JUSTIFYING CONCURRENT CLAIMS IN PRIVATE INTERNATIONAL LAW

Abstract Can claimants choose between contract and tort claims arising on the same facts with different jurisdictional and/or choice-of-law consequences? While domestic legal systems generally recognise concurrent liability, commentators object that its extension to private international law would be unprincipled and would threaten the field's values. This, however, contrasts with the position in common law and under EU Regulations, where concurrent claims are generally recognised with only narrow limits. This article justifies concurrent claims in private international law, arguing that the same premises supporting concurrent liability in domestic law exist in private international law, and that no field-unique concerns foreclose it.


I. INTRODUCTION
The classification of legal obligations is central to private law. 1 However, since in practice courts classify issues or claims into abstract legal categories, such categories may overlap on a single set of facts, effectively giving claimants a freedom to choose between different claims. The potential for concurrent liabilityliability under concurrent claims that arise on a single set of factsin contract and tort is particularly high, given the drastic expansion of negligence and economic torts in recent years. The common law's position on concurrent liability is an accepting one, with the House of Lords declaring it generally unobjectionable in Henderson v Merrett Syndicates Ltd. 2 But while this may be settled in domestic law, doubts persist whether concurrent claims should be recognised in private international law: there, the prevailing academic view is that such recognition would be unprincipled, and would undermine values or goals unique to private international law. 3 Yet, these concerns have been disregarded both by apex common law courts and the Court of Justice of the European Union, which have endorsed concurrent claims in contract and tort with few reservations, as the CJEU's decision in Wikingerhof GmbH & Co. KG v Booking.com BV 4 demonstrates.
In this article, we justify the recognition of concurrent claims in private international law. We argue that the same premises which justify concurrent liability in domestic private law equally justify concurrent claims in private international law, and that this is not foreclosed by additional concerns unique to the field of private international law, as a whole or in relation to its sub-fields of jurisdiction and choice-of-law. Private international law rightly adopts a principled approach to the characterisation of legal obligations which relies on substantive notions of private justice. Thus, just as domestic private law recognises that the principled classification of claims into legal categories may cause such categories legitimately to overlap on particular facts when doing so would vindicate their underlying principles, private international law should as well. After Section II outlines the apparent concerns arising from concurrent claims in private international law, Section III surveys cases from common law jurisdictions and European Union law, identifying an emerging judicial recognition of concurrent claims in private international law. Section IV then sets out and elaborates on our argument for recognising such concurrent claims. This article focuses on concurrent claims in contract and tort, largely due to the attention that concurrency has received in those quarters, although our conclusions should apply more broadly to other instances of concurrent claims in private international law, such as claims in contract and unjust enrichment. 5 Before we begin, three clarifications of scope should be made. First, we are not arguing that there is at present an identifiable international substantive theory of private law, requiring certain fact patterns to give rise to concurrent claims in every legal system. Our argument is only that, whatever the forum's theory of private law as applicable in international disputes might be, parties should in general be free to bring concurrent claims whenever they arise in accordance with that theory. Second, the concept of "concurrent claims in private international law" must be clarified. Where concurrent claims are brought in a private dispute with international elements, two questions can arise at different stages of proceedings. There is the initial question of whether the claimant should be able to choose between concurrent claims prior to determining the forum in which proceedings should be brought and the applicable law of the claim. This is a matter of characterisation in private international law. However, after proceedings are allocated to the appropriate forum and the governing law determined, another question may arise, this time on whether the governing law recognises concurrent liability and permits the claimant to choose between concurrent claims. This is a matter of domestic law. In this article, we are concerned only with the prior question, which raises the possibility of what we will call "concurrent claims in private international law". 6 Third, this article covers concurrent claims in contract and tort under EU Regulations on both choice-of-law and jurisdiction, despite the latter's current 7 inapplicability in English law, not least because the latter's interpretation significantly influences the former's. 8 However, we are concerned only with concurrent claims in contract and tort in general, and not with particular situations where the Regulations set out specific regimes, like facts which involve insurance, consumer or employment contracts. While the latter raises important questions, we lack the space to deal with them, 9 and the answers to those questions do not bear materially on the main question of how concurrent claims should be dealt with in general. 10

II. CONCURRENT CLAIMS: INTERNATIONAL DIMENSIONS AND CONCERNS
Classification is an essential part of private law adjudication: different issues or claims are classified under different juristic categories with different applicable rules. But a rapidly changing world poses difficulties for the neat classification of legal obligations. Because it is often hard to draw a clear-cut line between the various instances of business and human interactions, a single factual matrix may frequently raise multiple issues classified 6 The term "concurrent claims" rather than "concurrent liability" is used because rules of private international law do not themselves attribute liability, but instead allocate disputes to courts and laws which do. But the phenomenon described is the same as that which exists in domestic law: where claims in contract and tort are recognised as available on a single set of facts. We are grateful to the anonymous reviewer for this point. 7 And future, so long as the UK remains unable to accede to the Lugano Convention or any equivalent European agreement. 8 See text accompanying notes 98 and 99 below. 9 For a discussion, see M. Poesen, "Concurrent Liabilities and Jurisdiction over Individual Contracts of Employment under the Brussels Ia Regulation" (2020) 16(2) J. Priv. Int'l L. 320. 10 Ibid., at 322. 140 [2023] The Cambridge Law Journal under different traditional private law categories. 11 In most common law jurisdictions, courts generally accept the existence of concurrent liability in domestic law. The locus classicus here is the House of Lords' decision in Henderson. 12 There, underwriters at Lloyd's made losses due to the negligence of their syndicate managers in providing advice, but found their contract claim against their managers time-barred, and so sued in the tort of negligence. Lord Goff, who delivered the leading judgment, upheld the underwriters' claim on the primary grounds that contract and tort claims arose from different duties, imposed by the "will of the parties" and the "general law" respectively, 13 both of which were owed to the underwriters by their managers. Other common law jurisdictions have since followed suit, recognising that claimants are generally free to choose between concurrent contract and tort claims arising from a single factual matrix. 14 Concurrent contract-tort liability, however, is not absolute; courts have developed two forms of limits thereon, although neither limit effectively precludes such liability. The first are direct limits, limiting the existence of concurrent contract and tort duties ex ante. In Henderson itself, Lord Goff held that the concurrency was "subject . . . to ascertaining whether the tortious duty is so inconsistent with the applicable contract that, in accordance with ordinary principle, the parties must be taken to have agreed that the tortious remedy is to be limited or excluded". 15 Andrew Burrows breaks this proposition down into two further principles. There is the "exclusion principle", under which parties can agree to exclude and extinguish tort liability. 16 There is also the "independence principle", under which tort claims are precluded "if in essence the claim is one to enforce a contract". 17 Neither principle, however, precludes concurrent liability outright; indeed, they appear to have had little impact outside cases involving express exclusions of tort liability. The second form of limitations are indirect: they accept the general existence of concurrent liability but limit its effects ex post, by disapplying or circumscribing the "incidental rules" which ordinarily follow the cause of action instituted, 18 such as rules on remedies, remoteness and limitation periods. incidental rules, however, follows directly from classification: for example, choosing a tort over a contract claim leads to tort remedies and limitation periods, as Henderson itself demonstrates. In sum, then, the common law's endorsement of concurrent liability is not just in form but also in substance: not only are there few ex ante limits on concurrent contract and tort claims, there are also many instances where a claimant's choice between claims determines the applicable incidental rules ex post. These direct and indirect limits on concurrent liability may thus be called tangential limits: they are justified by different reasons from those which would preclude concurrent liability outright, and so do not effectively preclude but merely qualify such liability in select situations.
In this article, however, we are not chiefly concerned with the position on concurrent liability in domestic law, but with the implications of recognising concurrent claims in private international lawthat is, in private disputes with foreign elements that raise questions about its amenability to a foreign court's jurisdiction or the applicability of a foreign law to its merits. The complication is this: even if we accept that, within single legal systems (i.e. in domestic law) a given dispute can be classified in several ways, with claimants being free to choose the classification that best suits them, it does not necessarily follow that this should also be so across multiple legal systems (i.e. in private international law). The latter gives rise to apparently more drastic consequences, because of the nature of the incidental rules at stake. While a choice between claims in purely domestic disputes is a choice between rules governing remedies, limitation periods and remoteness, a choice between claims in private international disputes is a choice between rules governing international jurisdiction and the applicable law. In other words, it is a choice between legal systems rather than mere rules of law.
If these are the implications of recognising concurrent claims in private international law, what are the concerns it raises? Adrian Briggs, in a seminal article 19 which continues to define the prevailing academic opinion on the issue, 20 argued that the justification for concurrent claims in private international law is "an elusive thing". 21 But beyond arguing negatively that no good reasons exist for recognising concurrent claims, Briggs also argued positively that doing so would be contrary to the values of equality, 19

142
[2023] The Cambridge Law Journal uniformity and certainty prized by the field of private international law. 22 Several interrelated concerns may be identified.
First, recognising concurrent claims would disproportionately disadvantage defendants in private international disputes. Briggs points out that giving the claimant alone freedom to choose between multiple choice-of-law rules ("several bites of the choice of law cherry") introduces inequality between the litigating parties by disproportionately disadvantaging the defendant. A claimant would be able to invoke multiple applicable laws and get her desired outcome if even one of them would grant her claim, while the defendant would only escape liability if all the potentially applicable laws would absolve her. 23 Second, above and beyond disproportionate defendant disadvantage, recognising concurrent claims is said to threaten values and goals unique to private international law. A distinction may be drawn here between the relevant values and goals in jurisdictional and choice-of-law contexts. In the choice-of-law context, the recognition of concurrent claims is said to undermine the value of international uniformity. Uniformity is one of private international law's main "conflicts justice" values: it minimises the claimant's ability to influence a dispute's substantive outcome simply by choosing to commence proceedings in a particular forum. 24 The EU's Rome I and Rome II Regulations on contractual and non-contractual obligations respectively 25 ("Rome I" and "Rome II") reflect this goal: their recitals both state that "the same national law" should apply to similar claims "irrespective of the country of the court in which an action is brought". 26 However, as Briggs argues, the choice-of-law process' goal of obtaining "a result which is only minimally influenced by the identity of the forum" may be threatened "if a claimant is allowed a free and unconstrained choice from the menu of choice-of-law rules available under the lex fori". 27 In the jurisdictional context, the recognition of concurrent claims is said to undermine the values of predictability and anti-fragmentation. Predictability, or legal certainty, is a central goal of common law jurisdictional rules for international commercial disputes; 28 and is also central to the EU's Brussels Recast Regulation on jurisdiction ("Brussels Ia"), 29 with its recital stating that "rules of jurisdiction should be highly predictable". 30 22  Predictability and certainty, however, may be jeopardised when claimants are given a choice between concurrent contract and tort claims, with the result that different courts may have jurisdiction over the parties' dispute. Relatedly, rules of international jurisdiction also generally avoid the fragmentation of disputes. 31 This value undergirds common law doctrines like forum non conveniens, which allocates disputes with subject-matter connections to multiple jurisdictions to that which best suits "the practicalities of litigation"; 32 as well as provisions of Brussels Ia like Articles 29 and 30, which require the court first seised to exercise jurisdiction to the exclusion of all other courts and empowers those latter courts to stay related actions in favour of the court first seised. Recognising concurrent claims, though, means that a dispute can be slotted under distinct jurisdictional gateways, with different courts adjudicating upon similar facts and witnesses, which risks parallel litigation and fragmentation. Under EU Regulations, in particular, these values of predictability and anti-fragmentation support the position that Articles 7(1)(a) and 7(2) of Brussels Ia, which contain two special jurisdictional rules for "matters relating to a contract" and "matters relating to tort, delict or quasi-delict" respectively, should be interpreted in a "mutually exclusive" manner. 33 Admittedly, while the above position remains dominant, it has its opponents, who deny that recognising concurrent claims is any more problematic in private international law than in domestic private law. Richard Plender and Michael Wilderspin, for instance, argue that the identification of claims and thus the possibility of concurrent claims is a "threshold question" for the "procedural law of the forum", and so the forum's legal categories should be used. 34 Briggs's (convincing) reply is that the existence of claims for the purposes of private international law is a question for the latter's doctrine of characterisation rather than the forum's substantive private law categories. 35 Another more common response to the above objections to concurrent claims is one that agrees with the thrust of the objections, while nevertheless considering the relevant concerns assuaged by other rules of private international law. These other rules therefore operate as effective limitations on concurrent claims. These responses reject concurrent claims in substance if not in form, and only differ from the position taken by objectors like Briggs in that they believe that the relevant problems have been solved. Two such lines of reasoning are worth consideration.

144
[2023] The Cambridge Law Journal First, some argue that parties may, by agreement, subject tort claims arising from their commercial relationship to the same forum or governing law as their contractual claim. The common law has long encouraged such private ordering via jurisdiction agreements, an ethos which Brussels Ia Article 25 embodies as well. While such free choice is less effective in the choice-of-law context, some legal systems also support it; Rome II Article 14, for instance, allows parties to choose the governing law of their tort claims subject to certain limits. The logic here is that, even if concurrent claims have undesirable consequences, these are avoided because parties can simply agree to treat their tort claims similar to their contract claims for jurisdictional and choice-of-law purposes.
Second, some argue that, at least in choice-of-law, while the applicable rule for tort claims generally selects the lex loci delicti or the lex loci damni, many jurisdictions (with the notable exception of Australia) 36 recognise an exception to that rule which selects a law with a close connection with the tort, and this can include the law governing a closely related contract. A prime example is Rome II Article 4(3), which states that when "it is clear from all the circumstances of the case that the tort/delict is manifestly more closely connected with a country other than [the loci damni]"which applies by default under Article 4(1)that country's law will apply instead. Importantly, Article 4(3) itself states that a "manifestly closer connection . . . might be based in particular on a pre-existing relationship between the parties, such as a contract, that is closely connected with the tort/delict in question". Similar arguments have been made in relation to domestic common law rules. For example, the common law's traditional "double actionability" rule has a "flexible exception", which allows courts to apply "the law of the country which . . . has the most significant relationship with the occurrence and with the parties". 37 Similarly, the UK's Private International Law (Miscellaneous Provisions) Act 1995 ("PILA") s. 12 contains an exception which displaces the default lex loci delicti rule in section 11 in favour of a "substantially more appropriate" law. 38 Here, courts are directed to consider factors relating to "the parties", "any of the events which constitute the tort or delict in question" or "any of the circumstances or consequences of those events". 39 Both these tort choice-of-law exceptions appear wide enough to allow courts to select the governing law of a closely related contract, even if the latter law is not expressly identified as it is under Rome II Article 4(3). Overall, the logic here is again that, even if concurrent claims have undesirable consequences, these are avoided because 145 Justifying Concurrent Claims in Private International Law defendants can simply use these tort choice-of-law exceptions to subject tort claims to the governing laws of closely related contracts.

III. JUDICIAL RECOGNITION OF CONCURRENT CLAIMS IN PRIVATE INTERNATIONAL LAW
Thus, there is an academic consensus against concurrent claims in private international law, on grounds that its recognition would be unprincipled and would frustrate private international law's values. This remains the case even among commentators who consider these concerns largely assuaged by other rules of private international law which effectively limit concurrent claims. Courts, however, have been largely unmoved by these objections to concurrent claims, instead choosing to recognise such claims in private international disputes, and claimants' ability to choose between them. Moreover, courts have not merely recognised concurrent claims in form but also in substance, by placing only tangential limits on the effects of such claims. This part traces the growing consensus in domestic common law legal systems and under EU Regulations on the general recognition of concurrent claims in private international law.

A. Common Law
Common law courts have long endorsed concurrent claims in private international law. Sometimes, this occurs implicitly, when courts simply assume that Henderson applies in the international context, as in Australia, Canada and New Zealand. By contrast, English lawwhile still untouched by EU Regulationsexplicitly endorsed concurrent claims in private international law, at least in the choice-of-law context. In Coupland v Arabian Gulf Oil Co., 40 a Libyan company's British employee suffered a workplace injury at the company's Libyan premises and sued the company in negligence under English law. The company argued that its duties were limited to those under the Libyan law employment contract. 41 Goff L.J. disagreed: "[t]he plaintiff can advance his claim, as he wishes, either in contract or in tort", 42 and unless the contract itself contained a "clause which restricts or limits the plaintiff's right to claim damages in tort", the tort claim can be advanced. 43

146
[2023] The Cambridge Law Journal contract claim failed, 45 but the company argued that the breaches of tortious duties of care were governed by English law instead. The director countered that the tort claim should be characterised as contractual as well, since (citing Briggs) there was "only one duty which ha[d] been broken", and the accumulation of claims would disproportionately disadvantage him. 46 Tuckey L.J. agreed with the company, reasoning that "[d]omestic law allows concurrent claims in contract and tort and it has always been assumed that English private international law does so also". 47 In support, he cited Coupland, which he read as "echo[ing] precisely in a public [sic] international law context what Lord Goff of Chieveley was later to say in the context of the domestic law about concurrent claims in contract and tort in the Henderson case". 48 The common law's clearest endorsement of concurrent claims in choice-of-law is the Singapore Court of Appeal's decision of Rickshaw Investments Ltd. v Nicolai Baron von Uexkull. 49 Seabed, a German company, employed a marketing agent under a German law agreement to find buyers in Singapore for artefacts it had salvaged. Subsequently, Seabed's successor, Rickshaw, terminated the employment agreement and sold the artefacts to Singapore's tourism authorities. The agent sued Rickshaw in Germany for breach of the employment agreement, alleging that he was owed salary and expenses. Rickshaw then counterclaimed in Singapore in, inter alia, tort, alleging that the agent had withheld information from Seabed and divulged information to the authorities, jeopardising Rickshaw's position in negotiations with the authorities and forcing them to sell the artefacts at an undervalue. The agent applied to stay the Singapore proceedings on the grounds that Germany, not Singapore, was the natural forum, and argued that Rickshaw should be allowed to sue him only for breach of the employment agreement and not in tort. The Court of Appeal disagreed; citing Henderson, it held that "absent bad faith", there is "no reason why [claimants] should be denied the freedom of choice to frame their causes of action" in whichever way was "more advantageous" to them. 50 The agent's alleged tort had a "separate legal existence from his contractual existence and breaches thereof", and since "the latter does not expressly limit or exclude the former", Rickshaw could sue him "concurrently in tort and contract". 51 Common law courts have also recognised the claimant's ability to bring concurrent tort and contract claims with different applicable jurisdictional 45 Ibid., at [18]. 46  C.L.J.
147 Justifying Concurrent Claims in Private International Law rules, albeit less explicitly than they have in the choice-of-law context. The UK Supreme Court's two decisions on jurisdictional issues arising from Lady Brownlie's claims, regarding the accident Sir Ian Brownlie and his family met while holidaying at the Four Seasons hotel in Cairo, are examples. In Brownlie v Four Seasons Holdings Inc., Lady Brownlie sued the hotel's Canadian holding company in contract and in tort. Lord Sumption, with whom the entire Court agreed on this point, denied her permission to serve the holding company out of jurisdiction, because she lacked a good arguable case in contract, because Lady Brownlie had evidently not contracted with the holding company, and in tort, because the holding company had not operated the Cairo hotel. 52 At no point, however, was Lady Brownlie's ability to bring concurrent tort and contract claims on identical facts, with different applicable jurisdictional gateways, ever questioned.
Subsequently, in Brownlie v FS Cairo (Nile Plaza) L.L.C., 53 Lady Brownlie brought similar claims against the Cairo hotel's Egyptian operating company. Parties agreed that both contract and tort claims were governed by Egyptian law, 54 but the operating company resisted service on grounds that Lady Brownlie again lacked a good arguable case against it in contract, because that would be time-barred under Egyptian law, and in tort, because the Egyptian law doctrine of non-cumul precluded concurrent liability in contract and tort. 55 Lord Leggatt, with whom the entire Court agreed on this point, disagreed with the company on both points. On the contract claim, it was arguable that the Egyptian limitation period only ran from the date Lady Brownlie learnt the operating company's identity. 56 On the tort claim, although "the doctrine of non cumul" was "a basic principle of civil law", 57 it was unclear on the Egyptian law evidence whether Lady Brownlie's tort claim would be negatived by it, so the claim was arguable. 58 Lord Leggatt's judgment affirms that English law recognises concurrent claims in private international law even if the applicable law subsequently denies concurrent liability. The latter is a concern for trial, or at most for interlocutory determinations on whether claims should be summarily dismissed, 59 but it does not affect the claimant's ability to rely on the tort gateway at the jurisdictional stage.
Thus, common law courts generally endorse the claimant's ability to bring concurrent contract and tort claims on one set of facts, despite this having different consequences for jurisdiction and choice-of-law. 52 Brownlie v Four Seasons  [158]. 57 Ibid., at [159]. 58 Ibid., at [160]. 59 Ibid., at [99].

148
[2023] The Cambridge Law Journal However, they have also recognised that certain limits should be placed on the claimant's freedom to choose between concurrent claims. These limits, though, are nevertheless only tangential limits based on separate concerns from those which would preclude concurrent claims outright.
The first of such limits is Rickshaw's "bad faith" limit, under which concurrent tort claims may be excluded by parties' express agreement. This reflects the point raised earlier, that parties may, by agreeing to subject tort claims to the same forum or governing law as concurrent contract claims, effectively eliminate the effects of recognising concurrent claims. Jurisdiction and choice-of-law agreements, however, cannot comprehensively eliminate those effects. This is because the question of whether parties have subjected their tort claims to the same forum or governing law as their contract claim involves two questions: whether as a matter of contractual interpretation the jurisdiction or choice-of-law agreement's scope covers the tort claim; and whether as a matter of legal policy that agreement should be enforceable. 60 Both these questions are conceptually anterior to the question of whether concurrent claims in contract and tort should be recognised in the first place. They also involve different principled concerns (i.e. whether parties have and may validly exercise an actual choice over the forum and law which their tort claims will be subject to), and so cannot be invoked in every situation where concurrent claims arise.
For jurisdiction agreements, the main question will be the first one of contractual interpretation, on whether parties actually intended their jurisdiction agreement to cover a related tort claim as well. Such an intention, importantly, can never be taken for granted. Though since Premium Nafta Products Ltd. v Fili Shipping Co. Ltd. ("Fiona Trust") 61 English courts generally "start from the assumption that the parties, as rational businessmen" intend all disputes "arising out" of their commercial relationship "to be decided by the same tribunal", 62 there is doubt even in English law whether this is a true evidential "presumption", or a mere restatement of the common law's contextual approach to contractual interpretation. 63 Other apex common law courts in Australia and Singapore have explicitly favoured the latter position. 64  C.L.J. 149 Justifying Concurrent Claims in Private International Law effect, it is a rule of substantive contract law rather than private international law, and so applies only to jurisdiction agreements governed by English law. 65 For choice-of-law agreements, the second question of whether such an intention to subject tort claims to the contractual governing law ex ante, even if clear, should be enforceable, further prevents such agreements from limiting concurrent claims. This is because most common law jurisdictionslike Australia, Canada and the UK prior to Rome IIdo not generally consider choice-of-law agreements for tort claims enforceable. 66 Thus, jurisdiction and choice-of-law agreements can at best only place tangential direct limits on the effects of concurrent claims.
Another potential limit on concurrent claims would be tort choice-of-law exceptions, like the common law's flexible exception and PILA s. 12. These, however, also place only tangential rather than comprehensive limits on concurrent claims. The common law's flexible exception, for example, does not operate simply because the tort claim arises concurrently with a contract claim. For a start, it is unclear whether the exception may select any law other than the lex fori or the lex loci delicti. 67 And even if it can, the exception will only do so when both those laws are, as the Court in Rickshaw put it, "purely fortuitous". 68 In particular, for the lex loci delicti to be fortuitous, the tort claim must involve such a transnational array of facts that the very idea of territorial connections becomes arbitrary. This will not be the case in most disputes where contract and tort claims arise concurrently: in Rickshaw, for instance, the agent's alleged torts caused harm in Singapore, so the flexible exception could not be invoked to apply German law, which governed his employment contract. 69 PILA s. 12, by contrast, is more capaciously worded and thus apparently broader than the common law's flexible exception. In Trafigura Beheer B.V. v Kookmin Bank, 70 a South Korean issuing bank sued Dutch sellers of oils for failing to provide it with a full set of bills of lading. The bank was out of pocket since the buyers, to whom the goods were released, had become insolvent. The sellers had done this because they discovered errors in the original bills of lading issued by its Singapore office and went about correcting them there for reissuance.

150
[2023] The Cambridge Law Journal to receive valid bills of lading. 71 Aikens J. accepted that the bank's claim, which essentially involved the seller's acts at its Singapore office, was properly characterised as tortious, so Singapore law applied under the default lex loci delicti rule in section 11. 72 However, under section 12, "the law that the parties have expressly or impliedly chosen to govern their preexisting contractual relationship" which "g[a]ve rise to events constituting the alleged tort" could apply. 73 This was English law, since the sale contract, the letter of credit and other related contracts were all governed by English law. 74 Trafigura establishes three significant propositions about PILA s. 12. First, the relevant enquiry is a multifactorial one that necessitates a "valuejudgment as to the significance of factors that point to an applicable law other than that identified under [section 11]"; the lex loci delicti is displaced only when the former factors are "substantially" more significant. 75 Second, a contract between parties could be one such factor to displace the lex loci delicti. 76 Third, there is no pre-requisite that the tort's connection with lex loci delicti be insignificant or "fortuitous". 77 However, it goes too far to suggest, as Briggs did, that after Trafigura section 12 practically excludes the effects of recognising concurrent claims. 78 After all, Aikens J. said that the governing law of parties' contract is simply "a factor" under the broader multifactorial enquiry, and applied it because, on the facts, it was the most significant factor by far. 79 Properly understood, then, while PILA s. 12 is broader than the flexible exception, it does not limit the effects of concurrent claims outright. Instead, it only places a tangential limit on the effects of recognising concurrent claims, triggered when relevant factors, including but not limited to the parties' contractual relationship, substantially point toward a law other than the lex loci delicti.

B. European Union Regulations
The recognition of concurrent claims has often arisen for consideration within the EU jurisdictional context, when courts determine whether proceedings should be fitted under Article 7(1)(a) or 7 (2)  involving similar provisions under the Brussels Convention, the CJEU recognised concurrent claims in contract and tort, giving claimants a choice between both Articles. An investor made unprofitable investments with a Luxembourgian bank through a German intermediary on the latter's advice. The investor sued the bank in tort and contract, and the bank argued that the tort forum should also have ancillary jurisdiction over the contract claim. The CJEU disagreed: under the Brussels Convention's "autonomous" taxonomy of legal categories, "a court which has jurisdiction . . . over an action in so far as it is based on tort or delict does not have jurisdiction over that action in so far as it is not so based". 81 Thus, the investor's different claims could legitimately lead to "different aspects of the same dispute being adjudicated upon by different courts". 82 Kalfelis, however, was qualified by Marc Brogsitter v Fabrication de Montres Normandes EURL. 83 There, the CJEU laid out a highly ambiguous ratio which it did not apply to the facts (which are thus irrelevant for our purposes), which would later be read by subsequent courts in three different ways, qualifying a claimant's ability to bring concurrent contract and tort claims to different extents. The first reading of Brogsitter recognises and endorses concurrent claims in general subject to a minor limit, which may be described as the "indispensability" principle: claimants can choose between concurrent contract and tort claims unless the latter requires proof of the contract's existence and contents to succeed. This arises from paragraph 25 of Brogsitter: tortious conduct would raise "matters relating to a contract" only "where the interpretation of the contract which links the defendant to the applicant is indispensable to establish the lawful or . . . unlawful nature of the conduct complained of". 84 In Source Ltd. v TUV Rheinland, 85 Staughton L.J. held that tort claims should be characterised as contractual under the Brussels Convention if they could not be brought "independently of a contract"that is, if the contract had to be "established" for the claim to succeed. 86 A second reading of Brogsitter places a broader limit on the recognition of concurrent claims, which may be described as the "reasonableness" principle: a tort claim will be precluded if, on the facts, a contract claim could reasonably have been brought instead. Paragraph 26 of Brogsitter appears to support this: claims would fall within the Brussels Regulation's contract jurisdictional rule when their "purpose . . . is to seek damages, the legal basis for which can reasonably be regarded as a breach of the rights and obligations set out in the 81

152
[2023] The Cambridge Law Journal contract which binds the parties". 87 In Bosworth v Arcadia Ltd, Gross L.J. held that Brogsitter's "true ratio" was that a tort claim should be characterised as a "matter relating to a contract" if "the legal basis of the claim can reasonably be regarded as a breach of contract". 88 A third reading of Brogsitter rejects concurrent claims entirely. Part of paragraph 24 of Brogsitter supports this: a tort claim is contractual under Brussels if "the conduct complained of may be considered a breach of contract". 89 In Holterman Ferho Exploitatie BV v Spies von Büllesheim, the CJEU read Brogsitter as authority that tort claims would be characterised as raising "matters relating to a contract" if "the conduct complained of may be considered a breach of . . . contract". 90 If any set of facts which "may" give rise to a contract claim will do so to the exclusion of possible tort claims, concurrent claims are impossible.
Recently, in Wikingerhof, the CJEU confirmed the first "indispensability" principle reading of Brogsitter as correct. 91 A German hotel agreed to be listed on a Dutch company's online booking platform, subject to the latter's unilateral ability to redefine the terms of their agreement. The hotel later sued the booking platform for abusing its dominant position in contravention of German competition law. The hotel said that this raised "matters relating to tort" under Brussels Ia Article 7(2), giving German courts jurisdiction; the booking platform countered that the claim raised "matters relating to a contract" under Article 7(1)(a), giving Dutch courts jurisdiction. The CJEU agreed with the hotel. Although the "contract" and "tort" jurisdictional rules were "mutually exclusive", Brussels Ia "is characterised by the possibility which it confers on the applicant of relying on one of the rules of special jurisdiction laid down by that regulation". 92 The court had to apply a two-step analysis: first, follow "the applicant's choice whether or not to rely on one of those rules of special jurisdiction"; 93 second, assess whether the applicant's claim involved matters falling within that rule, "by reference to the obligation . . . which constitutes the cause of action". 94 On the latter point, a "claim" 95 would fall under Article 7(2) where it relies on a "breach of an obligation imposed by law, and where it does not appear indispensable to examine the content of the contract concluded with the defendant in order to assess whether the conduct of which C.L.J.
153 Justifying Concurrent Claims in Private International Law the latter is accused is lawful or unlawful". 96 Since the hotel's claim was that the booking platform used its dominant position to force it to contract, that contract was irrelevant to establishing the lawfulness of the booking platform's conduct, and so the claim was tortious. 97 After Wikingerhof, then, Brussels Ia generally recognises concurrent claims subject only to the "indispensability" principle. This position is mirrored in the choice-of-law context under the Rome Regulations. In ERGO Insurance SE v Gjensidige Baltic AAS, the CJEU held that, in interpreting the concepts of "contractual" and "non-contractual" obligations within the Rome Regulations, it was important to uphold "the aim of consistency . . . in the application of the Brussels I Regulation which . . . draws a distinction . . . between matters relating to contract and matters relating to tort, delict and quasi-delict". 98 Echoing Kalfelis, the Court then defined "contractual" obligations as "legal obligation[s] freely consented to by one person towards another", and "non-contractual" obligations as "all actions which seek to establish the liability of a defendant and are not related to a contract". 99 The Brussels case law, which now endorses concurrent contract and tort claims subject to Wikingerhof's indispensability principle, should apply to the Rome Regulations as well.
As with the common law, EU Regulations also contain limits on concurrent claims. But again, these tend to be tangential limits based on separate concerns, which thus only qualify concurrent claims in select situations. The first of these, Wikingerhof's indispensability principle, is already discussed. The second, the use of jurisdiction and choice-of-law agreements, is similar to the position under common law. The positive rules are admittedly different, at least for the second question of enforceability rather than the first question of scope, the latter being a question of national law allocated to the governing law of the relevant jurisdiction or choice-of-law agreement. 100 Brussels Ia is stricter about the enforceability of jurisdiction agreements for non-contractual claims than the common law: Article 25(1) requires those claims to be "connected" with or "proximate" to parties' legal relationship, 101 though this is unlikely to bar enforcement of any agreements covering concurrent contract and tort claims, 102 since those tort claims are likely to be considered proximate to the relevant contracts. Conversely, for choice-of-law agreements, Rome II is more permissive than the common law: under 96

[2023] The Cambridge Law Journal
Article 14(1)(b), parties "pursuing a commercial activity" can agree to subject tort claims to a chosen law if that agreement is "freely negotiated", expressed with "reasonable certainty" and does "not prejudice the rights of third parties". This, however, is still less permissive than Brussels Ia Article 25 (1): as Carr J. noted in Pan Oceanic Chartering Inc. v UNIPEC UK Co. Ltd., Article 14(1)(b) demands that defendants must have had a "genuine opportunity to influence [the agreement's] content . . . for [their] own benefit". 103 On the whole, however, under EU Regulations the same conclusion reached under common law can also be drawn: jurisdiction and choice-of-law agreements at best place only tangential direct limits on the effects of concurrent claims, since they involve different principled concerns from and thus do not comprehensively preclude the latter.
The third limit on the effects of concurrent claims is the indirect limit imposed by the exception to the default tort choice-of-law rule, Rome II Article 4(3). Significantly, the only enquiry under Article 4(3) is whether a "clear preponderance of factors" connects the tort to a law other than the lex loci damni; there is no additional requirement that the lex loci damni's connection with the tort be "fortuitous" or lack genuineness. 104 Many commentators have thus suggested that Article 4(3) eliminates the effects of concurrent claims: a tort claim brought on the same facts as a claim under a contract between parties "closely connected" with it will be governed by the latter's law. 105 However, it is unclear whether under Article 4(3) the fact that a closely connected contract between parties is governed by a law other than the law of the place of damage conclusively evinces a manifestly closer connection with that law, or whether, as with PILA s. 12, that fact is only one factor (albeit a potentially weighty one) that suggests a manifestly closer connection with that law. Importantly, while commentators have generally assumed the former, 106 the case law actually suggests the latter: as Miles J. noted in Lyle and Scott Ltd. v American Eagle Outfitters, 107 "the overall question under Article 4(3) is not merely whether the tort claim is connected with the contract . . . [but] whether the tort is manifestly more closely connected with the country other than that indicated in Article  108 This, as Linden J. emphasised in Owen v Galgey, 109 means that the connected contractual relationship is "but an example of the circumstances which may bear on identifying the country with the most manifest connection"; courts must still consider "all the circumstances of the case". 110 Significantly, the lex loci damni was not displaced in those decisions. 111 In Lyle and Scott, the English lex loci damni of a passing off claim was not displaced by Pennsylvanian law, which governed the parties' agreement on how their goods should be branded and marketed. 112 In Owen, the French lex loci damni of an occupier's liability claim could not be displaced by the English governing law of the agreement for the claimant to work the property. 113 Moreover, while courts have occasionally invoked Article 4(3) to replace the lex loci damni with the law governing a closely related contract between parties, they have never rested their reasoning on that fact alone. Instead, as Picken J. observed in Avonwick Holdings Ltd. v Azitio Holdings Ltd., 114 in those cases courts reason that a significant majority of all factors (including other territorial connecting factors like the location where the "alleged wrongdoing was planned, orchestrated and implemented"; where "direct or indirect" damage was suffered; and where assets "at the heart of the alleged wrongdoing" were 115 ) also point to that law which governs the parties' contract. 116 Thus, Article 4(3) requires a "clear preponderance of factors" that connect the tort to a law other than the lex loci damni, and the fact that parties' closely connected contract is governed by that law is only one (albeit potentially weighty) factor that points towards the other law, to be considered and weighed alongside other factors. So understood, Article 4 (3) is merely a tangential indirect limit on the effects of recognising concurrent claims similar to PILA s. 12: the governing law of a concurrent contract claim is only one factor to consider alongside others when invoking the exception.

IV. JUSTIFYING CONCURRENT CLAIMS IN PRIVATE INTERNATIONAL LAW
On the whole, then, common law courts and the CJEU have generally recognised concurrent claims in contract and tort in private international 108

156
[2023] The Cambridge Law Journal law, subject only to certain tangential limits. But courts have also provided little justification for their position. Then questions thus remain: why should courts recognise concurrent claims in general, despite the concerns raised in Part II above? Here, we provide a justification for concurrent claims in private international law: the same premises which ground the existence of concurrent liability in domestic law apply equally in the realm of private international law, and no additional concerns unique to the field, whether as a whole or in relation to its sub-fields of jurisdiction and choice-of-law, foreclose such an extension.

A. Principled Classification and Concurrent Liability in Domestic Law
Private law has a well-established tradition of classifying legal obligations on the basis of principle. Aristotle, for instance, adopted a division between voluntary and involuntary transactions which resembles the contemporary division between contract and tort law. 117 Kant likewise developed a taxonomy which divides private law into contracts, property and fiduciary obligations. 118 Both accounts proceed from first principles: voluntariness for Aristotle, rights for Kant. 119 Legal philosophy has since consistently recognised the significance of classification to private law: 120 classification helps the law maintain internal coherence and enhances the values of predictability and certainty. 121 As Burrows notes, the classification of legal obligations is fundamental to "principled reasoning" and the "essential requirement of the rule of law that like cases should be treated alike". 122 Of course, no consensus has been reached, even within single legal traditions, on how private law classification should be carried out. 123 However, regardless of how a legal system defines its categories of contract and tort, overlaps between categories on particular facts does not threaten, but is instead the inevitable and acceptable consequence of, a principled classification of private law. This argument holds regardless of the theory of private law one adopts. It flows naturally from a pluralist theory of private law, where different legal categories exist to vindicate different social values with different normative underpinnings. As Hanoch Dagan notes, to a pluralist "some overlaps are perfectly acceptable, or even desirable" since "life is messy and . . . different contexts, while distinct in some senses, often raise overlapping normative concerns". 124 Overlaps between legal categories on particular facts are also consistent with a monist theory of private law. There, as Jacob Weinrib argues, the legal classificatory exercise performs three functions: 125 it posits a unified and coherent normative basis of private law, divides that unified basis into particular legal categories according to its internal logic, and finally orders specific facts and events into each of those legal categories. To monists, then, an overlap between legal categories does not betray a lack of unified normative bases: "an adequate legal classification must articulate the conceptual interrelation between, for example, property and obligation", rather than assiduously assign "each legal issue to one concept alone". 126 Thus, as Paul Davies acknowledges, "[n]o matter what taxonomy is adopted when dividing the law up into discrete areas for exposition, it seems inevitable that there will be disputes which can be analysed as falling within one or more taxonomies". 127 The above discussion may be summed up in two propositions: first, that the principled classification of claims into legal categories is inherent to the very idea of private law adjudication; and second, that such legal categories may legitimately overlap on particular facts if demanded by their underlying principles. These two propositions undergirded Lord Goff's endorsement of concurrent liability in Henderson, which rested on the identification of different contract and tort duties with different normative justifications. 128 The important point there, however, is not so much the identification of duties but the basis for doing soeven if concurrent liability is a manner of concurrent "obligations", "claims" or even "rights", what matters is that the concurrence of these legal concepts necessarily rests on principle rather than pragmatism. For once overlapping legal concepts are identified and separately justified, the general availability of concurrent liability is inevitable. So well-established is this proposition in domestic private law that the debate among courts and commentators on concurrent liability no longer centres around its existence in general, but instead focuses on the appropriate limits that should be placed thereon. As mentioned, however, these limits on concurrent liability have heretofore been narrow and tangential: they do not effectively eliminate, but in fact implicitly support the existence of, concurrent liability. 129 158 [2023] The Cambridge Law Journal

B. Principled Characterisation and Concurrent Claims in Private International Law
Characterisation is as central to private international law as it is to domestic private law, being the prerequisite to the application of its jurisdictional and choice-of-law rules. However, characterisation in private international law cannot use the substantive private law taxonomy of any one system of domestic private law, because private international law's function is to allocate disputes to various different legal systems on an even-handed basis. Thus, if private international law chooses a given legal system's legal taxonomy (say, the forum's), it will inevitably be faced with the problem of having to justify that choice in the first place. Moreover, it is impossible to maintain that characterisation can be performed without any reference to law, 130 since characterisation for legal processes is an inherently legal exercise. 131 A solution to this chicken-and-egg problem requires that characterisation in private international law be performed by reference to an external, international, "autonomous" legal taxonomy of private law. The crucial question, then, is: how is this international legal taxonomy of private law to be constructed? The typical answer to this question is given in the form of a negative proposition: that characterisation should not take place in accordance with the strict legal taxonomy of any one state, especially the forum. But this is unhelpful, because it only begs the positive question: how should courts construct the applicable international taxonomy instead?
One might suggest that courts must construct that taxonomy using a different methodology from that which they use in domestic private law: while domestic private law looks to substantive theories of private justice to approach classification in a principled manner, private international law should be agnostic about theories of private justice and approach classification in a pragmatic and ad hoc manner. Thereunder, courts would look at comparative law datum on how the various legal systems of the world, or at least those systems pleaded as applicable, have characterised claims similar to those before them, and determine whether parties' claims are "contractual" or "tortious" on that basis. The pioneer of this approach was Ernst Rabel, who argued that the interpretation of choice-of-law rules required courts to rely on comparative law. 132 Shades of this pragmatic comparative approach are also evident in John Falconbridge's "via media" theory of characterisation, which posits that courts should first "engage in a process of provisional or tentative characterisation . . . informed [by] the content of each of the two laws which may be the proper law"; 133 and Otto Kahn-Freund's "enlightened lex fori" theory of characterisation, which refers courts to a "system of internationally accepted analytical concepts". 134 However, the problem with this pragmatic comparative law approach to characterisation is that it attempts to convert an "is" into an "ought", and thus cannot justify how private international law should operate as a field of law. As Mathias Reimann notes, although comparative law and private international law share an "intimate relationship", the two fields share very different goals: comparative law seeks to determine law as fact, but private international law must justify what law should do in practice. 135 If private international law abdicated its function to comparative law, it would surrender law's essential normative function of ordering legal relationships to achieve justice.
Thus, private international law must adopt the same methodology in characterisation as domestic private law does: it must ask itself how, in principle, private disputes should be characterised for the purposes of jurisdiction and choice-of-law. Of course, the principled taxonomy private international law constructs cannot be modelled on any one domestic law taxonomybut it remains the case that private international law must still ascribe to some theory of private law in constructing that taxonomy. This proposition, that characterisation in private international law remains a principled affair, is as old as the discipline itself: Savigny, for instance, believed that courts should characterise claims according to the nature of parties' "legal relations", which would reflect various substantive ideas of private interpersonal justice. 136 It also has support from recent high common law authorities: in Raiffeisen Zentralbank v Five Star Trading LLC, Mance L.J. noted that characterisation in private international law remains "a search for appropriate principles to meet particular situations", using legal categories which are defined and re-defined in line with "the purpose for which they were conceived". 137 This search must be undertaken in a "broad internationalist spirit", but it remains a search for "appropriate" principles rather than mere practice. 138 More recently, in Solomon Lew v Kaikhushru Shiavax Nargolwala, 139 Lord Mance, now sitting as an International Judge on Singapore's International Commercial Court, re-emphasised that private international law as a whole demanded "a

160
[2023] The Cambridge Law Journal principled, not simply a pragmatic, approach", under which "[t]he law is shaped to serve and give effect to the reasonable expectations of its users". 140 This principled approach is also endorsed by the authors of Dicey, Morris & Collins, who see characterisation as requiring courts to "consider the rationale of the English conflict rule and the purpose of the rule of substantive law to be characterised", and "[o]n this basis . . . decide whether the conflict rule should be regarded as covering the rule of substantive law". 141 Characterisation in private international law, just as in domestic private law, thus requires courts to ask: what normative concern(s) are implicated by the facts of parties' dispute, and which legal category (or categories) reflects them? This has tremendous ramifications for the recognition of concurrent claims in private international law: if the principled classification of disputes and issues into legal categories is central to private law, and if in domestic law such categories may legitimately overlap when doing so would vindicate their underlying principles, it is hard to see why the same conclusion should not be reached in private international law. In both situations, the only pre-requisite for recognising concurrent liability/ claims should be the same: that there are principled justifications for so defining the breadth of the legal categories that happen to overlap on particular facts.
Thus, while Briggs is right that concurrent claims should not exist in private international law disputes when there is "only one duty which has been broken, even though domestic law has a variety of ways of conceptualising it", he is too hasty in concluding that this is "often" the case, and that therefore concurrent claims should not be recognised in general. 142 Instead, the true position is that, in private international law just as in domestic private law, the question of the existence of multiple duties is a separate and logically anterior question of private law theory. As the Singapore Court of Appeal noted in Rickshaw, a claimant's tort claim need not necessarily have the same principled basis as a concurrent contract claim, because they may have "separate legal existence[s]". 143 Likewise, once the CJEU defined "tort" obligations in Wikingerhof, as "obligation[s] imposed by law", 144 and thus a category of legal obligations with independent normative underpinnings, 145 this led inexorably to its confirmation that concurrent tort and contract claims were generally permissible. 146 We should emphasise, however, that our argument is not that an identifiable international substantive theory of private law exists, either in EU law or across different national legal systems, under which certain fact patterns should give rise to concurrent claims in contract and tort; we take no position on this matter. Our argument is simply that concurrent claims should be recognised whenever, in light of whatever theory of private law the forum's courts adopt for international disputes, there are indeed two separate principled bases for imposing liability on the defendant. Accordingly, the court's task is nothing more and nothing less than to determine, according to its international legal taxonomy for private international law disputes, whether such separate bases exist on the facts.

C. The Limits on Concurrent Claims in Private International Law
Of course, just as in domestic private law, the fact that concurrent claims in private international law should exist does not mean that there should be no limits on their recognition and enforcement. But the conclusions reached above allow us to go further, in explaining how the current limits on the recognition of concurrent claims can be justified. Properly understood, the only real limit on concurrent claims in private international law (i.e. the only reason why courts should subordinate one claim to another despite there being separate principled grounds for both on the facts) is the uncontentious proposition that parties may, via a genuine expression of intent, agree to exclude tort jurisdictional or choice-of-law rules. Every other ostensible limit on the recognition of concurrent claims may be understood as simple reiterations of the pre-requisite for recognising concurrent claims: that both claims exist on the basis of independent normative concerns implicated on a single set of facts.
The first limitation, the "bad faith" principle recognised in Rickshaw, should be uncontroversial, because it is private international law's analogue of Burrows' "exclusion principle", that parties can use express exclusion clauses to limit concurrent liability. As mentioned, 147 today both the common law and EU Regulations allow parties to subject tort claims to the contractual forum under jurisdiction agreements, and, to a lesser extent, 146 One may argue that a duty or liability "imposed by law" may nevertheless be a "matter relating to a contract" within the meaning of Brussels Ia Article 7(1)(a), since general tort law imposes such duties or liabilities on parties by virtue of their being (or being close to being) in a contractual relationship. This, however, is doubtful as a matter of positive law under Brussels Ia; such "matters" only exist when the "obligation" pleaded by the claimant is one "freely assumed by one party towards another" (see Judgment of 8 May 2019, Kerr v Postnov, Case 25/18, EU:C:2019:376; Dickinson, "Provinces", 219). Even if correct, though, the point remains consistent with our argument: if under the Brussels regime the definition of "matters relating to a contract" were indeed so wide there would be little need for courts to recognise concurrent claimsbut such recognition would be unwarranted precisely because, given the legal taxonomy adopted by the Brussels regime, there would be no principled basis for recognising non-contractual liability in those circumstances. 147 See text accompanying notes 60-66 and 100-103 above.

162
[2023] The Cambridge Law Journal to contractual governing laws under choice-of-law agreements. Commentators also recognise that, at least when the tort claim arises from parties' commercial dealingswhich will always be the case when it arises concurrently with a contract claimthere is no real reason why such agreements should not be enforceable. 148 The only concern here is to protect weaker or third parties, which is met by requirements like those contained in Brussels Ia Article 25 and Rome II Article 14(1)(b), which ensure that agreements reflect a genuine consensus between equallysituated commercial parties. But this limit on the effects of concurrent claims is uncontroversial, because it turns on the intuitively-acceptable principle that parties can, via a genuine display of consent, voluntarily agree to relinquish their rights. And importantly, this recognition of concurrent claims would be the logical pre-requisite of this limit, because it is only by first recognising concurrent claims in contract and tort that it becomes meaningful for courts to subsequently ask whether parties have genuinely agreed to exclude tort claims. Enforcing jurisdiction and choice-of-law agreements thus cannot be said to threaten concurrent claims in private international law, any more than enforcing exclusion clauses threatens concurrent liability in domestic private law.
The second limitation, the "indispensability" principle recognised in Wikingerhof, is more controversial. Since what must be "indispensable" to the tort claim is a need to "examine the content of the contract", 149 the idea seems to be that if a sufficiently-particularised tort claim requires the claimant to plead and prove a contractual provision's content, the latter should prevail over the former. 150 However, it is unclear why a tort claim requiring proof of a given contract provision should be excluded in favour of the contract claimafter all, here there could still be two different normative grounds for imposing two different heads of liability. Conversely, it is unclear why the very fact that a contract provision need not be interpreted to determine the scope of tort liability should allow the latter to be asserted independently of a concurrent contract claimjust because one claim's formal provenance differs from another's does not mean the two are normatively independent. At face value, then, Wikingerhof's indispensability principle seems both over-broad and under-inclusive, which renders its application somewhat arbitrary. 151 This is probably unsurprising; like the doctrine of illegality's traditional "reliance" rule, rules which apply depending on whether the claimant has pleaded a particular fact tend in many situations to be too rigid to accommodate the concerns said to justify them. 152 148  A better (re)interpretation of Wikingerhof's indispensability principle, then, is one that ignores the technicalities of such formal reliance and instead goes straight to the CJEU's underlying concern in enunciating it: 153 the need to ensure that tort claims exist "independently of [a] contract". 154 This would then appear to align Wikingerhof's indispensability principle with Burrows' "independence principle", that only a tort claim which "in essence . . . is one to enforce a contract" should be precluded. 155 But if this is so, Wikingerhof's indispensability principle may be no limit on recognising concurrent claims at all. In illustrating how the independence principle operates, Burrows noted that concurrent liability should only arise when "the thrust of the plaintiff's claim[s]" are different: 156 if you hire a workman to clean your chandelier but he destroys it instead, you may sue him in negligence rather than breach of contract because you are aggrieved about the interference with property rather than the failure to perform. 157 But if all the independence/indispensability principle requires is that there be different normative grounds for a claimant's separate complaints, then it is not really a limit on recognising concurrent claims, but a mere restatement of its essential pre-requisite. The only claims precluded by Wikingerhof's "indispensability" principle, then, are tort claims founded on, and only on, the breach of a contractual term between partiesand there, there is no separate basis for liability other than the breach of the agreement, and thus no basis for recognising concurrent claims at all.
The third limitation on the effects of recognising concurrent claims, namely tort choice-of-law exceptions, is perhaps the most controversialif concurrent claims should be recognised when separate principled bases exist for each, how can a rule that may potentially collapse both claims into the contract claim be justified? As mentioned, these exceptions are triggered either by the fact that the default applicable law under the default rule was "fortuitous" (common law) or overshadowed by a "preponderance" of factors pointing to another law, including but not limited to the fact that a closely related contract is governed by that law (PILA, s. 12 and Rome II, Article 4(3)). Properly understood, then, neither of these exceptions are truly designed to preclude the effects of concurrent claims. Instead, they simply ensure that tort choice-of-law rules do not favour one territorial factor (the location of the act complained of or the location where the 153 For readings to this effect, see Dickinson, "Provinces"; M. Poesen, "Regressing into the Right Direction: Non-contractual Claims in Proceedings Between Contracting Parties under Article 7 of the Brussels Ia Regulation" (2021) 28 Maastricht Journal of European and Comparative Law 390. See also A. Briggs, "Wikingerhof: A View from Oxford", available at https://eapil.org/2020/12/07/briggs-on-wikingerhof/ (last accessed 8 April 2022), who based on this reading of Wikingerhof (unsurprisingly) called the decision "a disaster". 154 Wikingerhof v Booking.com, C-59/19, at [33]. 155 Burrows, Obligations, 21. 156 Ibid., at 23, emphasis added. 157 Ibid., at 25-26.

164
[2023] The Cambridge Law Journal damage is suffered) over other territorial or non-territorial factors which are more significant given the nature of the tort alleged. 158 Thus, the common law's flexible exception operates only when the connection between the tort and the initial applicable law is insignificant in absolute terms, because that connection is "fortuitous". Similarly, PILA, s. 12 and Rome II, Article 4(3) demand that the connection between the tort and the default applicable law be overshadowed by the connections that a "clear preponderance" of other factors have with the tort; this logically imposes a requirement that the first connection be insignificant, and the latter connections be very significant, in relative terms. 159 Thus, while differences exist between them, tort choice-of-law exceptions all share the same function of ensuring that torts are significantly connected with their applicable laws. Importantly, we can also see this significant connection requirement not as a limit on recognising concurrent claims, but as another reflection of the essential pre-requisite for concurrent claims: that a tort claim reflects genuine normative concerns independent of those which support a contract claim. After all, in identifying connecting factors, just as in the prior act of characterising issues, private international law seeks to give effect to the various substantive private justice ideals reflecting parties' reasonable expectations. 160 Default tort connecting factors thus reflect value judgments about the law most closely connected to a tort claim, assuming that the claim implicates certain normative substantive justice concerns (that, given the kind of wrongdoing alleged, the law of the territory where it occurs/causes damage should govern). 161 However, when the claim does not substantially raise those concerns, the default connecting factors are deemed insignificant and set aside. In their place, tort choice-of-law exceptions invoke alternative connecting factors to establish a connection between the tort and the new applicable law. And these alternative connecting factors, like the default connecting factor, are also selected on the assumption that the tort claim implicates certain other normative substantive justice concerns (e.g. that, given the kind of wrongdoing alleged, the governing law should be that of which various territorial and non-territorial connecting factors leans significantly in favour).
The selection of (default or alternative) connecting factors under (default or alternative) tort choice-of-law rules is therefore a principled affair much like the characterisation of issues: courts recognise connecting factors as relevant, and others as irrelevant, on the basis of the normative concern (s) implicated by the facts of parties' dispute. It follows that the only situation in which tort choice-of-law exceptions could choose the governing law of a closely related contract as the sole connecting factor is when the facts implicate only one normative concern: party autonomy. If other normative concerns apply, the exceptions would dilute that factor with others. 162 Only in that situation are the effects of recognising concurrent claims eliminated, because only then is the law governing parties' closely related contract selected to govern the tort claim solely because it is the law governing parties' closely related contract. However, that result in that situation is uncontroversial, because the essential pre-requisite of recognising such claims (i.e. that a claimant's separate claims be grounded on different normative concerns) is absent. Thus, properly understood, tort choice-of-law exceptions, like Wikingerhof's indispensability principle, are not really limits on the recognition of concurrent claims in private international law; the only situation in which they would have that effect is when the essential pre-requisite for recognising such claims does not exist.

D. Defending Concurrent Claims
In sum, courts should recognise concurrent claims in private international law as they do concurrent liability in domestic private law, because, in both, principled legal categories are necessary, and their occasional overlap is desirable to vindicate their underlying principles. This means that whenever a dispute implicates separate normative concerns in accordance with the forum's courts substantive theory of private (international) law, those courts should recognise concurrent claims in contract and tort. The only justifiable limit on concurrent claims is the parties' intent that jurisdiction or choice-of-law clauses in a related contract should cover tort claims as well; every other ostensible limit simply re-iterates the essential prerequisite for recognising concurrent claims, that tort claims be based on normative concerns separate from those which support contract claims.
But as noted in Section II, the opposition to concurrent claims in private international law does not stop herecommentators further argue that recognising concurrent claims would offend values and goals specific to private international law: general goals like preventing disproportionate defendant disadvantage; choice-of-law values like international uniformity; and jurisdictional values like predictability and anti-fragmentation. The argument here is that even if the existence of separate normative concerns could by itself justify concurrent claims, the need to uphold these values defeats that initial justification. Yet, properly understood, the need to uphold private international law's values requires no such thing.