Insight
Three simple rules for conducting a cross-examination
Seasoned arbitration practitioners are used to seeing various ways to conduct a cross-examination, not all of which are equally effective. It is in the interest of our clients, as well as, indirectly, the arbitration industry, that cross-examinations are conducted as efficiently and effectively as possible. Increased efficiency will result in less time and money being spent on cross-examination, whereas increased effectiveness is more likely to produce evidence that is favorable to your client’s case.
For the benefit of younger practitioners who can still avoid learning bad habits, I propose below three simple rules that may assist in preparing a script for a cross-examination.
- Rule 1: Cross-examination is an algorithm where the input is the counterparty’s evidence and the output is your client’s evidence.[1]The black box in the middle – the interval between the input and the output – is composed of a series of questions that you must custom-develop for each cross-examination, depending on the subject matter of the witness’s evidence.
What does this mean? It means that, as a rule, there should be no questions in a cross-examination that do not serve the function of producing evidence that supports your client’s case, or at least neutralizes the counterparty’s evidence.[2] However, in a longer cross-examination you may wish to add a few fireside chat questions, to relax the witness and make the process easier to digest for the arbitral tribunal.
- Rule 2: Break down the witness’s evidence into smaller units that are neutral, in themselves, for the counterparty’s case.
What does it mean? Breaking down the witness’s evidence into smaller units (more limited sets of facts) neutralizes the evidence by deconstructing its narrative structure. This allows you to put questions to the witness that are safe from the point of view of your client’s case: even if you do not get an answer that supports your client’s case, the answer is at best neutral for the counterparty’s case. The rule also allows you to establish control over the witness in terms of both the substance and scope of the evidence given.
- Rule 3: Organize your questions around disputed issues in a logical order such that they eventually lead to a question that, in light of the witness’s earlier answers, can only be answered in one way, and that way supports your client’s case.
What does it mean? Once you have deconstructed the narrative structure of the witness’s evidence into smaller units, you are able to reorganize these units in a new narrative structure that reflects your client’s case. Depending on how well the witness is prepared, they may or may not see what you are getting at – but there is nothing they can do to stop it; you are the one asking the questions. Thus, if the witness provides a truthful answer to the last question in the series, you will have achieved what you set out to achieve, i.e., you will have generated evidence that supports your client’s case. However, even if the witness understands what you are getting at and tries to avoid directly answering the last question in the series, or does not give a truthful answer, this will undermine or at least reduce the credibility of the witness, in light of their earlier answers. Thus, even in the worst-case scenario, the value of the witness’s evidence is reduced, if not eliminated, at least insofar it relates to the issue (disputed facts covered by the series) in question. You can strengthen the witness’s lack of credibility by saying: “I see,” or something similar, when the witness gives an untruthful answer, but you should not start arguing with the witness. An understated reaction is to be preferred, as it allows the arbitral tribunal to draw its own conclusions from the evidence. You can make it more likely that those conclusions are in fact drawn in your closing statement or post-hearing submission, but this is not a matter for cross-examination.
These simple rules should take you some way towards establishing a well-structured script for a cross-examination, however, they are not a panacea. They work only if you have done your own homework and know the evidentiary record better than the witness. This is the premise of all effective cross-examination.
[1] An algorithm can be described as a systematic procedure that produces, in a finite number of steps, the answer to a question or the solution of a problem. An algorithm that produces a yes or no answer (such as a cross-examination) is called a decision procedure, i.e., a process that terminates and is, as such, a logical rather than a mathematical process. For a brief description see, e.g. https://www.britannica.com/science/algorithm. By definition, an algorithm can be described but not strictly speaking defined since a definition is, by definition, itself an algorithm (as it provides a finite answer – a de-finition – to a question), and thus logically cannot define itself; see Alan Turing, On Computable Numbers, with an Application to the Entscheidungsproblem, in Proceedings of the London Mathematical Society 2 (42): 230-265 (1936) (proving that Hilbert’s Entscheidungsproblem cannot be solved).
[2] “Neutralizes,” that is, generates evidence that neither supports nor undermines the counterparty’s case. Such evidence tends to establish undisputed facts.
Back to listing