Responsibility and Reason-Responsiveness: Comments on John Gardner’s 2016 Chorley Lecture, ‘The Negligence Standard: Political, Not Metaphysical’, (2017) 80(1) MLR 1-21

Is it OK for the law to assign responsibilities to persons who lack the ability to respond to reasons? For John Gardner, the answer is 'normally, no'. Even when we do saddle those persons with responsibilities, we do so because we treat them, fictitiously, as if they are able to respond to reasons. Is that right?

Emmanuel Voyiakis

John Gardner argues that questions of responsibility in tort law are political, not metaphysical. By that, he means that we settle them by appealing not to metaphysical truths about agency and responsibility, but to a host of reasons that reflect the terms under which we choose to live with one another in a political community. The difference is important because the terms of our interaction are (in Gardner’s apt description) ‘pliable’ in a way that metaphysical truths are not. Consider liability for harm caused by defective products. Our tort law could say, as the common law does, that manufacturers of consumer products are liable to make repair only when they are negligent. Or it could say, as some modern regimes do, that manufacturers are liable whenever their product fails to provide the level of safety consumers may reasonably expect. Gardner’s point is that the choice between those regimes of liability does not turn on asking which fits better with our preferred metaphysical account of what constitutes a person and what makes a person responsible for some event or outcome. Metaphysical thoughts like these may constrain the range of acceptable tort liability regimes, and Gardner has important things to say on the matter, but they do not determine the choice between them. The case for making responsibility for harm caused by defective products negligence-based or stricter will instead be based on substantive arguments about the proper division of responsibilities between manufacturers and consumers, e.g. on the basis of their relative ability to avoid the accident, to bear the cost of repair, the distributive effects of different divisions of responsibility, and so on. Maybe neither term captures the nature of the reasoning involved in drawing up that division, but if I had to describe it as either political or metaphysical, I would not hesitate to share Gardner’s preference for the former term.

Gardner nevertheless thinks that there is a metaphysical constraint in the way tort law may assign responsibilities to people, and he traces that constraint to a particular sense of the idea of responsibility (at 7-11). It is the sense of responsibility as a composite of ‘response’ and ‘ability’; literally, the ability to respond. In the context of practical reason, the ability to respond becomes the ability to respond to reasons. Gardner calls this ability ‘basic responsibility’, because he regards it as the basis on which the law may assign responsibilities to a person. In the normal run of things, the law assigns responsibilities (duties, obligations, liabilities) to a person on the assumption that this person is ‘basically’ responsible, i.e. that they are able to respond to reasons. Gardner describes this normative relation as ‘the normal dependence of assignable responsibility upon basic responsibility’ (at 10). He argues that this relation obtains even when legal principles assign responsibilities to persons who lack the ability to respond to reasons, e.g. very young children or people with advanced dementia or other mental illnesses, inasmuch as the relevant principles treat those persons ‘as if’ they had that ability. As he puts it (at 12):

Assignable responsibility is sometimes assigned, rightly or wrongly, to people who are not basically responsible for whatever they are being assigned assignable responsibility for. However, it is assigned to nobody without at least the fiction of their basic responsibility for whatever they are being assigned the assignable responsibility for.

Having identified this metaphysical constraint, Gardner devotes the second half of the piece to explaining what the constraint does not do: it does not provide an argument in favour of making negligence the default standard of tort liability. The negligence standard, Gardner reminds us, is a reasonableness standard. It requires us to assess how well or badly a person did in responding to the reasons they had to take care. Basic responsibility is not like that at all. Debates about basic responsibility are a lot more straightforward in that they are conclusively settled once it is established that a person is able to respond to reasons. That responsibility is ‘confirmed … by the mere fact that the question of one’s negligence arose’ (at 16). It follows that those who try to defend the negligence standard by appealing to considerations of basic responsibility misunderstand both ideas. If basic responsibility is fundamentally strict, the justification of the negligence standard must rely on considerations of assignable responsibility alone.

Gardner’s discussion brims with characteristic sharpness, clarity, and elegance. One of the joys of the piece is seeing Gardner bring his thoughts about responsibility into closer alignment with his seminal thesis about the ‘continuity’ of reasons as the basis for the justification of duties of repair (3-6)—both themes are developed further in his more recent From Personal Life to Private Law (OUP, 2018). Another is the sheer breadth of moral insight and experience that he draws upon in developing his points. Textbook legal examples rub shoulders with Renoir’s La Regle du Jeu and Eliot’s Middlemarch in a way that shows lawyers and artists grappling with a common set of questions about our humanity and our moral agency.

My aim in this note is to raise some questions about Gardner’s central claim about the relationship between ‘basic’ and ‘assignable’ responsibility. I agree that the fact that a person lacks the ability to respond to reasons will normally be a conclusive ground for not assigning responsibilities to that person. Nevertheless, I want to suggest that the link between basic and assignable responsibility may not be as tight as Gardner suggests. My intuition is that there are situations where it is OK to assign responsibilities to a person who lacks the ability to respond to reasons, and doing that would not be treating that person ‘as if’ they had that ability. That is most clearly the case when the responsibilities in question are coupled with benefits that this person has reason to want to be able to receive despite their inability to respond to reasons.

Suppose that a person suffers from unresponsive wakefulness syndrome (UWS), a condition that older medicine referred to as a ‘persistent vegetative state’. UWS sufferers cannot respond to reasons. If Gardner is right, normally we may not assign responsibilities to those persons, unless we have good normative warrant to treat them, fictitiously, as if they were able to respond to reasons.

This makes a lot of sense as an account of our responses to normal cases. Say that, due to the onset of their condition, a UWS sufferer misses a delivery date. They may be liable for breach of contract, but that will be so because that liability had been assigned to them at a time when they were able to respond to reasons, namely the time they entered the contract. By the same token, as long as they remain unable to respond to reasons, we may not assign any new contractual liabilities to them. The same goes for tortious liabilities. Say a UWS sufferer owns a property with a water deposit in it. Soon after their condition kicks in, the deposit bursts and floods their neighbour’s land. It seems to me that the UWS sufferer would be required to bear the burden of repair under Rylands v Fletcher [1868] UKHL 1, even though they were unable to respond to reasons when the accident occurred, because that responsibility was assigned to them at the point they did have that ability, namely when they erected the deposit in their land. They would not be responsible if the deposit had been erected by someone else after their UWS had kicked in.

This much goes for responsibilities. Consider now whether it goes for benefits too. Gardner does not take up this question, but it seems clear that the answer should be no. The ability to respond to reasons is not a necessary condition for a person to be entitled to benefits. Others may benefit a UWS sufferer in many of the ways they might benefit a person who can respond to reasons. For example, they may bequeath them some property. We would not say that the bequest is invalid because the beneficiary is unable to respond to reasons.

But now suppose that we put responsibilities and benefits together. Say that the UWS sufferer inherits a large property that contains a water deposit. The deposit breaks and floods the neighbour’s land. Could the sufferer be liable, given that they were unable to respond to reasons both at the time title passed to them and the time the accident occurred? One might say that we need to distinguish benefits from burdens here, assign the benefit of ownership to the UWS sufferer, and keep the burdens with the estate. This might be a sweet deal for the UWS sufferer, but it will be a terrible deal for the estate, not only because the testator might not have been aware of the sufferer’s condition but also because, once title has passed, the estate will have lost control of the property (I suspect that if we adopted that rule, fewer people would be willing to name UWS sufferers as beneficiaries). So, given the good substantive objections of the testator and their estate against splitting benefits and burdens, let us grant that the only two options on the table will be these: either the UWS sufferer gets both the benefits and the burdens, or they get neither.

I find this a hard dilemma. Sure enough, I see some force in the view that being assigned responsibilities while one is unable to respond to reasons is morally unpalatable whatever the attendant benefits, and if being free of those responsibilities requires one to forego the benefits, so be it. But my own intuition is that we should make space for less categorical views. For example, it is plausible to think that we cannot address the dilemma usefully until we know more about the likely scale of the respective responsibilities and benefits. Call me a drooling middle-class fantasist, but if the Duke of Westminster plans to make me sole heir to his property portfolio just as long as I undertake its upkeep and the other responsibilities of ownership, I would not want to miss out just because I am suffering from UWS. To generalise, perhaps it might be OK to assign responsibilities to a person who is unable to respond to reasons as long as those responsibilities are small in scale (or that, if they are considerable, their incidence is likely to be rare) and, by that person’s lights, they are clearly outweighed by the benefits. Maybe in some cases it must also be the case that the responsibilities can be adequately met through the benefits, e.g. any rental income that the property generates, so that the beneficiary need not go out of pocket, and so on.

These judgments will obviously have lots of moving parts, and some of them may turn out to be wrong, e.g. because some responsibilities are such that no attendant benefits will outweigh them. But they are plausible judgments, and I am not sure that Gardner’s account can explain what makes them so. On the face of it, those judgments run counter to Gardner’s main thesis because they raise the possibility that sometimes a person can have reason to want certain responsibilities assigned to them despite the fact that they are unable to respond to reasons. In Gardner’s terms, they suggest that assignable responsibility (here, responsibility for repair of the damage caused by the water deposit) does not always depend upon basic responsibility.

I suspect that Gardner does not want to deny that the UWS sufferer who inherits from the Duke of Westminster should be liable under Rylands if the water deposit in one of the properties bursts and floods the neighbour’s land. For one thing, his account does not commit him to denying this. When he describes the normative relationship between basic and assignable responsibility as one of ‘normal’ dependence of the latter on the former, he allows that there may be abnormal cases, where things may differ. The inheritance example may simply be such an abnormal case, for which we have to make special allowance, treating the UWS sufferer, fictitiously, as if they are able to respond to reasons.

I hope that in future work Gardner will say more about how this change of tack from normal to special cases works. My impression is that the inheritance example is not normatively special at all. It is simply a more complex version of a question we face all the time, namely whether assigning a given responsibility to a particular person would be too burdensome for them, in the light of the normatively significant features of their situation. Sure enough, we can expect that when a person is unable to respond to reasons, we have reason not to assign responsibilities to them. The ability to respond to reasons is essential to our ability to make choices and determine the course of our life, and we have good reason to want our responsibilities to others to depend, at least in part, on our choices and our other responses to the reasons we have. The inheritance example does not run counter to this idea, it refines it. The upshot of that example, as I see it, is that while reason-responsiveness is central to responsibility, we should take care not to exaggerate its significance. A person may be unable to respond to reasons and still have reason to want certain responsibilities assigned to them. And sometimes those reasons will win the day, making it OK to assign those responsibilities to that person.

Moreover, I am not sure that, in holding the UWS sufferer liable under Rylands, we would be treating them ‘as if’ they could respond to reasons. I suppose the ‘as if’ fiction signifies that while the sufferer’s condition has several normative consequences, it need not have this or that particular normative consequence, e.g. that it does not disqualify that person from being entitled to certain benefits, or shield them from certain responsibilities. But suppose we conclude that the UWS sufferer is indeed entitled to certain benefits or ought to bear certain responsibilities. It would still not be right to say that we are treating this person as if they could respond to reasons. The ground of our conclusion would be narrower, namely that this person’s condition is not a reason for denying them the benefit, or for shielding them from the responsibility. In reaching that conclusion, we would not ask what would have happened if that person had an ability they lack, or treat them as if they have it. We would simply determine that we may justifiably assign the relevant benefits or responsibilities to someone with the disabilities that this person actually has.

Reason-responsiveness is an essential aspect of our agency. It is also central to our ideas about responsibility, and Gardner has done more than anyone in the field of private law to illustrate its significance on that score. However, that significance may not as basic as Gardner suggests. After all, our agency survives our inability to respond to reasons. That inability is clearly not a block to us acquiring new entitlements towards others. Why would it be a total block to us acquiring new responsibilities towards them?


Published 12.06.18
Response by Emmanuel Voyiakis