Transnational Law: A Legal System or a Method of Decision Making?
by EMMANUEL GAILLARD*
AFTER SOME 35 years of legal debate and countless applications of transnational rules by international arbitrators since far before the debate over the concept even began,1 it may seem surprising that general principles of law - also frequently referred to as transnational rules or lex mercatoria2 - remain such a divisive issue in the world of international arbitration. Publications on the issue are indeed just as passionate as they were when the phenomenon was first identified and labelled as lex mercatoria in the 1960s,3 or when it became more broadly acknowledged in the 1980s.4 A recent and challenging example of this ongoing interest is found in Klaus Peter Berger's contribution to the study of 'The Creeping Codification of the Lex Mercatoria'.5
Professor of Law, University of Paris XII; Partner of Shearman & Sterling, Paris; Head of the International
Arbitration Group of Shearman & Sterling. This article is based on a presentation made by the author at the
CENTRAL Conference on Transnational Law held in Muenster on 5 May 2000.
For an early example of an award rendered on the basis of general principles of law, see the 2 September
1930 ad hoc award Lena Goldßelds Ltd. v.  USSR, discussed in V. V. Veeder, 'The Lena Goldfields
Arbitration: The Historical Roots of Three Ideas', (1998) 47 Inťl & Comp. LQ 747.
On the issue generally and on the nuances of the terminology, see E. Gaillard and J. Savage (eds), Fouchard
Gaillard Goldman on International Commercial Arbitration (Kluwer Law International, The Hague, 1999),
para. 1443 et seq.
See  especially,   Berthold   Goldman,   'Frontieres   du  droit et  "lex  mercatoria"',   (1964)   Archives  de
philosophic du droit, Vol. IX, Le droit subjectif en question 177; Berthold Goldman, 'La lex mercatoria
dans les contrats et ľ arbitrage internationaux: realite et perspectives', (1979) 106/. D. Int. 475, and, for a
critical  view,   Paul   Lagarde,   'Approche   critique   de   la  lex   mercatoria',   in   Le  Droit  des  Relations
Economiques Internationales - Etudes Offertes ä Berthold Goldman (1982) 125.
5ee especially, for a critical view, Michael J. Musti.il, 'The New Lex Mercatoria: The First Twenty-five
Years', in (1988) 4 Arbitration International 86, and, for a reply, Berthold Goldman, 'Nouvelles reflexions
sur la Lex Mercatoria', in Etudes de Droit International en FHonneur de Pierre Lalive (1993) 241;
Emmanuel Gaillard, 'Thirty Years of Lex Mercatoria: Towards the Selective Application of Transnational
Rules', in (1995) 10 ICSID Review - EIL] 208.
Klaus  Peter Berger,   The  Creeping Codification  of the Lex  Mercatoria  (1999).   5ee also the  major
contributions to the debate by Filip J. M. de Ly, International Business Law and Lex Mercatoria (1992);
Dominique Bureau, Les sources informelles du droit dans les relations privées internationales (Thesis,
University   of Paris   II,   1992);   Felix   Dasser,   Internationale   Schiedsgerichte   und  Lex Mercatoria   -
Rechtsvergleichender Beitrag zur Diskussion über ein nichtstaatliches Handelsrecht (1989).
ARBITRATION INTERNATIONAL, Vol. 17, No. 1 © LCI A, 2001
59
60                      Arbitration International Volume 17 Number 1
It would be a mistake, however, to consider that the debate has gone around in circles, always dwelling on the same issues. On the contrary, it has been strongly renewed. Initially, the controversy focused on the very existence of rules other than those found in a given legal system, with the potential to be selected by parties and arbitrators. This solution was conceived as an alternative to the traditional choice-of-law approach which purports to identify, in international situations, the most closely related body of domestic rules to be applied to the case at hand. Certain scholars readily recognized and promoted the transnational rules alternative. Others, however, denied its existence; then, when confronted with the reality of its existence, challenged its advisability as an option available to the parties; and, when confronted with the wide acceptance of that option in practice, its availability as a choice open to arbitrators in the absence of any choice of law expressed by the parties. Today, this aspect of the debate has shrunk in scope to that last situation, with some arbitration laws accepting the arbitrator's option to select transnational rules when the parties remained silent on the applicable law,6 and others rejecting that possibility.7 Among practitioners, this initial controversy was for a time so inflamed that the positions taken seemed to be driven more by act of faith than by rational argument. These positions were all the more futile when, at the same time, the players of international commerce were already making full use of their options by selecting, where appropriate, transnational rules to govern their contracts.
Today, the debate has refocused on issues of sources and methodology. Indeed, transnational rules or lex mercatoria in whatever form are now sufficiently established for the heart of the controversy to have shifted, concentrating more recently on the establishment in further detail of the content of those rules or the more systematic assessment of the means to do so. As a result, very significant differences of opinion on how such goals may be achieved have emerged. A cynic might say that the proponents of lex mercatoria are now so numerous, and their cause so well accepted, that they can afford to fight among themselves. We will thus examine the terms of the renewed debate surrounding lex mercatoria (see (I) below) before reassessing, in light of this debate, the traditional inquiry of how lex mercatoria compares with a genuine legal order such as the law of a given country (see (II) below).
I. THE RENEWED DEBATE ON LEX MERCATORIA
In essence, two main issues cause supporters of lex mercatoria or transnational rules to differ fundamentally, and to have done so for some time, even if these
Most notably French law since 1981 (Article 1496 of the New Code of Civil Procedure); Dutch law since 1986 (Article 1054 of the Code of Civil Procedure); Swiss law since 1987 (Article 187 of the Private International Law Statute (LDIP)). 7     Most notably, the UNCITRAL Model Law (Article 28), English Arbitration Act 1996 (section 46), the German 1985 Arbitration Act (Article 1051(2) ZPO), all of which remained fairly conservative in this respect.
Transnational Law
61
differences were kept in the background in the past when the primary debate focused on the very existence and legitimacy of resorting to rules other than those of a given legal system. These issues are whether lex mercatoria is defined by its content or by its sources (see (a) below) and whether it should be restricted to a list or understood as a method (see (b) below).
(a) Is Lex Mercatoria Defined by its Content or by its Sources?
The first area of controversy among the supporters of lex mercatoria has to do with the extent to which transnational rules are characterized by their purported specificity, from a substantive standpoint, vis-a-vis rules found in national legal orders. For one school of thought, such specificity is the very raison d'etre of recourse to transnational rules, these rules having been conceived and developed in response to the perceived inadequacies of national legal orders.8 From this viewpoint, international transactions require added flexibility, which the requirements found in national laws would seldom accommodate. This school of thought is related to the theory of the 'specific needs of international business', which has subsequently been derided as a new form of the laissez-faire doctrine.9
Another view, which we believe to be the better one,10 finds the specificity of transnational rules to lie in the fact that these rules are derived from various legal systems as opposed to a single one, and more generally from various sources,11 rather than in their allegedly differing content. In other words, their specificity is one of source, not of content. Indeed, there is no reason to believe that national legal orders are unable to accommodate adequately the specific needs of international situations, for instance by creating a separate set of substantive rules to govern international situations. Numerous examples can be found of this approach to accommodating the 'specific needs of international business', in monetary relationships for instance,12 or in the field of arbitration whenever, as in France or Switzerland, international arbitrations are governed by a different set of rules from domestic ones.13 In this connection, it is important not to confuse a national legal order with its domestic, as opposed to international, substantive rules.
Admittedly, because they are chiefly derived from various national legal systems, transnational rules stand a better chance not to reflect the outdated rules which may still be found in certain legal systems. In that sense, they may help to meet the concerns of modern business, but this is not to say that, by nature, national laws cannot achieve the same result. On the contrary, it is because a
5ee e.g. Loquin, 'La réalité des usages du commerce international', in (1989) RID Eco. 163. 9     Mustill, supra n. 4, at p. 181. 5ee e.g. Gaillard, supra n. 4. 5ee infra, I.(b), second paragraph.
On the admissibility in international law of certain indexation clauses which are prohibited in domestic law, see e.g., in France, Cass, civ., 21 June 1950, (1950) 39 Rev. crit. de droit international přivé 609, Note Batiffol.
Article 1492 et seq. of the French New Code of Civil Procedure; Article 176 et seq. of the Swiss Statute on Private International Law.
62
Arbitration International Volume 17 Number 1
sufficient number of legal systems have adopted modern rules that transnational rules will be able to follow them in embracing the most appropriate solution.
These two ideas of the specificity of content and the formation from comparative law sources were present, in the most intricate way, in the early writings on lex mercatoria1'1 but they now deserve to be segregated if one does not want to exacerbate an artificial distinction between national legal orders - which are not confined to domestic rules and do not necessarily contain outdated rules -and transnational rules. This is why, in our opinion, lex mercatoria should be defined today by its sources, the details of which will be examined below,15 as opposed to its content.
(b) Is Lex Mercatoria a List or a Method?
The second issue on which opinions are divided within the pro-lex mercatoria camp concerns the means of assessing the contents of transnational rules. Are the contents to be found in a list or 'creeping codification', be it static or open-ended, or are they to be derived, on a case-by-case basis, using a specific methodology which may in turn make use of existing lists or restatements but which is by no means limited to these tools? Some proponents of lex mercatoria16 have expressed the view that lex mercatoria is to be found in lists, despite the fact that at the outset the presentation of lex mercatoria in terms of lists was in fact intended to be a criticism designed to show how scarce, contradictory and unpredictable transnational rules were.17 The elaboration of far more substantial lists, such as that of UNIDROIT, have reversed this perspective, such that today these lists are often presented as the principal if not the only, component of lex mercatoria.18 Such lists do present the advantage of being simple to use and of responding to the criticism of the alleged vagueness of transnational rules. In contrast with the initial perception, they provide the necessary predictability of the outcome which is valued by the parties in international commerce.
The other approach to defining the contents of transnational law is to view transnational law as a method of decision-making, rather than as a list. This approach consists, in any given case, of deriving the substantive solution to the legal issue at hand not from a particular law selected by a traditional choice-of-law process, but from a comparative law analysis which will enable the arbitrators to apply the rule which is the most widely accepted, as opposed to a rule which may
5ee in particular B. Goldman, supra n. 3.
5ee infra., I.(b), third paragraph.
5ee especially Berger, supra n. 5, at p. 218.
5ee in particular the list of some 25 principles discussed by Lord Mustill in his article on lex mercatoria,
supra n. 4, in an attempt to show how poor and inadequate this option was.
5ee e.g., Fabrizio Marrella and Fabien Gélinas, 'The UNIDROIT Principles of International Commercial
Contracts in ICC Arbitration - Introduction and Preliminary Assessment and Extracts from ICC Awards
referring to the UNIDROIT Principles', m (1999) ICC Bulletin Vol. 10, No. 2, at pp. 26-32, and extracts
from awards at pp. 33-109; Michael Joachim Bonell, The UNIDROIT Principles of International Law in
Transnational Law in Commercial Legal Practice (1999) at p. 7 et seq.
Transnational Law
63
be peculiar to a legal system or less widely recognized. This comparative law analysis is greatly assisted today not only by the extremely comprehensive compilations of principles previously discussed, but also by the existence of a number of international treaties which, whether in force or not, reflect a broad consensus, by the increasingly large number of published awards providing as large a number of precedents to international arbitrators and by the availability of extensive comparative law resources19 such as monographs on a large number of specific issues.
The transnational law method should thus, in our opinion, be conducted in the following three steps. First, the utmost attention should be given to the parties' intentions. They may have suggested a methodology themselves, for instance in limiting the comparative law analysis to two legal systems20 or to those of a region.21 They may have used clumsy terminology which arbitrators need to interpret in order to give effect to the parties' true intent.22 In all of these instances, the first task of the arbitrators will be to implement the parties' instructions. Second, the arbitrators will determine, on the basis of the comparative law sources mentioned above, whether the contentions made by the parties are supported by a widely accepted rule, or whether they merely reflect the idiosyncrasies of one legal system, in which case they should be rejected. This will be the case, for instance, of the French rule pursuant to which a subcontract will be void if certain conditions including the placing of a bond in favour of the subcontractor are not met,23 of the English rule denying the validity of agreements to agree, or of the Algerian rule prohibiting agents, all of which are fairly peculiar to the legal system in which they are found. Third, in determining whether the acceptance of a given rule is sufficiently wide for that rule to qualify as a general principle of law, the unanimous acceptance in all legal systems is by no means required. The unanimity requirement sometimes advocated by authors generally unfavourable to general principles of law24 would render that methodology meaningless. Indeed, if this was a requirement, general principles of law would be either useless, when they reflect a rule accepted in every law, or non-existent, failing such unanimous acclaim. The real function of the general principles method is, on the contrary, to enable
For an excellent example of thorough comparative law research, see Jerome Ortscheid, La reparation du dommage dans ľarhitrage commercial international (Thesis, University of Paris XII, 1999). 5ee e.g. clause 68 of the Eurotunnel construction contract of 13 August 1987, which refers to the principles common to French and English law. On the 'tronc commurí approach, see Mauro Rubino-Sammartano, 'The Channel Tunnel and the Tronc Commun Doctrine', in (1993) 10/. Inťl. Arb. 59. For examples of regional general principles, see Gaillard, supra n. 4, at p. 230 et seq. For example, the parties may have referred to transnational rules under the label of 'trade usages', which is not technically correct but which may have to be understood as a reference to general principles of law where there is reason to believe that the parties meant to have such principles apply. 23   Article 14 of the law of 31 December 1975, disregarded in ICC Award No. 7528 (1994), (1997) XXII Y B Comm. Arb. 125.
5ee e.g. P. Mayer, 'L'autonomie de ľarbitre international dans ľappréciation de sa propre competence', in (1989) 217 Recueil des cours de ľAcadémie de droit international 319.
64
Arbitration International Volume 17 Number 1
arbitrators to discriminate between rules which enjoy wide recognition, and those which are particular to one or to a limited number of legal systems.
This understanding of transnational law presents a distinct advantage over the view which reduces it to a list, for it eliminates the criticism based on the alleged paucity of the list.25 Any allegation made by a party in a given case will necessarily find an answer in the form of a generally accepted rule, even if no such rule is contained in any precedent award, international treaty or pre-established list. It will then be for the arbitrators to assess, if such situation arises, whether or not the contention made by the requesting party finds general support in comparative law. For example, one author applied this methodology to the very complex and detail-oriented field of damages (including quantum, interest rates, starting point of interest calculations, punitive damages, etc.), and through this research was able to describe the trends which could be applied to such specific issues by arbitrators having to apply transnational rules.26
The fact that, in recent years, several lists of general principles have come to light in the form of more complete and detailed restatements, has not modified this conclusion. Indeed, however extensive they may be, these lists will never render the role of arbitrators dealing with general principles of law a mechanical task. If anything, as these lists become more numerous, one cannot exclude situations in which the various lists will conflict. Just as they have dealt with conflicts of laws, arbitrators having to apply transnational rules may now have to deal with conflicts of lists. An example is found in the 'hardship' rule which is defined in a similar way in the UNIDROIT Principles, the Lando Principles on European Contract Law and the CENTRAL list of Principles, except on one particular issue. As opposed to the two other lists, UNIDROIT Principle 6.2.2(a) accepts that an event which existed before the contract, but which became 'known to the disadvantaged party after the conclusion of the contract', may qualify as a hardship event. Other legal systems deal with such a situation only as a case of error, which may lead to very different results. Another example of conflicting lists consists of the definition of the situation in which a party having broken off contractual negotiations may be held liable. Whereas the Lando and UNIDROIT lists of Principles require that the party having broken off the negotiations be shown to have continued negotiations when intending not to reach an agreement with the other party,27 the CENTRAL Principles shift the criterion to the aggrieved party and only require that such party be found to be 'justified in assuming that a contract would be concluded'. This latter requirement, which is closer to the théorie de ľapparence philosophy, is obviously much less demanding than the UNIDROIT and Lando ones. In such situations of conflicting lists, the task of the arbitrators will still be very clear should they follow the comparative law methodology. This task will be (with the assistance of the various lists available,
5ee Mustill, supra n. 4.
5ee Ortscheidt, supra n. 19.
UNIDROIT Principle 2.15(3); Lando Principle 2:301(3).
Transnational Law
65
arbitral precedents and, in some instances, international instruments) to assess whether or not the rule invoked by a party reflects a norm which genuinely corresponds to a trend enjoying broad international recognition. The UNIDROIT Principle is likely to be disregarded in the first example and to prevail in the second one.
This is not, however, to downplay the merits of the existing lists of principles. These lists will in many instances, in the absence of any conflict or ambiguity, enormously facilitate the task of arbitrators having to rule on the basis of transnational rules.28
After having clarified what we believe to be the correct methodology to be followed by arbitrators who were mandated by the parties to apply or, in the absence of any choice of law made by the parties, who have chosen to apply transnational rules, we may now revisit the longstanding query of how the transnational rules methodology compares with the application of a fully fledged legal order.
II. THE ISSUE OF LEX MERCATORIA AS A DISTINCT LEGAL SYSTEM REVISITED
Four characteristics are generally found to be the mark of a genuine legal system (ordre juridique): its completeness, its structured character, its ability to evolve and its predictability. We will examine in turn how the general principles methodology scores with respect to each of these criteria.
This analysis has lost any practical importance in the numerous cases in which arbitrators, even in the absence of an express submission of the matter in dispute to general principles by the parties, have been granted the option to select the 'rules of law', as opposed to 'the law', they deem appropriate where the parties have remained silent on the applicable law. This will be the case where the arbitrators are acting pursuant to French, Swiss or Dutch arbitration laws, for example, as opposed to English or German laws or under the UNCITRAL Model Law, which, as we have seen, has adopted a rather conservative approach in this respect.29 This will also be the case, even for arbitrators sitting in England, Germany or in a country having adopted the Model Law, when the arbitration rules chosen by the parties have enlarged the scope of the arbitrators' options by granting them the freedom to apply 'rules of law', as do the 1998 ICC Rules, the 1998 LCIA Rules or the 1997 International Arbitration Rules of the AAA. This language ('rules of law^, which was first used in the French law on arbitration in 1981, was in fact specifically intended to bypass the issue of whether lex mercatoria or general principles qualify as a genuine legal order.
For examples of the now very numerous awards referring to the UNIDROIT Principles, see e.g. Marrella and Géľmas, supra n. 18. 5ee supra, n. 6 and 7.
66
Arbitration International Volume 17 Number 1
However, in situations in which arbitrators are still required to apply, in the absence of a choice expressed by the parties, a 'law' and not mere 'rules of law', it may still be of interest to assess whether the transnational rule methodology could nonetheless qualify as a legal system and hence be applied as the 'law' selected by the arbitrators.
(a)  Completeness
Although there has been some debate on this issue,30 it is generally accepted that a genuine legal order is complete, i.e. able to provide an answer to any legal issue which may arise between the parties, even if in order to do so one has to resort to general principles of that legal system (good faith, legality of what is not expressly prohibited, etc.).
Understood as a list, be it a sketchy one or a lengthy restatement, general principles of law cannot meet this requirement. However, with the methodology approach to general principles of law, a solution can always be found to any given legal issue raised by a party. Indeed, the arbitrators will always be able to decide, on the basis of the comparative law elements presented by the parties, whether the allegation made by one of the parties is widely supported or not. No possible gap in any given list can prevent them from doing so. In a case where a party argues, for instance, that a framework agreement is devoid of any binding effect as an 'agreement to agree', although this issue may not be covered by the various lists available, using the comparative law methodology, the arbitrators will be able to determine that this rule is specific to a limited number of legal systems and, as such, does not qualify as a general principle of law.
(b) Structured Character
A crucial difference between a legal system and mere 'rules of law' lies in the fact that the former presents a degree of structure which is absent from the latter. A legal system is an organized set of rules, with various levels of generality and close ties between rules belonging to those various levels. This structure is the key to understanding the logic and values of the system as a whole, and to interpreting any given rule in that system. In a legal system, there are certain very general principles such as the binding force of contracts, good faith or the need to protect certain important values or fundamental rights, which, taken together, form the public policy of that legal system. There are also more specialized rules in any given field (contracts, property, etc.), as well as increasingly specific rules in selected areas of interest, such as consumer contracts or intellectual property, etc. The understanding of the links between these different levels of generality is essential when interpreting any given rule. To take only one example, the positioning of the rule to be interpreted vis-a-vis more general rules will have a
5ee especially P. Mayer, La distinction entre regies et decisions en droit international přivé (Dalloz 1973); Charm Perelman, Le probléme des lacunes en droit (Bruyland, Bnrxelles 1968).
Transnational Law
67
direct impact on its scope of application. A rule which derogates from a more general principle in a specific situation will be interpreted strictly, whereas a rule which is the mere application of a more general one will be expanded to similar situations by analogy. This distinction is at the very heart of the choice between an a contrario interpretation and analogous expansion, between which the interpreter may often hesitate. Similarly, the creation of new rules applicable to certain specific situations will have to accommodate, and somehow balance, the various general values found in the legal system at hand. This is generally the task entrusted to the legislature and, in some instances, to the judiciary, so far as judge-made law is concerned.
This characteristic of being composed of rules organized in different levels of generality is now present in the body of transnational rules comprising lex mercatoria. Of course, from a terminology standpoint, one should recognize that the expression 'general principle of law' has two very distinct meanings. It can be used to denote a rule which is very general in nature (e.g., good faith, pacta sunt servanda, etc.), or to refer to a rule found generally in many legal systems throughout the world. The first meaning refers to the positioning of the rule in the legal system, at a given level of generality, the second one to the source of the rule. It is essential for the understanding of the transnational rules methodology to recognize that these two meanings do not necessarily coincide. General principles, understood as principles derived from a number of legal systems through the comparative law approach, may be very general in nature (e.g., good faith) but can also be very specific (e.g., the duty to give notice of breach promptly in order to enable the co-contractor to remedy such breach whenever feasible). This is the reason why it is not nonsensical to refer to a 'specific general principle' in the context of a particular transnational rule. The expression 'transnational rules' has the advantage of avoiding any ambiguity although this is not, in our opinion, a sufficient reason to banish the widely used term 'general principles' by reference to the transnational source of these rules. In describing actual arbitral case law, several commentators have indeed noticed the increasing specialization of 'general principles'.31
Not only can general principles comprising lex mercatoria be found to have various levels of generality but, perhaps more importantly, they are interrelated just as they would be in a genuine legal order. An obvious example can be found in the various applications of the general principle - in both meanings of the term - of good faith. Good faith is a very general precept. It is the foundation of more specific rules such as good faith in the interpretation of contracts, good faith in the performance of contracts, etc. Good faith in the interpretation of contracts is, in turn, the foundation of even more specific rules, such as those according to which the true meaning of a provision overrides the letter, or which prevent the party having drafted a document from relying on its ambiguities in order to obtain a
31   See E. Gaillard, (1987) 114 /. D. Int. 137, at p. 141; Loquin, soprán. 8, at p. 190.
68
Arbitration International Volume 17 Number 1
particular benefit (the contra proferentem interpretation rule). This latter example is that of a general principle with respect to its source, albeit a very specific one as far as its scope is concerned.
Another example of interrelated general principles of law is that of pacta sunt servanda and, to the extent it is accepted as a general principle, rebus sic stantibus. Although general in scope, the binding force of contracts is by no means without exceptions. Just as an excessive penalty clause will be reduced or disregarded despite the pacta sunt servanda principle, a party cannot necessarily rely on a contract which has been dramatically rendered inoperable by unforeseen circumstances.32 Here the relationship is one of principle-exception, not one of principle-application. Thus, the rebus sic stantibus rule, where applicable, will be construed narrowly, whereas the rule according to which the intent should override the letter or the contra proferentem rule can be expanded by analogy. All of the above rules operate within the general principles arena exactly as they would in the context of a given national law and therefore of a genuine legal order.
In light of the above, it is not sustainable to argue that general principles are composed of vague and contradictory rules such as pacta sunt servanda and rebus sic stantibus.33 As we have seen, these two principles are not contradictory; they are interrelated in a principle-exception logic. Just like legal systems which recognize the imprévision rule frame it as an exception to the binding force of contracts,34 general principles can accept both, with one tempering the other. This is the reason why both principles are found in well-drafted codifications of general principles of law, such as the UNIDROIT Principles, the Lando Principles or the CENTRAL list of Principles.35 One should also recognize that, if understood as a comparative law method, general principles do reflect the various levels of generality which characterize a genuine legal order. This is not surprising to the extent that general principles of law draw their existence from a variety of genuine legal orders, which unquestionably present this characteristic. When implementing this approach, however, practitioners should be aware of the different levels of generality of the laws or of the rules found in various lists from which they may seek to derive the existence of new general principles of law, and not find contradictions where only derogations are intended.
(c) Evolving Character
A third characteristic of a genuine legal order is that it will necessarily evolve over time in order to take into account the needs of the society which it is designed to
On the issue of whether the rebus sic stantibus doctrine is a general principle of law, see Hans van Houtte,
'Changed Circumstances and Pacta Sunt Servanda', in Transnational Rules in International Commercial
Arbitration (E. Gaillard, ed.) (ICC Publication No. 480/4), at p. 105.
5ee e.g. Antoine Kassis, Theorie generale des usages du commerce (1984) at para. 349 et seq.
See e.g., in Egyptian law, which became the model of many laws in the Arab world, the provisions of Article
147 para. 1 (binding force of contracts) and para. 2 (imprévision).
All of these lists are reprinted in Berger, supra n. 5.
Transnational Law
69
regulate. Be it through its case law or through legislative action, it will contain certain rules governing the process by which such updating is to occur, as well as rules indicating to which situations any new rules will apply. The issue of whether transnational rules possess this evolutionary character depends in part on the school of thought to which one belongs. Where transnational rules are understood as a list of principles, they may or may not be evolutionary. This will depend on whether or not the list in question is updated regularly, organizations under the auspices of which such lists have been issued being generally aware of the need to do so. As a matter of fact, CENTRAL emphasizes the permanent updating of its list36 and UNIDROIT has already undertaken the updating of the Principles published in 1994.37 Nonetheless, a list will always remain the reflection of a comparative law effort taken at a given time.
In contrast, where transnational rules are understood as a methodology drawing from a number of sources pursuant to the comparative law approach, they will by nature be extremely responsive to the changing needs of international commerce. At any given time, this methodology will enable arbitrators to take into account the most current status of the various laws from which the principles are to be drawn. As a result, arbitrators having chosen to apply transnational choice-of-law rules as opposed to those of a given legal system38 may find that, in light of the recent at least partial conversion of certain common law countries,39 statute of limitations issues ought to be characterized as substantive rather than procedural. Indeed, where many legal systems have evolved on a given issue, the comparative law approach will lead to the application of the most modern trend, whereas the traditional choice-of-law approach may lead to the application of an obsolete rule, depending on the odds of the distribution of the various connecting factors. As such, the comparative law understanding of transnational rules does encompass a degree of bias in favour of the most current rules. Once again, this is not to say that lists are not to be used in applying this methodology but, precisely when the lists are ageing, arbitrators ruling pursuant to transnational rules are perfectly free, and indeed encouraged, to search for the most current status of the issue. The same philosophy is applied when arbitrators take into account international treaties recently negotiated among a number of countries as a source of general principles, even where such treaties have not yet entered in force.40
As a result, understood, as they should be in our opinion, as a method, transnational rules are just as evolutionary as any given legal system. In essence,
5ee Berger, supra n. 5.
Michael Joachim BonelL An International Restatement of Contract Law (2nd edition, 1997), at p. 171 et
seq.
On the application of the transnational rule approach to choice-of-law rules themselves see Gaillard, supra
n. 4.
5ee e.g. in English law, the Foreign Limitation Periods Act 1984.
5ee e.g. Jean-Paul Béraudo, 'The United Nations Convention on Contracts for the International Sales of
Goods and Arbitration', in (1994) ó ICC Bulletin 1, at p. 60; Andrea Giardina, 'International Conventions
on Conflicts of Law and Substantive Law', in ICCA Congress Series No. 7, Planning Efficient Arbitration
Proceedings; The Law Applicable in International Arbitration (ed. A. J. van den Berg) (1996) 459.
70
Arbitration International Volume 17 Number 1
recourse to the transnational rules method will erase the oddities remaining in certain laws, but will also avoid surprises arising from the adoption of an atypical rule in some recent statutes, thus playing, in some instances, a moderating role, and in others a modernizing one.
(d) Predictability
Finally, it is argued that predictability is a major advantage of genuine legal systems as only they enable the parties to assess the likely outcome of any diverging views that may arise. To the contrary, lex mercatoria would thus be utterly unpredictable. This view is generally summarized in the statement according to which, when asked by 'an ordinary businessman' what the lex mercatoria answer to a given issue is, it would be almost impossible to provide a specific answer. On this point, we submit that, contrary to common wisdom, transnational rules offer as much predictability, if not more predictability, than genuine legal systems. Obviously, the list approach has provided a first answer to the criticism that transnational law is hard to locate, almost in the physical sense of the word, as opposed to a neat leather-bound series of law reports, or a convenient electronic database, and is thus vague and unpredictable. More fundamentally, the criticism is based on a remarkably abstract perception of the law, totally detached from the realities of commercial transactions and the actual needs of 'ordinary businessmen'.
In actual practice, transnational principles are primarily used in three different situations. The first one occurs when the parties agree to the application of a given law, generally that of the State or the State-owned entity which is party to the transaction, but also agree to temper the application of that law by resorting to transnational rules of some sort. When dealing with a sovereign State, there is no doubt that the 'ordinary businessman' would favour transnational rules or, at a minimum, transnational rules tempering the laws of that sovereign State, over the laws of that sovereign State alone. The second situation arises when the parties could not agree on a given legal system, generally because of the perceived advantage that each party would obtain through the application of its own law, and thus they selected transnational rules to govern their transaction. In such a situation, the predictability of the outcome is to be assessed in comparison with the parties' other option of remaining silent on the applicable law. Assuming that all potentially applicable laws would lead to the same result on a given issue, the question is likely to be moot but, should they diverge on the validity of a provision of the contract, for example, would it really be more predictable to leave the determination of its validity to the discretion of the arbitrators to select the applicable law, as opposed to accepting, through general principles, to have the contract governed by the rules which are, for any specific issue, the most generally accepted in the world? The third situation is precisely where the parties have remained silent as to the law applicable to their transaction. There, the reasoning is the same as the one which sometimes leads the parties to select transnational rules when they cannot agree on a given law. For example, in an Anglo-Qatari situation,
Transnational Law
71
in which English law would consider a given provision of the contract invalid and Qatari law valid, would it really suit the needs of the business community to have the arbitrators decide the issue on the basis of the assessment of the various links of the matter with the various potentially applicable laws rather than to enable the arbitrators to decide the issue, if they see fit, on the basis of the principles most commonly accepted around the world? A poll of 'ordinary businessmen' asked the question in those terms would, no doubt, be most challenging. If predictability is a value, it is by no means certain that the traditional approach prevails over the transnational rules method on this account.
In sum, assuming that the above characteristics are indeed those of a genuine legal system, the list approach and, even more certainly, the comparative law approach to transnational rules performs quite well as far as completeness, structured character, evolving nature or predictability standards are concerned. As a result, if not a genuine legal order, transnational rules do perform, in actual practice, a function strikingly similar to that of a genuine legal system. Thus, one may be tempted to conclude that, where the relevant arbitration rules or arbitration statute mandates the arbitrators to select the 'law' applicable to the dispute, as opposed to mere 'rules of law', it is nonetheless open to them to select - this choice being particularly appropriate where the connecting factors are almost equally divided - transnational rules as 'the law' applicable to the dispute.