Abstract and Keywords
This chapter explores the relationship between philosophy and criminal law. It first describes the link between criminal law and action theory, with emphasis on the elusive notion of an action, when omissions can be interpreted as actions, actions that are voluntary, and actions that correspond to their legal descriptions. Attention then turns to the relationship between criminal law and the philosophy of mind, giving consideration to when and why a mind is deemed guilty. These considerations lead directly to the discussion of the morality of punishment, and to the analysis of retributivism and consequentialism. The chapter concludes by assessing the relationship between criminal law and political philosophy, focusing on liberalism and on the instances of wrongdoing that justify the state a in inflicting hard treatment on criminals.
I. Introduction to the Philosophy of Law
Philosophy, as is well known, is the quintessential “meta” discipline, in the sense that virtually every other discipline can be studied philosophically. Like virtually any other discipline, the criminal law, too, can profitably be approached from a philosophical perspective. But the criminal law is exceptional in that its relationship to philosophy is uniquely intimate. While there can be great biologists or historians who do not study biology or history philosophically, it is hard to imagine a criminal law scholar, or even a practitioner, who does not approach her subject—even if unawares—somehow philosophically. For, as we shall see, most fundamental discussions in the criminal law are philosophical through and through. This chapter will be divided into four sections, each focusing on a large area of philosophical activity which overlaps substantially with the criminal law.
Since, in principle, the criminal law can only be set in motion when a crime has been committed, and, in principle, a crime can only be committed when someone acts (i.e. when someone does something), Section II is devoted to the relationship between the criminal law and philosophical action theory. Themes to be discussed in this first section include the surprisingly elusive notion of an action, when omissions can be taken to be actions, when merely preparatory steps to commit a crime can be taken to constitute an attempt, when actions can be taken to be voluntary, and when actions correspond to their legal descriptions. (p. 132)
That someone has acted in a way that has been previously described as a crime, however, is not enough to claim that a crime has been committed, in the sense that the person who acted in this way is thereby liable to punishment. As the famous Latin slogan has it: actus non facit reum, nisi mens sit rea. In other words, we need both a “guilty act” and a “guilty mind.” Section III is thus devoted to the connections between the criminal law and the philosophy of mind. The crucial question that shall occupy us is: When is a mind guilty? A person cannot justly be said to have acted with a guilty mind if she was not capable of, say, distinguishing right from wrong, or appearances from reality—if she was, for example, a child. But even if she was fully capable, she may be more or less culpable (i.e. blameworthy), depending on the mental states that accompanied her action.
Assuming someone committed a crime, and that she not only was in control of her faculties, but committed it with a culpable mental state, then, in the standard case, she will be punished. But what exactly is punishment? Despite various disagreements, all definitions of punishment agree that its essence is to inflict something painful or unwelcome to the criminal. Of course, the assumption is that she has done something bad herself—that is precisely why she must endure pain. Still, the fact that punishment seeks to hurt people gives us plenty of reason to inquire about its moral justification. Why exactly is the state justified in hurting people? The morality of punishment will be discussed in Section IV.
Lastly, in Section V we will discuss the relationship between the criminal law and political philosophy. Even if the state is morally justified in hurting criminals, we still need to know which instances of wrongdoing merit such treatment. After all, not every bad thing that people do calls for the state to react punitively, or for the state to react at all. Even in cases in which it appears clear that the state ought to react punitively to wrongdoing, the question as to how much punishment ought to be inflicted deserves attention.
As will become clear, there are conspicuous overlaps between the different themes discussed herein. The fourfold structure of the chapter does not seek to suggest any precise boundaries between these broad areas—neither in philosophy generally nor in the criminal law however narrowly construed. The section headings merely seek to highlight the conceptual proximity between the concerns of the criminal law and well-established areas of philosophical investigation. Appreciating such proximity may both facilitate the exposition of the material and the advancement of the relevant debates.
II. Action Theory
The raison d’être of the criminal law is to punish criminals. But, as stated at the outset, a person cannot be punished unless she has (or is supposed to have) committed a crime. And a crime cannot be committed unless someone has done (p. 133) something—unless, that is, someone has acted. This is captured by the “act requirement,” also known as the “voluntary act requirement” of the criminal law. These rather straightforward assertions highlight the importance of the philosophical study of human action for the criminal law. But the assertions also immediately begin to adumbrate some of the complexities of the criminal law’s notion of action. The sense of “doing something” operative in the criminal law has to be quite general: crucially, criminal “doing” must include both “attempting” and “omitting.”
The first problem that confronts us when discussing attempts appears to be mostly descriptive. Think of someone who wishes to detonate a bomb at a theater. The suggestion that this mere wish, unaccompanied by overt action, is already an attempt to blow up the theater is surely misguided. But what about the person who not only merely wishes to detonate a bomb, but buys some ingredients for manufacturing bombs? Or what about one who buys all such ingredients? Or one who buys all the ingredients and who also scouts the theater where she will place the bomb? What exactly should we require in order to compellingly claim that someone’s preparations amount to a legal attempt, and that this person’s actions thus fall under the purview of the state’s punitive power?
These questions suggest that there may not be a fixed point in nature which neatly separates attempts from merely preparatory steps. To a certain extent, whether or not we are in the presence of an attempt is already a normative question: within some limits, different criminal law systems may decide differently when a certain behavior constitutes an attempt. As a matter of fact, this sort of imprecision holds for most other inchoate (or “incomplete” crimes, e.g. conspiracy and solicitation) crimes as well. The imprecision holds, too, for the most fundamental of questions: whether or not an event is an action. That the very term “action” is the result of a normative decision (the decision to ascribe responsibility) was a view that Hart once defended, though later in life he abandoned it.1
An easier case is that of completed attempts: when a person does everything (she thought) necessary for the completed crime.2 Imagine someone who intends to kill many innocent people, and who does everything reasonably necessary to have a bomb explode at a movie theater. Although she completes her attempt, due to some wholly unforeseeable event, the bomb somehow fails to explode. It is obvious that there is no completed crime here (the bomb did not go off, no one was injured, etc.), but it is no less obvious that there was an attempt, and, ex hypothesi, that the attempt was completed. (p. 134)
Completed cases allow us to see another even more unquestionably normative problem with attempts. Consider two agents, each endeavoring to blow up a (different) theater; they both have the same intentions and in fact perform the exact same actions; but where one succeeds, the other fails. Most jurisdictions recognize a difference between the punishment for attempted and completed crimes. But why should there be any difference between these two agents’ punishments? Punishing one more leniently than the other understandably strikes many as letting moral luck run amok.3 Our moral intuitions seem to converge on the idea that at least completed attempts deserve the same punishment as completed crimes.
Should uncompleted attempts be punished at all (and, if so, how sternly)? Here intuitions vary considerably, ranging from the view that only completed attempts should be punished to the view that some uncompleted attempts should be punished, albeit more leniently than completed attempts. The rationale for not punishing incomplete attempts stems from the already mentioned difficulty in fixing the exact point at which the incomplete attempt pops into existence. The rationale for punishing them, but more leniently, boils down to two interrelated arguments. First, by abandoning her project to commit a crime, the incomplete attempter thereby deserves less blame than had she gone on with it. Secondly, society should never remove incentives to desist from doing wrong—and a lesser punishment is such an incentive.
The other way in which the criminal law’s conception of action is broad concerns the fact that some omissions, that is, some failures to act, constitute actions. Consider two nurses who have been for a long time planning to kill a patient in the hospital in which they work; they agree that the first nurse’s signal to the second that it is time to kill their patient would be for her to stand completely still, “doing nothing.” Obviously, “doing nothing” is not here really doing nothing: it is a very real and concrete part in the nurses’ plan to kill their patient. It is hard to accept that one of the nurses should escape punishment because she “did not do anything.”
One obvious way to avoid such suggestion would be to insist that even if during the specific time of the shooting one nurse remained still, surely she participated in the planning of the murder—and this was to do something. This is true, but it highlights problems with the individuation of actions. Imagine that the nurses began planning their crime 20 years before the patient ever entered the hospital: it seems odd to say that the nurses’ action lasted 20 years. When do actions begin and end? How many other actions were included in, or overlapped with, the nurses’ action of killing their patient? Consider a different example. Someone turns the light on because she wants to read a book. How many actions has she performed? Has she (p. 135) spent money on electricity? Has she bothered her partner? Has she overloaded the circuit-breakers? Are all these different actions? Joel Feinberg famously referred to a “well-known feature of our language, whereby a man’s action can be described almost as narrowly or as broadly as we please” as “the accordion effect.”4 Are these different “descriptions” different actions? If so, then it would follow that we are constantly doing an infinite number of things—not an attractive thesis. If not, we would need an account of what exactly our basic actions (which could then be infinitely re-described) are—as I will show later, such account is rather elusive.
These sorts of uncertainties have led some to challenge the convenience—or the intelligibility—of considering action as a sine qua non condition of criminal liability.5 Admittedly, rigorously defining action is a perennial philosophical problem—and, like most philosophical problems, one which, in a sense, is not supposed ever to be resolved. But this should not cause skepticism about action in the criminal law. First, because the fact that there are complicated cases does not entail that there are not also perfectly unproblematic cases: some cases involving omissions and attempts are indeed very murky, but many others are not. Secondly, because paradigmatic cases of action, such as when someone rapes someone else, involve neither omissions nor attempts.
Even if we restricted our study to the most complicated cases, however, their complexity may not be enough reason to abandon the deep-seated intuition that to be “punished” (or, for that matter, rewarded) for something one has not done (in the broad sense just sketched) is absurd. For “punishing” people for really doing nothing is both conceptually hard to grasp, and normatively hard to accept. Consider the conceptual point first. Imagine a mother telling her daughter, “I am going to punish you”; the daughter asks “Why, what have I done?,” to which the mother replies “You have done nothing, but I am punishing you anyhow.” That this mother simply does not know what the word “punishment” means appears as the inevitable conclusion. For if she really believes that her daughter has done nothing, and yet she slaps her, she would be abusing, assaulting, or otherwise victimizing her daughter, not punishing her.
As for normative implications, consider some trends in contemporary criminal law whereby phenomena the connection of which to human action is increasingly tenuous—say, so-called possession crimes—are criminalized.6 These trends are worrying because they render people liable to punishment merely for being in possession of things (say, illegal drugs) found on, for example, their property—even if without their knowledge and despite their reasonable vigilance. In so doing, these trends further expand the already enormous punitive power of the state. So it seems prudent to continue (p. 136) endorsing the view that, at a very minimum, criminal punishment should require an action. Be that as it may, this does not by itself shed any light on the nature of action.
At least within the Anglo-American legal tradition, one specific philosophical theory of action has been particularly influential—it is, for example, expressly endorsed by the U.S. Model Penal Code7—in dictating what the criminal law should understand by “action.” This is the volitional theory of action, according to which actions are “willed bodily movements” (“acts of will” and “volitions” are synonyms). While volitionalism can be traced back for centuries, its most systematic and influential formulation is found in Austin (1869),8 and its most ambitious contemporary defense is found in Moore (1993).9 Volitionalism answers with surprising ease some of the vexing questions posed earlier. For example, when someone wishes to read a book, and thus switches the light on, bothers her partner, spends money, etc., for volitionalists there is no indeterminacy: these are not different actions. The only action is the person moving her body in a certain way (a movement that could then be re-described in myriad ways).
The price paid for such a neat answer to the problem of action individuation is high. Most obviously, the answer generates clear problems for the volitionalist regarding some omissions (in which bodies do not move). Either volitionalists would have (implausibly) to deny that these omissions are actions, or would have to continue to offer odd ways of individuating actions (say, a nurse’s action to stand still so as to signal to her partner that it is time to kill their patient would have started a couple of decades ago when they first began planning their crime). Neither alternative seems attractive.
Moreover, the phenomenological account of human action that the volitional theory of action presupposes is rather strange. Think again of someone who wants to sit on the sofa and read a book; according to volitionalists this is what happens: she has a garden-variety pro-attitude (a desire or an intention) to sit on the sofa and read a book, but, in addition, she also has a much mysterious “volition” to move her muscles in exactly the way necessary to read this book. Except for very few cases in which we focus on moving specific muscles in specific ways (say, when rehabilitating an injury), we hardly ever find ourselves willing to move muscles. Volitions, then, are either something very strange, so strange that we have never really experienced them, or else they happen to be just regular intentions. The former option is evidently unattractive. But so is the latter. For it is hard to understand why the intentional object of such allegedly garden-variety intentions has to be restricted to muscle movements.10
Moreover, if volitions are simply intentions, and all actions are preceded by volitions, then all actions are, in a sense, intended. In fact, volitionalists tend to believe that “involuntary action” is a misnomer—for if the putative action truly is involuntary, then it is a mere movement, and not an action. For them, all actions are voluntary, because all actions follow volitions. In this sense, the expression “voluntary (p. 137) action” is redundant: the famous “voluntary act requirement” should be recognized as simply the “act requirement”—and one can indeed agree with this specific point without thereby endorsing volitionalism.
Linking action to the presence of this or that mental state, as volitionalists do, is problematic in that the analysis of mental states has, for good reasons, traditionally belonged to a different area of the criminal law: culpability. For example, and as we shall see in the next section, the distinction between intended and unintended action is the cornerstone of our efforts to apportion blame. But if, as some volitionalists believe, it turns out that all actions are in a sense intended (because they are all preceded by volitions, and volitions are just intentions), this would create difficulties for determining the blameworthiness of these actions.
Before concluding this section, it is important to mention that in addition to requiring an action (whatever that may turn out to be) as a condition of liability, the criminal law also requires that such action must be antecedently, and statutorily, described as a crime.11 This is the actus reus requirement. Although it is plain that the actus reus and the act requirement are different, they tend to be confused, so a word about their difference is in order.
It is impossible to satisfy the actus reus requirement without thereby also satisfying an act requirement, although it is perfectly possible to satisfy an act requirement without satisfying any actus reus requirement. This asymmetry highlights a normative difference between these two requirements. The wrongness of “punishing” someone who has not acted is absolutely fundamental, as we saw previously. But punishing people for something that they were not antecedently told was punishable, in contrast, is, while normally wrong, not fundamentally wrong. The latter is a political value, captured in the famous principle of legality: nulla poena sine lege. While the central importance of the principle of legality is undeniable, one could imagine thought experiments in which a person may be justly punished for an act which had not been statutorily described as a crime beforehand. No thought-experiment could, however, yield an intelligible, let alone a just, act of “punishing” someone who did not act.
III. Philosophy of Mind
In order for someone to be liable to criminal punishment, in addition to satisfying both the act and the actus reus requirements, she needs to satisfy the mens rea requirement: her mind must be “guilty.” It is useful to break down the discussion of (p. 138) the mental requirement for liability into two general topics, one primarily related to mental capacities, and the other primarily related to mental states. If a person commits a crime while she was not capable of distinguishing right from wrong, or while she was unaware of what she was doing, then she is not liable to be punished. Typical examples of people lacking basic mental capacities include children and severely handicapped people. A 5-year-old child, for example, cannot really be blamed for firing a gun and killing someone—for she simply did not know what she was doing. Similarly, if unbeknownst to a person, someone puts a drug in her coffee which severely impairs her perception of reality, she is not to blame for firing a gun and killing someone. In cases like these, the agents are not accountable, or at least not as fully accountable.12
Admittedly, there are problems with thresholds: at what age exactly does a human being acquire enough mental capacities so that she becomes a full-blown moral agent? Imagine that we settle on, say, 18 years of age. Are there not 17-year-olds who are much more mature than some 21-year-olds? Similarly, how much impairment of your senses is enough to claim that you were not competent? The law—and not just the criminal law—cannot escape running into these line-drawing problems; lines have to be drawn somewhere: for voting rights, for buying cigarettes, and also for determining when someone is a competent moral agent. But the rationale for this requirement should be perfectly clear independently of these complications. It is unjust to punish a 5-year-old child, or someone who was unaware that she had been intoxicated, for their mental capacities are either not fully developed or not fully functional. (Getting intoxicated before committing a crime in order to then defend oneself by appealing to the intoxication will not work: this would be an instance of an actio libera in causa: roughly, you were accountable when you decided to render yourself “unaccountable.”)
Imagine then two full-blown moral agents, whose mental capacities are fully developed and functional, each committing identical homicides, independently from each other, but, ex hypothesi, in exactly the same circumstances, except for their own mental states. It may be unjust to punish them identically—if their mental states are different. For example, imagine that one of them intended to kill her victim, whereas the other one killed her victim negligently. Intuitively, the first killer is more blameworthy than the second. In more general terms, the intuition is that intended wrongdoing is more blameworthy than unintended wrongdoing. The force of this intuition makes it as good a candidate to a universal moral principle as any, and it has in fact undergirded virtually every penal code in history.13
The area of the criminal law which studies the mental states with which people commit crimes is known as culpability or as mens rea. In some cases we can use (p. 139) expressions such as “Someone is culpable” in order to indicate that she is guilty (as opposed to innocent); but in a more technical sense, when we say that someone is culpable what we mean is that she is to blame. One important characteristic of culpability is immediately obvious: culpability is the scalar property par excellence in the criminal law. The extent to which people could be accountable may sometimes be a matter of degree, although accountability is often an all-or-nothing condition, if for no other reason than as a result of the line-drawing problem just described. The law will treat a 1-year-old as if he is just as unaccountable as a 13-year-old, and an immature 19-year old as accountable as a very wise 40-year-old. Similarly, whether an event is an action or whether a given action has been statutorily defined as a crime are not matters of degree.
Culpability, in contrast, is scalar to its core. While both agents who commit homicide we are discussing are culpable—in the non-comparative sense that they are not innocent, and are indeed to blame—the interesting point is that one is more culpable than the other. Different legal systems, and different legal traditions, focus on different mental states in generating scales with different degrees of blameworthiness. But there is a remarkable convergence amongst theories of culpability regarding a pair of distinctions between mental states with which agents can commit crimes. One of these distinctions we have already introduced: that between committing crimes intendedly or unintendedly. The other distinction is that between committing a crime aware that one is committing it or unaware that one is committing it.
The interplay between these two distinctions gives rise to a fourfold map of modes of culpability, that is, of the different culpable ways in which crimes can be committed which, ceteris paribus, give rise to different degrees of blame (and, eventually, to more or less severe punishments). In decreasing order of blameworthiness: (a) someone may commit a crime both intending to commit it and knowing that she is committing it; (b) knowing that she is committing it, although not intending to commit it; (c) neither intending nor knowing that she is committing it, but more or less aware that there is a (substantial) risk she may be committing it; and, finally; (d) someone may commit a crime not intending to commit it, not aware that she is committing it, and not even aware that there is any risk that she may be committing it.14
While the decreasing blameworthiness in these four modes of culpability may be essentially unproblematic, the uninitiated may wonder how a person who neither intends to commit the crime nor even knows that she is (risking) committing it could possibly be culpable. The uninitiated would in fact be in the company of many criminal law scholars who suggest that an agent who commits a crime in such (p. 140) circumstances is blameless (and thus should not be punished either). The predominant position, however, continues to be that the agent who conforms to that fourth scenario is culpable (albeit less than in any of the other scenarios). Explaining why illuminates the contours of the lower limit of culpability.
Someone who neither intends to commit a crime nor knows that she is (risking) committing it may still be blameworthy if she should have known it. This would involve cases in which she failed to exercise a duty of care. It is because of her failure to know that we blame her. In cases in which we do not think that she should have known, then, indeed, there is no blame; such cases can be referred to as accidents. Contrast two agents; the first one forgets that she carries a loaded gun in her purse and when she reaches into it in order to grab her phone, she happens to pull the trigger, injuring a passerby. The second agent is writing on the board as she teaches a class when the chalk in her hand breaks and a piece flies off, injuring one of her students. (Arguendo, assume that the injuries in both cases are identical.) The second agent causing injuries to students in this way is truly a bona fide accident because, unlike the first negligent agent, we could not reasonably have expected her to behave any differently. The difference between accidents (which are by definition blameless) and those cases involving the lowest degree of blame is that in the former we do not expect the person to have known better, whereas in the latter we do.
The upper limit of culpability is reached when the person intends to commit the crime (and, a fortiori, knows that she is committing it). Of course, it is possible to commit a crime in even more blameworthy ways: say, when the criminal sadistically enjoys the harm that she is inflicting, or when she chooses a particularly cruel way of committing her crime. While these factors may indeed aggravate her overall blameworthiness, from the perspective of culpability more or less narrowly construed, committing a crime intendedly is sufficient to put a criminal in the most blameworthy category.
The distinction between intended and unintended wrongdoing is so central to our intuitive ways of thinking about blame that it has often been assumed, erroneously, that unintended wrongdoing is not merely less blameworthy than intended wrongdoing, but utterly blameless. Socrates seems to make this mistake in his defense:
either I have not a bad influence [on Athenian youth], or it is unintentional, so that in either case your accusation is false. And if I unintentionally have a bad influence, the correct procedure in cases of such involuntary misdemeanors is not to summon the culprit before this (p. 141) court but to take him aside for instruction and reproof, because obviously if my eyes are opened, I shall stop doing what I do not intend to do.15
Much more recently, another famous trial gave rise to a similar mistake:
foremost amongst the larger issues at stake in the Eichmann trial was the assumption current in all modern legal systems that intent to do wrong is necessary for the commission of a crime. On nothing, perhaps, has civilized jurisprudence prided itself more than on this taking into account of the subjective factor. Where this intent is absent…we feel that no crime has been committed.16
In addition to erroneously mapping the distinction between intended and unintended action onto the distinction between crime and no crime, Arendt errs in thinking that the importance of intent is a creature of modernity. Its importance can be traced back for millennia, and it was systematically analyzed already in ancient philosophy (as the passage from Plato’s Apology just quoted illustrates)17—an analysis that was further refined in Rome,18 and during the Middle Ages.19 These analyses of intentions and other mental states informed the work of foundational figures of modern common law (e.g. Austin;20 Bentham21), and have been the concern of jurists throughout time. Determining what exactly an intention is, how it differs from related conative states (like desires, wishes, and hopes),22 and why intentions are so central for our efforts to apportion blame23 have been, and continue to be, philosophically active research areas. Nowhere can a criminal lawyer wishing better to understand criminal culpability turn to more profitably than toward contemporary philosophy of action.
Independently of the particularities of its history, criminal law systems around the world require accountability and a guilty mind in order for punishment to be inflicted. Added to the act and actus reus requirements these are the essential factors that may, in the standard case, render people liable to punishment. Admittedly, however, some cases are not standard. First, because there exist “strict liability” offenses, in which people are punished even if they do not have a guilty mind. Secondly, because sometimes people who meet all requirements sketched previously can still avail themselves of defenses which will spare them from punishment, such as “self-defense, duress, autrefois acquit, and [the variegated forms of] immunity from prosecution.”24 (p. 142)
Amongst the many different definitions of punishment, Hart’s is by far the most influential. This is Hart’s version of what he calls the “standard” case of punishment:
(i) It must involve pain or other consequences normally considered unpleasant.
(ii) It must be for an offence against rules.
(iii) It must be for an actual or supposed offender for his offence.
(iv) It must be intentionally administered by human beings other than the offender.
(v) It must be imposed and administered by an authority constituted by a legal system against which the offense is committed.25
Hart also recognizes that there are “many other possibilities” in which punishment can manifest itself, which he calls “sub-standard”:
(a) Punishments for breaches of legal rules imposed or administered otherwise than by officials (decentralised sanctions)
(b) Punishments for breaches of non-legal rules or orders (punishments in a family or school)
(c) Vicarious or collective punishment of some member or social group for actions done by others without the former’s authorization, encouragement, control or permission.
(d) Punishment of persons (otherwise than under (c)) who neither are in fact nor supposed to be offenders.26
While one would expect the criminal law to be concerned with “standard” punishment, the scope and reach of criminal justice systems is such that forms of “sub-standard punishments” are becoming increasingly common. The boundaries of criminal punishment now overlap with the boundaries of administrative and, above all in affluent countries, immigration law, where all sorts of evidently punitive measures against illegal immigrants are stipulatively categorized as non-criminal.27 Moreover, “sub-standard” forms of punishment are also widespread within the context of the “war on terrorism”—where people are punished secretly and without trial. (Of course, one could challenge the view that what Hart calls sub-standard punishment is really punishment; for example, it is difficult to (p. 143) understand how hard treatment inflicted on persons who are not even “supposed to be offenders” could possibly qualify as punishment.28) The conceptual discussion, however, does nothing to alleviate the plight of those upon whom these treatments are inflicted.
The specific aspect of Hart’s definition of punishment whereby it is supposed to involve “pain or other consequences normally considered unpleasant,” meets with virtually universal agreement. Insofar as punishment is, then, by definition, something in principle unwanted, it stands in need of moral justification. Typically, to do things to people against their will is morally impermissible—much more so when those things are meant to be painful or unpleasant. The first step in the moral justification of punishment is to suggest that by committing crimes criminals render themselves legitimate targets of this hard treatment. It seems undeniable that, because of their crimes, criminals are different from innocent people, and that inflicting hard treatment on the former is thereby at least morally different from inflicting it on the latter.
But even if we recognize a difference between criminals and non-criminals, this does not entail that punishment as such is thereby justified. There are, after all, many examples of (non-criminal) laws the violation of which does not entail, or justify, punishment. For example, not paying your debts entails that your creditor can seize your property (to cover the amount that is owed), even by force. Being stripped by force of something you own can be painful and unpleasant, but what the law seeks is for creditors to be able to recuperate what is owed to them, not to make debtors suffer. Punishment, in contrast, seeks to make people suffer.
This is the first watershed moment in the philosophical justification of punishment. Those who deny that the difference between criminals and non-criminals is sufficient to justify punishment (or, those who, more daringly, deny that the very distinction has much, if any, significance) are known as abolitionists. They see punishment as morally illegitimate. In spite of all that can be said against punishment and against contemporary criminal justice systems, however, the vast majority of scholars are not abolitionists. Most philosophers believe that punishment can indeed be justified, and they differ as to the particulars of such justification.
An internal division amongst all those who believe that punishment is justified constitutes the second watershed moment in the philosophical justification of punishment. The two main justifications of punishment are consequentialism and retributivism, and they have been historically associated with teleological moral doctrines and deontological moral doctrines, respectively. Admittedly, the connection between retributivism and deontology, on the one hand, and that between consequentialism and teleology, on the other, ought to be qualified. The fact that someone endorses consequentialism as a justification of punishment does not entail (p. 144) that she has to endorse a teleological comprehensive moral doctrine; and the fact that someone endorses a retributive justification of punishment does not entail that she has to endorse a deontological comprehensive moral doctrine.29 But for the purposes of sketching the main characteristics of the most important justifications of punishment, it is useful to keep in mind that retributivist sensibilities are more at home in a deontological worldview, and that consequentialist sensibilities are more at home in a teleological worldview.
Consequentialist justifications of punishment are the more variegated, and, in a way, perhaps also the most widespread. These justifications focus on the painful, unpleasant, or otherwise unwanted aspects of punishment, and appeal to these characteristics in positing that punishment is itself an evil.30 The only way in which the state could be justified in inflicting this evil on criminals is if by so doing the state is likely to bring about good consequences which may outweigh punishment’s alleged intrinsic badness. For example, if by punishing a criminal we are likely to rehabilitate her, then this good result may outweigh, in the consequentialist’s eyes, the evil of punishment. Similarly for some of the other famous consequentialist concerns: deterring the criminal on trial—or the general population—from committing crimes of this sort again, incapacitating the criminal from harming others, prevention, and so on, may be seen as goods that outweigh the intrinsic badness of punishment.
Retributivism, in contrast, does not look at any further consequence punishment may bring about in order to justify it: punishment, for retributivists, is morally justified by its being deserved. If someone deserves to be punished, then, in principle, inflicting such punishment is justified. Clearly, retributivism need not be opposed to any of the aims that typically mobilize consequentialists (again: rehabilitation, incapacitation, or deterrence). But they deny that these other aims, laudable as they may be, belong to the discussion of the justification of punishment—they are “a happy surplus.”31 Retributivists link the justification of punishment to its inherent justice: people getting what they deserve is just, and it is a good thing in itself. It is just to award the highest grade to a student who deserves it, or a medal to a hero who deserves it, and similarly, it is just to award punishment to whomever deserves it.
Some strengths and weaknesses in each of these two justifications of punishment immediately become apparent. The main strengths of consequentialist approaches are the result of the palpable benefits which punishment would produce: it is hard not to hope for a safer world, or one with fewer criminals. Apart from the fact that the consequences consequentialists envision as resulting from punishment are not certain to obtain, the main weaknesses of consequentialism stem from the risk that focusing too much on those desired consequences may (p. 145) lead to treating the criminal herself unjustly. The main strengths of retributivism are related to the respect of the defendant’s dignity as a human being: she will not be used merely as a means in order to advance this or that goal (however laudable the goal may be): she would be treated justly. But in her self-imposed indifference to the consequences of punishment, the retributivist may in fact become obsessed with one value (giving people what they deserve), at the expense of other things which are valuable too.
These strengths and weaknesses, again, track some of the classical strengths and weaknesses of the comprehensive moral doctrines to which these justifications of punishment have historically been associated. Teleological moral doctrines define the right thing to do as a matter of maximizing the good: the paradigmatic example is utilitarianism, which defines the right thing to do as that which maximizes utility (the good). Deontological moral doctrines do not define the right in terms of maximizing the good at all: the right and the good are, rather, independent of each other. This independence invites objections, for if actions are not assessed by attending to their consequences, then how? Teleological doctrines also have their problems. A classical objection raised against them concerns punishment directly: utilitarianism may justify scapegoating an innocent person, if so doing were the only way to avoid great disutility.
A way of summarizing the strength and weaknesses of consequentialism and retributivism is to focus on the famous slogan used to distinguish them: consequentialism is forward-looking, whereas retributivism is backward-looking. It is, of course, good to look forward, but the past also has some claims on us. Think of the famous “never forget” injunction on the wake of horrible atrocities: the injunction is in tension with consequentialism, insofar as the latter may sometimes recommend forgetting past atrocities. Similarly, while remembering the past may be a dignified display of integrity, a concern with the past may have a tendency to become obsessive and thus to cloud our vision of the future.32
The obvious pros and cons of these justifications have given rise to a very active research agenda seeking to develop “mixed” justifications of punishment, which seek to capture the best of each family of justifications while avoiding their pitfalls. An example is the burgeoning field of “transitional justice,” which seeks to combine the need to do justice to massive human rights violations (retributivism) and the need to contribute to the viability of incipient democracies (consequentialism). Independently of transitional justice, the effort has been particularly active amongst philosophers during the last 50 or 60 years, inspired by the realization that “the view that there is just one supreme value or objective (e.g. Deterrence, Retribution or Reform) in terms of which all questions about the justification of punishment are to be answered, is somehow wrong.”33 Unfortunately, however, and in spite of (p. 146) their undeniable superficial attractiveness, efforts at mixing retributivism and consequentialism have failed. They have failed in at least two senses. First, they have failed in the sense that the concessions made to one camp remain unacceptable to the other camp. Secondly, they fail in the sense of not convincing the neutral, non-partisan observer that the best aspects of retributivism and of consequentialism have indeed been combined in a way that also succeeds in avoiding the pitfalls of each.
Consider one famous example of an unsuccessful mixed justification. John Rawls suggested that the debate between retributivism and consequentialism is in fact resolved “by the time-honored device of making them apply to different situations.”34 What he proposes is by now a common gambit: to ask two different questions, one of which to be answered along retributive lines, and another one along consequentialist lines. Rawls thus invites us to consider two different questions a young son, upon learning of someone’s punishment, might ask his father: (a) “Why was F put in Jail yesterday?” and (b) “Why do people put other people in jail?” And Rawls then suggests that the father would answer the first question along the lines of: “Because he [F] robbed the bank at B. He was duly tried and found guilty”; and the second question along the following lines: “To protect good people from bad people.”35 Moreover, Rawls suggests that each of these questions (and their respective answers) constitutes a point of view corresponding to a different branch of government: in one case the judiciary, and in the other the legislative. And then he claims that “one can say, then, that the judge and the legislator stand in different positions and look in different directions: one to the past, the other to the future,” and that “the justification of what the judge does, qua judge, sounds like the retributive view; the justification of what the (ideal) legislator does, qua legislator, sounds like the utilitarian view.”36
The main problem with mixed justifications such as Rawls’s, is that they “mix” much less than advertised. Granted, in Rawls’s solution government-as-a-whole can be seen as a system which includes both retributivist and consequentialist rationales. But each of the two branches remains remarkably unmixed: if you are a judge, you had better be a retributivist; if you are a legislator, you had better be a consequentialist. Not a very satisfying result.
Despite these efforts at combining retributive and consequentialist rationales, the debate over the philosophical justification of punishment remains as alive today as it has ever been.37 Just as the criminal law needs to continue to rely on “action” even though its account remains elusive, it also needs to continue to punish, even punishment’s justification remains a contested topic. (p. 147)
V. Political Philosophy
Independently of why in the end the state might be morally justified (if indeed it is) in punishing criminals, there is the further question as to what exactly should it be politically legitimate for the state to describe as a crime so as to subsequently punish. Quite obviously, there are many moral wrongs which ought not to be political wrongs—that is, wrongs which should not be made criminal. For example, betraying loved ones, whom we have led to expect our help and support, is, morally speaking, a bad thing to do. But it is very difficult to find a persuasive argument why lack of loyalty to loved ones (including here not only romantic partners, but relatives and friends) should be made criminal. In order to decide which acts ought to be criminalized we need to move beyond moral philosophy, and into political philosophy (or political morality).
The most appealing contemporary principles of criminalization are inspired by the basic ideas found in John Stuart Mill’s seminal On Liberty (1977). As is well known, Mill’s major preoccupation in that book is to protect the individual from potential abuses by the state. It is perfectly obvious that the relationship between the state and the individual is asymmetrical, in the sense that, compared to the individual’s, the state’s power is immense. Surprisingly, perhaps, this asymmetry holds more or less equally—that is, it is sufficiently large—independently of whether the state under consideration is an enlightened democracy or a reactionary or otherwise ruthless tyranny. Mill, in fact, was above all focused on the ways in which the rights of individuals could be violated in enlightened democracies. His worry concerned what his work transformed into a household expression: the “tyranny of the majority,” by which he meant to capture the ways that the legitimate interests and aspirations of individuals, even in democracies, can be illegitimately drowned by those of the masses (whom the state represents).
This problem, of course, does not concern only the criminal law: states can set back individuals’ legitimate interests and aspirations by favoring, say, some artistic or cultural manifestations and not others; or by (even tacitly) endorsing some lifestyles and not others.38 But nowhere can the power of government affect individuals more gravely than within the context of the criminal law. After all, in some jurisdictions individuals can, by way of criminal punishments, be subjected to physical pain, or even killed. Even in the least punitive and most humane jurisdictions, individuals are, by way of criminal punishment, stripped of their liberty, sometimes for the entirety of their lives. Criminal punishment causes more pain and suffering, and it generates more misery amongst individuals (and not only amongst the deserving criminals, but amongst their relatives and friends), than all other areas of government activity (p. 148) combined. Moreover, there exists a growing consensus amongst philosophers that criminal justice systems of even enlightened democracies routinely overcriminalize and over-punish— and that this overcriminalization affects above all the very poor and members of traditionally discriminated ethnic minorities.39
Part of an explanation—albeit clearly not a justification—for over-punishment may relate to the fact that even in enlightened democracies the widespread understanding of what punishment is supposed to be, and of what it is supposed to accomplish, is not itself very enlightened. In the United States, for example, slavery is, technically speaking, a permissible punishment: the U.S. Constitution clearly states: “neither slavery nor involuntary servitude, except as a punishment for crime whereof the party shall have been duly convicted, shall exist within the United States, or any place subject to their jurisdiction.”40 Slavery (even if conceptualized as a punishment), one would have hoped, would have by now been completely abolished in enlightened democracies. As we saw earlier, no one can seriously deny that punishment is supposed to be painful—but there obviously are many ways in which hard treatment can be painful without it being slavery.
Early in On Liberty, Mill claims that his essential goal is:
to assert one very simple principle, as entitled to govern absolutely the dealings of society with the individual in the way of compulsion and control, whether the means used be physical force in the form of legal penalties, or the moral coercion of public opinion. That principle is that the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their number, is self-protection. That the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant.41
It is a trite joke to point out how Mill then spends the rest of the book struggling with the many and complex qualifications needed to both make sense and justify that allegedly “very simple principle.” And there exists a very large scholarship continuing Mill’s struggles.42
For our purposes here, however, somewhat of a corollary to the spirit of what is now widely known as Mill’s “harm principle” can be stated simply. When it comes to the scope of the criminal law, that is, to the reach of the state’s punitive power, we ought to be cautious, and err on the side of prudence. One could propose an analog to the famous principle related to doubts about defendants’ guilt, in dubio pro reo, and suggest a similarly broad and simple principle of criminalization: in dubio pro libertas. This (p. 149) highlights the essential point in Mill’s position, universally known as liberalism: an important liberal insight is that the state should remain neutral as to theories of the good. The state should not endorse what it takes to be the good life—precisely because this may restrict the liberty of those who do not agree with such a conception of a good life. This point concerns liberalism in general, but its connection to criminal law in particular is obviously important. When the liberal defends the right of a person to harm herself by, say, by using drugs, she is not thereby endorsing the view that harming oneself with drugs is a good thing—she is merely erring on the side of liberty: in dubio pro libertas. The liberal’s caution leads her to adopt a minimalist criminalization principle, and something like Mill’s harm principle is as good a starting point as any.
Predicating punishment only on actions which harm (or risk harming) others prevents legislation of two types. First, it prevents legislation that would criminalize actions which only (risk) harm (to) the very agent. In this sense, liberalism is opposed to what is known as paternalism: namely, legislation that seeks to protect the individual from herself. A well-known difficulty, however, is that it is relatively easy to re-describe the intent of a piece of legislation that appears paternalistic as if it were non-paternalistic. So, while punishing competent adults for not wearing seat belts when operating motor vehicles may look like a clear case of paternalistic criminalization, it could be argued that the legislator’s intent is to protect others from the “harm,” say, of having to pay for picking up non-seat-belt-wearing bodies from the public roads in which they may land after accidents. To the extent that these sorts of re-descriptions appear contrived, the liberal will tend to reject them. A very important case study—and a timely one as well—concerns (recreational) drug use by adults.43 To the extent that such behavior does not (risk) harm to others, the liberal is likely to reject its criminalization—although opponents may, as in the example of seat belts, try to suggest that there is a rationale for criminalizing drug use which is not paternalistic.
The second type of criminal legislation prevented by the liberal stance is that which predominantly seeks to uphold the values of a society. The opponent of the liberal here is the moralist (or legal moralist, as she is also known). The moralist thinks that exercising the state’s punitive power in order to teach moral lessons, or to add to the cohesion of a given community is a legitimate enterprise. Moreover, moralists may see the teaching of moral lessons as the duty of responsible states. Perhaps it is impossible for the criminal law of any moderately complex society not to sometimes say something about its values; but the liberal opposes making this a primordial function. Probably the most famous and illustrative contemporary debate between liberalism and moralism is the Hart–Devlin debate,44 into which Ronald Dworkin, amongst many others, also entered.45 At issue was whether homosexuality (p. 150) (between males) could be criminalized, simply by virtue of its alleged immorality. Devlin, representing the moralist side, argued that homosexuality (amongst men) could be legitimately punished simply by virtue of its being considered immoral by the majority of people, whereas Hart (and Dworkin), taking the liberal side, rejected this position. Liberals reject moral majoritarianism.
There are contexts in which the alternatives to liberalism appear more plausible: for example, within the context that opposes liberalism to communitarianism, or within that of economic policy. There may be cases, moreover, in which a full rejection of paternalism seems naïve—individuals may need state protection from the also asymmetrical power relation between them and large corporations. Think, for example, of nutritional labels, or of health warnings on cigarettes: without these, individuals may be, to an extent, at the mercy of powerful advertising campaigns. Be that as it may, within the context of the criminal law, and in particular given the many excesses, power asymmetries, and injustices which the criminal law engenders,46 and in spite of the moralist resurgence of late (admittedly much more nuanced than Devlin’s),47 it is difficult not to endorse (something like) liberalism.
The liberal not only seeks to limit the number of moral wrongs which ought to be recognized as legal wrongs. She will also seek to limit the number of legal wrongs which ought to be recognized as specifically criminal wrongs. Insofar as prison and death constitute greater infringements of individual freedom than fines, confiscations, and forced fulfillment of contractual obligations, the liberal will hesitate to favor the former. She will tend to treat punishment as the measure of last resort—and this is not to thereby take sides regarding the debate between retributivism and consequentialism. A liberal may recognize that the fact that criminals deserve punishment is an important consideration. But, ceteris paribus, and as a political stance independent of the debate between retributivism and consequentialism, the liberal would like to criminalize less rather than more. And she would also like to punish less rather than more, for much of what holds for the problem of overcriminalization holds, too, for the problem of over-punishment.48
An age-old popular position is that punishment must fit the crime: those committing particularly horrible crimes ought thereby to suffer particularly horrible punishments. The liberal may resist such simplistic conclusion, and so may the retributivist.49 The main reason why some think that the retributivist cannot accompany the liberal in being suspicious of punishment is that, historically, retributivism has been erroneously equated with the famous lex talionis, the famous “eye for an eye, tooth for a tooth.” This is partly explained by the enormous influence that Kant, who held both positions,50 has exerted. There is, however, no reason (p. 151) why a retributivist must endorse lex talionis. All that her retributivist credentials require her to hold is that it is good when people (including criminals) get what they deserve, and that this goodness is relevant for the justification of punishment. Even a retributivist who also endorsed lex talionis may share the liberal suspicion of punishment. For she may understand that the fact that something is good does not entail that anyone has an all-thing-considered obligation to bring it about. She may also understand that moral goodness and political goodness do not always coincide.
American Law Institute, The Model Penal Code and Commentaries: Official Draft and Revised Comments (1985)Find this resource:
Austin, John, Lectures on Jurisprudence (1869)Find this resource:
Bentham, Jeremy, “The Principles of Morals and Legislation,” in John Bowring (ed.), The Collected Works of Jeremy Bentham, Vol. 1, (1962)Find this resource:
Berlin, Isaiah, Four Essays on Liberty (1958)Find this resource:
Feinberg, Joel, Doing and Deserving (1970)Find this resource:
Feinberg, Joel, The Moral Limits of the Criminal Law (1984, 1985, 1986, 1990), 4 vols.Find this resource:
Hart, H. L. A., Punishment and Responsibility (1st ed., 1968; 2nd ed., 2008)Find this resource:
Kant, Immanuel, The Metaphysical Elements of Justice (transl. John Ladd, 1965)Find this resource:
Mill, John Stuart, “On Liberty,” in J. M. Robson and Alexander Brady (eds.), Collected Works of John Stuart Mill, Vol. XVIII (1977)Find this resource:
Moore, Michael S., Act and Crime: The Philosophy of Action and its Implications for Criminal Law (1993)Find this resource:
Moore, Michael S., Placing Blame: A General Theory of the Criminal Law (1997)Find this resource:
(1) H. L. A. Hart, Punishment and Responsibility (2nd ed., 2008). See also H. L. A. Hart, “The Ascription of Responsibility and Rights,” in Gilbert Ryle and Anthony Flew (eds.), Logic and Language (1951), 171–194, especially 187 ff.
(2) See R. A. Duff, Criminal Attempts (1996), 119–127; Larry Alexander “The Philosophy of Criminal Law,” in Jules Coleman and Scott Shapiro (eds.), The Oxford Handbook of Jurisprudence and Philosophy of Law (2002), 832–840; and Larry Alexander, Kimberly Kessler Ferzan, and Stephen Morse, Crime and Culpability: A Theory of Criminal Law (2009), 197–216.
(3) A very useful collection of influential articles on moral luck is found in Daniel Statman (ed.), Moral Luck (1993).
(4) Joel Feinberg, Doing and Deserving (1970), 137.
(5) Douglas Husak, “The Alleged Act Requirement in Criminal Law,” in John Deigh and David Dolinko (eds.), The Oxford Handbook of Philosophy of Criminal Law (2011), 107–124.
(6) Markus D. Dubber, Victims in the War on Crime: The Use and Abuse of Victims’ Rights (2002).
(8) John Austin, Lectures on Jurisprudence (1869).
(9) Michael S. Moore, Act and Crime: The Philosophy of Action and its Implications for Criminal Law (1993).
(12) Aristotle, Nicomachean Ethics (transl. H. Rackham, 1926), 1110b1 ff.
(13) Leo Zaibert, Five Ways Patricia can Kill her Husband: A Theory of Intentionality and Blame (2005), 33–108.
(14) When it comes to culpability, the two main legal traditions in the West, the common law (Anglo-American) tradition and the European (German) tradition have much more in common than is often assumed. The fourfold map of culpability just sketched is no closer to, say, the fourfold Anglophone map of modes of culpability adopted by the Model Penal Code (purpose, knowledge, recklessness, and negligence) than it is to a generalized fourfold map of modes of culpability representative of the German tradition (dolus, dolus eventualis, conscious culpa (bewusste Fahrläsigkeit), and unconscious culpa). For a helpful comparison of these two traditions, see Markus D. Dubber and Tatjana Hörnle, Criminal Law: A Comparative Approach (2014) and George P. Fletcher, Rethinking Criminal Law (2000); for a comparison between these traditions and different philosophical approaches throughout history, see Zaibert (n. 13).
(15) Plato, “The Apology,” in Edith Hamilton and Huntington Cairns (eds.), The Collected Dialogues (1961), 26a.
(16) Hannah Arendt, Eichmann in Jerusalem: A Report on the Banality of Evil (1977), 275.
(18) David Daube, Roman Law: Linguistic, Social and Philosophical Aspects (1969).
(19) Thomas Aquinas, Summa Theologiae (1974).
(20) John Austin, Lectures on Jurisprudence (1869).
(21) Jeremy Bentham, “The Principles of Morals and Legislation,” in John Bowring (ed.), The Collected Works of Jeremy Bentham, Vol. 1 (1962).
(22) See G. E. M. Anscombe, Intention (1963), and John R. Searle, Intentionality: An Essay in the Philosophy of Mind (1983).
(24) John Gardner, Offenses and Defences: Selected Essays in the Philosophy of Criminal Law (2007), 141.
(25) H. L. A. Hart, Punishment and Responsibility (1968), 4–5. This is also known as the Flew–Benn–Hart definition, since the three authors more or less simultaneously (but independently) developed very similar accounts. See S. I. Benn, “An Approach to the Problem of Punishment,” (1958) 33 Philosophy 325–341, and A. G. N. Flew, “The Justification of Punishment,” (1954) 29 Philosophy 291–307.
(27) Leo Zaibert, “Uprootedness as (Cruel and Unusual) Punishment,” (2008) 11 New Crim. LR 384–408.
(28) Leo Zaibert, Punishment and Retribution (2006), 26 ff.
(34) John Rawls, “Two Concepts of Rules,” in Samuel Freeman (ed.), John Rawls: Collected Papers (2001), 22.
(37) See Zaibert (nn. 13, 28).
(38) Ronald Dworkin, A Matter of Principle (1985).
(39) See Douglas Husak, Overcriminalization: The Limits of the Criminal Law (2008) and Hyman Gross, Crime and Punishment: A Concise Moral Critique (2012).
(40) Thirteenth Amendment.
(41) John Stuart Mill, “On Liberty,” in J. M. Robson and Alexander Brady (eds.), Collected Works of John Stuart Mill, Vol. XVIII (1977), 223.
(42) Isaiah Berlin, Four Essays on Liberty (1958); Joel Feinberg, The Moral Limits of the Criminal Law (1984, 1985, 1986, 1990), 4 vols.
(44) See Patrick Devlin, The Enforcement of Morals (1968), and H. L. A. Hart, Law, Liberty, and Morality (1969).
(47) See Leo Zaibert, “The Moralist Strikes Back,” (2011) 14 New Crim. LR 139–161.
(50) Immanuel Kant, The Metaphysical Elements of Justice (transl. John Ladd, 1965).