“…State Police detectives will investigate whether the man’s behavior should be treated as a medical or a criminal matter.” – Graying duo keep passenger in check, Boston Globe, 6/5/07.
“Neither side disputes that Gomez, a diagnosed schizophrenic, is mentally ill.” - Clash on shooter’s free will, York Daily Record, 6/16/07.
Those with mental illness make up an increasingly large percentage of offenders incarcerated by our criminal justice system. Some are arguably insane, with depression or psychoses that put them at risk of suicide and self-mutilation. Others are basically in touch with reality, but with personality and behavioral disorders that were significant contributory causes of the crimes that landed them in jail. Whatever the severity of their condition, such individuals can be understood as having, presumptively, two sorts of problems, medical and moral. The medical problem is their mental illness, the moral their proven criminal misconduct, punishable by legal sanctions including imprisonment.
An obvious concern is that mental illness goes largely untreated, or poorly treated, by a criminal justice system that’s mostly dedicated to meting out punishment. Indeed, underlying behavioral and mental disorders are often exacerbated by institutional regimes that deliberately make life tough for inmates. Inflicting deserved suffering on offenders, after all, is the point of punishment at a time when retribution, not rehabilitation, is the primary principle of criminal justice in the United States. Such a mission is flatly incompatible with alleviating mental illness. As recent experience in Massachusetts (see note 1) and other states shows, prisons are inducing morbidity and mortality related to mental illness. So to the extent that mental illness is a cause of crime and intrinsically worth eradicating, prisons are making things worse, not better.
We tolerate this blatant dysfunctionality largely because retribution has pride of place in our approach to crime. We look for any kernel of rationality as the justification for assigning blame, even if the causes of an offense lie mostly in diagnosable behavioral disorders.
Why, one wonders, should retribution be our top priority? Why should a punitive moral regime, that of our criminal justice system, trump the curative medical approach when dealing with mentally ill offenders? Do they deserve to be punished? Well, some would say, they do to the extent they’re rational. But why should punishing the rational element of an offender take such precedence over curing, or at least ameliorating, his disorder, especially when it contributes significantly to his criminal propensities?
Such questions form part of a larger debate about medicine and morality. As we make progress in understanding the roots of human behavior and in gaining control over the neural and biological systems that produce it, the concern arises that we might medicalize the moral domain out of existence. We might go too far in treating people as fixable, improvable objects, as opposed to morally responsible agents. Something crucially important could be at risk – an essential dignity, autonomy or responsibility, maybe personhood itself. Analyzing and tinkering with the very mechanisms of the self and its choices might spell the end of agency as it’s commonly understood. In which case, perhaps, we should keep the scientific understanding of the person at bay.
Conflicts and Commonalities
To see if such fears are well-grounded, and to address policy questions such as how to treat mentally ill offenders, we must first define medicine and morality, then see how they might conflict when thinking about ourselves. Medicine, based in science, generally takes what philosopher Daniel Dennett calls the physical and design stances with respect to a person. If there’s a part or sub-system of the human organism that’s malfunctioning, doctors intervene and fix it, if possible. If treatment works, the sick person is restored to her normal capacities. Morality, on the other hand, takes what Dennett calls the intentional stance: persons are purposeful, goal-directed agents whose behavior in society, affecting others, inevitably gets evaluated on a normative, good-bad, dimension. We, not any of our parts, are subject to categorically moral responses such as praise and blame that are triggered by how we treat others. Morality assumes that we’re in basically good mechanical operating condition, such that we have standard capacities to respond to evaluation, or, as is often the case, to the prospect of being evaluated and being held responsible. This prospect helps to keep us in line, ethically speaking. If we’re basically sane, rational and know right from wrong, then we count as moral agents, that is, people who can be controlled (more or less) by the enforcement of social norms. If we’re not sane, for instance if we’re suffering serious delusions as a result of psychosis, then we don’t count as moral agents and shouldn’t be held responsible.
Both the medical and moral domains therefore have as a central focus the production and control of behavior, but of course they diverge in important respects. Medicine wants to reduce the suffering caused by disease and bring behavior-producing and controlling mechanisms back to normal so that persons have a full range of capacities. Morality, on the other hand, wants to keep behavior within normative limits, to “guide goodness” as law scholar Stephen Morse puts it, by treating persons as responsible agents subject to praise, blame and other moral responses. These can range from the casual expression of likes and dislikes (hey, knock it off!) to the full blown sanctions of criminal law (…for the term of your natural life).
Obviously, moral blame and punishment can impose considerable suffering on the part of the recipient, and indeed on one common conception of morality such suffering is deserved whether or not it brings about any good consequences, such as better behavior. This categorically non-consequentialist and retributive aspect of morality, which now largely drives our criminal justice system, sets it clearly apart from medicine, which never seeks to inflict suffering unless it contributes to a clearly defined positive outcome.
This suggests that conflicts between medicine and morality might be characterized as disputes about the source of bad behavior, and therefore the type of intervention or response that’s warranted. Do wrongful acts arise from a sub-personal, “mechanical” defect that’s perhaps medically treatable, or do they instead arise from a defect of the person as a whole, a moral disorder of, say, character? Should the person be fixed, or should she be punished, or both? The answer might depend, at least partially, on our conception of persons.
The Person as Construction
On a naturalistic, science-based understanding, persons – their characters, motives, and actions – are the complex result of their interacting sub-systems, especially discrete brain systems as they are now being mapped by neuroscience. The brain exists to control behavior for the benefit of the organism, and it is the unique, stable patterns of behavior, as well as unique physical characteristics (voice, visage and physique) that make up a unique individual. The brain’s systems can be spatially localized assemblies of neurons (amygdala, corpus callosum, hippocampus) or functionally specific global systems (neurotransmitter and endocrine production systems) that affect many parts of the brain and body.
Behavioral patterns that help define a person are a direct reflection of brain systems as they’ve developed from the interaction of genetic endowment with environments (uterine and otherwise) since conception. Neurally embodied predispositions – character traits, for instance – are amplified and expressed by the neural links to mechanisms controlling speech, facial expressions and movement. So the person exists as the coherent ensemble of what the brain and body do, inseparable from their functioning. There isn’t in addition an immaterial soul or categorically mental “I” in charge of the ensemble; at least science shows no evidence for such a thing. There’s of course the phenomenal, experienced sense of being a self, but that too is a construction of the brain as research on consciousness is beginning to reveal. Persons, as well as the feeling of being a person, are physically instantiated through and through, so in a significant sense we are our brains (and our bodies) as expressed in conscious experience and overt action.
This means the capacities and functioning of particular brain systems can determine particular characteristics of a person. More specifically, what we might consider a character defect might be a matter of being functionally defective at the sub-personal level. For instance, research indicates that the capacity for self-control – resisting impulses – is partially a function of the right prefrontal cortex (RPFC). If you have a defective RPFC, this might, for instance, cause you to have difficulty keeping your temper: you’re prone to angry outbursts in response to minor slights and inconveniences, outbursts you might immediately regret. (The story of Phineas Gage is the classic, extreme example of how brain damage can affect basic personality traits.) If people don’t know about this defect, you get a lot of flack for your impulsivity – in fact, you get morally blamed for it, perhaps even by yourself. Even though it’s a sub-personal functional deficit causing the problem, it’s you the person that’s on the receiving end of some rather unpleasant, person-level moral responses. After all, you’re sane and you know right from wrong, right? The problem, however, is that because of your RPFC deficit, you’re not particularly responsive to this sort of (aversive) control. Getting blamed, scolded or otherwise punished for your outbursts doesn’t much change your behavior or fix your self-control problem.
Medicalizing the Moral Domain
Now, let’s suppose doctors trained in neuroscience determine that your irascibility is indeed a direct result of your defective RPFC and that there’s a way to restore it to normal, the way we fix a broken leg. Suddenly, your character flaw becomes a solvable medical problem. You voluntarily agree to undergo treatment et voila, you’re no longer so disruptive at home and at work; life is better for you and those around you. In contrast to some types of moral interventions (blame, punishment), the medical approach deliberately tried to minimize your suffering. Further, it didn’t take your self or character as inviolate or immutable, but accepted the possibility that you might have to change, perhaps dramatically, to solve the problem.
This instance of medicalizing the moral domain seems relatively unproblematic, precisely because there’s a clear causal connection between a brain defect and bad character and behavior. There seems no good reason not to accord to medicine what morality can’t manage. Of course this example is blatantly fictional since we’re nowhere near having the technical capacity to precisely identify and fix all the brain systems responsible for particular characterological defects and behavioral disorders. But the principle seems unproblematic: a defective sub-personal part or system responsible for seriously objectionable behavior should be fixed, if the patient voluntarily consents to treatment.
This principle already uncontroversially applies in many cases of mental illnesses and behavioral disorders that have identifiable causes in brain system disregulation, such as addiction, depression, anxiety and obsessive-compulsive syndromes, and tumor-induced mania and sexual compulsion. We unhesitatingly use appropriate drugs, surgery, behavior therapy and other medical and psychiatric interventions even though there might be a moral dimension to the dysfunctional behavior in question, e.g., deception and thefts by addicts to support their habits. As we get better in tracing the connections between brain and behavior, and as our neuro-psychiatric techniques improve, the range of disorders to which such interventions are applicable will, and should, increase. They might include, for instance, diagnoses such as anti-social personality disorder, borderline personality disorder and psychopathy, assuming these turn out to have identifiable neural etiologies or can otherwise be addressed by medical and psychiatric interventions. At the same time, we won’t any longer see the point of punitive moral responses involving blame, social sanctions or punishment that have no behavior-guiding effect. Where medicine goes, morality retreats.
Barriers to Medicalization: Folk Dualism and Free Will
Or it would, were it not for the retributive aspect of morality mentioned above, combined with the fact that most people don’t take the naturalistic view of persons and selves as constructions. Many, perhaps most folk are what I’ll call folk dualists: they believe that we possess indivisible souls, or at least something categorically non-physical or mental in addition to the brain. Souls, it is widely thought, have a power of free choice – a contra-causal free will – that transcends the complex but ultimately mechanistic workings of the brain’s neurons and neurotransmitters. Whatever the physical states of the brain, body and environment were at the time of the choice, the essential person – the immaterial self-soul – could have chosen otherwise. As neurophilosopher Joshua Greene points out, it’s a commonplace for people to wonder concerning a crime: is it him or his brain that’s responsible?
Such self/brain dualism helps to reinforce the categorically retributive understanding of morality, which says the freely willing I deserves blame and punishment for choosing to act immorally or criminally. True, in cases of extreme mental illness with clear physical origins (a tumor-induced sexual obsession, for instance), many folk dualists will concede that the brain plays a role in generating bad behavior. But otherwise the default assumption is that the immaterial or mental self could have chosen not to commit the crime, whatever her brain was like and whatever her circumstances at the time. Since she did so choose, she deserves blame and punishment.
But commonsense self/brain dualism is under increasing attack from neuroscience, and as the belief in the soul and its contra-causal free will declines, the reach of medicalization will grow, and should grow, at the expense of retributive morality. Once we see that the brain does it all and that brains can be fixed, then, technology permitting, there’s no good reason not to take a medical approach to behavioral problems, including those that result in crime and abuse. The moral domain will shrink insofar as problematic behavior that used to be chalked up to a character defect, involving the exercise of contra-causal free will, becomes modifiable using drugs, surgery, behavioral and talk therapy and other sorts of medical-psychiatric interventions. The question becomes: why punish the person (often ineffectively) when they can be changed (voluntarily, of course)  for the better? Without the freely willing soul as the target of retributive blame, it’s difficult to justify the infliction of suffering that serves no non-retributive purpose.
But medicalization and the end of retributive morality doesn’t by any means spell the end of morality. If, thanks to the good offices of doctors, psychiatrists and neuroscientists, we’re maintained in good neural operating condition, then the primary means of keeping behavior within ethical boundaries still remains the normative framework of holding each other morally responsible. As sane, adult members of a community, we know the consequences of breaking promises and laws, and the prospect of being held responsible for breaching social contracts helps to keep us in line, often pretty effectively. Of course if we’re lucky we’ve been taught to want to conform to moral norms – not to lie, cheat, abuse, steal, rape or murder. But even for goody two-shoes, explicit codes of conduct and the threat of social and legal sanctions help to maintain good intentions when the temptation to misbehave arises, as it often does. So despite the encrochments of medicine in addressing what were once thought of as moral defects of the soul, the moral domain still has its rightful, essential place in our collective behavioral economy. It’s just that when it comes to punishment its rationale is now entirely consequentialist. Suffering need not and should not be inflicted for retribution’s sake, but only if it does some good and there’s no non-punitive alternative means of guiding goodness.
Public Health and the Epidemiology of the Criminal Mind
There are problems for consequentialist morality, however, even on its own terms. All too many mechanically sound, that is, basically sane and rational people, haven’t learned to be morally upright promise-keepers. For reasons having to do with their genetic consititution (e.g., an inherited predisposition to an impulse control disorder) and/or environmental circumstances (a chaotic or ethically lax household, school or community) some adults end up predisposed to be abusive, deceitful, and otherwise disinclined to obey moral and legal rules. Moreover, in some cases these same genetic and life circumstances conspire to make them less cognizant or susceptible (or both) to the threat of sanctions as deterrents to bad behavior. An individual with these propensities – a career gangster who enjoys beating people up, let’s suppose – is much like you were with your hypothetical temper control problem (see above): at high risk of flouting social norms, which then makes him a target for criminal sanctions which, unfortunately, may not be particularly effective because of these propensities.
The difference between you and him is that his morally problematic behavior – violent crime, chronic interpersonal abuse – isn’t directly traceable to a specific neural functional deficit or brain anomaly. Instead, it’s part of his character as acquired over the course of his life thus far, the complex interaction of genetic endowment and experience. His brain, which physically instantiates all his behavioral propensities, reflects the current net result of this interaction (see note 8). But there isn’t an appropriate target for medical intervention at the physical, sub-personal level since there isn’t an identifiable neural disorder. Rather, we’ve got a more or less normal brain – the neural hardware – instantiating morally problematic behavioral “firmware” installed by life experience – the criminal mind, let’s call it. Without a clear target for medical intervention, it seems we’re forced to take an exclusively moral, not medical, stance with respect to such individuals. But, because they aren’t that controllable by the threat of blame or punishment, many end up cycling repeatedly though courts and prisons, while others continue their careers of transgression undetected. The moral-legal domain unequivocally rules in these cases, but unfortunately doesn’t do a particularly effective job. Every crime or episode of abuse, after all, constitutes a failure of punishment to deter. Indeed, the evident shortcomings of moral-legal deterrence suggests that the moral stance, although a necessary component of our approach to crime and immorality, is hardly sufficient.
What to do? Medicine, broadly conceived to include public health and behavioral rehabilitation, can again come to the rescue. Crime, addiction, family violence and abuse can be understood from an epidemiological standpoint, as preventable and treatable threats to public health (see here and here, for instance). These behavioral syndromes have their origins in specific conditions that shape persons starting at conception. Were we to take a pro-active approach in controlling these conditions, we could likely significantly reduce the chances that individuals acquire propensities for crime and anti-sociality in the first place. Those who still end up acquiring them can, to the extent possible given developments in psychiatry and behavioral technology, be given new, productive behavioral repertoires. Either way, the criminal mind becomes the target of intentional, coordinated epidemiological and behavioral interventions that primarily seek to prevent and cure, not punish.
Personal and Social Environmental Protection
What would it take, concretely, to actually reduce the prevalence of criminal and abusive propensities in a given at-risk population? (Note that this is not the question of reducing the prevalence of crime tout court, since we’re seeking a solution that comes before engaging the deterrence, detection and punishment handled by criminal justice.) Well, it would take interventions that address the entire development of a person starting at conception, looking at each stage of life and each hour of the developing person’s day. The point of such programs would be to create personal and social environments that increase the chances of healthy physical and mental development while acquiring positive, productive behavioral repertoires. Stating the obvious: the fetus shouldn’t suffer the insults of maternal malnutrition, substance abuse, trauma or violence. Once born, babies and children need a good diet, protection from environmental toxins (e.g., lead in paint, water and air), exercise, opportunities for cognitive development, and non-violent, cooperative behavioral role models provided by parents, teachers and peers. Schools must provide the same environmental protections and learning opportunities, and the wider community – neighborhoods, workplaces, institutions, and government – must also provide environments that maximize the chances of healthy physical, emotional, and cognitive development and that reinforce norms of cooperation and non-violence.
Of course many families, schools, and communities in the United States already provide such environments as part of their socio-cultural heritage. The result, unsurprisingly, is far less dysfunction, crime, abuse and addiction. What’s being suggested here is the intentional creation or reconstruction of such environments as a coordinated public health strategy for families and communities without such a heritage. Depending on one’s politics (e.g., libertarian vs. liberal), the scale and reach of the contemplated interventions, and whether they are privately or publicly conducted, such a strategy may seem either too impractical or too intrusive or both. Piecemeal and occasional programs that teach parenting skills, provide after-school recreation, or find jobs for adolescents might be fine, but the soup to nuts approach of targeting all personal and social environments to promote healthy development and guide goodness seems beyond the pale for some. Yet it is only such a comprehensive and integrated approach that can turn the corner on entrenched patterns of crime and abuse stemming from historically dysfunctional families and communities. Piecemeal programs can help, but to the extent that any personal or social environment remains toxic (an abusive parent, chaotic school, dangerous neighborhood, etc.) to that extent the person’s chance of growing up good is compromised.
Again, it’s important to remember that an epidemiological approach to preventing the criminal mind (instantiated in the physical brain) doesn’t in the least obviate morality, since person-level responses involving praise and blame still play an essential role in shaping behavior. After all, the very process of socialization in families and schools, as well as everyday interactions in daily adult life, depends on deploying basic interpersonal rewards and sanctions. The key, however, is to make sure that individuals don’t grow up learning abusive styles of interpersonal control. Part of being good, after all, is to not punish unnecessarily and too harshly when others – children, spouses, neighbors, co-workers – need correcting. In short, the moral domain itself must not be needlessly punitive, and this too is a matter of acquiring humane behavioral repertoires. Seeing through the twin myths of the immaterial soul and contra-causal free will, then dropping retributive morality and the notion of deeply deserved, non-consequentialist punishment will make this all the easier. We’ll stop playing the blame game to the hilt, as we do all too often, and look to the wider context of what determines someone’s faults and virtues.
Mixed Cases: Medicine and Morality
Returning to the problem with which we started, what becomes of our mentally ill criminal offender in light of the accommodation between medicine and morality? We presumptively supposed he has two problems, one medical (mental illness), one moral (bad behavior). In our current criminal justice system, the retributive response to bad behavior (imposing deserved punishment) takes precedence over addressing mental illness. Under the proposal recommended here, our priorities should be reversed and the retributive component of the moral domain abandoned.  Treating mental illness, whatever its severity and etiology, should take precedence because it compromises the very capacity to behave well and be responsive to moral evaluation. A serious impulse control disorder, for instance, makes it harder to conform to norms of right and wrong of which the otherwise rational offender is probably fully cognizant. Only when sufficient capacity for self-control is restored does it make sense to apply moral rewards and sanctions that depend on that capacity to be effective.
To the extent that dysfunctional and harmful behavior is a consequence of an identifiable disorder, appropriate treatment should be made available on a voluntary basis (but see note 6). To the extent that someone is rational, sane and thus still a moral agent, consequentialist moral responses remain appropriate so long as they don’t compromise the goals of treatment. This is to say that sanctions, including fines, restrictions of liberties and involuntary detention still play a role in guiding goodness, but that the burdens imposed on the offender should not impede or conflict with the conditions necessary to restoring him to mental and physical health.
Practically speaking this is a tough line to walk, since after all the deterrent effect of criminal sanctions depends on their being at least somewhat punitive. How is it possible to deter the rational component of an offender via punishment while simultaneously fixing the irrational, impulsive, characterological components, should these exist? Answer: with considerable difficulty, perhaps. But facing this conflict, and not allowing the moral to trump the medical, is what’s required of us once we stop being retributivists. If we assiduously keep in mind that the goal of both medicine and morality (now revised) is producing good behavior while minimizing suffering, then we can design interventions, including involuntary detention, that work in both domains. Speaking to the moral domain, we must follow through on threats of legal sanctions for them to be effective deterrents, so we must in some cases restrict liberty (and of course such restriction can serve other goals as well, such as public safety or “quarantine” as philosopher Derk Pereboom puts it). But, speaking to the medical domain, because we must also fix what’s fixable such restriction obviously shouldn’t be further damaging to mental or physical health, and indeed must be conducive to its restoration.*
In a humane medical-moral system, the patient/client/inmate is given all possible medical-psychiatric help to restore normal control capacities, and if she’s judged to have a sufficiency of such capacities, she’s held to be morally competent: responsive to the prospect of being held responsible. Then, ideally, she participates in a maximally rewarding, minimally punitive moral community.
The design of such a system presents a formidable, but not insuperable challenge, involving further developments in behavioral technology, medicine, psychiatry and rehabilitation. It’s a project that should involve our best minds, in service to laudable goals. But the difficulty of simultaneously curing and deterring mentally ill offenders simply highlights the importance of prevention as suggested in the previous section: we’re far better off creating family and social conditions which reduce the incidence of mental illness, character disorders, behavioral pathology, criminality and abuse in the first place. Such prevention minimizes the need for both treatment and deterrence, and so is the best route toward a less punitive, less disordered society composed of healthy, ethical individuals. Plus, it’s probably a lot less expensive.
Conclusion: The Mutual Constraint of Medicine and Morality
The dawning understanding that persons and selves are not immaterial, irreducible souls, but psycho-behavioral constructions of brains and bodies helps to reveal the common behavioral focus of the moral and medical domains. Since persons are constructed, they can be reconstructed, which gives us many options besides praise and blame in our quest to become better persons. Morality, whether informal or codified in criminal justice, is one traditional means of shaping behavior, that which takes the person as the more or less immutable target of rewards and sanctions. On the other hand, medical science, augmented by psychology and sociology, understands the physical and functional constituents of persons and how they are constructed by nature and nurture. Such understanding gives us a wide range of alternative means of behavior control, in that the very mechanisms of personhood and the very development of persons can become targets of interventions to reduce the prevalence of crime, abuse and other morally problematic behavior. What were once seen as categorically moral defects can now be understood, treated and prevented as behavioral dysfunctions. Some dysfunctions have clear neural etiologies, amenable to drugs, surgery and other neuro-psychiatric techniques, techniques which are improving continually. But some are the result of life-long character development, and these are best prevented in advance by public health interventions at the family and social level. Absent a specific treatable brain disorder, altering the entrenched neural firmware of personality is difficult, in some cases impossible (although what’s impossible now may not be later).
Given that we’re constructions, not essences, the worry might arise that the self is up for grabs, that the very basis for personal autonomy and inviolability is at risk. But this is a non-sequitur. Learning that one is a neural construction doesn’t diminish the physical, psychological and moral salience of personhood, since after all we’re hard-wired to take being our particular selves very seriously. We’re not able to give up, nor should we give up, our root desires for personal autonomy and inviolability. This means that in the age of neuroscience persons still remain the primary objects of moral concern, that the deontological obligation to respect the autonomy rights (life, liberty, pursuit of happiness) of each and every person properly remains one of the core values in our moral economy. The default assumption unequivocally remains that it’s the person as currently constructed that gets to decide whether or not to undergo a treatment that might substantially change her. Likewise, policy on public health interventions remains subject to open democratic debate, which will prevent the compromise of essential liberties as judged by the participants. In short, there’s no reason or justification in the naturalistic, non-essentialist view of the person for medicine and public health to become Big Brother. Morality thus constrains the march of medicine.
But equally, medicine and science shape morality, since without the freely willing soul to take ultimate credit and blame, non-consequentialist moral retribution loses its primary justification. This will help keep the punitive side of morality in check. True, until we all grow up to be saints the necessity for moral-legal deterrence remains, so it isn’t as if there’s suddenly no role for morally justified punishment. But this will increasingly be seen as an unfortunate byproduct of the extent to which we haven’t gotten right the formation of humane, non-violent character and the non-punitive, reward-oriented control of behavior. Life will always give us hard knocks, but there’s no reason to add to them unnecessarily.
© Thomas W. Clark, July 2007
*Treating addiction is a classic mixed case. Addicts have altered their brain chemistry and structure in ways that make the prospect of moral evaluation less salient than their drug cravings. Yet they still retain some rational capacity to respond to person-level incentives (privileges, cash, vouchers, etc.) such as used in contingency management. So the trick is to do what's possible on the medical side to attenuate an addict's cravings, e.g., using medications such as methadone and buprenorphine, while providing a graduated series of behavioral interventions that engage his surviving rational-moral capacities. These capacities will regenerate if all goes well. Seeing that the addict didn't contra-causally choose to become an addict helps to keep the interventions from being unnecessarily punitive (more details here and here).
 Massachusetts has experienced a recent spate of inmate suicides, sparking a call for reforms in the treatment of offenders with serious mental disorders; see Inmate kills self as prison suicides continue. The rise of the mentally ill in prison populations is discussed in The mentally ill, behind bars (New York Times).
 Although the tide may be turning; see for instance Just say no to retribution by Edward Rubin.
 See for instance Thomas Metzinger’s Being No One: The Self-Model Theory of Subjectivity (Cambridge, MA: MIT Press, 2003).
 Resisting the power of temptations: the right prefrontal cortex and self-control, Daria Knoch and Ernst Fehr, Annals of the New York Academy of Sciences, January 28, 2007.
 I’ll call it the soul for convenience, but it’s the non-physical or mental aspect of the self that’s being refered to, nothing necessarily religious. Leon Kass, former chair of the President’s Commission on Bioethics, defends the soul here.
 In some instances of extremely disordered individuals who aren’t otherwise controllable, it might be allowable to intervene without consent, but such cases would require the utmost care and deliberation in light of possible human rights abuses.
 Nevertheless, some skeptics about contra-causal free will still hold retributive views, for instance law professors Stephen Morse and Michael Moore. Some think that the only check on excessive deterrent punishment is the retributive conception of giving people the punishment they deserve, and no more. But deterrence is quite plausibly constrained by the countervailing value of personal liberty. The limit on draconian punishments, therefore, need not rest on retributive desert, but simply seeing that the desire for liberty can outweigh the need for deterrence in our value hierarchy. Deciding where on our collective hierarchy these values should lie is of course a vexed political debate with no clear solution.
 Regarding the influence and interaction of environmental and genetic factors on crime, see for instance Caspi, et al, Role of genotype in the cycle of violence in maltreated children, Science, V 297, 2002.
 On this point see Greene and Cohen’s For the law, neuroscience changes nothing, and everything, and Mobbs, et al., Law, responsibility and the brain.
 Of course morally neutral interventions, including restrictions on liberty, may be necessary for public safety in certain cases, e.g., restraining the seriously delusional psychotic, or chronic sex offenders with a high probability of relapse. I don’t pretend to have addressed the vexed conflict between personal freedom and public security that such cases raise.
 About this worry, see Maximizing liberty: retribution, responsibility and the mentor state.