ICC report on the Accuracy of Fact Witness Memory in International Arbitration


In November 2020, the ICC’s Task Force on Maximising the Probative Value of Witness Evidence (the “Task Force”) published a report on the Accuracy of Fact Witness Memory in International Arbitration (the “Report”). The Report considers the science behind memory and whether modifications can be made to arbitral practice to enhance the probative value of fact witness evidence in international arbitration. This Law-Now provides a summary of the Report and considers its implications for the international arbitration community. 

Why is this Report important for arbitration practitioners?

Witness of fact evidence is a common feature of international arbitration and is often a key focus of the merits hearing.  The quality, and persuasiveness of that evidence is considered by some practitioners to be critical to the outcomes in proceedings.  Unsurprisingly, the preparation of written witness statements is time consuming, as is the exercise of preparing to examine (or cross-examine) those witnesses at the merits hearing. The Tribunal too will commit time reading the written witness statements in in advance of the hearing, so it can understand the evidence given in cross-examination and seek clarifications if necessary.  The time, therefore, committed to the preparation and examination of witness evidence is a significant factor in the overall cost of some arbitral proceedings.

As the Report itself notes, scientific studies and human experience show that witness memory is imperfect and is subject to distorting influences as soon as it is formed. The Report therefore considers whether the “best practices” adopted by arbitration practitioners do in fact produce witness evidence that has the expected probative value, and whether that justifies the time and expense that goes into this process.

What are the findings of the Report?

Reviewing existing research, and carrying out its own study, the Task Force found that memory distortion can occur in a commercial setting typical of arbitration proceedings. In particular, biasing a witness in favour of a particular party and exposing them to suggestive “post event” information affected their memory, leading witnesses to recall details in a way that better suited their own party’s case.

In summary, the Task Force found that:

  • The memory of an honest witness who gives evidence in international arbitration can become distorted, leading to the witness being less reliable than the parties (including the witness) expect;

  • There are steps that can be taken to reduce the distortion of witness memory and to better assess the weight that ought to be given to the evidence in light of such distortions; and

  • Witness evidence may be presented in arbitral proceedings for a number of different purposes, many of which do not rely upon the accuracy of witness evidence (for example, witnesses can be used to give evidence about the general background or to the “set the scene” in which the dispute arose, which may not concern disputed facts, and therefore will not necessarily be material to the tribunal’s determination of the issues in dispute).

As regards measures that may assist in preserving the accuracy of witness, the Report provides guidance in two sections:

(A)- measures that can be taken to reduce distorting influences and their effect including:

  • In-house counsel should establish procedures for teams to keep contemporaneous notes of issues being discussed, avoid setting a “party line”, and involve external counsel early in the process; and

  • Outside counsel should conduct witness interviews at the earliest opportunity when memory is likely more accurate, use questions which are unbiased and open-ended, avoid contaminating the witness’ memory with other documents such as party submissions, and ensure the first draft witness statement is completed in the witness’ first language.


(B) - steps that allow the parties and tribunal to identify and weigh the distorting influences that might exist, including:

  • Parties should educate themselves to better understand the workings of human memory, and understand that misleading information received after an event can interfere with or impair a witness’ memory of that event; and

  • The Tribunal might consider requiring each witness statement to include information about the way it was prepared, remind the witness that it is permissible to answer “I don’t recall”, or exclude witnesses from the hearing room until after they have given their evidence.

Conclusions and comment

The science of memory is a topic of considerable complexity, which has only recently started to influence the practice of evidence collection and presentation.  The scientific basis for the risk of evidence distortion is unlikely to be controversial, but the way in which arbitral institutions and arbitrators in any particular case approach that risk may be.

It is also likely that more effort will now be put into attempting to undermine witness evidence from a scientific perspective as well as challenging witnesses on disputed facts.  For the time being, however, factual witness evidence is unlikely to lose its status in international arbitration, particularly given its key role in merits hearings. It is noteworthy that the Task Force found that generally responders to their questionnaire (circulated to members of the ICC Commission on Arbitration and ADR) with a civil law background believed that fact witness testimony would benefit from being limited to facts that are disputed and relevant to the outcome of the arbitration, while those with a common law background tended to accept a wider approach to fact witness testimony as useful.  This suggests that the standing of witness evidence is far from settled and the Task Force’s report may be only the beginning of a new phase in dispute resolution exploring the quality of, and alternatives to, witness evidence.

The Report concludes that while the accuracy of witness evidence is undoubtedly important, there are steps that can be taken to reduce distortion, but this must be balanced against the usefulness of current practices. For example, although the preparation of witness statements by counsel can have a distorting effect on the memory of the witness, there is a significant benefit in so doing in helping a lay witness communicate with the Tribunal.

While the Report does not suggest fundamental changes to the way that evidence is collected and presented, it shows that an increased cognisance of the malleable nature of human memory, and taking steps to mitigate that risk within the framework of current practice, may well lead to more accurate factual witness evidence.

It is not only in the context of international arbitration that the accuracy, and indeed usefulness of, factual witness evidence is under examination. The Business and Property Courts of England & Wales have recently issued a new Practice Direction regarding witness evidence at trial.  The Practice Direction follows the work of the Witness Evidence Working Group which published its own report and recommendations in late 2020.  The Practice Direction will be applicable from 6 April 2021 and will require additional steps to be taken in the preparation of witness evidence for use at trial, a number of which are influenced by emerging concerns around memory distortion.

References: ICC Report on the Accuracy of Face Witness Memory in International Arbitration.