Saturday, February 1, 2014

Crime and Agency

One of my recent-ish posts on this blog addressed itself to retributive theories of justice.  I argued that "retribution" has no legitimate role to play in our theories of punishment, because it rests on the notion that the infliction of suffering on a wrong-doer can somehow serve as compensation to a victim.  In short, it implies that the suffering of the criminal is good as suffering-- a notion that very few of us would accept, when shown in this light.   In conversation and elsewhere, my co-blogger Ajay raised some interesting objections to the post, however-- not, I believe, on the basis of the argument so far, but on the step I made next.  I went on to argue-- or at least, to imply with gestures difficult to mistake-- that there can be no case in which punishment is good in itself.  Punishment is violence, after all, though we are apt to forget it in a society that has abolished the lash and the knout and the rack (though it holds on to the syringe and the solitary cubicle).  The act of depriving someone of her or his freedom is a violent one, and imprisonment is coercive by definition.  My suggestion in the post, not fully spelled out, was that such violence can never be justified on purely deontological (Kantian) grounds-- meaning that punishing someone is never a way of treating her as an "end" rather than a means.

If I wish to hold to his suggestion, it would entail rejecting retributivism, but also -- crucially-- its old liberal nemesis-- rehabilitation, another deontological basis for punishment.  I didn't address myself to the latter in the post, mostly because I didn't fully anticipate this line of argument in advance.  To state my position clearly now, I regard rehabilitation as certainly one of the goals that should be pursued once the criminal is imprisoned-- but not one that can justify locking someone up on its own.  In short, rehabilitation is necessary, but insufficient.  I will try to make clear below some of the reasons I am wary of it.

If we agree in rejecting such deontological theories, and if, further, we are to believe that punishment is still necessary in extreme circumstances (which I do), we will have to argue for the point on consequentialist grounds.  This means we must be willing to use people as means rather than ends, when that end is of truly overriding significance.  In the earlier post I laid out one consequentialist justification for punishment (there could be others-- deterrence e.g.) which I think is basically valid-- that of protection-- of the need to protect those who might be victims of violent crime in the future from those who commit them in the present.  You will notice that this justification does not regard punishment as good in itself or good for the criminal, but as good (or, perhaps, evil but necessary) only in regard to the result it will bring about for other members of society.

Ajay's objections to this, as I understand them, are two-fold: first, that we should be skeptical from the outset of arguments which justify using people as means rather than ends.  Such use of the human person is in itself a violation-- and perhaps, an anti-human act in the deepest sense.  I basically agree, which is why I am not a consequentialist in toto, but only in regard to extreme conditions. (I think I have made clear elsewhere on this blog, for instance, that with regard to nearly every form of military intervention we are likely to make abroad, I am solidly deontological.  I tend to think it is better not to commit evils yourself than to commit a great many of them in the name of preventing someone else' evils-- though this stance could also perhaps be justified on enlightened consequentialist grounds.)  Furthermore, I am a consequentialist about extreme circumstances precisely because I regard force as a violation and an evil in itself-- one which can only be justified when the costs of not employing it become intolerably high.

Ajay's second point-- again, to the extent I can reconstruct it without having the primary source in front of me-- was that there are some situations in which punishment can be regarded as good in itself-- or as a way of treating the criminal as an end rather than a means.  He gave as an example the role punishment can play in respecting the moral agency of a criminal.  Lenient regimes, the argument goes, tend to treat criminals by contrast as passive victims of social forces rather than as men and women with individual consciences.  The simplistic liberal (according to this version) thinks something like the following: "They (the criminals) couldn't help it, so if they are going to be punished, it can't be because they deserve it, but only because the rest of society needs to be protected from these actions they cannot help but perform.  In fact, if the whole criminal justice system could become a sort of vast psychiatric hospital rather than a place of punishment for wrongdoing, so much the better."

The trouble with this simplistic liberal argument, says Ajay, is that it doesn't treat criminals as people who choose to do right or wrong, but as objects buffeted about by material forces-- in short, as something other than fully human.  Political liberals, for instance, used to argue in the wake of the widespread looting that invariably followed a New York City blackout that such property crimes were the predictable result of the conditions in which ghettoized blacks in the city were forced to live.  The Neoconservative writer Midge Decter argued in response, in the pages of her husband's organ Commentary, that in so doing, liberals were actually displaying a deeper form of racism than the "law and order" conservatives, because they assumed that poor blacks were the pawns of economic circumstances--rather than willing and acting agents in their own right.  I don't find this persuasive, for reasons I will lay out below, but I find it a suggestion worth responding to.

If I have misrepresented Ajay's argument-- and I very well may have-- I invite him to correct me in the comment thread.  I should also say that since I don't have his argument in writing, I have chosen to trace its genealogy back through some more primordial and atavistic life-forms, such as Midge Decter, in order to have something concrete and quotable to respond to in the post.  This is a great deal like trying to draw conclusions about a gorilla from a rhesus monkey-- so understand that in criticizing writers like Decter, I am not necessarily criticizing Ajay-- I am simply going after an argument that I take to bear a superficial similarity to the objections he raised.

One of the first points to make is that Decter's accusation against liberal theories-- that they do not sufficiently regard criminals as "moral agents"-- is a very strange one, coming from a conservative, and seen in light of the history of ideas.  It's an accusation that could only be explained by a certain bizarre inbuilt tendency of the careers of all modern political ideologies-- that of gradually evolving, "through stages that each seem logical and necessary, into their opposites" (to quote Bertrand Russell).

We should remember, after all, that the original point of contention between Benthamite liberals and conservative theorists like James Fitzjames Stephen in the 19th century was the same as it is in Decter's argument-- it concerned the extent to which punishment should be justified by the good it did the criminal, rather than society as a whole-- but it was the Benthamite liberals who thought that it should be justified in this way, whereas conservatives resisted the notion.  This was the classic dispute between "retributivism" on the conservative side, and "rehabilitation" among liberals.  The language in which our society talks about criminal justice today suggests that the Benthamite liberals won out-- in theory, at least, if not in practice.  We are still in a land of "corrections" facilities (correcting whom?), of "reformatories" (where people are "reformed") and "penitentiaries" (i.e.-- places where one is "penitent," and gradually morally improved through the censure of one's own conscience).  Our official parlance still upholds the goal of rehabilitating the criminal to participate fully in society, however few resources we actually devote to this goal.

Of course, one could uphold rehabilitation on purely consequentialist grounds, and Bentham himself no doubt did so, for consistency's sake.  But it is clear from the popular rhetoric of liberal prison reformers of the 19th century that many of them were broadly deontological.  And certainly, whether utilitarian or deontological or both, they all believed that criminals were moral agents.  The popular novelist and reformer Charles Reade published a novel in 1856 about the prison system under the title It Is Never Too Late to Mend, which captures two important elements of this 19th century liberal sensibility in one go.  First: the evident earnestness and humanity of the reformers-- their kindly conviction that all people have moral value and are capable of doing the right thing.  And secondly-- frustratingly difficult to disentangle from the first-- their sinister Victorian priggishness and ruthless busy-bodying-- which point the way toward "moral cruelty" of a sort the Victorian mentality refined into a high art.

One hopes you can have the first of these two without the second-- but even if you could, theoretically, the Victorian reform movement didn't-- and that's the point worth making here.  The reformers were no doubt sincere in their belief that imprisonment could operate as a form of benevolent moral instruction.  But I'm not at all sure this was a good thing.  It seems to be a feature of human psychology that we allow ourselves a greater exercise of our sadism when we imagine that the violence we are inflicting is done for the victim's own good.  This conviction of our own rectitude in administering pain releases us from many of the constraints of bad conscience we would normally feel.  There is a good chance the Medieval Church would never have been so relentless in its quest to burn heretics at the stake, for instance, if it hadn't earnestly convinced itself that it was saving their souls by doing so.

Similarly, the Benthamites of the 18th and 19th centuries-- perhaps because they thought they were doing right by criminals-- invented some of the most elaborate, invasive, and totalitarian systems of control and sanction against prison inmates ever to be devised.  The symbol for this in modern debates has become the nightmarish image of the "Panopticon"-- though this was only one manifestation of Bentham's genius for moral cruelty. (It was largely in reference to Bentham, in fact, that Judith Shklar first coined the term "moral cruelty," which I have found so useful here and elsewhere.)  Gertrude Himmelfarb told us of another manifestation, about which she is rightly withering.  Keith Thomas, in a review of Himmelfarb, gives us a sense of its features:

Now [Himmelfarb] returns to the fray with a further essay on Bentham, this time devoted to his scheme of Pauper Management, whereby a million indigent people would be more or less compulsorily incarcerated in houses of industry run for profit as a private monopoly by a National Charity Company (in which Bentham once again would have been personally involved). The pauper children who inhabited these institutions were to be taught morality in the form of two propositions:
That the condition they are doomed to is as good a one, i.e., as favourable to happiness as any other.
That if it were not, no efforts which they could use by the display of collective force would have any tendency to improve it."
So the Victorian emphasis on rehabilitation and reformation could be either the highest manifestation of the humanitarian impulse of its age-- or the last spasm of an advanced sadism-- the sort which convinces itself that its violence is done out of pure love, like the Christian soldiers Augustine once imagined who would decapitate their foes with a prayer for their souls in their hearts.

Awaking to the danger posed by the latter possibility, liberals increasingly moved away from the emphasis on moral agency and thus, away from the idea of "reformation" as a goal of punishment-- neglecting in the process, it must be said, the opposite perils that they had once so percipiently identified in their enemies.  Stephen Jay Gould, one of the most persuasive critics of biological determinism in modern America, shows in detail in The Mismeasure of Man how deterministic ideas originated on the Left in the 19th- and early 20th century, among socialists and liberals who felt that if they could show that criminals couldn't help but commit crimes, due to their genetic programming, then society would be inclined to take pity on them, and this would lead to the creation of less punitive regimes.

We know of course -- Gould makes it clear if we hadn't already surmised it -- that this is not what actually happened, and that the ghost of this idea of "genetic determinism" transmigrated once again-- this time to the political right, in the form of scientific racism (remember that certain inbuilt tendency of ideologies I mentioned above?).  We have discussed before on this blog the attempts by Noah Millman at the American Conservative to make an argument similar to what the 19th-century liberal determinists alleged.  He has claimed that if differences in social outcomes between racial groups could be accounted for by genetic factors, it would lead to a less cruel outlook on behalf of the privileged toward the genetically handicapped.  I think we know enough now to say that this is implausible, to put it charitably.  In addition to such determinism being wrong on its scientific merits, the uses to which eugenics and scientific racist theories have been put to the past indicate that they are not great incubators of humane and compassionate values.

Meanwhile, even as this sort of determinism resurfaced on the right, a new Victorian emphasis on "moral reformation" likewise appeared among Neo-conservatives.  We have seen it in Midge Decter's article, which makes a case for the more stringent punishment of looters based on the theory that doing so treats them as moral agents, and therefore, full humans, in a way that the liberal determinist claim ("they are poor and suffering, so how could they have done otherwise?") does not.

There are numerous ironies in this, some of them peculiar to Decter's argument.  For instance, despite her castigation of "liberal racism," she evinces unmistakable contempt for the inhabitants of the inner-city ghettoes she describes.  It is impossible, that is to say, to imagine a black reader from this background coming away from her article with the impression that she has a "deeper respect" for her subject than the liberals she is criticizing.  Unfortunately, her article has retreated behind a paywall on the Commentary website since I first read it, so I am unable to provide direct quotation to cinch the point, but suffice it to say that Decter betrays most of the prejudices one might expect of her era, political persuasion, and social location.  She writes, for instance, of the poor as lay-abouts, whose biggest problem is a culturally-reinforced laziness.  She insists furthermore that the government already does far more than it should to help them escape poverty.  Responding to the liberal claim that New York City is partly responsible for the looting outbreak, because of its failure to provide for its poorer citizens, Decter points to various social programs already available to the poor (even though, as a Neoconservative, she regards these all as titanic failures).  This is the "Haven't we already done so much for these people?-- and look at the thanks we get!" school of thought.

The other ironies of Decter's piece are more general: such as the adoption of the Victorian moral reform agenda as a way of defending what once would have been its opposites: more punitive, traditional, and retributivist regimes.

So it transpires that these supposedly "fundamental" philosophical differences between left and right--  the debate between moral freedom and determinism, between deontological ethics and consequentialism, between means and ends-- are not really so "fundamental" after all, but rather are picked up by liberals and conservatives as they come to suit their agendas.  But what are these agendas if they are not concerned with precisely these foundational issues?  Whence the continuity of these ideas of "the left" and "the right"we all routinely employ if it does not reside in the underlying philosophy of each ideology? 

I come away from my historical overview with the conviction that the backbone of the liberal theory of punishment-- the agenda which provides it with its continuity across generations-- is the quest for a less punitive regime: less physical pain inflicted on the condemned, fewer final and irreversible forms of punishment (such as capital punishment), shorter terms in prison.  This was the broad agenda of the Victorian reformers (even of Bentham, in some sense), and for its sake they adopted an argument for the individual moral agency of the prisoner-- his or her capacity to become a better person.  But once the "moral reformation" idea took on a life of its own and became the justification for an ever more punitive system of "corrections," liberals began to turn on it.  They invoked a type of determinism, in the hope it would produce a more compassionate attitude toward prisoners.  Similarly, the principle that seems to anchor conservative theories of justice is not any abstract philosophical conviction about moral agency, but a basic drive toward more punitive punishment regimes.


The reason why philosophical foundations are used so capriciously by different ideologies in not that they are meaningless in the abstract.  Ideas-- even abstract ideas-- have consequences, viz. the careers of biological determinism and scientific racism in the 20th century.  Rather, they are used capriciously because in any of the binary divides we have been discussing above, both sides contain important glimpses of the truth-- and both sides become false in all manner of deadly ways when they are held to contain the exclusive truth.  Because both are true to an extent, then ideologies can choose one or the other to emphasize in a given historical moment.

Allow me to illustrate the "truth on both sides" claim, and to show that I don't mean this in a pat, false-equivalency sort of way.  Suppose, for instance, we ask whether the wrong in a crime lies in its effect on the moral character of the person who commits it, or in its effect on its victims.  If we answer "the effect on the victims" alone, then we are implicitly arguing that the inner moral state of the criminal is irrelevant, and all that matters is that we prevent him or her from inflicting violence on others.  

Science fiction has already provided us with the relevant thought experiments to show the problem with this claim.  Anthony Burgess' classic A Clockwork Orange portrays a futuristic society in which young hoods, such as the protagonist Alex, are subjected to the "Ludovico method," which has been devised by liberal determinists.  The method produces a forced nausea and pain reaction in the would-be criminal every time he contemplates the prospect of inflicting violence.  The book asks us, in effect, whether we truly believe the problem of crime has been "solved" by this method.  If the only objection one makes to crime is that is has negative effects on its victims, then of course it has been solved-- all hail Ludovico.  But I doubt any of us is likely to go this far, when Alex, the book's anti-hero, still wishes to commit violence, and has simply been rendered mechanistically impotent for the act itself.  There is something wrong with his moral stance, not just with his physical action-- and the fact that the latter was altered for him by force can only seem to the reader a profound violation of his humanity.

And if we retreat to the other pole and say that the only thing wrong with a crime is its effect on the inner moral state of the criminal, this commits us to a bizarre sort of Augustinianism in which any act can be justified, so long as it is undertaken with love in one's heart.  I find this hard to believe, just as I am unable to accept that the only problem with a murder is that it made the murderer a worse person, and not that it extinguished a human life.  To make this claim would be effectively to deny that the protection of life has any intrinsic value (which Augustine was willing to do-- but I suspect we are not).  

Thus, what is wrong with crime would appear to be both its effects on the criminal and its effects on the victim.  An obvious insight-- but, like all obvious insights, one obscured by ideology.

Similarly, and more to the point of this post, we can say that treating someone as a moral agent is in fact essential to treating her as a full human being, just as Decter asserts.  However, if we are to treat agency not just as a descriptive proposition but also as a normative ideal, we have to recognize that, should there be obstacles that stand in the way of the exercise of this human agency, then surely part of respecting people as agents means helping them to clear these obstacles-- such as poverty, the legacy of racial injustice, and poisons in the air we breathe that can diminish our capacity for rational choice.  

It is profoundly duplicitous to invoke "agency" to defend moral inaction with regard to such obstacles.  You cannot have your foot on someone's neck and then tell him you aren't going to move it because you respect his strength so much that you trust him to remove it on his own.  That is plainly a laughable misappropriation of the ideals of respect and of autonomy-- yet it is not so different from the conservative who maintains he respects the poor more than liberals do, because he views them as wholly responsible for their own fate rather than being in part the victims of social forces.

Given that in all these binaries (society vs. individual, agency vs. determinism, etc.) there is truth on both sides, I see nothing wrong with re-emphasizing one side when the other has gotten out of control in a given society.  But how do we know when a particular side needs to be reined in?  

I propose the following rule of thumb: whichever side of a binary is being used in a given context to justify a more punitive regime is the one that is has outgrown its appropriate sphere.  The task of people who write about crime and punishment, in consequence, should broadly be to try to make a society more lenient than it already is, and to emphasize the elements of the various binaries that will contribute to that goal of leniency in a particular context-- without-- hopefully-- emphasizing them in a way that seems to rule out their opposites entirely.  

I say this not because punishment is never valid or necessary in my view, but because most other forces in society will always be pushing away from leniency, and will require some corrective.  Criminals will always be regarded by the rest of the population as internal enemies-- as domestic foreigners.  We know enough about human psychology to say, in fact, that simply accusing a person of a crime is enough to render him or her in most of our minds a source of fear.  Witness the phenomenon of "Trial by Headline," as Jessica Mitford once dubbed it, whereby the accused are condemned in the court of public opinion before they ever stand trial-- and thus their jurors will almost invariably have formed a conviction of their own guilt before they enter the courtroom.  

Even if we are guilty of precisely the same crimes as the criminal, we still regard him as a moral alien as soon as he is caught.  Rush Limbaugh's remarks about drug addicts, for instance, in the days before it was discovered that he too was addicted to prescription medications, are simply one of the more spectacular examples of this widespread and familiar hypocrisy.  Societies will always want to do greater degrees of harm to their criminals than is morally permissible-- than is minimally necessary, in short, to prevent rapes and murders and armed robberies from taking place.  

I hope that the theoretical perspective I outlined in my previous post on accountability is in the interests of leniency-- but if it could be shown not to be, I would probably abandon it more readily than I would abandon my lenient tendencies.  We have seen already that the "humane" theory of one generation becomes the legitimizing ideology of the cruelest regimes of another.  I tend to think that rehabilitation is broadly a more humane ideal than retribution, for instance-- but we know that there could be Benthamite moral reformatories infinitely worse than the retributivist dungeons of yesteryear.  I have my theoretical perspective-- but I have it for the most part because I think it makes people less cruel and more clement than they would be otherwise.  I would probably trade it in for another if it were shown to me that this was not the case.


I have one final point to make.  As a result of the reasoning above, I am happy to make common cause with other theoretical perspectives, so long as they favor an agenda of greater leniency.  It seems to me that retributivists, Victorian-style reformers, liberal consequentialists and determinists alike can find something to criticize in America's current prison system and work together to ameliorate it.  For instance, I don't think there is any purely theoretical persuasion that necessitates mandatory sentencing for nonviolent crimes.  Nor is there one that necessarily views the debilitating effects of solitary confinement as proportionate to the crimes it punishes.  Or that thinks it is acceptable that 3% of all prisoners report being sexually abused in prison (including one in ten minors in juvenile lock-ups). Or that thinks the death penalty should be sought for a 20-year-old boy, whatever crime he committed.

We can agree from a variety of philosophical perspectives that these things are intolerable in a civilized society.  Given that this is the case, the partisans of mercy should not waste too much of their breath on their internal theoretical debates: they should save some of it, at least, for the partisans of punishment.


  1. It occurs to me in looking back over this post that, amidst all the pondering and quoting and musing, I don't make it very clear WHY, exactly, I favor consequentialist theories of punishment over deontological ones. I will try to do so now, but let me first re-emphasize that I regard these theoretical disputes as ultimately of secondary importance (though of primary INTEREST, perhaps, to a relentlessly unempirical thinker like me). Any abstract theory which regards such punishment as remotely legitimate can be manipulated to justify tyrannies and cruelties of various sorts, and consequentialist theories are certainly no exception to this. Of greater ultimate meaning is what one DOES with one's theories-- which is why I tried to offer the olive branch in the last few paragraphs to those who may have different theories, but a similar concrete agenda. This doesn't mean I don't have an opinion about theory, however, though it may have gotten somewhat lost in the foliage above.

    To take a machete to this tangle, I would say that there are only two significant and-- for me-- decisive reasons for favoring consequentialist punishment theories (if not consequentialism in all cases). 1) They do not commit us to the view that doing violence to someone, at however minimal a level, is ever a way of showing that person greater respect. This means we can avoid a doctrine (punishment as respect) that I regard as especially prone to abuse. Of course, there is a meaningful sense in which grabbing a child by the arm who is about to dart across the street is actually for her or his own good, and shows the child greater respect than letting her or him get hurt would do, but among adults possessed of all their mental faculties, the notion loses credibility. To take an example from elsewhere on this blog, I spent a fair amount of time complaining in an earlier post about orthodox Christian defenses of the hellfire doctrine, who claimed that hell was necessary in order to maintain a belief in human moral agency. God sends us to hell, the argument goes, because He respects our moral freedom. I won't go into another refutation of the point: I just want to suggest that it should, and probably does, outrage us to hear that damning someone eternally-- torturing someone eternally-- can ever be a means of showing him or her "greater respect." Obviously, this is an extreme and unusually devious invocation of the deontological justification for punishment, but I think the way in which it outrages us can tell us something important about the punishment as respect notion as a whole.

  2. And 2) consequentialist theories hold out the possibility of the eventual abolition of punishment in a way that deontological theories do not. Or rather, for punishment to cease in the deontological framework, we would have to obtain a world in which no one EVER commits a crime, which is hard to imagine. The consequentialist framework, meanwhile, makes it theoretically possible -- even without spelling out a clear path forward-- to eventually reach a society that finds ways of protecting people from violence without repaying violence in kind. I want to live in such a world. This is not because I think that crime is a fiction or that the rapist or murderer "couldn't help it." I have a post coming down the pipeline that deals with the mafia (at least as it is portrayed in movies), and I think it will be clear from what I write there that I don't view these guys as robin hoods-- nor do I think in any exclusive way that "society is to blame" for their actions. Criminals, like all people, do have moral agency, though it is often circumscribed by tremendous and unjust social limitations.

    Rather, I regard the punishment-free world as a desirable one because I have some residue of the Christian ethic in me, and I still think that the very highest moral stance-- although not one I ever seem to be able to achieve in my own life-- is to turn the other cheek, and not to repay evil with evil. I think we should view punishment the way we view war or any other form of violence. We should regard it as something we may have to resort to, with the world as it is and people as they are. But it is something we should still regard as wrong, and-- from the highest possible ethical stance-- something that should not exist.