If a litigant has actual knowledge that an opponent’s proposed search terms will not “hit” on responsive, proportional ESI, does the litigant have a duty to tell the opponent?[1]
In Advanced Magnesium Alloys Corp. v. Dery, 2022 WL 3139391 (S.D. Ind. Aug. 5, 2022), the plaintiff, Advanced Magnesium, alleged that defendant Dery conspired with defendant Alliance by using Advanced Magnesium’s trade secrets to “jump start” Alliance’s entry into the magnesium recycling market. Plaintiff and Alliance agreed to 159 search terms. The court wrote:
Plaintiff argues that one reason false negatives are likely in this case is that “Alliance’s top executives conducted themselves in a deliberately obfuscating manner in its internal communications, using code names for Dery … and complains that “[t]ellingly, Alliance did not offer to use Dery’s code names of ‘Mister X,’ ‘Alain X,’ ‘Mr. X,’ ‘Monsieur X,’ ‘M. X,’ or Dery’s initials as search terms when it proposed search terms for responsive ESI in this case.” [emphasis added].[2]
Did Alliance have a duty to “fess up?” The court did not directly address that question and, instead, ordered additional relevant proportional discovery. “As is so often the case in litigation, … we find ourselves in a world that is far from ideal…. The Court is left to fashion a resolution to this dispute that would have been far more efficiently resolved by cooperation between counsel.”[3]
Jason R. Baron, Esq., has directly addressed the question:
A document search now, when so many of the records exist only in digital form, must be done using (among other techniques) carefully selected search terms. The party “producing” the documents will have much better knowledge about what terms will capture the relevant material than the other side will have. What obligation is there, and what obligation should there be, for the side with the greater knowledge to assist its adversary in finding important material as part of the obligation to “meet-and-confer” in the discovery planning process? Note how this question raises directly, now in a digital context, the exact same dilemma that has been with litigators since the advent of discovery: where do my obligations to my client end and my obligations to the system, and even to an adversary, begin?
Patrick Emery Longan, “Foreword: Ethics and Professionalism in the Digital Age 9th Annual Georgia Symposium on Ethics and Professionalism,” 60 Mercer L. Rev. 845, 847 (2009)(emphasis added).
Jason’s symposium presentation was titled “E-Discovery and the Problem of Asymmetric Knowledge,” Transcript-Morning Session, 60 Mercer L. Rev. 863, 866 (2009). He explained:
I refer to asymmetry, or asymmetric, because in the usual course, one party (usually a corporation or an institutional defendant), has privileged access to its own enormous data set. This has always been true, even in the world of paper documents, but today there is a twist: given the impossibility of manual review to find smoking gun documents and other material documents, the only way parties will have access to vast collections of ESI is through making intelligent, reasonable requests for automated searches to be conducted. But now the stakes are higher in framing such requests, because small differences in the language of those requests make an enormous difference when using computers to search for strings of text.
Jason asked:
Does counsel have a duty to “assist” in the negotiation process if one side had made very poor requests for information in the form of search terms? If the value of the requests are poor because they are way over-inclusive, then … courts will expect the parties to engage in narrowing to filter out thousands or millions of irrelevant documents. What if the requests are on their face reasonable, but the other side … knows that they are poor requests based on special knowledge? Under these circumstances, to what extent should counsel to a party be obligated to “reveal”–and I use that word conscious of the ethical prohibition here–some aspect of her client’s electronic data set in the interest of furthering expedition of the litigation?
All is not so simple in the land of ESI, and the ethics of negotiating search protocols is very much an open area awaiting Judge Facciola and his brethren to weigh in.
Id. at 874-75 (emphasis added).
Jason continued:
Let me pose a modest hypothetical (or not so modest, because it has some length to it). Assume that Corporation X is required to produce all documents on a certain fungicide named “Benlate.” Assume further that a specific request to produce documents asks for all documents from the CEO of Corporation X, whose last name is “Bartlett,” that in any fashion relate to the sale of the fungicide Benlate. Also assume that the general counsel of defendant Corporation X has been on the job for the last ten years, during which time she has been involved in one hundred lawsuits involving discovery searches against Corporation X databases, including suits for CEO records. Also assume that although no suit specifically involved production of Benlate, the general counsel is extremely familiar with both the proprietary name Benlate and the generic name of “Benomyl” for the fungicide, as they are interchangeably and commonly used throughout the corporation. Finally, assume further that the general counsel knows that there are ten million OCR’d documents (that is, documents that have been scanned in using an Optical Character Recognition process with a known error rate), and that in the past there are many documents that have been uncovered scanned in with the name Bartlet, without the second “t” at the end.
Now, assume that the plaintiff’s counsel, without knowledge of prior litigation against the defendant, proposes a set of keywords to be searched. Further assume that on the plaintiff’s list of keywords is the term “Bartlett” (spelled correctly), as well as “Benlate.” Excluded from the list are either any variations on the name Bartlett, or any synonyms for the specific proprietary name “Benlate.” The parties are ordered by the court to meet-and-confer regarding the subject of search protocols to be used to find relevant evidence.
Question number 1: Does the defendant’s counsel (here, I necessarily include Corporation X’s general counsel assumed to be present sitting with outside counsel), at a meet-and-confer, have an ethical duty to inform the plaintiffs that searches of the term “Bartlett,” spelled correctly, would not necessarily produce all instances of documents related to the CEO, given the known presence of spelling errors in the collection?
Question number 2: Does the defendant’s counsel have a further duty to identify that a search for the term “Benlate” may be under inclusive, when she knows that a generic name for the fungicide, Benomyl, almost certainly exists in the collection?
I have an unequivocal response in answer to Question 1: Yes, it would be taking advantage of a factual mistake not to “fess up” to one’s prior knowledge, in the form of saying that there should be more than just a correct spelling of the keyword, so as to encapture all variations of the CEO’s name, Bartlett, either with one “t” and two “t’s” at the end, or any other fuzzy variations that are known. Failure to discuss expanding the search request to encompass known variations of the spelling of the CEO’s name would be tantamount to withholding a material fact from the other side during the negotiations.
On Question 2, the matter of Benlate vs. Benomyl, while I do not believe counsel has a general duty of hand-holding to make opposing counsel’s searches better, if the matter bore greatly on a material issue in the litigation, it would also be a mistake to fail to proffer the synonym as part of the negotiations. That is, I believe, the right answer is to provide one’s knowledge rather than wait for a later iterative cycle to reveal the term. This is more controversial, but I say this because (a) the second term, Benomyl, is almost certain to be found out anyway, and one gains credibility in being up front about it when there is no outcome-determinative difference perceived; and (b) failure to do so greatly increases the risk of a “do-over” in having to redo discovery. Whether I should or would do so only with the client’s consent is something which I take it would be Professor Freedman’s way out of this ethics conundrum–he would require me to go to the client to get permission for use of a synonym that the other side had not considered. But it is an issue that I would be quite interested in having my fellow panelists weigh in on with this caveat: In the real world, there is precious little time to bring clients in to micro-manage litigation decisions on this level. And so, as a practical matter one may have to assume that the lead lawyer in the representation will be by default “on her own” in deciding what course of action to pursue.
Id. at 875–76 (emphasis added).
Jason based his conclusion on the duty to cooperate, the attorney’s role as an officer of the court, the lawyer’s duty of fairness to the opponent, and wrote:
As applied in this brave new world of ESI, my argument is that failure to disclose special knowledge of the relevant characteristics of the ESI universe of your client, that would knowingly fail to optimize and impair an otherwise reasonable-sounding search strategy or protocol request, is tantamount to the suppression or concealment of evidence in conflict with one’s ethical obligations under Rule 3.4.
Id. at 877 (emphasis added); accord “The Case for Cooperation,” 10 Sedona Conf. J. 339, 344 (2009);; Dana A. Remus, “The Uncertain Promise of Predictive Coding,” 99 Iowa L. Rev. 1691, 1716 (2014).
Further, Jason also justifies his conclusion based on practical considerations. “[L]et me suggest that what I consider here as doing the right thing also makes good economic sense, and it can be sold to clients under that banner if a hymn to higher ethical standards fails to move them.” He argues that this duty to disclose avoids a “do over,” which is precisely what the court ordered in Magnesium Alloy.[4]
One court addressed the duty when a search for “Arthur Montana” failed to return relevant documents mentioning “Art Montana”:
The federal rules do not require that plaintiff identify the exact “magic words” to obtain clearly relevant discovery. The “obligation on parties and counsel to come forward with relevant documents not produced during discovery is ‘absolute’.” Tracinda Corporation v. DaimlerChrysler AG, 502 F.3d 212, 243 (3d Cir.2007). Defendant had an independent duty to produce key relevant requested discovery even if plaintiff did not specifically list the precise ESI wording or spelling in its list of search terms. Defendant may not delegate to plaintiff the duty to identify its relevant requested documents. As noted in Poole ex rel. Elliott v. Textron, Inc., 192 F.R.D. 494, 507 (D.Md.2000), “[t]he rules of discovery must necessarily be largely self-enforcing. The integrity of the discovery process rests on the faithfulness of parties and counsel to the rules—both the spirit and the letter.” Moreover, “the discovery provisions of the Federal Rules are meant to function without the need for constant judicial intervention, and … those Rules rely on the honesty and good faith of counsel in dealing with adversaries.” Hopei Garments (Hong Kong), Ltd. v. Oslo Trading Co., Inc., No. 87 CIV. 0932(MBM), 1988 WL 25139 (S.D.N.Y. March 8, 1988).
It is absurd to posit that because defendant’s e-mails mention “Art” or “Art Montana,” instead of “Arthur Montana,” that defendant did not have a duty to produce relevant e-mails between Lanzetta and Judge Connor that mentioned plaintiff. Defendant should have undertaken a search for e-mails mentioning “Art” or “Art Montana” even if plaintiff did not specifically identify these precise terms. If the search term “Arthur Montana” was not sufficient to pick up references to “Art” or “Art Montana,” defendant should have known this and taken steps to expand its search. It is stupefying that in December 2012, the Court had to Order defendant to conduct an ESI search for documents mentioning “Art” and “Art Montana,” and that the search identified relevant documents at that late date that were not already produced. The Court would be flabbergasted if plaintiff did not rightfully assume that his designation of the search term “Arthur Montana” did not also subsume a search for “Art”, “Art Montana” and “Montana.”
Montana v. Cnty. of Cape May Bd. of Freeholders, 2013 WL 11233748, at *9 (D.N.J. Sept. 20, 2013)(emphasis added). While a technologically competent attorney in 2022 likely should have a better grasp of designing Boolean search terms, the principle appears sound. “[E]ach party must sign its disclosures and certify that it has conducted a reasonable search.” The Raine Grp. LLC v. Reign Captial, LLC, 2022 WL 538336 *2 (S.D.N.Y. Feb. 22, 2022). The producing party “should also cooperate with the requesting party to the extent the requesting party believes that other search terms, custodians or locations may have relevant information when fashioning an ESI protocol, subject to Rule 26(b)’s limitations.” Id. at *1.
In sum, an ESI protocol and search terms work in tandem with the parties’ obligations under the Federal Rules and do not replace a party’s independent obligation to produce electronic (or paper) documents that are reasonably accessible, relevant, and responsive within the meaning of Rule 34.
Id. (emphasis added).[5]
____
[1] In Is There a Duty to “’Fess Up?”, I addressed that question in the context of inability to preserve information that should be preserved. There, the better rule is candor.
[2] Due to some confusion over the agreed terms, the court wrote that it “can only assume that Dery’s name was also searched for by Alliance and the parties simply neglected to include that fact in their briefs.”
[3] I addressed other aspects of Magnesium Alloys in Be Careful What You Agree To, But the Duty to Cooperate May Provide an Escape Hatch.
[4] Jason proposed an iterative search process and, among others, cited to ClearOne Commc’ns, Inc. v. Chiang, 2008 WL 920336, at *2 (D. Utah Apr. 1, 2008)(“the search protocol is not the ‘last word’ on electronic discovery in this case. The use of key word protocols is one step in the process which contemplates many more steps….”).
[5] “Make no mistake about it, however, it should not have been necessary to ‘spoon-feed’ 7-Eleven regarding the relevant and responsive Project P documents to produce.” Younes v. 7-Eleven, Inc., 312 F.R.D. 692, 701 (D.N.J. 2015)(discussing “Project Philly,” “Philly Project,” and “Project P”).