I. INTRODUCTION
In a recent article in this journal,Footnote 1 I argued that there is an important distinction between two types of claim in private law. Sometimes the law requires D to act in a certain way for C's benefit, and C's claim against D arises from D's breach of duty or wrong against C; and in other cases the law specifies a contingency for which D has responsibility vis-à-vis C, so that a claim arises for C against D in the event that the contingency materializes, though D has no duty to prevent the contingency from materializing and so can become subject to C's claim despite not having committed a wrong. Thus one can distinguish between claims arising from a primary duty and claims arising from what I described as a “primary liability.” In this article I attempt to show that this distinction is important across private law and not just in the unusual circumstances of Vincent v. Lake Erie,Footnote 2 on which I concentrated in the previous article. I also defend the view that the primary-liability claim cannot be represented under Wesley Hohfeld's scheme of legal relations. However, although I suggest that Hohfeld's scheme may be incomplete in this respect, my argument is not anti-Hohfeldian in spirit, and I also make some comments on the role of Hohfeldian analysis in private law.
II. CLAIMS ARISING FROM A PRIMARY DUTY AND THE CHARACTER OF THE REMEDIAL RELATION
It is helpful to begin not with examples of what I described as primary right-liability relations but with the ordinary case of the primary right-duty relation and the remedial relation that arises from it. According to the standard analysis of claims in private law, the primary relation is a right-duty relation. D has a duty to C—for example, a duty of care in tort or a contractual duty of performance—and C has a correlative right. The right here is a Hohfeldian “claim-right” or right in the strict sense.Footnote 3 If D breaches the primary duty, a secondary or remedial relation arises. This is also conventionally understood to be a right-duty relation, C having a claim-right to compensation, and D a duty to pay it.
Sometimes the remedial relation will be a debt for a certain sum, and a debt is always characterized in the same way as a right to payment correlated with a duty of payment.Footnote 4 In my view, this is a mischaracterization of the remedial relation or of a debt.Footnote 5 Consider the case where D, the debtor or party subject to the claim, is dead. He may have died in the act of committing the breach of the primary duty that gave rise to the claim. I take it that a dead person cannot bear a duty, and yet C's right to compensation nevertheless arises from the time of the breach and persists despite D's death. Once an executor or administrator of D's estate has been appointed, one can say that C's right is correlated with a duty on his part or on the part of the estate, but this does not explain the position before then. It is no more plausible to say that the estate bears a duty before anyone has been appointed who is capable of performing it on its behalf than it is to say that the dead person does. It has been suggested that this means that C has a right to which there is no correlate.Footnote 6 But the answer, it seems to me, is that the remedial relation is not a right-duty relation at all but a power-liability relation, and D's estate can be subject to a liability after D's death without any need for someone who can act on behalf of the estate.
Consider also the case where D, though alive, is incapable for some reason of taking steps to make the payment, say because his wealth is temporarily inaccessible or he is ill.Footnote 7 Does he then have a duty of payment although he is incapable of performing it? And what about the case where he does not know the facts that would reveal that he has such a duty and could not reasonably be expected to know them?Footnote 8 If he has a duty, it is a strict-liability duty, that is to say, a duty that can be breached without fault. Strict-liability duties are controversial, and I discuss them below. Some would say that this situation demonstrates that they are sometimes justified.Footnote 9 In my view, again, the better explanation is that the debt is a power-liability relation, which can subsist irrespective of whether D has a duty of payment, so that there is no need to rely on the idea of a strict-liability duty.Footnote 10
Of course, this is not to say that there cannot be a duty to pay a debt (or satisfy a remedial liability). The point is that there is a distinction between the debt itself and a duty to pay it. This is why one can distinguish, as the courts sometimes do, between the simple nonpayment of a debt, which is a state of affairs subsisting from the time when the debt accrues due, and the wrong of nonpayment of the debt. A debtor might commit this wrong when he knows or ought to know of the debt and has the means to pay it but still fails to pay. It is possible that he could be liable in respect of such a wrong not only in the measure of the debt itself but also for consequential loss arising from the nonpayment, and it is also possible that he could be punished for the wrong, but neither of these responses should be available where the proceedings are for the recovery of the debt as such. This analysis of the remedial relation is consistent with Hohfeld's scheme, although, as I say above, it is not the usual Hohfeldian analysis of a debt or remedial liability.
On my understanding, in ordinary legal usage—in the case law and doctrinal literature—the expression “claim” is used to refer to a remedial right, usually but not always a remedial right to compensation.Footnote 11 C has a claim once there is a cause of action, that is to say, once D has committed the breach of duty. This is the sense in which I use the expression here. The expression “claim” in this sense cannot be equated with the Hohfeldian “claim-right” or right correlated with a duty. A claim is a type of right but, as argued above, it is in the ordinary case a Hohfeldian power and not a Hohfeldian claim-right; and also, of course, a claim-right may be a primary right, whereas a claim is a remedial right.
It is worth briefly saying something about the nature of a power, since this raises issues that will be relevant below. The exercise of a legal power changes the legal position, but in my view someone has a legal power only when the rationale of the rule by virtue of which his act has the effect of changing the legal position is to enable someone in his circumstances to do this by acting in this way.Footnote 12 In other words, the law confers a power in order to enable the power-holder to decide whether to change the legal position in a certain way, and when he exercises the power, he acts with a view to making such a change. This is the position where, as a matter of ordinary legal usage, someone is said to exercise a legal power, for example, when a public official exercises a statutory power, or in private law when a party makes a contract, waives a claim or other right, transfers property or licenses its use, makes a will, or, acting as a trustee, transfers property to a beneficiary under a trust. This is the sense in which, in my view, a remedial right or claim is a power.
On another view, someone exercises a legal power whenever she performs a “volitional act” that changes the legal position.Footnote 13 The exercise of a power in the preferred sense above is a volitional act, but a volitional act need not be the exercise of a power in that sense. For example, on this volitional-act approach, when D commits a breach of a primary duty owed to C, say by breaching a contractual duty or committing a trespass or an assault, she thereby exercises a power to alter the legal relation between herself and C, and this is what generates the remedial relation.Footnote 14 These are volitional acts but they are not, it seems to me, correctly understood as exercises of a power. There is in any case no need to postulate the exercise of a power in order to account for the creation of the remedial relation in these cases. When D commits a breach of duty, it is simply the breach itself that gives rise to the remedial relation.Footnote 15
Of course, this raises the issue of how we should determine the correct understanding of a power in Hohfeld or more generally. It has been said that the concepts used by Hohfeld in his scheme are defined by stipulation, and any criticism of Hohfeld or his scheme based on a different understanding of them is simply misconceived: one can engage with Hohfeld only on his own terms, by adopting his stipulative definitions.Footnote 16 It is also suggested that the broader volitional-act understanding of a power is how the concept is understood by Hohfeld, and it is true that Hohfeld uses this expression to describe the exercise of a power, though the examples of powers that he actually gives are all powers in what I say is the true sense.Footnote 17 In any case, it is not in the end important how Hohfeld himself understands the concept. To the contrary, it does an injustice to Hohfeld to treat his scheme as exempt from criticism except on the ground of internal inconsistency in the light of the definitions stipulated by him for the concept of a power and the other basic concepts. The scheme is supposed to serve as a framework for understanding legal relations and formulating propositions of law, and it must be open to criticism on the ground that it disregards significant moral distinctions that the law should be capable of reflecting. Where there is a divergence of views, the better definition of the concept of a power or some other basic concept is the one that better captures the relevant moral distinctions so that it lends itself to the formulation of sound moral and legal propositions. The definition of a power suggested above is superior because there is an important moral distinction between the case where someone can change the legal position by choosing to do so, and the case where his act has this effect for other reasons. The same issue arises with respect to the concept of duty, which I come back to below.
In the discussion above, I assume that a claim arises from the breach of a primary duty. Now I move on to discuss some examples of claims arising from what I describe as a primary liability. In these cases, in my view, the form of the remedial relation is the same, that is to say, the claim is a Hohfeldian power correlated with a remedial liability; but, I argue, the claim does not arise from the breach of a primary duty in accordance with the standard pattern but from a primary liability, meaning that D bears the risk of a contingency as against C. This is not the Hohfeldian sense of “liability” by which a liability is necessarily correlated with a power (though I come back to this below). In the next section, I explain how the distinction between primary duties and primary liabilities is helpful in addressing the controversy over strict-liability claims in private law and particularly in tort.
III. PRIMARY-LIABILITY CLAIMS AND STRICT LIABILITY
A strict-liability claim is generally understood to be a claim arising irrespective of moral fault. Many writers have taken the view that strict liability is wrong in principle. The argument begins with what has been described as the “choice theory of responsibility,” or the “control principle,” the theory that moral fault or blameworthiness depends in the end on the choices or decisions of an agent about how to act.Footnote 18 Broadly speaking, on this approach, an agent can be morally at fault in respect of harm only if the harm was the consequence of a decision by the agent to act and the agent intended or foresaw the harm or the risk of it. This means that the agent cannot be morally at fault for failing to do an act that she is incapable of doing, and it means that moral fault should be understood and assessed from an internal or subjective perspective, the perspective of the agent, that is to say, relative to the agent's own abilities and circumstances, and in the light of what was known and understood by her.Footnote 19 Thus, on the choice theory, moral fault is to be equated with subjective fault.
This subjective approach to moral fault is widely accepted as appropriate in the criminal law for determining whether D should be punished and with what severity,Footnote 20 on the basis that it is justifiable for the law to take measures against D in respect of an act and its consequences only to the extent that D is morally responsible for them according to the choice theory. It seems to follow that liability to pay compensation in tort law should also depend on subjective fault, and some commentators do indeed reject strict liability in tort law for this reason. As Peter Cane puts it (though it is not his view): “In terms of the . . . choice theory of responsibility, liability without fault (. . . ‘strict liability’ . . .) lacks moral foundation.”Footnote 21
However, in tort law there are very well-established strict-liability claims, and many commentators who object to strict liability in criminal law do not regard strict liability in tort law as raising an issue of fairness in the same way. A good example is the famous rule in Rylands v. Fletcher,Footnote 22 which holds that loss caused to C by the escape of a potentially harmful thing held on his neighbor D's land, such as a dangerous animal or water in a reservoir, gives rise to a remedial liability of D to compensate C, correlated with a claim for C (strictly speaking, in my view, a power, as discussed above). D is not necessarily morally at fault according to the choice theory, since there may be nothing that he failed to do that he could have done to prevent the loss, once he had decided to have the dangerous thing on his land, and this he was entitled to do. Similarly, excluding strict liability would appear to rule out the doctrine of vicarious liability, which makes an employer liable to compensate for loss caused by acts of his employees. The law of negligence is normally said to be fault-based, because liability arises from the failure to meet a standard of care rather than simply from the infliction of loss or harm in specified circumstances.Footnote 23 But the standard of care is defined as the standard to be expected of a reasonable person, which is an objective standard. It does not adopt D's perspective or take account of D's weaknesses, and the standard may be beyond D's capability. Thus fault in this sense is not subjective fault, or moral fault according to the choice theory.Footnote 24
Some writers argue that tort law recognizes a wider legal concept of fault, so that it is possible to be at fault according to the law when one is not morally at fault. One can be legally at fault for failing to act as a reasonable person would act and even for failing to do what one could not have done. Legal fault is based on an objective determination of the balance of interests as between C and D and takes no special account of D's subjective perspective.Footnote 25 However, although it may in the end be justified to adopt an evenhanded approach between parties in private law that does not take account of D's particular weaknesses, and although it may be quite reasonable to use the word “fault” to refer to conduct that falls short of an objective standard laid down by the law, it is nevertheless unclear how this approach manages to overcome the objection based on the choice theory: that moral fault or blameworthiness, or in other words, subjective fault, should be required for the imposition of measures against D by the law.Footnote 26 Possibly the assumption may be that legal fault can play the role in law of moral fault in morality; but the point is that the law must be morally defensible and, if the argument from the choice theory is sound, it is subjective fault that is required for legal liability to be morally defensible.
Does this mean, then, that if we retain strict-liability claims, we must reject the choice theory; or conversely, that if we accept the choice theory, we must reject strict-liability claims? Or can strict-liability claims be reconciled in some way with the choice theory? As mentioned, some writers reject strict-liability claims because of their incompatibility with the choice theory. By contrast, other accounts appear to defend strict-liability claims by explicitly rejecting the choice theory of responsibility, including the implication of the choice theory that someone cannot have a duty that he is incapable of fulfilling.Footnote 27
At this point, it is helpful to return to the distinction between the two types of claim made at the start of the article: (1) where a claim arises because the law imposes a duty on D to act in a certain way for the benefit of C and she fails to do so; and (2) where a claim arises because the law stipulates a certain contingency for which D bears responsibility vis-à-vis C, so that its occurrence will generate a claim for C against D, though D has no duty to prevent the occurrence and may not be able to do so. In the first case, a claim arises by virtue of a primary duty, and in the second case a claim arises by virtue of a primary liability. The choice theory of responsibility requires that a duty be understood in the subjective sense, so that subjective fault is a condition of a claim arising from a breach of duty. This means that the content of a duty owed by D to C must take account of D's own abilities and circumstances, and there is no room for strict-liability duties, including duties to meet an objective standard of conduct.Footnote 28 But this concerns only claims arising from a primary duty. It does not imply that a claim can arise only from a breach of duty, and so it says nothing about primary-liability claims. A claim under the rule in Rylands or under the doctrine of vicarious liability arises from a primary-liability relation that allocates to D the risk of loss to C without imposing on him a duty to prevent it. The law of negligence, applying an objective standard of care, can be understood to allocate to D the risk of loss to C caused by D's failure to reach the standard, without imposing on her a duty to attain it. The recognition of primary-liability claims does not involve abandoning the subjective understanding of duty, and primary-liability claims are not in this sense against the choice theory of responsibility. Thus we can accept the possibility of strict-liability claims without rejecting the choice theory, provided they are understood as primary-liability claims, not primary-duty claims.
A critic of strict liability might be inclined to dismiss this argument and insist that the moral issue is simply whether claims arising without subjective fault are morally objectionable and that this cannot depend on what might appear to be a technical distinction concerning the proper way to formulate the law. However, this glosses over the nature of the moral objection to strict liability. If the objection is that an agent should not be hold to have acted wrongfully in the absence of subjective fault, the distinction is crucial, since a claim arising from a primary liability is not open to this objection. The choice theory may, however, impose other constraints on claims arising from a primary liability. It may, for example, require that D should be able to find out when he may be subject to a primary liability, so that he can take steps to avoid the possibility of a claim or reduce its likelihood, or insure against it. This may introduce subjective considerations, but it does not involve a requirement of subjective fault on D's part for a claim to arise. In private law, the legal position is routinely expressed in accordance with the standard pattern above, under which a claim always arises from the breach of a primary duty, and the distinction at the heart of tort law, between fault-based claims and strict liability claims, is taken to be a distinction between two types of duty-based claim. It is the characterization of strict liability claims as claims arising from the breach of a strict liability duty that brings them into conflict with the choice theory and makes them contentious.
The distinction between right-liability relations and right-duty relations is a distinction of the same sort as that between right-duty relations and power-liability relations or liberty-no-right relations. It is a matter of what might be called the “form” or “modality” of legal relations. It is, in other words, a distinction of the sort that Hohfeld is concerned with in his scheme, though the right-liability relation as I describe it does not appear in the scheme. As argued above in connection with the concept of a power, moral argument is relevant to the understanding of the basic concepts in Hohfeld's scheme. In particular, whether strict-liability duties are objectionable is relevant to the understanding of the concept of duty in a classification of legal relations according to form or modality. But in arguing for the significance of the distinction between right-liability relations and right-duty relations, I am not advancing a substantive theory to justify the availability of claims based on primary duties or primary liabilities or to show when they should arise. I am merely pointing to something that has to been taken into account by any such substantive theory, including a theory of strict-liability claims in tort law. It seems to me that the discussion of strict-liability claims in the literature is often flawed for failing to do this. For example, as mentioned above, there is no reason why the choice theory has to be abandoned to make room for an objective standard in tort.Footnote 29
Consider also Tony Honoré's much-cited discussion of strict liability.Footnote 30 As I understand it, Honoré's argument has two distinct but related parts. In one part of the argument, Honoré says that “we are responsible for what we do, including what we do without intending to or without foreseeing the consequences.”Footnote 31 This is the essence of his concept of “outcome responsibility.” In support of it, he argues that the consequences of an agent's acts are central to his identity, whether or not intended or foreseen. Each of us would have no “continuing history or character” unless outcome responsibility is ascribed to us in this way.Footnote 32 This part of Honoré's argument implies that an agent should be understood to have or be capable of having a duty to avoid acts with harmful consequences, even when the harm is not foreseeable. Under the choice theory, the association Honoré makes between the identity of the agent or his history or character and the consequences of his acts would apply only to intended or foreseen consequences. Thus here Honoré's approach is directly contrary to the choice theory.
In a related but distinct part of his argument,Footnote 33 Honoré points out that it is an inevitable part of life that we suffer the bad consequences of our acts just as we benefit from the good consequences, irrespective of whether these consequences were foreseen or intended. We have to “take the rough with the smooth.”Footnote 34 Furthermore, we are “entitled to insist” that the legal system should operate a regime that in the same way imposes liability on each person for the losses caused by his acts to others.Footnote 35 This is, as Honoré says, a risk-allocation argument. This risk-allocation argument, in conjunction with the argument that an agent is capable of bearing responsibility for the unforeseen consequences of his acts, is offered as a justification for strict liability. But the point for present purposes is that although, presented in this way, the risk-allocation argument appears to draw on the concept of a strict-liability duty,Footnote 36 it would be better to understand it in terms of primary liabilities. It does not require that people should be held to have acted wrongfully without fault; in this respect, the argument is consistent with the choice theory. As suggested above, however, the choice theory may require that the conditions for a primary-liability claim to arise should be such that it is possible to plan for them,Footnote 37 and Honoré's approach may be open to criticism on this basis.Footnote 38
IV. “REMEDIALISM” AND “DUTY SKEPTICISM”
On one understanding of private law, it is always appropriately expressed without any reference to a primary relation, for example simply in the form: “If harm has resulted to C from D's doing X, then C has a claim against D for compensation.” This might be described as a “remedialist” approach. On this understanding, it adds nothing to say that D had a primary duty to C not to do X, and there is plainly no point in making a distinction between two types of primary relation along the lines suggested above.
One idea behind this approach is simply that it is a straightforward way of looking at the law from the standpoint of someone who is interested only in how the law may be employed against him (or for that matter employed by him for his own benefit), an idea that was famously captured by Oliver Wendell Holmes in his “bad man” theory of law.Footnote 39 From this standpoint, the issue is generally whether the occurrence of certain events has given rise to a claim for compensation, and it seems unnecessary to ask whether this is by virtue of a primary right-duty or a primary right-liability relation. As a matter of fact, even from this standpoint, the distinction can be important if some other remedy is at stake, such as an injunction requiring the defendant to perform her duty, as considered below.
The remedialist approach implies not so much that there is no primary duty but that from a certain perspective it is not important whether this is the case or not. However, remedialism is naturally associated with the complete rejection of primary duties in private law. Some commentators have distinguished between “duty-accepting” and “duty-skeptical” theories as competing theories of tort law or private law in general.Footnote 40 On the “duty-accepting” approach, tort law imposes duties on one person (or class of people) for the benefit of another person (or class). A claim always arises from the breach of a duty owed by D to C, that is to say a wrong by D against C. Thus tort law creates bilateral relations.Footnote 41 These bilateral relations are readily understood to be based on or to reflect principles of interpersonal morality, such as the moral principle that one should not assault others or, more generally, that one should take reasonable care not to cause harm to others. This seems to be broadly the traditional approach,Footnote 42 and it is presupposed in the standard Hohfeldian analysis of private law.
According to the “duty-skeptical” approach, tort law, properly understood, is not based on duties owed by D to C, and a claim does not arise from a wrong committed by D against C. Instead, it is concerned with imposing sanctions on D as a means of minimizing aggregate loss in pursuance of efficiency or utility or with administering payments from D to C to secure compensation for C as part of the redistribution of losses across the community in pursuance of a scheme of distributive justice. On this understanding, tort law does not give effect to or reflect principles of interpersonal morality. It is fundamentally a matter of public policy rather than private justice, and the rationale for allowing C to make a claim is to harness his pursuit of his own interest in compensation to the implementation of public policy.Footnote 43 Although the law may be conventionally formulated in terms of a duty owed by D to C, on this approach this is not a genuine duty but a fiction, or at least a device for signifying a condition for a claim to arise and the person in whose favor it arises.Footnote 44 A more apt formulation might be the remedialist one.Footnote 45
The argument for primary right-liability relations might be thought to be a version of remedialism or duty skepticism, and it is true that proponents of a duty-skeptical approach often describe the law as liability-based rather than duty-based.Footnote 46 But the primary right-liability relation should not be understood in this way. Recognizing the possibility of primary liabilities is not inimical to the recognition in principle of primary duties or bilateral relations. The primary right-liability relation is a bilateral relation, different from the primary right-duty relation but playing an equivalent role in private law; like the primary right-duty relation, it subsists as a matter of justice between C and D, but instead of representing a requirement for D to act for the benefit of C, it represents an allocation of responsibility for a contingency to D vis-à-vis C. The suggested approach is a version of the traditional approach based on the assumption that private law is properly formulated in terms of bilateral relations, but with the concept of a bilateral relation extended to include right-liability relations as well as right-duty relations.
There is some support for remedialist or duty-skeptical approaches in the common law,Footnote 47 but for the most part the language and the bilateral structure of private law support the traditional approach, and it gains support also from the obvious plausibility of the link between private law and principles of interpersonal morality. However, if it recognizes only primary right-duty relations and not primary right-liability relations, as discussed above, the traditional approach has difficulty explaining strict-liability claims.
V. PRIMARY RIGHTS, REMEDIAL RIGHTS AND CLAIMS IN CONTRACT
It is always said that contract involves strict-liability duties: D can have a contractual duty to achieve a result that proves to be impossible (and may indeed always have been impossible), and she can be held to have committed a breach of duty for failing to achieve it.Footnote 48 It has been said that even if strict-liability duties are problematic elsewhere, there can be no objection to them in contract because the contracting party has chosen to accept such a duty.Footnote 49 But why would anyone choose to accept a duty that might prove impossible to perform, so that she might end up acting wrongfully without any fault? No one should be expected to do that. This is the last place one would expect to see a strict-liability duty. It would be better to say that contracting parties accept that they may incur remedial liabilities in certain circumstances despite not having breached any duty. This does not mean that they choose to incur a liability but that they choose to accept the risk that they may do so. On this understanding, the parties' agreement serves to allocate the risks arising from their transaction without imposing a duty of performance, and this gives rise to a primary right-liability relation.
Furthermore, there is a problem in explaining the position in contract even if we accept the possibility of strict-liability duties. This point concerns the relationship between primary and remedial relations. In principle it applies in tort as well as contract, but it is more apparent in connection with contract. In the usual case in tort, it makes no difference to the nature of the remedy whether the claim is a primary-liability claim or a primary-duty claim; in either case the remedy is compensation. But consider the availability of specific performance in contract. Take an ordinary contract for the supply of goods. If the defendant fails to supply the goods, the court will not order him to do so (even if the claimant wants such an order). It is invariably said that the defendant D has a duty to supply the goods, but specific performance is not available as a remedy for the breach of duty. C is confined to a claim for compensation. This seems to me difficult to make sense of. Of course there are often cases, in tort more commonly than contract, where, once D has committed a breach of duty, it is impossible or excessively burdensome for him to perform the original duty or even a variant of it, and he is not ordered to do so—for example, where he has caused personal injury or property damage through an accident. But in this case of a supply contract, the circumstances when the order is denied may be exactly as they were when the original breach occurred. What reason can there be to deny an order to perform in such a case, if there was indeed originally a duty to perform? Is the implication not that D actually has a liberty not to perform? This would mean that the claim for compensation arises from a primary liability.
It is often said that D has a primary duty to perform the contract but that he should not be compelled to perform by an order of specific performance, since compulsion is a different matter from the existence of the duty. In other words, there is a case of what is sometimes described as a “right to do wrong.” D does not have a legal liberty not to perform—she acts wrongfully by not performing—but she is entitled not to perform in the sense that she has a right not to be subject to compulsion, though she is liable to pay compensation.Footnote 50 But this is not an apt description of the legal position. It may be that there is a moral duty to perform and also a moral objection to compelling D to perform. For example, it might be said that although D has a moral duty to perform, compulsion to perform cannot be justified because it offends against the “harm principle”;Footnote 51 or it might be said that although D has a moral duty to perform, it is against the public interest for contacting parties to be compelled to perform, because this prevents them from taking up other more lucrative alternatives (subject to the payment of damages), and it is on the whole for the public benefit for this to happen.Footnote 52 But if this is the position, it would be better to say that there is no legal duty to perform. There is no reason to equate the primary duty in law with a moral duty free of considerations concerning compulsion; the justifiability of compulsion is part of what determines whether there should be a legal duty.
Possibly one might argue that if the reason why D is not compelled to perform is not an objection of principle but simply the practical difficulty involved in compelling performance, then it is still appropriate to say that D had a primary duty in law though he is never compelled to perform. But even if this is right—which seems to me open to doubt—it is not the type of objection generally leveled against the availability of specific performance, as the arguments mentioned above indicate. Furthermore, on the interpretation suggested above—that in making a contract, the contracting parties choose to allocate the risks of their transaction without generally accepting duties of performance—the implication is that there is not generally any moral duty to perform either, and the absence of compulsion follows naturally from what was agreed between the parties.Footnote 53
Thus in contract it would be better to say that generally specific performance should not be granted because D has no legal duty to perform, not that specific performance is unavailable as a remedy for D's breach of duty. In other words, the primary relation involves a primary liability, an allocation of risk with respect to performance, rather than a duty to perform, and D has a liberty not to perform. The reason why it is thought that there must be a primary duty is that it is assumed that a claim for compensation must arise from a breach of duty.
In tort, as mentioned above, an injunction to prevent D from causing harm to C is usually not in issue, because usually the claim is for personal injury or property damage that has been inflicted and cannot now be prevented or undone.Footnote 54 But in principle there could be a case, just as in contract, where C has a claim against D for compensation for harm, but C cannot have D enjoined from acting in the way that caused the harm, even when an injunction is entirely practicable. A possible case might be where D is required to compensate her neighbor C for loss caused by an activity carried out by D on her property, but D is permitted to continue the activity.Footnote 55 Here again, the claim cannot be understood as a claim arising from a breach of duty.
I do not consider here why there is usually no primary duty in contract, or when or why there is sometimes a primary duty, or, for that matter, why or when there should be primary duties or primary liabilities in tort. As mentioned above, these questions are not the subject of the present discussion. Hohfeldian analysis is about the form or modality of legal relations, and my argument at this point is that it is necessary to appreciate the distinction between primary duties and primary liabilities in order to understand fully the relationship between primary relations and remedial relations.
VI. CLAIMS ARISING OUT OF THE OWNERSHIP OF PROPERTY
The Unauthorized Use of Property
Where D uses C's property without permission, C has a claim for reasonable payment, the measure of which is the sum that might reasonably have been agreed as a license fee, typically a proportion of D's benefit, exceeding any loss to C. This claim is well established in the case law, and I refer to it as the unauthorized-use claim. Assuming that the claim is justified, the question is: What is the primary relation from which it arises? In a case following the standard pattern, where D breached a primary duty, the standard remedy is compensation in the measure of C's loss. This is because the function of the remedy is to rectify the wrong, defined in terms of the primary relation. Thus the remedy should put C in the position he would have been in if the primary duty had been performed. This is not to say that logic dictates this outcome. It would not be logically inconsistent to say that the measure of compensation should be less than this because the full measure is excessively burdensome (though this argument has not been influential in practice). But it would surely be inconsistent for the measure of recovery to exceed the loss caused by a wrong if it is designed as a remedy for the wrong, though C might be allowed a greater measure for some other reason—for example, to deter wrongdoers or to remove a profit wrongfully made by D.Footnote 56
In the case of the claim for unauthorized use, if the claim arises from a primary duty not to use the property, as is usually assumed, the measure of compensation should be C's loss resulting from the unauthorized use. But the deemed license-fee measure will generally exceed any loss suffered by C, and indeed C may have suffered no loss at all. I cannot see how this measure can be understood as a remedy for the breach of a primary duty not to use the property. The measure is explicable, however, if the primary relation is what I call a right-liability relation. The relation is, from C's point of view, the entitlement of C as owner to all the value of her property, including its use-value. In the event of unauthorized use by D, C has a claim for a reasonable payment, in the form of the deemed license fee, representing the contribution to D's profit derived from his unauthorized use of the property. The relation is a right-liability relation because it is simply D's unauthorized use and the benefit obtained through it that justify the claim, whether or not the unauthorized use was a breach of duty.
There is a distinction, in other words, between a claim arising from unauthorized use per se and a claim arising from the breach of a duty not to use the property. Usually D does commit a breach of duty in using C's property without authorization. But unauthorized use may not be wrongful. For example, say D uses C's property in an emergency; it would be reasonable here for D to be free to use the property, but for him nevertheless to become subject to an unauthorized-use claim. This was how I suggested Vincent v. Lake Erie should be explained.Footnote 57 In this case, D moored his boat at C's dock in an emergency and was found liable to pay for the damage he caused to the dock, even though he was also said not to have acted wrongfully. The puzzle over this case has arisen from the assumption that a claim must be based on the breach of a primary duty, but for an unauthorized-use claim, just as in contract, a claim can arise against someone from an act that she is at liberty to carry out.
Invalid Transfers of Property
Finally, take the case where money or property originally belonging to C comes into the hands of D without a valid transfer from C to D, that is to say, without a valid exercise of a power of transfer by or on behalf of C. At the moment when the property enters D's estate, C acquires a claim against him to recover it, or its traceable proceeds, or its value.Footnote 58 The primary relation might be thought to be C's right in respect of her property, correlated with the duty of D and others to refrain from taking or using it or being in possession of it. But if the claim arises from the invalid transfer itself, which may not involve any act of D's at all, the primary relation cannot be a right-duty relation unless the duty is a strict-liability duty. This is indeed how C's claim is sometimes understood.Footnote 59 But again, it is more plausible to say that the primary relation is a right-liability relation. This is not to say that there is no duty not to take C's property or use it without permission; the point again is that this duty, and the claim that arises from its breach, should be distinguished from the primary right-liability relation and the claim arising by virtue of it from an invalid transfer or an unauthorized use.Footnote 60
VII. CAN PRIMARY-LIABILITY CLAIMS BE EXPRESSED IN ACCORDANCE WITH HOHFELD'S SCHEME?
If there can be strict-liability duties, then strict-liability claims can be explained in terms of the breach of a primary duty. But even with this concession, some of the claims discussed above cannot be, because the claim arises from an act that D was at liberty to perform. In my view, we should accept that a claim can arise from an event (which may or may not be an act of D) that is not a breach of duty by D, that is, the claim arises from a primary right-liability relation, or primary liability for short. D's “liability” signifies that he bears a responsibility in respect of a contingency, or a direct allocation of risk.Footnote 61 If the position before a claim arises is properly described in the way I suggest, it seems to me that it cannot be adequately described in terms of any combination of the relations found in Hohfeld's scheme. However, it is worth considering this issue further.
Does a Primary-Liability Claim Arise from the Exercise of a Power?
According to the “volitional act” understanding of a power, which is discussed above,Footnote 62 someone exercises a power when she performs a volitional act that changes the legal position. On this understanding, some of what I describe above as primary-liability claims can be explained, consistently with Hohfeld's scheme, as arising from a conventional power-liability relation. For example, if D uses C's property without authorization, or fails to perform a contract, and C acquires a claim as a result, the claim has resulted from a volitional act that on this approach is the exercise of a power, whether or not D has acted wrongfully.Footnote 63
However, the volitional-act view is rejected above. Properly understood, the exercise of a power involves a decision by the power-holder to act with a view to changing the legal position. In the cases that I characterize as primary-liability claims, the reason the claim arises is not that the law empowered D to determine whether a claim would arise and he chose to confer a claim. In Vincent, for example, D may or may not have intended to compensate C for damage or pay for his use; he is unlikely to have given any thought to the issue. If he did think about it, he would certainly not have thought that his liability to C was something he could determine through the exercise of a power, in the way that he might make a contract or transfer property. It was the mere fact of the unauthorized use that generated the claim. On the correct understanding of a power, in none of the cases I described as primary-liability claims is it plausible to say that the claim arose from the exercise of a power by D. Furthermore, sometimes there is no volitional act at all and so nothing that could constitute the exercise of a power, even on the volitional-act approach. For example, in the case where a claim arises against D because of his receipt of an invalid transfer of money or property from C, typically no volitional act on D's part is involved at all.Footnote 64
Is a “Right-Liability” Relation Merely a Conditional Remedial Duty or Liability?
In a case of the standard form of claim discussed above, in doing X, D breaches a duty owed to C not to do X and in consequence becomes subject to a claim by C for compensation for loss caused by doing X. Sometimes the legal position may be expressed in the way described above as remedialist, in the form of a condition for a claim for a certain remedy to be available: “In the event of loss to C caused by D's doing X, C has a claim against D for compensation.” This formulation disregards the primary relation, or one might say it expresses the primary relation as a conditional remedy. As suggested above, this is surely inadequate as a description of the primary relation because it makes no reference to the fact that doing X is a breach of duty.
What about the cases I describe as primary-liability claims? Here, can the primary relation be expressed simply as a conditional remedy? In these cases, the claimant C acquires a claim on the occurrence of some event that is not a breach of duty by D. As discussed above, the usual understanding of a claim is that it is a remedial right correlated with a duty to provide a remedy on the part of D. Before the claim arises, it appears, one can say simply that there is a conditional right-duty relation, the condition being whatever act or event triggers the claim. On my view, the claim is a Hohfeldian power, that is, the remedial relation consists of a power correlated with a Hohfeldian liability, but again, it seems that one can say that the position before the claim arises is just a conditional power-liability relation. This might appear to provide an adequate description, within Hohfeld's scheme, of the primary relation, but I am not sure whether this is as straightforward as it may appear.
In some circumstances, it seems entirely apt to describe the legal position in terms of a conditional relation. For example, say that under a contract A has a duty or liability to make a payment if a certain event occurs. Here it would be customary and surely appropriate to say that before the event, A is subject to a conditional duty or liability. But in the cases described above in terms of right-liability relations, a conditional formulation seems less apt. It seems wrong to formulate a primary right as a conditional right to a remedy. For example, take the case where on the receipt of an invalid payment from C, D incurs a liability or duty to repay C. It would hardly be apt to say that before D has a received an invalid transfer, she is subject to a conditional duty or liability to return the invalid transfer or repay its value. Similarly, in the case of a strict-liability tort claim by C against D, such as a claim under the rule in Rylands, it would hardly seem apt to say that before he suffers any harm, C has a conditional claim for compensation and D a conditional liability to pay it. In contract, it is suggested above that the primary relation between contracting parties C and D is appropriately expressed in terms of a right-liability relation that allocates risk between them without imposing a duty to perform, so that when D departs from her contractual performance, she does not commit a breach of duty but nevertheless becomes liable to pay compensation to C. If this analysis is accepted, it would still not seem apt to say that during the ordinary subsistence of a contract, before D has departed from her performance, she is subject to a conditional duty or liability.
Is this just a matter of usage, or is there a significant distinction between what I call a primary-liability claim and the type of conditional relation exemplified by the case of an ordinary contractual condition? In the case of a contractual condition, the condition relates to a particular, well-defined event, and the position arising from it is also well defined. The legal position before the event occurs is defined by reference to the position if it does occur, in anticipation of it, which is why it is aptly given a conditional formulation. By contrast, in the primary-liability cases there are many possible events that could give rise to a claim by virtue of the primary relation, and the remedial position will vary accordingly. If the legal position is uncertain, the question whether such an event has occurred, and if so, what the remedial position is as a result, should in principle be addressed by considering the content of the primary relation, that is to say, what the contingency is and what D's responsibility is in respect of it. Thus it is not helpful nor illuminating to define the primary relation by reference to the remedial position, and this is why it seems wrong to express the primary relation as a conditional remedy.
Possibly this distinction corresponds to the distinction between “external” and “internal” conditions.Footnote 65 Say that in the event of X, P will have to pay Q the sum S. It is said that if the condition is external, the position should be expressed as follows:
If X occurs, Q has a right to the payment of S by P.
But if the condition is internal, the formulation should be:
Q has a right [to-be-paid-S-by-P-if-X-occurs].
In the case of an internal condition, before the event X occurs Q has a conditional right to the payment, whereas if there is an external condition Q has no right at all unless X occurs. Thus in the former case, the same legal relation subsists before and after X, whereas in the latter case X generates a new legal relation; before X occurs, there must be a distinct legal relation subsisting between the parties, by virtue of which the right arises on the occurrence of X. Usage of the conditional formulation is apt for internal conditions rather than external conditions. In the case of a contractual condition, the condition is internal, whereas in the case of a primary-liability claim, the condition is external.
Lastly, even if I am wrong to deny that what I refer to as a primary-liability claim can be adequately expressed in the form of a conditional remedy, this does not undermine the basic distinction between claims based on a primary duty and claims not based on a primary duty; and furthermore, it remains an appropriate and convenient formulation to describe the latter as a claim based on a primary liability.
VIII. CONCLUDING COMMENTS ON PRIMARY-LIABILITY CLAIMS AND HOHFELDIAN ANALYSIS
I argue above that some claims in private law that are conventionally understood to arise from a primary right-duty relation would be better understood as arising from what I describe as a primary right-liability relation. Whether this is right does not depend on whether the concept of a right-liability relation can be found, expressed as such, in the case law. It would certainly fail this test. I discuss claims in tort, contract, and property, but not for the purpose of identifying examples of the formulation of the law in terms of primary-liability claims. To the contrary, I emphasize that claims are customarily expressed to arise from a wrong or breach of duty. My point is that sometimes claims that should be understood as primary-liability claims are mischaracterized as primary-duty claims.
I put forward two general arguments in support of the recognition of primary-liability claims, the first concerning the nature of duties and subjective fault, and the second concerning the relationship between primary and remedial rights. The first argument began with the point that if, as the choice theory of responsibility implies, the content of a duty should be assessed from the subjective standpoint of the duty-bearer, there is no room for strict-liability duties. It follows that if all claims in private law arise from the breach of a primary duty, no claim should be permitted in the absence of subjective fault. This would rule out many well-established claims in contract and property as well as tort. The recognition of the right-liability relation would allow such claims to be admitted without any contradiction of the choice theory. Those who reject the choice theory may be indifferent to the argument, though surely for some of them it is the difficulty of explaining strict-liability claims that is behind their doubts about the choice theory.
The second argument relies on the fact that a remedial relation must depend on the primary relation. If D had a primary duty, it does not follow that he should always become subject to a remedial duty to perform the original duty or some variant of it, enforceable by an order of specific performance or an injunction. However, there are some circumstances in which the absence of such a remedial duty implies that D did not originally have a primary duty at all, but a liberty not to perform. A standard case in contract illustrates this above. In such circumstances, either such an order should be available, or the claim was based on a primary liability, not a primary duty. If we are satisfied that specific performance or an injunction should not be available, we are bound to accept that the claim is a primary-liability claim. If we are not prepared to accept the possibility of primary-liability claims, we are bound to conclude that the current remedial law is sometimes indefensible. Even when an injunction or specific performance is not at stake, it is still inconsistent both to characterize a claim as based on a wrong or breach of duty and also to insist that D was at liberty to act in the way that generated the claim; this is the point made in connection with the case of Vincent. Also, I point above out that the recognized measure of the pecuniary remedy for the unauthorized-use claim cannot be reconciled with the assumption that the primary relation is a right-duty relation.
These arguments for recognizing the concept of a right-liability relation at the same time demonstrate the role that the concept can play in legal argument. The arguments do not in themselves indicate what primary duties and primary liabilities arise in private law. That depends in the ordinary course on arguments of authority, and more fundamentally on arguments drawn from a substantive theory of tort or contract or private property that offers a moral foundation for and guidance in the interpretation and development of these areas of law and the claims they generate. But the concept of a primary-liability claim will sometimes be involved in such arguments. As mentioned above, the concept is a matter of the form or modality of legal rules, which is what Hohfeld's scheme of legal relations is concerned with. Hohfeldian classification provides the tools for the precise formulation of legal rules and helps to expose certain types of fallacious argument resulting from the use of imprecise terminology.Footnote 66
I also pointed out that the argument for the recognition of primary liabilities is not a form of “duty-skepticism” in private law. The recognition of primary liabilities is not inimical to the recognition of primary duties. Neither is it at odds with an understanding of private law in terms of correlativity and bilateral relations; to the contrary, it presupposes that private law should be understood in this way. In its adoption to the Hohfeldian principle of correlativity in private law,Footnote 67 and in its recognition of the importance of the analysis of the form or modality of rules, the approach is Hohfeldian. However, if it is right that the recognition of the right--liability relation is required for an adequate account of private law, Hohfeld's scheme is incomplete, since no such relation is found there.
Finally, even if I am wrong in saying that there is a lacuna in Hohfeld's scheme, it is nevertheless the case that there is an important and underexplored distinction in private law between what I describe as primary-duty claims and primary-liability claims.
