The stakes

To distinguish between these two concepts is a matter of considerable concrete importance. Decisions of tribunals which do not respect jurisdictional limits may be invalidated by a controlling authority. But if parties have consented to the jurisdiction of a given tribunal, its determinations as to the admissibility of claims should be final. Mistakenly classifying issues of admissibility as jurisdictional may therefore result in an unjustified extension of the scope for challenging awards, and frustrate the parties' expectation that their dispute be decided by the chosen neutral tribunal.

Of course, national laws may explicitly provide that arbitral disposition of issues of admissibility are not final. But then again, national laws may explicitly provide that all decisions by arbitrators are subject to full appeals, including findings of fact or conclusions of law. Indeed, national laws may forbid arbitration altogether. Yet that is emphatically not the modern trend. This essay proposes an approach consistent with an international consensus that decisions of arbitrators having jurisdiction are final.

The problem in a nutshell

A routine commercial dispute may serve as a convenient illustration. Vekoma B.V. (Netherlands) agreed to deliver certain quantities of coal to Maran Coal Corp. (US). Their contract contained a clause providing that disputes would be submitted to arbitration in Switzerland under the Rules of the International Chamber of Commerce. That clause also required that any arbitration must be initiated 'within thirty days after it was agreed that the difference or dispute cannot be resolved by negotiation'. A dispute arose. Maran wrote to Vekoma on Day T that unless Vekoma agreed to a suggested accommodation by Day T+8, Maran would initiate arbitration. Vekoma never responded. On Day T+83, [Page601:]

Maran wrote a reminding letter. On Day T+93, Vekoma rejected Maran's complaint. On Day T+122, Maran initiated arbitration. Vekoma argued that Maran's right to arbitrate the claim had lapsed. The ICC arbitral tribunal (presided by the well-known French arbitrator Jean-Louis Delvolvé) dismissed this objection on the grounds that Vekoma should in good faith have answered Maran's letter of Day T, and that the thirty-day period began to run only on Day T+93, when Vekoma made its position known. The commencement of arbitration on Day T+122 was therefore timely. An award of some US$ 650,000 was rendered in favour of Maran. Vekoma sought annulment. The Swiss Federal Tribunal upheld the challenge, holding that the agreement to arbitrate was subject to a condition subsequent, namely the thirty-day limit, and that that condition failed when Maran neglected to initiate arbitration within thirty days of day T+8-i.e. the limit Maran itself had established. 1

The annulment was met with sharp criticism. 2Some challenged the Swiss court's analysis of the parties' agreement and conduct. Everyone is of course free to criticise judicial reasoning. But if the issue is whether an arbitral tribunal had jurisdiction to decide a dispute, there can be no question that the court was entitled to reconsider the arbitrators' conclusion as part of its control function.

Others questioned the judgment on the basis that it disregarded a presumption that parties do not intend to make an invalid stipulation of arbitration. But this objection misses the mark, since the Swiss court did not question the validity of the arbitration clause.

The fundamental error of the annulment was rather that it misunderstood the nature of the challenged arbitral decision. The arbitrators had made a decision as to the admissibility of the claim. The parties had agreed that all disputes under their contract would be decided by this particular tribunal, and as noted the validity of the arbitration clause was not at issue. The arbitrators therefore decided the admissibility issue in the exercise of their jurisdictional authority. The Swiss court was simply not entitled to review their decision in this regard.

This simple example has parallels in current cases involving far greater stakes, particularly in the area of arbitration under international investment treaties. [Page602:]

Participants in such proceedings are sometimes tempted to invoke comments about admissibility made by judges of permanent international tribunals whose decisions are not subject to any kind of review. This is perilous. The International Court of Justice (ICJ) is a forum of first and last resort. Its pronouncements about jurisdiction and admissibility are, generally, pure abstractions. 3The classification of an issue as one of jurisdiction or admissibility may serve to explain the ICJ's ordering of its own procedure. 4But no matter what the ICJ says, its word is by definition final.

Nor is much at stake when most national courts consider the nature of preliminary objections. Here the opposite cause has the same effect: whenever all decisions are subject to appeal, review is available whether the initial decision is deemed to have related to a matter of jurisdiction or one of admissibility. Either way, the higher court's opinion will prevail.

By contrast, a dominant feature of arbitration is that jurisdictional decisions are reviewable, but not others. (This essay is not concerned with extreme categories of 'decisions' that violate due process, or disregard imperative rules of public policy.) Hence it is vital to understand the fundamental distinction between the two concepts. They are indeed as different as night and day. It may be difficult to establish the dividing line between the two. There is a twilight zone. But only a fool would argue that the existence of a twilight zone is proof that day and night do not exist. 5

Night and day

As stated, the occasional terminological digressions of international courts of last resort tend to be unhelpful, because they have no reason to see the importance of the distinction in terms of reviewability. And so, it seems, they fire harmless shots from the hip in reaction to particular circumstances. [Page603:]

The judges of the ICJ would of course be perfectly able to tell night from day-if it really mattered to them. When Gerald Fitzmaurice applied his mind to the subject in The Law and Procedure of the International Court of Justice, he revealed an acute perception of the distinction in a single sentence:

But an unsuccessful jurisdictional plea leaves open the possibility that a finding on the ultimate merits may still be excluded through a decision given against the substantive admissibility of the claim. 6

Thus 'substantive admissibility' may arise as an issue after jurisdiction has been established. Fitzmaurice did not go on to state explicitly that such issues of admissibility, like other substantive matters, are not subject to review once decided by a tribunal having jurisdiction. 7But then his frame of reference was still that of the international proceduralist Shabtai Rosenne, who had written in 1957 that internationally there is 'no hierarchy of courts with pre-determined jurisdiction [but a] haphazard multiplicity of courts, with no preference of schematic hierarchy between them, each having jurisdiction to the extent specified in the international treaty by which it was established.' 8

That vision-never wholly justified-must be abandoned in light of modern realities. Today, most international tribunals are subject to the review of hierarchically dominant bodies. This is true in the rapidly burgeoning field of investment arbitration, where arbitral determinations are subject to annulment[Page604:]

by national courts or statutory mechanisms like the ad hoc committees of the International Centre for the Settlement of Investment Disputes (ICSID), or refusal of recognition and enforcement under the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards. And such review has long been a fundamental factor in international commercial arbitration.

International law can no longer ignore the implications of recognizing the distinction between jurisdiction and admissibility from the perspective of reviewability. Moreover, national authorities can no longer ignore the desirability of harmonisation whenever they deal with cases having an international dimension.

Fitzmaurice himself should have seen that Rosenne's vision was doubtful, given that his own work, in a section entitled precisely 'Jurisdiction and Admissibility', prominently featured the Ambatielos case-decided by the ICJ as long ago as 1952 on the clear premise of a hierarchical relationship between itself and an international arbitral commission created under a bilateral treaty. The Court was faced with a challenge to its jurisdiction to decide the dispute (between Greece and the UK). It found that it had the authority to determine whether under a particular treaty the case should be referred to arbitration. The Court had also been told by the UK that the claim could not be heard because Mr Ambatielos had not exhausted local remedies and because the claim was being presented after undue delay. These objections, so the Court held, were 'arguments in defence directed to the admissibility of the Ambatielos claim . . . the Court expresses no view concerning the validity or legal effect of these arguments'. 9These objections were thus left to be decided by an arbitral award. 10

There is no relevant conceptual difference between the Ambatielos scenario (a prior determination by the ICJ that a given case must be heard in international arbitration) and instances where arbitral awards are subsequently reviewed on jurisdictional grounds. The same deference to determinations of admissibility should be shown in both situations.

In the dawn of the twenty-first century, awards in the burgeoning field of investment arbitration are entering the public domain at a dizzying pace. They are curiously inconsistent in their treatment of our two concepts. But instead of dwelling on what frankly appear to be heedless incongruities, we would do well to focus on the most illuminating precedent. [Page605:]

SGS v. Philippines11is in this respect a model of clarity. The two parties had entered into a service contract containing a stipulation to the effect that disputes arising out of it should be referred to a Philippine court. Claiming a failure of payment, SGS nevertheless sought ICSID arbitration under the bilateral investment treaty between its home country, Switzerland, and the Philippines. The Philippines asserted that the ICSID tribunal had no jurisdiction given the fact that SGS's essential claim was an alleged failure of payment under the contract, and the disputes clause in the contract required such matters to be taken to the Philippine court and nowhere else. The tribunal noted that the broadly-worded treaty encompassed contract claims and that the claimant had specifically alleged breaches of the treaty. It therefore concluded that it had jurisdiction. 12Nevertheless, it did not accept that a BIT would 'automatically override the binding selection of a forum . . . to determine their contractual claims,' and that as long as 'the essential basis' of the claim is a breach of contract, 'the tribunal will give effect to any valid choice of forum clause in the contract'. 13

This led the tribunal to devote a section of the award to the 'distinction between jurisdiction and admissibility'. Although there was arbitral jurisdiction, there was an impediment to the claimant's reliance on the contract as the basis of its claim before the BIT tribunal 'when the contract itself refers that claim exclusively to another forum'. 14A decision under the treaty would be premature, and the claim was therefore inadmissible.

The implications are clear. If the Philippine court were to establish that a given amount was due but its judgment were then disregarded, the claim before the international jurisdiction would become mature. A claim would also be ripe if the Philippine court rejected the contractual claim but did so in a manner alleged to be wrongful under international law, as a denial of justice.

Whether an arbitral tribunal asserts or declines jurisdiction, its decision on this account should be reviewable. In the SGS v. Philippines case, therefore, only the Philippines was theoretically in a position to challenge the award on jurisdiction, because the tribunal asserted jurisdiction. (This the Philippines was[Page606:]

unlikely to do, having prevailed since the matter was referred to its courts.) As the claimant, SGS could hardly complain when the arbitral tribunal agreed that it had jurisdiction. Nor could it then raise a challenge when the tribunal, in the exercise of its jurisdiction, found that there was an 'impediment' and that the claim was 'premature' - two words that denote inadmissibility.

This lucid approach may be contrasted with the unsatisfactory terminology used in the operative part of the jurisdictional award in Methanex v. US. The ICSID tribunal explained its disposition of 'the USA's several challenges based on the "admissibility" of Methanex's claims' and of 'the USA's jurisdictional challenges'. 15The careful reader observes the absence of inverted commas around the word jurisdictional. It is surely no accident. In deference to the distinguished arbitrators, one may prefer to interpret the treatment of 'admissibility' as a sign that the tribunal was adopting the terminology used in the pleadings without necessarily endorsing it.

The terminology was in fact wrong. The USA's final pleadings made clear that its conception of 'challenges to admissibility' was that 'taking all of the allegations of fact made to be true, including uncontested facts, … as a matter of law, there can be no claim, and the claim is ripe for dismissal at this stage for that reason'. 16This may be a very good definition of a motion to dismiss for failure to state a cause of action, or, to use the expression current in England, a strikeout application. The tribunal quite correctly took the view that this was not a jurisdictional challenge. Regrettably it did not clear the air by adding that it was not an objection of inadmissibility either. This can be easily demonstrated. By its own definition, the USA's challenge required consideration of the 'matter of law' which would preclude the claim. The merits of a case are not limited to issues of fact. The USA may have been seeking a knock-out blow on the merits before any hearing of facts, but however early such a defence may be pleaded and considered, it is a defence on the merits and not a matter of admissibility. The USA was not arguing that the case was unhearable, but that it was legally hopeless. That is precisely how one should understand the difference between a challenge of inadmissibility and a strike-out application.

Even if facts are assumed true as pleaded (and there is therefore no resolution of factual controversies), a strike-out application involves a consideration of the merits of the case; the objective of the application is precisely to secure a[Page607:]

determination that the legal basis of the claim is meritless. A ruling on such an application is therefore as unappealable as decisions on the merits generally. In this respect it is a harmless mistake to refer to such applications as dealing with admissibility. Yet the confusion of terms comes at a cost when it blurs such fundamentally distinct concepts.

In Enron v. Argentina, the arbitral tribunal said quite correctly that 'a successful admissibility objection would normally result in rejecting a claim for reasons connected with the merits'. 17But the arbitrators did not see the implications of this deceptively simple phrase, for they also asserted that the distinction between jurisdiction and admissibility does 'not appear to be necessary in the context of the ICSID Convention, which deals only with jurisdiction and competence'. Surely the arbitrators did not really mean that the ICSID Convention deals with jurisdiction and competence in the sense of establishing their distinctive characteristics; we are still waiting for someone to do that. 18 They obviously were conveying nothing more than the observation that the ICSID Convention expressly mentions both jurisdiction and competence. So it does, and one surmises that this belt-and-braces drafting was intended as a means of ensuring that the intended effect did not become a hostage to murky doctrinal debates. 19

On the other hand, although the ICSID Convention does not expressly use the word 'admissibility,' its all-important Article 52 mandates annulment of awards for excess of power-and so, contrary to the statement just quoted, it does indeed become 'necessary' to understand the difference between objections to be finally decided by the arbitrators and objections subject to review.

The twilight zone

One can understand why a superficial reading of cases leads to the temptation to lump together all objections which, if upheld, would end the case. This common characteristic may explain why indices in case reports tend to be[Page608:]

hopelessly inconsistent. It would be caddish to disparage the thankless efforts of index preparers, and the problem is so widespread that it would be unfair to single out a particular publisher. Suffice it to say that in the most recent volume of an otherwise invaluable series of case reports one finds under the heading 'admissibility' two decisions which are indeed properly matters of admissibility (whether additional claims may be raised once the initial pleadings have been submitted, and whether the claim was formally submitted within the time limits provided for in the relevant treaty), and three others which are not. These three concern either 'the nature and subject-matter' of the dispute or the period of time when the alleged cause of action had arisen. These are recognisable as classical jurisdictional problems, since the relevant treaty that created arbitral jurisdiction was circumscribed both ratione materiae and ratione temporis. Meanwhile, under the heading 'jurisdiction', one finds references to decisions dealing with the alleged failure to exhaust local remedies and the alleged failure of new claims to remain within the scope of the initial notice: issues properly understood as ones of admissibility.

We can live with a twilight zone, but not if it lasts until dawn-and certainly not if dawn never comes. Unfortunately the terminology perpetuated by the US Supreme Court seems to condemn us to permanent twilight. The problem is encompassed in a single vaporous locution: arbitrability. It is a matter of wonderment that the Court's orders in concrete cases tend to be quite sensible- as though the judges have a kind of night vision which penetrates the shadowy abstractions. But all judges and arbitrators do not have the same discernment as the judges of the Supreme Court, and many are led astray.

The persistent abuse of the word arbitrability has led to international disharmony, because elsewhere that word has an established meaning that is narrow and therefore useful. The American use of the word conflates admissibility and jurisdiction. This has created a vast muddle. With all respect due to the eminent judges of the Supreme Court, one cannot resist wishing that they would retrieve their copybook and start over.

It is necessary to worry about language. Language is liberating; language ensnares. Words illuminate; words conceal. The attempt to seek a sensible accommodation of the words arbitrability, admissibility and jurisdiction is a worthy endeavour that could eliminate much waste of legal resources.

In the very first edition of their Law and Practice of International Arbitration, published in 1986, Alan Redfern and Martin Hunter wrote: 'The concept of[Page609:] arbitrability is in effect a public policy limitation upon the scope of arbitration as a method of settling disputes. Each state may decide, in accordance with its own public policy considerations, which matters may be settled by arbitration and which may not.' 20Arbitrabilité, wrote Fouchard and Gaillard, 21is the question whether the dispute is of a type (une matière, e.g. divorce or paternity) susceptible to being resolved in arbitration or whether particular parties (e.g. certain public law entities) are legally entitled to agree to arbitration.

When leading French and English authors agree to the meaning to be ascribed to a Latinate term of considerable significance, it seems most unfortunate for American authorities to promote an entirely different meaning. 22Of course, the US Supreme Court has just as much right as Humpty Dumpty to assert that 'when I use a word, it means just what I chose it to mean, neither more nor less'-but that does not, at least on this side of the looking glass, make it a welcome idea. Fouchard and Gaillard referred to the US Supreme Court's usage as 'génératrice de confusions' and properly noted that the startling American usage is at odds with the one that underlies the great worldwide effort at harmonisation we know as the UNCITRAL Model Law. The only official version of the Model Law is in the English language with American spelling- but not, it appears, with an American meaning.

Article 1(5) of the Model Law provides: 'This Law shall not affect any other law of this State by virtue of which certain disputes may not be submitted to arbitration or may be submitted to arbitration only according to provisions other than those of this Law.' The section of Holtzmann and Neuhaus' Guide to the Model Law dealing with Article 1(5) is simply entitled: 'Arbitrability'. [Page610:]

The authors write: 'Article 1(5) was the result of consideration of a variety of proposals aimed at excluding nonarbitrable subject matters from arbitration under the Model Law.' 23These two US lawyers deal repeatedly with 'arbitrability' throughout their 1307-page volume, but never in the sense of the US Supreme Court.

When the US Supreme Court decided First Options in 1995, 24one could only conclude that neither the judges nor their clerks had made themselves familiar with the relevant international literature. True, it was a domestic case and the written pleadings of the parties therefore did not have an international focus. At any rate, the Court perpetuated the unfortunate usage of the word arbitrability as interchangeable with either jurisdiction or admissibility. For a fuller exposition of the fuzziness of the Court's terminology, Professor Park's study remains valuable. 25 For present purposes, a summary will suffice.

Mr and Mrs Kaplan owned a small investment firm. It became indebted to First Options, a clearing house. Some difficulties arose, and a work-out agreement was signed by First Options and the firm. The agreement contained a broad arbitration clause. First Options subsequently decided to bring an arbitration to enforce payment terms of the work-out agreement. The difficulty arose because First Options wanted to pursue the Kaplans as well as their firm, but of course they were not signatories to the document containing the arbitration clause.

The Kaplans objected. An arbitral tribunal was nevertheless empanelled. It found itself to have authority to decide the issue of 'arbitrability', and ordered the Kaplans to answer on the merits-which the Kaplans did, and lost. They then took the matter into the US court system. In the end, their position was vindicated by the US Supreme Court. This was as it should have been. Yet there was, one might say, madness in the Court's method.

Instead of seeing that this was an issue of jurisdiction, and that the issue of whether a party has subjected itself to the authority of an arbitral tribunal can never be finally decided by the relevant arbitrator(s), the Supreme Court[Page611:]

explained that this was an issue of arbitrability, and that the answer depended on the 'fairly simple' question: 'Did the parties agree to submit the arbitrability question itself to arbitration?' 26

It may be 'fairly simple' to imagine that parties might address arbitrability explicitly, for example, as Kenneth Pierce has suggested, by providing that 'any controversy regarding the scope or meaning of this arbitration clause shall be settled in the arbitration'. 27Yet it is something so rare or accidental in international practice that one cannot accept that the possibility of such a stipulation should serve as a reasonable basis on which to expect parties to regulate their conduct. And above all, First Options was a case about jurisdiction. Once one so perceives it, the only question is whether or not the Kaplans could be held to have consented to arbitrate the claim against their firm. Nothing would have changed if the separate corporate entity had made a special 'arbitrability stipulation'.

In the aftermath of First Options, the US Supreme Court has also considered that the following cases involved issues of 'arbitrability':

(i)In Howsam v. Dean Witter Reynolds, Inc. 28the question was whether a claimant could invoke the arbitration provisions of the National Association of Securities Dealers Code, which enables customers to seek arbitration within six years of the event giving rise to the dispute. The Court said that it did not believe that the parties had intended to have such an arbitrability issue decided by the courts rather than the arbitral tribunal. 29Instead of purporting to read the minds of parties who assuredly had not been thinking about this matter at all, it would have been better to say that this was about admissibility, and therefore exclusively a matter for arbitral determination.

(ii)In Green Tree Financial Corp. v. Bazzle, 30numerous owners had taken out home improvement loans from Green Tree. All loan agreements contained the identical broad arbitration clause. A class action was brought, alleging that Green Tree had violated consumer legislation. A judge ordered a single arbitration between Green Tree and the borrowers as a class (individuals could opt out if they wished). An award of some $32 million[Page612:]

was rendered against Green Tree, which applied for its annulment on the grounds that the imposition of class-wide arbitration was in violation of the individual arbitration clauses. The Court said that this was not about whether the parties had agreed to arbitration, but 'what kind of arbitration they had agreed to'-a matter for the arbitral tribunal to decide. Once again, rather than leaving the reader to ponder the meaning of the concept of 'kind of arbitration,' it would have been better to say that this was a matter of admissibility.

'Not so fast!' it may be objected. The jurisdiction/admissibility dichotomy may be conceptually unassailable, but was the classification correct in these cases? What about Green Tree's argument that it had agreed to arbitrate any dispute relating to Contract No. 1 with Client A, and to arbitrate any dispute relating to Contract No. 2 with Client B, and so on until Contract No. N with Client X- but not to face them en bloc? And what about Dean Witter Reynolds' argument that they had consented to arbitrate only timely claims? Why are these not matters of jurisdiction?

Even the cases of alleged failure to respect a condition precedent to participate in an attempt at negotiation might be argued to belong to the jurisdictional category. True, the French Supreme Court has held that a claim is inadmissible (irrecevable) if it arises out of a contract stipulating that the parties must submit to a conciliation procedure before initiating legal action, and that condition precedent had not been fulfilled. 31Mr Justice Colman of the English High Court decided to the same effect in Cable & Wireless plc v. IBM (UK) Ltd. 32In her study of nine ICC arbitral awards entitled 'Multi-Tiered Dispute Resolution Clauses in ICC Arbitration', Dyalá Jiménez Figueres showed that the arbitrators viewed this issue as one of admissibility, to be accepted or rejected depending on whether the terms of the stipulation were obligatory or optional, and in the former case whether the procedure had been attempted. 33Fine and good- but what happens when the decision of the arbitral tribunal goes against the respondent, who insists: 'I consented only to arbitrate claims that had been the object of a bona fide, structured attempt at settlement?' Why is this not a matter of jurisdiction, reviewable by the courts? [Page613:]

What about the issue of locus standi? What if a respondent State contends that a foreign investor does not have standing to sue on account of its rights in a company independently of that company's own rights? 34What about a challenge, as in South West Africa, 35on the ground that the plaintiff cannot establish 'any legal right or interest in the subject matter of the claim'?

Are all of these issues in the twilight zone, and therefore hostage to whatever presumptions individual decision-makers may wish to make about unexpressed intentions?

The lodestar

Having done quite a lot to get us into trouble, the US Supreme Court now comes to our rescue. It may be convincingly demonstrated that it would have reached a result opposite to the Swiss Maran Coal judgment. The Court's way of dealing with timeliness issues is clear and instructive. 36

Once again, the US Supreme Court deserves applause for its analysis and not for its choice of words; it refers to timeliness issues as ones pertaining to 'procedural arbitrability'. (It is hard to improve on Professor Rau's comment: 'But who could possibly think that linking together these two words-each with a troubled history, and each notoriously manipulable and vague-is calculated to increase intelligibility?' 37) The two leading cases are Wiley38and Howsam. 39The former involved a dispute as to whether a mandatory two-step negotiating process had been accomplished as a precondition to arbitration; that issue, said the Court, should be left to the arbitrator. Building on this precedent, the Court emphasised thirty-eight years later in Howsam-see above-that 'procedural questions which grow out of the dispute and bear on its final disposition are presumptively not for the judge, but for an arbitrator to decide.' 40[Page614:]

Could the stubborn objector nevertheless insist that these two cases, like Maran Coal, do relate to the fundamental question of jurisdiction, in the sense that the consent to arbitration applies to timely claims and not others? How do the words 'grow out of the dispute' allow us to skip over the question whether the dispute was entrusted to the arbitrators in the first place? It would be wrong- and pointless-to deny that this argument has a defensible logic. Hard cases precisely involve a collision of logic, not the encounter of logic and illogic. 41Indeed, in Howsam the 10th Circuit Court of Appeals, very much like the Swiss Supreme Court in Maran Coal, had accepted this thesis, reasoning that the NASD rule imposed a 'substantive limit on the claims that the parties have contracted to submit to arbitration'. 42The Supreme Court's reversal is all the more emphatic, and seems more consonant with legitimate expectations, as well as with a policy of avoiding a multiplicity of proceedings. 43

In his ground-breaking and comprehensive analysis of whether judges or arbitrators should make the ultimate decision when it comes to challenges to arbitral authority, Professor Rau suggests the following focal point. Since the fundamental question is whether the parties have consented to arbitral authority, he reasons, we should not rely on labels or metaphors, 44but rather enquire whether in a given case the parties should reasonably be considered to have intended that contentions regarding any particular issue, including threshold problems which might preclude consideration of the merits, should be decided conclusively by the arbitrators. In his development of this idea, Rau quotes a US court which asked whether the challenge was 'relevant to the nature of the forum in which the complaint will be heard'. 45[Page615:]

There is promise in the notion of 'relevance to the nature of the forum'. It enables us to see that the nub of the classification problem is whether the success of the objection necessarily negates consent to the forum. Our lodestar takes the form of a question: is the objecting party taking aim at the tribunal or at the claim? Was the thirty-day deadline in Maran Coal intended to be a limitation on the tribunal or on the claim? The answer seems clear if one puts the question thus: in the event the thirty-day limitation was exceeded, was it the parties' intention that the relevant claim should no longer be arbitrated by ICC arbitration but rather in some other forum, or was it that the claim could no longer be raised at all? Opting for the former conclusion would mean that the objection is jurisdictional, but it is hard to imagine that there would be many adherents of such a thesis. The purpose of the limitation was clearly to ensure that disputes would not linger. No reasonable purpose would be served by stipulating that cases brought within thirty days are somehow suitable for ICC arbitration, but others should not be subject of any stipulation at all-i.e. exposed to the vagaries of international conflicts of jurisdiction.

Following this lodestar will make it easy to classify objections in many cases, and should make it easier in all. Timeliness issues, or conditions precedent such as participating in a conciliation attempt, pose no problem. The same goes for contentions of extinctive prescription; waiver of claims; mootness; or absence of a legal dispute or of an indispensable third party. There is even less difficulty with issues of ripeness à la SGS v. Philippines. (The exhaustion of remedies, on the other hand, goes beyond mere ripeness; the ICJ's treatment of exhaustion in Ambatielos as a matter of admissibility is subject to doubt, except in the particular area of denial of justice. 46) As for challenges to locus standi, the answer would depend on whether the issue was germane to the scope of jurisdiction contemplated in the relevant international instrument (e.g. the definition of an 'investor' in a BIT). 47[Page616:]

Conclusion

To understand whether a challenge pertains to jurisdiction or admissibility, one should imagine that it succeeds:

- If the reason for such an outcome would be that the claim could not be brought to the particular forum seized, the issue is ordinarily one of jurisdiction and subject to further recourse. 48

- If the reason would be that the claim should not be heard at all (or at least not yet), the issue is ordinarily one of admissibility and the tribunal's decision is final.

Specific applications of this concept may be subject to discussion. Yet the aim promoted by the approach suggested above should provide a focus for such debates. Once it is established that the parties have consented to the jurisdiction of a particular tribunal, there is a powerful policy reason-given the multiplicity of fora which might otherwise come into play internationally, with hugely different practical outcomes-to recognise its authority to dispose conclusively of other threshold issues. Those are matters of admissibility: alleged impediments to consideration of the merits of the dispute which do not put into question the investiture of the tribunal as such.

This essay does not treat private and public international law as separate worlds. It deals with a generic problem without troubling itself with abstractions which are rapidly becoming obsolete, particularly in the field of investment arbitration under international treaties. Like Robert Briner throughout his career, the international legal process confronts issues that require an understanding of national law, but are ever less confined to it. [Page617:]

1
Transport- en Handelsmaatschappij 'Vekoma' B.V. v. Maran Coal Corporation, judgment of 17 August 1995, ASA Bulletin 1996, 673; Revue suisse de droit international et de droit européen 1996, 573.

2
P.D. Friedland, 'The Swiss Supreme Court Sets Aside an ICC Award' (1996) 13:1 J. Int. Arb. 111; W.W. Park, 'Determining Arbitral Jurisdiction: Allocation of Tasks Between Courts and Arbitrators' (1997) 8 The American Review of International Arbitration 133. See also P.A. Karrer & C. Kälin-Nauer, 'Is there a Favor Iurisdictionis Arbitri? Standards of Review of Arbitral Jurisdiction Decisions in Switzerland' (1996) 13:3 J. Int. Arb. 31.

3
In his comprehensive work on The Law and Practice of the International Court of Justice 1920-1996 (1997), Shabtai Rosenne writes, at p. 883: 'Neither the case-law nor the writings of publicists display any certainty or unanimity over the categorization of preliminary objections.'

4
In the Right of passage case (Preliminary Objection), I.C.J. Reports 1957, this resulted in a remarkable cascade of six sequential decisions on objections.

5
I am indebted to this memorable passage from Methanex Corporation v. United States of America, Partial Award on Jurisdiction and Admissibility, 7 August 2002, 7 ICSID Reports 239 at 271 (para. 139): 'it is perhaps not easy to define the exact dividing line, just as it is not easy in twilight to see the divide between night and day. Nonetheless, whilst the exact line may remain undrawn, it should still be possible to determine on which side of the divide a particular claim must lie.' The issue there was the meaning of the words 'relating to' in NAFTA Article 1101(1), which establishes that arbitrators may adjudicate claims with respect to 'measures adopted or maintained by [a host State] relating to . . . investments . . .'.

6
G. Fitzmaurice, The Law and Procedure of the International Court of Justice, vol. 2 (1986) at 439. Fitzmaurice's attempt to distinguish jurisdiction and competence is less impressive. He insisted that the former relates to the general 'field' of authority, while the second concerns its pertinence to particular cases. He concluded that there was thus a 'double requirement' which in turn would be multiplied by two or three (ratione materiae, ratione personae and ratione temporis), ibid at 435. This seems to be sterile categorisation for its own sake. The concept of authority can be articulated as a single requirement: that a claim fall within the scope of the tribunal's legal power to decide. All aspects of that requirement pertain to jurisdiction. Fitzgerald's struggle to seek a special place for 'competence' seems vain. Things do not become clearer when we consider this pronouncement: 'Jurisdiction relates to the capacity of the Court to decide a concrete case with final and binding force. Competence, on the other hand, is more subjective. It includes both jurisdiction and the element of the propriety of the Court's exercising its jurisdiction in the circumstances of the concrete case.' (S. Rosenne, supra note 3 at 536) The words 'subjective' and 'propriety' cannot fail to provoke misgivings. Surely questions of procedure should find a firmer ground. And what Rosenne is actually telling us is that competence is jurisdiction, only bloated with 'elements of propriety' to be evaluated on a case-by-case basis by an ephemeral majority of judges. We really must do better.

7
Fitzmaurice did not discuss 'procedural' admissibility. He likely took it for granted-as he should have-that procedural questions are by necessity to be decided by the relevant forum and its rules.

8
S. Rosenne, The International Court of Justice (1957) at 249.

9
I.C.J. Reports 1952, 23.

10
Reports of International Arbitral Awards 1956, 83.

11
Société Générale de Surveillance v. Republic of the Philippines, ICSID Case ARB/02/6, 29 January 2004, 8 ICSID Reports 518.

12
'It is, to say the least, doubtful that a private party can by contract waive rights or dispense with the performance of obligations imposed on the States parties to those treaties under international law.' (para. 154).

13
At para. 153, citing Vivendi (2002), 6 ICSID Reports 340 at 366 (para. 98).

14
Para. 154.

15
Supra note 5 at 279 (para. 172).

16
Ibid. at 262 (para. 109).

17
Enron Corp. and Ponderosa Assets v. Argentina, ICSID case ARB/01/3, 14 January 2004, <www.asil.org/ilib/Enron.pdf>, at para. 33.

18
Professor Schreuer writes that under the ICSID Convention's terminology the word 'jurisdiction' refers to the requirements of Article 25. This is not a definition, nor can it be; Article 25 contains no explanation of ratione materiae or ratione temporis. Such matters are left to be resolved by reference to the scope of 'consent' as given in particular cases. Schreuer goes on to assert mysteriously that 'competence' concerns 'the narrower issues confronting a specific tribunal, such as its proper composition or lis pendens'. He does not explore the matter further, but throws up his hands: 'The two terms are often used interchangeably.' C.H. Schreuer, The ICSID Convention: A Commentary (Cambridge: Cambridge University Press, 2001) at 538.

19
See supra note 6.

20
At p. 105. This study does not concern arbitrability in what should be the proper sense of the word, i.e. whether a matter may lawfully be resolved by arbitration. If a matter is not arbitrable, it does not help that the arbitrators unquestionably have jurisdiction-and it therefore does not matter whether the arbitrability problem is categorised as one of jurisdiction or admissibility. If a matter is not arbitrable, an award is not saved by the arbitrators' finding that the arbitrability issue was raised too late, even though the lateness of other types of objections might be non-reviewable issues of admissibility.

21
Ph. Fouchard, E. Gaillard, B. Goldman†, Traité de l'arbitrage commercial international (Paris: Litec, 1996) at 326.

22
As to usage in the most widely used Latin language, see p. 97 of F. Gonzáles de Cossio's comprehensive treatise Arbitrage (Mexico: Editorial Porrúa, 2004): 'El texto "un asunto que pueda ser resuelto por arbitraje" del artículo II(1) de la Convención de Nueva York implica que existen controversias que no puedan ser resueltas por arbitraje. Es decir, non son arbitrables.' ('The phrase "a subject matter capable of settlement by arbitration" in Article II(1) of the New York Convention implies that there exist disputes which may not be resolved by arbitration. In other words, they are non-arbitrable.') (Emphasis added.)

23
H.M. Holtzmann & J.E. Neuhaus, A Guide To The UNCITRAL Model Law On International Commercial Arbitration (Deventer: Kluwer Law & Taxation, 1989) at 38-39.

24
First Options of Chicago Inc. v. Kaplan, 514 U.S. 938 (1995).

25
W.W. Park, 'The Arbitrability Dicta in First Options v. Kaplan: What Sort of Kompetenz-Kompetenz Has Crossed the Atlantic?' (1996) 12 Arbitration International 137.

26
Supra note 24 at 943.

27
K.R. Pierce,'Down the Rabbithole: Who Decides What's Arbitrable?' (2004) 21 J. Int. Arb. 289 at 290. (The AAA Commercial Arbitration Rules, amended explicitly in light of First Options, use a similar formulation, in Rule 7(a), intended to save the parties ink-or perhaps mental energy.)

28
537 U.S. 79 (2002).

29
Ibid. at 86.

30
123 S. Ct. 2402 (2003).

31
Cass. Ch. mixte, 14 February 2003, Poiré v. Tripier et al, J.C.P. La Semaine Juridique Entreprise et Affaires 810 (8 May 2003).

32
[2000] EWHC 2059.

33
(2003) 14:1 ICC ICArb. Bull. 71. See also E. Jolivet, 'Chronique de jurisprudence arbitrale de la Chambre de commerce internationale (CCI): arbitrage CCI et procédure ADR', Gaz. Pal., 16/17 November 2001, p. 1783.

34
See Enron & Ponderosa v. Argentina, supra note 17; CMS Gas Transmission Company v. Argentina (2003), ICSID case ARB/01/8, 7 ICSID Reports 492, <www.asil.org/ilib/cms-argentina.pdf>; GAMI v. Mexico (2004), <http://ita.law.uvic.ca/documents/Gami.pdf>.

35
(Second Phase), I.C.J. Reports 1966, 6.

36
Timeliness issues are unrelated to jurisdiction ratione temporis, which limits the scope of the tribunal's authority to disputes having their origin in-or after or before-a particular time period. See Maffezini v. Spain, ICSID case ARB/97/7, 25 January 2000, 5 ICSID Reports 396, <http://ita.law.uvic.ca/ documents/Maffezini-Jurisdiction-English_000.pdf> and Lucchetti v. Peru, ICSID case ARB/03/4, 7 February 2005, <www.worldbank.org/icsid/cases/lucchetti-award.pdf>.

37
A.S. Rau, '"The Arbitrability Question Itself"' (1999) 10 The American Review of International Arbitration 287 at 322.

38
John Wiley & Sons, Inc. v. Livingston, 376 U.S. 543 (1964).

39
Supra note 28.

40
Ibid. at 84.

41
This difficulty, Cardozo wrote in his famous monograph, is a fundamental challenge of legal decision-making: 'One principle or precedent, pushed to the limit of its logic, may point to one conclusion; another principle or precedent, followed with like logic, may point with equal certainty to another.' B.N. Cardozo, The Nature of the Judicial Process (1921) 40.

42
261 F.3rd 956 at 965.

43
In Cardozo's paradigm case: 'in the end, the principle that was thought to be most fundamental, to represent the larger and deeper social interests, put its competitors to flight' (supra note 41 at 42).

44
'Severability' is but a metaphor, 'voidness' but a label , he argues forcefully, and neither should be a substitute for analysis of relevant, possibly countervailing considerations; A.S. Rau, 'Everything You Really Need to Know About "Separability" in Seventeen Simple Propositions' (2003) 14 The American Review of International Arbitration 1.

45
Ibid at 70, quoting Great Western Mortgage Corp. v. Peacock, 110 F. 3d 222 (3d Cir. 1997), where a party challenged the validity of a contractual exclusion of punitive damages; the Circuit Court held that that issue-'unrelated to the question of forum'-should be decided by the arbitrators. In Military and Paramilitary Activities In and Against Nicaragua, the United States argued (unsuccessfully) that the ICJ was not entitled to hear the case because it came under the exclusive authority of the Security Council. This objection was wrongly characterised as one of admissibility; Nicaragua v. US, I.C.J. Reports 1984, 392.

46
See J. Paulsson, Denial of Justice in International Law (Cambridge: Cambridge University Press, 2005), c. 5. In the absence of exhaustion of local remedies, the underlying claim may still be perfectly ripe and appropriate for adjudication, but not before a particular forum that requires exhaustion as a precondition of access to it. The problem is one of jurisdiction. But denial of justice in international law contains a substantive requirement of exhaustion of local remedies, because the delict is not consummated unless it is shown that the national legal system was given a reasonable opportunity to correct it. A failure to exhaust is therefore ordinarily fatal to the claim no matter where it might be heard.

47
For similar reasons the otherwise useful discussion of 'Admissibility' in J. Collier & V. Lowe, 'The Settlement of Disputes in International Law' (Oxford: Oxford University Press, 1999) at 155ff., should be considered with some care because it focuses squarely on the ICJ. In cases where the State parties have given broad consent to ICJ jurisdiction, arguments about nationality or exhaustion of local remedies may be classified as matters of admissibility because the Court's authority in a general sense remains. If an ephemeral arbitral tribunal is established under a treaty which contains requirements as to the nationality of private claimants, or as to their prior exhaustion of local remedies, the claims as such are perhaps subject to no impediment but the forum seized is lacking one of the elements required to give it life in the first place. For such a tribunal these are matters of jurisdiction.

48
One type of exception is posited by Professor Rau in his remarkable essay, to wit: it may be permissible to conclude that contracting parties intend even such issues to be decided conclusively by arbitrators; supra note 44.