Prejudices, emotions and misattributions

In my earlier forays into the theory and science of emotion, there was one thing that struck me as extremely potent as an explanation: misattribution. Misattribution (frequent appeal to which is made by social psychologist Jonathan Haidt and colleagues) often goes like this: You have an emotional reaction, positive or negative, and you look for a reason for why you might have this reaction by scanning the environment for salient differences that might account for it. Haidt calls this ”post-hoc rationalisation”.  Post-hoc rationalisation results in misattribution when the reason you take to account for your emotional reaction does not correspond to what in fact caused it.

This is a quick, often unreflected, process and it seems to be quite widespread. But people differ enormously in what type of rationalisations and attributions they tend to make. Some will often blame their own flaws for any negative reaction to a situation, others will blame the food, their company, the climate, or just the nearest person. The process is also often very useful: we need to explain our negative and positive reactions, and we need generalised explanations if we are to make plans for how to live our lives if we are to avoid these unpleasant experiences and make the pleasant ones more frequent.

Now, our emotional reactions are caused by a vast combination of factors. Some we are aware of, or can become aware of, some are welcomed, and some we are reluctant to accept. I like avant garde jazz, but I also very much like the fact that I like it. It’s part of my self-image. This being true, any unpleasant encounter with avant garde jazz tends to be blamed on the circumstances. In fact, even if my last five, or ten encounters would have been unpleasant, I would be unlikely to attribute this to my tastes having changed.

If you are prejudiced against certain people (this based on group or individual characteristics), you are likely to attribute the valence of any negative emotional reaction you have encountering these people to them. If you are unaware of your prejudice, or unaware of that it is a prejudice (perhaps because you are reluctant to accept it), you are likely to try to find some rationalisation of your reaction that correspond to your considered view of what constitutes a proper reason for an emotional reaction.

Discrimination very rarely proceed by someone being ruled out on basis of group membership. All stops pulled apartheid is very rare. Rather, everyday discrimination proceed by people having an averse reaction to a person or situation, and then looking for something that could be treated as an acceptable reason to disfavour that person.

Let’s say I am interviewing people for a position as a research assistant, and one of the applicants is female. Let’s say I’m prejudiced against women, but I don’t think I am. So I have an averse reaction (this is my prejudice being manifested) and I start looking at the applications for a reason why I might have this reaction. And it turns out the female applicant’s typing skills are somewhat worse than the male applicants. ”Ah – typing! Typing is very important for a research assistant”. This is a proper reason, even if it’s not my reason and it’s not a good enough reason to determine who get’s the job.

Prejudices, in other words, often work by making the prejudiced person more likely to find some acceptable reason on the basis of which he/she may discriminate against the target group. This sort of discrimination is probably quite common, but exceedingly hard to prove, especially for the person who exhibit this strategy (very often not knowing it).

The phenomena on which this is built – post hoc rationalisation/explanation, is, as mentioned, a very useful cognitive feature and we wouldn’t want to get rid of it. In fact, generalizations are often very useful, and generalizations and prejudiced are quite clearly related. What we need, of course, is better generalizations, and making sure that this process properly correspond to the reasons we accept. I’m guessing (because the jury is still very much out on what works for prejudice-reduction) that what’s required is that we, contrary to inclination, approach that to which we have averse reactions, to find out more about the proper cause of that reaction, hoping to calibrating our reactions to what actually matters. (This may, for all I know, be what Gordon Allport meant by the ”contact-hypothesis”, btw).

Punishing causes

”Tough on crime, tough on the causes of crime”, remember? I remember very little of the substantive debate but I do remember that it spurned a debate on the use of non-committing slogans in political speech. The immediate appeal of slogans should not be dismissed as a mere trick, however.

There is a quite obvious and quite relevant issue hidden in the slogan: How does, or should, our approach to crime relate to our understanding of, and approach to, the causes of crime? Presumably our toughness or softness on crime should be modulated on the basis of our perception of the causes in question because clearly, all crimes do not share causal histories and clearly, this matters to how we assign responsibility.

Quite generally, causes are relevant to responsibility and to criminal punishment as punishment is meted out on basis of, and in proportion to, the harm agents cause. A highly valid defense to the allegation of murder is to say that you didn’t cause the person to die. Or, actually, accurately and more precisely: that you didn’t do it. To murder someone is not only to cause the death of another, but to do so while trying to do so.

The most obvious causal component of relevance to responsibility/culpability is the decision. We are condemned for the things we decide to do, and decisions have consequences.  Reasons and considerations are presented to us, or thought up by us, and then we make a decision to act on some of them. We are then held responsible for at least the causal consequences that we had reason to believe would follow. But decisions are not where causal chains end. And while deciding to do something that will cause harm when there is no reasonable amount of compensation is surely blameworthy in it’s most paradigmatic form, it’s not the end of blameworthiness either.

Yet, there are further moral judgments to be made which goes beyond the decision. For instance: what considerations did you act on? How was the case for and against acting presented to you? Was it greed? Was it vengeance? And now the question becomes: can we add to a criminal sentence on the basis of pre-decision causes? We seem to be able to subtract from a sentence on basis of certain causal pre-cursors, such as ignorance or a mental episode. But can we add?

Now to the hate crime context. For theoretical reasons as well as practical ones further down the line, it’s important to distinguish between the reasons for the support of hate crime legislation and the justification of that legislation. My reasons to favor higher taxes may be that I would gain from it in the long run, but that’s not sufficient as justification as tax rates are not in place to satisfy my interests. It does mean, however, that I’m more likely to look for, and find, further reasons for higher taxes.

It’s very likely that support for hate crime legislation is at least in part grounded in the intuition that some pre-decision causes are worse than others. We dislike, and we are right to dislike, prejudices, vengefulness, greed etc. But it is not clear that we should punish on the basis of the moral objectionableness of pre-decision causes. Even when we are somehow responsible for having become bad people, we can’t be punished for being bad people, only for doing what bad people tend to do, and which makes them bad: harm. If there are other justifications, we should identify them. But we should be very clear that our acceptance of those justifications is not wholly founded in our independent, warranted, but legally invalid, moral stance. Or, of course, we must make the case that these pre-decision causes are reasonable grounds for punishment enhancement. Which means much more work.

Two lessons to draw from this:

1)We may maintain that decisions are where culpability starts, but that the picture is more complex then previously recognized. Decisions may be judged as worse not only on the basis of the harm intended, caused or risked, but on the considerations that was deemed sufficient for the decision to be made. To treat a prejudiced view as a decisive reason, then, is worse, for instance, than treating greed as such a reason because it is a worse reasons. Still, we are not punished for this reasons, but for treating it as a decisive reasons. (Leaving, for know, the question whether the cause and/or reason for our accepting this consideration as a (decisive) reason should influence the extent to which we are culpable…)

2) Pre-decision causes are clearly relevant for effective preventive measures. We should concentrate a lot of effort at counteracting them and the most effective means may not be punishment enhancement, or any other means available to the criminal law.

Being ”tough” on causes of crime doesn’t necessarily, or primarily, involve punishing people for having certain beliefs attitudes or dispositions, but effectively counteracting the conditions under which such beliefs, attitudes or dispositions arise.

The christmas day hate broadcast

‘this the season and all that and whoever is on the non-news part of your state-subsidised radio or television is instantly promoted to the status of National Treasure. And about bloody time.
To be serious for just a minute, though (even though I don’t really want to): on sunday the 25th, swedish radio P1 will broadcast an episode of the soaringly popular program Filosofiska Rummet. This episode features yours truly in conversation with the magnificient police officer and educator Jeanette Larsson and professor of Criminal Law, Per-Ole Träskman. The topic is hate crime, it’s nature and moral status, and the point and justification of hate crime legislation. I may sound like a sceptic on the show, but that’s mostly a group-dynamics kind of thing.

There goes the neighbourhood

What counts as a hate crime?
Among the first questions you should ask when being introduced to a new category-concept is this: what does it cover? What qualifies as an ”X”? When given a fairly informative definition, you then quickly go on to fringe cases, and ask whether they qualify or not. This technique will make you seem polemic and lacking in seriousness. If you’re a decent philosopher, that’s at least partly true.
We need fringe cases in order to lure out the differences between theoretical models, and also to test the plausibility of these models.

Hate crimes, as we know, invoke a combination of factors, mainly crime, prejudice and the small but very significant and problematic notion of ”because of”. A crime is a hate one when it is committed because of a prejudice against a certain group (membership in which is based on some to-be-determined characteristic important to the victims’s, or somebody’s ”identity).

Now, it’s important to note that committing a crime because of a prejudice against a group, is more narrow than committing a crime because of group membership. I may believe (rightly) that the elderly are less able to catch me, running from the scene of the crime, and therefore target them. According to the so called ”victim selection model” for hate crimes, this might qualify simply because group mebership was part of the reason why I picked out this particular victim. According to the animus-model, it wouldn’t count unless I also held some negative and unfounded views about the elderly and that was part of the motivation for the crime.

The victim selection model allows for cases where the perpetrator has no prejudice, and prejudices in general do not enter the explanation why the crime took place. Semantically, this puts the label ”hate (and bias) crime” in question. But consider a further case:

A person assaults a group of immigrants that have recently moved in to the neighbourhood. The reason is that he/she (rightly) thinks that the arrival of immigrants have lowered the economic value of his/her house, and thus intend to scare them into moving away. Let’s imagine this is the sole reason, and the perpetrator can deny any prejudice by saying that he/she would’ve assaulted anyone whose presence in the neighbourhood had that effect.
What does the animus model say? Well, no animus is present, so this seem not to be a hate crime. The victim selection model, on the other hand, would presumably rule it in.

However: the peculiar feature of this case is that the crime is still committed because of prejudice. It’s just not the prejudice of the perpetrator, but that of the potential byers of property. On one interpretation of the ”because of” clause in the hate crime definition, then, these sorts of crimes would count. In many respects, these crimes seem to be at least as shady, and in some sense cowardly, as crimes committed on the basis of a prejudice that you have yourself. They presumably do as much damage. But should they count as hate crimes?

It depends, you might say (and rightly so), on the proper analysis of the

because of

. It’s just that this analysis depends on the plausibility of what’s included, like the fringe case above. And that in turn depends on the moral foundations of hate crime legislation. A question that, as reader of this blog may have noticed, is far from settled.

Hate Speech as Hate Crime

A number of States have laws criminalizing speech on the basis of content. ”Hate speech”, as it is often known, is a regulation prohibiting certain views from being expressed. This is distinct from direct incitement to criminal acts, or, for that matter, causing physical harm by expressing a view very loudly in someones ear, by the emphasis on content. (Lets leave for now the crucial question of how to individuate content in context).
Now, Hate Speech and Hate Crime are usually kept apart. The former is much more controversial and not embraced by as many states, or by as many scholars. Indeed, its not uncommon to come across strong advocates of Hate Crime legislation that are simultaneously in strong opposition to Hate Speech legislation.

The key difference, it is claimed, is that Hate Crimes require a ”base offence”. This means that in order for a Hate Crime to exist, there must be an act that would be criminal even absent the hate motive. But in the case of Hate Speech, it is said, there wouldn’t be an offence absent the motive or content.

There is a clear weakness in this argument, and it depends on the conflation of hate motive and hate content. I can express a hateful view without actually harboring the hate expressed. Linguistic content is not a relation between my internal state and the words I use, but between linguistic conventions/functions and the words I use. If Hate Speech is a crime based on content, it is a crime that can be committed with any motive. This means that there is a ”base offence”, independent of hate/bias motive, which can then be turned into a Hate Crime, if such motives are present.

This does not mean that all states with hate crime laws should start punishing hate speech acts. It only means that what acts can be a hate crime depends on what acts are criminal in the state in question. If speech based on content is such a crime, there is no theoretical hurdle to stop it from being a Hate Crime.

Future-oriented and customized punishment

Legal punishment is normally justified by appeal to Wrongdoing (the criminal act) and Culpability (”the guilty mind”). These are features focusing on the perpetrator, which makes sense as it is he (nearly always a ”he”) who will carry the burden of the punishment. We want to make sure that the punishment is deserved.

But it is also typically justified by appeal to societial well-being. To protect citizens from harm, to promote the sense of safety, to reinforce certain values, to prevent crime by threatening to punish, to rehabilitate or at least contain the dangerous. According to so-called ”Hybrid” theories, punishment is justified when these functions are served, but only when it befalls the guilty, and in proportion to their guilt (this being a function of wrongdoing and culpability). Responsibility/culpability constrain the utilitarian function. Desert-based justification is backward-looking, while the utilitarian, pro-social justification is forward-looking. (Arguably, the pro-social function is dependent on the perceived adherence to the responsibility-constraint.)

Neuroscientist and total media-presence David Eagleman had a very interesting article in The Atlantic a while ago, pointing out that revealing the neural mechanisms behind certain crimes tends to weaken our confidence in assigning culpability. Rather than removing the justification for punishment, Eagleman suggests that we move on from that question:

Continue reading Future-oriented and customized punishment…

Hate Crimes and the Unfair Distribution of Harm

Hate Crimes Hurt More is the title of Paul Iganski’s 2001 article in the American Behavioral Scientist, and while this title could have done with an added question mark, subsequent work by Dr Iganski and others offers ever increasing support for the central claim. The harm caused by hate crimes is not just the harm done to the victim, but to the group membership to, or association with, which was why he or she was targeted, and to society as a whole. Calculated in mere aggregative terms, then, there seem to be plenty of reasons to prioritize hate crimes, to prevent them from happening and to help victims (primary, secondary and tertiary) cope with the consequences. It is important that we work out why hate crimes hurt more. The reasons may be extrinsic to the crimes in question, for instance, and thus be addressed by other means than criminal sanctions. So far, so utilitarian, so within the wrong-doing – culpability paradigm. Hate Crimes are worse than other crimes if (and only if) they tend to do more harm.

Hate crimes Hurt Less?

Now, while there is evidence for the claim that HC’s hurt more, it is far from complete, and as data collection is far from standardized across nations, it’s hard to sustain the universal claim. So now I’m going to make a bold suggestion: Hate Crimes actually hurt less and that’s what’s wrong with it.

If we focus on the harm done to the victim’s group, there seem to be at least two mechanisms at work:

1) If people like me are targeted, I’m more likely to be next.

2) I feel more empathy for people like me, so when someone with whom I share an important characteristic, I feel it more.

This means that if another group/characteristic is targeted, I’m unlikely to be targeted, and I won’t, presumably, care as much. I guess most people feel like this about gang-related violence. If totally random acts of violence occur, everyone would, presumably, feel equally threatened and thus the harm would be even greater (even if the likelihood of you being next would be inversely proportional to the group within which the random violence occur).

The suggestion here, then, is that hate crimes hurt some people more than others. And that’s what’s so wrong about them. It’s discriminatory, and prioritizing hate crimes would be an act of distributive justice, independent of claims about the extent of harm caused by these crimes. This is in line with an Harel and Parchomosky (On Equality and Hate, 1999) who claim that recognizing the moral significance of hate crimes in criminal law depends on extending the wrong-doing – culpability paradigm with what they call the Fair protection paradigm.

Equality vs Utility

This brings to the surface a very basic conflict in moral and political philosophy: It’s uncontroversial that harmfulness counts, but does equality of distribution of that harm (and benefits) count as well?  If it does, there can be conflicts. Making hate crime a priority may be very costly in terms of police efforts, for instance, and require resources to be taken from investigations addressing other crimes. In principle, we can have a situation where a smaller number of crimes, and a lesser aggregate of harm, takes place but a disproportional portion befalls a minority of the population. That scenario comes in roughly three types, testing our commitment to fair and or equal distribution:

1) The number of crimes and amount of harm befalling the disadvantaged group is smaller than it would otherwise have been – I.e. the total harm is smaller, and the harm caused to the group is smaller than in the original scenario, but the proportion carried is larger. Preferable in absolute terms, not preferable in equality terms.

2) the number of crimes and amount of harm befalling the disadvantaged group is the same, but that of the majority is lowered.

3) The number of crimes and amount of harm befalling the disadvantaged group is larger than in the original scenario.

Our commitment to equal distribution is shown by which of these we find preferable to the status quo, or if none of them are.  If, in a Rawlsian spirit, you want to look at how well, in absolute terms, those are doing who are worse of, 1) is preferable to the status quo.  (The difference principle)

Hate Crimes hurt more because they hurt some more than others

The argument made, and the mechanisms posited above, works, if at all, on the group level. There is a further argument to be made, which works on the level of society as a whole: What’s wrong with (some) inequalities of distribution is that they tend to lead to suboptimal outcomes. That is, if an unfair proportion of crimes hits certain, typically already disadvantaged, groups, this is correlated with societal discord, hostility and impoverished inter-group relationships. It makes successful interactions less likely by raising suspicions.

If this is true, a purely harm/utility driven account of the badness of hate crimes can appeal to harmfulness in absolute terms. This harm may result even if the primary and secondary harm caused by a targeted hate crime is actually smaller than that of a ”completely random” crime, or any other crime for that matter.

Changing the focus to this level does, I think, get the general problem right. And it’s not just a problem with hate crimes, but with hate and prejudice in general (I believe Barbara Perry would agree). It does, however, mean, that demonstrating the relevant harm caused by individual crimes is exceedingly difficult. And this might be right, too: while the direct harm is caused by the criminal, the Hate Crime specific harm may not be, and thus the problem should be addressed not with criminal sanctions but by other means. We could still hold an individual responsible for targeting a group that carries more than it’s share of the burden, however – it would be a crime dependent on the disvalue of unfairness, rather than of harm.

the Philosophy of Hate Crime Map

Should perhaps be framed by areas marked ”Here be Dragons”.

The Philosophy of Hate Crime Symposium

Tomorrow it starts: the Philosophy of Hate Crime Symposium, the 2nd in a series of symposia in the When Law and Hate Collide project. The Symposium, as far as I know, is the first to concentrate on philosophical aspects of Hate Crime and Hate Crime Legislation. (There has been a Law and Philosophy special issue, however, on Hate Crime Legislation back in  2001).

It is also quite unique insofar as, amazingly, I’m hosting it.

It is a symposium about hate. As a counter-weight to that other, more famous,  2400 year old symposium.

Here’s the schedule. As you can see, it’s all very interesting stuff

Monday 26/9

Introduction: How Law and Hate Collide

Mark Cutter, University of Central Lancashire and Christian Munthe, University of Gothenburg

Moving Beyond “Hate” Crime

Barbara Perry, Department of Social Science and Humanities, University of Ontario Institute of Technology

How hate hurts. The moral philosophical basis of ‘hate crime’ laws

Paul Iganski, Department of Applied Social Science, Lancaster University

Targeting Vulnerability: A Fresh Set of Challenges for Hate Crime Scholarship and Policy?

Neil Chakraborti, Department of Criminology, University of Leicester

The OSCE and its Work on Hate Crime

Joanna Perry OSCE

Panel Discussion

Tuesday 27/9

Criminalizing Hate, Criminalizing Character

Heidi Hurd, University of Illinois, College of Law

Hate as an Aggravating Factor in Sentencing

Mohamad Al Hakim, Department of Philosophy, York University

Two Kinds of Expressive Harm

Antti Kauppinen, Department of Philosophy, Trinity College Dublin

Philosophy of Hate Crime – a Conceptual Framework – Morality, Law and Public Policy

David Brax and Christian Munthe, Department of Philosophy, Linguistics and Theory of Science, University of Gothenburg

General Discussion

The symposium will be filmed and made available to the public as soon as possible. Watch the upcoming webpage: www.h8crime.eu

Hate Crime and Interchangeability

Hate and Interchangeability

A key concept in distinguishing Hate Crimes from other crimes is Interchangeability. This seems to be a main feature of group hatred: treating members of the target group as interchangeable: any person belonging to that group could be selected. In their 2002 book Hate Crimes Revisited Levin and McDevitt argue that this means that there is an amount of randomness in these crimes. Any bearer of the targeted characteristic is a potential victim, and this is at least partly why the psychological impact of hate crimes on the targeted group is so severe. Systematically ”random” acts of violence influence the feeling of security of populations more widely then does, say, violence between criminal gangs.

The hate in hate crime is generalized hate, it is not intended to encompass hatred for individual persons based on individual characteristics. But it is not entirely general: the crime cannot be totally random, the victim not, totally interchangeable. Then it will not be a hate crime. Two types of hate motivations then seem to fall out of the hate crime picture

1) The perpetrator does not hate black people, just this person

2) The perpetrator does not hate black people, but all people

If the psychological effects of random acts of violence are such as Levin and McDevitt suppose, then, presumably, the latter alternative should be even worse. Or possibly we are not that bothered by complete randomness (we are very unlikely victims), but by randomness within a smaller group to which we belong, either because we are more likely victims (our feeling of insecurity would then be inversely proportional to the size of the group) or because we sympathize more with victims of our own group.

The Central Park Jogger

In the highly publicized ”The Central Park Jogger Case” back in 1989 these sorts of considerations came into view. The victim, a 28 year old white female investment banker, was jogging in central park when she was assaulted, raped and beaten nearly to death by a gang of youths. Evidence suggesting that she was targeted because she was female, or upper middle-class, poured in, suggesting race-, gender- or class-based hatred. But so did evidence that the gang was looking for a random victim, based for instance on earlier assaults by the gang. Interchangeability, then, plays a double role here:

1) Would they have assaulted anyone coming along at that time?

2)Would they have assaulted any woman/ white person /investment banker?

There is an important discussion going on whether rapes should more commonly be understood as gender-based hate crimes, circling around this question: is the hate involved general enough? Is the reasons for targeting women exclusively reducible to gender-hatred?

Hate and acquaintance

Hate crimes are normally understood as being committed by a member of one group against a member of another, distinct group. Recorded hate crimes tend to be cases where the victim and the perpetrator does not know each other. This is in contrast to other types of crimes (Levin and McDevitt, again). But while this is so, it might be due to how we conceptualize and understand hate crimes. If there is a personal relation between victim and perpetrator, we are less likely to understand the hatred as generalized. Spousal abuse, while terrifyingly common, seems to be understood as based in the particulars of the relation, and thus does not qualify as hate crimes. The interchangeability condition is not satisfied in the ”right” way. Alternatively, it is too difficult to distinguish the general from the particular in crimes within personal relations.

This might mean that hate crime legislation has a blind spot, however. In a recent report from RFSL (The Swedish Federation for Lesbian, Gay, Bisexual and Transgender Rights) about ”HBT and Honour”, it is pointed out that many HBT-people are assaulted by family members because of their group membership. Even if this too involves violence between distinct groups, being distinct in one characteristic is compatible with being overlapping in others, such as family membership.

Group hatred may drive violence even within personal relations. Many instances of rape, even in relationships or between acquaintances, may be caused by gender-based hatred. It is hard, however, to show the interchangeability aspect in acquaintance cases. This may also mean, rightly or wrongly, that the self-interested psychological effects of perceived randomness does not occur when we are informed about these crimes.

Expanding the Label

Should the hate crime label be attached to a wider set of cases? Would the victims of assaults within a family, say, or victims of rape, benefit from being recognized as the victims of hate crime? While perceived ”randomness” may mean that more people will feel threatened by the occurrence of these crimes, it might also provide some solace to the victim: the assault was, in fact, nothing personal. It could have happened to anyone.

But these are complicated psychological matters: It depends on how much I identify with the relevant group membership – being targeted because of being female, say, might not feel ”random” at all, but strike at the very heart of my identity. Even if I do feel it’s random, this very fact may make the event ”senseless”, and I may feel even more unfortunate because it happened to me for no particular reason. Some victims may benefit from finding some ”meaning” in the assault, but others may suffer from it. Especially in cases of rape, ”blaming the victim”, even by the victim her/him-self may seem to make sense of the particular assault, but result in more suffering. And making sense of the particular assault by focusing on the characteristics of the perpetrator, especially if the victim stands in a close personal relationship to him/her, may tie them closer together, at great personal cost.

If the Hate Crime label is effective in offering support and protection to especially vulnerable, disadvantaged or frequently targeted groups, there seems to be no objection to expanding it to cover cases which may at the surface look as instances of personal hostility. Whether it is effective in this regard is another question.