He then gave an order: 'Cut the living child in two and give half to one and half to the other.' The woman whose son was alive was filled with compassion for her son and said to the king, 'Please, my lord, give her the living baby! Don't kill him!' But the other said, 'Neither I nor you shall have him. Cut him in two!' Then the king gave his ruling: 'Give the living baby to the first woman. Do not kill him; she is his mother.' (Bible, I Kings 3, 25-27, New International Version)

I. Introduction

Few topics in international arbitration send more shock waves along the perceived fault line separating civil and common lawyers than that of discovery. The rumblings are felt most clearly in arbitrations involving 'mixed' parties or tribunals, i.e. cases in which the parties, or the members of the arbitral tribunal, hail from different legal traditions. In such cases, clashes over the use and extent of discovery techniques - and even their moral rectitude - are commonplace.

The debate between the defenders and detractors of discovery in international arbitration has found its way into the academic literature, where it has likewise tended to be framed in terms of the common law/civil law dichotomy. Discovery-minded common lawyers are alternatively depicted in the literature as harbingers of progress in international arbitral procedure, or as barbarians at the gate of the ancien continental regime.

In our view, the ongoing debate on discovery in arbitration has produced more heat than light, the primary reason being that it starts one step too late in the analysis. 'Discovery', in the general sense of legal processes used for bringing out the facts of a dispute, is present in all legal systems. Indeed this has always been the case, with the well-known story of King Solomon, quoted above, being perhaps the earliest recorded example of a sole arbitrator's wisdom in employing an effective (if risky) discovery device.

The issue, then, is not whether 'discovery' should be permitted in international arbitration, for the answer to that is clearly yes. Rather, the appropriate questions are the extent to which discovery should be allowed, and how to manage the process effectively. We submit that these questions can best be answered in an eminently practical, case-by-case manner. [Page25:]

II. The contours of the debate

It is not the purpose of this paper to survey the broad literature touching on the question of discovery in arbitration.1 For present purposes it will suffice to note that the debate is most often cast in terms of the supposed 'Great Divide'2 between the common and the civil law. Indeed, discovery is frequently treated as a symbol - or perhaps more appropriately, a symptom - of the Great Divide itself.

Common lawyers tend to follow Lord Mustill in seeing discovery as a 'powerful instrument for justice'3 that should be incorporated to a greater or lesser extent into international arbitration procedure. As once described by Andrew Rogers Q.C. at an ICCA Congress, discovery is 'an essential feature' of the arbitral process that is, at least in some cases, indispensable in order to ensure that a just result is reached. 4

On the other side of the debate (for the most part) stand commentators from the civil law tradition, who are typically reluctant to see the incorporation into international arbitral practice of discovery in general, and document discovery in particular. As one civil law commentator has put it:

We feel that the principle onus probandi incumbat allegandi excludes the possibility of obtaining the help of the court to extract evidence from the other side. We react to the notion of discovery, be it English or, worse, American style, as an invasion of privacy by the court, which is only acceptable in criminal cases, where the public interest is involved. . . .

Civil lawyers' negative reaction to discovery is probably due to their experience with discovery American style, that looks to us to be definitely inquisitorial. Is it not typical of the Inquisition (be it practised in XVth century Spain or XXth century Soviet Union) to solicit, by various devices, the help of the accused in insuring his own demise? 5

Or as perhaps more dispassionately observed by another commentator: 'The idea of producing all documents, both those which are helpful and unhelpful to one's case, is something most unpalatable to civil law practitioners.' 6

The great debate about the Great Divide plays itself out not only in the literature, but also in the beliefs and expectations of legal practitioners and their clients. For instance, while giving a lecture on arbitration to a group of civil lawyers, one of the authors was struck to find that the word 'discovery' ranked second only to 'punitive damages' in terms of its capacity to strike terror into the civil law hearts of the audience. In a later presentation to a predominantly common law audience, the reaction was just the opposite: 'How can it be that broad discovery is generally not available in international arbitration proceedings? How else will we find the "smoking guns"?'

Knee-jerk reactions along the lines of these anecdotal examples are not simply matters of interest from a comparative law perspective - they also have consequences for parties and tribunals in real cases. In one recent case in which the authors were involved, the tribunal ordered English-style discovery. Our clients dutifully complied by producing several boxes of documents to the adversary, while the other party (which was from a civil law jurisdiction) produced only a handful of documents in response to an equally broad request. Only the prompt intervention of the tribunal avoided the prospect of real unfairness in that case. In another recent case involving [Page26:] 'mixed' parties, disputes about document discovery ballooned into a veritable litigation of its own, vastly increasing the costs of the arbitration to the parties and requiring the tribunal to spend literally hundreds of hours delving into, and ruling upon, complicated questions of relevance and privilege. The difficulties in both cases were exacerbated by the near-instinctive reaction of the common law sides to press for as much discovery as would be available in their own national systems, combined with the equally instinctive belief on the part of the civil law parties that they should not have to disclose anything that they could keep to themselves if litigating at home.

The foregoing leads us to two criticisms of the discovery debate as generally formulated in the literature and in the arbitration hearing room. The first is that the axis of the debate may itself be fallacious. As Jan Paulsson has written, perhaps 'we are doing no more than camouflaging everyone's true belief that "my way is the right way" and "your way is the wrong way" by expanding the notion of "my way" into an imposing family of "us" which in fact does not exist. Perhaps there is no Great Divide at all.' 7

The second - and for present purposes more important - criticism is that the participants in the discovery debate tend to start one step too late in the analysis by assuming that discovery is a monolithic, and wholly common law, concept. It is not, as explained in the next section, and the mere recognition of that fact may go a long way towards de-mystifying (and de-politicizing) the concept.

III. Defining the problem

Black's Law Dictionary defines 'discovery' as follows:

In a general sense, the ascertainment of that which was previously unknown; the disclosure or coming to light of what was previously hidden; the acquisition of notice or knowledge of given acts or facts . . . 8

A more practice-specific formulation can be found in the Dictionary of Arbitration and its Terms, which defines 'discovery' as:

A legal procedure invoked before a trial to inform both parties of the facts in a dispute in order to narrow the issues and save time and expense. 9

These two sources define discovery in its most general - but frequently overlooked - sense. At base, discovery is no more than a legal process for bringing the facts of a case to light, in particular so that the issues in dispute can be clarified and (hopefully) limited. It is, in short, neither innately a common law nor a civil law concept, and certainly not the stuff of which Great Divides are made.

Discovery in the general sense is an inevitable feature of every legal system that attempts to decide disputes on a rational basis. The reason is simple: in every such legal system, one way or the other, the facts must be brought out. This is perhaps especially true in international arbitration cases, which typically tend to be fact-driven. 10

That is not to say, however, that the same methods for bringing out the facts are employed in the common law and civil law worlds. Let us begin with the usual suspect [Page27:] where discovery is concerned - the common law world - and in particular the Anglo-American jurisdictions.

In England and in America, discovery is a highly-developed feature of the civil trial system. 11 Frequently justified as a compensating factor for the common law judge's lack of inquisitorial powers12 - or somewhat more cynically as a mechanism to encourage the settlement of cases by forcing both parties to turn over their 'smoking guns' - discovery has many facets in the common law systems.

The vision that generally comes to mind is boxes of documents being packed up and shipped off to one's adversary, and indeed document discovery is one of the main elements of Anglo-American discovery practice. But it is not the only one. The Rules of Civil Procedure that apply in the U.S. federal courts, for example, provide for no less than five discovery mechanisms in addition to compulsory document production. These are: (i) depositions of witnesses; (ii) written interrogatories; (iii) physical inspections of objects or property; (iv) physical and mental examinations of persons; and (v) requests for admissions. 13

All of this combines to produce complicated and time-consuming pre-trial proceedings, and the U.S. system of discovery has been criticized (even by its participants) on this ground. In particular, the criticism has been levelled that pre-trial discovery can develop into a mini-litigation, as parties battle over the questions whether full compliance has been made with their discovery requests, whether privileges may be invoked to shield certain documents, whether a witness may decline to answer certain questions in an oral deposition, and so on. Disputes over these and other discovery-related issues have to be heard and decided by someone. The crowded dockets of the U.S. federal courts largely preclude judges from this task, and therefore a separate judicial authority - the Magistrate Judge - has been established to deal with such matters. Magistrates are non-life-tenured judges whose primary function is to hear and resolve discovery disputes, thereby keeping the wheels of pre-trial discovery turning and allowing cases to proceed towards an eventual trial (or settlement). 14

The specific discovery mechanisms discussed above are largely unknown in the civil law world. That does not mean, however, that 'discovery' - in the general sense of legal mechanisms for bringing forward the facts - does not form a part of civil law procedure. Quite the opposite: in nearly every civil law system, lawyers produce relevant supporting documents to the courts (and their opponents) with their written pleadings. 15 This is of course a form of discovery, in the broader sense, because it establishes and narrows the facts that are in dispute at an early stage of the litigation.

Furthermore, judges in civil law countries typically possess inquisitorial powers that are unknown in the common law world. The judge can call for witnesses to be heard at any stage of the proceeding, 16 can put questions to them directly and, in many countries, can order a party to produce relevant documents or other evidence. 17 The fact that these processes are controlled by the judge, rather than by the parties as in the common law systems, does not make them any less methods of 'discovery' - i.e. mandatory mechanisms for bringing forward the evidence and establishing the pertinent facts of the case.

One civil law mechanism - the court-appointed expert - merits particular attention. Experts are individuals or organizations appointed by the court (either at the request [Page28:] of a party or at the instance of the court) to investigate a factual or legal aspect of a case, particularly where it concerns complex or technical elements. 18 The expert is given terms of reference - sometimes themselves a source of dispute between the parties - and is then sent off to investigate the case and report on the facts. The parties are usually placed under an obligation to cooperate with the expert's endeavors, even if this means revealing facts detrimental to their claim or defense. Such cooperation may, for instance, include permitting scrutiny of pertinent documents and examinations of the machinery or other physical objects involved in the dispute, 19 although it typically stops short of requiring general access to corporate records. Failure to cooperate with a court-appointed expert may result in the drawing of negative inferences against the non-cooperating party or other sanctions. 20 Altogether, it seems, a relatively interventionist (and potentially time-consuming) process for discovering and establishing the facts.

The mechanism of the tribunal-appointed expert has found its way into arbitration practice as well, 21 being employed by, amongst others, the Iran-U.S. Claims Tribunal. Consider, for example, the case of Shahin Shaine Ebrahimi. 22 In that case an expert was appointed by the Tribunal to value an expropriated Iranian company in which the claimants were shareholders. The process of drawing up the terms of reference, and identifying an expert with whom both parties could live, in itself took nearly seven months. The expert delivered his draft report three months later, and another six months later an expert hearing was held at which the parties were given the opportunity to respond to the contents of the expert's report. The final award, based heavily on the results of the expert inquiry, was issued almost exactly one year later. By the time all was said and done, some 16 months, excluding the time that elapsed between the expert hearing and the final award, were consumed by this characteristically civil law procedure for bringing out the facts.

What emerges from the foregoing discussion is that 'discovery', in the general sense, is present in both common law and civil law legal systems, and in both it can lead to protracted and intrusive proceedings. The mechanisms employed to discover and bring out the facts vary significantly between the two systems, of course, and any given practitioner (or arbitrator) is likely to prefer the mechanisms that are characteristic of his or her own system. 23 But this does not imply the existence of any 'Great Divide'.

Indeed, the simple expedient of replacing the word 'discovery' with a less politically charged and more descriptive term - the best choice probably being 'disclosure' - might go a long way towards removing the needless philosophical debate that tends to accompany the subject. 24 One might then profitably distinguish between the voluntary disclosure of evidence, which is present in all legal systems, and 'mandatory' or 'coercive' disclosure, which is used in both civil law and common law systems but is more characteristic of the latter.

It follows from what has been said above that the real question is not whether 'discovery' should be allowed in arbitration, but rather the extent to which the mandatory disclosure of evidence should be required in a given case. The follow-up question is how to manage the chosen disclosure regime effectively.

In our view, as explained in the following two sections, both questions must be answered in eminently practical ways, with a view to promoting efficiency and economy in the arbitral proceedings, and a just result at the end. [Page29:]

IV. How much disclosure to require

In addressing the question of the extent to which disclosure of evidence should be required in international commercial arbitration proceedings, the logical starting place is the major systems of institutional arbitration rules. After surveying these rules - which provide little specific guidance - we will suggest a checklist of factors to be considered by counsel and tribunals in deciding upon the scope of mandatory disclosure in a given case.

A. Arbitration rules: discretion but little concrete guidance

All of the major systems of arbitration rules provide for 'discovery' in the basic sense of the production of documents by a party in support of its own case. This is done in language ranging from advisory25 to mandatory, 26 depending on the particular set of rules.

Furthermore, and again without exception, the major systems of rules provide the possibility for other and less voluntary forms of disclosure. The ICC Rules, for example, state that '[a]t any time during the proceedings, the Arbitral Tribunal may summon any party to provide additional evidence'. 27 Somewhat broader and more detailed are the WIPO Arbitration Rules, which state:

At any time during the arbitration, the Tribunal may, at the request of a party or on its own motion, order a party to produce such documents or other evidence as it considers necessary or appropriate and may order a party to make available to the Tribunal or to an expert appointed by it or to the other party any property in its possession or control for inspection or testing. 28

Perhaps reflecting their common law origins, the new LCIA Rules, adopted in 1998, are more liberal still. Article 22.1 grants the tribunal the power:

to order any party to produce to the Arbitral Tribunal, and to the other parties for inspection, and to supply copies of, any documents or classes of documents in their possession, custody or control which the Arbitral Tribunal determines to be relevant;

and also to

order any party to make any property, site or thing under its control and relating to the subject matter of the arbitration available for inspection by the Arbitral Tribunal, any other party, its expert or any expert to the Arbitral Tribunal.

Lastly we come to the UNCITRAL Arbitration Rules - a creation of the United Nations and in some sense a set of arbitration rules intended to span the perceived Great Divide. 29 Article 24(3) of the UNCITRAL Rules provides that '[a]t any time during the arbitral proceedings the arbitral tribunal may require the parties to produce documents, exhibits or other evidence within such a period of time as the tribunal shall determine'. 30

Two conclusions emerge from the brief overview above. First, all of the major systems of international arbitration rules provide the tribunal with the authority to order various forms of mandatory discovery - i.e. disclosure in addition to the voluntary [Page30:] submission of documents believed by a party to support its own case - ranging from the forced production of documents to the disclosure or inspection of tangible evidence. 31 Second, none of the systems of rules provides substantial guidance to the tribunal (or the parties) regarding the extent or manner in which this authority should be exercised. Instead, tribunals are left to fall back on the essential but general injunction that the parties must be treated 'fairly and equally' at all stages of the proceedings and that each must be given a 'fair opportunity' to present its case. 32

The consequences of this are two-fold: first, the tribunal's decision on the extent of the required disclosure will be highly discretionary; and second, that decision will be essentially unreviewable, since in order to prevail in a setting-aside proceeding (or resist enforcement of the resulting award) a party would have to demonstrate that the effect of permitting (or foreclosing) a particular disclosure mechanism was so extreme as to deny that party a fair opportunity to present its case. 33 Consequently, both from the standpoint of the parties and of doing ultimate justice in the case, it is essential that the right choices be made in the first instance.

B. Factors to be considered by the tribunal in exercising its discretion

What, then, is to be done when the question of discovery is raised - as it is more and more often these days - in a concrete case? Should the tribunal permit 'discovery' in the sense of mandatory disclosure of evidence? If so, how much should it require?

In our view, there are no right or wrong answers to these questions, and the appropriate responses will depend almost entirely on the particular circumstances of the case. It is possible, however, to identify several factors that should be taken into account, on a case-by-case basis, in deciding the extent to which the mandatory disclosure of evidence should be required. These are listed and briefly described below.

1. Nature of the case

A primary consideration will be the nature of the case itself - i.e. whether it is fact-driven or primarily concerns a legal issue such as the interpretation of a contract provision. In the former instance, where the facts are primarily dispositive, discovery in the sense of mechanisms for the mandatory disclosure of evidence is more likely to be useful and appropriate. It is a simple fact of litigation life that parties will never knowingly disclose damaging documents (or other evidence) if given a choice; properly tailored disclosure mechanisms can level the playing field in this regard, thereby promoting a just result at the end of the process and possibly also enhancing the prospects of settlement in the meantime. 34 On the other hand, where a dispute primarily concerns a legal issue, the scope of disclosure can safely be curtailed severely - e.g. to contemporaneous documents relating to the negotiation of the contractual clause at issue - or even dispensed with altogether.

To provide one anecdotal example, in a recent arbitration in which many millions of dollars were at stake, the dispositive issue was the interpretation of a single clause in a [Page31:] massive contract. In that case the parties, both represented by common law counsel, agreed to forego document discovery altogether in the interest of speeding up the arbitral process - a wise choice properly endorsed by the tribunal in that case.

2. Amount in dispute

The amount in dispute should also be considered in the disclosure calculus. Even limited and narrowly-tailored disclosure mechanisms can be expensive to apply, both in terms of legal fees to the parties and time spent by the tribunal members (for which the parties ultimately have to pay). Economy and efficiency augur in favor of permitting little or no discovery in cases where the amount in dispute is small. 35

3. Nature of the claims

The nature of the claims asserted by the parties in the arbitration may also have an impact on the appropriate extent of discovery. In some cases, the documents or other evidence capable of proving or refuting a claim may reside solely in the possession of the other party. This might, in turn, justify a broader disclosure regime to ensure that a just resolution is reached.

At this point the civil law reader's objection reflex will no doubt have been triggered by the complaint that a party should have its evidence in place before bringing a claim, and there is indeed merit to this view. Claims or defenses that appear to have been conjured from the air, opportunist claims developed in the course of the proceedings, or even a party's failure to exercise proper diligence in record-keeping must surely weigh against a call for extensive disclosure mechanisms in an arbitration. At the same time, there will be some cases where a party is evidence-poor for reasons beyond its control; and indeed nearly every practitioner has at least one war story about a case originally appearing to be a 'dog with fleas' that turned out to have a pedigree after a proper evidentiary bath. 36 We posit simply that where it appears that a party lacks access to the necessary evidence for reasons beyond its reasonable control, more expansive disclosure mechanisms may be justified.

4. Background of the parties

It is arbitration gospel that the success of the process necessarily depends upon the consent and cooperation of the parties. For that reason the background of the parties should be considered in determining the appropriate extent of disclosure in a particular case. This is not to advocate a return to the Great Divide, but only to say that it must be recognized that civil law and common law parties (and arbitrators) have markedly different attitudes towards most forms of discovery. Especially in cases involving 'mixed' parties, the efficiency of the process is unlikely to be well-served by dragging a party kicking and screaming into a disclosure regime that it does not support and may well not understand. Indeed, even the internal cohesion of the tribunal could be threatened in such circumstances. As is so often true in arbitration, compromise and judicious restraint should be the watchwords when dealing with disclosure issues in a case where the parties or the panel members hail from markedly different legal traditions. [Page32:]

5. Availability of various discovery techniques

A final important consideration is the potential utility of the various possible disclosure mechanisms. Lawyers generally think of discovery as consisting of document exchanges, and indeed that is the most common form of compulsory disclosure of evidence (both in the common law system and in arbitration). But it is by no means the only one, as already described, and tribunals and parties should think more expansively when considering the question of discovery.

To cite only one example, a discovery device common in the United States is the so-called request for admissions, 37 in which a party puts to its opponent a series of factual and/or legal propositions which the opponent is required to admit or deny. This technique is merely a more formalized version of something that happens in arbitrations (and in civil law courtrooms) all the time: a party supports a particular proposition by arguing that the other party has failed to contest it. 38 The request for admissions accomplishes the same result more directly and unambiguously. It therefore merits consideration by tribunals and litigants as a mechanism for clarifying and limiting the issues in dispute at an early stage of the proceedings.

C. Conclusions

The foregoing list of considerations is certainly not comprehensive. Furthermore, it is entirely possible that in a given case the considerations noted above may point in different directions, thereby requiring a balancing of interests for which the arbitral process is frequently commended. Nonetheless, consideration of the listed factors may provide at least rough guidance in dealing with the thorny question of 'discovery' in the sense of the mandatory disclosure of evidence.

Finally, it should not be overlooked that the decision concerning the extent to which disclosure mechanisms will be employed in an arbitration is, in the first instance, a question for the parties themselves. They are the masters of the procedural process under all major systems of arbitration rules, and it would be remiss of counsel not to attempt to reach an agreement on this score at the outset of the proceedings (or even when drafting the arbitration clause). Only failing such agreement should the tribunal be called upon to plug the procedural gap.

V. How to control the process

Once it has been decided whether and how much disclosure will be required in a particular case, the next essential issue to address is how to control the process. The discovery regime in common law jurisdictions has rightly been lampooned for its propensity to generate fishing expeditions, mini-litigations and various other insidious consequences. 39 We have argued in this paper that 'discovery' in the general sense - i.e. mechanisms for the disclosure of evidence relevant to the case - is a necessary feature of every arbitration. But that is not to say that all forms of disclosure are equally easy to manage. The basic civil law scenario, in which disclosure is limited to the documents upon which each party wishes to rely, will lead to few if any disputes. 40[Page33:] However, once one ventures beyond that - and in our view it will frequently be desirable or necessary to do so - disputes inevitably will arise.

The key to controlling the discovery process in an arbitration is two-fold: (i) identifying the standards to be applied to the chosen disclosure regime, and (ii) identifying a procedure for dispute resolution. It bears repeating that both aspects should ideally be discussed between the parties at the outset, before a specific issue has arisen and is before the tribunal, with the attendant hyperbole that may thereupon come into play.

A. Standards

The preferred approach, then, is for the parties to reach agreement on the appropriate standards to govern the disclosure regime that they (or the tribunal) have decided to implement. By 'standards' we refer simply to the rules that will govern the disclosure process - i.e. what is to be done, by whom, and when.

Arbitration practitioners frequently advise clients that it is preferable to refer to a set of arbitration rules in their contracts rather than trying to draft in ad hoc provisions, which carries with it the risk of leaving 'gaps'. The same advice applies with respect to disclosure mechanisms. Simply saying 'we will exchange document requests', for instance, leaves a lot of important questions unanswered - such as when this will occur, whether categories of documents can be requested (as opposed to specifically identified documents), which, if any, privileges may be invoked and how any disputes are to be resolved. Such gaps can lead to extensive mini-litigations during the proceedings, as in fact occurred in the case referred to earlier in this paper. 41

A better idea is for the parties to agree upon (or the tribunal to direct) the application of an integrated set of rules to regulate whatever disclosure mechanisms are adopted. Consider, for example, a hypothetical case involving two parties who wish to engage in full American-style document discovery. In agreeing to this, such parties could reasonably consider making explicit reference to Federal Rule of Civil Procedure 26, which provides for an integrated system of document disclosure including timetables, rules as to scope and the invocation of privileges, and possible sanctions for abusive practices. The advantage is that the rules will be clear and known in advance, thereby promoting voluntary compliance and minimizing the need for intervention by the tribunal.

In most cross-border cases the parties will likely prefer an 'international' standard to the U.S. civil procedure rules, or indeed those of any other national jurisdiction. In that event a sensible choice would be to refer to the Rules of Evidence promulgated by the International Bar Association ('IBA Rules'). 42 The new version of the IBA Rules, 43 promulgated in June 1999, contains comprehensive and integrated provisions governing document discovery as well as the presentation of evidence by fact and expert witnesses. The Rules attempt to strike a balance between the attitudes of common law and civil law practitioners on these topics and therefore constitute an attractive model for adoption by parties or the tribunal in international arbitrations.

Cases may of course arise in which the parties are unable to agree on a pre-existing standard to govern the disclosure process. In that event the parties or the tribunal may have to design ad hoc provisions to regulate whatever mechanisms will be employed. [Page34:]

Let us take as an example the most common form of mandatory disclosure used in arbitration proceedings - document requests and exchanges. Once again without intending to be comprehensive, set out below are a number of issues that should be considered in designing a document disclosure process. 44

? Scope of the permitted discovery. The primary question here is whether document requests will be limited to specific documents, or whether requests for categories or classes of documents are permitted. In the new 1999 Rules, the IBA has opted for the broader variant. 45 The practice of the Iran-U.S. Claims Tribunal, on the other hand, is to permit only requests for specifically identified documents. 46

? Definition of relevance. Some definition of relevance will have to be agreed upon by the parties or supplied by the tribunal. This can be very broad - i.e. any documents relating to the dispute - but in many cases it may be able to be limited by agreement or arbitrator decision to the dispositive issues in the case. The new IBA Rules, for instance, use the criterion that the documents requested must be 'relevant and material to the outcome of the case'. 47 This already introduces an element of flexibility into the breadth of the required disclosure, since the IBA's formulation could be read as restricting discovery to documents that are potentially outcome-determinative.

Such a reading might, however, be unduly restrictive in many cases. The better reading would appear to encompass not only documents potentially determinative of the outcome of the case as a whole, but also documents relevant to specific issues (defined either by the parties or the tribunal) that bear upon the outcome of the case.

? Efforts made by the requesting party. Some arbitral panels, for example the Iran-U.S. Claims Tribunal, require a party to show that it has made reasonable efforts to acquire the evidence sought before it can request it from the other party. 48 Incorporating such a requirement obviously has the effect of narrowing the disclosure obligation substantially.

? Confidentiality. In some cases documents in the possession of one of the parties may contain trade secrets or otherwise be confidential. In that event, it must be decided whether such documents are subject to disclosure and, if so, how the confidentiality of those documents will be protected. 49

? Privilege. In many legal systems, certain information may be withheld from disclosure on grounds of privilege. The most common privileges are attorney-client, husband-wife and attorney work-product, but in certain jurisdictions (notably the United States) more esoteric privileges have been developed - including notably the so-called 'deliberative process' and 'investigatory' privileges, which protect the internal decision-making and fact-finding processes of organizations in some circumstances. 50 In any arbitral document discovery procedure, rules will have to be devised concerning which privileges can be invoked and how such objections must be raised. 51

? Memorializing the rules. Whatever standards are put into place, it is essential that they be memorialized in written form. The most likely mechanism would be a procedural order from the tribunal regulating in detail the matters noted herein. [Page35:]

Although the considerations listed above have been described in the context of document discovery, they could also provide useful guidance with respect to the pre-hearing disclosure of other kinds of evidence - for example, physical evidence and fact or expert witness testimony.

To summarize, discovery in arbitration can only work smoothly when the ground rules are clearly set out by the parties or the tribunal, preferably at the outset of the arbitration proceedings. Reference to an integrated standard - such as the IBA Rules - will often be preferable in order to avoid both the need to craft ad hoc procedures and the risk of 'gaps' in the regime that will frequently lead to later disputes. Apart from the possibility of direct adoption, the IBA Rules also provide useful guidance in identifying the issues relevant to designing a disclosure regime.

B. Resolving disputes

Even with the application of proper foresight in the design and implementation of a document disclosure (or other compulsory discovery) regime, disputes may nonetheless arise. Therefore it is essential for the parties and the tribunal to decide - at the outset of the process if possible - how disputes regarding the chosen disclosure regime will be handled.

There is ultimately no substitute for self-regulation by the parties, and a panel should always try to encourage counsel to resolve problems among themselves. This has the advantage of being faster, less ponderous and for the parties probably cheaper than resolution by the panel (in which case a party may be obliged to pay both attorney's fees and, under some sets of arbitration rules, additional fees for the tribunal's time). Nonetheless, since self-regulation can never be assured, prior agreement on the decision-maker is an important pre-requisite for the success of the process.

Who might the decision-maker in disclosure disputes be? The following options range from the obvious to the creative - and each has its good and bad points.

? The full arbitral panel. This has the advantage of bringing to bear on often tricky disclosure issues the full weight and experience of those entrusted to resolve the broader dispute. It is also, however, expensive and time-consuming and unless properly managed can lead to a mini-arbitration on disclosure issues.

? One member of the panel, most likely the chairman. This will cut down on the hours spent in resolving disclosure disputes, thereby saving the parties money and speeding up the process. On the other hand, it means that one member of the panel has access to information about potential evidence that is denied to the other members, if the contested evidence is ultimately not admitted in the proceeding.

? An outside expert. The out-sourcing option takes its inspiration from the Magistrate device employed in the U.S. court system and from the civil law expertise procedure. 52 This approach would take the burden from the tribunal itself, and might provide a further benefit of bringing special expertise to bear on issues of a technical or specialized legal nature to which the tribunal members may have had limited exposure (e.g. prevailing confidentiality practices in a particular industry area, or issues of legal privilege). These advantages, however, could be counter-balanced by the time and expense required to identify and involve a third [Page36:] party in the process. Further potential problems may also be created. Confidentiality of the arbitration will be diminished. Most worryingly, this option runs the risk of effectively amounting to the appointment of a fourth arbitrator with decision-making power over issues having the potential to influence the outcome of the arbitration.

? The tribunal secretary. While, of course, the tribunal secretary is traditionally and properly focused on administrative tasks, in nearly every case that person is also legally-trained, familiar with the file, available and cheap. As with an outside expert, however, confidentiality issues may arise and real decision-making power is effectively transferred away from the arbitrators. Many administrative secretaries would have the added problem of being associated with one of the arbitrators, thereby creating the risk that information on disputed evidentiary material could be revealed (intentionally or unwittingly) to one but not all of the panel members.

Whoever the ultimate decision-maker, it is vital that that person should have the time and resources to resolve discovery issues in a speedy and efficient way. Prior decision on the standards to be applied, as well as careful drafting of procedural orders governing the chosen disclosure regime, will also greatly facilitate the process.

VI. Conclusion

Practicality and flexibility are two of the hallmarks of arbitration that, it is widely believed, have led to its great success as a method of settling international commercial disputes. Too often, however, the day-to-day business of adjudicating cases becomes bogged down in larger philosophical controversies over the common law/civil law 'Great Divide', with the result that arbitral efficiency can be compromised. Nowhere is this more apparent than with the politically charged issue of discovery.

We submit that at least where discovery is concerned, the ongoing debate about the Great Divide is an irrelevant distracter. It should be left outside the hearing room, and parties and arbitrators should instead focus on finding practical, case-by-case solutions to the problems associated with the disclosure of evidence. We have suggested a number of factors that may assist in finding common-sense solutions to these problems, both with respect to the extent to which disclosure should be required in a given case and how the process can be effectively controlled.

The only fatal error for parties and panels will, perhaps, be reacting rather than planning ahead. This may indeed be the ultimate lesson to be gleaned from the tale of King Solomon, whose apparently ad hoc discovery technique, while ultimately successful, could just as easily have ended in disaster. Even in normal commercial disputes, where issues of life and death are happily not at stake, the participants will be well advised to take the lesson to heart: think broadly and creatively, but in advance.

1
A good overview of the diverging opinions on the subject can be found in the papers on the presentation of evidence in arbitration presented at the ICCA 12th International Arbitration Conference held in Vienna in November 1994. See Albert Jan van den Berg, ed., Planning Efficient Arbitration Proceedings - The Law Applicable in International Arbitration (Kluwer, 1996) 112-173 [hereinafter Van den Berg]. See also Richard B. Lillich, ed., Fact-Finding Before International Tribunals (11th Sokol Colloquium, 1992) 147-236 [hereinafter Sokol Colloquium].

2
Jan Paulsson, 'Overview of Methods of Presenting Evidence in Different Legal Systems' in Van den Berg, supra note 1 at 112.

3
Michael J. Mustill & Stewart C. Boyd, The Law and Practice of Commercial Arbitration, 2d ed. (Butterworths, 1989) [hereinafter Mustill & Boyd] at 324.

4
Andrew Rogers, 'Improving Procedures for Discovery and Documentary Evidence' in Van den Berg, supra note 1, 131 at 132. Having championed the availability of arbitral discovery, Mr Rogers goes on to advance a number of recommendations for controlling possible abuses. See id. at 138-39.

5
Claude Reymond, 'Civil Law and Common Law Procedures: Which is the More Inquisitorial? A Civil Lawyer's Response' (1989) 5 Arbitration International 357 [hereinafter Reymond] at 360-361. In the same vein, see Thomas E. Carbonneau, 'Darkness and Light in the Shadows of International Arbitral Adjudication' in Sokol Colloquium, supra note 1 at 153.

6
Paolo Michele Patocchi & Ian L. Meakin, 'Procedure and the Taking of Evidence in International Commercial Arbitration: The Interaction of Civil Law and Common Law Procedures' [1996] IBLJ/RDAI 884 at 891.

7
Jan Paulsson, 'Overview of Methods of Presenting Evidence in Different Legal Systems' in Van den Berg, supra note 1 at 114.

8
Black's Law Dictionary, 6th ed. (West, 1990) at 466.

9
K. Seide, ed., Dictionary of Arbitration and its Terms (Oceana, 1970) at 73.

10
See Arthur Marriott, 'Evidence in International Arbitration' (1989) 5 Arbitration International 280 at 281.

11
This remains the case in England notwithstanding the Woolf Reforms, which have, however, established more stringent controls on the discovery process. See generally Chancery Guide (revised Nov. 1999) paras. 6.2-6.5 at 10.

12
Mustill & Boyd, supra note 3 at 324. Common law authors generally put forward three basic rationales for having discovery: (1) to assist in the search for the truth; (2) to save time by narrowing the issues for trial; and (3) to avoid the unfair surprise that may occur at trial if both parties are not aware of the other's basic evidence. See Monica P. McCabe, 'Arbitral Discovery and the Iran-United States Claims Tribunal Experience' (1986) 20 Int'l Lawyer 499 [hereinafter McCabe] at 503.

13
Federal Rules of Civil Procedure 27-28, 31, 33-36.

14
See 28 U.S.C. §§ 631, 636(b)(1)(A); see also 'Magistrate Judges', 32 Am. Jur. 2d §§ 161, 185, 199 (1999).

15
See generally Christian Borris, 'The Reconciliation of Conflicts Between Common-Law and Civil-Law Principles in the Arbitration Process' in Stefan N. Frommel & Barry A.K. Rider, eds., Conflicting Legal Cultures in Commercial Arbitration: Old Issues and New Trends (Kluwer, 1999) 1 at 11.

16
See e.g. French New Code of Civil Procedure ('French NCCP'), Art. 218 (granting a court the power to summon witnesses to appear, either on its own motion or at the request of the parties); Dutch Code of Civil Procedure ('Dutch CCP'), Arts. 192, 198, 199, 214, 429j (conferring similar powers).

17
See French NCCP, Art. 11 ('At the request of a party to the proceedings, the judge may order the other party to produce evidence in its possession; refusal to do so may result in a fine.'); Dutch CCP, Arts. 19a and 110. It should be noted, however, that requests to a court to order the production of documents are, in practice, rarely granted in France or the Netherlands.

18
See e.g. Dutch CCP, Arts. 221-225; Swiss Federal Code of Civil Procedure, Arts. 57-60; Code of Civil Procedure, Geneva, Arts. 255-269; Italian Code of Civil Procedure, Art. 61. With respect to French practice, see Hans Smit, 'Roles of the Arbitral Tribunal in Civil Law and Common Law Systems with Respect to Presentation of Evidence' in Van den Berg, supra note 1, 161 at 163 ('In France, . . . an expertise is used to discover information that in common law systems would be obtained through pretrial discovery.').

19
See id.; see also Dutch CCP, Arts. 226-232 (providing for the power of the judge to order site inspections by experts).

20
See e.g. French NCCP, Art. 11 ('The parties are under the obligation to cooperate with the judge in his enquiries and to the extent that they fail to do so, the judge may draw such inferences as he sees fit.'). This principle also exists in Dutch law, see Hoge Raad (Supreme Court) 12 June 1953, NJ 1954, 61; and in Italian law, see Italian Code of Civil Procedure, Art. 116.

21
See e.g. UNCITRAL Arbitration Rules, Art. 27. Paragraph 2 of Article 27 requires the parties to 'give the expert any relevant information or produce for his inspection any relevant documents or goods that he may require of them'.

22
Shahin Shaine Ebrahimi, et al. and The Government of the Islamic Republic of Iran (12 October 1994), Award No. 560-44/46/47-3, at 15-19.

23
See Durward V. Sandifer, Evidence Before International Tribunals (University Press of Virginia, 1975) at 8.

24
This change in nomenclature has been adopted in English procedure as a result of the Woolf Reforms. See Part 31 of the 1999 English Civil Procedure Rules ('Disclosure and Inspection of Documents').

25
See UNCITRAL Rules, Arts. 18(2) and 19(2) (providing that the claimant and respondent 'may annex' all documents they deem relevant to the Statement of Claim and Statement of Defence); ICC Rules, Art. 20(2) (referring to the 'written submissions of the parties and all documents relied upon').

26
See LCIA Rules, Art. 15.6 ('All Statements referred to in this Article shall be accompanied by copies . . . of all essential documents on which the party concerned relies . . .'); WIPO Arbitration Rules, Arts. 41(c) and 42(b) (providing that the Statement of Claim and Statement of Defense shall 'be accompanied by' the relevant documentary evidence upon which each party relies); see also NAI Arbitration Rules, Art. 28(1).

27
ICC Rules, Art. 20(5). Compare NAI Arbitration Rules, Art. 28 ('The arbitral tribunal shall have the power to order the production of specific documents which it deems relevant to the dispute.').

28
WIPO Arbitration Rules, Art. 48(b). This perhaps reflects these Rules' intended function in resolving intellectual property disputes. Compare also ICSID Rules of Procedure for Arbitration Proceedings, Rule 34(2) (providing that the tribunal 'may, if it deems it necessary at any stage of the proceeding: (a) call upon the parties to produce documents, witnesses and experts; and (b) visit any place connected with the dispute or conduct inquiries there').

29
As expressed by Professor Pieter Sanders, they were intended 'to be acceptable in both capitalist and socialist systems, in developed and developing countries and in common law as well as civil law jurisdictions' (quoted in W.G.O. Morgan, 'Discovery in Arbitration' (1986) 3:3 J. Int'l Arb. 9 at 22).

30
Compare AAA International Arbitration Rules, Art. 19(3) ('At any time during the proceedings, the tribunal may order parties to produce other documents, exhibits or other evidence it deems necessary or appropriate.').

31
In this respect, the major systems of rules mirror the underlying arbitration acts in the most frequent venues for international arbitrations, which generally afford arbitral tribunals wide discretion in determining how the proceedings will be conducted, including the power to order various forms of disclosure. In the common law countries, see Section 34 of the English Arbitration Act (1996) (permitting arbitral tribunals to order the production of documents); Section 7 of the United States Federal Arbitration Act, 9 U.S.C. § 7 (same). In the civil law host countries, see French NCCP, Art. 1460 (giving the tribunal power to order a party to produce 'an element of proof' and obliging the parties to 'cooperate in all matters regarding evidence' on pain of the drawing of negative inferences); see also Dutch CCP, Art. 1039(4) (granting the arbitrator 'the power to order the production of documents'). In addition, both the German and the Dutch Codes further provide for extensive powers to tribunal-appointed experts. See German Code of Civil Procedure, § 1049(1); Dutch CCP, Art. 1042(2).

32
See UNCITRAL Rules, Art. 15(1); ICC Rules, Art. 15(2); LCIA Rules, Art. 14(1); NAI Rules, Art. 23(1); ICSID Arbitration Rules, Rule 6(2) (obligation of the tribunal to 'judge fairly'); AAA International Arbitration Rules, Art. 16(1); WIPO Arbitration Rules, Art. 38(b).

33
See Charles N. Brower, 'Discovery and Production of Evidence in the U.S.: Theory and Practice' in Taking of Evidence in International Arbitral Proceedings/L'administration de la preuve dans les procédures arbitrales internationales (ICC Publishing, 1990) 7 at 18.

34
See id. at 11.

35
See McCabe, supra note 12 at 505.

36
See e.g. Rogers, supra note 4 at 133-134.

37
See Federal Rule of Civil Procedure 36; see also supra text accompanying note 13.

38
Indeed, the equivalent may already be done by judges in some civil law jurisdictions. See Reymond, supra note 5 at 362 (noting that a civil law judge may query counsel for the parties in order to 'list the disputed issues . . . [and] to try to reduce them to a minimum'). Compare Dutch CCP, Art. 176 (providing that a proposition put by a party is considered to be admitted if not specifically and properly denied by the counter-party).

39
Even Warren Berger, the former Chief Justice of the U.S. Supreme Court, was once moved to comment that the civil litigation system in the U.S. had become 'too costly, too painful, too destructive, too inefficient for a truly civilized people'. Address (Mid-Year Meeting of American Bar Association, 1989) 52 U.S.L.W. 2471 at 2471.

40
The only disputes are likely to relate to the authenticity of the documents produced, or perhaps to their relevance.

41
See supra, pp. 25-26.

42
See Jan Paulsson, 'Cross-Enrichment of Public and Private Law Dispute Resolution Mechanisms in the International Arena' (1992) 9:1 J. Int'l Arb. 59 at 63 (noting that the IBA Rules 'reflect a workable accommodation of diverse procedural traditions which greatly resembles what skilled arbitrators do in practice on a daily basis').

43
International Bar Association, Rules on the Taking of Evidence in International Commercial Arbitration (1 June 1999) [hereinafter IBA Rules], reprinted in (1999) XXIVa Y.B. Comm. Arb. 410.

44
Many of these have been addressed in the new IBA Rules, and therefore a citation will be provided in the footnotes to the relevant IBA provision where applicable.

45
See IBA Rules, Art. 3(3).

46
See Islamic Republic of Iran and United States of America, Case No. B1 (Claims 2 and 3), Order of 18 March 1998; Islamic Republic of Iran and United States of America, Case No. A15 (I:D and I:H), Order of 24 May 1994, para. 5; Joan Ward Malekzadeh, et al. and Islamic Republic of Iran, Case No. 356, Order of 12 August 1993.

47
IBA Rules, Art. 3(b).

48
See McCabe, supra note 12 at 515-517. See also IBA Rules, Art. 3(3)(c) (requiring the party making a document request to certify that 'the documents requested are not in the possession, custody or control of the requesting Party').

49
See id., Art. 9(2)(e). Compare WIPO Arbitration Rules, Art. 52.

50
See K.L v. Edgar, 964 F. Supp. 1206, 1208 (N.D. Ill. 1997) (deliberative process privilege); Raphael v. Aetna Casualty and Surety Co., 744 F. Supp. 71, 74 (S.D.N.Y 1990) (investigatory privilege).

51
Compare IBA Rules, Art. 9(2)(b) (permitting objection against a document request based upon a 'legal impediment or privilege under the legal or ethical rules determined by the Arbitral Tribunal to be applicable').

52
See supra text accompanying notes 14-19. Art. 6 of the IBA Rules likewise provides for this possibility, as do the WIPO Rules. See WIPO Arbitration Rules, Art. 52(d) (providing for the appointment of a 'confidentiality advisor' to review materials claimed to be confidential and therefore not subject to disclosure). Notably, ICC also has an optional expertise procedure. See ICC Rules for Expertise (in force from 1 January 1993).