Monday, November 28, 2016

Racism and a presumption of credibility

What I want to try to argue for here is a presumption of credibility for claims of racism.

For the purpose of this post, let’s define a presumption Px as a deliberative advantage in favor of a finding that x is true based on supporting justification for Px, such that, in a dispute involving x, the person trying to show that x is false (x-opponent) bears the burden of proof and, if x-opponent offers no evidence, the person trying to show that x is true (x-proponent) succeeds. A presumption is a deliberative advantage, but not a required conclusion; an initial deliberative advantage may be overcome by sufficient evidence to the contrary.

Presumptions usually are based on some supporting justification. A very common example is the presumption of innocence. This presumption operates in favor of a finding that the x-proponent, a criminal defendant, is innocent and the x-opponent, the prosecutor, bears the burden of proof. The supporting justification is the value in American jurisprudence that it is better for nine guilty men to go free than for one innocent man to be wrongly convicted.

Here my supporting justification for a presumption of credibility for claims of racism is that, because whether one recognizes racism depends crucially on one’s first-personal perspective, the person claiming racism should be deemed credible, absent sufficient evidence to the contrary.

Claims of racism generally present unique challenges for proof. Consider the new movie, directed by Jeff Nichols, based on Loving v. Virginia (1967).[1] The trailer for the movie (here) includes a scene where Virginia state police officers barge into the Lovings bedroom and, when Mildred Loving explains that she’s Richard Loving’s wife, one officer responds, “that’s no good here.” The scene portrays well the racist attitudes that may have been prevalent at that time in certain places.

One thing to clarify here is that Virginia’s law and enforcement of its law in the Loving case is a separate wrong from the officer’s attitude toward the Lovings in the scene mentioned above. The case involved Virginia’s miscegenation law, but my focus here is the officer’s attitude in the movie scene.

If we were examining a charge of racism against the officer, it would be difficult to prove. You often can’t find racism from a professor’s armchair or a judge’s bench. In real life, the scene is not reconstructed for us exactly as it was. Instead, you have the words or actions, the features of the surrounding circumstances, and the accused racist before you in his best Sunday clothes (see Bannon).

Even if the scene was reconstructed for you exactly as it was, from a detached third-personal perspective, unless you yourself have experienced similar attitudes, you may not see racism. We may see the identical scene, but our perspectives are shaped by our prior observations and experiences. It is as if we see the world through our own theoretical lens, shaped by our past, which picks up certain features and ignores others.

But, for the person experiencing racism, the attitude expressed and the wrong left in its wake is as real as real can be. It's not about the exact words spoken or any particular feature of the circumstance. If we examined the words or actions, they could be interpreted as innocent. If we examined the features of the circumstance, they may seem morally insignificant. If we saw the person again in a different setting, he may appear to be the farthest thing from what we would expect from a racist. From the first-person perspective of the target of racism, the attitude conveys to its target: you don't belong here or there is something deeply wrong with you. From those with other perspectives: you’re just making the whole thing up.

If this is accurate and racist attitudes are seen differently from first-personal and third-personal perspectives, then this seems to be a good reason to presume the credibility of a first-personal claim of racism.

Again, a presumption of credibility is not proof of credibility or proof of racism. A presumption is usually defeasible. It simply places the burden of proof on the other party to prove otherwise.

Three objections come to mind. I’ll take them in the order from easiest to the most difficult, leaving the third for further discussion in the comments.

First, what about false claims? A presumption is not proof and can be overridden with evidence to the contrary, including evidence that the claim was false.

Next, what about innocent “racists”? People are accused of racism all the time and they’re merely good people who were brought up to view the world a certain way; they don’t intend any harm. As stated by Shelby and others, I don’t think racism requires specific intent. Racial bias and prejudice may be “subtle, implicit, or even unconscious.”[2] Individual attitudes or behaviors usually are the product of institutional racism and, first and foremost, it is our institutions that require critical examination and reform.

Finally, the nature of racism and the fact that racist attitudes are seen differently from first-personal and third personal perspectives, for some, may not seem like sufficient justification for a presumption. My response for now is this: against the backdrop of our nation’s history of slavery, segregation, internment, and dispossession, the fact that the targets of racism are in a unique position to see the wrongness of racism, which may not be visible to others, again seems a good reason to give them the benefit of the doubt.

Chong Choe-Smith
Department of Philosophy
Sacramento State

[1] Loving v. Virginia (1967) 388 U.S. 1.
[2] Tommie Shelby, “Race and Social Justice” (2004) 72 Fordham L. Rev. 1697, 1706.


  1. It looks like the supporting justification for the presumption of innocence conflicts with your supporting justification for the presumption of credibility of claims of racism. That's not an objection to the latter, but I'm curious which you think is weightier.

    Maybe the presumption of innocence only applies when legal punishments are on the table, but that doesn't seem plausible. Informal social punishment is pretty serious, even if it's perhaps less serious than formal legal punishment. If that's right (and I'm not sure it is), then if it's better that 9 guilty people go free than for 1 innocent person to be wrongly convicted by a court of law, maybe it's only something like 4 guilty people who should be left alone than for 1 innocent person to be wrongly convicted by public opinion.

    But that still supports *some* presumption of innocence. How should we think it stacks up against someone's charge of racism?

  2. The conflict is not readily apparent, but, I suppose, if we’re thinking generally about proceeding with charges of racism cautiously, I still would place the burden for any risk of error on the one being charged than the one making accusations.

    The conflict initially is not readily apparent because the presumption of innocence only applies in criminal cases or when criminal punishments are at stake and claims of racisms are adjudicated generally in civil cases and or other social contexts. The justification, as stated, is incomplete. The justification for the presumption of innocence is that it is better for nine guilty men to go free than for one man to be wrongly convicted because criminal punishments involve the deprivation of life or liberty, which are among the most fundamental interests possessed by citizens of a liberal democracy. This criminal presumption of innocence, by the way, has not been extended to apply outside the context of a criminal trial (I actually wrote a published opinion on this issue when I worked for the court of appeal, People v. Beeson (2002) 99 Cal.App.4th 1393, which held that a defendant in a civil involuntary commitment proceeding was not entitled to a presumption of innocence instruction—although, when writing the opinion, I thought he ought to be entitled to an instruction, his life and liberty are at stake, but all the precedent went the other way: courts are very reluctant to extend the rights given to criminal defendants in criminal trials to defendants in even quasi criminal cases—see Guantanamo Bay!). A person charged with racism may face significant social stigma, but his life and liberty are not directly or obviously at stake.

    The only time that I think the presumption of credibility for claims of racism would be in conflict with the presumption of innocence is in a criminal case involving a hate crime enhancement. There, the presumption of innocence would prevail and the jury would have to find that the defendant intentionally targeted a person based on race or other protected classification beyond a reasonable doubt. In other words, the burden would be on the prosecutor and not the defendant to prove the facts necessary to support the enhancement.

    Otherwise, I don’t see a conflict. If we’re thinking generally about proceeding with charges of racism cautiously, any social harm experienced by the one who stands accused seems far less significant than the social harm of racism (or letting racism go unpunished or being a society that doesn’t take racism seriously). There is a risk of error here and someone has to bear it. It seems, given the backdrop of racism in this country, the person who should bear the burden for any risk of error is the one being accused. He's not entitled to a presumption of innocence, but he can rebut the presumption by showing that the accusation is false.

    Also, when speaking of racism, I’m usually thinking not of individual defendants, but of systemic racism or racist norms entrenched in our social, political, or legal institutions. When those who experience racism day in and day out say that certain institutions treat them differently because of their race, I think the presumption should apply and those claims of racism should be believed. Many who do not have anything like the same kind of experience may jump to the conclusion that groups like Black Lives Matter have no business rioting in the streets, they’re overreacting, they’re making the whole thing up. When claims of racism are made about certain institutions by those with first personal experience, we should evaluate those institutions and, if necessary, do the very hard work of dis-entrenching any racist norms or stereotypes.

    1. But the law is an ass. Again, it doesn’t seem even remotely plausible to think that a presumption of innocence against accusations applies only when criminal punishments are at stake. And it makes it too easy to avoid the conflict. You’ve just defined it away.

      In fact, the only reason to think a presumption of innocence applies in criminal cases is because of a more general principle that the imposition of any form of punishment is serious moral business and so requires an accuser to meet a justificatory burden. Again, maybe informal social punishments are, in general, less serious than formal legal punishments, but I don’t even think that’s obvious. Mill, for example, devoted much more time to worrying about the former than about the latter, like here:

      “Society can and does execute its own mandates: and if it issues wrong mandates instead of right, or any mandates at all in things with which it ought not to meddle, it practises a social tyranny more formidable than many kinds of political oppression, since, though not usually upheld by such extreme penalties, it leaves fewer means of escape, penetrating much more deeply into the details of life, and enslaving the soul itself.”

      Yes, being killed or (literally) imprisoned is bad, but so is this. Any form of punishment is bad. It’s to do things to people that aren’t ordinarily ok to do anyone. Therefore, doing them requires special justification, which is a burden the accuser should discharge if the punishment is to be legitimate.

      “Protection, therefore, against the tyranny of the magistrate is not enough; there needs protection also against the tyranny of the prevailing opinion and feeling; against the tendency of society to impose, by other means than civil penalties, its own ideas and practices as rules of conduct on those who dissent from them….”

    2. I guess I’m persuaded by the supporting justification for a presumption of innocence in criminal cases and I think the justification doesn’t apply in noncriminal cases. The justification basically being to level the playing field for a criminal defendant against a state far more powerful that he when life or liberty hangs in the balance.

      Here are two related points. First, I probably didn’t make this clear but sufficient supporting justification for a presumption is not the same as sufficient proof for a proposition. Whether a supporting justification is persuasive or not will depend on our background assumptions, here, the background assumptions of our system of law. Not everyone, especially those who don’t accept the background assumptions of our system of law, will accept the supporting justification.

      Second, generally, no one gets a deliberative advantage. Everyone must prove his or her own case. Your comments seem to be suggesting a broader presumption of innocence in all cases (a deliberative advantage for the accused). We can also interpret your comments as rejecting a presumption of credibility in cases involved racist accusations (no deliberative advantage for the accuser).

      I can see how, in criminal cases, the accused would need some help to level the playing field. I don’t see how the same justification applies in a noncriminal setting. There may be much at stake (social punishment); every accusation carries some consequences. Nevertheless, the general rule applies: no one gets a deliberative advantage. I guess I would need to hear more to be persuaded to extend a presumption of innocence to all cases.

      I can also see how, in cases involving racist accusations, the accuser would need some help to level the playing field (the point of my main post). If we are persuaded, we allow another exception to the general rule.

      You may just disagree with the idea of presumptions (giving deliberative advantages to anyone) generally: either a presumption of innocence in all cases or none at all.

  3. Thanks Chong, obviously a very timely piece, but also very provocative and interesting.

    I think there is something missing from this argument. We can agree that a phenomenon can only be detected from a particular perspective without agreeing that it can be reliably detected from that perspective. So we need to add that people who feel or believe they are experiencing racism are indeed likely to be experiencing it.

    This would be a safe assumption in lots of environments. For example, if racist social interactions were the predominant kind of social interaction, then this is something you can claim safely even without any racism detection abilities.

    But when racist interactions are relatively infrequent, then it is a statistical fact that even people with very good racism detectors are wrong more often they are right. As you know, this is the base-rate problem which we try to teach our inductive logic students to avoid using Baye’s rule of belief revision.

    So we also need to add the assumption that racist social interactions are frequent enough that false positives are relatively infrequent. That is something that will be true in some places and untrue in others. I think it is probably untrue, for example, at Sacramento State. Unless we employ a definition of racism that is so permissive that it is pointless to accuse anyone of it, the vast majority of mundane social interactions on campus are not racially motivated and for any randomly selected one that is perceived as such, it is probably not. Of course, this isn’t to deny that there are tons of racist incidents. It’s also not to deny that there are specific incidents that are unmistakably racist, though these are the ones that are so flagrant that they undermine the initial assumption that their reliable detection requires a particular perspective. And, even if I am wrong about SacState, the fact of environmental variation would cast doubt on the idea that we can universally presume accusations of racism to be credible.

    There are other things that militate against reliability. One of them, which I discussed in a recent post, is the Baader-Meinhof Phenomenon.. To use Bayes’ rule at all we must have a belief about the prior probability that a particular event will occur, in this case a racist interaction. Unfortunately, most people do not get these from meaningful statistics, but from the frequency with which such things are reported on the news. When the media are especially interested in racist events- such as now in the wake of Trump’s election- it will produce an exaggerated sense of the frequency of these interactions. Hence, this will erode everyone’s detection ability in the direction of more false positives.

    Finally, I have to admit I have never understood this business of it being intrinsically worse to convict one innocent person than to let 9 guilty people go free. To me this only makes sense on the basis of certain quite plausible views about the secondary and tertiary effects of a system that is run differently. And I think these consequentialist arguments are actually very compelling. One of these probable effects is that when we presume the accused to be guilty, it will tend to increase the frequency of false accusations and consequently the rate of false convictions. The reason for this is that so many of these accusations (just as with attempted rape, medical malpractice, etc.) occur when the most significant evidence available is the testimony of the plaintiff and the defendant. Under those circumstances, if I am properly motivated to cause someone harm, I have a very serious incentive to make a false accusation. And, by definition, these are just the sorts of cases for which your reply to your first objection is not adequate.

    Finally, a question for you: Would you be willing to generalize your suggestion to other areas, such as sexual harassment? Your perspectival argument seems to apply here a fortiori.

    1. It seems your comment falls into the third category of objections: the justification is insufficient without an additional assumption that racism can be reliably detected from a first personal perspective. My initial reaction is to reiterate that a presumption is a deliberative advantage (establishes from the outset the preliminary question of who bears the burden of proof) and not a required conclusion. We can presume credibility without an assumption or evidence that racial social interactions are frequent enough in the context in question to avoid false positives. That’s the point of a presumption: placing the burden on the other party to poke holes in the claim, maybe including that actual instances of racism are so infrequent in the context in question to render such perceptions or judgments unreliable. So I would be reluctant to add the assumption that racist social interactions are frequent enough to avoid false positives. Showing that such things (that racist social interactions occur) or the frequency of their occurrence brings us right back to the reason for the presumption (those who experience racism will say instances of racism are frequent and those who do not experience racism will say that instances of racism never or rarely occur).

      Measuring racist social interactions is not like measuring how many people walk through the doors of Mendocino Hall. (Maybe you have access to reliable statistics on racism that I don’t know about.)

      My next reaction is that, even if we added the assumption that racist social interactions are frequent enough that false positives are relatively infrequent, I would say that your generous assessment of Sac State may have been too hasty. This is not because Sac State ranks high in racist social interactions, but because racism is so pervasive (that it may be the norm) in most places. The assumption may apply almost everywhere in America.

      What comes to mind is Peggy McIntosh’s article, “White Privilege: Unpacking the Invisible Knapsack,” where she writes that “white privilege is like an invisible weightless knapsack of special provisions, maps, passports, codebooks, visas, clothes, tools, and blank checks.” If we define racism as unjustified discriminatory treatment on the bases of race (whether socially constructed or otherwise), the concept of white privilege suggests that racism occurs all the time, everywhere. Some people are advantaged and other people are disadvantaged all the time, everywhere.

      Some people can “turn on the television or open to the front page of the paper and see people of [their] race widely represented” and I would add positively and not only stereotypically. Some students can walk into classrooms and see instructors of their race widely represented. Some students can write a paper and make a couple grammatical mistakes and not have the mistakes attributed to their Hispanic heritage or the fact that English was their second language. In my short time at Sac State, I’ve heard from students who have been targeted by law enforcement in Sacramento and who have been assumed to be low performers in classes because of their Hispanic heritage. This is not a problem with Sac State; this is problem all the time, everywhere.

    2. I honestly have never given this much thought to the presumption of innocence. I gave it as an example here because people generally are familiar with it.
      There is usually some supporting justification for a presumption. There are certain values that underlie our system of law, including the values of life and liberty. There is also the value that individuals should be protected against arbitrary and capricious abuses of power by due process of law or the rule of law. A presumption of innocence instruction is one part of the bundle of due process rights a criminal defendant enjoys in the context of a criminal trial. Presumptions in the law sometimes operate to level the playing field. One may think that everyone is on equal footing and all that is needed is some principle of equal opportunity. But most people realize that we don’t start on equal footing and that there are asymmetries of power and resources (see Unequal Opportunity Race video, thanks to my students who pointed this out). The criminal defendant, who is usually indigent, is represented by an overworked public defender with limited resources against a prosecutor representing the state. If the criminal defendant was not entitled to counsel, notice of the charges against him, a speedy trial, a jury of his peers, and, yes, an instruction that he should be presumed innocent, he would likely spend years languishes in a jail cell awaiting trial with little hope of success. The presumption of innocence levels the playing field in the criminal context, making it possible for a criminal defendant to have a fighting chance against a powerful state to defend himself against a deprivation of his life or liberty.

      Turning now to racism, against a backdrop of a society recovering from slavery and segregation and steeped in racial bias and stereotypes, a presumption of credibility merely levels the playing field for the targets of racism.

      As to sexual harassment, I would have to give it more thought. At first glance, it seems to me that something like a presumption of credibility should apply to those experiencing discriminatory treatment on the basis of any suspect classification (particularly vulnerable groups who have been historically subject to invidious discrimination). I think in places where corroborating evidence is required (in other parts of the world), that should be changed. I would have to give more thought to whether the background conditions are sufficiently similar and whether the first personal and second personal perspectives are so different as to justify privileging a first personal perspective.

  4. I’m aware that the media may influence our perception of racism or the frequency of instances of racism. But most people that I know who are people of color don’t think racism is a thing today because of what they hear in the media; they believe that racism is a thing because of what they’ve experienced personally throughout their entire lives and the lives of their family members and friends.

    I’m also aware that there’s a lingering worry here about reliability, but, why should the targets of racism have to try to prove something that other people, from their third personal perspective, are unable to experience and, hence, reluctant to believe. It seems their testimony is often discredited or not given enough credit when there is every reason to give them the benefit of the doubt.

    Maybe I'll say more tomorrow to respond to your other comments.

    1. Chong, thanks, these are interesting replies.

      I agree with much of what you say. I continue to think that the only compelling basis for shifting the burden of proof in this way would be compelling evidence that charges of racism are more likely to be correct than their denials. I'm not sure if you agree with this and just think the case can probably be made, or disagree and think that there are reasons for presuming credibility that are insensitive to questions of reliability.

      You're right that measuring prejudice is difficult, but there are some interesting empirical approaches to it. For example, in one sort of experiment subjects are told that an ugly fake scar is going to be put on their forehead and then they are asked to conduct a series of interviews. Later they are asked if they felt that they were being treated differently because of their scar. A significant percentage are very sure. But it turns out there was no scar applied at all.

      In any case, the measurement problem is part of reality. It can't be impossible to measure racism but unproblematic to charge people with it. And we philosophers get really uncomfortable when people claim to have privileged access to phenomena then repudiate efforts to verify it.

      Of course it goes without saying that none of my skepticism is about the prevalence or pernicious effects of racism itself. Part of what makes our worst problems our worst problems is the difficulty of fairly and accurately ascertaining when they have occurred in particular cases.

  5. I had made a recent observation about the current lack of distinction between rationalization and intimidation which may add to your argument. Rationalization appears to me to be a self-comforting behavior when considering uncomfortable positions. The intent behind rationalization would then be to formulate a way to avoid or confront discomforting ideas. Presuming our conscience involves a calculation of the possible judgment of others, what is the difference between rationalization and conspiratorial intimidation?

    If we imagine ourselves as retreating not into nature or the world when we 'run away' or 'avoid', then we are simply running *back into society*. Avoidance itself then seems to imply that we enter the intimidating protection of our society to avoid facing the response by another. If the concerns of that other person were legitimate, it leaves them 'high-and-dry' without any recourse that perhaps should have been granted to them. In this way, avoidance becomes its own sort of intimidation. In the more obvious conspiratorial intimidation sense, one formulates a way to confront unpleasant ideas in some kind of sophistry, manipulation to accuse the other so as to invoke the wrath of either themselves or the social structure behind them. Of course, in this way, we see how society commits its own sort of violence by the toleration of violence.

    In regards to the culpability of the one accused of racism being such that they must show sufficient evidence they aren't, the justification for such an argument comes from the fact of racism permeating throughout current society. The same way that someone claiming "You're human." would be seen as uncontroversial within the dwelling of humans, the claim "You're racist." is uncontroversial in a racist society. The burden of proof falls on the one being accused to deny the claim, as you suggest.

    1. Russell, thanks for your comment. I don’t think I entirely understand it, but here are my initial thoughts. It depends on how we define rationalization and intimidation. I would define rationalization as a self-interested explanation rather than as self-comforting behavior (this may be because I tend to be a rationalist, as opposed to a sentimentalist). Intimidation may be defined as the use of physical or psychological coercion or pressure to achieve certain ends. So I don’t see exactly how the two link up as you suggest.

      After the fact, one may rationalize words or actions that may have been perceived as racist, and this may have the effect of intimidation. But, more directly, it may perpetuate a social structure that is pervasively biased. It seems to me any intimidation is an effect (maybe an unintended effect) of perpetuating an unjust social structure. This is how I would describe the relation (at least in many cases): rationalization leads to perpetuating an unjust social structure which has the effect of intimidation.

      I haven’t spent time thinking of these terms or concepts as much as it sounds that you have, so I probably haven’t responded adequately to your comment. The concepts are interesting and may be helpful, after giving them some more thought.

      For all: I want to be clear that a claim of racism is a serious and nontrivial one. Racial bias may be pervasive, but it is not the case that everyone and everything is racist in our society. The threads of racial bias may be so interwoven into our social fabric that they may be hard to disentangle, but they are threads and not the entire cloth.