their experiences with the police, however, many white people remained skeptical.

In fact, a Pew Research Center poll taken in 2016 showed that 50% of white people polled thought that black Americans were treated less fairly by the police compared with 84% of black people polled, while 41% of whites polled thought that black and white Americans were treated about equally by the police compared with 14% of black people polled (Pew Research Center 2016).Footnote ⁴ This poll was taken 13–22 months of heightened news attention after the polls mentioned above. While there has been some movement toward acknowledging racial disparities in policing, according to these polls, white Americans are still almost three times as likely to believe that black and white Americans are treated equally by the police.

In addition to this lingering skepticism, these white Americans often seem to grasp for alternative explanations of incidents of police violence. For example, the Washington Post–ABC News poll cited above found that “while 6 in 10 whites see the Ferguson [Michael Brown] and Staten Island [Eric Garner] killings as isolated cases, 3 in 4 blacks say they are a sign of a broader problem in the treatment of African Americans by police” (Balz and Clement Reference Balz and Clement2014).Footnote ⁵ The “isolated incident” theory, which often takes the form of asserting that the police officer in question is an exception to the norm, is only one of the many alternate explanations given by skeptical whites. Other examples include, for instance, the theories that black people commit more crimes and so are rightfully targeted by police, that the officers' fears were justified in such incidents, that these incidents wouldn't happen if black people had complied with police orders, and so forth. However, the “isolated incident” explanation is an especially problematic one because it loses sight of the wider context of disparities in policing practices and the justice system that helps us make sense of these incidents of police violence.

As Brittney Cooper points out, the use of the “isolated incident” response to black Americans being killed by the police stands in contrast to the assertion by some, especially in the conservative media, that we are now witnessing a “pattern” of police being murdered by black Americans after the killings of Texas Deputy Darren Goforth in 2015 and of New York Officers Wenjian Liu and Rafael Ramos the prior year. Moreover, this supposed pattern has been attributed to the protests of the Black Lives Matter movement. Some white Americans thus seem to be more willing to see a pattern in cases of police killed by black citizens than in cases of black citizens killed by police, even though the evidence in favor of the latter is considerably stronger. As Cooper asks, “How is it that the random acts of two mentally unstable Black men who had no formal or informal relationship with the Black Lives Matter movement constitute a trend, but the two dozen police killings of unarmed Black citizens again remain a collection of unfortunate but isolated incidents?” (Cooper Reference Cooper2015).

The phenomenon on display in the above case seems to be that of a significant portion of white Americans failing to change their beliefs to reflect readily available evidence. It is precisely this kind of failure that my concept of agential insensitivity gives us the tools to understand. Agential insensitivity thus gives us a way to conceptualize the resistance to evidence that epistemologists of ignorance have observed, and to explain important social phenomena like the one just discussed. I will now examine this concept further.

2. AGENTIAL INSENSITIVITY

Epistemic insensitivity is rooted in the relationship between an epistemic agent and their environment. When an agent is sensitive, their epistemic tools are sufficient for giving them an accurate view of their environment (whether physical or social), and they use these tools effectively in order to obtain this accurate view. When an agent is insensitive, however, either the epistemic tools they possess are insufficient to give them an accurate view of some aspect of their environment, or they do not use them effectively.Footnote ⁶

By epistemic tools I mean both the means by which human beings access evidence (such as senses, attention, memory, and the testimony of others), and the means by which we interpret evidence (such as our linguistic framework and culturally shared narratives and images, as well as “informal patterns of reasoning, current standards of evidence, currently accepted theories and background assumptions, and particular techniques of measuring and investigating”) (Grasswick Reference Grasswick2004: 103). I will call these access tools and interpretive tools, respectively.Footnote ⁷ Our access tools, then, roughly determine what we see, while our interpretive tools determine how we see it. In general, access tools are shared by all humans, and interpretive tools are shared by particular communities.Footnote ⁸

In this section I will distinguish agential insensitivity, which is rooted in the failure of an agent to appropriately use the access and interpretive tools at their disposal, from other forms of epistemic insensitivity that are rooted in the insufficiency of the tools themselves.

Epistemic insensitivity can be caused by the inability of the agent to form an accurate understanding of their environment given the access tools that they have. For example, in his account of knowledge as sensitivity, Robert Nozick mostly examines cases of insensitivity in which a person has a true belief but, because of some quirk about their environment that they are unable to detect using the access tools possessed by humans, their belief state is not appropriately connected to the world. For Nozick, sensitivity is a relationship between beliefs and facts in which beliefs change in response to changes in facts. A person knows that p, on his account, if both of the following subjunctive conditionals are true: “If p weren't true, S wouldn't believe that p” and “if p were true, he [S] would believe it” (Nozick Reference Nozick1981: 172, 176). Nozick calls this particular form of subjunctive sensitivity tracking the truth. He says, “Knowledge is a particular way of being connected to the world, having a specific real factual connection to the world: tracking it” (Nozick Reference Nozick1981: 178). If a belief is insensitive, on the other hand, it does not track the truth in this way.

One of Nozick's examples of insensitivity is that of a man in a tank who is made to believe through direct stimulation of his brain that he is in a tank. This person, according to Nozick, does not know he is in a tank because, while he possesses a true belief to that effect, this belief is not sensitive. “It is not true of him there that if he were in the tank he would believe it; for in the closest world (or situation) to his own where he is in the tank but they don't give him the belief that he is (much less instill the belief that he isn't) he doesn't believe he is in the tank” (Nozick Reference Nozick1981: 177). In this case, it is the combination of having his brain manipulated, and possessing faculties of perception that don't allow him to detect this, that prevents the man from knowing he is in the tank. Even if his access tools are otherwise reliable (i.e., in other circumstances he could form the same belief using the same method, and this belief would be sensitive), they are not capable of detecting the particular environmental quirks involved in this case. His access tools are thus inadequate for giving him an accurate picture of his environment.

Sometimes, however, it is our interpretive tools that prove inadequate for accurately making sense of the world. In this case, the problem is not rooted in what we see, but in how we see it. For example, it is often the case that a culture's shared interpretive tools are disproportionately shaped by members of dominant groups within the culture, since they have more power to develop and distribute these tools than do members of marginalized groups. (Consider, for example, the overrepresentation of white people in books, movies, television, science, and the academy, particularly in gate-keeping roles in these industries.) As a result, the available interpretive tools predominately reflect the experiences of dominant groups, and are often unable to accurately describe the contours of the social environment or the experiences of members of subordinated groups in that social environment.Footnote ⁹ As Miranda Fricker says, this makes it possible for the powerful to “have appropriate understandings of their experiences ready to draw on as they make sense of their social experiences, whereas the powerless are more likely to find themselves having some social experiences through a glass darkly, with at best ill-fitting meanings to draw on in the effort to render them intelligible” (Fricker Reference Fricker2007: 148). Fricker understands cases of the second type to be instances of a distinctive sort of epistemic wrong, which she calls “hermeneutical injustice”.

Fricker uses the case of Carmita Wood to illustrate how women suffering sexual harassment before it was a widely acknowledged phenomenon were victims of hermeneutical injustice (Fricker Reference Fricker2007: 149–50). Wood was an administrative assistant who was sexually harassed by a professor working in the same department. When she quit to escape the abuse, she was unable to collect unemployment because she had no way to express what had been happening to her, and she was dismissed as leaving for personal reasons. When Wood later met up with a group of women who had had similar experiences, however, they were finally able to make sense of what was happening to them and name what they had all been suffering in silence – sexual harassment.

Wood's case is a straightforward example of hermeneutical injustice, and also of insensitivity due to insufficient interpretive tools. Due to the absence of the necessary collective interpretive tools, Wood and the other women were unable to fully understand their own experiences. Moreover, they were unable to make others understand them, and hence unable to effectively seek redress for the wrongs they suffered. Finally, the faulty collective interpretive tools in question proved inadequate for helping any agent involved form an accurate picture of their social environment.

The type of epistemic insensitivity that primarily interests me is caused by an agent ineffectively using their epistemic tools – either their access or interpretive tools – rather than by the inherent inadequacy of the tools themselves. I call this kind of insensitivity agential insensitivity because it stems from the agent's epistemic behavior, being a matter of how they use the tools at their disposal. This behavior is rooted in features of the agent, such as their particular cognitive-psychological profile, experiences, desires, beliefs, emotions, and interests – all of which are influenced by their wider environment, including their social, political, historical, and physical environment. Agential insensitivity is thus a product of the complex interactions between an agent's individual features and their social world. I'll briefly examine two features of agents that can affect their access and interpretive tools – motives and social location – in order to illustrate.

As I mentioned above, research in cognitive psychology has demonstrated that our motives (desires, needs, goals, or interests) often influence our cognition, including our patterns of attention and interpretation of evidence.Footnote ¹⁰ Psychologists have observed, for example, that we have a tendency (known as confirmation bias) to pay more attention to evidence that supports what we want to believe. Confirmation bias can involve selective attention to confirming evidence over other evidence, as well as cutting off searches for evidence prematurely after finding confirming evidence. In each of these ways, we often manage to effectively steer our attention away from evidence in favor of undesired conclusions. Psychologists have also identified a tendency (known as disconfirmation bias) to hold evidence that supports undesired conclusions to higher standards of evaluation than evidence that supports desired conclusions. This cognitive tendency enables an agent to prevent disconfirming evidence from changing their beliefs to the extent that it should by essentially interpreting away the evidence. If an agent works harder to discredit evidence in favor of an undesirable conclusion, then they can feel justified in not changing their beliefs to reflect this evidence.

However, the motives of individual agents do not arise in a vacuum, but can be impacted by their social location. Our location within social hierarchies can directly influence what we are motivated to know and ignore about our social world, for example. As Alcoff says, “social identities can confer motivations, or not, to develop a critical consciousness toward conventional beliefs and values” (Alcoff Reference Alcoff, Sullivan and Tuana2007: 44). In other words, by determining whether or not the status quo benefits us, our social location can motivate us to be more or less skeptical of the status quo. Even the kinds of individualistic motives that cognitive psychologists focus on, such as positive self-evaluation, can still be shaped by our location in social hierarchies. As it becomes increasingly unacceptable to be racist, sexist, homophobic, etc. in the public sphere and in many places in the U.S., for example, the ability of many members of dominant groups to see themselves as good people becomes intertwined with their ability to see themselves as not possessing these socially-disapproved traits. Moreover, many members of dominant groups will experience a tension between self-interested motives to preserve, or at least not disturb, the status quo, and motives to see themselves as a basically good people who don't harm others. These tensions can in turn give many members of dominant groups motives to ignore the ways in which they and their country behave in ways that are racist, sexist, homophobic, etc., and the ways in which status quo social arrangements harm many members of marginalized groups. These motives are both individual and socially shaped.

Moreover, like our interpretive tools, our patterns of attention (an access tool) are affected by social forces. As Gaile Pohlhaus Jr. says, “social position has a bearing on what parts of the world are prominent to the knower and what parts of the world are not” (Pohlhaus Jr. Reference Pohlhaus2012: 717). This is because, in addition to affecting who is in our community and thus what interpretive tools we have at our disposal, our social position affects what is salient to us by determining which experiences we are more likely to repeatedly have. This can lead to the formation of “habits of expectation, attention, and concern, thereby contributing to what one is more or less likely to notice and pursue as an object of knowledge in the experienced world” (Pohlhaus Jr. Reference Pohlhaus2012: 717). These patterns of attention also double back to shape what interpretive tools we develop, a point to which I return in Section 5 below.

3. ATTENTION FAILURES

Agential insensitivity is thus caused by an agent ineffectively using their epistemic tools as a result of the complex interaction between features of that individual agent and their environment, including their social environment. Agential insensitivity can take either of two forms. The first form is that of attention failure, in which an agent uses their access tools ineffectively. The second is that of uptake failure, in which they fail to make appropriate changes or additions to inadequate interpretive tools, or fail to effectively use adequate tools that are already available. I examine each of these forms in more detail in this section and the next, beginning with attention failures.

An attention failure occurs when an agent fails to attend to epistemically relevant and available facts or evidence, and thus does not use their access tools effectively. I take someone to be attending to evidence if they are consciously aware of it, and this evidence is foregrounded in their awareness. An object is foregrounded in this way when one's focus is directed toward it. For example, as I type, I am consciously aware of the sun shining on my table, but my computer screen is foregrounded in my awareness. According to my use of “attention”, then, I would be attending to my computer screen, not to the sun shining on my table. To attend to something in this way involves taking the object of one's attention to be important or relevant enough to direct one's cognitive energies towards. If something is attended to, it is a candidate for further investigation or thought.

One thing to note here is that the facts or evidence that are overlooked in an attention failure must be epistemically relevant.Footnote ¹¹ People cannot attend to every piece of information that crosses their path – there are simply too many. I take what evidence is epistemically relevant to be judged relative to the particular inquiries that an agent is engaged in. A person's inquiries encompass the questions that have become live for them and have captured even a mild bit of interest. There can be different levels of engagement in these inquiries, from thoroughly researching a topic, to a casually and intermittently reading about it. It is possible for evidence that is epistemically relevant to some inquiry not to seem so to a particular agent engaged in that inquiry. An attention failure occurs when evidence that actually is relevant is overlooked, and not simply when evidence that seems relevant from the agent's perspective is overlooked.

In many cases, however, there will still be more epistemically relevant evidence than a person can attend to. I will thus add the further constraint that the epistemically relevant evidence which is overlooked in an attention failure must also be available to the agent. I take a piece of evidence to be available to an agent if they are able to access it without an inordinately large amount of effort, given their faculties, time, and resources. What evidence is available to an agent should be judged based on what the agent would be able to access if they put in a reasonable amount of effort, not what an agent thinks is available to them. What counts as a reasonable amount of effort will be context-dependent. For example, if someone is writing an article on a particular topic, it would be reasonable to expect them to devote more time to scouring all pertinent sources of information than if a person is just trying to be a well-informed citizen. Availability is thus context-relative, since what is considered available to an agent depends on their faculties, resources and projects, but is not determined by what an agent thinks is available to them.

This requirement for the overlooked evidence to be available to an agent marks a major distinction between cases in which our access tools are insufficient and cases involving attention failures. In the former, the agent fails to attend to certain facts or evidence because they are not accessible to the agent. In the tank case, for example, due to the man in the tank's environment and limited faculties, there is no available evidence which will allow him to know he is in the tank. It is not simply that he does not attend to epistemically relevant evidence, but that he cannot. An attention failure occurs, on the other hand, when the agent could attend to evidence but does not. The agent's epistemic tools are sufficient for obtaining accurate information about the environment, but they are not applied to the relevant objects in the right ways.

We can see how attention failures could play a role in the policing disparities case discussed above. Much of the information needed to understand the wider context of individual incidents of police violence is readily available to most skeptical white Americans. Incidents of police violence, protests in reaction to these incidents, Department of Justice reports on racial bias in several police departments (including Ferguson, Baltimore, and Chicago), and national efforts at sentencing reform, for example, are being increasingly reported in mainstream news sources. If one expends a little more effort, much more information is accessible online. It seems, then, that much white ignorance about racial disparities in policing practices is not a matter of skeptical white Americans’ being unable to access this information, but being unwilling to do so.

Moreover, this information is epistemically relevant to questions such as whether black and white Americans receive equal treatment by the police, whether highly publicized incidents of police violence are isolated incidents, and what is the correct way to go about policing. In some cases, these may not be lines of inquiry which some white Americans have engaged in.Footnote ¹² However, these questions are being raised more frequently and more openly as the coverage of these incidents increases, along with the number of public debates about what should be done. As this happens, these questions become live for an increasing number of white Americans.

Moreover, many skeptics of racial disparities are explicitly taking up some of these questions. In some of these cases skeptics are not simply failing to investigate the issue at all, but over-attending to misleading information. Doing so can lead them to cut off their searches for relevant information prematurely by causing them to feel like they have done sufficient research.Footnote ¹³ When looking to explain episodes of police violence, many skeptical white people focus, for example, on any crimes or misbehavior the victims of police violence may have committed, and how dangerous and stressful policing is. Sometimes attention is simply diverted by issues that are not directly relevant to the question at hand as well.

Let's return, for instance, to the example above of the attention paid to the few incidents in which police officers have been the victims of targeted killings. This can be seen most recently when a sniper killed five police officers at a Black Lives Matter protest in Dallas in July 2016. The sniper was unaffiliated with Black Lives Matter, but Black Lives Matter came under fire for supposedly encouraging anti-police sentiments that could lead people to want to injure or kill police. For example, the National Association of Police Organizations (2016) asserted that “We are currently in the midst of war on cops”, and blamed both protesters and politicians who support them (“senseless agitators and gutless politicians”) for the deaths of the Dallas officers. Many have responded to the supposed war that Black Lives Matter is waging on police by voicing their support for the police using the counter-slogan “Blue Lives Matter.” These kinds of reactions serve to divert attention away from police violence and the responding protests, and toward the police and their struggles. In this case, attention is not only not paid to evidence in favor of the existence of racial disparities in policing, but is diverted elsewhere.

The lack of attention paid to, and active diversion of attention away from, available information about racial disparities in policing by many skeptical white Americans suggests that attention failures may be playing a role in their skepticism.

4. UPTAKE FAILURES

Agential insensitivity can also involve uptake failures. An uptake failure occurs when an agent has attended to epistemically relevant facts or evidence, but fails to change their beliefs in response to this evidence because they fail to use their interpretive tools effectively. The problem here is not that evidence is not accessed, but that it does not have the impact it should on the agent's beliefs because it is misinterpreted, explained away, not followed to its conclusions, or the like. It is also not the case that the agent is simply missing the interpretive tools they need to understand the evidence correctly, as in cases of hermeneutical injustice, but that they are not effectively using the interpretive tools that are appropriate and available to them (in the sense described above). When an uptake failure occurs, the agent could use existing interpretive tools to understand the evidence correctly, but does not. This can occur because they fail to correctly apply the tools they have, or because they fail to update their interpretive tools in light of new evidence. I will focus on the latter for the purposes of this paper because it seems the most relevant for understanding the racial disparities in policing.

Ideally, interpretive tools help us understand our environment and experiences. However, in order for them to play this role, they have to be appropriately updated in response to an agent's circumstances. Pohlhaus Jr. sums up this process as follows.

This recalibration can take the form of participating in the development of new interpretive tools, which is necessarily a communal activity, or searching out already developed tools that are not in one's personal epistemic toolbox. When an agent fails to recalibrate their interpretive tools in this manner, this constitutes an uptake failure.

As in the examples of hermeneutical injustice mentioned above, this recalibration does not always happen when the experiences in question are primarily those of marginalized groups. However, it needn't be the case that new interpretive tools have not yet been developed, as in Fricker's sexual harassment example. It sometimes happens that such tools are devised by, and disseminated within, a marginalized group, but that they fail to be taken up by the dominant culture. In these cases, the interpretive tools necessary for understanding a particular set of experiences are available, but many members of dominant groups are not using these tools.

Pohlhaus Jr., for example, examines this failure of members of dominant groups to adopt relevant interpretive tools (or, as she calls them, epistemic resources) developed by members of marginalized groups. She does so using the example of the white jury's failure to understand the testimony of Tom Robinson, a black man wrongfully accused of raping an impoverished white woman, in Harper Lee's novel To Kill a Mockingbird. In particular, Pohlhaus Jr. focuses on Robinson's claim that he stopped to help the woman with chores because he felt sorry for her. The jury interprets Robinson's testimony not as an expression of genuine sympathy for a poor young woman struggling to take care of her family on her own, but rather as an assertion of his superiority to her.

Pohlhaus Jr. claims that the white jury failed to understand Robinson's testimony because “it has not begun to use epistemic resources that would allow it to know what Robinson knows: those things that appear saliently in the experienced world of a black man in 1930s Alabama” (Pohlhaus Jr. Reference Pohlhaus2012: 727). However, she points out that such resources aren't simply nonexistent, as in cases of hermeneutical injustice. After all, Robinson is able to understand and describe his own experiences, unlike the victims of sexual harassment in Fricker's case. The problem is also not that it is impossible for the white jury to use the resources developed by the black community to understand these experiences, but that they refuse to do so – and moreover they have the power to refuse to do so. Pohlhaus Jr. says, “there is nothing forcing them to use faulty epistemic resources; rather, they lack such resources due to a prejudice against taking seriously the experienced world outside of white men and a refusal to enter into truly cooperative interdependence with knowers situated outside dominant social positions” (Pohlhaus Jr. Reference Pohlhaus2012: 725). It is their refusal to acknowledge the fact that a black man like Robinson may understand and experience something that they do not, and to work with him to develop the ability to understand that experience, that prevents the white jury from adopting new interpretive tools. This is not inevitable, argues Pohlhaus Jr., but willful. This suggests that the case is one of uptake failure, not hermeneutical injustice.

We can also see how uptake failures could play a role in the policing disparities case. Many times the relevant problem seems not to be one of accessing the right information, but rather of how information is interpreted once it is attended to (i.e., whether connections are made to other information, what these connections are taken to be, and so forth). The “alternative explanations” problem discussed above seems to be rooted in just this kind of failure. In this case, it is not that skeptical white Americans are not attending to incidents of police violence, but that they are interpreting them as, for example, the isolated actions of one “bad apple” police officer, or as reasonable reactions on the part of police officers to a frightening/non-compliant individual, rather than as part of a wider pattern of racial disparities in policing. Furthermore, the split between the interpretations of this information favored by black and white Americans suggests that these incidents are being viewed using different interpretive tools.

The explanations offered by many white Americans of incidents of police violence rely on interpretive tools that make sense of their own experiences with the police. Due to the very disparities discussed above and the high degree of segregation in much of the United States, many white and black Americans have radically different experiences of the police. White communities, especially middle- and upper-class ones, often have a lower police presence and experience the police as being fair and helpful. Aside from an occasional traffic ticket, an upper-/middle-class white person may not have had any interactions with the police at all, giving the impression that police only stop a person when they do something wrong. These experiences with police are reinforced by a white-dominated media that portrays crime as what happens elsewhere on the nightly news, depicts the police as the heroes who track down criminals on television shows, and emphasizes any misbehaviors that a victim engaged in during or prior to an incident of police violence. All of this can contribute to making dominant interpretive tools seem appropriate, and the interpretations they enable seem plausible to skeptical white Americans.

Many black Americans, on the other hand, have to grapple with experiences of the police like the ones discussed above – experiences of being targeted and stopped without reason, of being watched by police who are always patrolling their neighborhood, and of having friends, family members, or neighbors who have been harassed, injured, or killed by police. It should be no surprise, then, that alternative interpretive tools would be developed within targeted black communities in order to express and make sense of their experiences, and that these tools would give them a different view of incidents of police violence, and also of disparities in policing.

Importantly, however, there are readily available interpretive tools that could enable skeptical white Americans to better understand the occurrence of police violence, and its connection to wider racial disparities in policing. For example, there is the language that has been invented to talk about racial disparities in policing and the justice system (i.e., “black lives matter,”Footnote ¹⁴ “say her name,”Footnote ¹⁵ police brutality, driving while black, prison-industrial complex, school-to-prison pipeline, etc.). Moreover, nation-wide protests have served not just to call attention to incidents of police violence, but to provide the tools for understanding these incidents. For example, by portraying Michael Brown as a child who deserved protection, and whose death was worth mourning, protesters are providing counter-narratives to dominant portrayals of Brown as a dangerous and frightening criminal, and as having provoked his own death. Using phrases like “black lives matter” provides another tool for understanding this same vision of victims like Brown. Protesters and activists also often draw connections between incidents of police violence, and between these incidents and wider trends of racial disparities in policing and the justice system, that might not be visible using dominant interpretive tools.

Finally, these very same protests put pressure on skeptical white Americans to update their interpretive tools. By repeatedly providing different interpretations of events and people and highlighting patterns that cannot be accounted for using dominant interpretive tools, widespread and continued protests reveal the limitations of these tools and call on those who use them to either recalibrate them or to adopt new tools. The failure to do so in light of this pressure constitutes an uptake failure.

5. CONCLUSION: ADDRESSING AGENTIAL INSENSITIVITY

In this paper, I developed a taxonomy of epistemic insensitivity that allowed me to identify a form of epistemic insensitivity I call agential insensitivity, and distinguish it from other forms of insensitivity. According to my taxonomy, all epistemic insensitivity involves the failure of an epistemic agent to change their beliefs to reflect available epistemically relevant facts or evidence, but the different causes of this failure mark distinct forms of insensitivity. In cases of agential insensitivity, this failure is due to the agent not effectively using the epistemic tools at their disposal; whereas in the other cases of epistemic insensitivity I examined this failure is rooted in the insufficiency of the agent's epistemic tools. I further identified two manifestations of agential insensitivity, namely attention failures and uptake failures. These forms of agential insensitivity occur when an agent fails to properly use their access tools to attend to epistemically relevant evidence that is available to them, and when they fail to properly use the interpretive tools at their disposal in order to accurately make sense of evidence in their possession, respectively.

My account provides a framework for understanding some of the ways that agents can be unwilling or unable to make proper use of the evidence at their disposal, and remain ignorant as a result. One benefit of my framework is that it helps shed light on why different strategies will be required to address agential insensitivity than other forms of epistemic insensitivity, and also on the unique challenges that will likely be faced when attempting to do so. While giving a detailed account of possible solutions to agential insensitivity is beyond the scope of this paper, I would like to end by gesturing at some implications that my account has for thinking about how agential insensitivity can and cannot be remedied.

Before I jump into this, I would like to briefly discuss when and why agential insensitivity is problematic and in need of remedying. Not all cases of agential insensitivity seem equally worrisome. For example, say I am casually engaged in inquiries about marine life – it is a subject that I periodically enjoy reading about, but I'm not trying to become an expert. If I choose not to search out information that is relevant to my inquiries about marine life in spite of this information being easily available to me on the internet, this seems innocuous even though it is an attention failure.Footnote ¹⁶

However, the agential insensitivity witnessed in the policing disparities case is not innocuous. This is because agential insensitivity about racial disparities in policing practices often plays into a wider pattern of collective white ignorance about these practices, and about racist social structures in general.

First, the agential insensitivity in the policing disparities case is more epistemically harmful. In one sense, any case of agential insensitivity comes with some epistemic cost, since an opportunity to gain knowledge is lost. However, it doesn't seem like all cases of agential insensitivity come with the same epistemic cost.Footnote ¹⁷ For example, an attention failure about marine life is more localized than one about policing disparities. What I mean by this is that attention failures about marine life are not likely to connect, reinforce, or lead to agential insensitivity about related phenomena, like oceans or other kinds of animals. Attention failures about racial disparities in policing, however, can encourage and be encouraged by attention failures about other kinds of racial disparities. Thus, attention failures about racial disparities about policing will usually be part of a broader agential insensitivity about the domain of racial disparities in general. This is because the roots of agential insensitivity about racial disparities about policing are systemic, being entwined with wider structures of collective white ignorance, such as those discussed by epistemologists of ignorance. Mills, for example, argues that on matters related to race, white people's thinking is characterized by “a particular pattern of localized and global cognitive dysfunctions (which are psychologically and socially functional), producing the ironic outcome that whites will in general be unable to understand the world they themselves have made” (Mills Reference Mills1997: 18). In Section 2, I explained some of the motivational and cognitive aspects of this pattern. People are not just looking away on a whim, as in the marine life case, but have ready-made excuses to look away. They have ready-made narratives to explain away evidence. Thus, one reason that agential insensitivity about marine life is less epistemically harmful than that about disparities in policing is that the epistemic costs are limited, and will not ripple out to encourage and reinforce more ignorance over a wider domain.

Regardless of whether the attention failures we see in this case are epistemically bad or not, they are also distinct from the marine life case in that they are ethically/politically bad. A failure to attend to the abundant information available to me about marine life is not just epistemically innocuous, but ethically/politically innocuous, since it has no harmful social consequences. In the racial disparities in policing case, however, the attention failures of individuals play into a pattern of collective ignorance about these disparities that has harmful consequences for black Americans, as I demonstrated in the first section. Because of this, it is important to seek remedies for harmful types of socially-supported agential insensitivity like this one – whether they are deemed epistemically harmful or not.

Having established what kinds of agential insensitivity require remedy and why, I will now examine what these remedies could – and could not – look like. It should be clear from my above argument that different forms of insensitivity will, in general, require different remedies. In cases of insensitivity involving inadequate epistemic tools, changes need to be made to the tools themselves, or to environmental conditions that are beyond the agent's ability to track. In the sexual harassment case, for example, the development of new interpretive tools can restore the sensitivity of the agent. With the concept of sexual harassment at their disposal, Wood and others have been able to develop a more accurate view of occurrences in their work environment. Similarly, if we removed the man from the tank and set him free to roam the lab, his beliefs about his situation would be set aright fairly quickly. This is because his access tools would be sufficient to give him an accurate picture of his new environment.

This is not the case with agential insensitivity. This form of insensitivity can't be remedied by simply moving the agent to a new setting or introducing them to new interpretive tools. This is because in cases of agential insensitivity, there is something amiss with the agent and their epistemic behavior, and not just with their epistemic tools. The insensitivity occurs here because they fail to use available and appropriate tools in a way that would enable them to acquire knowledge about their environment.

Agential insensitivity's complicated roots make it much less straightforward to correct. For one, the ability to push others to correct their ignorance when this ignorance is rooted in agential insensitivity will be limited. Simply providing more information to agentially insensitive subjects will not always work, due to their ability to divert attention away from this evidence or interpret it away. It is also particularly difficult to convince a person to adjust their patterns of attention or adopt new interpretive tools because their current tools often feel suitable for making sense of their own experiences. Evaluations of evidence will always take place against a background of current beliefs, assumptions, experiences, and collective interpretive tools, which usually go unquestioned and which we may be reluctant to change. As Alcoff says, “What I already know and believe will have a privileged place in my judgments by affecting my determinations of coherence, consistency, and plausibility, and this is in fact good epistemic practice. … the law of entropy operates in the realm of belief: we tend toward conserving the beliefs we have until forced to call them into question” (Alcoff Reference Alcoff, Sullivan and Tuana2007: 44–5). Unless an agent feels the failure of their own epistemic tools acutely enough, they will likely not be motivated to change them. Another person can issue challenges to these tools, as I argued that continuing protests have the potential to do, but such challenges may well be ineffective.

Attention and uptake failures can also reinforce each other in a way that complicates this process even further. As Pohlhaus Jr. argues, for example, it is difficult to call attention to the parts of the world that those in dominant social positions do not experience, because they are not using the interpretive tools that would make sense of these parts of the world. But it is also difficult to convince them to update their interpretive tools because they do not experience and attend to the parts of the world that warrant the use of new interpretive tools. She says,

Attention failures thus make it more difficult to see the need to update one's interpretive tools. Uptake failures in turn make it more difficult to see the need to attend to different evidence. Finally, these problems are all exacerbated when the parts of the world that are being called attention to, and the interpretations being offered of them, are ones that make the agent uncomfortable and challenge their positive conception of themselves and their world.

We see these obstacles to the correction of agential insensitivity at work in the case of policing disparities, for example. As I observed in the last section, there is often a stark difference between white and black Americans’ typical experiences with the police. Gaining knowledge about racial disparities in policing is difficult and unlikely if one is relying on interpretive tools and patterns of attention based on typical white experiences. On the other hand, the tools developed through typical black experiences would seem suspicious to many skeptical white Americans, who haven't shared in the experiences in question. Finally, many skeptical white Americans are strongly motivated to believe that racism, including racial disparities in policing, is a thing of the past. This belief allows them to justify not taking action to rectify racial disparities, and rest comfortably in the knowledge that they and their country are good. This can cause many skeptical white Americans to not only be suspicious of the tools developed through typical black experiences, but to be motivated not to use these tools because they do not want to see the version of the world that these tools make visible.

Given this understanding of the obstacles that attempts to address agential insensitivity will face, what kinds of strategies might actually work? Again, laying out a detailed program for addressing agential insensitivity is beyond the scope of this paper. However, I would like to briefly examine some of the conditions that would need to be satisfied in order for such strategies to work.

In order to address agential insensitivity, the agent's epistemic behavior will have to change. Some philosophers thus focus on what epistemically responsible behavior looks like, or what epistemic virtues will need to be cultivated to combat such phenomena. José Medina, for example, gives a sophisticated account of the epistemic virtues, including humility, curiosity/diligence, and open-mindedness, that subjects will need to be responsible epistemic agents in the face of active ignorance (Medina Reference Medina2012). However, as we saw above, an agent's epistemic behavior is rooted in many factors such as an individual's motives, emotions, experiences, and background assumptions. These factors are in turn shaped by an agent's environment, including their social environment. As a result, solutions to agential insensitivity that call on the agent to change their epistemic behavior while ignoring these roots will likely be ineffective. Because of this, I would like to focus on the individual and social conditions that would have to be in place in order to cultivate the kinds of epistemic virtues that philosophers such as Medina have discussed.Footnote ¹⁸

If we were to imagine a scenario in which the agential insensitivity of one of the skeptical white Americans from the previous case was resolved, for example, what would have to change? In this scenario, the white American would have to, at minimum, trust the testimony of black citizens impacted by the policing practices in question (or their activist advocates), seek out and take seriously reliable reports about current racial trends in policing practices, and adopt the necessary interpretive tools to make sense of this testimony and data. However, the conditions for making these changes would also have to have been developed in our imagined scenario. These conditions would likely involve giving the skeptical white American repeated experiences that are not race/class-typical (say, living or working in different communities, having trusting relationships with people of other races, etc.). They would likely also involve a shift in internal attitudes, such as changing their motives (e.g., so that they are not invested in believing that racism is a thing of the past or that their sense of self-worth is grounded in fighting against racism rather than in merely appearing not to be racist). In other words, our imagined scenario would require significantly changing both the agent and their environment.

While these are conditions that the skeptical white American could work to create even in the presence of unjust social structures, realizing these conditions would be made easier through structural changes, such as changes to residential patterns and educational curricula. In other words, perhaps the easiest way to create our imagined scenario is to create a world that fosters the appropriate kind of epistemic behavior.Footnote ¹⁹ Even attempted individually in unchanged social structures, creating the conditions for making these changes requires the cooperation of others and is thus a necessarily social activity. As Pohlhaus Jr. says, just learning to use the interpretive tools developed by others “is not a disembodied, apolitical exercise … Becoming attuned to when and how to use epistemic resources requires engagement with practitioners skilled in their use, placing oneself in encounters where it makes sense to use them, making mistakes and being corrected” (Pohlhaus Jr. Reference Pohlhaus2012: 721). Laying the conditions for changing epistemic behavior in this case will require creating communities in which this behavior can develop and flourish.

Understanding agential insensitivity thus helps shed light on why strategies that help correct some forms of ignorance will not necessarily work to correct ignorance rooted in agential insensitivity, and also calls for thinking of creative new ways to address agential insensitivity. This will likely require both interventions aimed at changing people's emotions, desires, motives, and so forth at the individual level, as well as interventions aimed at changing social-political structures.

This call for new solutions becomes more urgent when we consider the role agential insensitivity plays in such phenomena as the disproportionate skepticism of white Americans about racial disparities in policing practices and the justice system. Ignorance such as this is not just an abstract epistemic problem, but one with concrete social and political consequences. Because of this, it is even more important to understand the contours and causes of agential insensitivity, and how our tactics must change to address this kind of insensitivity. It is my hope that my account aids us in these endeavors.Footnote ²⁰