Examining the Structure of Remedial Law

Conventionally, a judicial private law remedy (JPLR) is understood as a court order made following two types of events: a violation of a legally recognised right, or a threatened violation of a legally recognised right.1 When there is a ‘rights-violation’, the prevailing view is that the innocent party obtains (at least) a ‘secondary’ remedial right against the defendant to be placed in as near position as possible as if the ‘primary’ right had not been violated.2 When there is threatened rights violation, the innocent party may receive a power to obtain a judicial ruling directing the defendant to comply with its correlative ‘primary’ duty.3 On this view, the principal reason that most judicial JPLRs are awarded is to declare, or provide defendants with further reasons to comply with, the ‘primary’ or ‘secondary’ legal duties that defendants already owe to claimants. In Rights, Wrongs, and Injustices: The Structure of Remedial Law4 (Structure), Stephen Smith rejects this view of JPLRs. Smith argues, first, that although certain JPLRs are made on the ‘grounds’ of threatened rights violations, many of the JPLRs that courts issue are instead made on the ‘grounds’ of ‘wrongs’ or ‘injustices’. Secondly, Smith argues that the idea of ‘secondary legal duties’ is largely5 misplaced and that, prior to a judicial ruling, rights violations only give rise to legal liabilities. The book, which is the culmination of many years’

work, 6 is an original and thought-provoking work of scholarship demanding serious engagement.
Structure is replete with arguments deserving close attention, ranging from broad claims about the structure and philosophical foundations of remedial law to more specific claims about the nature of various judicial 'orders', the distinctiveness of 'legal' and 'equitable' judicial 'orders', and a thought-provoking final chapter on 'defences'. This review is inevitably highly selective; its focus being on Smith's classification of the 'grounds' upon which courts issue JPLRs -'rights-threats', 'wrongs', and 'injustices' -and the specific remedies that Smith places within these categories. 7 Detailed attention is given to these issues because Smith's classification of the 'grounds' upon which courts issue JPLRs is both novel and central to his interpretation of the 'structure' of 'remedial law'.
This review makes three principal claims:first,the reasons that Smith provides against the existence of a 'secondary' legal duty to pay 'compensatory damages' are unpersuasive; second, the taxonomy of damages awards that Smith advances should largely be rejected; and third, Smith's argument that all restitutionary orders following 'defective transfers' are made on the 'grounds' of 'injustices' is doubtful. As explained below, these criticisms substantially undermine the 'structure' of 'remedial law' that Smith defends.

METHODOLOGY AND SUBJECT
Smith's stated aim is to provide an explanatory theory of the scope, foundations, and structure of 'remedial law' in common law jurisdictions. 8 He pursues this task by addressing four main questions: (1) the content of remedial law; (2) why courts issue remedies at all; (3) the specific 'grounds' upon which remedies are issued; and (4) how the remedies issued relate to the 'substantive law'. 9 Regarding the first question, Smith says that he is not concerned to reform or evaluate the law, 'but simply to understand it'. 10 In relation to the second question, Smith claims that explanatory theories of remedies can be categorised according to whether and how they address two questions:the 'normative'question of what principles or values underlie the law; and the 'analytic' question of how remedies relate to other parts of the law. 11 Smith's specific focus is the relationship between 'remedial' and 'substantive' law. By 'remedial law', he means the legal rules relating to when courts issue 'directive rulings' as part of (2021) OJLS (forthcoming). 8 Structure n 4 above, ch 1. 9 ibid, 1-6. 10 ibid, 11. 11 ibid, 25-26. the resolution, whether permanent or temporary, of a private law dispute. 12 By 'substantive law', Smith means the legal rules determining the (primary) legal duties persons owe independently of any judicial event. 13 A 'directive ruling' or 'order' is a judicial order made as part of the resolution of a private law action directing a defendant to act or refrain from acting. 14 To understand the relationship between substantive and remedial law, Smith focuses on the 'grounds' upon which remedies are ordered and on how the rights and duties that JPLRs impose relate to the prior right-duty relations between claimants and defendants. In addressing these two issues, he discerns a 'structure' to remedial law based upon the reasons why courts make certain kinds of rulings and how the rights and duties these rulings impose differ from or resemble,the substantive law right-duty relations between claimants and defendants.

SMITH'S PROPOSED STRUCTURE
According to Smith, courts issue JPLRs on three distinct 'grounds': 'rightsthreats', 'wrongs', and 'injustices'. 15 By 'grounds', Smith claims to denote the 'summary descriptions of the facts that claimants must establish to obtain remedies'. 16 But Smith also occasionally refers to the 'grounds' for a remedy to denote the reason(s) for that remedy being awarded. 17 This equivocation arguably reveals an ambiguity as to precisely what Smith is seeking to explain. 18 As noted above,this view departs from previous expositions for why JPLRs are awarded. 19

Rights-threats
According to Smith, a 'rights-threat' occurs 'whenever it is likely that, if the defendant's plans do not change' the defendant will infringe, or continue to infringe, one of the claimant's 'substantive rights'. 20 A 'substantive right' is a legal right that a person possesses either because they reside in a jurisdiction where such rights exist generally or because an 'ordinary' right-creating event has occurred. 21 Such rights can be contrasted with 'remedial rights', which are rights 12 ibid, 6-9. For Smith, 'a remedy is a judicial ruling' and a ruling is a 'legally operative judicial pronouncement': ibid, 6, 19. 13 ibid, 2, 7. 14 ibid, 14. We shall refer to 'direct rulings' as 'orders'. arising from a judicial event, such as the making of an order. 22 Smith claims that on proof of a rights-threat, courts normally make 'replicative' orders, 23 which are orders directing a defendant to do what they already had a substantive legal duty to do vis-à-vis the claimant. 24 Smith argues that 'replicative orders' made on the grounds of 'rights-threats' include: orders to pay sums due under a contract, ejectment, delivery up, certain injunctions, specific performance, and 'substitutionary damages' awards, which he conceives as 'monetary substitutes for specific relief '. 25 The reason courts make 'replicative orders' in response to 'right-threats' is that if a defendant is unwilling to comply with a substantive duty owed to a claimant, the reasons that the substantive law has provided the defendant for not violating the claimant's correlative right must be insufficient. 26 Courts, therefore, make 'replicative orders' on the 'grounds' of 'rights-threats' to provide defendants with further reason to respect the claimant's threatened right.
The usual way that claimants establish a 'rights-threat'is by demonstrating the defendant's engagement in an ongoing rights-violation, which courts properly regard as evidence that the defendant is unwilling to respect the claimant's substantive rights. But although defendants engaged in ongoing rights-violations have committed a legal wrong, Smith argues that 'replicative orders are not responses to wrongs qua wrongs.' The object of an order to, for example, not trespass on the claimant's land is to ensure that, by providing the defendant with a further reason for complying with their substantive duty not to trespass, a defendant does not in future (or continue to) trespass on the claimant's land.
Smith provides three main reasons supporting his claim that 'replicative orders' are made on the grounds of 'rights-threats' rather than 'wrongs'. 27 First, courts may issue 'replicative orders' notwithstanding that a defendant is yet to have breached any substantive duty it owes to a claimant. 28 For example, on proof that a defendant is imminently intending to infringe a restrictive covenant, a court may issue a quia timet injunction directing the defendant to comply with the covenant. This shows that the occurrence of a wrong is not a necessary precondition to a 'replicative order' being made.
Secondly, courts generally refuse to issue 'replicative orders' where a defendant has committed a wrong, but the wrong is unlikely to be repeated. 29 For example, an injunction is unlikely to be issued in response to a one-off trespass. This shows that the commission of a wrong is not sufficient for the making of a 'replicative order'.
Finally,it is difficult,says Smith,to think of any reason that could justify courts issuing 'replicative orders' other than that the claimant's right was threatened. 30 He suggests, contrary to the view that he attributes to Birks, that the 'the cause of action for 'replicative orders' is simply proof that the defendant has a substantive duty to do what the order requires', cannot account for, among other things, quia timet orders or the fact that courts will refuse to order defendants to comply with negative contractual duties merely on proof that they exist. More fundamentally, however, the Birksian view 'supposes that courts will issue orders regardless of whether this serves any practical purpose'. 31

Wrongs
The second 'ground' upon which courts issue JPLRs is, Smith argues, that a 'wrong' has occurred, which is where 'something that [the law says] ought not to happen', happens. 32 Although Smith accepts that most private law orders are directed at persons who have committed wrongs, he argues that it does not necessarily follow that when a remedy is directed at a wrongdoer the remedy is issued on the 'grounds' of the wrong. 33 So, for example, Smith claims that whilst compensatory damages are generally conditional upon a wrong's occurrence, the wrong is not the reason why such awards are made. 34 Orders made by courts merely on proof of a wrong's occurrence are, according to Smith, roughly the private law equivalent of criminal punishment since, like criminal punishment, these orders are made in response to a wrong without any need for the victim to prove that 'consequential loss' has been suffered. 35 Furthermore, JPLRs that are ordered on the 'grounds' of wrongs are, like criminal punishments, set at a monetary figure that is ultimately 'a matter of choice and convention'. 36 On proof of 'wrongs', courts respond by making 'creative orders', 37 which bring rights into existence that are different from the 'substantive rights' violated. 38 According to Smith, examples of 'creative orders' made on the 'grounds' of wrongs are orders to pay exemplary, nominal, and 'vindicatory' damages. 39 Significantly, the final category includes 'user damages', 40 'market-price damages', 41 'gain-based damages', 42 and 'damages for wrongs to the person'. 43 Contrary to orthodoxy, Smith claims that these awards are not responses to 'right threats' because there is no substantive legal duty to pay damages before the order is made.

Injustices
A novel claim made by Smith is that there is a third 'ground' upon which courts issue JPLRs: 'injustices', which is 'an unfair loss or gain arising from a transaction between the claimant and defendant'. 44 More accurately, an 'injustice' is an action or state of affairs that is unfair 'because a loss or gain … has been unfairly allocated, distributed, or allowed to persist', 45 though Smith leaves open what makes a gain or loss 'unfair'. On proof of an 'injustice', courts respond by making 'creative orders', 46 the content of which depends upon the specific 'injustice' in question. The purpose of such orders is to correct the 'injustice'.
Smith claims that there are two main remedies that courts issue on the 'ground' of 'injustices': 47 orders to make restitution for 'defective transfers' and orders to pay damages for 'consequential losses'. Smith understands the latter to include balance-sheet losses caused by a breach of duty and certain 'cost-ofcure' awards following contractual breach. 48 As discussed below, Smith's classification of restitutionary awards and damages awards for 'consequential losses' as injustice-responding orders rests primarily on two observations: first, that restitutionary awards for 'defective transfers' and damages awards for 'consequential losses' are orders creating new duties; and secondly, that the 'ground' for such orders is neither 'rights-threats' nor 'wrongs'. 49

UNDERSTANDING THE LAW OF DAMAGES
From the perspective of 'remedial law', the central question posed by the law of damages,claims Smith,is whether orders to pay damages 'confirm [pre-existing] substantive duties to pay damages … [or] create new duties'. 50 This is because if 'duties to pay damages only arise once courts order defendants to pay … then the entirety of damages law is part of remedial law'. 51 For Smith, the distinction between the 'duty-confirming view' and 'duty-creating view' of orders to pay damages is therefore fundamental. This makes examining his reasons for favouring the 'duty-creating view'essential to assessing his account.This section undertakes that task. The discussion commences by considering Smith's claim that the positive law in common law jurisdictions does not recognise a prejudgment legal duty to pay damages. Next, Smith's specific objections against the most plausible theoretical explanation for the 'duty-confirming view', 'the continuity thesis', are examined. Finally, we consider an important implication of Smith's preference for the 'duty-creating view' of damages awards: his proposed classification of different kinds of damages awards. 6 Is there a pre-judgment legal duty to pay damages?
It is conventionally thought that orders to pay damages enforce a secondary legal duty of reparation that arises upon the breach of a primary legal duty. This is what Smith calls the 'duty-confirming view' of damages awards. In Structure, Smith reiterates certain arguments he has previously advanced against the duty-confirming view. Briefly, these arguments are: (1) that common law judges typically, though not exclusively, speak in terms of wrongdoers being 'liable' to pay damages as opposed to having pre-judgment 'duties' to do so; 52 (2) that a failure to pay damages until ordered to do so by a court is never a source of a further duty (or liability) to pay damages; 53 (3) that paying damages prior to the making of such a court-order is no defence to a claim for damages; 54 and (4) that the existence and content of any such duty cannot be determined in advance of a court order compelling payment. 55 Smith's thesis has some prominent supporters. Goldberg and Zipursky, for example, have argued that a defendant's liability to pay damages upon the commission of a civil wrong is not grounded in any legal (or moral) duty arising upon the violation of a primary right; rather it is imposed by the court via a principle of 'civil recourse'. 56 By contrast, Penner and Quek have argued that while Smith is correct that 'there is no legal duty on a wrongdoer to pay … damages to the victim of his wrong prior to any court order … a wrongdoer does owe her victim a moral duty of repair', which the authors characterise 'as a Kantian duty of virtue'. 57 More recently, Gardner, 58 and Steel and Stevens, writing together, 59 have issued forceful responses to earlier versions of Smith's arguments against the 'duty-confirming view'. These responses cannot be comprehensively examined here, but three specific observations are worth making. First, although, as Gardner explains, Smith presents his objections as 'doctrinal', he generally uses 'the black-letter law … to generate something more like a philosophical objection', 60 indicating that the dispute is at least partially theoretical. For example, Smith's concern that the content of the alleged duty to pay damages cannot be determined in advance of a court-order and is therefore 'unknowable' seems to rest on a deeper conviction that the law (generally) upholds the maxim that ought implies can. However, as Gardner, and Steel and Stevens, explain, private law does not conform to this principle; the objective negligence standard perhaps providing the clearest example. Moreover, even if Smith's concern here is more about retrospectivity than indeterminacy, Gardner  retroactive legal duties do not conform to the ideal of the rule of law, law failing to conform to this ideal 'is no surprise and no rarity'. 61 Secondly, some of Smith's doctrinal claims are inaccurate or misleading. For example, Smith highlights the fact that a failure to pay damages until ordered to do so by a court is never a source of a further duty (or liability) to pay damages. But Steel and Stevens provide good reason to downplay the significance of this, noting that this may ultimately be attributable to the way obligations are individuated. 62 Moreover, as they (and Gardner) also note, in some common law jurisdictions, interest payable on damages awards is calculated from the date when the wrong occurred rather than the date of judgment, 63 indicating that the black-letter law is more equivocal than Smith admits. Smith also claims that paying damages prior to the making of such a court-order is no defence to a claim for damages, citing the Court of Appeal's decision in Edmunds v Lloyds Italico & l'Ancora Compagnia di Assicurazione e Riassicurazione SpA. 64 But as Steel and Stevens explain, Smith has misinterpreted this case because, properly understood, it stands only for the narrower proposition that an obligation to pay money requires the payee's co-operation to fulfil. 65 Finally, responding to Smith's observation that judges often speak in terms of a 'liability' to pay damages, Stevens and Steel correctly observe that this is explicable on the basis that, following breach, the wrongdoer has both a duty and a liability to pay damages. Since legal duties are not always enforceable (for example an unenforceable contractual obligation), it is necessary that the beneficiary of the duty also have a Hohfeldian 'power' to enable enforcement (via a court order) of the duty they are owed. 66 Gardner makes essentially the same point, stating that there is actually 'no such thing as a liability to pay damages' as opposed to 'a liability to be required to pay (a specified sum in) damages'. 67 All this provides good reason to doubt Smith's claim that the positive law does not recognise a pre-judgment legal duty to pay damages. However, one difficulty in assessing Smith's view is that he does not clearly identify what he regards as the necessary conditions for a duty to be 'legal'. It is therefore likely that at least some of the disagreement between Smith and others is attributable to holding different conceptions of a legal duty. But whatever conception of legal duties one adopts, Smith's claim that there is no pre-judgment legal duty to pay damages is questionable because it suggests that the breach of a legal duty may discharge the wrongdoer from the need to conform to the reasons that justified that duty, which is implausible. The (predominately) doctrinal dispute just described reflects a deeper disagreement between proponents of corrective justice as an explanation for damages awards, like Gardner, Steel and Stevens, and others, like Smith, who reject this explanation. The core of this disagreement revolves around the plausibility of the explanation for damages awards that Gardner labels 'the continuity thesis', 68 which Smith admits provides the 'most influential theoretical argument for the duty-confirming view'. 69 In Structure, Smith defines 'the continuity thesis' in broad terms as the view that: 'when we breach a substantive duty, the duty (or the reasons underlying the duty) continues in force, albeit partly or wholly in the form of (or giving rise to) a duty to pay damages'. 70 Smith claims that while this thesis 'presents a prima facie compelling prescriptive theory of damages', 71 certain 'conceptual' and 'doctrinal' objections demonstrate its falsity.Hence,a more fundamental objection Smith raises against the 'duty-confirming view' is that its most plausible version(s) are based on a false premise. Given what was said above, the persuasiveness of Smith's critique of the 'duty-confirming view' therefore ultimately appears to rest on the persuasiveness of his arguments against the continuity thesis.

Which Version of the Continuity Thesis?
The formulation of the continuity thesis that Smith adopts is designed to remain neutral between the two main rival versions of the 'duty-confirming view', which are that the primary duty breached itself continues after the breach; 72 and that the reasons for the primary duty persist following breach, helping to ground a new 'secondary' duty to pay damages. 73 Since Smith aims to show why all 'duty-confirming' views are invalid, it is unsurprising that he does not investigate the relative merits of these rivals. However, when properly understood, at least the second view, and possibly the first, 74 is capable of defusing Smith's objections.

Smith's Conceptual Objection to the Continuity Thesis
The 'primary' objection Smith raises against the continuity thesis is the 'conceptual' one that 'the duties that … duties to pay damages ceased to exist before they could be transformed'. 75 Smith cites the termination of a contract upon serious breach to support this claim. 76 But, as he acknowledges, other than when performance has become 'futile', 77 serious breach alone normally does not terminate a contract; the promisee must 'elect' to do so. Smith's real concern must therefore be that 'it is not clear how' the promisee's decision to terminate a contract can 'transform [the defendant's primary] duty into a duty to pay damages'. 78 In reality, there is no great mystery here. Following breach, the law of contract is generally concerned to maximise the innocent party's autonomy subject to this not unjustifiably prejudicing the breaching party. Upon the occurrence of a sufficiently serious breach, it is therefore unsurprising that the innocent promisee can choose between insisting upon performance, accompanied by the option to claim damages for recoverable loss, and electing to terminate the parties' reciprocal arrangement and accept the outstanding performance in the form of a monetary substitute. 79 For Smith, this second possibility is conceptually puzzling, but its solution can be found in the version of the continuity thesis favoured by Gardner, drawing on earlier work by Raz. 80 To explain further, in addition to converging upon a particular version of the continuity thesis, both Raz and Gardner adopt a specific conception of what a legal (and moral) duty actually is. For these authors, to say that one owes another a duty means that the beneficiary of that duty has an interest of sufficient strength to provide both a reason for the duty-bearer to protect and promote that interest and a reason not to act for at least some reasons that may count against protecting or promoting that interest. 81 Importantly, Gardner also explains that in determining what action(s) the breaching party must undertake following a breach of duty, new considerations 'may countervail' and assert normative force. 82 It is accordingly the (normatively significant) events of the breach's occurrence and, where choice is available, the promisee's decision to seek a monetary substitute in lieu of actual performance that,together with those reasons that justified the original duty and still await 'next-best satisfaction', 83 ground the new 'secondary' duty to pay damages.

Smith's Doctrinal Objections to the Continuity Thesis
Smith's conceptual objection to the continuity thesis thus dissolves upon a proper understanding of Gardner's account. But Smith also raises the 'doctrinal' objection against the continuity thesis that it 'is inconsistent with the rules governing the assessment of compensatory damages'. 84 The 'main inconsistency' here, claims Smith: 'arises in respect of the various rules that limit defendants' liability for consequential losses, such as mitigation, contributory negligence, collateral benefits, exoneration of liability clauses (limitation and exclusion clauses), and "remoteness"'. 85 Smith believes that these rules are 'inconsistent with the continuity thesis' because: 'if paying damages is meant to make the world as if the defendant had performed its original duty, the fact that the claimant was careless or could have avoided the loss or, more generally, the loss was not reasonably foreseeable, seems irrelevant'. 86 This might initially seem like a forceful objection against the continuity thesis, but it is not because it misinterprets the idea of 'next-best conformity' with the primary duty. 87 The content of the reparative duty grounded by the continuity thesis following a breach of legal duty is not to put the victim of the breach in exactly the same factual position as if the breach had not occurred. To see why this is so, consider that the defendant's breach of duty may influence numerous later decisions of the claimant ranging from whom to contract with to what modes of transportation to engage in, with all the attendant further consequences that may flow from these decisions. Even limiting our concern only to those consequences that have accrued by the time of trial, 88 Smith cannot be suggesting that, when assessing damages, the court must take an exhaustive ledger of all such costs and benefits no matter how tenuous the causal connection?
Smith's error here involves misunderstanding what the continuity thesis actually entails. The continuity thesis does not require the breaching party to reconstitute the innocent party's non-breach factual position in every possible respect, but instead to achieve 'next-best conformity' with the reasons that justified the relevant primary duty subject to the constraints imposed by any new considerations now exerting normative force. 89 Put another way, what the idea of 'next-best conformity' entails is not 'next-best' factual conformity with the claimant's non-breach position, but rather giving effect via a secondary duty to pay damages, so far as can now be justified, to the reasons grounding the primary legal duty. Although this typically involves making good some of the breach's adverse consequences for the innocent party, it does not necessarily require making good all such consequences and may even have the effect of placing the innocent party into a better factual position than it would have been in had the breach not occurred.
The truth of this final observation is demonstrated by cases where a postbreach causally related benefit of the breach accrues to the innocent party and is Decisions of this kind are controversial and there is scope for reasonable disagreement regarding precisely when a post-breach benefit should be considered sufficiently 'collateral' to the breach to be disregarded for the purposes of damages assessment. But unless one takes the view that post-breach benefits that, no matter how tenuously related to the breach's occurrence, would not have accrued to the innocent party 'but for' the breach should always reduce the sum awarded, a damages award will sometimes result in the innocent party being placed into a better financial position than if the breach had not occurred.
Smith rightly observes that there is a puzzle here, but he draws the wrong conclusion from these cases, claiming that, like rules limiting damages awards, the 'collateral benefits' doctrine demonstrates the falsity of the continuity thesis. 93 He also 'more tentatively' suggests that the continuity thesis 'has difficulty explaining compensation for lost profits or incidental out-of-pocket expenses'. 94 However, Smith's simultaneous objection to the continuity thesis for failing to explain why certain consequential losses are recoverable and the fact that not all consequential losses are recoverable is puzzling. It is true that on a purely factual interpretation of what the continuity thesis entails, any rules having the effect of placing the innocent party into a worse (or better) factual position than had the breach not occurred are problematic. But on this (incorrect) interpretation of the continuity thesis, one aspect of the law not requiring explanation is the availability of 'compensation for lost profits or incidental out-of-pocket expenses'. 95 Thus, assuming that the aforementioned 'collateral benefits' decisions can be justified, what they really demonstrate is that Smith's interpretation of what the continuity thesis entails is incorrect. Rather than aiming to place the victim of the breach in exactly the same factual position as if the wrong had not replacement costs for a thing or service that was destroyed, damaged, lost, wrongly retained, or not provided as promised', 189. The answer is that doing the next-best thing to complying with the original duty normally also requires making good certain adverse consequences resulting from the original non-compliance. Admittedly, and as discussed further below, the precise limits the law draws are far from self-explanatory. occurred, the continuity thesis seeks to achieve 'next-best conformity' with the reasons justifying the primary duty in the way that best accommodates those new considerations now exerting normative force. Some have argued that in both the contractual and tortious contexts this generally involves substituting for the primary right infringed and making good those adverse consequences of the breach falling within the scope of the breaching party's reparative responsibility. 96 But one need not endorse such an account to support (a version of) the continuity thesis. 97 To elaborate, when determining the scope of a breaching party's compensatory liability, it is not only the content of the relevant primary duty that has normative salience. For example, that the victim of the breach could have taken reasonable steps to reduce the adverse consequences of the breach is relevant in determining whether such adverse consequences are now properly the responsibility of the wrongdoer or the victim. A fortiori, the fact that the victim of the breach has, by their careless behaviour, exacerbated the adverse consequences of the breach is relevant to determining how far the wrongdoer's reparative responsibility extends. Additionally, the fact that the victim assented to a clause limiting the scope of the other party's compensatory liability for breach will be relevant in determining what the breaching party's liability is, albeit that it is clearly necessary to have rules, as all sophisticated modern legal systems do, to determine whether such assent was voluntarily made.
More difficult to explain are rules of 'remoteness' in both contract law and the law of torts. Focussing on contract law, 98 it is, as Smith observes, puzzling why 'reasonable foreseeability' at contract formation provides the sole criterion for determining the boundary of a breaching party's reparative responsibility. One possibility is that the extent of this responsibility should depend, at least in part, upon the nature of the breach. 99 A more compelling view is that the Hadley v Baxendale test of contractual remoteness 100 is just a proxy for some other principle (or set of principles) that demarcates the adverse consequences of a contractual breach that the breaching party can justifiably be held liable for. Stevens has recently advanced this view, 101 though it may be that he has not fully captured all of the complexity present. In addition to whether the particular adverse consequence suffered was within the purpose of the duty breached, 102 it is arguable that the law is concerned with the nature and degree of the causal link between the breach and the particular adverse consequence 96  suffered, 103 and, perhaps amongst other relevant considerations, the degree of disproportion between this adverse consequence and the contract price. 104 Finally, a similar response may be advanced against Smith's additional claim that the decision in Ruxley Electronics and Construction Ltd v Forsyth (Ruxley) to deny the claimant homeowner the cost of rebuilding the pool on the ground that it was 'out of all proportion' to the benefit to be obtained from doing so 105 'is difficult to reconcile with the continuity thesis'. 106 But given that under (a plausible interpretation of) the continuity thesis reasons other than those grounding the primary legal duty breached may exert normative force when determining the content of the secondary duty to pay damages, it would be surprising if the remedy that most closely replicates the primary duty was always appropriate. That the breaching party made a contractual promise to do something is not the only consideration exerting normative force post-breach in a case like Ruxley. Other considerations, arguably even 'public interest' concerns, 107 may legitimately be taken into account in deciding whether to enforce, via a monetary substitute for performance, the parties' original agreement in view of the (often relatively minimal) benefit that will result from curing the breach. While rational people may disagree on which side of this 'reasonableness' boundary the facts in Ruxley fall, 108 that there is such a boundary cannot reasonably be disputed.

Smith's proposed taxonomy of damages awards
If sound, the defence of the continuity thesis just provided was important not only because it discredits Smith's claims about damages awards but also because it undermines his distinction between 'rights-threats' and 'wrongs'. Recall that Smith claims that there is a distinction between the basis for specific relief (for example orders to pay debts or specifically perform contracts) and the basis for orders to pay various kinds of damages awards. While the former respond to 'rights-threats',the latter respond to 'wrongs'or 'injustices'.But if the continuity thesis is correct, this alleged distinction becomes difficult to maintain because at least some coercive orders share with damages awards the objective of giving legal effect, so far as can be justified, to the reasons grounding the primary duty breached in the post-breach world. 109 It is simply that while this goal is sometimes best achieved by coercive relief, in other cases an award of damages is, all relevant matters considered, more appropriate.
That Smith divides damages awarded in lieu of specific relief, which he calls 'substitutionary damages', 110 into two categories:awards responding to a 'rightsthreat' and awards responding to an 'injustice', 111 should put us on notice that something is amiss within his 'structure'. Also somewhat unusual is Smith's classification of 'market-price damages' (ie awards of 'the difference in market value between the goods [or services or land] … delivered and the goods [or services or land] … promised') as fundamentally different from specific relief. 112 As Smith correctly observes, within the law of contract 'market-price damages are available regardless of whether the claimant suffers a [balance-sheet] loss'. 113 But rather than differentiating these awards from specific relief, this feature unites market-price damages and specific relief because in both cases the court's order aims to facilitate 'next-best conformity', in the sense explained above, with the claimant's primary right to performance irrespective of the eventual balancesheet consequences of the breach.
Less obviously, the viability, or at least utility, of Smith's concept of an 'injustice' to explain certain damages awards may be undermined by the correctness of the continuity thesis. One of Smith's central examples of a JPLR responding to an 'injustice' is an award of damages for consequential financial losses caused by a breach of legal duty, which he calls 'compensatory damages'. 114 While Smith admits that such awards might be explained as responding to 'wrongdoing', he claims that this explanation cannot account for the 'limitation rules' on these awards. 115 It was just argued that this is incorrect. But Smith's reasoning here also appears to fail to clearly to distinguish legal and moral 'wrongdoing'. 116 As Gardner and Birks have explained, legal 'wrongdoing' merely denotes the breach of a legal duty, 117 which may (or may not) coincide with the violation of a moral duty. Smith apparently thinks that if 'compensatory damages' respond to 'wrongs', this entails that the extent of a breaching party's compensatory liability must correspond, or be proportionate, to the moral culpability of the defendant's conduct. 118 While Smith is correct to observe that the law does not uphold this principle, this would remain true without the 'limitation rules' on 'compensatory damages' that Smith criticises because the adverse financial consequences accruing to the victim of a breach of contract generally bear no relationship to the breaching party's moral culpability. 119 Smith's new scheme, at least as regards damages awards, therefore ultimately makes a mountain out of a mole hill. This is perhaps most clearly revealed by his positive explanation for 'the rules governing the assessment of compensatory damages' -that these rules determine 'when it is fair to attribute responsibility for the claimant's losses to the defendant' 120 -since this is precisely what proponents of the continuity thesis also believe. 121 As was explained above, in determining what constitutes 'next-best conformity' with the primary duty in the post-breach world, various considerations other than the reasons that justified the primary duty breached are also relevant. These considerations include whether the victim's action (or inaction) has exacerbated the breach's adverse consequences as well as, perhaps, the breaching party's moral culpability or the extent of disproportion between the losses suffered and the value of the original contract (or other right) infringed.
A final observation about Smith's taxonomy of damages awards concerns his classification of awards for the non-pecuniary consequences of legal wrongdoing as responses to 'wrongs'. 122 Smith's view is that these awards are fundamentally different from awards for the pecuniary consequences of legal wrongdoing, which he claims -incorrectly, we have argued -respond to 'injustices'. It is true that awards for the non-pecuniary consequences of a legal wrong are distinct from its pecuniary consequences since only the latter involve actually making good the breach's adverse consequences. These two kinds of awards nevertheless share something fundamental, which is that they both attempt to make the world closer to conformity with the reasons that justified the primary duty.
Smith also expresses a secondary concern that some awards for 'nonpecuniary loss' are provided without any evidence that such 'loss' has occurred, citing the 'loss of amenity' award in Ruxley as an example. 123 Ruxley is a controversial case, but there seem to be at least two plausible explanations for the House of Lords' decision to award £2,500 for the homeowner's 'loss of amenity'. One is that the award was indeed for the homeowner's loss of enjoyment with the Court simply taking an understandably relaxed attitude to proof of this non-pecuniary loss because it could not be financially quantified. The other explanation is that the award was not really for the non-pecuniary consequences of the breach at all, but an attempt to value the abstract (or normative) loss of performance entailed by the breach itself in circumstances where there was no other feasible way to value this lost performance since awarding the 'cost of cure' was considered 'unreasonable' and there was no difference in market value between the performance promised and that provided.

RESTITUTION FOR 'DEFECTIVE TRANSFERS' AND 'INJUSTICES'
One of Smith's most novel arguments in Structure is that some JPLRs are awarded on the grounds of 'injustices'. In addition to 'compensatory damages', the most notable JPLRs that Smith argues are ordered on this 'ground' are 'restitutionary orders' following 'defective transfers'. By a 'restitutionary order' responding to a 'defective transfer', Smith means an order directing a defendant 120 Structure n 4 above, 265. 121 Admittedly, Smith's conception of 'fairness' may be different, which highlights the unhelpfulness of this explanation. 122 Structure n 4 above, 216-220. 123 ibid, 218. to return 'money or property' or 'a sum equal to the property's value' 'transferred' by a claimant whose autonomy was 'impaired'. 124 A claimant typically shows autonomy impairment, claims Smith, by demonstrating that the 'transfer' in question was the result of -amongst other events -mistake, duress, undue influence, incapacity, compulsion, or necessity.
Smith observes that restitutionary orders following 'defective transfers' are not made on the grounds of 'wrongs' because 'the beneficiary of a defective transfer may be an entirely innocent participant in the transfer'. 125 Nor are such orders made on the 'ground' of 'rights-threats' since they do not confirm pre-judgment legal duties. There is no pre-judgment legal duty to make restitution of 'defective transfers', Smith argues, because: this duty would be unknowable; 126 damages are not awarded for a failure to make restitution; and proprietary theories of unjust enrichments are false. 127 This final observation means that a restitutionary order to return a 'defective transfer' is not made on the 'ground' of a threatened violation of any subsisting imperfect property right that the transferor possesses.
There are two key, controversial, assumptions underpinning Smith's discussion of restitutionary orders following 'defective transfers'. The first is that if there is no pre-judgment legal duty to return mistaken payments, the same follows in respect of all (or most) restitutionary orders following 'defective transfers'. The second is that the primary normative justification for restitutionary orders following 'defective transfers' is that the 'transfer' was made in circumstances where the claimant's autonomy was impaired. The correctness of the first assumption is doubtful. 128 As to the second, such theories of 'defective transfers' are at best incomplete. The main deficiency, as explained below, is that these theories fail to explain why a claimant's impaired autonomy should impose a legal duty on the defendant to return the value 'transferred'.
The reasons Smith provides for why there is no pre-judgment legal duty to make restitution for 'defective transfers' are unconvincing. For reasons already outlined, we do not regard either the observation that a duty is unknowable, or that there is no duty to pay damages for a failure to perform a duty, as persuasive evidence for the absence of a pre-judgment legal duty to make restitution. Nor does Smith's 'unknowability objection' provide a compelling normative reason against recognition of a pre-judgment legal duty since, at best, it implies only that a duty to make restitution should not arise until a defendant could reasonably have known of the 'defective transfer'. The positive law does not, in our view, provide a clear answer to whether there is a pre-judgment legal duty to make restitution of 'defective transfers'. Hence, the discussion that follows briefly examines whether the law should recognise a pre-judgment legal duty to make restitution in certain scenarios. We will argue that, within a Kantian conception of private law, Smith's claim that there should not be a pre-judgment legal duty to make restitution of mistaken payments is plausible. We nevertheless maintain that the reasons given for the absence of a pre-judgment legal duty to make restitution of mistaken payments probably do not apply to all 'defective transfers'.

Mistaken payments
Proponents of a Kantian interpretation of private law claim that the principal scenarios where it is generally justifiable for JPLRs to be awarded are when there is a violation, or threatened violation, of a 'duty of right', a duty that a person has voluntarily assumed,or a property right. 129 'Duties of right'are duties that follow from Kant's principle of right, which states that '[t]he fundamental principle applicable to the interaction of self-determining beings is that action should be consistent [or co-exist] with the freedom of whomever the action might affect'. 130 On a Kantian view of private law, unless one accepts a 'proprietary theory' of mistaken payments, there is probably no pre-judgment legal duty to make restitution of mistaken payments. First, there appears to be no compelling Kantian justification for a payee having a 'duty of right' to return the value of a mistaken payment to a payor. Secondly, the legal duty to make restitution of a mistaken payment is not voluntarily assumed by a defendant. Finally, absent judicial or statutory pronouncement to the contrary, when there is merely a 'duty of virtue' to , there is no legal duty to .
If C pays D money under a mistake of fact or law, provided that the mistake is sufficiently serious, 131 D can generally claim the payment's value back from D. 132 The usual justification for this claim, which Smith adopts, 133 focuses upon the payor's defective consent. 134 But such explanations are incomplete because they fail to account for why C's defective consent should matter to D, the payee, in such a way that justifies imposing a legal duty upon D to make restitution. Scholars are aware of this problem, 135 and Wilmot-Smith has convincingly demonstrated that no persuasive Kantian justification has been provided for the payee's legal duty to make restitution. 136 No such justification seems likely to exist either because it is difficult to identify any right that the mistaken payor possesses that the payee violates, or threatens, by retaining the payment's value.
In response to Wilmot-Smith, Penner has argued that restitutionary orders following mistaken payments should be understood as legalising a 'duty of virtue' to return the value of mistaken payments. 137 'Duties of virtue', unlike 'duties of right' which encompass obligations not to infringe upon the 'equal freedom' of others, denote obligations that persons have to concern themselves with the lives of others. 138 Penner's argument is an explanatory claim. He does not suggest that the legalisation of a 'duty of virtue' to return mistaken payments is normatively justified; only that this is the best explanation for restitutionary orders following mistaken payments. 139 Significantly, Penner also does not claim that the legal duty to return the value of a mistaken payment arises prior to a court order.
From within a Kantian framework, 'duties of virtue' cannot be legally justified duties. Moreover, it does not follow that because the law may have legalised a 'duty of virtue' by allowing judges to order restitution of mistaken payments that the law should recognise a pre-judgment legal duty to do so. Thus, at least from within a Kantian framework, the existence of a pre-judgment legal duty to return the value of a mistaken payment prior to a court order should probably be rejected as Smith claims.
There are two principal interpretations of when a 'failure of condition' may provide a right to restitution. The first is parties, on an objective analysis, 142 agreed that the claimant's performance of an act was undertaken on a condition that has failed. 143 The second is the defendant, on an objective analysis, knew that the claimant intended to make a 'transfer value' on a condition that has failed. 144 The requirement that the parties objectively agreed that the claimant's act was conditional, or that the defendant objectively knew the claimant intended its 'tranfer of value' to be conditional, demonstrates that the justification for ordering restitution for a 'failure of condition' cannot be simply that the claimant's consent to a 'transfer of value' 145 was vitiated. 146 An alternative (and more plausible) interpretation of restitutionary awards for a 'failure of condition' is that such orders derive, in some way, from the normative value of legally recognising certain agreements. On this view, the law of agreements is not co-extensive with the law of contract.
Under an agreement-based account of 'failure of condition', the form of agreement required for restitution to be ordered may take at least the following different forms: (T1) there is a legally recognised contractual or non-contractual agreement that the claimant's act is conditional; (T2) there is a legally recognised contractual or non-contractual agreement that the claimant's act is conditional as well as a contractual or non-contractual agreement that if that condition fails, the value of the claimant's act will be returned in monetary form. According to (T2), an award of restitution for a 'failure of condition' simply requires the defendant to do what they legally agreed to do, late. (T1) is different because the restitutionary order does not specifically enforce the parties' agreement. But such an order still gives effect, albeit perhaps less directly, to the parties' agreement that the claimant's act was performed subject to an agreed condition that has failed because it reverses,so far as money can,the act 147 that the claimant conditionally performed.
'Failure of condition' is arguably not a homogenous 'unjust factor' and different cases where courts have ordered restitution following a 'failure of condition' might be best explained by reference to different kinds of agreements. For instance, (T2) seems more plausible in respect of claims for restitution of a complete performance by a payor of a contractual payment obligation on a condition that fails. 148 But (T2) may be a less plausible explanation of cases where restitution for a 'failure of condition' is awarded to an innocent partialperformer of an entire service obligation. 149 A complete Kantian theory of 'failure of condition' obviously cannot be developed here. But if restitutionary orders following a 'failure of condition' are understood as giving effect to legally recognised, though not necessarily contractual, agreements it may be that restitutionary orders in response to a 'failure of condition' are susceptible to Kantian justification. According to Kant, contracts are a legal institution through which legal persons vary their rights and obligations vis-a-vis one another, 150 which enable them 'to set and pursue their own purposes interdependently'. 151 The rights arising from contracts 'are the juridical manifestations of the freedom inherent in selfdetermining activity'. 152 The same appears true of other forms of legally recognised agreements. Legal agreements can, from this perspective, be understood as manifestations of the parties' agency and, to the extent that the law values the forms of self-determining activity that these agreements make possible, it should generally enforce the rights and duties they create. 153 It is therefore arguable that, from a Kantian perspective, restitutionary orders following a 'failure of condition' may be legally justified because they restore the claimant and defendant, so far as money can, to a position where their joint agency, manifested in the form of their pre-existing agreement, determines their respective positions. The restitutionary order achieves this by imposing a duty upon the defendant to reverse, as far as possible in monetary form, the act that the claimant performed on an agreed condition that has failed. If so, the argument just developed for why there should probably be no pre-judgment legal duty to return the value of a mistaken payment does not apply to restitutionary orders following a 'failure of condition'. This does not demonstrate that the law should, all things considered, recognise a pre-judgment legal duty to make restitution for a 'failure of condition'. But it does, at least from within a Kantian framework, suggest that Smith's conclusion that there should be no pre-judgment legal duty to make restitution following all or most 'defective transfers' is too quick because it is plausible to regard restitutionary orders in response to 'defective transfers' as having different justifications.

CONCLUSION
Structure is an intensely original and thought-provoking book that makes a highly significant contribution to an important and under-theorised part of private law theory. However, whilst Smith may be correct that private law remedies are not ordered solely on the 'ground'of 'right-threats',he has in our view failed to substantiate key claims upon which certain of his more specific arguments rest, which significantly undermines the overall structure of 'remedial law' that he proposes. Relevantly, the arguments that Smith provides against the 'dutyconfirming view' of damages awards are largely unpersuasive and it is doubtful that he has demonstrated that all restitutionary orders following 'defective transfers' are made on the 'ground' of 'injustices'. Despite these specific criticisms, we trust that our engagement with just a few of the book's many novel ideas reveals our admiration for what, in writing Structure, Smith has accomplished. 22