Judgment in Triumph Controls UK Ltd & Anor v Primus International Holding Co & Ors  EWHC 176 (TCC) was delivered by Coulson J on 7 February 2018, the same day as the very good session explaining the new disclosure rule of which I wrote here. There are interesting crossovers between the two developments, links between Triumph Controls and other cases criticising disclosure management, and other factors which make Triumph Controls part of a continuing thread focusing on how disclosure is performed.
The advantage of sitting tight for a bit before writing about judgments is that someone else will leap in and do a summary, saving me the wearisome task of paraphrasing the thing. We have had summary reports from Masood Ahmed (who is on the Rule Committee), from Ed Spencer and Lizzie Hancock at Taylor Wessing and from Pinsent Masons Out-law.com and I do not intend to cover the same ground – the judgment is only 42 short paragraphs long anyway.
What might be more helpful is to pick on some specific points which have implications beyond the bare decision.
Although this judgment appears as a victory for the defendants, it did not go all their way. Their original application was for “wide-ranging orders [which on the face of it] required the claimants to redo, from scratch, the search exercise that they had undertaken”.
This was described in the judgment as “unrealistic” and the result was an application for “two more focused orders”. It was the defendant who got the first black eye on the proportionality front.
Then, of those two “more focused orders”, the claimants successfully beat off the first. The judge said of this application:
I consider that the claimants have always been clear about what they have done about the Farnborough shared drive; any complaint about it now raised by the defendants is very late; the evidence shows that the claimants’ approach has been reasonable and proportionate; and the defendants are unable to demonstrate that any particular folders/file paths have not been disclosed.
Unnecessary, late, ignoring the reasonableness and proportionality of their opponents, and lacking in evidence of harm; if that had been their only application, the defendants would have lost. Have a quick skim through Denton before bringing that kind of application on its own.
The second application involved an agreement to review 450,000 “responsive documents”. I put “responsive documents” in quotation marks because the reference is not to an agreed final delivery but to “potential documents” resulting from a search using agreed keywords. The claimant’s EDQ (Electronic Documents Questionnaire) indicated that these documents would be “manually reviewed” and there was no reference to what the judgment refers to as a “Computer Assisted Review (‘CAR’)”.
TAR v CAR v Whatever – it does actually matter what you call the process
Let’s just pause to beg for some consistency in the terminology used to describe the various means by which computer technology is used to reduce volumes and arrive at a disclosure set. In Pyrrho Investments Ltd v MWB Property Ltd & Ors  EWHC 256 (Ch) the Master said
the term ‘predictive coding’ is used interchangeably with ‘technology assisted review’, ‘computer assisted review’, or ‘assisted review’. It means that the review of the documents concerned is being undertaken by proprietary computer software rather than human beings.
…and went on to use the term “predictive coding” in four of the ten factors listed at the end of the judgment.
The Registrar in Brown v BCA Trading Ltd & Ors  EWHC 1464 (Ch) didn’t bother with definitions but plunged straight in with “…using predictive coding or via a more traditional keyword approach instead”, and said that “predictive coding must be the way forward”. In Triumph Controls, as I have said, Coulson J talks about “Computer Assisted Review (‘CAR’)”.
If these judgments are to be of any use in deciding what has and has not been approved by the court, then we need some uniformity of nomenclature or, rather, some explanation of what is meant by any of the terms used which does not depend on saying that it means something a bit like some other term. Triumph Controls is not really about using the “wrong” technology or using it wrongly but about breach of an agreement to do a manual review, about lack of transparency as between the parties, and (it seems) a failure to explain things properly to the court. Nevertheless, we ought to be given some idea of what “CAR” is intended to mean, or what Coulson J understood it to mean in this case.
Let’s go back to Pyrrho and that unfortunate phrase used by the master:
It means that the review of the documents concerned is being undertaken by proprietary computer software rather than human beings.
That black-and-white “the job is either done by software or by humans”, when coupled with vagueness as to what type of tools were actually used, will come back to haunt us as it is idly copy/pasted from judgment to judgment. If it did not matter much in Triumph Controls (because the failing was at a higher level), it will matter in future cases, both at the stage of agreements or orders (when parties casually use expressions without clarifying what they mean) and on subsequent enforcement applications.
Those drafting the new disclosure rule need to take notice of the potential difficulty here, which will cause problems even before considering that different software providers use different methods and different terminology. This is not the place for a full disquisition on the subject, but we need to make the point.
I am reminded of the 2013 case in the Alabama Admiralty Court which I wrote about here, and its links to Ralph Losey’s article Poor plaintiff’s counsel, can’t even find a CAR, much less drive one. In that case, the lawyers reached an agreement using terminology which, as it turned out, neither of them understood. Even the best solicitors (and the claimant’s solicitors in Triumph Controls are among the best at this game) can stumble when an agreement uses terms which may mean different things to different people. The Pyrrho quotation above which unintentionally implies that “manual review” is the straight opposite of TAR, CAR, predictive coding or whatever, with no shades of grey between them, is unhelpful here.
The importance of protocols
Coulson J refers approvingly to the TeCSA/TECBAR edisclosure protocol developed for the Technology and Construction Court, and specifically to its provisions for agreements between the parties. He gently knocks down Master Matthew’s assertion in Pyrrho that the TCC eDisclosure protocol is “only a protocol and [it] has no normative force”. This, Coulson J says, is a
rather old-fashioned approach [which] misunderstands the importance of protocols in the TCC, where the court’s overarching requirement is that the parties cooperate and comply with them, unless there is a very good reason why not.
The drafting of protocols is a non-trivial task. They require that the terms be clear and unambiguous whilst leaving room for changed circumstances. In this context, you might like to look at the protocol ordered by Maura Grossman (she who first came up with the term “technology assisted review”) in a US case called In Re Broiler Chicken Antitrust Litigation. There is an excellent article about that case by Michele Lange on the ACEDS blog here and the order itself can be found on the Catalyst blog here
I won’t pause now to analyse it, but it is commendable both for its clarity and for its recognition that things may change. I am not setting it up in rivalry to the TeCSA/TECBAR edisclosure protocol, merely offering it as a source for ideas despite its US context. Time spent in negotiating these things is time well spent.
Three key sentences
There is a lot in this judgment about the apparent defects in the claimant’s exercise, the weakness of the sampling technique and so on. These have been amply reported elsewhere and I want to focus on three sentences.
what they did is not what they said they would do in the EDQ, which promised a manual review of all documents responsive to the keyword searches. Neither is what they did at all clear from their Disclosure List.[#27]
There were perhaps 10 paralegals for associates involved in the searches. It is not apparent that there was any overseeing senior lawyer. [#30]
…given that both the CAR and the sampling were unilateral, I am slightly surprised that there is not now better evidence as to what actually happened. [#29]
Taking each of those in turn:
One can, perhaps, envisage circumstances in which failing to comply with an agreement, protocol or order might actually produce a better outcome than that which would have resulted from strict compliance, and there is something slightly dangerous in the assumption that you are stuck with whatever you originally agreed to, come what may. Parties ought to be encouraged to think that they can reopen agreements (as in Maura Grossman’s protocol referred to above) if circumstances change and that, if they cannot persuade their opponents, they can go back to the court. In the absence of such further agreement or order, however, Coulson J’s proposition that parties should abide by their agreements is not wrong.
One difficulty here, and one which I raised at the recent public discussion about the proposed new rules, is that the idea of going back to the court is all very well provided that you can get a hearing date within a reasonable time.
The “overseeing senior lawyer”
Coulson J’s comments about the number of people involved in the disclosure exercise without “any overseeing senior lawyer” echoes a comment made by Hildyard J in RBS Rights Issue Litigation  EWHC 3433 (Ch) when he referred to ““…the use of vast armies without any sufficient focus…” and to:
an unfocused disclosure process, which has fanned out exponentially and extravagantly without sufficient control and direction; (b) the commitment of increasing resource to the identification of documents, leaving a diminished resource for their assimilation, without properly taking stock as to whether the process had overtaken the purpose and/or whether a more confined process should be adopted.
The disputes in the two cases, though both about disclosure, were not of the same kind, but the general observation about project management arises in both. Who is actually driving this CAR?
“..better evidence as to what actually happened” in the disclosure exercise
The “overseeing senior lawyer” should also be the one who heads off Coulson J’s comment about there being “not now better evidence as to what actually happened”.
When I first passed this judgment around, a US commentator wrote to me saying: “my reading suggests a better defense of what was done, and why, would have won the day”. It is hard to tell, of course, from the sparse words of a judgment, particularly one whose core is about inadequate explanation, but I wonder if the outcome would have been different if an “overseeing senior lawyer” had been available to articulate clearly what had been done.This touches on another point raised during the recent session about the new rules, covered in my report on it here
Judges don’t have time for detailed assessment, Judge Waksman went on. Perhaps eDisclosure providers might be called upon to explain what is involved. It would be good to hear from the lawyer who is actually in charge of disclosure, he said, not necessarily the barrister or partner – it may be the associate actually running the process, for example.
So there’s an idea – when disclosure issues are being argued before the court, bring along someone, or evidence from someone, who is actually involved in the exercise – not merely responsible for it but actually doing it. Defence, as well as attack, may benefit from “..better evidence as to what actually happened”.
At one level, Triumph Controls tells us nothing new. The existing Practice Direction 31B, lauded by Coulson J, has been around for a while and its obligations as to cooperation and agreements go back further still – it underlay the conclusion in Digicel as Coulson J observes.
There has been plenty of notice that the courts have switched their fire from technical breaches of the rules to criticism of the management of the exercise – not just in the RBS case referred to above but in another Coulson J case The Dorchester Group Ltd (t/a the Dorchester Collection) v Kier Construction Ltd  EWHC 3051 (TCC) where he described Kier’s disclosure process as “cumbersome, and inadequate” and said that “the disproportionate costs of the Kier disclosure exercise are due to the way in which the exercise itself has been carried out”. Again, the specific points at issue were not the same, but the principle – that someone identifiable should be in charge of the disclosure exercise – appears from both judgments.
The idea that eDisclosure exercises be explained by someone actually involved in them seems too obvious to warrant too much comment. It is interesting that the same point should have arisen on the same day in Triumph Controls and in the talk about the proposed new rule. Perhaps the rule-makers might consider providing expressly for this, to head off the possibility of a judge declining to hear such evidence – and anyway, it can’t hurt to have the evidence to hand.
If it tells us nothing new, Triumph Controls gives us a handy compendium of principles and points which may help when evaluating if there really is something to complain about, whether in our own disclosure or that of our opponents.