INTERPRETING MULTIPLE DISPUTE-RESOLUTION CLAUSES IN CROSS-BORDER CONTRACTS

Abstract

Cross-border contracts often contain a clause which purports to reflect the parties’ intention regarding how disputes arising from their agreement should be resolved. Some such contracts might feature a “jurisdiction clause”, thus signifying the parties’ wish to subject their disputes to litigation before the courts in a specific state. Others may include an “arbitration clause”, meaning that claims arising from the contract should be subjected to an arbitral hearing. More unusual are cases in which the parties have included a jurisdiction and an arbitration clause in the same cross-border contract. This article seeks to assess English law’s approach to determining the parties’ preferred mode of dispute resolution in these more difficult cases. As it seeks to demonstrate, the current practice in this area is not always easy to defend. The article advances an alternative basis for determining which of the two competing clauses should prevail.

I. Introduction

Dispute-resolution clauses purport to reflect the parties’ intention as to how claims arising from their agreement are to be resolved. Jurisdiction and arbitration clauses are the most common examples of such provisions. ¹ A jurisdiction clause designates the court of a particular state as having competence to entertain the parties’ disputes. The parties can afford the chosen court exclusive jurisdiction, thus undertaking not to initiate a claim anywhere else. Alternatively, they can give the chosen court non-exclusive jurisdiction, thereby, retaining the option of commencing a claim in a forum other than the one designated in the agreement. By contrast, an arbitration clause evidences the parties’ intention to refer their disputes to what can be described as “a private system of adjudication”, ² rather than court-based litigation.

Typically, parties opt for either a jurisdiction or an arbitration clause in their agreements. In these instances, whether the claim falls within the scope of the relevant dispute-resolution clause hinges on the law governing that term. Under the law in England, ³ the scope of a jurisdiction clause is usually determined according to the law governing the underlying contract. ⁴ With regard to an arbitration clause, the question whether a dispute falls within it has to be determined by reference to the proper law of the arbitration agreement, ⁵ which may differ from the law applicable to the underlying contract. Ever since the House of Lords’ landmark ruling in Fiona Trust & Holding Corp. and others v Privalov and others (“Fiona Trust”) ⁶ in 2007, jurisdiction or arbitration clauses that are governed by English law, or a foreign law that is not proved to be different in content, have been construed broadly. This liberal approach is premised on the assumption that, as business people, parties are likely to have intended for all (and not just some) aspects of their disputes to be heard by the selected court or arbitral tribunal. ⁷ Consequently, the ruling in Fiona Trust is understood to be endorsing a presumption in favour of “one-stop dispute resolution” under English law. ⁸

Less typical, but more difficult to address, are cases in which the parties have included a jurisdiction and an arbitration clause in the same cross-border contract. Here, the main challenge is to determine the parties’ preferred mode of dispute resolution, which could require that the clauses be ranked. The objective of this article is to evaluate English law’s approach to such clauses. This issue has received some attention in the existing literature, perhaps most notably in an article by Richard Garnett, entitled “Coexisting and Conflicting Jurisdiction and Arbitration Clauses”. ⁹ Published over a decade ago, the article offers an assessment of how the common law courts treat multiple dispute-resolution clauses. ¹⁰ The forthcoming discussion draws on many of the same cases as those examined by Professor Garnett and shares some of the same reservations expressed in his article about the courts’ approach to interpreting multiple dispute-resolution clauses. However, this article catalogues and explains the cases differently, and, most significantly, it advances an alternative proposal for addressing the confusion that could arise from the inclusion of these clauses.

Instances where courts in England, and across the Commonwealth, ¹¹ have had to make sense of multiple dispute-resolution clauses have been surprisingly more frequent than it may have been expected. As the assessment of these cases highlights, the task of discerning the parties’ favoured mode of dispute resolution is not always as daunting as it might initially appear. In some instances, which are considered in Section II, it is possible, through a reading of the dispute-resolution clauses, or based on general principles of contractual interpretation under the relevant applicable law, to interpret which of the two provisions should give way. As such, the relevant jurisdiction and arbitration clauses in these cases can be regarded as being “naturally reconcilable”. However, there are also other instances, analysed in Section III, where the parties’ agreement features dispute-resolution clauses that are “not naturally reconcilable”. In such cases, where dispute-resolution clauses tend to take the form of exclusive jurisdiction and mandatory arbitration clauses, the intentions of the parties cannot be meaningfully ascertained by reference to the wording of the agreement or other canons of construction. As the discussion below seeks to demonstrate, while the English courts’ attempts at interpreting naturally reconcilable dispute-resolution clauses have been generally defensible, their approach to making sense of clauses which are not naturally reconcilable leaves much to be desired. Particularly questionable has been the reasoning that underpins the courts’ efforts to determine which of the two conflicting provisions should prevail – which has been said to signify “a clear but unspoken pro-arbitration bias” on their part. ¹² In response, Section IV examines possible means by which to address the confusion arising from the inclusion of multiple dispute-resolution clauses that are not naturally reconcilable. It is argued that, instead of seeking to establish the parties’ intentions in these cases – an endeavour which, because of the innate irreconcilability of the clauses, is bound to prove fruitless – a more prudent course of action would be to develop a rule of domestic law that imposes a solution to the question of which of the two dispute-resolution clauses should prevail. Based on this rule, and subject to certain narrowly-defined exceptions, priority should be attached to the designated court or arbitral tribunal first seised that is competent to entertain the claim. It is argued that this “first-seised approach” would provide a clear, pragmatic and balanced mechanism for addressing dispute-resolution clauses that are not naturally reconcilable.

II. Jurisdiction and Arbitration Clauses That Are Naturally Reconcilable

When examining the cases in which courts have had to make sense of multiple dispute-resolution clauses in the same contract, it becomes clear that there are instances where it is possible to determine the parties’ intentions. In other words, in these cases, seemingly conflicting modes of dispute resolution are indeed capable of reconciliation. These cases can be placed into two broad categories. The first category comprises cases where the question of how the two modes of dispute resolution should be read can be answered by looking at the wording of the agreement. The second category includes cases where, even though the pecking order for the jurisdiction and arbitration clauses cannot be determined from the text of the agreement, it can nonetheless be identified based on the application of general principles of contractual interpretation under the relevant applicable law. As the assessment of the cases that can be placed in each of these categories highlights, except for at least two decisions ¹³ outlined in Section III below, the courts’ treatment of naturally reconcilable dispute-resolution clauses has been generally defensible.

A. Order of Priority Discernible from the Wording of the Agreement

There are at least five different types of case where the parties’ wishes regarding the interpretation of the jurisdiction and arbitration clauses in their contract can be deciphered from the wording of those provisions. The first (and perhaps most obvious) type contains cases in which the parties have expressly outlined which of the jurisdiction or arbitration clause should prevail. ¹⁴ A helpful example here is the English Court of Appeal’s ruling in Naviera Amazonica Peruana S.A. v Compania Internacional de Seguros del Peru. ¹⁵ The dispute in this case arose in the context of a marine insurance contract between a Peruvian shipowner and a Peruvian insurer. A jurisdiction clause in favour of courts in Peru was included in the general printed terms of the contract. The terms of the agreement were subsequently varied by a typed endorsement which contained a London arbitration clause, resulting in there being two dispute-resolution clauses. However, the general printed terms contained a clause stating that, in case of conflict between the printed and typed stipulations, the latter were to take precedence. As such, the English court (rightly) afforded primacy to the parties’ choice of arbitration in London over their choice of litigation in Peru. ¹⁶

The second type of case in which it is possible to reconcile the jurisdiction and arbitration clauses based on their wording is where the parties have specified which disputes should be subjected to litigation in the chosen court and which should be referred to arbitration. ¹⁷ Consider, for instance, the Australian case of Seeley International Pty Ltd. v Electra Air Conditioning B.V. ¹⁸ Under a distribution agreement, E appointed S as its exclusive distributor of a brand of air conditioners in Australia and New Zealand. Clause 20.1 of the agreement stated that, in the event of a dispute, parties should first try to resolve them through discussions for a period of 30 days. However, the clause added that, if those discussions proved fruitless, the dispute was to be referred to arbitration in Melbourne. At the same time, Clause 20.3 outlined that the parties were at liberty to seek “injunctive or declaratory relief in the case of a material breach or threatened breach of this Agreement”. ¹⁹ A few months into the agreement, S brought court proceedings in Australia, seeking a declaration that E had acted in breach of contract and in breach of section 52 of the Trade Practices Act 1974 (Cth). In response, E pointed to Clause 20.1 of the contract and asked the court to stay its proceedings, pursuant to section 7(2) of the International Arbitration Act 1974 (Cth). In the Federal Court of Australia, Mansfield J. rejected E’s application. His Honour noted that, while Clause 20.3 did not talk about injunctive or declaratory relief from a court, it should nonetheless be understood as stating that such matters had to be determined by the designated court and were not to be referred to arbitration. ²⁰ Consequently, because S’s claim concerned declaratory relief, as stated under Clause 20.3, Mansfield J. (rightly) prioritised the choice of litigation over the choice of arbitration.

The recent decision of the English Court of Appeal in Adactive Media Inc. v Ingrouille ²¹ is another example where it could be said that, by setting out which claims should be dealt with through litigation and which by means of arbitration, the parties themselves have made clear the internal hierarchy of the dispute-resolution clauses in their agreement. AMI, a US company incorporated in Delaware, had entered a consultancy contract with I, a British citizen resident in England. Clause 15 of the contract stated that Californian law governed the parties’ agreement and that “[a]ny case, controversy, suit, action, or proceeding arising out of, in connection with, or related to this Agreement shall be brought in any Federal or State court located in Los Angeles County, the State of California”. ²² At the same time, Clause 17 of the agreement specified that “all claims, disputes, controversies, differences or misunderstandings between the parties arising out of, or by virtue of this Agreement or the interpretation of this Agreement” were to be resolved by arbitration in Los Angeles, except for any claims started by AMI concerning breaches of provisions in Clauses 7 and 8 of the agreement. Clause 7 outlined I’s obligations of confidentiality to AMI; Clause 8 included detailed terms which sought to protect AMI’s interest in I’s “work product”. ²³

Following a dispute, AMI commenced court proceedings in California, accusing I (and other parties) of a series of wrongdoings, including breaches of obligations under Clause 7 (but not Clause 8). Eventually, the Californian court handed down a default judgment in AMI’s favour. Subsequently, AMI took steps to enforce the Californian judgment against I in England. I, in turn, sought to rely on section 32(1)(a) of the Civil Jurisdiction and Judgments Act 1982 ²⁴ as a defence to the enforcement proceedings, by arguing that the Californian judgment had been obtained in breach of the arbitration clause. Effectively, I claimed that the competing dispute-resolution clauses in his agreement with AMI should be interpreted by prioritising the choice of arbitration in California over the choice of litigation before the courts in that US state. At first instance, H.H.J. Russen Q.C. rejected I’s attempt to rely on the statutory defence. The judge found that, by virtue of the carve-out expressed in Clause 17, AMI was entitled to litigate matters in California that concerned (but were not limited to) allegations relating to breaches of confidentiality obligations under Clause 7. ²⁵ Additionally, he ruled that the arbitration clause was ineffective, as it was inconsistent with the Californian jurisdiction and the choice-of-law clauses that preceded it. ²⁶

The Court of Appeal, ²⁷ however, allowed I’s appeal, holding that he could rely on the statutory defence in order to resist the enforcement of the Californian judgment. In arriving at this conclusion, the Court of Appeal construed the apparently competing dispute-resolution clauses by giving primacy to the choice of arbitration. David Richards L.J. gave the court’s only reasoned judgment. ²⁸ To begin with, His Lordship found that Clause 17 had indeed created an effective provision for arbitration because the jurisdiction and arbitration clauses were not inconsistent. ²⁹ David Richards L.J. then proceeded to examine I’s submission that the Californian judgment should not be enforced in England, because the proceedings leading up to it had been brought in breach of the arbitration provision. ³⁰ It is this aspect of the ruling that is of special interest for present purposes, as it shows the court’s approach to the construction of multiple dispute-resolution clauses in the same agreement. David Richards L.J. accepted that the exception under Clause 17, read together with the jurisdiction agreement in Clause 15, allowed AMI to initiate court proceedings in California concerning breaches of obligations under Clauses 7 and 8 of the contract. Moreover, His Lordship acknowledged that AMI’s claims did in fact include alleged breaches of the confidentiality obligations under Clause 7. Nevertheless, he was unpersuaded that the claims in California fell within the exception under Clause 17. According to David Richards L.J., for AMI to benefit from the right to bring litigation in California, as envisaged for it in the express carve-out under Clause 17, AMI’s claim had to be confined only to allegations regarding breaches of obligations under Clauses 7 and 8. In other words, the choice of Californian jurisdiction could have trumped the choice of arbitration in that state only if AMI’s allegations had been restricted to breaches under Clauses 7 and 8. However, in addition to accusing I of misuse and unauthorised disclosure of confidential information, AMI had accused him of breach of contract, breach of fiduciary duty, misappropriation of funds and conspiracy to undermine its business interests. Consequently, David Richards L.J. (rightly) concluded that the Californian judgment was not enforceable in England because it related to an action brought in breach of the arbitration clause.

The third type of case in which jurisdiction and arbitration clauses in the same contract are, upon closer inspection, naturally reconcilable is where one clause affords exclusive competence to the designated court or arbitral tribunal, while the other does not. ³¹ Suppose that A and B have entered an agreement governed by English law, which contains a non-exclusive jurisdiction clause in favour of courts in England and a mandatory London arbitration clause. In such a case, the non-exclusive English jurisdiction clause does not preclude the parties from bringing their claim elsewhere; the mandatory London arbitration clause does, however, do so. Thus, if the parties were to disagree about the mode of dispute resolution, because of its mandatory nature, the choice of arbitration in London should take precedence over the choice of litigation before the English court, which is optional. Similarly, in the reverse situation, where the agreement contains an exclusive English jurisdiction clause and an arbitration clause stating that A and B may resolve their disputes through arbitration seated in London, there can be little scope for questioning the finding under English law that the parties had intended for the choice of litigation in England to outrank the choice of arbitration in London.

The fourth type of case in which multiple dispute-resolution clauses can be said to be naturally reconcilable is where the wording of the agreement makes clear that disputes are to be resolved through one mode of dispute resolution, unless either party chooses to utilise the other. One such case where the parties had outlined their intentions along these lines was Fiona Trust. In this case, the English court was not asked to make a pronouncement about which of the two modes of dispute resolution should be prioritised. Instead, the question was whether the parties’ agreement to refer their dispute to arbitration remained effective and binding, notwithstanding the claim from one of the contracting parties that the underlying contract between them should be rescinded due to bribery. The dispute-resolution clause in Fiona Trust had stated that

(b) Any disputes arising under this charter shall be decided by the English courts to whose jurisdiction the parties hereby agree.

(c) Notwithstanding the foregoing, but without prejudice to any party’s right to arrest or maintain the arrest of any maritime property, either party may, by giving written notice of election to the other party, elect to have any such dispute referred … to arbitration in London. ³²

In contrast with the clauses discussed above, where the choice of litigation and arbitration are expressly ranked within the agreement, the relevant clause in Fiona Trust did not explicitly state which of the two forms of dispute resolution should receive priority. Nevertheless, it made it reasonably clear that the parties had envisaged litigation before the chosen court as the option of last resort, should they both decide not to refer the dispute to arbitration. Put differently, the choice of arbitration ranked ahead of litigation, in the event that either party wished to utilise it.

A similar type of clause featured in Lobb Partnership Ltd. v Aintree Racecourse Co. Ltd. ³³ However, unlike Fiona Trust, the court in Lobb was actually asked to determine which of the two modes of dispute resolution should give way. A employed L, a firm of architects, to provide services in relation to a construction project. A dispute arose between the parties, which A sought to resolve by means of arbitration. The relevant dispute-resolution clause stated that “[d]isputes may be dealt with as provided in paragraph 1.8 of the RIBA Conditions [of Engagement] but shall otherwise be referred to the English Courts”. ³⁴ Paragraph 1.8.1 of the RIBA Conditions sets out that “any difference or dispute … shall be referred by either of the parties to arbitration by a person to be agreed between the parties or, failing agreement … a person to be nominated … by the President of the Chartered Institute of Arbitrators”. ³⁵ L applied to the court for, inter alia, a declaration, pursuant to section 67 of the Arbitration Act 1996, that the arbitration agreement was void, on the basis that the use of the words “may” in the dispute-resolution clause and “shall” in paragraph 1.8.1 of RIBA gave rise to ambiguity. L contended that the consent of both parties to an arbitral hearing was needed for the reference to be valid. Colman J., however, rejected L’s arguments and application. While His Lordship accepted that the dispute-resolution provision in the agreement was “a somewhat unusually-worded clause”, ³⁶ he was still persuaded that it could be read as meaning that “either party could insist on arbitration if he chose to do so, rather than that disputes might be arbitrated only if both parties agreed upon that course”. ³⁷ As such, A’s choice of arbitration (rightly) prevailed, because the clause made it reasonably clear that either party had the right to resort to arbitration – and, failing that, the matter could be litigated before the English courts. ³⁸

The fifth type of case where the parties’ wishes regarding the interpretation of the jurisdiction and arbitration clauses are discernible from the wording is where the agreement gives party A (but not party B) the option to insist on one form of dispute resolution, thus affording priority to that mode over the other in cases initiated by A. An example of this kind of dispute-resolution clause can be found in NB Three Shipping Ltd. v Harebell Shipping Ltd. ³⁹ The case centred around a disagreement between the charterer and the shipowner regarding payments under the charterparty. Clause 47 of the charterparty contained the parties’ wishes regarding the mechanisms for resolving disputes under the charterparty. In particular, Clause 47.02 stated that “the Courts of England shall have jurisdiction to settle any disputes which may arise out of or in connection with this Charterparty but the owner shall have the option of bringing any dispute hereunder to arbitration”. ⁴⁰ The charterers brought a breach-of-contract claim against the ship-owners before the English court. Soon thereafter, the ship-owners stated their intention to refer the dispute to arbitration and sought to obtain a stay of the court proceedings pursuant to section 9 of the Arbitration Act 1996. The court granted the stay, stating (rightly) that the parties’ choice of arbitration should trump their choice of litigation. ⁴¹ Morison J. found that the dispute-resolution clause in this case should be read as being an exclusive jurisdiction clause in relation to claims initiated by the charterers, but a non-exclusive one for actions brought by the ship-owners, as it reserved them an option to refer any disputes regarding the charterparty to arbitration. Therefore, in His Lordship’s view, “Clause 47 is designed to give ‘better’ rights to owners than to charterers” ⁴² and “once owners exercise their option the parties have agreed that the disputes should be arbitrated”. ⁴³

B. Priority Through Application of General Principles of Contractual Interpretation

The second subcategory of case in which a jurisdiction and an arbitration clause in the same agreement can be naturally reconciled is where general principles of contractual interpretation under the applicable law can be relied on to ascertain which clause the parties might reasonably have intended to prevail. The following is one such rule of interpretation under English law, which is particularly relevant to the present discussion ⁴⁴ : “[G]reater weight should attach to terms which the particular contracting parties have chosen to include in the contract than to pre-printed terms probably devised to cover very many situations to which the particular contracting parties have never addressed their minds.” ⁴⁵

Therefore, in a contract governed by English law, or a foreign law that is not proved to be different in content, if the standard terms contain one form of dispute resolution, but the clauses specifically agreed by the parties include another, the latter mode should take precedence. The application of this English canon of construction can help to explain the ruling in Indian Oil Corp. v Vanol Inc. ⁴⁶ The sale-of-goods contract in contention incorporated the plaintiffs’ standard terms and conditions, which included a clause stating that parties should attempt to resolve their disputes “by mutual consultation” or, failing that, by referring them to arbitration in India. ⁴⁷ The contract also included additional provisions expressly agreed by the parties, which stipulated that “all disputes arising [under the agreement] shall be submitted to the jurisdiction of the English courts”. ⁴⁸ The main issue for consideration was whether the plaintiff’s claim was time-barred by means of the parties’ agreement. Nevertheless, before addressing this question, Webster J. relied on the general rule that a specifically agreed term prevails over a conflicting provision in the general terms, observing (obiter, and rightly) that the English jurisdiction clause in this case trumped the Indian arbitration clause. ⁴⁹

In sum, there are many instances where the parties’ seemingly conflicting modes of dispute resolution are, in fact, naturally reconcilable. As a result, when attempting to construe multiple dispute-resolution clauses, the first step ought to be to examine whether it is indeed possible to decipher the parties’ intentions as to how their disputes are to be resolved either from the wording of the agreement or through the application of general principles of contractual interpretation under the relevant applicable law. It is argued that, to the extent that the parties’ intentions as to how (and where) their disputes are to be resolved can be determined by using these techniques, they should be honoured.

III. Jurisdiction and Arbitration Clauses That Are Not Naturally Reconcilable

However, there are other instances where the parties’ intentions concerning their preferred mode of dispute resolution cannot be meaningfully determined by reference to the wording of the agreement or the relevant canons of construction. In these instances, the dispute-resolution clauses – which tend to take the form of an exclusive jurisdiction clause and a mandatory arbitration clause – are incapable of reconciliation. The discussion now focuses on how the English courts have sought to make sense of these clauses. As it sets out to demonstrate, the courts’ approach in this area is open to criticism and ripe for reconsideration.

The starting point in understanding the courts’ treatment of jurisdiction and arbitration clauses that are not naturally reconcilable is the first-instance ruling in Paul Smith Ltd. v H&S International Holding Inc. ⁵⁰ Paul Smith concerned a licensing agreement under which the plaintiffs, designers and manufacturers of clothing, granted a licence to the defendants to manufacture, promote, distribute and sell their products in North, Central and South America. Clause 13 of the agreement, headed “Settlement of Disputes”, stated that:

If any dispute or difference shall arise between the parties hereto concerning the construction of this Agreement or the rights or liabilities of either party hereunder the parties shall strive to settle the same amicably but if they are unable to do so the dispute or difference shall be adjudicated upon under the Rules of Conciliation and Arbitration of the International Chamber of Commerce by one or more Arbitrators appointed in accordance with those Rules. ⁵¹

This provision was followed by Clause 14, entitled “Language and Law”, which stated that “[t]his Agreement is written in the English language and shall be interpreted according to English law. The Courts, of England shall have exclusive jurisdiction over it to which jurisdiction the parties hereby submit”. ⁵²

The plaintiffs accused the defendants of breach of contract. Following negotiations, the defendants referred the claim to an arbitral hearing under ICC Rules, which made a number of preliminary determinations – namely, that the arbitration agreement was valid, that three arbitrators should be appointed and that London was the place of arbitration. Not long thereafter, the plaintiffs commenced court proceedings in England, seeking, inter alia, an injunction to restrain the defendants from continuing with the arbitration. In this respect, they argued that, because of the inconsistency between Clauses 13 and 14, the arbitration agreement should be ruled ineffective. Steyn J. was, however, unpersuaded by this submission, regarding it to be a “drastic and very unattractive result … [as it] involves the total failure of the agreed method of dispute resolution in an international commercial contract”. ⁵³ Instead, His Lordship considered that the two provisions focused on different matters: the arbitration provision was concerned with the mode of dispute resolution, while the jurisdiction agreement represented the parties’ intention as to the law governing the conduct of the arbitration. ⁵⁴ Put differently, as far as the resolution of substantive disputes was concerned, the parties’ choice of arbitration was given priority over their choice of litigation. Accordingly, Steyn J. dismissed the plaintiffs’ application, with the result that the arbitration could proceed.

Steyn J.’s reasoning in Paul Smith has received much criticism within the academic literature. For example, in an article published in the Lloyd’s Maritime and Commercial Law Quarterly, Paul Tan has advanced the following observations as to the questionability of the rationale underpinning the ruling in Paul Smith:

First, in cases where the arbitration clause already identifies the seat of the arbitration, it is entirely artificial (and unnecessary) to construe the competing clause conferring jurisdiction on the courts as identifying the courts having supervisory jurisdiction over the arbitration. Although courts will not acknowledge it, upholding the arbitration clause in such circumstances in effect amounts to striking down the competing jurisdiction clause. Second, the decision to uphold the arbitration clause at the expense of the competing jurisdiction clause is often made in cases where the evidence may have been at best neutral as to which of the dispute resolution mechanisms the parties objectively intended. Indeed, the evidence usually pointed away from arbitration. Decisions between competing jurisdiction clauses thus appear to be influenced by a clear but unspoken pro-arbitration bias too often too mechanically applied in disregard of the overriding purpose of contractual interpretation, which is to give legal effect to the objective intentions of the parties. ⁵⁵

In the same vein, Professor Garnett has considered that Steyn J.’s reading of the dispute-resolution clauses was “plainly at odds with the parties’ agreement” because, “by the time the matter was heard, the ICC had confirmed London as the seat of arbitration and so, strictly speaking, English procedural law [had become applicable] in any event”. ⁵⁶ Indeed, as suggested in a leading practitioner text, while “it may be possible” to arrive at the sort of construction as the one preferred by Steyn J. in Paul Smith, such an interpretation is not “really plausible, because it does not need saying” that the courts of the place where the arbitration is seated have supervisory jurisdiction over the arbitration. ⁵⁷

A few years after the decision in Paul Smith, Steyn J.’s approach to the construction of exclusive jurisdiction and mandatory arbitration clauses in the same contract was somewhat modified by Moore-Bick J. in Shell International Petroleum Co. Ltd. v Coral Oil Co. Ltd. ⁵⁸ The contract between the parties contained two dispute-resolution clauses. Under Clause 13, entitled “Applicable Law”, the parties had stated that “[t]he Agreement, its interpretation and the relationship of the parties hereto shall be governed and construed in accordance with English law and any dispute under this provision shall be referred to the jurisdiction of the English courts”. ⁵⁹ Clause 14 of the agreement, headed “Arbitration”, then set out that the parties’ disputes “shall be finally and exclusively settled by arbitration by three arbitrators in London England in accordance with the rules of the London court of International Arbitration”. ⁶⁰ The defendant had set out to commence proceedings in Lebanon, which prompted the plaintiff to apply for an anti-suit injunction on the basis of the dispute-resolution clauses. In these circumstances, the issue for the court’s consideration was whether the inclusion of a London arbitration clause and an exclusive jurisdiction clause in favour of courts in England meant that the parties’ dispute-resolution clauses were void for inconsistency and removed the ground for the plaintiff to obtain the injunction. In answering this question, Moore-Bick J. drew on Steyn J.’s judgment in Paul Smith. However, unlike Steyn J., who had all but entirely focused on the parties’ agreement to refer their dispute to arbitration at the expense of the exclusive English jurisdiction clause, Moore-Bick J. assumed some role (albeit a more limited one) for the jurisdiction clause. Accordingly, His Lordship sought to reconcile the two provisions by stating that “the parties did intend substantive disputes between them to be referred to arbitration”, but their disagreements regarding the proper law should be addressed by means of litigation in London. ⁶¹ Having concluded, on the basis of this reading of the clauses, that the dispute-resolution clauses in the agreement were not inconsistent, Moore-Bick J. granted the injunction to restrain the defendant from bringing a claim in Lebanon.

Notwithstanding Moore-Bick J.’s efforts to give some role to the jurisdiction clause for dispute-resolution purposes, it is argued that the reasoning in support of his construction of the clauses remained artificial and, hence, unconvincing. His Lordship simply prioritised the choice of arbitration in London over litigation before the English court in relation to the determination of the parties’ substantive disputes and held that the English court’s role was confined to hearing disputes “about the proper law”. Other commentators, too, have been critical of Moore-Bick J.’s reasoning. For example, Professor Garnett has considered that the judgment in Shell International “ignored a clear inconsistency between the clauses” and that Moore-Bick J.’s “strategy for reconciliation of the clauses is even more questionable” as “[i]t [was] not at all clear what ‘disputes about the proper law’” entailed. ⁶²

Despite these shortcomings, the approaches in Paul Smith and Shell International remained influential in shaping subsequent legal developments in this area. The next significant case, which drew (but also built) on those decisions, was Ace Capital Ltd. v CMS Energy Corp. ⁶³ Unlike the dispute-resolution clauses in Paul Smith and Shell International, the jurisdiction and arbitration clauses at the centre of the dispute in Ace Capital were, in fact, naturally reconcilable. Decided in 2008, the dispute in Ace Capital arose from insurance policies involving an English underwriter, A, and the insured, C, a Michigan corporation. The agreement between A and C contained, inter alia, a mandatory London arbitration clause and a service-of-suit clause in favour of courts in the US. ⁶⁴ C brought court proceedings against A in Michigan, seeking a payment under the insurance policy. In response, A initiated a court claim in England, asking for an injunction to restrain C from continuing with the court proceedings in Michigan.

In many respects, the dispute-resolution clauses in Ace Capital fall within the third type of case identified in Section II(A) above, where one clause affords exclusive competence to the designated court or arbitral tribunal, while the other does not. The dispute-resolution clauses in the parties’ agreement were, therefore, naturally reconcilable. The court could have allowed the claimant’s injunction application, based entirely on the wording of the clauses, and given primacy to the mandatory choice of arbitration in London over what was effectively a non-exclusive choice of litigation before a US court. Although, in the end, Christopher Clarke J. granted the injunction, thus effectively ruling that the US jurisdiction clause should give way to the London arbitration clause, His Lordship did not reach this conclusion based on a natural reading of the provisions. Rather, he interpreted them much like Moore-Bick J. had construed the dispute-resolution clauses in Shell International. ⁶⁵ Accordingly, Christopher Clarke J. held that substantive disputes between the parties should be referred to arbitration in London. ⁶⁶ However, His Lordship afforded a somewhat wider (but still limited) scope to the US jurisdiction clause, meaning that the US courts had competence “to declare the arbitrable nature of the dispute, to compel arbitration, to declare the validity of an award, to enforce an award, or to confirm the jurisdiction of US courts on the merits in the event that the parties agree to dispense with arbitration”. ⁶⁷

Notably for present purposes, in the course of his reasoning Christopher Clarke J. pointed to the “strong legal policy on both sides of the Atlantic in favour of arbitration” ⁶⁸ as a factor which prompted him to find that the choice of arbitration in London prevailed over the choice of litigation in the United States. He also drew on the House of Lords’ decision in Fiona Trust, where, as mentioned above, the House of Lords had ruled that dispute-resolution clauses should be read liberally, observing that:

the principle of liberal interpretation in favour of arbitration encourages, as it seems to me, not only an expansive reading of what an arbitration clause includes but also a restrictive reading of any other clause which is said, notwithstanding an arbitration clause providing for all disputes to be referred to arbitration, to exclude particular disputes from arbitration (either generally or at one party’s option), without expressly saying so. ⁶⁹

In short, Christopher Clarke J. considered that the wider pro-arbitration policies required that, in cases containing the construction of provisions akin to those before him, the arbitration clause should be given a broad meaning and any other dispute-resolution clause that is within the same agreement should be interpreted restrictively.

It is argued that this specific ground for giving primacy to the mandatory arbitration over exclusive jurisdiction clauses is not particularly persuasive. It is true that courts have almost invariably honoured the parties’ agreement to refer their disputes to arbitration, limited any interference with arbitral process and all but routinely enforced arbitral awards. However, it is hard to accept that these wider pro-arbitration proclivities justify prioritising the choice of arbitration over the choice of litigation when the parties have not made it clear which should prevail. Jurisdiction and arbitration clauses are both examples of party autonomy. ⁷⁰ Therefore, while giving primacy to the choice of arbitration in cases involving mandatory arbitration and exclusive jurisdiction clauses may seem like a pro-arbitration practice, it is a step that could be harmful to arbitration more generally, in as much as it contravenes party autonomy – a principle of significant importance in international commerce. ⁷¹

Be that as it may, the same pro-arbitration considerations that supported Christopher Clarke J.’s reasoning in Ace Capital have continued to be influential in subsequent cases where courts have been called upon to make sense of jurisdiction and arbitration clauses that are not naturally reconcilable. In this context, perhaps the most illustrative example is the decision in Sulamerica Cia Nacional de Serguros S.A. & others v Enesa Engenharia S.A. ⁷² In this case, decided in 2012, E, a Brazilian engineering company, was engaged in the construction of a large hydroelectric facility in Brazil. To protect itself against possible legal liabilities, E took out an insurance policy with S, a Brazilian insurance provider. Among other terms, the insurance policy contained an exclusive jurisdiction clause in favour of courts in Brazil and a mandatory London arbitration clause. Following a number of incidents, E made claims under the insurance policy. S, in response, refused to accept liability, arguing that the losses in question did not fall within its scope. Subsequently, S initiated arbitration proceedings in London, seeking, inter alia, a declaration of non-liability. E, however, chose to start court proceedings in Brazil, claiming that S had no right to commence the arbitration. In due course, the Brazilian court awarded an injunction to restrain S from pursuing the arbitration. For its part, S applied for an injunction from the English court to restrain E from continuing with the court proceedings in Brazil.

One of the key issues which the English court had to address was which of the two modes of dispute resolution – namely, arbitration in London and litigation in Brazil – should prevail. ⁷³ In answer to this question, Cooke J. found that the choice of arbitration in London outranked the choice of litigation in Brazil. In arriving at this conclusion, His Lordship drew on Christopher Clarke J.’s pronouncements in Ace Capital, stating that:

The English courts, when faced with an exclusive jurisdiction clause and an arbitration agreement, look to the strong legal policy in favour of arbitration and the assumption that the parties, as rational businessmen, are likely to have intended any dispute arising out of the relationship into which they have entered to be decided by the same tribunal. Unless expressly provided otherwise, the parties must be taken to have agreed on a single tribunal for the resolution of all disputes. A liberal approach to the words chosen by the parties in their arbitration clause must now be accepted as part of our law. I follow in this regard the comments of Christopher Clarke J. ⁷⁴

Cooke J. then proceeded to read the clauses in much the same manner as Christopher Clarke J. had in Ace Capital. Consequently, Cooke J. ruled that all disputes or disagreements of substance had to be determined through arbitration in London. According to His Lordship, the jurisdiction clause afforded competence to Brazilian courts to decide on “the arbitrable nature of the dispute, to compel arbitration, to declare the validity of the award, to enforce the award, or to confirm the jurisdiction of the Brazilian courts on the merits in the event that the parties agree to dispense with arbitration”. ⁷⁵

In short, courts in England appear to have settled on a clear approach to interpreting multiple dispute-resolution clauses that are not naturally reconcilable. Based on this approach, they have almost always prioritised arbitration in the designated venue over litigation in the chosen court, thus essentially ignoring the presence of an exclusive jurisdiction clause in the agreement. Notwithstanding the problems with the reasoning underpinning the courts’ treatment of multiple dispute-resolution clauses that are not naturally reconcilable, some may still insist that, for the following reasons, the criticisms of the current approach are somewhat exaggerated.

First, it may be considered that the choice of a mandatory arbitration clause should prevail over an exclusive jurisdiction clause because of section 9 of the Arbitration Act 1996. The provision allows a party who has been sued in England in breach of an agreement to refer the disputes to arbitration – be it in England or elsewhere – to apply for a stay of proceedings. ⁷⁶ When faced with such an application, the court must grant a stay, provided that the defendant has not taken any steps to respond to the substantive claim in the court proceedings ⁷⁷ and that the arbitration agreement is not found to be null and void, inoperative, or incapable of being performed under its proper law. ⁷⁸ Second, it may be observed that, in the type of cases considered in this part, it is defensible, as a matter of construction, to prioritise the choice of arbitration because the inclusion of a mandatory arbitration clause could be read as signifying the parties’ intention to resolve their disputes by a means other than the default mechanism of litigation before a court.

However, it is argued that both of these reasons for maintaining the existing treatment of multiple dispute-resolution clauses that are not naturally reconcilable are themselves open to challenge. Take the first observation that, by virtue of section 9 of the Arbitration Act 1996, a mandatory arbitration clause should always prevail. As noted by Professor Garnett, “such an observation is correct where court proceedings are brought in breach of an arbitration agreement and no competing jurisdiction clause exists”. ⁷⁹ However, it is argued that the provision cannot be used as a basis for the proposition that arbitration should prevail as a mode of dispute resolution where the agreement contains a mandatory arbitration clause and an exclusive jurisdiction clause that are not reconcilable, as it says nothing about this issue. To say that an exclusive jurisdiction clause should always come second to a mandatory arbitration clause is also hard to defend, given that English law upholds exclusive jurisdiction clauses with much the same vigour as mandatory arbitration clauses. For example, where the parties have chosen the courts in England to have exclusive jurisdiction and the jurisdiction agreement falls within the Hague Convention, ⁸⁰ the court must hear the matter unless the agreement is null and void under English law, including its choice-of-law rules. ⁸¹ Likewise, in cases where proceedings are initiated in England in breach of an exclusive jurisdiction clause in favour of a court within a Contracting State to the Hague Convention, the English court must ordinarily stay its proceedings. ⁸² Even at common law, where the English courts have a discretion whether to hear proceedings commenced in England in breach of an exclusive foreign jurisdiction clause, they have almost invariably upheld the parties’ jurisdiction agreement by granting a stay. ⁸³ Indeed, such has been the level of respect shown at common law towards exclusive jurisdiction clauses that the editors of Dicey, Morris & Collins on the Conflict of Laws have observed that the Hague Convention “does little to alter the existing common law”. ⁸⁴ Equally questionable is the observation that the inclusion of a mandatory arbitration agreement points to the parties’ intention to swap the default mode of dispute resolution (i.e. courts) for an alternative one (i.e. arbitration). Surely such an intention can be ascribed to the parties in those cases where they have included only an arbitration clause in their agreement and not those where they have included a mandatory arbitration clause alongside an exclusive jurisdiction clause. In view of the above analysis, it is argued that the current treatment under English law of multiple dispute-resolution clauses that cannot be naturally reconciled is unsatisfactory and ought to be rethought. To this end, the next part of the discussion focuses on finding an alternative (and more defensible) basis for addressing the conflict inherent in these provisions.

IV. Rethinking the Approach to Jurisdiction and Arbitration Clauses That Are Not Naturally Reconcilable

A number of possible options present themselves when searching for an alternative means by which to make sense of jurisdiction and arbitration clauses that are not naturally reconcilable. At one end of the spectrum, there is the option to treat these clauses as being void for inconsistency. An example where the dispute-resolution clause was deemed to be so vague as to be ineffective is the English Court of Appeal’s ruling in E.J.R. Lovelock, Ltd. v Exportles. ⁸⁵ Lovelock was not concerned with how multiple dispute-resolution clauses in the same contract should be treated. Rather, it was a case where the problem of interpretation arose due to the confusing wording of a single arbitration clause. Nevertheless, it would be helpful to consider the case, as it exemplifies the type of situation in which a dispute-resolution clause could be ruled invalid for inconsistency.

The dispute in Lovelock was in relation to a sale-of-goods contract involving a Russian seller and an English buyer. The contract contained a two-part arbitration clause. The first part purported to refer “[a]ny disputes and/or claim” to arbitration before English arbitrators, while the second stated that claims should be subjected to arbitration in Russia in accordance with the Russian Chamber of Commerce Arbitration Rules. The buyers started breach-of-contract proceedings in England against the sellers. The sellers applied to the English court to stay those proceedings, arguing that the dispute should be determined by means of arbitration in Russia. In rejecting the sellers’ stay application, the Court of Appeal found that the arbitration clause was void for inconsistency. ⁸⁶

Lovelock is, however, an exceptional decision and should not be seen as representing the English courts’ general attitude to interpreting confusingly-worded dispute-resolution clauses, including those which contain both a jurisdiction and an arbitration clause. Indeed, an assessment of the case law, and associated academic commentary, highlights that the preferred approach in England (and elsewhere in the common law world) has always been to try to give effect to dispute-resolution clauses, however incoherently or clumsily they have been articulated in the contract. ⁸⁷ The following passage in Steyn L.J.’s judgment in The Star Texas – a case concerned with the interpretation of an arbitration clause – serves to show the existence of this stance on the part of the English courts:

The spectre of a catalogue of possible alternative constructions may at first glance seem to confront us with a daunting task. The reality is different. The fact that a multiplicity of possible meanings of a contractual provision are put forward, and that there are difficulties of interpretation, does not justify a conclusion that the clause is meaningless. The Court must do its best to select, among the contending interpretations, the one that best matches the intention of the parties as expressed in the language they adopted. And, in a case where there are realistic alternative interpretations of an arbitration clause, the Court will always tend to favour the interpretation which gives a sensible and effective interpretation to the arbitration clause. ⁸⁸

Therefore, when asked to construe vaguely-phrased (including multiple) dispute-resolution clauses, the courts’ principal task is to determine the common intention of the parties. At the same time, however, courts should be wary not to entertain interpretations which deprive “ordinary words of their ordinary meaning and unwarrantably [seek] to foist upon them a quite extraordinary connotation”. ⁸⁹ For these reasons, it is not entirely plausible to suggest that jurisdiction and arbitration clauses that are not naturally reconcilable should be read as though they cancel each other out.

At the other end of the spectrum is the option of allowing for both arbitration in the chosen venue and proceedings in the designated court to be brought concurrently. This means of treating the provisions might be seen as an improvement on the status quo, as it does not simply overlook the exclusive jurisdiction clause. Nevertheless, for a host of (mostly practical) reasons, it is not an appealing alternative and should not be adopted. The following passage in Professor Garnett’s article sums up why it would be unworkable (and, indeed, unwise) to allow for both litigation and arbitration to proceed in parallel: “Not only do parallel proceedings lead to significant additional costs and inconvenience for parties but there is also a great risk of inconsistent results between the tribunals leading to great complexity at the stage of enforcement of any judgment or award.” ⁹⁰

In his 2013 article, Professor Garnett proposed an alternative basis for prioritising dispute-resolution clauses that are otherwise incapable of natural reconciliation. Characterised by Professor Garnett as the “more appropriate forum test”, this approach sits between the two extreme approaches already outlined (and dismissed). It seeks to make sense of these clauses by ranking the choices of litigation and arbitration “on the basis of which forum is more appropriate to resolve the dispute, with recognition to be given to the tribunal which has been first seised by a party”. ⁹¹ The passage below in Professor Garnett’s article highlights how the more appropriate forum test would operate:

If a party sues first in the courts of country X pursuant to a contractual clause which stipulates such courts, then its choice should generally be given deference, assuming the suit is brought in good faith and the forum has some connection with the parties and the subject matter of the dispute. Likewise, if a party seeks to trigger arbitration by issuing a notice to arbitrate or seeking an order from the court to appoint the arbitral tribunal before any court proceedings on the merits have been filed, then the arbitration clause should be given priority, absent any vexatious conduct by the party. Commencing litigation in a non-stipulated forum, however, carries no weight. ⁹²

There is much to recommend the more appropriate forum test as a basis for determining which of the two chosen modes of dispute resolution should prevail. The test does not suffer from the shortcomings afflicting the other two possible options discussed above. Additionally, it is consistent with the wider approach at common law in England which draws on discretionary principles in addressing jurisdictional disputes. Finally, and most importantly, unlike the English courts’ current practice, the more appropriate forum test seeks to afford equal treatment to both modes of dispute resolution. In many ways, therefore, its adoption would represent an improvement on the English courts’ existing approach to the construction of jurisdiction and arbitration clauses that are not naturally reconcilable.

Despite these advantages, the test suffers from at least two major drawbacks which, it is argued, render it a less than compelling model to embrace. First, it is not entirely clear how the more appropriate forum test would be enforced. This problem would arise in cases where, even though the chosen forum in England has concluded that it is more appropriate, the designated foreign forum in the other clause nonetheless decides to proceed with hearing the dispute. The second drawback with the more appropriate forum test concerns the wider practical implications arising from its application. While an important element under the test is the forum that was first seised, the weight attached to it, for the purpose of the test, is merely greater as opposed to decisive. Other considerations, “including the purpose behind such proceedings and the connections between the tribunal and the parties and subject matter” are also significant in enabling the contracted court or arbitral tribunal in England to decide whether to hear the matter or to order it to be dealt with by the other agreed court or arbitral venue. ⁹³ Therefore, the more appropriate forum test would be applied in much the same manner as the English forum non conveniens doctrine. The range of possible considerations in play when applying the more appropriate forum test, along with the enquiry’s fact-specific and discretionary nature, are bound to lead to a number of problems. For one, parties would be encouraged to gather and present excessively large volumes of evidence and witness testimony in order to persuade the chosen court (or arbitral tribunal) that the other designated forum is more appropriate. In such circumstances, it is entirely conceivable that the process of deciding which of the two irreconcilable modes of dispute resolution should prevail could become unduly protracted and overly costly. Moreover, and given the potentially wide range of factors that could be consulted in determining the prevailing mode of dispute resolution under the test, the enquiry is bound to generate unpredictability for commercial parties, as it would be very difficult to tell with reasonable certainty, from the outset, which of the two would be given priority.

It is argued that a better approach to the construction of multiple dispute-resolution clauses that are not naturally reconcilable would be one that possesses the advantages of the more appropriate forum test, without having its drawbacks. For this purpose, the first step has to be to accept that, because of the innate irreconcilability of the clauses, it is not possible to discern the parties’ intentions meaningfully. Instead, a more prudent course of action would be to develop a rule of domestic English law that imposes a solution to the question of which of the two dispute-resolution clauses should prevail. Based on this rule, and subject to certain limited exceptions outlined below, as a starting point, the designated court or arbitral tribunal first seised that is competent to entertain the claim, should outrank the court or arbitral tribunal stipulated in the other clause. Such a basis for deciding the internal hierarchy of the clauses can be labelled the “first-seised” approach. The following two hypothetical scenarios seek to show how the general rule under the first-seised approach would work in practice:

1. A and B are parties to a contract which contains an exclusive jurisdiction clause in favour of the courts in England. The contract also includes a term which states that the parties’ disputes should be referred to arbitration in Hong Kong. A accuses B of breach of contract and commences court proceedings against B in England, before B has taken any steps to refer the matter to arbitration. The English court has competence to hear the matter under the law governing the jurisdiction clause. B applies to the English court to stay its proceedings, pursuant to section 9 of the Arbitration Act 1996. Under the first-seised approach, the choice of litigation in England should take precedence. The choice of arbitration in Hong Kong should be deemed to have given way, meaning that B has no basis to apply for a stay under section 9 of the Act.

2. X and Y enter a contract containing a clause which states that all disputes relating to it should be subjected to the exclusive jurisdiction of courts in England. The contract also includes a clause which stipulates that the parties should refer their disputes to arbitration in London. Subsequently, X accuses Y of breach of contract, and proceeds to seek a court order to appoint an arbitral tribunal to hear the dispute, as stipulated by the arbitration clause within the contract. The arbitral tribunal has competence under the proper law of the arbitration agreement. According to the proposed approach, the choice of arbitration in London should trump the choice of litigation before the English courts. As a rule of English domestic law, X’s engagement of one mode of dispute resolution (here, arbitration in London) results in the other (i.e. litigation in England) becoming outranked.

The first-seised approach might be regarded as being hard to enforce and, ultimately, unconvincing. More specifically, it could be observed that a court or arbitral tribunal specified in the other clause may proceed to hear the claim, even though the matter is already pending before the contracted court or tribunal first seised. In such circumstances, there is a real risk of parallel proceedings and, potentially, irreconcilable court or arbitral rulings. Doubts about the enforceability of the proposed approach are bound to be strongest in those cases (such as the one in scenario 1 above) where the court or arbitral tribunal specified in the other dispute-resolution clause is abroad. ⁹⁴ Notwithstanding these reservations, it is argued that the first-seised approach can be enforced by granting anti-arbitration or anti-suit injunctions. At first blush, it might not seem appealing to rely on these measures because, in these cases, the court or arbitral tribunal which becomes second seised has been stipulated in the other clause. However, granting injunctions to uphold the first-seised approach is not as drastic or unreasonable as it might initially appear. To support this observation, it would be helpful to return to scenario 1. Suppose that, after A’s commencement of court proceedings against B in the stipulated forum in England, B nonetheless proceeds, pursuant to the other dispute-resolution clause, to initiate an arbitral hearing against A in Hong Kong. In such a situation, the English court can issue an anti-arbitration injunction, ⁹⁵ since, under the proposed approach, upon the English court’s assertion of jurisdiction over the matter, the choice of litigation in England would be deemed as having trumped the choice of arbitration in Hong Kong. It is argued that commencing the action in Hong Kong, after the other contracted forum (in England) has been seised, is akin to bringing the claim in a non-contracted forum and, for that reason, the party initiating it can be restrained from pursuing it by means of an injunction.

Likewise, injunctions can be used to enforce the first-seised approach where, after the commencement of an English arbitral hearing, court proceedings are brought before the chosen foreign court. The facts in scenario 1 can be altered to illustrate this point. Suppose that, this time, the agreement between A and B contains a mandatory London arbitration clause and an exclusive Hong Kong jurisdiction clause, and that B commences court proceedings in Hong Kong after A’s initiation of arbitration in London. The fact that the arbitration in the designated venue was engaged ahead of litigation before the stipulated court, and the arbitral tribunal is competent to hear the dispute, can provide the basis for A to seek an anti-suit injunction. Again, the subsequent court claim in Hong Kong is, by virtue of the operation of the first-seised approach, tantamount to an action brought in a non-contracted forum, thus enabling the defendant in those proceedings to apply to the English court for an anti-suit injunction. ⁹⁶

In most cases, prioritising the choice of dispute resolution that is first seised, as a matter of course, would provide a workable and unbiased method for treating jurisdiction and arbitration clauses that are not naturally reconcilable. Nevertheless, there are likely to be limited occasions where insisting on this general rule could lead to outcomes that are unwelcome. In such instances, it would be reasonable to allow the contracted court or arbitral tribunal in England a limited discretion to decline to uphold the general rule under the first-seised approach. This discretionary power is far more limited in scope than the one envisaged under the more appropriate forum test and would only come into play in very exceptional circumstances. As such, its incorporation into the first-seised approach would not lead to the sort of problems that would be generated by the application of the more appropriate forum test. The list of exceptions where it could be warranted to depart from the general rule under the first-seised approach on a discretionary basis is not closed. However, the following are circumstances where allowing a potential departure from it would seem most apt.

The first exceptional situation could arise where, having assumed competence over the dispute, for reasons outside the control of the parties, the chosen court or arbitral tribunal first seised is unable to proceed with the matter. Suppose that a contract between A and B contains a Utopian jurisdiction clause and a London arbitration clause that are naturally irreconcilable. A commences breach-of-contract proceedings against B in Utopia. At some point thereafter, civil war breaks out in Utopia, with the result that Utopian courts are closed indefinitely. If B subsequently initiates proceedings in England, with the view of initiating an arbitral hearing, it would be sensible to allow the court a discretion to prioritise the choice of arbitration in London, even though it was not first seised, provided that B can establish that the Utopian court is unable to hear the matter.

It would also be imprudent to insist on the general rule under the first-seised approach where, owing to developments outside the control of the parties, at least one of them would be discriminated against on racial, religious or political grounds in the forum first seised. Consider the case in which C and D are parties to a contract that contains an exclusive Narnian jurisdiction clause and a mandatory London arbitration clause. C initiates court proceedings in Narnia against D, pursuant to the jurisdiction clause. Suppose that in the period between entering into the agreement and the commencement of the Narnian proceedings, the ruling party in Narnia has adopted openly discriminatory policies towards individuals or corporations from D’s home state. In these circumstances, in the event that D chooses to commence court proceedings in London, in order to appoint an arbitral tribunal, there would be merit in allowing the court a narrow discretion to give priority to the choice of arbitration, if it is persuaded that the contracted forum in Narnia cannot dispose of the dispute justly.

The third instance where a relaxation of the general rule under the first-seised approach would be warranted could arise in the context of multi-party disputes. Assume, for example, that A and B are parties to a contract which contains an exclusive jurisdiction clause in favour of courts in England and a mandatory Ruritanian arbitration clause. Before the courts in England, B initiates a negative-declaration claim against A. Subsequently, A starts arbitral proceedings in Ruritania against B and co-defendants, C and D, alleging that they are all parties to a conspiracy against A. In the English proceedings, A challenges the court’s exercise of jurisdiction, by pointing to the Ruritanian arbitration clause and arbitral proceedings. In such a situation, it would be advisable to allow the English court a limited discretion to give way to the Ruritanian arbitral proceedings, in so far as it is satisfied that such a step is necessary to avoid the risk of inconsistent rulings.

Finally, it would be prudent to depart from the general rule under the first-seised approach where related proceedings involving the same issues are already ongoing between one of the parties and a third-party in the other forum. Suppose, for instance, that X and Y are parties to a contract which features multiple dispute-resolution clauses that are not naturally reconcilable: an Arcadian jurisdiction clause and an English arbitration clause. Pursuant to the arbitration clause, and in order to appoint an arbitral tribunal, X commences court proceedings in England against Y. Y challenges the order, pointing to the Arcadian jurisdiction clause and the fact that it has already started litigation in Arcadia against Z concerning allegations that are related to those in the dispute between Y and X. Here, too, there would be merit in allowing the English court to give priority to the choice of litigation in Arcadia, on a discretionary basis, if it is satisfied that doing so would avoid the risk of irreconcilable decisions being rendered.

The first-seised approach is not without shortcomings of its own. Its introduction could tempt parties to press ahead with their preferred mode of dispute resolution under the agreement sooner than they otherwise might have done, making it less likely for their disagreements to be resolved amicably. Nevertheless, it is argued that, on the whole, its adoption would represent an improvement on the status quo. By stating that, as a general rule, priority should be given to the tribunal first seised that has competence to hear the matter, the proposed approach in this article would provide a clear, predictable and unbiased mechanism for ascertaining the internal hierarchy of dispute-resolution clauses that are not naturally reconcilable. Moreover, by allowing for departures from the general rule in very exceptional circumstances, the approach seeks to avoid some of the more serious problems that could arise as a result of the strict adherence to that starting position, while at the same time ensuring that the overall predictability and simplicity of the proposed model are not undermined.

At this juncture, it might be helpful to examine how some of the decided cases examined in Section III above, which contained irreconcilable jurisdiction and arbitration clauses, would be decided under the first-seised approach. In all of these cases, the general rule under the proposed approach would apply, as none of them falls within the sort of exceptional circumstances that would warrant a departure from it. It is suggested that cases such as Paul Smith and Sulamerica would be decided in the same way: the parties’ choice of arbitration would trump their choice of litigation. After all, in each case, the relevant steps in order to start arbitral proceedings in London were taken before the commencement of the court claims in England and Brazil, respectively. Of course, the outcome would have been different had the court proceedings in these cases been brought ahead of the arbitral hearing. The outcome would also be the same in Shell International. In this case, the court was not asked directly to determine how the relevant jurisdiction and arbitration clauses should be ranked. Instead, the question was whether the parties’ insertion of a mandatory London arbitration clause and an exclusive jurisdiction clause in favour of courts in England meant that their dispute-resolution choices were void for inconsistency. Under the first-seised approach, the court would still grant an injunction, because the proceedings in Lebanon would have been in breach of the jurisdiction and arbitration clauses that featured in the agreement.

However, there is at least one case that is likely to be decided differently based on the first-seised approach: Melford Capital Partners (Holdings) LLP v Digby. ⁹⁷ Although cited, this recent English High Court ruling was not discussed in Section III above, as the judgment largely reiterated the English courts’ approach in cases like Paul Smith, Shell International and Ace Capital. The facts of the case are complex and need not be recited here in detail. It suffices to say that, fundamentally, the dispute concerned whether the claimants, two partnerships, had validly expelled the defendant, one of their former partners. The relevant partnership agreement for present purposes contained both an exclusive jurisdiction clause in favour of courts in England (Clause 27.2), ⁹⁸ and a London arbitration clause (Clause 28). ⁹⁹ After a breakdown in relations, the claimants expelled the defendant. They then commenced court proceedings in England, accusing the defendant of breach of confidence and seeking an injunction to restrain him, among other things, from using information that they claimed belonged to the partnerships. In response, the defendant brought a defence and counterclaim, inter alia, questioning the validity of his dismissal. A few months later, the claimants brought arbitration in London, pursuant to Clause 28 of the relevant partnership agreement. In those proceedings, they sought to obtain a determination that the defendant had been validly dismissed from the partnerships. The claimants then applied to the English court, seeking for the defendant’s counterclaim to be stayed, pursuant to section 9 of the Arbitration Act 1996, as it concerned matters that should be determined through arbitration. In response, the defendant argued that the stay application should be dismissed because of the conflicting dispute-resolution choices in the relevant agreement.

Charles Morrison, sitting as a Deputy Judge of the High Court, rejected the defendant’s argument, stating that the parties’ substantive disputes should be referred to arbitration in London, with the English court merely having a supervisory jurisdiction over any arbitration. ¹⁰⁰ Therefore, he allowed the claimants’ application for a stay of proceedings under section 9 of the Act. In arriving at this conclusion, the judge relied on reasoning concerning the interpretation of multiple dispute-resolution clauses in the cases discussed above, in particular Paul Smith, Shell International and Ace Capital. ¹⁰¹ It is argued that this case would be decided differently under the approach proposed in this article: the parties’ choice of litigation before the English court would outrank their choice of arbitration in London, as the claimants brought the court proceedings in England prior to initiating the London arbitration. Thus, under the first-seised approach, substantive disputes between the parties would be resolved through litigation before the English court.

V. Conclusion

This article has sought to evaluate English law’s approach to interpreting multiple dispute-resolution clauses that appear in the same cross-border contract. As the discussion has sought to illustrate, the occasions where the courts have been called upon to construe these terms have been far from isolated. Predictably, it is generally more difficult to discern the parties’ intentions as to how claims arising from their agreement are to be resolved when they have chosen two modes of dispute resolution rather than just the one. That said, at least in some instances, the exercise has proven to be less daunting than at first may have been feared. These instances include cases where the jurisdiction and arbitration clauses can be regarded as being “naturally reconcilable” – that is to say, where it is possible to decipher the parties’ intentions as to how their disputes are to be resolved either from the wording of the agreement or through the application of general principles of contractual interpretation. It is argued that, to the extent that the internal hierarchy of the dispute-resolution clauses can be determined by using these techniques, courts should uphold it.

However, the existing case law reveals also instances in which multiple dispute-resolution clauses are “not naturally reconcilable” – particularly where the agreement contains an exclusive jurisdiction clause and a mandatory arbitration clause. As this part of the discussion has sought to demonstrate, the reasoning underpinning the courts’ approach to determining which of the two provisions should give way is hard to defend, not least because it points to a clear pro-arbitration stance on the part of the courts. As a result, they have almost routinely prioritised the choice of arbitration, while showing little regard for the presence of the jurisdiction clause in the agreement.

Against this backdrop, a number of possible alternative approaches for making sense of multiple dispute-resolution clauses that are not naturally reconcilable were examined. It was argued that, because of the inherent conflict between these clauses, it would be futile to seek to ascertain the parties’ intentions as to which one of them should take precedence. Instead, a more fruitful course of action would be to develop a rule of domestic law that would state how the clauses should be ranked. According to this rule, and subject to a limited range of exceptions, the designated court or arbitral tribunal first seised that is competent to entertain the claim should outrank the court or arbitral tribunal stipulated in the other clause. This “first-seised” approach would provide a clear, predictable and unbiased basis for addressing the confusion regarding which of the two conflicting dispute-resolution choices in the parties’ agreement should prevail.