In this post I’m moving away from A Debate over Rights to develop some thoughts inspired by a couple of papers by John Gardner. I’m not going far from the book, though – the first section of this post is relevant to the question of how we conceptualise rights, while the second relates to the question of the morality of law (which two of the book’s authors have been debating for some time).
1. Oh you shouldn’t do that
The opening paragraph of John Gardner’s 1996 paper ‘Discrimination as Injustice’ makes an interesting claim about torture – the wrongness of torture, in particular.
Reasons of justice are reasons for or against altering someone’s relative position. The word ‘relative’ is of the essence here. One may have reasons to alter someone’s position which do not make any essential reference to anyone else’s position. For example, the fact that a prisoner is being tortured is reason enough by itself to write letters of protest, with the aim of improving the prisoner’s treatment. Torture is inhumane. But isn’t torture also unjust? Doesn’t one also have a reason of justice to protest? Perhaps. As part of one’s protest, one might relate the position of the torture victim to the position of other people (other prisoners, people of different political views, the torturers themselves, the torture victim’s victims, the government, etc). In that case one may be trying to give a reason of justice for the torture to desist. It may buttress the reason of humanity. But of course it may also fail to do so. The authorities inflicting the torture may accurately reply, in some cases, that they are inflicting it with impeccable justice. Yet still, on grounds of its inhumanity, the torture should cease, and the protests should go on if it does not.
Gardner returned to this point more recently, in his 2011 paper ‘What is tort law for? Part 1. The place of corrective justice’.
Norms of justice are moral norms of a distinctive type. They are norms for tackling allocative moral questions, questions about who is to get how much of what. Some people think of all moral questions, or at least all moral questions relevant to politics and law, as allocative. But that is a mistake. As a rule, allocative questions are forced upon us only when people make competing claims to assignable goods. Many morally significant goods, including many relevant to politics and law, are either not competed for or not assignable. They include goods such as living in a peaceful world and not being tortured. … Of course it does not follow that there are no questions of justice that bear on the resort to torture or on the quest for a peaceful world. The point is only that many moral questions about the resort to torture and the quest for a peaceful world are not questions of justice. If, for example, we say of someone who was tortured by the secret police that her treatment was unjust, she might well say, if her moral sensitivity has been left intact, that this misses the point and marginalizes her grievance. She is not complaining that she was the wrong person to be picked out for torture, that she was a victim of some kind of misallocation by the secret police, that she of all people should not have been tortured. She is complaining that torture should not have been used at all, against anyone. Her complaint is one of barbarity, never mind any incidental injustices involved in it.
Torture is inhumane or barbaric – there are other words we could use, such as ‘degrading’ or ‘brutalising’; the core meaning has to do with attacking or invading another person’s humanity or personhood. Morally, it should stop, both universally and in any given case – but it is not, of itself, unjust. The moral question raised by torture isn’t a question of allocating it justly. One distribution of torture may be prima facie less just than another – the torture of randomly-stopped motorists would arouse more outrage than the torture of convicted rapists – but the less unjust distribution is not less immoral. A regime which reserved torture for people found guilty of heinous crimes would still be morally repugnant. Any torture – for anyone – is bad torture; in an absolute sense, any torture – for anyone – is as bad as any other torture.
Gardner sets torture alongside position-relative justice, and the freely competing subjects of law-governed society, to make a point about the limits of allocative justice. No distribution of torture (or of absolute poverty, polluted air, reduced life-expectancy, etc) is more just than any other. This is both because torture is not a good to be appropriately allocated and, more importantly, because the absence of torture is not an assignable good and hence not subject to constraints of scarcity. The question of who should be exposed to torture, instead of the current victim, doesn’t arise. There is no reason, in principle, why there should not be enough non-torture for everyone – and, here and now, it will always be better if our actions do not add any more people to those already suffering it.
But there’s a bit more going on here than that. There are any many ills whose absence is not an assignable good. To put it another way, there are any number of areas in which life could in principle be made better for everyone, or (to put it in less ambitious terms) where making life better for one person doesn’t require making it worse for another: health, clean air, peace, Pettit’s ‘dominion’ (a condition of resilient non-intererference’). Depriving someone of a non-assignable good is morally wrong, without necessarily being unjust. Allocative thinking in a negative form may well be involved in the infliction of such an ill: it may be motivated precisely by the desire to improve one’s own relative position at the expense of the victim. However, allocative questions do not have to be involved in their rectification: there is in principle no shortage of clean air, so the harm of air pollution is not rectified by ensuring that the air the company directors have to breathe is equally polluted.
Actions of this type are, by definition, characterised by a lack of respect for the equal entitlements of others and ourselves. Since they don’t profit the person carrying them out (also by definition), they tend to have a character of gratuitous or vindictive malice. The definition does not, however, imply that such acts are all inhumane or barbaric. If I jammed my neighbour’s TV reception so that they were unable to receive BBC 4, this would certainly be a maliciously cruel act, but it would be a stretch to classify it as barbarity. Indeed, much of what tends to fall under the heading of anti-social behaviour consists precisely of the deliberate or reckless deprivation of others of non-assignable goods – goods like the ability to sleep undisturbed by noise or to walk to the shops unperturbed by vandalism. Depriving others of non-assignable goods is a bad thing to do, and there is no situation in which we should not, morally, strive to do less of it – but it is not generally barbaric or inhumane.
Obviously torture makes a much better example for Gardner’s purposes than anti-social behaviour, both because it’s more extreme and because it’s commonly carried out by state authorities rather than by next-door neighbours. But I think the use of torture as an example also points to a different argument about justice and moral wrongs. Consider the first sentence quoted above: “Reasons of justice are reasons for or against altering someone’s relative position.” Norms of justice, Gardner argues in the second extract, are appropriate for tackling those questions which we face “when people make competing claims to assignable goods”. There’s a fundamental concept of personhood lurking here: a person, we can infer, is someone whose position (however defined) can be measured relative to the positions of other people; someone who can successfully claim assignable goods; someone whose self-interested claims can compete with those of other people; and someone whose disputes with other people can be adjudicated, and whose relative position can be altered, through the process of law, in other words by applying public norms using socially recognised procedures. And – at the risk of sewing a shirt onto a button – a law-governed society is a society composed of such individuals; and when we say ‘law’, we mean the kind of law through which such a society, and such individuals, can govern themselves. Clearly, the terms Gardner used would not work well in a feudally-ordered society, or a society run along religiously-validated caste lines, or the small-c communist society which was to follow the withering-away of the socialist state. We are talking about a society composed of formally equal individuals, differently endowed with personal resources, but each capable of making claims to assignable goods; entitled to expect that those claims will be respected; and entitled to attempt to vindicate them through the law.
We can see how this model of personhood relates to an allocative model of justice by looking at some scenarios. If my neighbour encroaches on my back garden, I may sue him and let the courts adjudicate our competing claims to the assignable good behind my house. If he takes our dispute personally and steals my property or assaults me, justice is involved in a different sense. Restitution will certainly be required, bringing allocative justice into play; however, my neighbour is also transgressing in a more serious way, improving his relative position by socially disallowed means. Theft and personal violence can be seen as ways of gaining an unfair advantage or nobbling the competition. (Gardner also suggests that criminal justice is allocative in the sense that it turns on the correct allocation of the status of criminal, which seems valid if rather ingenious.)
What about if my neighbour gets his revenge by a more indirect route, swearing at me in the street or disturbing my rest with loud music (or jamming my BBC 4 signal)? In such a case, given that the good in question is non-assignable, justice in Gardner’s terms may not be involved. Even so, the courts are likely to take the view that my entitlement to a non-assignable good has been needlessly infringed. (Not that this is a simple proposition, as we can see if we remember Hohfeld. If I am entitled to quiet nights – and why should I not be? there is, in principle, no shortage – does this mean that I hold a privilege as against all my neighbours, with a correlative duty on each of their parts not to disturb my rest? Can this be generalised, to cover mutual obligations among neighbours and entitlements to other forms of domestic tranquillity? I think this would be very problematic. Make these duty/privilege relationships unwaivable and everyone involved would be encumbered with a vast array of duties to abstain from potentially disturbing behaviours. Make them waivable, on the other hand, and the effect would be to destroy the universality apparently offered by the discourse of rights: all we would do would be to translate different individuals’ widely varying levels of entitlement and grievance into the language of waived and unwaived rights.)
Setting these broader considerations aside, the main point here is that deliberate deprivation of a non-assignable good can be grasped in terms of (allocative) justice, essentially by assimilating it to the ‘unfair advantage’ model associated with criminal justice. Indeed, we could rework the ‘unfair advantage’ model itself in terms of the deprivation of a non-assignable good. Laws criminalising physical violence, for instance, can be seen as protecting the non-assignable good of bodily integrity. In terms of acquisitive crime, if individuals A, B and C are all planning to bid for a valuable object at an auction, but are prevented from doing so when I steal it, what I have deprived them of is precisely the non-assignable good of a fair competition. A similar argument could be developed for the theft of an article on sale, or (less directly) of something in private possession. (We can see here, incidentally, how far removed the principles of allocative justice are from any redistributive model of social justice; in allocative terms, mere ownership of a resource at a given point cannot be unjust. Allocative justice and social justice must always be in tension, this side of the revolution.)
The principle here is that the autonomous, self-interested individuals on which our legal model is predicated need – and hence are entitled to – certain non-allocative goods if they are to play their competitive, law-governed part in society. One such good is the rule of law itself; others are bodily integrity and property rights. We can extend this model of entitlement – and hence of rights which can be vindicated in the courts and disputes which can be adjudicated according to law – to other non-assignable goods, including the good of eight hours’ sleep or an evening in front of BBC 4. In practice, many non-assignable goods are difficult to deal with in this way, as witness the vagaries of anti-social behaviour legislation: the baseline entitlement to a non-assignable good (such as peace and quiet), the level to which others are responsible for upholding that entitlement and the degree to which offending behaviour infringes it are often hard to establish. However, this is not to say that relationships between one person’s anti-social behaviour and another’s unmerited suffering can never be established; in practice they very often can. My neighbour is not going to be able to fly under the law’s radar by making sure that all he deprives me of is the non-assignable good of a good night’s sleep – any more than if it were the non-assignable good of an unbroken nose.
But what is my neighbour doing in the (mercifully, highly unlikely) case that he tortures me? Here, I think, a different relationship between justice and personhood obtains. If we think of bodily integrity as a non-assignable good (and certainly your good health does nothing to impair mine), then the victim of torture has been deprived of a non-assignable good, and may be unable to play a full part in society as a result – but, as stated, this is no less true of the victim of a random assault at pub closing time. We can say that torture is more likely to have traumatic effects, and this seems significant: certainly if we think of other experiences which are likely to produce trauma (rape, battlefield stress, partner abuse) the word ‘torture’ is never far away. Torture, then, is one of the things that inflict trauma, in a way that a beating in the pub car park generally isn’t. But why is this a significant distinction? The point, I think, is that torture is an attack on my personhood. Personal violence can often be understood in terms of enhancing the attacker’s relative position by depriving the victim of a non-assignable good, making it harder for that person to play a role in society. Pace Gardner, the immorality of torture is not grounded in its depriving the victim of a non-assignable good. Torture is not about enhancing the torturer’s position relative to the victim, even with respect to the non-assignable good of freedom from pain. Torture – and other forms of traumatic assault – can be seen as an attack, not on the victim’s capacity to function in society, but on the victim’s basic recognition as a person who might be entitled to any such capacity. More simply put, causing pain for no reason is not something one person does to another; torture thus situates the victim as less than a person. It’s interesting, in passing, that Mill characterised rape in very similar terms – “the lowest degradation of a human being, that of being made the instrument of an animal function contrary to her inclination”. To commit rape, in this line of thinking, is not to deny someone the good of freedom from rape, but to deny her the status of a person entitled to freedom from rape (and entitled, as a second-order right, to live her life on the basis of an assumed freedom from rape).
I think Gardner’s distinction between the immorality of torture and the wrongs which can be understood in terms of allocative justice is valid and powerful, although not quite in the way that he uses it. What I think it points to is the ways in which people can be reduced to something below the status of personhood – through torture or brutalisation, but also through homelessness, institutionalisation or becoming a refugee – and the powerlessness of the language of justice to address these very basic, fundamental wrongs. If the law is about justice, and justice is defined in terms of the correct adjudication of competing claims among autonomous individuals, how can it address – how can it fail to overlook – those people who are shut out of the game entirely, by being denied the status of person in the first place? And if the law can’t be invoked, what can?
2. Did you read the trespass notices, did you keep off the grass?
A bit more Gardner, from the 2011 paper on tort law. It’s quite a complicated thought, so the quote has to be on the long side:
Let’s allow … that tort law often helps to constitute the correctively just solution. What doesn’t follow is that tort law’s norm of corrective justice should not be evaluated as an instrument. On the contrary, to fulfill its morally constitutive role, tort law’s norm of corrective justice must be evaluated as an instrument. It must be evaluated as an instrument of improved conformity with the very moral norm that it helps to constitute. To see why, think about some other laws that are supposed to lend more determinacy to counterpart moral norms.
Quite apart from the law, for example, one has a moral obligation not to drive one’s car dangerously. The law attempts to make this obligation more determinate by, for example, setting up traffic lights, road markings, and speed limits. If the law does this with sound judgment, the proper application of the relevant moral norm is changed in the process. A manoeuvre that would not count as dangerous driving apart from the legal force of the lane markings at the mouth of the Lincoln Tunnel may well count as dangerous driving – and hence a breach of the moral norm forbidding dangerous driving – once the lane markings are in place. But this holds only if the law proceeds with sound judgment. It holds only if relying on the lane markings assists those who rely on them to avoid violating the original moral norm. If the mouth of the Lincoln Tunnel has profoundly confusing lane markings, reliance on which only serves to make road accidents more likely, failing to observe the lane markings is not a legally constituted way of driving dangerously. It is not immoral under the ‘dangerous driving’ heading. That is because, if the lane markings are profoundly confusing, driving according to the lane markings does not and would not help to reduce the incidence of dangerous driving.
The lesson of the case is simple. A legal norm cannot play its partly constitutive role in relation to a moral norm unless it also has some instrumental role to play in relation to the same moral norm, unless conformity with the legal norm would help to secure conformity with the moral norm of which the legal norm is supposed to be partly constitutive.
We start with the “moral obligation not to drive one’s car dangerously”. Laws – embodied in road markings – are put in place to support this moral norm. In doing so they also constitute it, make it “more determinate”: if road markings are being generally observed, failing to observe them may amount to driving dangerously in and of itself. However, road markings – and laws – may defeat their own purpose. If road markings are so confusing that attempting to rely on them would make the driver more dangerous to other road users rather than less, failing to observe them will not amount to driving dangerously. Similarly a law may instantiate a moral norm, but do so in such a “profoundly confusing” way that someone attempting to observe the law will be more likely to violate the norm. If this is the case, anyone committed to observing the norm will be best advised to disregard the law which purports to embody it. “A legal norm cannot play its partly constitutive role in relation to a moral norm unless it also has some instrumental role to play in relation to the same moral norm”: road markings put in place to help prevent dangerous driving may themselves define dangerous driving, but only if observing them actually leads to less dangerous driving.
Three relationships between moral norms and the law are envisaged here. In one, the law embodies and gives substance to a moral norm. In the second, the “proper application” of the norm is redefined by reference to the law, leading to a changed perception of the norm itself. The third is identical to the second, except that in this scenario the “proper application” of the norm has been redefined to the point where the law does not assist observation of the norm, and may even impede it.
There’s a problem here, relating to that word ‘instrumental’. It seems to me that there’s something inherently problematic in judging the success or effectiveness of laws in consequentialist terms – in terms of the outcomes which they produce or appear to produce. Firstly, assuming that the moral norm to which a law relates can be straightforwardly identified, there is the question of what should be counted as success. Bad road markings, in Gardner’s image, are those for which “driving according to the lane markings does not and would not help to reduce the incidence of dangerous driving”. However, it is a commonplace of debates on sentencing that the criminal law can modify behaviour – both individually and at the level of society – in many different ways; what type(s) of behaviour modification should be counted as success is an open question. Is a law prohibiting practice X at its most effective if the incidence of X-ing is reduced to zero? Or is the effectiveness of the law to be judged by the appropriateness of the punishment dealt out to X-ers, or by the opportunity it gives the community to express their repugnance at X-ing, or by the degree to which it raises awareness of the plight of victims of X-ers? A case could be made out for any of these, not all of which can be reconciled easily or at all. Secondly, it’s not always clear that the moral norm underlying a law can in fact be readily identified, still less the body of moral norms underlying the law (or an area of the law, such as the criminal law or the law of tort). The point here is not that the law is necessarily obscure, but that it is necessarily multivocal: it’s always possible for different and competing claims to be made as to the underlying moral rationale of a law or laws. This in turn raises the question of who is to do the identifying – and whether what they identify can change over time. Suppose that an elected government, facing a long-term economic depression, declares that poverty is a higher priority than crime, and that the law should generally not be used to impoverish poor offenders further. Or suppose that an elected government, facing a rise in crime figures, declares that the chief menace facing the country today is lawless behaviour by immigrants, asylum seekers, Travellers and people of no fixed abode, and that wrongdoing by individuals with no stake in a local community should be treated more harshly. Would these programmatic announcements represent authoritative clarifications of the body of moral norms instantiated by the law, the criminal law in particular? Would we expect the judiciary to ‘read down’ legislation to ensure compliance with these policy stances? If not, why not?
As in the case of torture considered as deprivation of a non-assignable good, I think Gardner’s analogy here pulls in a different direction from his stated argument. Road markings modify behaviour in a distinctive way and in a distinctive context, neither of which maps easily onto the law in general. To drive a vehicle is to put others at risk and accept the risk imposed by others; driving safely rather than dangerously benefits both the driver in question and other road users, in a way which is true of few other ‘virtues’ in driving. In effect, driving safely is the solution to the key co-ordination problem posed by collective road use – and it is a simple, readily available and generally acknowledged solution. Moreover, road markings constitute the moral norm of driving safely in a peculiarly authoritative way, which is perhaps only possible because the norm itself is so generally agreed. Road markings do not typically take the form of recommendations or advice; even to call them instructions would understate the force they have in practice. Rather than advise (or instruct) a driver to make certain choices, road markings typically operate by excluding certain choices altogether: they do not influence behaviour so much as structure it. As such, road markings are not open to be technically observed or observed in spirit or ingeniously circumvented: they are observed or not. Both the moral norm underlying road markings and the criteria for their observance are self-evident, in a way that is seldom true of the law.
Are we committed to abandoning any ‘instrumental’ evaluation of the law, or of individual laws, by reference to their outcome? This conclusion would be unfortunate; not only would it necessitate abandoning Gardner’s insight on the reflexive relationship between laws and norms, it would make it impossible to say whether any law was making the world a better place. A narrower reading of Gardner’s analogy may provide a solution. The situation in which road markings are “profoundly confusing”, such that “reliance on [them] only serves to make road accidents more likely”, can be interpreted in a number of different ways. The implication could be that the road markings are so confusing that it is effectively impossible for any one driver to follow them. Alternatively, it could mean that the markings can be followed, but only at so great a cost in time and attention as to force the driver to disregard other road users, so that observing the markings made his or her driving more rather than less dangerous. Lastly, it could mean that the markings are confusing in the sense of allowing widely diverse readings; markings which could plausibly be followed in multiple different ways would not make any one person’s driving more dangerous, but would greatly increase the likelihood of accidents.
All these forms of confusion can be readily envisaged as flaws of badly-made laws or legal systems: the law so complex and confusing that it is impossible to observe; the law whose demands are so extensive as to make it hard to carry on the activity the law is intended to regulate; the law whose vague or contradictory wording causes more social conflicts than it resolves. Any one of these flaws will make a law less effective, either in guiding individual behaviour or in resolving co-ordination problems; as a result, the moral norm underlying the law will be less effectively constituted in social practice, or (at worst) not constituted at all. However, these are all formal flaws: the failure of the law to constitute a moral norm can be inferred from the failure of the law as law. The realisation of the moral norm underlying the law does not need to be measured as an outcome – indeed, it is probably better if this is not attempted, for the reasons given above.
What I draw from Gardner’s analogy, in short, is a restatement of the intimate connection between morality and the formal virtues of law. To say that a law or body of laws is coherent, comprehensible and followable is not simply to say that it is well-made. A well-made law is also one which is well suited to embody a moral norm – and, crucially, to refine and specify the proper application of the norm in social practice, playing “[a] partly constitutive role in relation to a moral norm”. Whether or not the formal virtues of law have any moral content in themselves, I think this argument suggests that there is at least an irreducible affinity between law and morality.