Re: Morse and Retribution

University of Pennsylvania law professor Stephen Morse sticks to his retributive guns in the light of naturalism. Why? Because he thinks retribution is intrinsically good, increases human dignity, and therefore produces justice. However, the intrinsic good of retribution and how punishing people without regard to consequences increases human dignity are left obscure on Morse's account. Under naturalism, we need not appeal to either retributivism or deontology to protect human autonomy rights, which are the driving desiderata on a consequentialist understanding of morality.

Speaking at a recent American Enterprise Institute conference on The New Neuromorality, University of Pennsylvania law professor Stephen Morse argued that neuroscience will have little impact on our moral and criminal responsibility practices. Current neuroscience offers no radically new insight into ourselves that might force us to reconsider our notions of personhood or culpability, or that might suggest revisions to our rationales for punishment.[1] In particular, Morse says retribution should remain a central principle of criminal justice. His deflationary view of the implications of neuroscience for the law contrasted sharply with that offered by Princeton neurophilosopher Joshua Greene. Greene argued that since neuroscience definitively rules out the existence of the soul and contra-causal free will, retributive justifications for punishment are called into question. Instead, we should move toward a deterrence-based system of criminal sanctions, a system he sees as more humane.

Both Morse and Greene agree about the non-existence of the soul and free will, and both think that lawyers, judges and juries should stop talking about free will (“silly talk,” says Morse). But why does Morse think that retribution still has a claim on us, even though, as he puts it, he’s a “good-enough-for-government-work determinist,” a materialist, and even a moral non-realist?

Rationality and responsibility

As a materialist and determinist, and therefore a naturalist, more or less, Morse properly notes that our conception of responsibility can’t rest on the supposition that we achieve moral agenthood by virtue of being uncaused in some respect. For persons are always and everywhere caused to be as they are, and to act as they do, as neuroscience makes clear in spades. No, moral agency inheres in the fact that people are by and large rational, guided in their behavior by reasons, and the law and morality provide behavior-guiding rules and sanctions. Our responsibility practices are targeted to those agents who are capable of being thus guided, and that capacity defines them as moral agents. Those who aren’t guidable or who act under sufficient duress we excuse. There’s nothing in all this that requires any exemption from determinism, and in fact an exemption would make people less guidable and less rational, since the prospect of criminal sanctions wouldn’t help determine their behavior.

Deontological retributivism

This all makes perfect sense, but Morse goes on to argue that there are two components of our responsibility practices in the West. One is consequentialist, which seeks to produce social benefits from punishment (e.g., public safety, deterrence, rehabilitation, etc.), and the other is deontological, which enjoins us to do what is right, independent of any costs or benefits that may accrue. Morse says that retributive punishment – giving people what they deserve – is essentially deontological, not consequentialist:

Some things… we do only because it’s right to do them. For example, we give people what they deserve. Why? Because it’s right to do so. Not because we’re going to produce good consequences. If we do produce good consequences, terrific. But that’s not necessarily why. We do it because we think they deserve it.

Deontological guidance

But deontological retributivism seems to conflict with Morse’s account of criminal justice. After all, if the function of our punishment practices is to guide moral agents in their behavior, that seems very much a consequentialist rationale, since it’s good behavior that we’re seeking as a consequence of imposing criminal sanctions. So how does consequence-ignoring deontology fit in? Addressing this issue, Morse assimilates the deontological component of the law to his guidance view, saying:

So when we look at our conduct guidance systems, we see deontological conduct guidance – don’t torture small infants who are innocent – even if you think you’re going to produce good consequences from doing so. And we see consequential reasons, let’s see a net decrease in crime.

Now, if retribution is part of the deontological component of the law, as Morse claims, this must mean it plays a role in deontological conduct guidance. But what is this role? It can only be to guide our conduct in imposing punishment in ways that are presumably good. Retribution – giving people what they deserve – enjoins us to punish the guilty independent of any social benefits or costs that might accrue, as Morse explicitly points out (see his first quote above). The deontological retributive principle guides us towards the presumptive good of giving people their just deserts, just as the injunction to not torture babies guides us toward the good of happy babies. In neither case, says Morse, should consideration of external consequences influence our behavior; only the principle of doing what’s right has force.

The autonomy right

It’s intuitively the case that babies, not to mention children and adults, shouldn’t be tortured or unjustly harmed to achieve social benefits. We don’t hang an innocent man to satisfy a lynch mob that would otherwise torch the jail. Why? Because at or near the top of our Western values hierarchy is the belief that human beings have an inviolable right to their basic autonomy – call it the autonomy right. Each of us deeply desires to be treated as an end in ourselves, not instrumental to the ends of others or the state. This desire is encoded in a basic principle of morality and the law: do not unjustly compromise the autonomy right, except in extraordinary circumstances.[2] As Morse suggests, our behavior is strongly guided by this principle, which is often thought of as irreducibly deontological. But in fact we can straightforwardly reconstrue it in consequentialist terms: we very much desire our autonomy rights, so we have nearly unbreakable guidance rules that produce the beneficial consequence that such rights are preserved.[3]

The costs of retribution

But what about retribution? Is it a good that, like the autonomy right, should trump other goods in guiding our behavior? It isn’t clear that it is. Morse plays the apologist for retribution, saying:

Retribution, by the way, is not revenge; retribution is giving people what they deserve. And the theory of retribution is, when people intentionally and without justification or excuse inflict harm on their fellow human beings they deserve to have some kind of negative reaction… there’s meant to be some negative sanction. It doesn’t mean you’re harsh or you’re nasty… it just says that they should get what they deserve. Now suppose, as I do, that if you treat people as potentially desert-bearing creatures that you increase human dignity, that you increase the notion of a life worth living, is that something that science tells me I have to give up? It doesn’t and it can’t.

For Morse, retribution is not only right, it needn’t be harsh; moreover, it’s dignity-enhancing and makes life worth living in a way it wouldn’t otherwise be. But in the real world does retribution substantially contribute to human dignity and lives worth living? Despite Morse’s soft-pedaling of retributive harshness, the “negative sanction” of just deserts as carried out in our criminal justice system often involves very harsh punishments indeed - long prison terms in punitive environments without rehabilitative amenities, or even the death penalty. Retributive suffering, as Morse points out, need not result in any social benefit, and often has significant personal and social costs. Prisoners primarily subject to retributive punishment are (quite logically) thought not to deserve safe and clean living conditions, education, counseling, mental health services or skills development, and are often returned to society in worse shape than when convicted, posing a continuing threat to their communities and a drain on resources. How, one wonders, does all this increase dignity or make lives more worth living? (Keep in mind that possible side effects of exacting retribution, such as respect for the law, public safety and deterrence, can’t count in calculating deontological dignity.)

Perhaps the otherwise benefit-free suffering of the offender confers dignity on the victims of crime and makes their lives more worth living, since retributive satisfactions are, after all, not inconsiderable. But this isn’t what Morse has in mind, since he specifically says that giving people what they deserve need not involve the nasty pain or death that ordinarily conveys such satisfactions. So it isn't clear from his remarks what dignity or meaning inheres in retribution, especially something that should trump other tangible goods. He simply says it is right to punish culpable, unexcused wrongdoers, and its rightness makes punishment a deontological good that should guide our behavior as punishers. But since the offender’s suffering need entail no social benefit, and since affording retributive satisfactions isn’t its object, it’s hard to see what’s good about it, especially if we believe, as does Morse, that there is no deeply deserving soul or free will that merits punishment. To say that inflicting consequence-free suffering is right just begs the question of why it is right.[4]

Abandoning retributivism

It is true that inflicting certain deprivations and suffering on offenders can sometimes serve a legitimate social purpose, namely to guide their and other’s behavior towards goodness, as Morse puts it. But this is not retribution, it’s consequentialism. On a humanitarian consequentialist analysis, we should only impose suffering to achieve the goals we desire when no other means suffice, making sure not to violate the autonomy right, since that’s primary among the goods we seek to protect and advance. But if the goal of the law, as Morse believes, is to guide those in its custody (and the rest of us potential miscreants) toward goodness, the retributive imperative to punish irrespective of the effects on such guidance seems irrational, a purposeless violation of the autonomy right. Morse might say retribution has it’s own separate, possibly conflicting good that it achieves when it guides us as punishers to impose just deserts. But again, why should we think that retributive suffering, when it achieves no behavior-guiding purpose for the subject of punishment or any other potential offender, is a good?

On Morse’s deontological retributive principle, only rational and thus guidable agents deserve punishment when they voluntarily break the law; so clearly desert tracks an offender’s guidability. Yet Morse and other retributivists such as law professor Michael Moore must deny that desert is equivalent to guidability, since after all that would reduce deontological retribution to consequentialism. On the one hand, they hold that desert accrues only to agents who possess just those characteristics that allow them to be sensitive to anticipated consequences, but on the other deny that such behavior-guiding sensitivity is the rationale for desert. Guidability, based in rationality, is doing all the work in picking out moral agents, yet retributivists insist that such agents are proper targets of punishment even if no guidance, or negative guidance, ensues.[5] But since there’s no clear deontological good about punishment (except that it’s “right” to punish, which begs the question), retributivism seems difficult to justify. By contrast, the consequential goods of punishment (when they obtain), goods which actually depend on an agent’s guidability, can justify a non-retributive, more humane rationale for criminal justice, as for instance recommended by Joshua Greene. On a naturalistic, consequentialist reconceptualization of desert, offenders deserve no more punishment than what’s necessary to guide goodness.[6]

As Morse says in his last quote above, there’s nothing in science that can force him to abandon retributivism, but having abandoned the supernatural soul and contra-causal free will as the locus of ultimate, buck-stopping responsibility, there seems no good reason to hold onto it. He rightly says that neuroscience doesn’t undermine our notion of reasons-responsive personhood, but rather helps to explain it. But treating people as persons doesn’t logically entail or otherwise require retributive punishment, even though it does require respect for their fundamental rights. Morse’s challenge, should he choose to accept it, is to demonstrate the positive, dignity-enhancing aspect of suffering that serves no consequential benefit. Otherwise, he should abandon retributivism as an unnecessary and inhumane holdover from a time in which offenders were really thought to be self-made, and thus really deserving of suffering, whether or not it reforms or deters.

TWC, August 2005


In personal correspondence, Morse says that a purely consequentialist understanding of morality and the law would necessarily permit treating people instrumentally, and that therefore deontology will inevitably be smuggled back in to block utilitarian excesses. In particular, he thinks that to deny the deontological claim of retribution necessarily implies a pernicious, rights-denying consequentialism, what he calls a "good bacteria, bad bacteria" view of human beings, so we mustn’t abandon retribution. But as I've argued above, our rights don’t depend either on deontology or retribution. Rather, our fundamental desire to be treated as ends in ourselves – as persons – is a cardinal value that guides behavior, such that we act in order to preserve the autonomy right. It’s widely agreed that this is an unequivocal good to be achieved, a primary consequential outcome of our moral and legal practices. But it isn’t at all clear what fundamental right or good is inherent in retribution, even though Morse sticks with his claim that it somehow (but how, precisely?) contributes to human dignity and “produces justice.”


[1] Curiously, Morse thinks that a thorough understanding of the neural mechanisms of consciousness, rationality and intentionality (perhaps achieved in the distant future, he speculates) might show we’re not really persons. But for now, “until that revolution in the sky comes, we don’t have to give up our personhood.” This is puzzling because as a naturalist he already believes there are mechanisms that instantiate personhood, so why should discovering their details pose a further threat? About persons, brains and responsibility, see The responsible brain.

[2] A fictional example of such circumstances: The TV series MASH aired an episode involving a Korean mother hiding under a tarp with her newborn baby and several other refugees. In order to prevent the group’s discovery by enemy soldiers nearby, and certain death for everyone, she smothers her baby’s crying, knowing that this must kill it.

[3] Whether deontology ultimately reduces in all cases to consequentialism is a question well beyond the scope of this paper, but see note 6 below.

[4] See here, here and here for other examples of begging this question.

[5] In his paper Reason, Results and Criminal Responsibility Morse writes that “…moral and legal rules are action-guiding, at least in large part, and thus they are concerned with reasons for action. Moral condemnation and desert and punishment should, therefore, apply only to phenomena capable of being guided by reason” (p. 377). If desert only applies to action capable of being guided by reason, one wonders why desert still applies in the absence of any guiding function, as Morse says it should.

[6] Steven Pinker, speaking at the same conference, suggested that the deontological principle to punish, no matter what the consequences, can be understood as the pragmatic implacability of the law. That is, the deterrent threat of punishment is a credible threat only if punishment is actually carried out, whatever the countervailing considerations. This is another example of how deontology might reduce to consequentialism.