Leading Questions with Rishabh Gandhi 
Leading Questions

Marshalling of evidence in arbitration: What arbitrators look for while writing awards

Marshalling of evidence in arbitration is not only about helping the tribunal decide; it is also about helping the award explain itself.

Rishabh Gandhi

In this ‘Leading Questions’ section, Rishabh Gandhi explains the importance of issue‑wise marshalling of documentary and witness evidence in arbitration.

Question: What does “marshalling of evidence” mean in arbitration?

Answer: For an arbitrator, marshalling of evidence is the discipline of turning a crowded record into a sequence of findings. It is not about compiling papers. It is about organising proof. The expression is familiar to Indian adjudication, and courts, including the Supreme Court, have used it while referring to orderly evaluation of evidence before findings are recorded. In arbitration, that discipline becomes even more important because Section 19 of the Arbitration and Conciliation Act, 1996 makes it clear that the tribunal is not bound by the Code of Civil Procedure or the Indian Evidence Act, while Section 19(4) expressly empowers it to determine the admissibility, relevance, materiality and weight of evidence.

In practical terms, marshalling is what allows a record to be converted into a reasoned award. It also serves another function. It separates adjudication from impression. In many arbitrations, one party appears stronger because the record is bulkier, the pleadings are more assertive, or the correspondence is more voluminous. But adjudication does not proceed on impression. It proceeds on tested propositions. Marshalling is the step by which a tribunal moves from narrative to proof, and from proof to findings.

Question: Why does marshalling matter more in arbitration than in court litigation?

Answer: Court proceedings come with an internal discipline of their own. Pleadings, issues, stages of evidence and appellate habits impose structure. Arbitration often does not. In most commercial arbitrations, the problem is rarely lack of material. It is excess of material. Contracts, addenda, emails, minutes, site records, account statements, technical reports, expert material. Once all of that is put on file without issue-wise organisation, the tribunal is left to reconstruct the case for itself. That affects the award directly. Section 31(3) requires reasons, save in limited cases.[2] Reasons do not emerge from bulk. They emerge from structure. A well-marshalled record helps the tribunal do what it is expected to do: identify the issue, locate the evidentiary support, weigh it against the contrary material, and record a defensible finding. This is also where arbitration differs from ordinary court habit in a practical sense. In court litigation, procedural stages themselves often force some discipline over time. In arbitration, especially in document-heavy commercial matters, a badly structured case can remain badly structured all the way until final arguments. By then, the difficulty is no longer merely presentational. It affects comprehension, time spent by the tribunal, and ultimately the quality of the reasoning reflected in the award.

Question: What is the evidentiary value of documents in arbitration? Do they prove themselves?

Answer: Not necessarily. A document does not prove the case merely because it is filed, marked, or placed in a compilation. Its evidentiary worth depends on more basic questions. Who authored it? Who signed it? Is execution admitted or disputed? Is the document relied on for the truth of its contents, or only to show that a communication was made? Does any witness speak to its origin, preparation, circulation, or use? Does it fit the contractual and factual sequence in which it is being invoked?

That is where mode of proof matters. One document may go to authorship. Another may go to execution. A signature may establish endorsement, but not necessarily the truth of every statement contained in the document. The contents may still require explanation, admission, surrounding correspondence, or witness linkage. Section 19 does not dilute this exercise. It gives the tribunal control over admissibility and weight. The tribunal still has to decide what the document proves, and just as importantly, what it does not prove.

From the standpoint of writing an award, the real question is always the same: what proposition does this document establish? A few contemporaneous and properly connected documents usually do more work than several volumes of unstructured paper.

The law on interference with arbitral awards also indirectly reinforces this point. Decisions such as ONGC Ltd. v. Saw Pipes Ltd. and later the more restrained formulation in Associate Builders v. Delhi Development Authority and Ssangyong Engineering & Construction Co. Ltd. v. NHAI show that arbitral findings must remain anchored in the contract and the evidentiary record. The tribunal is not expected to mechanically reproduce documents. It is expected to evaluate them, connect them, and explain why they matter.

Question: How should evidence not be marshalled in arbitration?

Answer: Evidence is often lost not because it is weak, but because it is unstructured. The mistakes are familiar. Large compilations are filed without linking documents to issues. Witness statements repeat pleadings instead of proving facts. Adverse documents are skirted around as if the tribunal will somehow miss them. The record becomes a warehouse: everything stored, nothing traced.

Experience on the Bench at the trial stage shows that the real difficulty is rarely lack of evidence, but lack of structure in how it is presented. In one arbitration, despite filings running into several volumes, a simple question from the tribunal “Which document proves this?” was met with hesitation and cross-referencing across compilations. The difficulty was not absence of material. It was absence of structure. The claim may have been sustainable, but it weakened because the evidentiary path was unclear. That moment is more common than lawyers like to admit. If the answer to that question cannot be given directly, the case has already begun to lose force.

There is another recurring mistake. Parties often marshal only the material that supports their case and leave the inconvenient record unaddressed. That is rarely effective. A tribunal writing an award has to deal not only with what supports a claim, but also with what undermines it. Ignored adverse material has a way of reappearing at the worst possible stage — usually when the tribunal is testing whether a finding can genuinely be sustained.

Question: How should evidence be marshalled so that the award survives scrutiny?

Answer: The safest method is issue-wise marshalling. For each issue, the tribunal should be able to see four things: the fact asserted, the evidence relied upon, the evidence pointing the other way, and why one version is being preferred. That is what makes reasons possible. It is also what makes challenge harder.

At the Section 34 stage, courts do not sit in appeal over facts and do not set aside awards merely by reappreciating evidence. But that is also where vulnerability arises. In Ssangyong Engineering & Construction Co. Ltd. v. NHAI, the Supreme Court explained that findings based on no evidence, or findings rendered by ignoring vital evidence, may attract interference under the patent illegality standard in domestic awards. That logic was reinforced in Delhi Metro Rail Corporation Ltd. v. Delhi Airport Metro Express Pvt. Ltd. The lesson is simple. An arbitral award is not a catalogue of exhibits. It is a chain of reasoning. Every important finding must trace back to an identifiable evidentiary foundation. That is what an arbitrator looks for while writing the award. And that is what a court looks for when the award is challenged.

A useful way to think about it is this: marshalling is not only about helping the tribunal decide; it is also about helping the award explain itself. A reasoned award need not resemble a court judgment in length, but it must still show that the tribunal has seen the real dispute, identified the decisive material, and arrived at a conclusion through a rational evidentiary path. That is usually the difference between an award that survives challenge and one that invites it.

Rishabh Gandhi is a former judge, arbitrator, and arbitration counsel specialising in commercial and construction disputes. Gandhi is also the founder of Rishabh Gandhi and Advocates.

Judicial Appointments Tracker: March 2026

Israel-Iran war: Plea before Bombay HC to bring back mortal remains of Indian seafarer killed off Oman coast

Mukesh Kumar Choudhary joins Unity Small Finance Bank as National Head Legal

Why India must pass the pending Facial Recognition Technology Bill

TN Elections: PIL in Madras HC to bar candidates not professing Hinduism, Sikhism or Buddhism from SC seats

SCROLL FOR NEXT