Marc J. Goldstein Arbitrator & Mediator NYC
November 02, 2023

Gotham Novit Curia

If you insist that international arbitrators sitting in an environmentally common law case cannot properly look at or consider any legal authorities other than those submitted by the parties, stop reading here. I am here to preach to the convertible.

Let’s begin this discussion with a nod toward a legal doctrine that has essentially no shelf space in US arbitration law or US practice of domestic or international commercial arbitration: Jura Novit Curia (“JNC”), or in its arbitral variant, Jura Novit Arbiter (“JNA”).  We start there because among those who arbitrate in a procedural culture called the adversary system, which accounts for probably the vast majority of international arbitration whether contract-based or treaty-based, we are well-advised to have in our grasp the historical and theoretical foundation for the arbitrator to take any initiative to figure out the content of the law she is charged with applying. And by “initiative” we do not mean opening and reading the authorities submitted by the Parties. We mean anything other than that, including what you simply remember.

One contemporary commentator purports to credit Aristotle as a pioneering advocate for  JNC, and quotes the Great Greek from Stagira as having said at least once:

[A] litigant has clearly nothing to do but to show that the alleged fact is so or is not so, that it has or has not happened. As to whether a thing is important or unimportant, just or unjust, the judge must surely refuse to take his instructions from the litigants: he must decide for himself all such points as the law-giver has not already defined for him.

[as quoted in Fried Rosenfeld, “Iura Novit Curia in International Law,” European Int’l Arb. Rev.  Vol. 6, Issue 1 (2017)].

Granted this view of the judge as oracle predates by well over a millennium the emergence of an adversary system based on advocacy, never mind the refinement of the adversary system into the party-created body we now know as the arbitral tribunal. But even if it there is only negligible support for an arbitrator-as-oracle model in contemporary international arbitration, does that imply that there is no space for oracular behavior within the adversarial framework of arbitration? Do we, as avowed non-oracles, simply keep our fingers off the keyboard and our noses out of the legal databases where authorities the parties did not furnish to us might be found? Do we, moreover, purge our minds of whatever legal knowledge of arbitration law and applicable substantive law that (presumably) qualified us in the eyes of the parties to be selected as arbitrators, and approach each case as voracious neophytes craving legal education from the parties?

To answer this question, it is probably helpful to have a working knowledge of what JNA is, and what it isn’t. You can spend a good portion of your remaining lifetime reading volumes that have been written on the subject. Or trust your Commentator’s reductionist version: It is the principle that an arbitral tribunal may — consistent of course with due process — decide a case on the merits by furnishing legal reasoning for the outcome that has been fashioned by the Tribunal rather than by any of the parties.  And the due process right, in reductionist terms, is the right of the parties to learn of the proposed Tribunal reasoning from the Tribunal, and to endorse or deplore it, and to endorse or deplore the very fashioning of it (that is to say, the adoption of JNA methodology) before the Final Award is issued.

My source for these reductive nuggets on JNA? — The writing of a respected comparative law scholar in her chapter of a recent major academic compilation on JNA. To quote Professor Guiditta Cordero-Moss:

It is widely recognized that the will of the parties plays a determinant role in international commercial arbitration. The aim of this book is to investigate to what extent the arbitral tribunal may nevertheless develop its own legal reasoning. An independent legal reasoning will not necessarily be based on the will of the parties as manifested in the terms of the contract, in the law chosen in the contract or in the legal arguments presented by the parties.

[G. Cordero-Moss, General Report on Jura Novit Arbiter, in F. Ferrari and G. Cordero-Moss eds., Iura Novit Curia in International Arbitration (Juris 2018), Ch. 17 at p. 463].

Another contemporary commentator, possibly channeling his inner Aristotle, also suggests that the quest for the right legal reasoning to resolve the merits is at the core of JNA:

“Knowledge of law” must be understood in a manner related to the specific case. The arbitrator is not obliged to know all the laws, but his mission is to use the legal tools (within the rules of the game) to settle the dispute in accordance with justice. Is it not precisely for this reason that they were appointed by the parties? It would be absurd to say that a tribunal exercises its powers to settle the dispute only on the basis of what has been said by the parties.

If within their analysis of the laws the arbitrators identify some rule or principle that has been ignored by the parties (involuntarily or intentionally) and that could be crucial for the decision of the case, should the tribunal simply ignore said rule or principle because the parties did not mention it? The task of the tribunal to submit an award strictly according to law is not bound by what has been said by the parties.

[C. Collantes, Give Me The Facts and I’ll Give You the Law: What Are the Limits of the Iura Novit Arbiter Principle in International Arbitration?, Kluwer Arbitration Blog, Jan. 10, 2019].

If the core content of JNA is the power of the arbitrator to fashion the legal reasoning that resolves the merits, then the Americanized version of the JNA question — roughly, “can we as arbitrators ever do our own legal research?” — asks a question that is not the same as the JNA question, and while that question may intersect with JNA it is very far from a concentric overlap with JNA  [See, e.g., P.B. Marrow, Can an Arbitrator Conduct Independent Legal Research? If Not, Why Not? (May 6, 2013). Available at SSRN: or]

The troubling implication of the JNA doctrine’s core principle is that it is possible, not likely but certainly possible, that an arbitral tribunal could see its way to reject a legal outcome based on the submissions of either party, determine the outcome based on entirely on legal theories and authorities identified by the Tribunal, and salvage its award from annulment by sharing those theories and authorities with the parties for them to provide comments in advance of the award. In that case, so much for party autonomy beyond picking the panel. That JNA core principle, however rare might be its implementation by an arbitral tribunal in actual practice, is rejected by legal cultures based on the adversary system, which is to say most common law legal cultures.

The difficulty JNA presents for arbitrators sitting in US seated international cases or foreign seated cases governed by US substantive law is what guidance for our conduct do we take from our rudimentary understanding that the adversarial system rejects JNA (or, stated differently, that we are presiding in an adversarial, not an inquisitorial process)? In the perennial unresolved debate in our community about the proper role if any for arbitrators in identifying legal authorities other than those presented by the parties, the understanding that we reject JNA divides us into more or less two camps: the absolute rejectionists, who see no role for any legal inquiry by the Tribunal beyond the submitted authorities of the parties, and the moderate rejectionists who permit that such legal inquiry may be made but in narrow windows whose opening does not fundamentally alter the party-driven character of the overall process.

Consider this Commentator to be a moderate rejectionist, and let’s look at the content of moderate rejectionism.  Several practical examples can lead the way.

EXAMPLE NO. 1. In a New York-seated arbitration, the contract provides that New York law governs and that mediation shall be completed before arbitration is commenced. The Federal Arbitration Act is clearly applicable. Even before the Tribunal is fully constituted, Claimant files (with the provider organization) a motion to dismiss Respondents’ counterclaim, arguing that there is no arbitral jurisdiction over the counterclaim until a mediation of the counterclaim is completed. The Claimant cites only New York state court non-FAA cases concerning when mediation is treated by a state court a condition precedent to its subject matter jurisdiction. Respondent responds, also citing only New York State non-FAA cases, and argues that mediation is not a condition precedent to jurisdiction under that case law as properly understood and as applied to the present facts.

It would be apparent to many arbitrators that case law under the FAA treats an agreement for pre-arbitration mediation as a procedural condition whose fulfillment is a procedural aspect of the arbitration and  not  a condition precedent to the parties’ consent to arbitration. (BG Group, anyone? Maybe Howsam?) The presiding arbitrator taps into her preferred legal research tool, retrieves and reads the leading cases, shares the cases with counsel, and invites written comments on how they bear (if at all) on the jurisdiction motion.

Fair or foul?

Has the Tribunal exceeded its powers, or otherwise overstepped a procedural norm of arbitration? The parties remain free to instruct the Tribunal that they are in agreement to have the obligation to mediate be enforced as it would be enforced in a non-arbitrable case in the New York courts. If they do not have such an agreement, they are free to comment upon and argue for or against the treatment of mediation indicated by the federal cases. Whether the Tribunal has impermissibly intervened in favor of the party opposing the jurisdiction objection is more complicated, but only slightly so. The Tribunal’s powers are defined by the contract, and the contract by necessary implication (because it involves interstate and international commerce and specifies a New York seat) calls for the Tribunal to apply federal and New York arbitral procedural law. The Tribunal would exceed its powers by applying New York judicial civil procedure law to a question of arbitral jurisdiction, unless the parties, modifying their written agreement, now agree that the Tribunal should do so.

Has the Tribunal committed a due process violation by saving one party from the outcome that might have resulted from proper application of the wrong body of law? I think not: The Tribunal was motivated to apply the right content of the agreed applicable law,  not to help one party win.  At least one commentator agrees, so I am not alone;

[A]n arbitrator should not be deterred from seeking the information he needs to correctly resolve a case just because that information may weigh in favour of one of the parties’ positions. In requesting such information, the arbitrator is not seeking to help one party over another. He’s trying to reach the right result. The right result will necessarily go in favour of one party and against the other, but that fact alone does not convert an arbitrator’s request for further information into a due process violation.

[J. Kirby, How Far Should an Arbitrator Go to Get it Right?, in The Powers and Duties of an Arbitrator, Liber Amicorum Pierre A. Karrer, Ch. 20 p. 193 at 197 (Kluwer 2017)].

I should add that I do not read the quoted commentator  as limiting “information” to evidence and excluding information about the content of the law. Nor do I read the position as limiting the request to a request to the parties and not to an online legal library at rue arbitrator’s disposal. The transparency-due process issue needs to be separated from the question of the legitimacy of the arbitrator’s quest for the content of the law.

EXAMPLE NO. 2. The Tribunal in a partial final award has decided that it will impose a sanction by entering a contemporaneous procedural order, and specifies in the PFA that relief from the sanction may be obtained only upon the fulfillment of certain conditions. The entire subject matter of this PFA is what if any sanction should be imposed and what should be the conditions for relief from the sanction.  No merits claims are resolved in this PFA. The affected party applies for relief from the sanction, admits that only some of  the PFA conditions for relief have been met, but advances no legal arguments about how the functus officio doctrine (or the limits in the agreed arbitration rules on modifying an award) impacts the Tribunal’s power to grant the requested relief. The presiding arbitrator taps into her preferred legal research tool, refreshes her recollection on the leading cases defining the contours of the functus officio doctrine and its exceptions, and with agreement of the co-arbitrators after sharing those cases, invites the applicant to submit written comments on how the making of an award bears on the Tribunal’s power to grant the requested relief under the functus officio doctrine or other limits on Tribunal powers to modify an award.

Has the Tribunal exceeded its powers or otherwise encroached on a procedural norm by “doing its own research”? I put the term in quotes here, because what the Tribunal has in fact done is not to embark on independent research for a Tribunal-motivated legal solution that the movant has not addressed, but has simply resorted to its recollection about a well-established universe of case law concerning a legal doctrine affecting the power of the Tribunal that obviously ought to bear on the decision. It seems unlikely that the functus officio doctrine is inapplicable just because the parties fail to mention it, although they might agree to dispense with it.

In this instance the Tribunal elects not to present specific cases for comment because relevant principles might be found in various cases and the applicant, having the burden of persuasion, does not deserve to have the Tribunal conduct legal research for (or against) her. Depending on what is received from both parties, there would be another opportunity if the briefing is inadequate to ask both sides to comment on specific cases that the Tribunal might be inclined to cite in a ruling. This is not to suggest, however, that a Tribunal must always revert to the Parties for comment on specific cases if it wants to cite a case that neither party has cited. One consideration is that if the decision involved is a procedural matter, an aggrieved party may seek to have the order vacated or modified by the Tribunal and may argue that the case law the Tribunal cites is inapposite.  Another consideration is that if the Tribunal’s question seeking comments is adequately phrased, the case law the Tribunal cites is unlikely to be obscure and reasonably ought to be found in competent research by the parties. (Unpublished decisions known to Tribunal members for unique reasons may be viewed differently).

EXAMPLE NO. 3. Suppose one party has sought issuance of an arbitral subpoena that would require a witness in California to testify via Zoom in a New York-seated arbitration. The other party objects to the subpoena in various respects but raises no issue about the Tribunal’s power to command Zoom testimony from a non-party. But one or more Tribunal members, tapping the databases of their own recollections in the first instance, know there is case law on both sides of the question of arbitral power to require non-party Zoom testimony under FAA Section 7. In an email to the Parties, the Tribunal mentions a few of those cases and directs the parties to comment on what they believe to be the right approach. This recourse to case law that has not been mentioned by either party has nothing to do with the merits directly — although the non-party’s evidence might well affect the outcome. Rejection of JNA in a common law environment should not mean that arbitrators must pretend to forget about an unsettled and controversial issue of arbitral power that is well known to them but possibly much less familiar to the parties.

EXAMPLE NO. 4. The fourth and final example for today’s Commentary relates to how common law court decisions are reported in certain online legal research tools: with hyper-linked citations to precedents cited (approvingly or otherwise) in the opinion. Do the arbitratorsimpermissibly “do their own research”, in violation of our Rejection-of-JNA,  by reading the court-cited precedents within the party-cited precedents to obtain a more complete understanding of the legal doctrine invoked by the party? This Moderate Rejectionist would say NO.  My first defense: this conduct does not seek a legal theory to resolve the case that neither party has mentioned. The arbitrator is seeking clarity about the contours of the legal theory invoked by a party, contours that might be clearer from reading the embedded precedents within the party-cited case. My second defense: If a legal theory is advanced by a party, the Tribunal should apply that legal theory (or principle) as it exists in the law, not only as it exists in the renditions of the law advanced by the Parties. The prescriptive norms of arbitration as an adversarial process will lead the Tribunal to give procedural primacy to the submissions of the Parties. But a quest for correctness in the content of the law, if the Parties’ renditions are not fully satisfactory, should not to be conflated with adoption of a Tribunal-conceived legal theory. This is not the importation of JNA into common law arbitration.

If I am an outlier in believing the arbitrator has a power of inquiry, and that the fairness issue is separate and is addressed by transparency to the parties about the results of the inquiry, then at least I have some good company. As the International Law Association stated in its 2008 Report on Ascertaining the Contents of the Applicable Law in International Commercial Arbitration:

[Arbitrators] should inquire about the applicable law within the general parameters of the arbitration defined by the parties and, considering costs, time and relevance issues, may conduct their own research, provided the parties are given an opportunity to be heard on material that goes meaningfully beyond the parties’ submissions.

We often debate the implications of what provider rules say or fail to say on a given issue. And on this issue we find mostly silence. The AAA Commercial Rules are and have long been silent. The CPR Non-Administered Rules are also silent save that “[s]ubject to these Rules, the Tribunal may conduct the arbitration in such manner as it shall deem appropriate.” (Rule 9.1). This leaves the door open. The JAMS Comprehensive Rules are also silent. This is a curious institutional silence, given that there is considerable sentiment in the domestic arbitrator ranks in the USA that when it comes to the law, the panel should take what the parties give them and seek no more.

When we look to the leading international rules, our attention goes first to the LCIA, whose Rule 22.1 (iii) (in the 2020 version)  has provides expressly — and has done since at least the 1998 version of the Rules that:

The Arbitral Tribunal shall have the power, upon the application of any party or … upon its own initiative, but in either case only after giving the parties a reasonable opportunity to state their views and upon such terms (as to costs and otherwise) as the Arbitral Tribunal may decide … (iii) to conduct such enquiries as may appear to the Arbitral Tribunal to be necessary or expedient, including whether and to what extent the Arbitral Tribunal should itself take the initiative in … ascertaining … the law(s) or rules of law applicable to the Arbitration Agreement, the arbitration and the merits of the parties’ dispute.

Let’s call the LCIA Rules approach — not widely adopted as far as I can find — Ambivalent Moderate Rejection of JNA. Ambivalent in the sense that it gestures in the direction of party autonomy by requiring input from the parties to be considered (but without requiring party acceptance) before the Tribunal starts foraging. That ILA Report formulation, which came ten years after the LCIA Rule, does not require this preliminary party-consultation step. And we find that most international rules leave the door wide-open for whatever position the Tribunal may wish to take. See for example ICC Rule 21(1): “In the absence of any [party] agreement, the arbitral tribunal shall apply the rules of law which it determines to be appropriate.” The Singapore Centre Rules list among the “Additional Powers of the Tribunal” absent agreement of the parties limiting such power : “to … conduct such enquiries as may appear to the Tribunal to be necessary or expedient” (SIAC Rule 27(c)) and also “to … decide, where appropriate, any issue not expressly or impliedly raised in the submissions of a party provided such issue has been clearly brought to the notice of the other party and that other party has been given adequate opportunity to respond” (SIAC Rule 27(m)). Broadly speaking, the landscape of international arbitration rules is marked by approaches more flexible than the LCIA model, leaving the Tribunal broad discretion to conduct inquiries into the content of the applicable law without previewing with the parties the prospect that such inquiries will or may be made. The ultimate criteria universally applicable are post-inquiry: was there equal treatment of the parties and were the parties given an opportunity to be heard on any outcomes of the Tribunal research into the content of the law that might affect the outcome in a material way?

Let’s address one more objection from the JNA literature before you leave.

Some commentators object in principle to Tribunal initiatives to ascertain the law, saying that the JNC principle as applied in civil law courts, i.e. the version inherited from Aristotle, rests on the premise that judges in the courts of a state are protectors and purveyors of a law of their forum (lex fori), and that arbitrators, having no comparable position and instead serving the parties not the state, have “no forum .”  [See, e.g., P. Landolt, Arbitrators’ Initiatives to Obtain Factual and Legal Evidence, Vol. 20 No. 2 Arb. Int’l at pp. 184-85 (2012)]. Not to argue the issue exhaustively, I would only point out by way of examples: (1) if the Parties in a international case governed by New York substantive law constitute a Tribunal of two party-appointed retired New York judges and as Chair one lioness of the New York litigation bar who has also given long service as an international arbitrator, did they not intend to have a Tribunal function on New York substantive law as a New York court would?, and (2) if the Parties to an international case governed by the law of Ontario constitute a Tribunal composed of two lions of the Toronto litigation bar and one retired Justice of the Supreme Court of Canada, did they not intend to have the Tribunal function on Ontario law as an Ontario court would?

It is time for some takeaways, as you have other blogs to read and other work to do!

The commentator who writes in favor of flexible arbitral power in the search for the content of applicable law risks to be mis-branded as a zealot and a maverick, as a closeted autonome perpetually foraging for his own legal guidance in disrespect of the parties’ autonomy, not to mention the competence of their legal representatives. I hope to be acquitted of these charges. But it seems that we as a community can do better than the broad homilies that buttress polar positions: “arbitration is a party-driven process” at one extreme, and “due process is all that matters” at the other.  Perhaps we can draw some guiding principles from the examples and the discussion provided in this post.

  1. Occasions for input on the applicable law from the Tribunal do not arise only when a merits award is being written. Applicable law may come into play at any number of mileposts, including threshold objections to jurisdiction, proposed issuance of arbitral subpoenas, etc. Parties may not even be aware of applicable law affecting the proposed procedural step, as in the case of a proposed arbitral subpoena. If the Tribunal asks the Parties “what about these cases concerning our power to command a Zoom appearance?” it is not taking sides between the parties, but is seeking in good faith to exercise its powers in regard to non-parties within applicable legal limits. If the Tribunal, in the mediation-as-condition-precedent example, asks “what about the FAA case law on this subject from the US Supreme Court?” it fairly raises the question of whether the Parties’ counsel truly wish to mutually exclude application of the FAA, or rather are skilled courtroom advocates who misconceived the application New York civil procedure law in an FAA-applicable arbitration.
  2. The context for arbitrator research may be critical to how we view it. Suppose a Tribunal decides during deliberation that it is important to study the cases cited, discussed and quoted within the cases cited by the Parties, and the Tribunal proceeds to do so, then deliberates further about whether the older precedents are sufficiently material to warrant asking the parties for further comments. By doing this the Tribunal carries out, in a manner that is fair to both sides, a process that may generate a more precise rendition of the law the parties have agreed to have the Tribunal apply to decide the case, and legal theories under that law that they have specifically invoked. On the other hand, a party-appointed arbitrator who conducts research in pursuit of a legal analysis not articulated by either party, based on case law not cited by either party, but that would advantage the party that appointed her if embraced by the Tribunal, generally speaking is acting inappropriately. But the infraction is not of a rule or a proscriptive norm against arbitrator research. It is a violation of the broader obligation to behave impartially.
  3. Timing is crucial. When arbitration is in a common law environment, the parties expect the development of claims and defenses, the legal theories and authorities supporting those theories, to be primarily their counsels’ handiwork. If an arbitrator in such a case suggested at a preliminary conference that case A v. B might well be controlling on the merits, the parties would naturally be dismayed. On other hand, parties arguably do have a reasonable expectation that the Tribunal might at the post-hearing stage of deliberations and award-drafting check to see if there have been new developments in a line of authorities relied upon by the parties, or appellate reversals of authorities cited by the parties. If this were not aligned with the expectations of the parties, we might expect to see party initiatives to stipulate that the Tribunal will not, or will not without the parties consent, conduct any legal research.


4 Responses to “Gotham Novit Curia”

  1. Steven Reisberg says:

    Well done!
    Best regards

  2. admin says:

    Thank you Steven. Hope all is well.
    Best wishes,

  3. Great writing along with brilliant and interesting analysis. What else could one want on a Tuesday evening. Love your blog. Pls keep at it!! Sss

  4. admin says:

    Steven thanks so much. Hope to see you soon in person. All the best.

Leave a Reply