1 Introduction

International contracts are often drafted in a rather standardised manner, making use of so-called boilerplate clauses that aim at regulating the interpretation and operation of the contract. In addition, they often contain an arbitration clause that submits all disputes to arbitration, thus excluding any involvement of national courts in disputes arising out of the contract.

Standardising contract terms, including a boilerplate legal framework and referring disputes to a private resolution mechanism are elements that seem to indicate an intention to render the contract self-sufficient. By including a detailed and extensive regulation of the legal relationship between the parties, the contract aims at making national law dispensable. If national law is not relevant, and the only basis for regulating the parties’ legal relationship is the contract, it becomes possible and meaningful to standardise contract terms even when contracts are intended to be implemented in a variety of legal systems, without the need to adapt them to the legal framework of the specific transaction. The impression of self-sufficiency is enhanced by the exclusion of national courts and the referral to arbitration instead. A closed circuit is created, dominated by the will of the parties: the relationship is regulated by terms of contract agreed to by the parties, and disputes are solved by a private body that must follow the parties’ instructions. External sources, including national law, may seem redundant.

Self-sufficiency may seem a realistic goal as long as the legal relationship remains within the borders of the closed circuit. This assumes that the legal relationship at any time is subject to the terms and legal framework agreed upon between the parties.

There are, however, situations in which this assumption may prove false: for example, if a difference arises between the parties, and they disagree on what the legal framework is (notwithstanding that they may have agreed in the past, prior to the conflict); if third parties’ interests or public interests are affected, and mandatory rules or policies override the parties’ agreement; or if the agreed terms or legal framework may be interpreted in more than one way or need specification by external sources. In these situations, the closed circuit is interrupted and recourse to external sources becomes necessary. To a certain extent, guidance may be sought in non-national, non-authoritative rules that may permit a uniform, transnational solution and thus reinstate the closed circuit. Where such uniform guidance is not available, again the closed circuit is interrupted. When a full closed circuit cannot be assumed, party autonomy may be limited.

To assess the limits of party autonomy, therefore, it will be necessary to analyse the above-mentioned situations where interference with the closed circuit may occur. Section 2 briefly discusses to what extent the legal framework provided by the contract and possibly given effect to in arbitration may resist control and interference by national law. Section 3 discusses to what extent the terms of the contract are capable of being interpreted in a uniform manner. Section 4 discusses to what extent transnational sources may provide a uniform legal framework capable of replacing national governing law. Section 5 investigates the extent to which the principle of faithful interpretation to contract wording may be a guiding principle for arbitral tribunals.

2 External Limits to Party Autonomy: Court Control

The closed circuit described above meets the contracting parties’ expectations of self-sufficiency as long as the arbitral tribunal gives effect to the will of the parties as embodied in the contract and the award is complied with by the losing party or enforced by the courts. The closed circuit fails when an arbitral award becomes invalid or unenforceable as a consequence of having given effect to the contract terms.

As is well known, international arbitration is an alternative method to solving contractual disputes based on the consensus of the parties. If the parties agree to submit disputes between them to arbitration, then the ordinary courts will have to decline jurisdiction over those disputes. The only possible mechanism to solve the dispute will be the arbitration chosen by the parties. If, on the contrary, the parties have not entered into an arbitration agreement, disputes between them will have to be solved by a national court that has jurisdiction. An arbitral tribunal, in other words, bases its existence upon the parties’ agreement. Moreover, the parties determine the composition of the arbitral tribunal, the procedural rules which it is obliged to follow, the scope of the tribunal’s competence and its power. The arbitral tribunal is bound to follow the instructions of the parties; otherwise, it exceeds the power that the parties have conferred on it. If the arbitral tribunal exceeds its power, neither its jurisdiction nor its award are founded on the parties’ agreement, and there is, consequently, no legal basis for either two. These basic elements of arbitration are based on the 1958 New York Convention on Recognition and Enforcement of Foreign Arbitral Awards, ratified by about 150 countries and reflected in most national arbitration laws, as well as in the UNCITRAL Model Law on International Commercial Arbitration, issued in 1985, revised in 2006 and adopted in about 60 countries.

Arbitration’s dependence on the parties’ will, which is uniformly recognised, is an important factor strengthening the opinion that (i) arbitration is a private matter between the parties, (ii) the arbitral tribunal is bound to follow the parties’ instructions, and (iii) national courts or State laws cannot interfere with the parties’ will. This opinion is certainly confirmed by the observation that the vast majority of arbitral awards are voluntarily complied with by the losing party. The parties agree to submit the dispute to arbitration, then they instruct the arbitral tribunal as to the scope of the dispute, the rules to be applied, etc., then the losing party recognises the arbitration’s result and complies voluntarily with the award. In such situations, the totality of the arbitration takes place in the private sphere of the parties. There is no point of contact between the national courts and the arbitration. Consequently, there will be no national judge who may decide to override the parties’ contract or expectations by considering an agreement invalid because it, say, violates European Union (EU) competition law, or by considering a contract not binding because, for instance, one of the parties did not have legal capacity under the law to which it is subject. The arbitrators may or may not decide to apply these rules, but, as long as the losing party accepts the result of the arbitration, there will be no possibility for a judge to verify the arbitrator’s decision. In these cases, therefore, limits to party autonomy are relevant only to the extent that the arbitral tribunal is requested by the parties, or elects by its own volition to apply State law to the dispute.

When the losing party does not voluntarily comply with the award, the courts will intervene. In these cases, the closed circuit is interrupted and limitations to party autonomy may become relevant.

The formal framework for arbitration grants it a relative autonomy, actually giving the appearance of an autonomous system. The main instrument upon which arbitration is founded is the already mentioned New York Convention. The Convention binds the courts of the States Parties to recognise arbitration agreements and thus dismiss claims that are covered by these agreements, as well as to recognise and enforce arbitral awards without any review of the merits or of the application of law. Only a restrictive and exhaustive list of grounds on which to refuse recognition and enforcement exists. Another important instrument is the previously mentioned UNCITRAL Model Law, which has contributed to a considerable harmonisation of the areas of arbitration law that are not covered by the New York Convention. The UNCITRAL Model Law is, in turn, based on the same principles as the New York Convention, which means that, together, these instruments create a harmonised legal framework for arbitration. Both instruments give a central role to the will of the parties. The power of the arbitral tribunal in fact derives from the agreement of the parties; therefore, the arbitral tribunal is obliged to follow the parties’ instructions in respect of the scope of the dispute, the applicable law, the potential remedies granted and so forth.

All this confirms to a large extent the understanding of arbitration as an autonomous system, based on the will of the parties and detached from national law. However, both the New York Convention and the UNCITRAL Model Law refer to national, non-harmonised legislation in a number of instances and thus reduce in few, but significant respects the detachment of arbitration from national laws. Thus, national law defines: (i) what may be subject to arbitration; (ii) when an award is deemed to conflict with public policy; (iii) the criteria according to which an arbitration agreement is binding on the parties; (iv) which mandatory rules of procedure apply; and (v) when an award is valid. In these situations, the closed circuit is interrupted.

To name a few examples: a contract between a Norwegian and a Ukrainian party was submitted by the parties to Swedish law. After a dispute arose and arbitration was initiated, the Ukrainian party maintained that it was not bound by the contract, because its representatives had signed the contract in such a way that it did not meet the formal requirements of Ukrainian law. The arbitral tribunal chose to follow Swedish law as stipulated by the contract, considered that the contract had been validly signed according to Swedish law and disregarded Ukrainian law as irrelevant. The arbitral tribunal, therefore, fulfilled the closed circuit; however, the award was set aside by the courts of the country where it was rendered, Sweden, because the legal capacity of a party is subject not to the law chosen by the parties in the contract, but to the law of each of the parties. Here, the closed circuit was interrupted and party autonomy restricted.

In another example, the European Court of Justice (ECJ) found that an award would be invalid and not enforceable for violation of public policy, if it gave effect to a contract that did not comply with competition law. Were the arbitral tribunal willing to follow the terms of the contract in full, the award would not be valid or enforceable; this is, therefore, another limitation to party autonomy.

Another example is a decision by a Russian court, refusing to enforce an award that had given effect to a shareholders’ agreement among the shareholders of a Russian company. The agreement regulated the parties’ rights and obligations in a manner that did not comply with Russian company law, and the court found that enforcing the award would have violated Russian public policy.

The harmonised framework for arbitration is, therefore, subject to national law in several significant respects, and this may have an impact on the enforceability of arbitration agreements and of arbitral awards, which in turn restricts the effects of party autonomy.

3 Terms of Contract: Absolute and Uniform Interpretation?

Bearing in mind the exceptions seen in Section 2, the liberal framework for arbitration permits to recognise and enforce awards even if an award were based on a wrong interpretation of a contract or the evidence, or were it to apply the applicable law wrongly or apply the wrong law. Therefore, if the award gives effect to a regulation contained in the contract, it will most likely be recognised and enforced even though the contract may disregard and violate the applicable law. Arbitration, therefore, to a large extent seems to permit relying on the assumption of the closed circuit. This, however, does not imply that party autonomy is absolute.

An absolute party autonomy, unaffected by external elements, assumes that the terms of the contract have a uniform meaning flowing from the words, and that they therefore may be interpreted equally in all legal systems. It is, however, not uncommon that contract terms need to be understood in light of assumptions and effects founded in the applicable legal framework.

Even plain words may acquire different meanings, depending on the culture and tradition of the interpreter. Take an apparently self-explanatory expression such as ‘summer nights’. If read by an Italian, it will create associations with a dark and warm night, possibly with crickets singing and a sky full of stars. If read by a Norwegian, it will evoke a bright and chilly night, with the sun as the only visible star. If the meaning of ordinary words is affected by the context, contract terms are even more vulnerable to this as they refer not to a natural phenomenon, but to legal effects that are created and supported by legal systems, which in turn use words as the most important means to create and define those legal effects. It then becomes even more difficult to separate the legal effects from the words used to express them. In this situation, it may be illusory to expect that the terms of a contract have an absolute meaning, fully independent of any legal framework or legal tradition.

National legal systems may differ from each other in many respects that are relevant to a contract, even when the systems belong to the same legal tradition or so-called legal family. Legal differences will be even greater across legal families, such as between the common law and the civil law systems. Modern comparative law research is inclined to consider this divide as overrated and largely overcome by a common core of European contract law. Elsewhere I argue that the common core reveals a certain synchrony between the systems on an abstract level, but that it does not necessarily lead to harmonised solutions on a specific level. Awareness about a common core may show that a certain principle may be recognised and a certain result may be achieved in a plurality of legal systems, albeit by employing different legal techniques. In a specific case, however, it is the particular legal technique employed in the contract that counts, and not the abstract possibility of achieving the desired result, if only the right legal technique had been adopted.

3.1 Force Majeure clause

An example of a contract term that may have different legal effects depending on the legal framework, is the so-called Force Majeure clause. This clause is meant to excuse a party’s non-performance of its obligation, if fulfilment were prevented by an event which was beyond that party’s control, unforeseeable and could not be reasonably overcome. One question is how the requirement of ‘beyond control’ shall be interpreted. Interpretation may be influenced by the legal system’s understanding of the assumptions about liability.

An illustration of this situation is where a producer cannot fulfil its obligations because it did not receive raw materials from its supplier. The question is whether failure by a party’s supplier may be deemed as an event falling outside of that party’s sphere of control. To answer this question, it is necessary to understand the purpose of the Force Majeure clause.

There may be several goals for a regulation on exemption from liability for non-performance. In some legal systems, the aim is to allocate between the parties the risk for supervening unexpected events according to which one of the two parties is closer to bear that particular risk. This approach assumes a strict liability, which is triggered irrespective of the conduct of the party that was prevented from performing its obligations.

According to an alternative approach, the risk for unexpected events should not be borne by a party, as long as that party has acted diligently and cannot be blamed for the occurrence of the impediment - even if in an objective allocation of risk that party would be closer to bear such risk.

The legal systems that follow the criteria of the strict liability and the allocation of risk between the parties according to the respective spheres of control, would consider the choice of supplier to be an event falling within the sphere of control of the seller. Certainly this impediment would not fall within the sphere of the buyer and, since all risks have to be allocated between the parties, it follows that it must fall within the sphere of the seller. That the producer has been diligent in selecting its supplier and cannot be blamed for the supplier’s failure to deliver is not relevant.

This is the approach taken by English law. German law, however, has a different approach. According to §276 of the German Civil Code (Bürgerliches Gesetzbuch; BGB), if the prevented party is to be blamed for the impediment or its consequences, it cannot be excused from liability. If, however, the prevented party can prove that it has not acted negligently, it will be excused from liability. If the seller has operated with diligence in the choice of supplier, it would not be considered liable for non-performance due to failure by the supplier.

The distinction between common law and civil law in the context of liability for non-performance can be explained by the English system’s inclination to privilege predictability for the sake of ensuring that business is carried out smoothly, rather than ensuring that an equitable justice is arrived at in a given case. Common law allocates the risk of non-performance between the parties according to where it is most closely to expect that the risk should be borne. This objective rule is not to be defeated by subjective criteria such as lack of negligence, because it would render the system less predictable. Civil law systems privilege (in different degrees) the subjective elements of a specific case, in order to ensure that an equitable solution is reached.

The interpretation of the Force Majeure clause, therefore, is not absolute and uniform, but depends on the governing law.

3.2 Entire Agreement clause

Another example of a term of contract that may be interpreted differently depending on the legal framework, is the so-called Entire Agreement clause. This is a recurring clause in contract practice, and states that the document signed by the parties contains the whole agreement and may not be supplemented by evidence of prior statements or agreements.

The purpose of the Entire Agreement clause is to isolate the contract from any source or element that may be external to the document. This is also often emphasised by referring to the four corners of the document as the borderline for the interpretation or construction of the contract. The parties’ aim is thus to exclude that the contract is integrated by terms or obligations that do not appear in the document.

The parties are, of course, entitled to regulate their interests and to specify the sources of their regulation. However, many legal systems provide for ancillary obligations deriving from the contract type, a general principle of good faith or from a principle preventing an abuse of rights. This means that a contract would always have to be understood, not only on the basis of the obligations that are spelled out therein, but also in combination with the elements of the applicable law which integrate it. A contract, therefore, risks having different content depending on the governing law: the Entire Agreement clause is meant to avoid this uncertainty by barring the possibility of invoking extrinsic elements. The Entire Agreement clause creates an impression of exhaustiveness of the written obligations.

This is, however, only an illusion: first of all, ancillary obligations created by the operation of law often may not be excluded by the contract. Moreover, some legal systems permit bringing evidence that the parties’ agreement creates obligations which differ from those contained in the contract. Furthermore, many civilian legal systems openly permit the use of pre-contractual material to interpret the terms written in the contract. Finally, a strict adherence to the clause’s wording may, under some circumstances, be looked upon as unsatisfactory even under English law, in spite of the formalistic interpretation style that English law may employ in respect of other clauses.

The effect of the clause, therefore, does not flow from its simple words, but is the result of a combination of the clause and of the governing law.

3.3 No absolute and uniform effects

The examples set out above show that national law may be relevant even in situations where the assumption of the closed circuit is not challenged by judicial control. Even though arbitral tribunals in these situations are allowed to consider exclusively the terms of the contract without running the risk of triggering invalidity or unenforceability of the award, they may find that the terms of the contract are not a sufficient basis for the decision and must be integrated by external elements.

4 Transnational Law: A Uniform Legal Framework?

We have seen above that contract terms are not capable of being interpreted without making reference to the applicable legal framework. This interrupts the closed circuit. Admittedly, arbitration may (to a certain extent, as was seen in Section 1) be capable of giving effect to the regulation agreed to by the parties in the contract without being obliged to comply with the peculiarities of the applicable law. However, the terms of the contract are not self-explanatory and have to be interpreted in light of the applicable legal framework, as was seen in Section 2. That the arbitral tribunal is free to interpret the contract and to decide how, if at all, the contract shall interact with the applicable law, does not give an answer to the question of how to interpret terms that are not self-explanatory. This may result in different interpretations of the same contract terms depending on the arbitrator’s background and inclination, and thus impacts on party autonomy.

It is worthwhile exploring whether the idea of an absolute party autonomy may be reinstated by including a uniform legal framework into the closed circuit. It is often proposed that transnational sources may give a uniform legal framework for international contracts. Transnational sources are concerned with giving effect to commercial practice without abiding by the peculiarities of the various legal systems; this could be deemed to make national laws redundant.

The differences among the various national legal systems have prompted various initiatives to formulate transnational sets of rules, in part developed spontaneously by business practice and in part restated and codified by branch organisations, international organisations, academic fora, etc. This complex of sources goes under various names, such as (new) lex mercatoria, transnational law or soft law. If transnational sources gave an exhaustive and harmonised regime, it would be possible to include these sources as the only applicable legal framework for the contract and thus reinstate the closed circuit.

As I argue elsewhere, however, transnational sources are not sufficiently precise and systematic to replace national laws – not to mention the formal circumstance that transnational sources may not, as a matter of private international law, govern a contract to the exclusion of any State laws.

Some of the most recognised transnational sources – in particular, the UNIDROIT Principles of International Commercial Contracts (UPICC) and the Principles of European Contract Law (PECL), are heavily based on a general principle of good faith. Good faith is a legal standard that needs specification and there does not seem to be any generally acknowledged legal standard of good faith that is sufficiently precise to be applied uniformly, irrespective of the governing law.

Moreover, these instruments grant the interpreter much room for interference regarding the wording of the contract – based on the central role given to the principle of good faith. This seems to contradict the very intention of standard contracts. International contract practice is meant to be exhaustive and self-sufficient, and not influenced by the interpreter’s legal tradition. Any correction by principles such as good faith would run counter to the expectations by the parties.

Neither is a solution provided by the instruments developed so far in the progressing work on a European contract law. The Academic Draft Frame of Reference, the Acquis Principles and the 2011 proposal for a Regulation on a Common European Sales Law, all largely based on the PECL, have a double approach to commercial contracts: they extend rules of consumer protection to commercial contracts (including an extensive and mandatory principle of good faith), and they then moderate these by reserving for contrary good commercial practice. Reference to good commercial practice as the only concretisation of the principle of good faith assumes that the interpreter is in a position to define good commercial practice and to assess its content. What constitutes good commercial practice, however, is not clear. It may be assumed that it coincides with the abovementioned spontaneous or academic transnational sources that often are deemed particularly apt to govern international contracts and go under the name of transnational law or lex mercatoria: scholarly works on the convergence of legal systems, general principles, restatements and trade usages. As shown below, these sources are not capable of giving a clear and harmonised picture of the transnational law of commercial contracts; hence, they do not clarify what good commercial practice is. Reference to good commercial practice, therefore, is not a sufficient concretisation of the standard of good faith.

Transnational sources, thus, do not always give a uniform solution. An arbitrator who needs to interpret contract terms will not find a definitive and uniform standard of interpretation in these sources, and will need to make recourse to other sources. This will again interrupt the closed circuit.

4.1 Force Majeure clause

To test the ability of transnational law to overcome the disparity of legal traditions, we can look at the examples given in Section 2 above. We saw that the expression ‘beyond the control’ in Force Majeure clauses may be interpreted differently depending on the governing law. Does transnational law offer a uniform solution?

One of the most successful instruments of contract law harmonisation is the 1980 Vienna Convention on Contracts for the International Sale of Goods (CISG). It is ratified by over 60 countries and looked upon, especially in some academic circles, as embodying principles that are generally recognised and reach well beyond the convention’s scope of application.

According to Article 79 of the CISG, a party is not liable for failure to perform its obligations, if it proves that the failure was due to an impediment beyond its control, which was unforeseeable and that could not reasonably have been overcome.

The CISG does not contain any reference to the diligence of the affected party as a criterion for exempting it from liability; in another context, the Convention confirms that diligence is not a criterion for excuse: Articles 45(1) (b) and 61(1)(b) provide that each party may exercise contractual remedies for non-performance against the other party without having to prove any fault or negligence or lack of good faith on that party, nor do they mention that any evidence of diligence would relieve the other party from its liability.

The Secretariat Commentary does not address the question of how the criterion of the sphere of control shall be interpreted, whether literally, or as a reference to the diligent conduct of the seller. Bearing in mind that the CISG shall be interpreted autonomously, without reference to domestic legal systems, it seems appropriate to interpret the Convention literally and see Article 79 objectively dividing the landscape into two spheres, that of the seller and that of the buyer, without reference to specific actual possibilities to exercise control. This is confirmed by case law and doctrine: procurement risk falls within the seller’s sphere of risk Therefore, failure by the seller’s supplier does not fall outside of the seller’s sphere of responsibility (unless the relevant good has disappeared completely from the international market). In the comment to the second paragraph of Article 79 on use of sub-contractors, the Commentary specifies that this special rule does not include suppliers of raw material or of goods to the seller.

However, this is not the only way of understanding the criterion of ‘beyond the control’. Article 79 of the CISG may be interpreted differently, depending on the interpreter’s legal tradition – something that has been designated as ‘troubling’.

Norway implemented the CISG with the Act on Sale of Goods. The latter introduced in §27 the concept of impediment beyond the control of the prevented party, with a literal translation of Article 79 of the CISG. By introducing this concept, the legislator intended to mitigate the then existing regime, which was based on strict liability.

Norwegian legal doctrine interprets the criterion of ‘beyond the control’ not as having an abstract understanding of each party’s sphere of control, but on the basis of the actual sphere of control of each party. Only if one party actually has the possibility of influencing a certain process, events caused by that process are deemed within the sphere of control of that party. That a party has started a process does not, in itself, mean that any events occurring in the course of that process are in that party’s sphere of control. The test must be if that party actually had the possibility of influencing the part of the process in connection with which those events occurred. Hence, in the case of procurement risk the interpretation of what is ‘beyond the control’ is opposite to the outcome under the CISG: the producer chose its supplier, and this choice is certainly within the producer’s sphere of control (it could have chosen another supplier, and then the default would not have happened). However, the producer has no actual possibility of influencing the performance of the supplier; therefore any impediment in connection therein shall be deemed outside of its sphere of control.

In conclusion, the CISG does not seem to provide a uniform standard for the interpretation of Force Majeure clauses.

4.2 Entire Agreement clauses

The other example of a contract term with inconsistent legal effects made in Section 2 above, is the Entire Agreement clause.

This clause is recognised in Article 2.1.17 of the UPICC and Article 2:105 of the PECL, with some restrictions: the provisions specify that prior statements or agreements may be used to interpret the contract. This is one of the applications of the general principle of good faith; it is, however, unclear how far the principle of good faith goes in overriding the clause inserted by the parties. If prior statements and agreements may be used to interpret the contract, does this mean that more terms may be added to the contract? Say, for example, that the parties have discussed certain specifications at length during the negotiations and this has created in one of the parties the reasonable expectation that these specifications would be implied in the contract even though they were not included in the final contract text. Article 1.8 of the UPICC would seem to indicate that this would be the preferred approach under the UPICC. According to this provision, a party may not act in a way inconsistent with reasonable expectations that it has created in the other party. This is spelled out in the PECL Article 2:105 (paragraph 4) of which states that ‘A party may by its statements or conduct be precluded from asserting a merger clause to the extent that the other party has reasonably relied on them.’

According to this logic, the detailed discussion during the phase of negotiations of certain characteristics for the products may create the reasonable expectation that those specifications have become part of the agreement even if they were not written in the contract. Their subsequent exclusion on the basis of the Entire Agreement clause may be deemed to be against good faith.

According to the opposite logic, however, the very fact that the parties have excluded from the text of the contract some specifications that were discussed during the negotiations indicates that no agreement was reached on those matters. Exclusion of those terms from the contract, combined with the Entire Agreement clause, strongly indicates the will of the parties not to be bound by those specifications. Their subsequent inclusion on the basis of the good faith principle would run counter the parties’ intention.

The foregoing shows that the application of the UPICC and of the PECL requires a specification of the principle of good faith. Is it intended as an overriding principle, possibly creating, restricting or modifying the obligations that flow from the text of the contract? Or is it meant to take the text of the contract as a starting point, ensuring that the obligations contained therein are enforced accurately and precisely as the parties have envisaged them? This represents the dichotomy between, on the one hand, the understanding of fairness as a principle ensuring balance between the parties-notwithstanding the regulation that the parties may have agreed on- and, on the other hand, the understanding of fairness as a principle ensuring predictability, and leaving it to the parties to evaluate the desirability of their contract regulation. This dichotomy characterises the different approaches of the common law and the civilian tradition. To enhance the ability of the UPICC to harmonise contract law, in 1992 UNIDROIT has created a data base collecting court decisions and arbitral awards on the various provisions of the UPICC. This is, therefore, the best source to turn to when inquiring how to interpret the Entire Agreement clause under the UPICC.

As per 2013, the Unilex database contained five decisions on Article 2.1.17 of the UPICC. Elsewhere, I have analysed these decisions, revealing that they are not based on a consistent understanding of the standard according to which the clause shall be applied. The Unilex database shows two approaches to Article 2.1.17 of the UPICC: one advocating the primacy of the contract’s language, and the other assuming that the UPICC provides for the primacy of the parties’ real intentions, which in turn may lead to considerably restricting the effect of the Entire Agreement clause.

Evidently, this is not sufficient to provide guidance as to which approach to choose when addressing the conflict between the contract’s language and the principle of good faith. This leaves so much room to the discretion of the interpreter, that it seems unlikely for UPICC Article 2.1.17 to give a harmonised regulation of its subject-matter. The UPICC, therefore, does not contribute considerably to a harmonised standard of interpretation.

4.3 No uniform legal framework

The examples made above show that submitting a contract to the CISG, the UPICC or other transnational sources does not necessarily ensure that the contract will be interpreted independently of the applicable legal tradition. The transnational law, therefore, does not ensure the closed circuit.

5 Arbitral Tribunals: Faithful to the Intention of the Parties

Above we have seen that the arbitral tribunal may, to a large (but not unlimited) extent, disregard the governing law without consequences for the validity and enforceability of the award; we have also seen that this is not a sufficient answer to the question of how to interpret contract terms when they are not self-explanatory; and we have further seen that it is not always possible to find a uniform standard of interpretation in translation sources. A principle that is often invoked in these circumstances is that the arbitral tribunal has a duty to be faithful to the will of the parties. Does this give sufficient guidelines? The arbitral tribunal may certainly not be inclined to let the terms of the contract be overridden by the formalities of the various national legal systems, but that does not give an answer to the question of how contract terms shall be interpreted.

We can assume a long-term loan agreement with an Early Termination clause permitting immediate termination of the contract and consequently the immediate repayment of the whole principal upon breach of the obligations contained in a certain clause. A literal interpretation of the Early Termination clause permits termination even when the breach is insignificant – for example, when the borrower has submitted its financial statements to the lender with one day delay. The breach may have had no consequences on the borrower’s creditworthiness, on its ability to repay the loan, or on the lender’s ability to verify these matters. The real reason for the lender to terminate the loan may have been that the interest rates had increased since the time of signing the loan, and that the lender sought to use the threat of early termination as an effective leverage for negotiating a higher interest rate. This would not be relevant in a literal interpretation: the clause would be considered applicable without regard to the real reasons for which it is invoked.

A purposive interpretation of the clause takes into consideration the purpose of the clause and tries to assess whether the particular situation may be deemed to fall into the scope of the clause. This may lead to considering the clause inapplicable in a situation where the reasons for which it is invoked do not correspond to the purpose of the clause.

What is more faithful to the intention of the parties: a literal implementation of the clauses that may permit speculative or abusive conduct, or an integration of the clauses with considerations of business purpose, good faith and trade usages? There seems to be no absolute answer to the question of which interpretation better meets the expectations of the parties: a strictly literal interpretation of the terms of the contract, or an integration of the contract with principles of good faith and commercial sense based on law, trade usages, transnational principles or other sources. The former would better reflect the parties’ expectations, assuming that the parties have consciously intended to achieve specific legal effects with each word that they have written in the contract. This, however, does not reflect the reality of how contracts are drafted and negotiated, as is demonstrated below.

5.1 The dynamics of contract drafting

As described elsewhere, the dynamics of contract drafting often involve inserting some of the clauses in a contract without the parties having given any particular consideration to their content or their effects under the applicable law.

This practice may be surprising, considering the importance that the governing law has for the application and even the effectiveness of contract terms, as was seen above. However, the practice of negotiating detailed wording without regard to the governing law, or even of inserting contract clauses without having negotiated them, is not necessarily always unreasonable. From a merely legal point of view, it makes little sense, but from the overall economic perspective, it is more understandable. The gap between the parties’ reliance on the self-sufficiency of the contract and the actual legal effects of the contract under the governing law does not necessarily derive from the parties’ lack of awareness regarding the legal framework surrounding the contract. More precisely: the parties may often be conscious of the fact that they are unaware of the legal framework for the contract. The possibility that the wording of the contract is interpreted and applied differently from what a literal application would seem to suggest, may be accepted by some parties as a calculated risk.

Considerations regarding the internal organisation of the parties are also a part of the assessment of risk. In large multinational companies, risk management may require a certain standardisation, which in turn prevents a high degree of flexibility in drafting the single contracts. In balancing the conflicting interests of ensuring internal standardisation and permitting local adjustment, large organisations may prefer to enhance the former.

It is, in other words, not necessarily the result of thoughtlessness if a contract is drafted without having regard for the governing law. Neither is it a symptom of a refusal of the applicability of national laws. It is the result of a cost–benefit evaluation, leading to the acceptance of a calculated legal risk. The sophisticated party, aware of the implications of adopting contract models that are not adjusted to the governing law and consciously assessing the connected risk, will identify the clauses that matter the most, and concentrate its negotiations on those, leaving the other clauses untouched and accepting the corresponding risk.

A faithful interpretation of the contract assumes an understanding of this uneven approach to contract drafting.

5.2 The need for predictability

While contract drafting may be uneven, predictability is extremely important in commercial contracts. The parties are interested in enforcing their rights, and for this purpose they depend on one or more national legal systems and their courts. Once a contract is finalised, therefore, parties are interested in its enforceability and in the predictability of the parameters according to which enforcement may be achieved.

Litigation lawyers carefully analyse the specific contract and its effects under the governing law, and try to assess as precisely as possible the possibility of winning a case in court or in arbitration on the basis of the contract wording, the applicable law and the degree of factual background that the governing law allows to bring into the dispute. Thus, on the one hand, drafting lawyers, while negotiating a contract, may have willingly disregarded the legal effects of some clauses. On the other hand, litigation lawyers, while assessing enforceability of the same contract, will carefully study its legal effects under the governing law. The varying degree of awareness during negotiations must accordingly be considered in light of the need for predictability once a dispute arises.

Furthermore, contracts are often meant to circulate, for example, because they are assigned to third parties, are used as security or serve as a basis for calculating insurance premiums. In these situations, it is essential that contracts are interpreted strictly in accordance with their terms: third parties are not aware of and should not be assumed to take into consideration the relationship between the original parties to the contract, what the original parties may have assumed or intended, or any circumstances that relate to the original parties and that may have had an impact on these parties’ interests. It is, therefore, expected that a contract is interpreted primarily, if not exclusively, in light of its terms – without considering things such as what a fair balance between the parties’ interests would be or what one party’s expectations might have been.

5.3 How to square the circle: the applicable law

The arbitral tribunal is expected to understand the dynamics of negotiations in order to properly give effect to the intention of the parties. Blindly applying the wording of the contract without any regard to the principles of the governing law or, to the extent that they are determinable and applicable, of transnational law, would not necessarily reflect the true intention of the parties if the clause that is being applied literally is one of the boilerplate clauses that the parties did not consider. Yet integrating or correcting a clause with national or transnational principles might not necessarily reflect the parties’ intention either, if the clause that is being interpreted is one of the clauses that the parties carefully negotiated.

Leaving broad discretion to the interpreter, however, runs the risk of undermining predictability, if the criteria for exercising such discretion are not clearly determinable. As was seen above, interpretation of the contract should take into consideration the need for predictability. Overriding the terms of the contract in the name of principles of good faith or equity would hence lead to results that are not compatible with the expectations of international business practice, if the standards that are applied are not clearly determinable. From the overview made in section 3 above, it seems that the standard of good faith is not sufficiently determinable on a transnational level. It may therefore be advisable to take in to consideration the criteria developed in the applicable law.

5.4 Variety of approaches

There is no uniform answer to the question of what interpretation is the most faithful to the parties’ intentions. A seminar organised at the University of Oslo in 2011 discussed the arbitrators’ approach to the interpretation of contracts and identified a variety of approaches.

Contracts are not necessarily always applied in strict accordance with their terms. There are different degrees of interference and the sources of the interference also vary quite considerably. There is a scale moving from a strict application of the governing law to integrate the contract, via interpretation of the contract terms in the context of transnational soft law principles such as the UPICC and the PECL (which are heavily based on the principle of good faith and may give rise to a substantial possibility of interfering with the contract language), to interpretation of the contract on the basis of its own terms combined with the parties’ interests and trade usages, to interpretation of the contract solely on the basis of its own terms. There is also a further approach to interpretation of the contract, which goes under the label of ‘splitting the baby’. This Solomonic approach consists of rendering an award in the middle range between the claims of each of the parties. This is not necessarily based on a literal consideration of the contract terms or on an integration of the contract with other sources, but simply on the desire to accommodate the interests of both parties. Interestingly, there does not seem to be a uniform perception of the frequency of this approach. A recent empirical study shows that the parties to arbitration perceive that they received a Solomonic award in 18–20% of the cases, whereas the arbitrators perceive that they take this kind of equitable decision in only 5% of the cases. This, therefore, adds a new variable to the equation of the interpretation of contracts. Not only is it uncertain whether the arbitrators will interpret the contract literally, whether they will use sources of law or whether they will apply transnational principles to give a more purposive interpretation; it is also possible that the decision will be influenced by equitable considerations that are not based on the contract or on other legal sources.

6 Conclusion

Party autonomy is limited in international arbitration, in spite of the widespread opinion that contracts are self-sufficient and that, together with arbitration, they create a closed circuit that manages to leave national law out.

To be sure, the legal framework for arbitration ensures that arbitration enjoys a significant autonomy, but this autonomy is not unlimited. If the losing party decides not to comply with the arbitral award, courts of law may exercise judicial control. Judicial control on arbitration is restricted, but there is room for overriding party autonomy in several respects.

Furthermore, even within the area where no judicial control may be exercised and arbitration is autonomous, the necessity may arise to integrate contract terms with external sources. Contract terms do not always have an absolute meaning with legal effects flowing directly from the words, and recourse to a legal framework may be required to interpret the terms and to define their legal effects.

To the extent that transnational sources provide a uniform legal framework, they may integrate the contract and reinstate a certain self-sufficiency. Where transnational sources are not sufficient, however, the arbitral tribunal will have to integrate the contract with external principles and rules, primarily stemming from the governing law.

All the above constitute limitations to party autonomy in arbitration.

  • 1
    A violation of EU competition law is, according to a controversial decision of the European Court of Justice (ECJ), deemed as a violation of ordre public and therefore prevents enforcement of the award under the New York Convention: see Case C-126/97 Eco Swiss China Time Ltd v Benetton International NV [1999] ECR I-3055.
  • 2
    That each of the parties’ own law governs their legal capacity, quite irrespective of which law the parties chose to govern the contract, is regulated by the New York Convention and the UNCITRAL Model Law and was also confirmed by a Swedish Court of Appeal (Svea Hovrätt) in State of Ukraine v Norsk Hydro ASA, T 3108-06, 17 December 2007; see Institute for Transnational Arbitration (ITA) Monthly Report, May 2008, vol. VI, no. 5.
  • 3
    For a more extensive analysis see Luca Radicati di Brozolo, ‘International arbitration and domestic law’ in Giuditta Cordero-Moss (ed), International Commercial Arbitration: Different Forms and their Features (Cambridge University Press 2013) ch 2. See also Giuditta Cordero-Moss, ‘International arbitration is not only international’ in Giuditta Cordero-Moss (ed), International Commercial Arbitration: Different Forms and their Features (Cambridge University Press 2013) ch 1.
  • 4
    State of Ukraine v Norsk Hydro ASA (n 2). For a more extensive analysis, see Giuditta Cordero-Moss, ‘Legal capacity, arbitration and private international law’ in Katharina Boele-Woelki, Daniel Girsberger, Talia Einhorn, Symeon Symeonides (eds), Convergence and Divergence in Private International Law – Liber Amicorum Kurt Siehr (Eleven International Publishing 2010) 619–33.
  • 5
    Eco Swiss China Time Ltd v Benetton International NV (n 1).
  • 6
    Arbitral award on a shareholder agreement between, among others, OAO Telecominvest, Sonera Holding BV, Telia International AB, Avenue Ltd, Santel Ltd, Janao Properties Ltd and IPOC International Growth Fund Ltd, Federal Commercial Court of West Siberia, Decision of 31 December 2006.
  • 7
    For a more extensive analysis of the matter, see Giuditta Cordero-Moss, International Commercial Contracts (Cambridge University Press 2014) ch 5. See also Giuditta Cordero Moss, ‘International arbitration and the quest for the applicable law’ (2008) 8(3) Global Jurist 1. A research project at the University of Oslo has analysed the limits that this may impose on party autonomy: <http://www.jus.uio.no/ifp/english/research/projects/choice-of-law/index.html>.
  • 8
    Giuditta Cordero-Moss (ed), Boilerplate Clauses, International Commercial Contracts and the Applicable Law (Cambridge University Press 2011). This book is based on a research project that the author ran at the University of Oslo from 2004 to 2009, and shows that the same contract wording may lead to diametrically different legal effects, depending on the governing law. See particularly part 3 of the book, as well as its conclusion and ch 3.
  • 9
    Edwin Peel, Treitel’s Law of Contract (13th edn, Sweet & Maxwell 2011) para 17064.
  • 10
    For a more extensive discussion and references, see Giuditta Cordero-Moss (n 8) ch 3.
  • 11
    For France, see Xavier Lagarde, David Méheut and Jean-Michel Reversac, ‘The Romanistic tradition: application of boilerplate clauses under French law’ in Giuditta Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 9 (section 2); for Italy, see Article 1347 of the Civil Code and Giogio De Nova, ‘The Romanistic tradition: application of boilerplate clauses under Italian law’ in Giuditta Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 10 (section 1), as well as the general considerations on Article 1135 of the Civil Code in De Nova (ibid) section 1; for Denmark, see Peter Møgelvang-Hansen, ‘The Nordic tradition: application of boilerplate clauses under Danish law’ in Giuditta Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 11 (section 1).
  • 12
    See the general principle on good faith in the performance of contracts in §242 of the German BGB. See Gerhard Dannemann, ‘Common law-based contracts under German law’ in Giuditta Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 4 (sections 3.2 and 3.3) for examples of its application by the courts.
  • 13
    For Russia, see Ivan S. Zykin, ‘The East European tradition: application of boilerplate clauses under Russian law’ in Giuditta Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 16 (section 1).
  • 14
    See, for France and Italy, footnote 11 above. For Finnish law, see Gustaf Möller, ‘The Nordic tradition: application of boilerplate clauses under Finnish law’ in Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 12 (section 2.1).
  • 15
    For Germany, see §309 No. 12 of the BGB, prohibiting clauses which shift the burden of proof to the disadvantage of the other party; see too Ulrich Magnus, ‘The Germanic tradition: application of boilerplate clauses under German law’ in Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 8 (section Italy, on the contrary, does not allow oral evidence that contradicts a written agreement, see De Nova (n 11) ch 10 (section 1).
  • 16
    In addition to Germany (see previous footnote), for France see Lagarde, Méheut and Reversac (n 11) ch 9 (section 2); for Italy see De Nova (n 11) ch 10 (section 4); for Denmark, Møgelvang-Hansen (n 11) ch 11 (section 2.1); for Norway, Viggo Hagstrøm, ‘The Nordic tradition: application of boilerplate clauses under Norwegian law’ in Cordero-Moss(ed), Boilerplate clauses (n 8) ch 13 (section 3.1); for Russia, Zykin (n 13) ch 16 (section 2.1). The situation seems to be more uncertain in Sweden, see Lars Gorton, ‘The Nordic tradition: application of boilerplate clauses under Swedish law’ in Cordero-Moss (ed), Boilerplate clauses (n 8) ch 14 (section; and Finland is more restrictive, see Möller (n 14) section 2.1.
  • 17
    See Edwin Peel, ‘The Common law tradition: application of boilerplate clauses under English law’ in Giuditta Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 7 (section 2.1).
  • 18
    Literature on the subject matter is very vast. Among the works most frequently referred to are Filip De Ly, International Business Law and Lex Mercatoria (T.M.C. Asser Institute 1992); Klaus Peter Berger, The Creeping Codification of the Lex Mercatoria (Kluwer Law International 1999; 2nd ed 2010), and Ole Lando, ‘The Lex Mercatoria in International Commercial Arbitration’ (1985) 34 International and Comparative Law Quarterly 747-768. For extensive references, see Roy Goode, Herbert Kronke, Ewan McKendrick, Transnational Commercial Law: Texts, Cases and Materials (Oxford University Press 2007) 24ff.
  • 19
    For a more extensive discussion, see Giuditta Cordero-Moss, ‘Does the use of common law contract models give rise to a tacit choice of law or to a harmonised, transnational interpretation?’ in Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 3 (section 2.4).
  • 20
    For a more extensive discussion see ibid.
  • 21
    For a more extensive discussion, see ibid.
  • 22
    For a thorough analysis of the enormous impact of the CISG on scholarship, see Franco Ferrari (ed), The CISG and its Impact on National Legal Systems (Sellier 2008) 436ff. Ferrari also shows, however, that the level of awareness about the CISG in the business community and among practicing lawyers is strikingly low, see ibid 421ff.
  • 23
    United Nations Secretariat’s Commentary to the UNCITRAL Draft Convention (14 March 1979) A/CONF./97/5, also available at the Pace Law School’s database, by clicking on the specific convention’s Articles: <http://www.cisg.law.pace.edu/cisg/text/cisg-toc.html>.
  • 24
    See Dionysios Flambouras, ‘The Doctrines of Impossibility of Performance and clausula rebus sic stantibus in the 1980 Vienna Convention on Contracts for the International Sale of Goods and the Principles of European Contract Law: A Comparative Analysis’ (2001) 13 Pace International Law Review 261 and references to literature and case law cited therein. See also Ingeborg Schwenzer (ed), Commentary on the UN Convention of the International Sale of Goods (CISG) (Oxford University Press 2010) Article 79, paras 11, 18 and 37, although para 27 seems to embrace the Germanic tradition.
  • 25
    United Nations Secretariat’s Commentary (n 23) 172
  • 26
    Schwenzer (n 24) Article 79 para 11, footnote 28.
  • 27
    Sale of Goods Act of 13 May 1988 (no. 27).
  • 28
    Ot.prp. nr. 80 (1986–87) 38ff. For extensive analysis of the preparatory works in this context, see Viggo Hagstrøm, Obligasjonsrett (2nd ed, Universitetsforlaget 2011) section 19.4.2.
  • 29
    Hagstrøm, ibid. For a more extensive analysis, see Giuditta Cordero-Moss, Lectures on Comparative Law of Contracts (University of Oslo, Institutt for privatretts stensilserie bd. 166, 2004) 151ff.
  • 30
    Hagstrøm (n 16) section 5.3. Hagstrøm’s interpretation is based on a Supreme Court decision rendered in 1970, long before the implementation of the CISG in the Norwegian system. However, the Supreme Court’s decision is still correctly referred to as incorporating Norwegian law after the enactment of the Sales of Goods Act, as the reference made by Hagstrøm confirms. See also Anders Mikelsen, Hindringsfritak (Gyldendal 2011) 33. A Supreme Court decision (Rt 2004, 675) affirmed that liability is strict when the goods delivered are generic. The test will then be whether the defects objectively are within the sphere of control of the seller. In this context, therefore, the Supreme Court has rejected the test of actual control and is more in line with the regulation contained in the CISG. This approach is consistent with the German tradition, that distinguishes between generic obligations (where liability is strict) and specific obligations (where the criterion of diligence applies). This distinction was abandoned with the 2002 reform of the BGB.
  • 31
    ‘Merger clause’ is another definition of the Entire Agreement clause, which may also be called ‘Integration clause’.
  • 32
    For a more extensive discussion see Cordero-Moss, Boilerplate Clauses (n 8) ch 3.
  • 33
    Ibid section 2.4.
  • 34
    The borrower’s obligation to submit its financial statements is commonplace in loan agreements and is generally found in the section of the so-called covenants. It is meant to make it possible for the lender to control the borrower’s continued creditworthiness.
  • 35
    See generally eg David Echenberg, ‘Negotiating international contracts: does the process invite a review of standard contracts from the point of view of national legal requirements?’ in Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 1.
  • 36
    See more extensively ibid.
  • 37
    See more extensively, Maria Celeste Vettese, ‘Multinational companies and national contracts’ in Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 2.
  • 38
    See the survey by the School of International Arbitration, Queen Mary, University of London, International Arbitration Survey: Choices in International Arbitration 13; available via: <http://www.arbitration.qmul.ac.uk/>. See also the survey published by the Law Society of England and Wales, Firms’ Cross-Border Work, December 2010, <http://ec.europa.eu/justice/news/consulting_public/0052/contributions/224_en.pdf> 8. For further references, see Celeste Vettese (n 37) section 1.
  • 39
    The programme for the seminar, the list of panel participants and the transcript from the panel discussions are available at <http://www.jus.uio.no/ifp/english/research/projects/choice-of-law/events/2011/2011-arbitration-and-the-not-unlimited-party-autonomy.html>.
  • 40
    See Cordero-Moss (ed), Boilerplate Clauses (n 8) ch 3 (section 7). See also Giuditta Cordero-Moss, ‘Interpretation of Contracts in International Commercial Arbitration: Diversity on More than One Level’ (2014) 22 European Review of Private Law 13–35.
  • 41
    This appears in the 2012 Survey of the School of International Arbitration of Queen Mary, University of London (n 38) section 7.
  • 42
    Ibid 38.
Copyright © 2017 Author(s)

CC BY 4.0