Introduction: The Logic of Criminal Law
Introduction: The Logic of Criminal Law
Abstract and Keywords
This introductory chapter first sets out the purpose of the book, which is to explore some of the philosophical foundations of criminal law. It then outlines the themes covered by the book including the relationship between ‘real crimes’ and ‘mere regulatory offences’; the size and shape of the ‘general part’ of criminal law; the relationship between substantive criminal law and the law of criminal evidence and procedure; the nature of criminal ‘wrongs’, criminal ‘harms’, and criminal ‘responsibility’; and the existence and importance of the criminal lawyer's distinction between actus reus and mens rea. An overview of the subsequent chapters is also presented.
1. Critique without Philosophy?
Criminal law has been described as a ‘species of political and moral philosophy’.1 We would not go so far as to say that criminal law itself is a form of philosophy, nor that its philosophical dimension is exhausted by concerns with morality and politics. But we would say that it can and should be the subject of philosophical study. The aim of this book is to explore some of the philosophical foundations of the criminal law. Modern textbook writers often claim to be laying bare these foundations in their accounts of criminal law, but their attempts tend to be thin and insubstantial. If this book gives some depth and substance to such claims in the future, then it will have served its purpose.
But some may find that purpose too narrow. There are many urgent questions to which academic criminal lawyers now commonly feel that they must address themselves, such as the question of the correlation between ‘law in the books’ and ‘law in action’, and the question of whether the criminal law plays a key role in sustaining institutionalized forms of injustice and oppression. These are indeed important issues, and academic lawyers ignore them at their peril. Devoting a book to the philosophical foundations of criminal law may seem insignificant by comparison, like fiddling while Rome burns. None the less, the philosophical foundations are of crucial significance. The concepts of ‘law’, ‘criminal’, ‘injustice’, and ‘oppression’ are not self-explanatory, and different accounts of them compete for our allegiance. Revealing the nature and source of such differences and their impact upon our thinking cannot be done without philosophy. A central task of philosophy is to break down conventional false associations and establish true ones in their place. It is to challenge our preconceptions and reconstruct with greater candour our patterns of thinking about the world, if not the world itself.
Yet some people may think that the essays in this book fail to go deep (p.2) enough into our preconceptions and patterns of thinking. Impatient to get on to the grand issues of injustice and oppression, such critics will regard the concentration on seemingly mundane concepts like action, intention, foreseeability, and responsibility as little more than an ideological smokescreen, giving far too much credence to the complacent self-image of the criminal law. This is a mistake. It ignores the fact that a social practice has its own internal logic. One cannot properly understand its strengths and weaknesses until one understands its workings from within. Consider, for example, the ferocious contemporary debate over the denial to battered women of a defence to murder on the ground of provocation when they kill their violent abusers after what the law regards as a ‘cooling-off’ period. In order to bring out the injustice of this denial, one might point out that it reflects a failure to appreciate the predicament of women trapped in violent relationships. That casts the denial as a denial of an excuse. But it is easy to drift into a different argument, in which the denial is portrayed as a failure to see that the death of an abuser is a reasonable price to pay for the ending of abuse. This is a justificatory drift: it is an attempt to justify the conduct rather than excuse it. If what one is interested in exploiting is the gendered aspect of the denial, the second argument is evidently a weaker one. One can use the stronger first argument, however, only if the provocation defence, when it is granted, takes the form of an excuse. For there is no injustice in refusing to extend the excuse to abused women, if there is no excuse to extend. Thus one must begin by exploring the nature of the provocation defence, establishing that it is indeed an excuse and not a justification.2 That is exactly the sort of distinction which some would be inclined to dismiss at first sight as remote from the grand issues which ought to concern criminal lawyers today.
2. Beyond Punishment
One of the reasons why some have come to resist the philosophical study of the criminal law relates to an underestimation of the range of philosophical problems to which the criminal law gives rise. In recent years, self-consciously philosophical discussion of the proper scope and structure of the criminal law has been treated by many writers as a mere adjunct to discussion of the justification of punishment.3 It is widely thought that the (p.3) conditions for legitimate criminalization are largely dictated by the conditions for legitimate punishment. This is in many ways unfortunate. In the first place, the theory of punishment has become a battleground in which the most fanatical moral outlooks are routinely pitted against one another: uncompromising Benthamites against unyielding Kantians. Such crude extremism readily spills over to create rival caricatures of the criminal law as a whole. Second, the annexation of criminal-law theory to the theory of punishment has led some to take a highly selective view of what they need to justify when justifying the workings of the criminal law. There are many forms of sentencing which are not punitive. There is also much of moral consequence which takes place in criminal proceedings before the question of sentence is reached. These matters have been almost studiously neglected in the literature. Finally, there is a dangerous feedback, whereby the study of punishment is itself corrupted by using this highly selective view of the criminal law as the exemplar of all punitive practice. In reality, much punishment takes place outside the criminal justice system—among friends, in families, at school—and such punishment calls for justification no less than criminal punishment.
Take, for example, the problem of infancy. The punishment of children is an everyday practice; yet the criminal liability of children begins only at age 10. Those who append the theory of criminal liability to the theory of punishment must see this as a difficulty. If they want to defend the asymmetry, they must either portray the legal practice or the extra-legal practice as anomalous. Either the legal practice is an institutional concession, or the extra-legal practice is marginal, not strictly speaking punishment, but something that looks like it. If, on the other hand, the asymmetry is to be condemned, then the source and the object of the condemnation will be predictable. Some will present the extra-legal practice as a utilitarian device, while others will see the legal practice as a retributivist hang-up. In these responses to the asymmetry, the baneful influence of the annexation of criminal-law theory to the theory of punishment is all too evident.
It will come as no surprise, therefore, that the contributors to this volume have been asked to consider the philosophical foundations of criminal law from perspectives other than that of argument about punishment. They have made commendable efforts to comply with this (as some would have it, unreasonable) demand. The result has been a group of essays which, while in many respects diverse, are joined together in more than one union. There are small-scale unions that have frequently allowed us to group essays together around a particular topic. Those brave enough to attempt a beginning-to-end reading of the volume will be struck (if we have succeeded) by the common concerns of neighbouring essays. But there are also larger unities of theme, broadly concerned with the inner logic of the criminal law, which cut across the localized commonalities of (p.4) topic. These include the relationship between ‘real crimes’ and ‘mere regulatory offences’, the size and shape of the ‘general part’ of the criminal law, the relationship between substantive criminal law and the law of criminal evidence and procedure, the nature of criminal ‘wrongs’, criminal ‘harms’, and criminal ‘responsibility’, and the existence and importance of the criminal lawyer’s distinction between actus reus and mens rea.
3. Value in Action
The contours of several of these themes are charted in Alan Brudner’s essay. Although he avoids dwelling on the theory of punishment, his starting-point is with the Kantian and Benthamite extremes so beloved of many who write about punishment. At the heart of his essay lies the idea that there need be no irreconcilable conflict between instrumental and intrinsic concerns in the criminal law, between values which the criminal law must promote and values which it must honour.4 For Brudner, it is welfare which must be promoted and agency which must be honoured. The removal of any impression of irreconcilable conflict is achieved, not by attempting to subsume one under the other, as some have tried to do, but by showing that each depends upon the other. Whereas welfare would be meaningless without agency, agency would be unconsummated without welfare (p. 50).5 This is said to have significant pay-offs for the structure of the criminal law. The most important pay-off is that it shows how and why ‘true crimes’ are to be distinguished from merely regulatory offences: for the reconciliation of the conflict between agency and welfare generates more than one set of principles for the criminal law (pp. 50–1). Here, Brudner gives us a welcome reminder that the interesting moral issues raised by the criminal law are not exhausted by a study of killing, rape, and pillage; pollution, milk-adulteration, and speeding are no less rich in morally intriguing detail. Too often it has been suggested or implied that philosophical method gives rise to a myopic concentration on a narrow band of crimes which are not representative of the criminal law as experienced by most of those who fall foul of it. Brudner shows what an ill-considered criticism this is.
What is striking about Brudner’s argument, however, is that he seeks to rectify this myopia by relying on an unexpectedly parsimonious set of abstract premisses. Brudner’s starting-point seems to be that it is agency that falls to be honoured, and welfare that falls to be promoted; but why limit our moral concerns to these? Morality is a good deal more plural (p.5) than this. In protecting endangered species, for example, one might be promoting natural diversity for its own sake, irrespective of agency or welfare. In decrying the desecration of graves, meanwhile, one is perhaps concerned with honouring the dignity of the dead rather than honouring the agency or promoting the welfare of the living. Brudner does not say how these values (and values they undoubtedly are) might fit into his moral outlook. Of course, there may be reasons to doubt whether the criminal law should be concerned with these values. One might argue that the killing of endangered species or the desecration of graves should be prohibited, if at all, only on grounds other than the values of natural diversity and the dignity of the dead. This is the job that the traditional liberal ‘harm principle’ is supposed to do. Brudner has some trenchant criticisms to make of this principle, which he sees as excessively welfarist, in his sense (p. 25). But the harm principle is an exclusory principle, whereas what Brudner offers us appears to be an inclusory principle. The harm principle explains why some activities should not be criminalized, whereas Brudner’s ‘dialogic community’ explains why certain activities should be criminalized. It may be that ‘dialogic community’ is meant to be exclusory as well, but Brudner gives no arguments for thinking that this is so.
4. From Action to Responsibility
Even if morally selective, Brudner’s focus on agency does draw attention to a recurrent theme in some of the essays which follow. This is the concern with action, the unit of agency. While several other contributors draw connections between problems about the concept of action and problems about the criminal law, Brudner is the most explicit in offering a reason for making such connections. It is that agency is at the heart of value, and the value of the criminal law therefore depends upon the honouring of agency as well as the promotion of its consummation. This has ramifications for the structure of criminal law, for its enforcement, and not least, as Brudner stresses, for its proper scope.
When the concept of agency is introduced into discussions of criminal law, the emphasis is often not on these issues, but on issues of responsibility. It is often supposed that the key to understanding responsibility will be found in a study of action. This supposition is apt to mislead. Normal breathing does not involve action. It has a semi-automatic aspect which means that it cannot be put on all fours with writing a letter or eating lunch. Nevertheless, we can be straightforwardly responsible for our normal breathing. Suppose someone with healthy respiratory powers is told (p.6) by the doctor to hold his breath for a few seconds while she conducts a tricky test. If the test fails because he forgetfully breathes again too soon, he is responsible for the test’s failure. That is because he is responsible for the breathing, although it is no action. Nor is this merely the derivative sense of responsibility sometimes used as a substitute for ‘cause’ (as where one says that a storm was responsible for damage to a house); rather, this is normatively significant responsibility, of the sort that can be labelled legal or moral. He is responsible in this latter sense because he has, for a limited period, control over his breathing. What this case suggests is that control, while at the heart of responsibility, has no special role to play in action that it does not play elsewhere in our lives. This reveals that there is nothing special about action that provides the key to understanding responsibility. Indeed, as we will discover later, a focus on what makes action special will seriously distort one’s account of responsibility. Yet Jennifer Hornsby and Antony Duff focus on precisely this. Neither of them suggests that action is a necessary condition of responsibility. For Duff it is the ‘paradigm’ (pp. 77–8); for Hornsby there is merely a ‘connection’ (p. 70). Even these relatively modest proposals, however, seem to us to herald the same difficulties. Both authors also mention the importance of control for responsibility, but neither takes the natural further step, when thinking about responsibility, of shifting attention away from what is special about action to cases of control more generally.
Hornsby stresses the logical link between action and intention: ‘an action is a person’s doing something intentionally’ (p. 55). Her main point is that while intention may be central to action, and understanding action may be the key to understanding responsibility, it does not follow that one is responsible only or even paradigmatically for that which one intends to do or intentionally does. That is because when one acts, one does many things: every action, as other philosophers have put it, bears many descriptions. While at least one of these things that is done is in each case of action done intentionally, it does not follow that all are. Therefore, it may be possible to define actions by reference to the fact that in each action one thing at least is done intentionally, and it may be possible to use the concept of action thus defined to help with one’s account of responsibility, without committing oneself to the position that one is responsible only or paradigmatically for what one intends doing or intentionally does. One’s account of responsibility may extend, even paradigmatically, to the many things one unintentionally does in action.
Although Hornsby here works with an account of action in which intention is central, she is eager to show that different accounts of action need not be rivals. Thus, actions for her are not only things done intentionally, but also attemptings or tryings, as well as (in some sense) bodily movements (p. 60). Elsewhere she has concentrated on the element of (p.7) attempting or trying;6 and her views on that subject are briefly reiterated in her essay. It is to accounts of action in which trying plays a pivotal role that Duff addresses his contribution to this volume; but he interprets these accounts, as Hornsby herself clearly does not, as ventures to identify a single, exclusive essence of action. The focus may be on ‘trying’ accounts of action, but the real target is what Duff sees as a reductive tendency throughout the philosophy of action. One of the practical dangers in such reductiveness, according to Duff, is the effect that it can have upon our understanding of the criminal law. If, for example, one insists that actions are essentially tryings, and nothing else, then one will end up exaggerating the extent to which the criminal law should focus attention on our tryings. So, it is said, a trying-based account of action will tend to lend false support to an extreme form of subjectivism in the criminal law (p. 89). Here Duff tries to set the trap which, according to Hornsby, he has himself fallen into on other occasions. In his earlier ground-breaking work, Duff notices the link between action and intention, but then goes on as a result to exaggerate the pervasive importance of intention in the criminal law.7 Now he accuses those who emphasize trying, rather than intention, of getting into the very same kind of trouble.
This is not to say that Duff overlooks the distinction between an action and its descriptions, or, as Hornsby reconstructs it, the distinction between an action and things done in that action. On the contrary, he points out that this distinction is a matter of some significance for the criminal law. The general requirement that there be a ‘voluntary act’ does not require, for example, that the defendant acted voluntarily under the particular description which has legal salience, but only that the defendant acted voluntarily tout court (p. 77). Thus, Duff’s opposition to essentialism about action is seemingly not an opposition to distinguishing actions from their descriptions. On the other hand, Duff appears to remain attracted by a holistic view of action, according to which the description applicable to an action makes a difference to what action it is (p. 105). In the last analysis, Duff stands by his instinct to deny the possibility of distinguishing actions, in anything other than name, from their circumstances and consequences (i.e. from the facts and events in terms of which they are described). This may be Duff the moralist speaking, rather than Duff the analyst. He may be reacting to the tendency of some criminal lawyers, faced with the possibility of distinguishing an action from its circumstances and consequences, to exaggerate the normative significance of the distinction. He may be reacting, for example, to the tendency to play down the relevance of consequences to moral and legal responsibility.
(p.8) 5. Counting Consequences
Andrew Ashworth’s essay not only exemplifies but defends that tendency. He believes that actions are essentially ‘tryings’—the analytical point.8 He also believes that, other things being equal, criminal responsibility should (because moral responsibility does) extend to and only to what the criminal tries to achieve—the normative point. But he does not mix up the two points. It is merely that the same thought underlies both of them: that trying is all that we have control over. We think this fails to secure the analytical point—not on Duff’s grounds, but on the simpler ground that there are events in our lives over which we have control but which are not actions (remember the ‘forgetful breathing’ example). There is no reason to define actions in some artificially wide way to accommodate such cases. Meanwhile, the normative point is also affected by such cases, because the things which we have control over other than actions (such as the forgetful breathing), are also not tryings. They are cases in which, if one had tried, one would have made a difference: that is the element of control. But the fact is that one did not try. So it cannot be the case that the dependence of responsibility upon control also makes responsibility depend upon trying.
Ashworth might reply that even though there can be responsibility without trying, in cases in which trying is in play responsibility only goes as far as the trying does (i.e. when one tries, one is not responsible for what one does not try to do). Ashworth appears to equate the view that responsibility only goes as far as trying with the view that once one gets to descriptions of one’s actions which refer to unforeseen (or at any rate unforeseeable) consequences, one’s responsibility runs out. What is not made clear is how one leaps from questions about trying, to questions about foresight and foreseeability. What about the foreseen and foreseeable things which we do without trying to do, which are merely side-effects of the things that we try to do? Even if we set this problem aside,9 however, Ashworth faces deeper difficulties. He wishes to combine his ‘trying’ principle, which is supposed to make criminal law insensitive to certain important kinds of luck, with another principle of which he was the original articulator, namely the principle of ‘representative labelling’ (or ‘fair labelling’, as he now calls it).10 The principle of representative (p.9) labelling requires that the names given to criminal offences, and the labels accordingly attached to convicted offenders, should give an accurate picture of the offence of which they have been convicted, including its gravity. One might assume, then, that it would be important to include in the label a clear indication of the harm, if any, which the offender is supposed to have brought about. In deploying the trying principle, however, Ashworth resists precisely this. Take his example of the batterer whose battery causes death (p. 118). English law would label her a manslaughterer, and at first sight it is hard to think of a more representative label; but Ashworth’s trying principle would make her, under certain conditions, a batterer and no more.
Perhaps there is simply a conflict here between Ashworth’s two principles, the trying principle and the representative labelling principle. In his contribution, however, Ashworth seems to envisage that they are in harmony. Harmony is presumably achieved only at the expense of a modification to the representative labelling principle itself. Not only must the label give an accurate representation of what the defendant did, but it must also give a representation which makes no reference to features of what the defendant did which lie beyond what he tried to do. That may explain why Ashworth has now accepted the name ‘fair’ labelling for his principle, instead of representative labelling. The extra work is now done by the idea of ‘fairness’, an idea which is not explicitly elaborated in Ashworth’s contribution.
Ashworth assumes throughout that the law can in principle (if not in practice) be shaped so as to eliminate the important kinds of luck, sensitivity to which would fall foul of his ‘trying’ principle. Michael Moore’s essay seeks to cast doubt on the coherence of some of the law’s attempts to narrow down the field of luck by the use of foreseeability criteria. His argument echoes several of the points made by Hornsby. Once we have realized that there are many things done in every action, we can see that any other eventuality, be it a death or a solar eclipse or a gas explosion, may also be both one thing and many things. Each is one thing in the sense that it is a particular incident occurring at a particular place and time, but also many things in the sense that it can be variously described. Think of a death. In a sense, a death is something that happens, with its own spatio-temporal location: it is what Moore and Hornsby, following philosophical convention, term a ‘particular’ (pp. 56, 127). Particulars may be variously described. Depending on the circumstances, a death may be described as a bereavement, a murder, a poisoning, a tragedy, a blessed release, or the supreme self-sacrifice. Or it may simply be described as a death. The result is that when we say ‘that death was intended (or foreseen or foreseeable)’, the question arises: was it the death qua particular that was intended, foreseen, or foreseeable, or was it the death, understood as just one of many descriptions which that particular may bear? The answer is that (p.10) mentalistic terms like ‘intended’, ‘foreseen’, and ‘foreseeable’ cannot by their nature be predicated of particulars. They introduce traces of what are often called ‘propositional attitudes’ (i.e. mental states the content of which takes the form of a proposition). Moore claims that this gives rise to problems in certain areas of the law. Sometimes, if we are to be legally liable, what the law requires us to intend or foresee, or what the law stipulates must be foreseeable, is merely ‘harm’ or ‘damage’. It seems to Moore that the law cannot mean ‘harm’ or ‘damage’ to count as a description of something: these terms are too open-ended to give any useful guidance as to the limits of legal liability, and yet the law claims to find useful guidance in the intention-foresight-foreseeability tests which take this form. Moore runs through the other things that ‘harm’ could be if it is not a description, but shows that the law gives too little specification for it to be a useful test, however it is meant.
Moore takes this situation to yield a fundamental ‘incoherence’ in certain mentalistic tests of legal liability, notably those used in determining legal causation (pp. 154–5). This way of putting the point is reminiscent, perhaps, of those sceptics that Moore describes (citing Morris Cohen) as the ‘stray dogs of the intellectual world’ (p. 130). His talk of incoherence seems in fact to be a dramatic way of expressing the point that the law could sometimes be sharpened up. We are all familiar with situations in which legal tests are ripe for more detailed specification. There may be doubts, for example, about the level of generality of the law’s description of the damage which must be foreseen or foreseeable for an arson conviction. Is it merely the destruction of the building which must have been foreseen or foreseeable, or is it the destruction of the building by fire?11 The law is open to interpretation here, but (unless you are a stray dog) it is hardly a case of incoherence.
6. Knowing Right from Wrong
Perhaps this is, however, to overlook a further subtlety, with which Moore does not concern himself. The subtlety concerns what it means for one to foresee something under its legally salient description. It could mean that one foresees it under the description which happens, unbeknown to one, to have legal salience. One knows the facts but not the law. Or it could mean that one actually appreciates the legal salience as well. One knows the facts and the law. Should it matter, for the purposes of criminal liability, (p.11) into which category one falls? English law generally says no: ignorantia juris neminem excusat. Some critics have linked the emergence of this doctrine to a time (surely imaginary) when English criminal law was broadly coterminous with morality.12 This would suggest that ignorantia juris neminem excusat is a doctrine which is easier to defend concerning crimes mala in se (crimes which would be wrongs even if there were no applicable legal prohibition) than it is concerning crimes mala prohibita (crimes which are wrongs only because of the applicable legal prohibition). This might be thought of as the traditional view. Douglas Husak and Andrew von Hirsch take a different view, envisaging a dividing line between those who can avail themselves of a mistake-of-law defence and those who cannot, that would cut across the mala in se/mala prohibita distinction. The central plank of their argument is that certain mistakes of law affect culpability. They recognize that criminal liability does not always turn on culpability, although they are, we think, too inclined to play down the incidence of strict liability and the importance of regulatory offences (p. 157). This may be wishful thinking: by contrast with Brudner, Husak and von Hirsch appear to think that a unified doctrine of culpability should prevail across the whole criminal law.
According to the doctrine of culpability which Husak and von Hirsch prefer, the effect of mistakes of law depends on a number of considerations. In the first place such mistakes should only furnish a defence if they are, in a stipulated sense, reasonable. But even that will not suffice by itself. The defendant must also have made a mistake, with some rational basis, as to the ‘injuriousness’ of her allegedly criminal conduct (p. 168). The relevant sense of ‘injuriousness’ is evidently heavily moralized: even the seemingly implausible view that euthanasia is not ‘injurious’ is among those that might count as ‘plausible’ on the Husak and von Hirsch account (p. 165). ‘Injuriousness’ appears to mean something like ‘all things considered wrongness’. The requirement that one make a ‘plausible’ mistake as to ‘injuriousness’ in this sense rules out the use of Husak and von Hirsch’s mistake-of-law defence in many, but not all, mala in se crimes. One can use the defence even where what one did was wrong apart from the law, so long as one had the requisite ‘plausible’ belief that it was not. The position with crimes mala prohibita is a little more complicated. The whole point about these crimes is that they would not be wrongs apart from the law. That means, if we read Husak and von Hirsch right, that the ‘injuriousness’ element in such crimes would be lacking, apart from the law. If one has no reason to know the law, then one has no reason to know of the ‘injuriousness’. Leaving aside those who find ‘injuriousness’ on spurious grounds, then, those who reasonably mistake the law concerning a (p.12) malum prohibitum crime appear automatically to meet the conditions for the Husak and von Hirsch defence. This all assumes, however, that the offence in question carries a culpability requirement: for if it does not, the Husak and von Hirsch argument does not apply anyway. Given the profusion of mala prohibita crimes which are strict liability crimes, this should not be dismissed as a minor caveat. Its effect is to generate many counterexamples to the traditional view that a mistake-of-law defence is easier to justify for mala prohibita crimes than for mala in se crimes.
Husak and von Hirsch’s essay illuminates the case of the defendant who did not know that he was breaking the law, but admits the moral wrong. They expressly decline to take a view on the converse case: the defendant who knows she is breaking the law, but denies the moral wrong (p. 165). That case, they say, can only be dealt with once one considers the arguments for a general moral obligation to obey the law. George Fletcher, on the other hand, feels able to discuss this case without going into that wider problem. His strategy is to distinguish the function of the legislature in criminal law from that of the judiciary. The legislature must strive for certainty in virtue of the nullum crimen sine lege principle. Judges, he thinks, are also bound by this principle to the extent that they may not alter the definition of offences as such. They may, however, respond to the facts of particular cases by altering the scope of defences. This they must do in accordance with what Fletcher calls ‘the Right’, a moralized account of legality (p. 177). One’s nominal legal wrongdoing may be eclipsed in court by some justification derived from that moralized account.
This suggests that justifications need not come to light until the deed is done. It might be thought, however, that justifications must, by their nature, be accessible ex ante to potential offenders as well as ex post to judges. Fletcher himself claims that whether one will benefit from a justification depends on whether one knows of the justificatory facts (p. 179). Sometimes he seems to be going even further. There are hints that what really matters is whether one acted by reason of the justificatory facts, and not merely with knowledge of them: Fletcher speaks of justificatory ‘intent’ (p. 179). If the argument can be pushed that far, it is hard to see how justifications can be held to arise independently of defendants’ reliance upon them, as Fletcher later suggests that they do (p. 186). One way out of this apparent dilemma would be to distinguish the justificatory facts from the fact of the justification. In other words, any defendant who is to benefit from a justification must act for the reason that the justificatory facts obtain, but need not act for the reason that they count as justificatory, which is an ex post matter for the judge. It is hard to see, however, why anyone would act for the reason that justificatory facts obtain if she did not also act by reason of the justificatory force of those facts. It is true that she might not act for the reason that they constitute a legal justification, (p.13) and thus the law would not be relied upon. Instead, she might be acting on the moral justificatory force of the facts. Thus Fletcher’s talk of Recht could be recast in more positivist terms, as the judges’ giving ex post legal recognition to ex ante moral arguments, or at any rate ex ante moral arguments which can be fitted into established legal categories.
7. Distinguishing Prohibition from Responsibility
Paul Robinson also lays much stress on the distinction between ex ante and ex post aspects of the criminal law, between the prohibitions addressed to potential perpetrators and the rules of responsibility addressed to judges. Robinson is concerned with the merits and demerits of the traditional lawyer’s distinction between the actus reus of a crime and its mens tea. One of the merits of this distinction, in Robinson’s eyes, is that it attempts to mark the important line between the definition of the prohibited conduct and the rules of responsibility. One of the demerits of the distinction is that it fails to mark this line properly (p. 206). Robinson is right to find a failure here, and right to think that many criminal lawyers too readily identify mens rea with responsibility.13 The actus reus-mens rea distinction cuts across the distinction between prohibition elements and elements of responsibility. In the English law of theft, for example, the requirement of an ‘intention permanently to deprive’ is plainly part of the mens rea of theft, if anything is. Nevertheless, and equally plainly, it goes to define the nature of the wrong itself rather than what makes the wrongdoer responsible for the wrong. The test for determining which of these it goes to (a test which Robinson does not spell out) relates to the nature of the justification for including in the definition of the crime the element of intention permanently to deprive. We all have certain special reasons, symbolic and instrumental, not to take others’ property permanently, which are supplementary to our reasons not to take it temporarily. These give rise to further reasons, again symbolic and instrumental, for us not to intend to deprive permanently. So the reasons which drive this part of the definition of theft are reasons for us, in so far as we are potential perpetrators, rather than for judges. They are ex ante rather than ex post. In recognition of all this, and unable to find any other use for the actus reus-mens rea distinction, Robinson argues that the criminal lawyer would be well advised to dispense with the distinction altogether. It would be better, he thinks, to divide criminal-law definitions up directly into the elements of prohibited conduct, on the one hand, and the rules of (p.14) responsibility on the other. The former he calls ‘rule articulation’ elements, while the latter are described, depending on how exactly they operate, as rules of ‘liability’ and of ‘grading’. The problem for Robinson is that he would like to preserve the kind of organizational tidiness which the actus reus-mens rea distinction has always been used to supply. He wants to be able to say something general (i.e. crossing the boundaries between different crimes) about what kinds of things will be found in each of the three classifications. ‘Future-conduct intention’ can bear on ‘rule articulation’, whereas ‘present-conduct intention’ will be found in the ‘liability’ category and ‘future-result culpability’ will matter for ‘grading’. The truth is that things are less tidy (even) than this. Consider the example of homicide. Robinson uses the distinction between murder and manslaughter to illustrate ‘grading’, a matter of ex post responsibility (p. 208). On his analysis, the distinction relates to grading because it is a distinction of ‘future-result culpability’, i.e. a distinction between intention and recklessness as to the death. But it is really a distinction between two different wrongs, between two different prohibitions. The distinction between murder and manslaughter is analogous, in all relevant respects, to the distinction between theft and a crime of temporary deprivation (e.g. ‘joy-riding’); yet Robinson would happily classify the latter as a difference of rule articulation because it involves what he calls future-conduct intention (p. 209). Once again, the point is that we all have special reasons, symbolic and instrumental, not to intend to kill, which are supplementary to the reasons we have for not taking the risk of death, knowingly or otherwise. In the end, Robinson’s classificatory system faces the same difficulty which makes the actus reus-mens rea distinction a distinction of limited use. It is that little can be said, at least of a kind that will help criminal lawyers, about the logical structure of crimes in general. Those things that one can say about the logical structure of crimes in general are doubtless of great philosophical interest, but they will not, as Robinson hopes, help to make the textbooks better.
Richard Tur’s arguments bear this out. He characterizes the law in general, and the criminal law in particular, as a set of ‘defeasible normative conditional propositions’ (p. 214), but illustrates how this is compatible with there being criminal offences of many different configurations. Thus, what is part of the definition of one crime (i.e. bears on conditionally) may be merely the absence of a defence in another (i.e. bears on defeasibility).14 As it stands, this is merely a formal philosophical distinction. In order to (p.15) make it a matter of concern to criminal lawyers one must add certain normative premisses. To some extent, in Tur’s essay, these normative premisses are moral propositions independent of law (p. 222). In large measure, however, the normative premisses are themselves propositions of law. Tur’s view is evidently that some aspects of the criminal law must be taken as ‘givens’ when others are being contested. It is only because and to the extent that the law already happens to make something turn on the distinction between the elements of an offence and the conditions of a defence that it is a matter of moral consequence whether one is a ‘definitional maximalist’ or a ‘definitional minimalist’: whether, in other words, one is predisposed to count any particular legal doctrine bearing on criminal liability as going to conditionality or to defeasibility. So, for example, if there are general principles of law which make the whole actus reus subject to mens rea, and dictate what kind of mens rea that will be, then the only way to get away from the general requirement of mens rea may be to shift a certain part of the actus reus into a defensive role where the general principles of law dictating mens rea do not apply.
Tur himself is willing, however, to subvert the ‘givens’ on which he relies. Having built his definitional minimalism on the back of the assumption that putting elements on the ‘offence side’ constrains one to take a strict view of the mens rea required, he goes on to present a view of the criminal law in which things are apparently much less strict. There can be crimes of ‘partial’ mens rea, or no mens rea at all: there are many permutations. That being so, it is not so clear why one need have a general view about how much or how little is to be included in the definition of a crime. It is not clear, in other words, why one needs to choose some general position on the scale from definitional minimalism to definitional maximalism. One motivation for so choosing might be that, like Fletcher, one considers the definition of crimes to be a matter for the legislature (by virtue of the principle of nullum crimen sine lege), and one also considers that judges are better placed to deal with certain problems of responsibility.15 In that case, one might need to say that those problems of responsibility belong to the realm of defences rather than to the definition of offences. There are certainly hints of this argument to be found in Tur’s essay. He indicates that the list of defeating circumstances is always open to change, presumably in contradistinction to the element of conditionality in criminal offences, and he indicates a preference that the openness to change of the former be exploited: judicial justice over legislative certainty (p. 215).
(p.16) The contrast between conditionally and defeasibility also plays a major role in Stephen Morse’s essay. He points out that the symptoms of mental illness can be significant on both sides of the equation. They can be used both to deny mens rea and to provide a positive defence. In the former role they have an instrumental (evidentiary) significance; that is to say, they can help one to establish lack of mens rea, but do not constitute lack of mens rea. In the latter role, however, they are of intrinsic significance. The fact that one suffers from certain symptoms of mental illness actually constitutes the defence. The puzzle about Morse’s proposal, at least from the point of view of an English lawyer, is that he finds no constitutive significance for mental illness on the conditionality side of criminal liability. It is a commonplace among English criminal lawyers that voluntary conduct is a (general) positive requirement of criminal liability, quite separate from any element of mens rea which may be required by the definition of particular crimes.16 It is also a commonplace that the so-called ‘defence’ of automatism is in reality no more than a denial of voluntary conduct. And there can be no doubt that, whatever automatism may be, it is sometimes a symptom of mental illness; hence ‘insane automatism’. If all this is correct, then (pace Morse) the mere fact that a symptom of mental illness is pleaded constitutively rather than instrumentally should not lead one to think that it necessarily belongs to the realm of defeasibility. It is not necessarily, in other words, being treated as a positive defence.
It may be thought that this matter is important only because of the twists and turns of criminal procedure and evidence: all that turns on whether something is a matter of conditionality rather than a matter of defeasibility are things like the evidentiary burden and who is allowed to raise the issue. But there is also a moral distinction at work. Morality deals with responsibility in two stages. There is prima-facie responsibility, and then there is all-things-considered responsibility. The questions which arise at the stage of prima-facie responsibility are about one’s susceptibility to moral judgments in the light of what one did, irrespective of what exactly it was one did. They concern, in other words, whether one falls (for the time being) under the judgmental jurisdiction of morality. All-things-considered responsibility, on the other hand, is sensitive to what one actually did. It is here that questions of excuse come in.17 That excuse belongs here and not at the stage of establishing prima-facie responsibility is demonstrated by the familiar fact that some wrongs are inexcusable.
Morse holds that when the symptoms of mental illness are not used instrumentally (to help deny mens rea), they can only be used as excuses. They are in that respect (although apparently not in other respects) (p.17) analogous to coercion (p. 265). The result of this is that the symptoms of mental illness can never be enough to take one outside the province of criminal liability altogether, whatever the crime may be. They cannot be used to eliminate liability for inexcusable crimes. Are there any such crimes in force today? In recent times, English law has given a firm ‘yes’. There is the now familiar rule that murder cannot be excused by duress, a rule which the judges purport to justify by presenting murder as always inexcusable.18 Morse’s argument would suggest that if the judges are right here, then mental-illness arguments should not be available in murder cases either. The diminished-responsibility defence, which operates in England in relation only to murder, is therefore turned on its head. Oddly enough, on Morse’s argument, murder is the very last place in which the defence should operate.
9. Science and Reason
Bill Fulford shares Morse’s preference for describing mental illness as an excuse. He also shares with Morse a view according to which the exculpatory significance of mental illness lies in its impact upon practical reasoning. But whereas Morse defends this view against a naïve moralist’s insistence upon assimilating mental illness to coercion, Fulford defends it against a naïve psychiatrist’s insistence upon reducing mental illness to disease. Fulford is anxious to distinguish illness from disease. Illness is the sufferer’s experience of disease, and disease the medical expert’s account of illness. If we accept this reclassification, then we have been guilty of pleonasm in speaking (up to now) of ‘the symptoms of mental illness’: the mental illness actually is the symptoms, as experienced by the sufferer; it is not something which has symptoms. What does not follow, but what psychiatrists are said to take too often for granted, is that our interest in mental illness must reflect its role in evidencing the presence of mental disease. On the contrary, argues Fulford, it is the illness itself which matters, and the search for an underlying disease should be viewed as a matter of derivative concern. If a disease can be found, so much the better, since that will open up new channels for treatment and cure, as well as proof of the illness. But if, as in the case of many a psychosis, the illness is unsupported by any (known) disease, we should not be led to doubt for a moment that (p.18) it can truly qualify as an illness. Nor, therefore, should we be led to doubt that it is capable of affecting responsibility. As Fulford notices, the criminal law typically errs here in treating the disease as pivotal.19
In shifting our attention from the disease to the illness, Fulford also brings our thoughts back to questions with which we began, questions about the role of action and value in criminal law. According to Fulford, the disease paradigm restricts our attention, with scientistic zeal, to matters of ‘fact’ and ‘functioning’, sidelining the discourse of action and value in terms of which the moral and legal significance of mental illness must be cast. Here Fulford works with a sharp distinction between fact and value, which has been at the heart of his earlier work on conceptual problems in medicine.20 To this he adds another distinction, perhaps less sharp, between matters of mental and physical functioning on the one hand and matters of reasoning and action on the other. The shift of attention from disease to illness brings with it a shift of attention from the discourse of functioning to the discourse of action. Here Fulford associates himself with a view that we criticized in our discussion of Hornsby and Duff. He appears to identify the boundaries of action with the boundaries of responsibility, and also appears to identify the boundaries of intention with the boundaries of action (pp. 300, 305). These false identifications have led some to subscribe to distorted accounts of responsibility in which questions of cause and effect are paramount. That is because the causes of actions (some say, the intentions which cause them) are what distinguish actions, first and foremost, from other kinds of conduct (such as the conduct of the forgetful breather), and so viewing the presence of action as the key to questions of responsibility makes questions of responsibility seem to turn, in part, on the presence of certain causal factors. Fulford, however, does not let his false identifications lead him down such a dangerous path. One of the main achievements of his essay is to distance questions of responsibility from questions of cause and effect. He does this by observing that the causal role of mental disease in disrupting practical reason is a contingency; that is to say, to ask whether one is rational or not is different from asking how it came about that one is rational or not. And what matters for responsibility, Fulford argues, is one’s rationality or irrationality. The causal origins are neither here nor there. On the same grounds, the causal origins of one’s conduct are neither here nor there: the forgetful breather is responsible for his breathing, even though that breathing is caused entirely by spontaneous reflex movements in his chest cavity. And so would he be, as Fulford’s argument reminds us, even if those reflex movements were in turn caused by some disease. So long as he is in control (p.19) of the breathing, his responsibility is intact. As in the body, so in the mind. It is not hard to see where the confusion creeps in. The control relation is in a loose sense causal. It belongs to a family of relations, of which the relation of cause and effect in a stricter sense is simply another member. There are some family resemblances. Both relations have counterfactual elements and both have what may be called a ‘direction of operation’.21 So the two are easily mistaken for one another. Nevertheless, there are important differences. To be in control of something is not necessarily to be its cause, nor vice versa. Take the example of a guided missile. The missile’s guidance system may be in control of the missile—the missile may be under control—even though its movements are so far entirely caused by the launch momentum. The control is lost if no adjustment in the guidance system would make a difference to the missile’s movements.22 But so long as no adjustment is necessary, the control of the system may be maintained throughout the operation of numerous external causes. It does not even matter, for the guidance system to be in control, what caused the working of the guidance system. If we understand ourselves as complex guidance systems, which is how we must be understood if the guidance of morality and law is to apply to us and be applied by us, then it is not hard to see how the causal history of our reasoning and our conduct can be left on one side when our responsibility is being assessed.23 Those who think that real responsibility is hostage to the facts about determinism are often accused of misunderstanding determinism or its role in the world. Our suggestion is much more mundane. They have merely misunderstood the conditions of responsibility.
10. Liberal Values
The misunderstanding is compounded by the introduction of a third concept, that of choice. Those who reject the modern criminal law and its philosophical foundations often do so because they think it is permeated with assumptions about human freedom which liberalism makes but determinism unseats. The rejection is ill-considered on two fronts. The preoccupation with responsibility, and with the grounds of responsibility which we find in the modern criminal law, are not peculiarly liberal. And neither the preoccupations of liberalism nor the foundations of the modern (p.20) criminal law are especially challenged by determinism. Liberal ideology is dominated by a concern with choices, ranging from small choices about what to do here and now to big choices about how to lead one’s life. Certainly the criminal law is not immune to such concerns. They serve to justify the harm principle, for example, if anything does. They play a large part in the shaping of criminal prohibitions. But they do not play such a large role as critics are wont to claim in our doctrines of criminal responsibility. One’s conduct may be voluntary even though one had no options; and without options, of course, one has no choices. Choices entail alternatives, but basic criminal responsibility is largely unaffected by the presence or absence of alternatives. Nor is moral responsibility much different. It depends primarily on control of conduct and self-control, not on choice. An account of it compatible with that gestured towards here has been defended at least since Aristotle, and throughout the works of many (Aristotle included) who could scarcely be described as liberals. It may be fashionable to lay all the problems of our criminal justice system at the door of our ideological tradition, but one of the aims of this book—and certainly of this introduction—is to show that the philosophical foundations of criminal-law doctrine are far from being that simple. Liberalism claims to be open to many values, to many moral concerns which are not dictated by liberalism itself. Nowhere is that openness more vividly illustrated than in the diverse preoccupations of criminal-law doctrine.
We would like to thank Christopher Shields and Andrew Simester for their comments on earlier drafts of this introduction. We have also benefited from the comments of many of the contributors to this volume. Jennifer Hornsby gave particularly helpful advice.
(1) George Fletcher, Rethinking Criminal Law (Boston 1978), p. xix.
(2) A case where the facts raise this issue is R v. Pearson (William)  Criminal LR 193, considering the position of close relatives who kill an abuser.
(3) e.g. Nicola Lacey, State Punishment: Political Principles and Community Values (London 1988); Alan Norrie, Law, Ideology, and Punishment: Retrieval and Critique of the Liberal Idea of Justice (Dordrecht 1990); C. L. Ten, Crime, Guilt, and Punishment (Oxford 1987).
(4) Philip Pettit, ‘Consequentialism and Respect for Persons’, (1989) 100 Ethics 116.
(5) Throughout this introduction, page references to the essays in this volume are shown in brackets in the text.
(8) Although he denies making the mistake which Duff warns against, of viewing a ‘trying’ as some kind of mysterious mental act detached from the actual things that one does (p. 109).
(9) Elsewhere Ashworth has attempted to fill this gap by using what he dubs the ‘belief principle; see A. Ashworth, ‘Belief, Intent and Criminal Liability’, in J. Eekelaar and J. Bell (eds.), Oxford Essays in Jurisprudence, Third Series (Oxford 1989).
(10) The principle is introduced in Ashworth, ‘The Elasticity of Mens Rea’, in Colin Tapper (ed.), Crime, Proof, and Punishment (London 1981), 45 at 53. The name change was proposed for no obvious reason by Glanville Williams in ‘Convictions and Fair Labelling’, (1983) 42 Cambridge LJ 85.
(11) Compare J. C. Smith and B. Hogan, Criminal Law (7th edn., London 1992), 708: ‘D must intend or be reckless as to destruction or damage by fire’ with Marise Cremona, Criminal Law (London 1989), 193: ‘The serious offence of arson requires no additional mens rea’ presumably beyond that required for ordinary criminal damage.
(12) See Glanville Williams, Textbook of Criminal Law (2nd edn., London 1983), 451.
(13) See e.g. C. M. V. Clarkson and H. M. Keating, Criminal Law: Text and Materials (2nd edn., London 1990), 149–50.
(14) On defeasibility, see H. L. A. Hart’s ‘The Ascription of Responsibility and Rights’, in (1948) 48 Proceedings of the Aristotelian Society 121, which he overhastily disowned in the preface to Punishment and Responsibility (Oxford 1968).
(15) See J. C. Smith, Justification and Excuse in the Criminal Law (London 1989), 5–6. Smith supports the view taken by the Law Commission in its 1985 proposals for codification of the English criminal law, that no attempt should be made to codify all defences to crime, even though all offences should be codified.
(17) Just as justification comes in when we move from the question of prima-facie wrongdoing to that of all-things-considered wrongdoing.
(18) See DPP v. Lynch  AC 653, Abbott v. R  AC 755, R v. Howe  AC 417, R v. Gotts  1 QB 660. It might be thought that the view that murder is always inexcusable is inconsistent with the doctrine of provocation. In the view of Lord Diplock, provocation is ‘an anomaly in English law’ (see DPP v. Camplin  AC 705, 713). Another explanation of the provocation doctrine would point to the fact that it is only a partial excuse, and would reinterpret the view that murder is always inexcusable as a view that there is no complete excuse for murder.
(19) Fulford correctly cites, in this regard, Bratty v. Attorney-General for Northern Ireland  3 All ER 523, and R v. Quick  3 All ER 347.
(20) K. W. M. Fulford, Moral Theory and Medical Practice (Cambridge 1989).
(22) See Harry Frankfurt, ‘The Problem of Action’, in his The Importance of What We Care About (Cambridge 1988).
(23) Of course, the causal history of many other things is of vital importance for responsibility. Hence our earlier reservations about the tendency to play down the relevance of consequences to responsibility.