Jump to ContentJump to Main Navigation
Torts and Other Wrongs$

John Gardner

Print publication date: 2019

Print ISBN-13: 9780198852940

Published to Oxford Scholarship Online: February 2020

DOI: 10.1093/oso/9780198852940.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 17 January 2022

Torts and Other Wrongs

Torts and Other Wrongs

(p.1) 1 Torts and Other Wrongs
Torts and Other Wrongs

John Gardner

Oxford University Press

Abstract and Keywords

This chapter focuses on the law of torts, not in the United States, but in other major common law jurisdictions (England and Wales, Canada, Australia, and New Zealand) in which tort cases are normally adjudicated by judges sitting without juries. It considers the so-called classical interpretation of the common law of torts by John Goldberg and Ben Zipursky, and how they tend to equivocate on an important point of law in a way that puts them at odds with some writers with whom they would do better to make common cause. It suggests that this equivocation is where the law of the United States parts company with the law in the rest of the common law world. The problem, an English lawyer might then teasingly say, is with American tort law rather than with the Goldberg and Zipursky rendition of it.

Keywords:   tort law, common law, John Goldberg, Ben Zipursky, private law, law of obligations


It is hard to think of any contemporary writers who have done more than John Goldberg and Ben Zipursky to reassert and reinvigorate what might be called the classical interpretation of the common law of torts. I for one am greatly in their debt. They have taught me a great deal, not only about torts, but also about how to combine legal argument felicitously with philosophical insight and historical scholarship. Like them, and partly because of them, I believe that the classical interpretation is the correct one. Other familiar ways of explaining what is going on in the common law of torts, while often illuminating in their own ways, are parasitic. They rely on the classical interpretation, often surreptitiously, for their appeal or even their intelligibility.

Given their influence over my own thinking, it hardly comes as a surprise that I have few significant disagreements with Goldberg and Zipursky. I do think, however, that they tend to equivocate on an important point of law, in a way that puts them at odds with some writers with whom they would do better to make common cause. This charge of equivocation is one that I make only with grave misgivings. That is not only because Goldberg and Zipursky stand out, among theorists of tort law, for their detailed knowledge of and sustained interest in legal doctrine, which they usually go to great lengths to spell out clearly. It is also because I am not sure to which body of law, exactly, their treatment should be held answerable. They give the impression that they are writing about the tort law of the United States of America. But it is doubtful whether there is such a thing. There are at least 51 legal systems in the United States with tort jurisdictions. Even if a national tort law can be distilled from this melée, it might well be hard to work out what it says on the point of law I have in mind. US legal systems, unlike other common law legal (p.2) systems, tend to use jury trials for the bulk of their tort litigation. It may be that trial judges tend to pass the buck to the jury on the point in question, and appellate judges then stay clear of it, with the result that the legal position is indeterminate. In that case there is no point of law for Goldberg and Zipursky to report, equivocally or otherwise.

In what follows, I will focus my attention on the law of torts, not in the United States, but in some other major common law jurisdictions (England and Wales, Canada, Australia, and New Zealand) in which tort cases are normally adjudicated by judges sitting without juries. So far as I am aware, these legal systems are all determinate, and all agree, on the point of law in question. If tort law in the United States is not determinate, or is determinate but does not agree, then what I am calling Goldberg’s and Zipursky’s equivocation is really something else. It is really their testimony to a respect in which the law of the United States has parted company with the law in the rest of the common law world. The problem, an English lawyer might then teasingly say, is with American tort law rather than with the Goldberg and Zipursky rendition of it. If that turns out to be the right diagnosis, then it goes a long way to explain why (as Goldberg and Zipursky lament) ‘academics [in America] have lost their feel for this basic legal category [viz. tort].’1 For, as I hope to show, one cannot have a proper feel for the category until one takes to heart this point of law that has only faltering or grudging recognition in the Goldberg and Zipursky analysis.


Goldberg and Zipursky famously say that ‘civil recourse’ is a defining feature of the law of torts.2 I could not agree more. To (p.3) spell the claim out: There is no law of torts if there is no legal power for someone who holds herself to have been on the receiving end of a tort (a ‘plaintiff’) to summon the alleged tortfeasor (a ‘defendant’) before the courts, and to do so unilaterally (without the leave of any official), with a view to obtaining a court-imposed or court-approved remedy against that defendant. It is then for the defendant, not any official, to apply for the case to be struck out. Otherwise the case marches on to trial, and in the absence of any defence marches on at the sole instance of the plaintiff to judgment against the defendant, and to the award of a remedy against the defendant. This feature (or set of features) amply distinguishes the law of torts from the criminal law, and from most if not all of public law.3 In criminal law, even where private prosecutions are possible, they can be taken over (and discontinued) by the prosecutorial authorities. In public law a cause of action is available, in general, only with leave of the court. The unilateral power of the plaintiff is limited to applying for leave of the court, which issues the writ (traditionally certiorari, mandamus, prohibito, or habeas corpus) on the application of (or ‘ex parte’) the applicant. Public law and criminal law therefore lack the defining apparatus of civil recourse.

Yet there is nothing here to help us distinguish the law of torts from the rest of private law. Indeed civil recourse, as I have just (p.4) explained it, is a defining feature of private law as a whole in the common law systems. It is equally a feature of the law of breach of contract, the law of unjust enrichment, and the law of equitable (including trustee) liability. What do Goldberg and Zipursky say to help us understand how the law of torts is different from these other departments of private law?

Here is one thing they say. They say that torts are wrongs, so that the tort plaintiff seeking recourse necessarily alleges (whether expressly or by implication) that she has been wronged by the defendant—that the tortfeasor has violated her rights—and that she is entitled to have a remedy against him on that very ground.4 Again I could not agree more. Torts are legal wrongs, i.e. breaches of legal duty or obligation5, and they are legal wrongs against particular people, i.e. they violate particular people’s legal rights. But this feature (or set of features) still does not help to distinguish the law of torts from the law of breach of contract or from the law of equitable liability. It only helps to distinguish it from the law of unjust enrichment. Although a wrongdoer may be unjustly enriched by his wrong, a claim against him for unjust enrichment does not depend for its success on his having been a wrongdoer.6 It is enough that, were he to keep his gains at the plaintiff’s expense, that would be a wrong against the plaintiff. Or, to put the contrast differently, the law of unjust enrichment offers remedies for what would be wrongs if not remedied. Whereas the law of torts, like the law of breach of contract, offers remedies for what are already wrongs at the time of the remedy, and on the ground that this is what they are.

On this point, as Goldberg and Zipursky point out, many have abandoned the classical interpretation of the law of torts. Following in Holmes’ footsteps they have said that the law of torts and the law of breach of contract are, in the respect just discussed, no different from the law of unjust enrichment. The nomenclature notwithstanding, it is not wrong to commit a tort or (p.5) to breach a contract. The only wrong (if there is one) is not to pay the price when one does.7 Goldberg and Zipursky will have none of this. They argue that the supposed counterexamples to the thesis that torts are wrongs are not counterexamples at all (the trespass in Vincent v Lake Erie, for example),8 or, if they are, that they should be regarded as peripheral (the innominate tort in Rylands v Fletcher).9 Here Goldberg and Zipursky stand shoulder to shoulder with those who have exposed for the mistake it is the Holmesian heresy (later an economistic orthodoxy) about breach of contract.10 In the process they implicitly maintain the classical continuity between the law of torts and the law of breach of contract, and draw the classical contrast between these two areas of law and the law of unjust enrichment.

So Goldberg and Zipursky agree that the law of breach of contract and the law of torts share the twin defining features mentioned so far (civil recourse, and for wrongs). They presumably think, and they would be right to think, that torts and breaches of contract differ from each other in some other yet-to-be-identified respect. But they say nothing at all about what I have called equitable liability, at least some of which is liability for wrongdoing. A breach of trust, a breach of fiduciary duty as a company director or attorney or agent, a breach of confidence: all these are equitable wrongs for which there is undoubtedly civil recourse against the wrongdoer. Goldberg and Zipursky ignore these (p.6) wrongs. Indeed they do not seem to leave any separate logical space for them. They quote the characterization of tort law found in Prosser and Keeton’s classic textbook. Torts are ‘civil wrong[s], other than breach[es] of contract, for which the court will provide a remedy’.11 As part of their move away from the classical interpretation and towards the newfangled Holmesian alternative, Prosser and Keeton ended up rejecting this as a way of demarcating tort law, but Goldberg and Zipursky endorse it.12 In the light of this endorsement, we may legitimately ask them: Where do equitable wrongs fit into the classification? They are clearly ‘civil wrongs … for which the court will provide a remedy’. Equally clearly, they are not breaches of contract. Trusts, non-trust fiduciary duties, and confidences may but need not be contractually created, and when they are contractually created, an action for breach of contract exists alongside, and does not replace, the action for breach of trust or breach of fiduciary duty or breach of confidence (as the case may be). So should these breaches, these wrongs, be regarded as torts? No. Let me explain why not.


[C]ompensation is an equitable monetary remedy which is available when the equitable remedies of restitution and account are not appropriate. By analogy with restitution, it attempts to restore to the plaintiff what has been lost as a result of the breach, i.e. the plaintiff’s lost opportunity. The plaintiff’s actual loss as a consequence of the breach is to be assessed with the full benefit of hindsight. Foreseeability is not a concern in assessing compensation, but it is essential that the losses made good are only those which, on a common sense view of causation, were caused by the breach.13

(p.7) Let’s temporarily bracket what McLachlin J says about causation and foreseeability. Our immediate concern is with the trio of remedies that she associates with breach of equitable duty. Allow me briefly (and crudely) to expand on her taxonomy.

(a) Restitution.

As McLachlin J uses the term, restitution is the return of an asset in specie. It reunites one party with the very thing that she lost to the other party, thereby extinguishing—in one fell swoop—the one’s loss of that thing and the other’s gain of it. An everyday situation (usually arising out of an unhappy contractual transaction rather than breach of an equitable duty) is that of mutual restitution. You have the car; I have the $1000 you paid me for it. Mutual restitution involves your returning the car to me and my returning the $1000 to you. The abstractness and fungibility of money means, of course, that it usually makes little sense to ask whether the $1000 I returned to you is the same money that originally changed hands, as opposed to just the same amount of money. This fact has licensed a wider loosening of talk about restitution, so that the word is often used to refer to returns of assets no longer in specie. You passed the car to a third party in exchange for a painting; so now, instead of the car, can I please have the painting? A court will sometimes ‘trace’ my claim on the car through to the painting. The problem is that, as the onward transactions build up, my loss and your gain increasingly come apart. You exchanged a bad car for a good painting, so if I get the painting in place of the car, on the basis that you transferred my $1000-worth of car for it, I take a $2000 gain off your hands even though I only made a $1000 loss on the car. That is no longer restitution in the strict sense. In fact there can no longer be restitution in the strict sense unless your transaction with the third party is unwound as well. If you have to pay me your full $2000’s worth of gain, as distinct from the $1000’s worth of loss that I suffered, it is usually better to think of that as a different kind of remedy, to which we now turn.

(b) Account (of profits).

A trustee has a duty to restore to the trust accounts any assets that have been displaced from them by his or (p.8) her breach of trust. Often there can be restitution in the strict sense, because the trustee still has the asset. But a trustee also has a duty to surrender to the trust accounts any profits he or she made on a trust asset, whether or not it was made by a breach of trust. So if she swapped the trust’s car for a painting for herself, she has to hand over the painting even though it is worth $2000, and the car was only worth $1000. And she has to hand over the extra $1000 even if she already sold the painting for $2000 and used $1000 of the proceeds to buy back the car and return it faithfully to the trust. Why? Because any profit on dealings with a trust asset is automatically impressed with the trust. Disgorgement of this profit, as already indicated, is not restitution in the strict sense.14 It extinguishes a gain, but it does not necessarily extinguish a loss. Any off-accounts profit made by the trustee is repayable to the trust even if it would not have accrued to the trust but for the off-accounts dealing, i.e. even if the trust is no worse off than it would have been had the off-accounts dealing not occurred.15 So there need be no loss involved except the notional (‘accounting’) loss that is constituted by the absence of the gain. That gain, qua gain, is what equity chases under the heading of ‘account’, as McLachlin J is using the term.

One short path leads off from here to equity’s handling of unjust enrichment; the trustee, to repeat, need commit no wrong to incur the liability to disgorge a profit, and in that respect he or she is in the same position as someone unjustly enriched (e.g. by the mistake of another). But a different short path connects the disgorgement liability of the trustee to the equitable liability of confidants and non-trustee fiduciaries. True, one cannot literally extend beyond the trust context the idea that the profit (p.9) on a trust asset is automatically impressed with the same trust. Information can be held in confidence, but profit cannot. So when equity chases the gain made from trading in confidential information, or from fiduciary double-dealing not involving any use of a trust asset, it does so, not on the footing that the gain on an asset automatically belongs to whomever the asset belongs to (there is no asset here16) but on the different footing that the gain is ill-gotten. There is a duty not to make unauthorized profits from these equitable relationships, and the remedy for its breach is disgorgement of the very profits made.17 What we have here is disgorgement as a remedy for equitable wrongs.

(c) Compensation.

What McLachlin J calls ‘compensation’ is the mirror image of the disgorgement remedy just described. The latter remedy aims to eliminate gains made by a wrongdoer even when he inflicted no losses; the former aims to eliminate losses inflicted by a wrongdoer, even when she made no gains. A tort lawyer might be tempted to say that what we are dealing with here are compensatory or reparative damages. But ‘[w]oe betide’ the lawyer who speaks of ‘damages for breach of trust’ or ‘damages for breach of fiduciary duty’.18 McLachlin J offers one reason why these are solecisms. The rules for the assessment of equitable compensation, she says, differ from the rules for the assessment of reparative damages at common law. In equity, she says, ‘[f]oreseeability is not a concern in assessing compensation’. This (p.10) claim is contentious,19 and more generally the differences in respect of assessment between common law damages and equitable compensation are unsettled.20 Less contention, however, attaches to another difference—also brought out by McLachlin J’s remarks—between reparative damages in tort law and compensation for equitable wrongs.

According to McLachlin J, equity prefers restitution and/or disgorgement, and resorts to a reparative award only where these other remedies are ‘not appropriate’. Or at any rate, equity prefers restitution, and resorts to reparation only ‘[b]y analogy with’ restitution. Either way, reparation does not have pride of place as a remedy for equitable wrongs. In this respect the law of torts is different. Reparative awards against tortfeasors can be supplemented by the court and eschewed by the plaintiff. But still they have pride of place as a remedy for tortious wrongdoing. They are not the remedy of last resort that they are for equitable wrongdoing; they are the remedy of first resort in the law of torts. Why this difference in remedial primacy? Here is one natural explanation, which is the explanation that I favour. The main reason why the defendant has duties to the plaintiff in the law of torts is to protect the plaintiff from losses, and the main mode of recourse that a plaintiff has against the defendant in a tort case is therefore recourse in respect of his losses.21 This contrasts with (p.11) the position in equity where the main reason for the defendant’s duties is to secure that the defendant’s dealings are conducted for the plaintiff’s advantage, and not for the defendant’s own. So the emphasis in equity is on the diversion of advantage—in the form of assets or profits—as opposed to the causation of loss, which is tort law’s first concern.

This explanation of the difference between tort remedies and remedies for equitable wrongs may touch a nerve for Goldberg and Zipursky. They contrast those who ‘have placed losses at the center of tort theory’ with ‘those who wish to retain a tort theory centered on wrongs’.22 They are not sure, on the evidence of my writings, where I stand with respect to this supposed divide.23 You can now see why. In my view, the divide is artificial. My view is that the wrongs of tort law (including those that are actionable without proof of loss) exist primarily in order to protect people from loss. This view, which calls for a lot of careful unpacking, is both wrong-centred and loss-centred. For present purposes, however, further unpacking is unnecessary. We are not concerned here with how to explain the different remedial priorities of the law of torts and the law of equitable wrongs. What concerns us here is simply that there are such different remedial priorities. Why does this concern us here? Because Goldberg and Zipursky call it into question. Since they say nothing at all about equitable wrongs, they obviously do not query the claims I just made about the primacy of gain-based remedies for equitable wrongs. But they do make it their business to query the claims I just made about torts. They do not exactly deny, but they certainly put up resistance to, the idea that there is a primary or pride-of-place remedy in the law of torts, and they resist, in particular, the idea that reparative damages are it.

(p.12) IV

I say that Goldberg and Zipursky ‘do not exactly deny’ the primacy of reparative damages because they agree that ‘[a] longstanding principle of remedies for nonwillful wrongs sets make-whole [reparative damages] as the default remedy’.24 This turns out, however, to be both a limited and a double-edged concession. First, ‘default remedy’ is too weak to capture the special place of reparative damages in the law of torts. Second, ‘for nonwillful wrongs’ already sets the scene for the introduction of a rival view. Those who think of the primacy of reparative damages as extending more widely are portrayed by Goldberg and Zipursky as falling into the trap of mistaking the law of negligence for the law for the whole law of torts.25

The rival view is not long in coming. Goldberg and Zipursky resist the idea that reparative damages have a more general remedial primacy in the law of torts by observing that diverse remedies, indeed bewilderingly diverse remedies, are commonly granted in tort cases.26 These include injunctions and exemplary (Americans say ‘punitive’) damages as well as reparative damages. Although Goldberg and Zipursky do not mention them, we could also add gain-based damages to the list.27 We could also amplify some of their points. For some torts—notably private nuisance and trespass to land—plaintiffs are very often less interested in obtaining an award of damages than they are in obtaining an injunction against continuation or repetition of the tort. Indeed, it is common for plaintiffs in nuisance and trespass to land cases not even to claim damages, or rather to claim damages only in lieu of an injunction should the court for some reason decline to issue the injunction sought. Should we conclude that private nuisance and trespass to land are not torts? Clearly not. So why, ask Goldberg and Zipursky, should we think that torts are characterized by the (p.13) primacy of reparation? Different torts, different scenarios, different remedial devices.

This argument proceeds from good observation of legal practice, but it misses the point. Even if almost all tort litigation ended in injunction, because plaintiffs did not seek damages, reparative damages would still be the common law’s remedy of first resort in the sense that we are after. For this is the only remedy against a tortfeasor that the successful plaintiff enjoys as of right. Injunctions (an equitable remedy)28 and damages in lieu of injunctions (originally a statutory equitable remedy29) are available only in the (p.14) discretion of the court. Likewise exemplary damages30 and disgorgement of profits.31 This is one way in which the law, in all the jurisdictions I am familiar with, distinguishes torts from equitable wrongs. Whether equitable wrongs attract any remedy as of right is debatable. Possibly, as McLachlin J hints, all equitable remedies (or all equitable remedies except for proprietary remedies against trustees) are discretionary, and possibly that includes even what McLachlin J casts as equity’s ‘default remedy’, restitution of an asset in specie. But it is not a matter of any significant debate—it is ‘trite law’—that torts do attract a particular remedy as of right, and that that remedy is an award of reparative damages.32

I say that this is not a matter of any significant debate without meaning to suggest that it attracts no sceptical reactions. Andrew Burrows, for example, notes the received wisdom that ‘equitable remedies are discretionary, while common law remedies are not’, but brushes the distinction aside as ‘highly misleading’:

It surely cannot seriously be suggested that the law is less certain—that a judge has more discretion—in deciding whether specific performance [of a contract] should be ordered than in deciding, in relation to damages, whether a loss is too remote or whether an intervening cause has broken the chain of causation or whether the claimant has failed in its duty to mitigate its loss.33

(p.15) That’s quite true, but again it misses the point. A right to reparative damages at common law is a right to some reparative damages. The job of the court is, inter alia, to determine what the amount is to be. The tort plaintiff claims unliquidated damages and the court liquidates them, i.e. crystallizes them into a money sum, which can then be recovered as a judgment debt. I would be the first to agree that in doing so the court exercises discretion about a range of matters bearing on the quantification of loss.34 It even has some discretion about how to conceptualize ‘loss’ for these purposes. But it has no discretion about whether to exercise these discretions. It is legally bound to make a reparative award of some kind if one is sought by the successful plaintiff, because he or she has a right to such an award.35 By contrast, the court is not legally bound to issue an injunction of any kind, or to order any kind of disgorgement of profits, or to award any kind of exemplary or punitive damages (even at common law).

How about restitution, a remedy known to common law as well as to equity? There are many important tort cases (mainly in conversion and trespass to land) in which damages have been awarded on an ostensibly restitutionary basis,36 and occasionally the successful plaintiff is said to have a right to elect that basis of (p.16) award.37 The proper interpretation of these cases is disputed. At least some of them, it is sometimes said, should be reclassified as discretionary disgorgement cases.38 Or in at least some of them, it is said, the award, although genuinely restitutionary, should be understood as remedying an unjust enrichment which occurred in tandem with the commission of the tort, thus not as a tort remedy at all.39 Or, it is said, there was indeed a restitutionary award for the tort itself in at least some of these cases, but it can be reanalysed, in one way or another, as reparative.40

Such a reanalysis would not be totally surprising since, as noted above, restitution strictu sensu annuls the defendant’s gains and the plaintiff’s losses in one fell swoop. Equity is more interested in chasing the defendant’s gain; so in equity, restitution fades into disgorgement. The common law of torts, by contrast, is more interested in remedying the plaintiff’s loss, and so tends to emphasize the loss side of the restitutionary equation. It is true that in property-use cases the courts sometimes have to go to considerable lengths to present what they tellingly41 call ‘restitutionary damages’ as loss-rectifying, and in doing so they may even be forced to abandon ways of thinking about loss that are standard elsewhere in tort law, even resorting to fictions.42 (p.17) That they do so is not a sign of cracks appearing in the doctrine of the primacy of reparation in the common law of torts. It is, on the contrary, testament to the power of that doctrine. The courts are under pressure to find a loss, however understood, to which they can attach a quantification, because reparative damages are the damages to which the successful plaintiff has a right.

Where a tort can be committed without causation of loss (e.g. trespass or libel) there is, of course, a possible scenario in which a plaintiff proves the tort but has no losses to recover, even by an unorthodox accounting of losses. Here the court is bound to award nominal damages. Goldberg and Zipursky regard this as another sign that reparative damages are not as special as they are made out to be.43 But again it points the other way. The courts do not regard themselves as legally free to deny reparative damages to the plaintiff even when there is nothing to repair. The court can however express the point that there is nothing to repair by awarding a symbolic sum. Indeed the court can go further. It can award such a conspicuously trifling sum that it thereby expresses contempt for the plaintiff who was such a stickler for his rights that he insisted on upholding them expensively and painfully in court even when the defendant’s violation of them caused him no loss. If the plaintiff had no right to damages, nominal and contemptuous damages would not be needed and would not carry the meanings that they do. The court could simply declare that a tort was committed and be done with it. And if the defendant’s right to damages were not a right to reparative damages, how would we explain the fact that nominal (and contemptuous) damages are to be awarded only in cases in which the plaintiff proved no loss, i.e. had nothing relevant to repair? The explanation is that they are not just nominal damages but nominally reparative damages. In that way, the existence of nominal damages confirms rather than refutes the primacy of reparative damages as a remedy in the law of torts.

(p.18) V

In buttressing the Goldberg-Zipursky case against the primacy of reparative damages as a remedy in the law of torts, Zipursky (writing alone) is at great pains to deny that there is a legal duty on tortfeasors to pay reparative, or indeed any other, damages. ‘Rather,’ he says, ‘the law imposes liability upon a tortfeasor.’44 I am not sure what he means. Sometimes he seems to be saying that even after an award of reparative damages has been made by the court—even after the court has liquidated the plaintiff’s unliquidated claim—there is no legal duty on the tortfeasor to pay those damages.45 If that is what he means then he is helping himself to a false contrast between ‘duty’ and ‘liability’. The primary liability of tortfeasors is none other than a liability to be placed under a duty by the court to pay a liquidated sum in reparative damages. Paying that sum is what the court orders the defendant to do, on pain of further recourse against him by the plaintiff in an action, now not for tort, but for recovery of an unpaid judgment debt.46 That this action for recovery of a debt is an action for enforcement of a duty to pay can only be doubted by way of a further Holmesian heresy with which (I take it) Zipursky would not wish to be associated, according to which one has no legal duty to pay one’s legally recognized debts.

Fortunately, however, Zipursky more often appears to mean something more limited and less Holmesian by his ‘no legal duty to pay’ thesis. He appears to mean only that the tortfeasor has no legal duty to pay any reparative damages unless and until ordered to pay them by the court.47 I disagree, and with reason,48 but happily the (p.19) disagreement is irrelevant here. Even if the tortfeasor lacks a legal duty to pay reparative damages before the court awards a liquidated sum in reparative damages against her, that does not alter the fact that the court has a legal duty to award a liquidated sum in reparative damages against her if the tort is proved. This legal duty exists because the successful plaintiff has a legal right to reparative (now taken to include nominally reparative) damages. The plaintiff’s right grounds a legal duty on the court to impose a new legal duty on the tortfeasor, a legal duty that is also grounded (by the court, and by the law) in the plaintiff’s right.


Their resistance to the idea that reparative damages have a special primacy in the law of torts combines with their scattered remarks partly conceding the opposite to yield a pattern of equivocation in Goldberg and Zipursky’s work. If I am right it is equivocation on a point of law on which (in the common law systems I am familiar with) no room for equivocation exists. This equivocation by Goldberg and Zipursky is patently connected—I am not sure whether as cause or as consequence—with their wish to keep their distance from what they call ‘corrective justice’ explanations of tort law. They regard a ‘corrective justice’ explanation as a rival to a ‘civil recourse’ explanation; they would rather set up in competition to than in league with those, such as Jules Coleman,49 Ernest Weinrib,50 Stephen (p.20) Perry,51 and Arthur Ripstein,52 who emphasize the element of corrective justice in tort law. That seems to me to be a mistake—not a legal one but a philosophical one. It is a mistake because ‘corrective justice’ is an answer to a different question from ‘civil recourse’. Once we know that the law of torts involves civil recourse, we naturally want to know what form the recourse takes. ‘Corrective justice’ tells us that it takes a primarily corrective form, that the law of torts attempts, through the recourse it gives, to put one or both of the parties to the proceedings into the same position that they would have been in had the wrong, the tort, not been committed. Compare that, if you like, with civil recourse that takes a primarily punitive form, attempting to impose loss on the defendant because of some wrong he did whether or not it thereby corrects any gain or any loss that came of the wrongdoing itself. There is no corrective justice in that. That is purely retributive justice. Possibly in some jurisdictions the retributive and the corrective are given equal pride of place in the award of damages in (what are still called) tort cases. But to go this way is to have ‘lost [one’s] feel for this basic legal category’.53 Or so I have claimed in this paper.

So it is tempting to render what I have claimed in this paper in the following (more theoretically highfalutin) terms. To distinguish the law of torts adequately from other branches of law we need to combine the Goldberg and Zipursky ‘civil recourse’ theory of tort law with some kind of ‘corrective justice’ theory, such as Coleman’s. This restatement of what I have said would give the right general impression, but it is in several respects misleading. For a start, everything I have said above about the law of torts applies equally (mutatis mutandis) to the law of breach of contract, so the manoeuvre of combining ‘civil recourse’ with ‘corrective justice’ would still not be adequate to distinguish the law of torts from the law of breach of contract. To do that, we would clearly need to add a further criterion on top of ‘civil (p.21) recourse’ and ‘corrective justice’. Otherwise we accidentally, and mistakenly, classify breach of contract as a tort.

More significantly, however, the combination of ‘civil recourse’ and ‘corrective justice’ doesn’t distinguish tort law from the law of unjust enrichment or from the law of equitable wrongs. These departments of private law are equally about correcting things, about putting one or both of the parties to the proceedings into the same position that they would have been had something not gone amiss. They just have different things to correct, and different ways of correcting them. Corrective justice can be reparative but it can also be disgorgative (excuse the coinage) or restitutionary. So ‘corrective justice’ doesn’t really help with the specific problem that has been the focus of this paper, namely that of how to isolate tort law from its various private law neighbours. By the time we get to that problem we are already in the neighbourhood of corrective justice. Two criteria we need to add to differentiate tort from two of its close neighbours in the neighbourhood (I have argued) are these: that the correction involved in tort law is the correction of wrongs (unlike in the law of unjust enrichment), and that it is correction that takes a primarily reparative form (unlike in the law of equitable wrongs). But all of this is corrective justice.

We could sum up the position that I have sketched in this paper by saying that the law of torts is a law of

  1. (a) civil recourse

  2. (b) for wrongs

  3. (c) in which primarily corrective justice is attempted

  4. (d) in a primarily reparative mode

  5. (e) in response to claims for unliquidated sums

  6. (f) where the duties breached are non-contractual.

Criterion (a) serves to distinguish the law of torts from criminal law and most if not all of public law. Within private law, criterion (b) keeps it apart from the law of unjust enrichment and some parts of the law of trustee liability. Criterion (c) makes the contrast between the law of torts and a conceivable branch of law in which there is, say, a primarily retributive aim to civil recourse. Criterion (d) keeps the law of torts distinct from the law of equitable wrongs. Criterion (e) is needed to draw the (p.22) distinction—which I have only mentioned in passing—between the original proceedings to obtain an award of damages against the tortfeasor and subsequent proceedings for enforcement of the award as a judgment debt (a liquidated sum). More generally, criterion (e) means that proceedings for debt recovery are not tort proceedings. Finally, criterion (f) serves as a placeholder for some future explanation of why breach of contract is not a tort even though it meets all of the other conditions.54

Criteria (a), (b), and (f) are endorsed by Goldberg and Zipursky, whereas criteria (c), (d), and (e) are not. I suppose they may think that (e) is redundant on the basis that actions for debt already fall outside tort law thanks to criterion (b). I suggested at the end of section V that this is a mistaken thing to think; actions for debt, including actions for enforcement steps to be taken towards recovery of a judgment debt, are premissed on wrongful non-payment of the debt. More central to this paper, however, has been my argument that Goldberg and Zipursky’s resistance to (d) involves them in equivocations about a settled legal doctrine, and that this is bound up with the philosophical error that Goldberg and Zipursky make in trying to distance themselves from (c). Tort law cannot be understood without understanding the primacy of the reparative, which entails the primacy of the corrective, in its remedial doctrines. The result of their omitting these criteria is that, in the work of Goldberg and Zipursky, the law of torts is incompletely differentiated from the rest of private law. It follows that the classical interpretation is incompletely, even though valiantly and insightfully, represented.


You may think there is something decidedly vain, or academic in the pejorative sense, about the task of identifying the defining features of the law of torts, as my list (a) to (f) purports to do. If (p.23) true, this is bad news for Goldberg and Zipursky as well as for me. So let me end by explaining some of the project’s importance.

First, it is important in the law. Legally, designating some wrong as a tort is a way of taking quite a significant body of doctrine off the shelf and applying it to that wrong. Not only do we import the right to reparative damages, with all of its implications; we also import, as McLachlin J suggests, various standard common law doctrines about causation, remoteness, quantification, mitigation of damage, exclusion and limitation of liability, and so on. It is a big question, in drafting legislation that creates new civil wrongs, and in interpreting such legislation, whether to make the new wrongs part of the law of torts (with all of these potential common law implications) or whether to make them freestanding.55 You may say that if this is true then I should add various other defining features to my list (a) to (f) above. I should add criteria about causation, remoteness, quantification, mitigation, etc. But that doesn’t follow. Not everything that is a legal implication of making something a tort is by that token a defining feature of the law of torts. There are some legal implications that could be present in one legal system but absent in another, although both have a law of torts. There are also some implications that could change over time without the law of torts being thereby abandoned or ‘lost sight of’. That is important, because when legislation designates some wrong as a tort, one consequence is that the legal treatment of the wrong will tend to vary with future changes in the law of torts as a whole (i.e. changes in its standard common law doctrines). There would come a point, however, at which such future changes would destabilize the category, by removing defining features in such a way as to make the decision to designate something as a tort unintelligible to those who come afterwards. Perhaps that has happened in some jurisdictions in the United States? That would help to explain the otherwise inexplicable rise of the view—which seems illiterate to those of us (p.24) trained in the law of other common law jurisdictions—that the law of torts is somehow reimaginable as ‘the law of accidents’.

You will notice that I have just opened up some space for Goldberg and Zipursky to resist my charge of equivocation about the law. ‘True,’ they might say, ‘reparative damages are the only remedy available as of right in the common law of torts. We never meant to deny it. But it does not follow that this is a defining feature of the law of torts. Some legal implications of making something a tort, as you say, are not defining features of the law of torts.’ This is a possible line of response. I have tried, however, to pre-empt it in this paper by suggesting that if we do not include the right to reparative damages as one of its defining features, the law of torts is insufficiently differentiated from some nearby branches of law, notably but not only the law of equitable wrongs. Goldberg and Zipursky may reply that this is a question-begging enterprise. It is an open question how much differentiation one needs, or at least how much differentiation one should expect to achieve before concluding that the rest is an accident of history. I agree. In a sense, indeed, everything we are discussing is an accident of history. The law could have developed in other ways and we could have ended up with differently differentiated areas of law. There need never have been a law of torts; it is an accident of history that there is, and that there is a law of contract, of trusts, of unjust enrichment, etc. All I would add is that the law of torts, for all its happenstance, has not simply bequeathed a more or less miscellaneous list of torts. It has also bequeathed a more or less miscellaneous list of features in virtue of which, according to the law, these qualify as torts, a list which I have tried to summarize as (a) to (f). Why add ‘accident of history’ to the list? The list itself is the accident of history.

This brings us to a second reason, related to the first, for stopping to identify the defining features of tort law. As areas of law take shape, they also take root. Each develops its own way of differentiating itself and its own related way of accounting for itself, a rationale, or more often a set of connected but partly competing rationales, for its own existence. There are then, as it were, local narratives available to the courts in justifying particular features of the law in that area, as well as in justifying changes to those features of the law. Sometimes these changes include changes to (p.25) the very criteria by which the relevant area of the law is identified, and thus differentiated from other nearby areas. There is no timelessness and no inevitability to these differentiations. But that is not to say that they can be casually dispensed with. You may think that if some rational resources are good for the law’s development, then more rational resources must be better, so that breaking down the boundaries between areas of law can only increase the richness of the available narratives, and hence make for law that is increasingly responsive to reality, or decreasingly cut off from it. Not so. Narrative coherence is not scalable.56 Growing indifferent to the established legal differences between torts and equitable wrongs, or becoming rationalistically fixated with harmonizing contract with tort, is apt to create an incomprehensible brouhaha: not a more unified narrative, not a bigger choice of narratives, but no intelligible narrative at all. And with no narrative, no future for development of the law by legal argumentation.

This is not a manifesto for keeping the law as it is. The law of torts is the common law’s main offering for dealing with quite a wide range of social problems and personal conflicts. We may reasonably wonder whether it is (still) up to the task of dealing with them, or whether it needs to be reformed or replaced. Before we can work out whether it is up to the task, however, we need to know what exactly it is. This brings us to a third reason for caring about the defining features of tort law. How can there be tort law reform if we have lost sight of what tort law is? If politicians and pundits think it is ‘the law of accidents’ won’t they just run amuck with their so-called reform proposals, aiming at the wrong targets, firing in crazy directions, and doing more collateral damage to the law in the process than will ever be compensated by any improvements that they may secure? Yes, they will, and they do. Many well-meaning people proposing improvements to the law of torts fall at the very first hurdle of not having much idea what the law of torts is, maybe thinking of it as another name for ‘compensation culture’ or ‘ambulance-chasing lawyers’ or ‘class-action racketeering’. I do not suggest for one moment that Goldberg and Zipursky’s conceptual work, let alone my own, will help to (p.26) solve this problem. Possibly the people who harbour such wild misconceptions are not much interested in being put right, or do not have the time or inclination or patience to read the scholarly literature, or even to have it explained to them. But scholarly writing about the law does not have to justify itself by showing its actual influence on anybody. It has to justify itself by showing that its influence would be good if, by chance, it had any. Subject to the narrow criticisms I have made in this paper, Goldberg and Zipursky’s joint work on tort law amply meets this condition.


(1) John C P Goldberg and Benjamin C Zipursky, ‘Torts as Wrongs’, Texas Law Review 88 (2010), 917 at 919. I am relying principally on this article as a conspectus of Goldberg’s and Zipursky’s views, since it is a recent restatement. They have openly changed their minds over time on a number of points, so trawling through their extensive oeuvre may sometimes lead one astray. Having said that, I will also refer in §V below to Zipursky’s important sole-authored article ‘Civil Recourse, Not Corrective Justice’, Georgetown Law Journal 91 (2002), 695 which deepens the theoretical foundations.

(2) ‘Torts as Wrongs’, above note 1, e.g. at 946.

(3) Some statutory causes of action are sui generis and hard to classify as public or private, so I am hedging my bets with ‘most if not all’. For example, the cause of action against public authorities under the Human Rights Act 1998 s7, is not a cause of action in tort, and ‘damages’ under s8 are not to be measured on tort-law principles: R v Secretary of State for the Home Department ex parte Greenfield [2005] UKHL 14, [2005] 1 WLR 673. ‘The 1998 Act,’ said Lord Bingham, ‘is not a tort statute’ (para 19). Nevertheless, there is civil recourse under the Act in the Goldberg-Zipursky sense. For similar examples and dicta see Maharaj v Attorney-General of Trinidad and Tobago (No 2) [1979] AC 385, 399; Kearney v Minister of Justice [1986] IR 116; Simpson v Attorney-General [1994] 3 NZLR 667; Taunoa v Attorney General [2007] NZSC 70, [2008] 1 NZLR 429. But compare Crossman v R (1984) 9 DLR (4th) 588, where a violation of a Canadian Charter right is classed and treated as a tort. Goldberg and Zipursky lead us to believe that the latter is also the Federal United States position (‘Torts as Wrongs’, above note 1, 939) although the leading case of Bivens v Six Unknown Federal Narcotics Agents 403 US 388 (1971) seems to me to point the other way. The Goldberg and Zipursky talk of ‘constitutional torts’ in this context may, of course, incorporate their own (I will suggest insufficiently discriminating) analysis of what a tort is.

(4) Goldberg and Zipursky, ‘Torts as Wrongs’, above note 1, e.g. 918.

(5) In this chapter I will mainly speak of ‘duty’ but I could equally have chosen ‘obligation’, which means exactly the same.

(6) For painstaking analysis, see Peter Birks, Unjust Enrichment (2nd ed, Oxford 2005), 11–38.

(7) Oliver Wendell Holmes, ‘The Path of the Law’, Harvard Law Review 10 (1897), 457 at 462. I should mention that, although I repeat it here, the attribution of this view to Holmes himself is contested: Joseph M. Perillo, ‘Misreading Oliver Wendell Holmes on Efficient Breach and Tortious Interference’, Fordham Law Review 68 (2000), 1085.

(8) John C P Goldberg, Anthony J Sebok, and Benjamin C Zipursky, Tort Law: Responsibilities and Redress (2nd ed, New York 2008), at 783–90, discussing Vincent v Lake Erie Transportation Co 124 NW 221 (Minn. 1910). The discussion is cited approvingly in ‘Torts as Wrongs’, above note 1, at 953.

(9) ‘Torts as Wrongs’, above note 1, 951–2, discussing Rylands v Fletcher (1868) LR 3 HL 330. Personally, I would tend to class Rylands with Vincent as not a counterexample at all. See chapter 5 below for explanation.

(10) Particularly effective exposés of the Holmesian mistake in the breach-of-contract context are Ian R Macneil, ‘Efficient Breaches of Contract: Circles in the Sky’, Virginia Law Review 68 (1982), 947 (where a useful catalogue of economistic true-believers may be found), and Daniel Friedmann, ‘The Efficient Breach Fallacy’, Journal of Legal Studies 18 (1989), 1.

(11) ‘Torts as Wrongs’, above note 1, 920 citing W Page Keeton et al, Prosser and Keeton on the Law of Torts (5th ed, St Paul MN 1984), 2. 11

(12) ‘Torts as Wrongs’, above note 1, 920.

(13) Canson Enterprises v Boughton (1991) 85 DLR (4th) 129 at 163. Although McLachlin J was speaking for a concurring minority in Canson, the passage above was quoted and endorsed as ‘good law’ by Lord Browne-Wilkinson in his lead speech in Target Holdings Ltd v Redferns [1996] 1 AC 421, and also found favour at the highest level in (e.g.) Cadbury Schweppes Inc v FBI Foods Ltd (1999) 167 DLR (4th) 577 para 93, and Youyang Pty Ltd v Minter Ellison Morris Fletcher (2003) 212 CLR 484 para 40.

(14) Compare a restitutionary claim against the trustee for the market rental value of the car during the period when it was, so to speak, missing from the trust. The two remedies are compared by Sarah Worthington, ‘Reconsidering Disgorgement for Wrongs’, Modern Law Review 62 (1999), 218. In writing this section, Worthington’s analysis has helped me more than any other. However Worthington’s analysis in turn owes much to James Edelman’s path-breaking Gain-based Damages: Contract, Tort, Equity, and Intellectual Property (Oxford 2002), where the restitution/disgorgement contrast takes pride of place.

(15) A good example is Bishopsgate Investment Management Ltd v Maxwell (No 2) [1994] 1 All ER 261.

(16) For discussion, see P M North, ‘Breach of Confidence: Is There a New Tort?’, Journal of the Society of Public Teachers of Law 12 (1972), 149.

(17) On disgorgement for breach of confidence, see Peter Pan Manufacturing Corp v Corsets Silhouette Ltd [1963] 3 All ER 402; Lac Minerals Ltd v International Corona Resources Ltd [1989] 61 DLR (4th) 14. On disgorgement for breach of fiduciary duty, see Furs Ltd v Tomkies (1936) 54 CLR 583 at 592; Regal (Hastings) Ltd v Gulliver [1967] AC 134; Gwembe Valley Development Co Ltd v Koshy [2003] EWCA Civ 1478; Glazier v Australian Men’s Health (No 2) [2001] NSWSC 6. Note that a trustee dealing in assets can be made liable on the ill-gotten gains theory too, for she too—indeed par excellence—is a fiduciary bound by the duty not to profit. For discussion of the relationship of these two bases of liability see Steven B Elliott and Charles Mitchell, ‘Remedies for Dishonest Assistance’, Modern Law Review 67 (2004), 16 at 31–2.

(18) Sir Peter Millett, ‘Equity’s Place in the Law of Commerce’, Law Quarterly Review 114 (1998), 214 at 225.

(19) This was the key point on which McLachlin J’s concurring minority in Canson differed from the majority. See also Lionel Smith, ‘The Measurement of Compensation Claims Against Trustees and Fiduciaries’ in Elise Bant and Matthew Harding (eds), Exploring Private Law (Cambridge 2010) where McLachlin J is taken to task (I think wrongly) for thinking that ‘a common sense view of causation’ can do without remoteness restrictions.

(20) See Bristol & West Building Society v Mothew [1998] Ch 1 at 17 and Bank of New Zealand v New Zealand Guardian Trust Co Ltd [1999] 1 NZLR 664 at 681 for the suggestion that different equitable wrongs go with different assessment rules, some more like the common law rules than others.

(21) The ‘therefore’ in this sentence is licensed by what I have elsewhere called ‘the continuity thesis’, according to which reasons for not having φ‎d survive one’s φ‎ing and are now reasons for taking remedial measures. See chapter 2 below for further elaboration. As I emphasize there, it would be better still if one’s φ‎ing could have been prevented. Prevention is analytically better, ceteris paribus, than cure. So we should welcome anything that the courts can offer by way of tort-prevention (including injunctions, exemplary damages, etc.). What I am adding in this paper is only that all of this tort-prevention rides opportunistically (and discretionarily) on the coat-tails of the civil recourse that tort law offers as of right to those who have already been wronged, which is a specifically reparative kind of recourse.

(22) ‘Torts as Wrongs’, above note 1, 920.

(23) Ibid, 928, citing a draft of what now appears as chapter 2 below.

(24) Ibid, 947.

(26) ‘Torts as Wrongs’, above note 1, 955.

(27) Attorney-General v Blake (Jonathan Cape Ltd Third Party) [2001] 1 AC 268, a contract case but one in which analogies to tort were suggested obiter.

(28) Law Debenture Trust Corporation v Ural Caspian Oil Corporation Ltd [1995] Ch 152 per Saville LJ at 172 (where the point was of the essence); Attorney-General v Blake, above note 25, per Lord Nicholls at 284–5; Virgin Atlantic v Premium Aircraft [2009] EWCA Civ 1513, [2010] FSR 15 at para 23; Cardile v LED Builders Pty Ltd (1999) 198 CLR 380 at 396; Maggbury Pty Ltd v Hafele Aust Pty Ltd [2001] HCA 70, (2001) 210 CLR 181 at para 76; Roberts v Rodney District Council [2001] 2 NZLR 402; Delta Hotels Limited v Okabe Canada Investments Company Limited [1992] ABCA 176 at para 10.

(29) Under the Chancery Amendment Act 1858, also known as Lord Cairns’ Act. Held to have survived the repeal of the Act in Leeds Industrial Co-operative Society Ltd v Slack [1924] AC 851. You may ask how, other than in being discretionary, these damages in lieu of injunction differ from ordinary reparative damages that the successful plaintiff enjoys as of right. To see what differentiates them, think about cases in which the defendant’s trespass, if allowed to continue, would amount to an expropriation of property (e.g. cases of trespass by building on a neighbour’s land). In such cases injunctions are often granted on the basis that reparative damages would not be an adequate remedy, this being one of the established criteria for the discretionary issue of an injunction. People are free, say the courts, to decide not to alienate their property even if they would be amply compensated for doing so; that freedom is of the essence in distinguishing property rights from some other rights. But where a trespassory building is already in place the courts may be reluctant to order its demolition (which may strike them as too wasteful) and may instead award damages in lieu of the injunction. These are equitable damages and may sometimes be calculated on a disgorgement basis: Wrotham Park Estate Co. Ltd v Parkside Homes Ltd [1974] 1 WLR 798. Notice that this all begins from the thought that these expropriation-like cases are cases in which the primary reparative measure of tort damages is not enough, and that could not be true were reparative damages not the primary measure. This all holds equally true in the other common law jurisdictions to which the 1858 Act was exported: Madden v Kedereski [1983] 1 NSWLR 305; Roberts v Rodney District Council [2001] 2 NZLR 402 (describing the disgorgement measure as ‘anomalous’ in the tort context); Semelhago v Paramadevan [1996] 2 SCR 415.

(30) Kuddus v Chief Constable of Leicestershire Constabulary [2001] UKHL 29; A v Bottrill [2002] UKPC 44; Couch v Attorney-General [2010] NZSC 27; Whiten v Pilot Insurance Co [2002] 1 SCR 595.

(31) Attorney-General v Blake, above note 25, per Lord Nicholls at 284–5. It is arguable that Blake awards, available only where an injunction is denied, are better classified as damages in lieu of an injunction. Be that as it may, there has been little enthusiasm for the decision. Much (award-avoiding) emphasis has been placed on Lord Nicholls’ remark that disgorgement awards for common law wrongs are ‘exceptional’. See Hospitality Group Pty Ltd v Australian Rugby Union Ltd [2001] FCA 1040, (2001) 110 FCR 157; Bank of America Canada v Mutual Trust Co (2002) 211 DLR (4th) 385.

(32) Recent judicial remarks in which this is treated as a trite point not requiring citation of authority include Anufrijeva v Southwark LBC [2003] EWCA Civ 1406, [2004] QB 1124 paras 50, 56; Majrowski v Guy’s and St Thomas’ NHS Trust [2006] UKHL 34, [2007] 1 AC 224 paras 21–2 per Lord Nicholls, para 46 per Lord Hope; Taunoa v Attorney General [2007] NZSC 70, [2008] 1 NZLR 429 paras 252, 318; Commonwealth v Mewett [1997] HCA 29, (1997) 191 CLR 471 per Gummow and Kirby JJ.

(33) Andrew Burrows, ‘We Do This at Common Law but That in Equity’, Oxford Journal of Legal Studies 22 (2002), 1 at 2.

(34) Why would I be the first? Because I am what they call a ‘hard legal positivist’ who thinks that when the law invites courts to judge the ‘reasonableness’ of something (or to use other evaluative criteria) it necessarily invites courts to step outside the law to complete their deliberations, except to the extent that the evaluative criteria have already been unpacked by law into non-evaluative ones. For more detail of this view, see my ‘Justification under Authority’, Canadian Journal of Law and Jurisprudence 23 (2010), 71.

(35) There is a general lesson here about rights, both legal and moral. As Joseph Raz says, ‘[a]ssertions of rights are typically intermediate conclusions in arguments from ultimate values to duties. … [T]he implications of a right … and the duties it grounds, depend on additional premisses and these cannot in principle be wholly determined in advance. … Because of this rights can be ascribed a dynamic character. They are not merely the grounds of existing duties. With changing circumstances they can generate new duties.’ Joseph Raz, The Morality of Freedom (Oxford 1986), 181 and 185–6.

(36) I am thinking of miscellaneous cases including Livingstone v Rawyards Coal Co (1880) LR 5 App Cas 25; Carr-Saunders v Dick McNeil Associates Ltd [1986] 1 WLR 922; Inverugie Investments Ltd v Hackett [1995] 1 WLR 713; Costello v Chief Constable of Derbyshire [2001] EWCA Civ 381, [2001] 1 WLR 1437; Severn Trent Water Authority Ltd v Barnes [2004] EWCA Civ 570.

(37) United Australia Ltd v Barclay’s Bank [1941] AC 1.

(38) See Attorney-General v Blake, above note 25, at 284–5 per Lord Nicholls.

(39) See Strand Electric Co v Brisford Entertainments [1952] 2 QB 246 at 254–5 per Denning LJ; Greenwood v Bennett [1973] QB 195 at 202 per Lord Denning MR; Kuwait Airways Co v Iraqi Airways Co (Nos 4 & 5) [2002] 2 AC 883 paras 67 and 79 per Lord Nicholls.

(40) See Stoke City Council v W & J Wass [1988] 1 WLR 1406 at 1416; Law Reform Committee, 18th Report (Cmnd 4774, London 1971), para 91.

(41) Harvey McGregor, ‘Restitutionary Damages’ in Peter Birks (ed), Wrongs and Remedies in the Twenty-First Century (Oxford 1996), 203: ‘damages for damage’.

(42) For lively discussion (with rival proposals for sorting out the mess) see Edelman, Gain-based Damages, above note 14, at 114–36; Robert Stevens, Torts and Rights (Oxford 2007), 79–84. Stevens suggests that we should think of the damages in most such cases as ‘substitutive’, not reparative, a distinction that is also drawn by Edelman in ‘Money Awards of the Cost of Performance’, Journal of Equity 4 (2010), 122 using the terms ‘substitutive compensation’ and ‘reparative compensation’, and crediting the distinction to Steven Elliott. I regard Elliott’s proposed ‘substitution’ measure of compensation, not as a rival to the reparative measure, but as just one among many possible reparative measures, the choice among which depends on the logic of ‘next-best conformity’ as sketched in my ‘What is Tort Law For?’, chapter 2. (On my view, in summary, substitution stands to reparation as terrier stands to dog. Reparation in turn stands to correction as dog stands to canine.)

(43) ‘Torts as Wrongs’, above note 1, 955.

(44) ‘Civil Recourse, Not Corrective Justice’, above note 1, 719, cited with apparent approval in Goldberg and Zipursky, ‘Torts as Wrongs’, above note 1, at 954.

(45) ‘Civil Recourse, Not Corrective Justice’, above note 1, 698: ‘tort judgments represent liabilities, not duties to pay’.

(46) Which brings with it a new raft of remedies, such as attachment of earnings and seizure of assets.

(47) ‘Civil Recourse, Not Corrective Justice’, above note 1, 720: a tortfeasor does not have ‘a freestanding legal obligation to pay independent of any action against her’.

(48) In England and Wales, statutory interest may (and in some circumstances must) be awarded on damages awards for the period between the cause of action arising and the award of damages by the court, on the basis that ‘a judgment against the defendant means that he should have admitted the claim when it was made and have paid the appropriate sum as damages’: Law Revision Committee, Second Interim Report (Cmd 4546, London 1934). This rationale was approved by the Court of Appeal in Jefford v Gee [1970] 2 QB 130 at 144 and associated with earlier (common law) dicta on the same point. On this view an award of tort damages places a retrospective legal duty on the defendant, a duty to have paid the award before it was awarded. Consider, in broad support of this view, McLachlin J’s dictum in Canson, above note 13, that ‘viewing the issue of compensation from the date of the breach [is] required at common law’ (but not, she adds, in equity, which looks at losses instead from the date of the trial).

(49) See e.g. Jules Coleman, ‘Tort Law and the Demands of Corrective Justice’, Indiana Law Review 67 (1992), 349.

(50) See e.g. Ernest Weinrib, ‘Understanding Tort Law’, Valparaiso Law Review 23 (1989), 485.

(51) See e.g. Stephen Perry, ‘The Moral Foundations of Tort Law’, Iowa Law Review 77 (1992), 449.

(52) See e.g. Arthur Ripstein, ‘As if It Had Never Happened’, William and Mary Law Review 48 (2007), 1957.

(53) Zipursky may well agree. See his critical remarks on the culture of punitive damages in his ‘Civil Recourse, Not Corrective Justice’, above note 1, 751–2.

(54) Although the details cannot be entertained here, my explanation would link the distinction between breach of contract and tort to the source of the obligations breached (do they or don’t they have their source in a contract?), rather than to the remedies available for the breach.

(55) A plum example is Simpson v Attorney-General [1994] 3 NZLR 667 where, had it been classified as an action in tort, an action for redress against a public authority under the New Zealand Bill of Rights Act 1990 would have failed owing to crown immunity under the Crown Proceedings Act 1950.

(56) For useful reflections, see Barbara Levenbook, ‘The Role of Coherence in Legal Reasoning’, Law and Philosophy 3 (1984), 355.