Communications of the ACM
Cerf's Up
On Expert Testimony
I recently had the opportunity to participate in a two-day discussion on expert testimony, sponsored by the U.S. National Academy of Science and the U.K. Royal Society. By remarkably good fortune, I am a member of both. The primary topic of discussion was "science in the interests of justice." In a world in which computers, computing, information technology, data science, and computational-X—for all values of X—are becoming the norm, expert testimony on the use and dependence on software artifacts is inescapably on the increase. If you accept that premise, then the role of ACM members, as professionals in the computing realm, makes them candidates for the provision of expert testimony in both criminal and civil judicial proceedings. In the case of the former, especially, accurate and factual testimony may make the vital difference between serving and miscarrying justice.
I learned many things at this event, but one thing really stood out for me: In the U.S. system of justice, experts are hired by each side (plaintiff and defense) to argue for their side. In the U.K., experts often advise the court rather than advocate for one side or the other. U.S. judges dealing with highly technical testimony are often confronted with "dueling experts" and forced to try to make sense of and decide between conflicted testimony. In Scotland, I learned, it is more common that the expert functions in an advisory capacity. In fact, the experts are allowed to participate in pre-trial discussions to inform the court of relevant factual information drawn from the expert's knowledge and experience.
I learned that some U.S. court judges turn to experts to learn more about the subject matter. They do not discuss the specifics of the case but seek to gain appreciation for the technical aspects of the case. One might imagine getting a tutorial on DNA matching, software vulnerabilities, misinformation in social media, hallucinating chatbots, and methods for verifying the correct operation of software applications. The whole idea is to gain an acquaintance with technology and vocabulary, partly to avoid being confused or misled by jargon or worse.
I confess that the Scottish practice greatly appealed to me, especially given the desire that expertise should lead to justice-serving outcomes. It is not clear, however, whether such a practice could be instantiated in the U.S. where litigation is common, and experts are brought in by both sides to try to persuade the judge or jury of the merits of each side's arguments. In highly technical cases, it seems unlikely that judge or jury will have specific expertise to evaluate competing testimony. Hence, a practice that allows the court to draw on its own experts seems attractive. In the adversarial format, it is not unknown for one side to attack the other side's experts, not on merits, but by eroding the credibility of the expert witness. "So, Mr. Expert, when did you stop faking the data in your published papers?"
In the U.S. system of justice, experts are hired by each side (plaintiff and defense) to argue for their side. In the U.K., experts often advise the court rather than advocate for one side or the other.
It occurred to me that ACM could be an ideal source of experts who could be called upon, not to advocate for a point of view but to inform the courts of factual information that would allow the evaluation of evidence for accuracy and credibility. I imagine this is not the first time such an idea has surfaced. Perhaps readers will have experience in this space and would share their opinions on best practices in the interest of justice.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2023 ACM, Inc.
Comments
John Grundy
Many thanks for this piece Vinton. I have several times been an expert witness in New Zealand and Australia court cases - alleged patent infringement, alleged poor quality software delivered, alleged theft of software design/code, and also review of e.g. R&D tax credit claims (i.e. is it really R&D). I find them very interesting - if time-consuming - and its quite a different use of my expertise/experience than my University teaching and research work.
In Australia and New Zealand we use the British approach - the expert witness is an officer of the court i.e. doesn't matter who is paying you; you are advising the court not a particular litigant. You comment (collegially) on each others' reports to the court. Sometimes you even prepare a report together / go on the stand together with the "other side's" expert witnesses as a little expert panel for the barristers and even occasionally the judge themself to question on technical matters. As yet, my credibility as an expert has never been questioned... (there is always next time I suppose). I have no idea how this might operate (or if it could operate) in the US context however.
I do like the idea of ACM being a source of non-partisan experts advising on the factuality of claims, deep domain knowledge, etc.
Best regards,
John
Jan Collie
Thanks, also, for this most apposite article. I firmly believe that expert witnesses that are essentially hired guns have no place in the administration of justice. If the assignment of guilt / culpability or innocence is truly for the court to decide, then the court needs impartial guidance on matters that require specialist knowledge. As you say, accurate and factual testimony can make the difference between serving and miscarrying justice. In criminal matters, in particular, it should never be forgotten that someone's liberty is at stake. It's not a game.
The adversarial nature of litigation axiomatically encourages attempts to discredit witnesses, perceived experts included. In my opinion - based on almost 20 years experience acting as a digital forensics expert in criminal and civil proceedings in England and Wales - I believe It would be more appropriate and effective to emulate the Scottish model. I very much support the idea of ACM being a source of non-partisan experts.
Kind regards,
Jan Collie
Displaying all 2 comments