Disagreements about scope and process in e-discovery shouldn’t split between plaintiffs’ and defendants’ interests. After all, everyone is a requesting and producing party, whether north or south of the “v.” Yet, the reality is that most defense counsel see themselves as producing parties, and most plaintiffs’ counsel identify with requesting parties. That unfortunate alignment poisons our ability to set aside allegiances and be officers of the Court mutually determined to find the most effective and efficient means to discover evidence illuminating the issues.
Cooperation in e-discovery is derided as naive in an adversarial system of justice, and “discovery about discovery” is vilified as a diversionary tactic, a modern take on the maxim, “if you can’t try the case, then try your opponent.” Counsel for responding parties are quick to note that no party is obliged to deliver a perfect production. They’re absolutely right. Perfection is not the standard. But, is a producing party entitled to fail before a requesting party may inquire into the scope and process of e-discovery? Must we wait until the autopsy to question the care plan?
These questions bring the clash of attitudes between plaintiffs’ and defense counsel into sharp relief.
Plaintiffs’ counsel laments:
“You gave us junk. You weren’t forthcoming about sources of ESI, and your searches were awful. You used keywords instead of common sense, missing patently relevant information. Important documents for key custodians turned up after we took their depositions. You didn’t understand the simplest questions about sources or forms, and you used insane estimates from vendors to assert disproportionality. Defer to your competence? Not when we saw no evidence of it.”
With great umbrage, defense counsel responds:
“You sued us. Your vague and over broad requests sought all manner of disclosure, and we spent a fortune collecting and producing information. Because you couldn’t make a case through your fishing expedition, you now question the integrity of what we did, seek to invade our work product and challenge our competency. Where’s your proof that we failed?”
In the days of paper discovery, we wondered whether our opponent would hand over the documents that hurt them or helped us. We didn’t always trust them; yet, we took it for granted they knew where to look for responsive documents and knew how to search them. The process of production was rarely scrutinized because it was straightforward and well-understood. Counsel enjoyed deference because we could presume competence.
The processes of electronic discovery are neither straightforward nor well-understood by most attorneys. We cannot presume competence. Moreover, counsel are not making judgments about most of what’s implicated in discovery. Huge swaths of responsive information are made invisible by poorly-designed and -implemented processes. Most data and documents are never scrutinized by lawyers; so the esteemed lawyer judgment that work product evolved to protect doesn’t come into play. This is assuredly true of keyword search and remains so with respect to predictive coding.
Although data volumes demand the use of technology as a proxy for lawyer judgment, we should not equate the two nor afford mechanized culls the same protected stature we gave to lawyers reading every document. The more we learn about the unreliability of linear human review, the more it seems unwise to grant human review much deference at all. Let’s not compound that error by extending work product protection to mechanized culling and search.
Moreover, not everything a lawyer does warrants work product protection. Foremost among these unprotected areas should be activities outside a lawyer’s ken and competency. It’s unrealistic to presume lawyer competency in e-discovery. We just haven’t gotten there. It’s delusional to defer to a competency in counsel that experience tells us is not there.
If we define success in discovery as getting to the most probative information with as little waste as possible, then our best hope for success in e-discovery is a transparent, collaborative process. We must not embrace a “right to fail.” Instead, we must facilitate agreement–or challenge–early in the litigation. We must strive to expose deficiencies in scope and process while they can be rectified easily and cheaply. Judiciously implemented, discovery about discovery is not a cul-de-sac; it’s an express lane. Too, it’s an education for all. The more courts allow discovery-about-discovery, the less we will need to do it, because what we will discover along the way are better, cheaper ways to succeed.
Craig, As usual I would agree with your observations. In terms of binary classification, aka predictive coding, there has to be transparency. There is always a tradeoff between false positive error and false negative error. The defense would prefer a greater false negative figure, which would potentially keep relevant evidence out of the hands of the requesting party and the prosecution would prefer a lower false negative figure which would mean the production is more inclusive; but now the defense has produced irrelevant documents which they would argue they had no obligation to produce under the rules. Ultimately, the parties have to negotiate over a cut off threshold using a measure such as ROC and the only way that can be accomplished is transparency and full disclosure. Importantly, it may be helpful to remind ourselves that document classification is a science which demands reliability, transparency and repeatability.
William Hamilton said:
Thanks for this timely post, Craig. While we all welcome the coming new rules (see the post today on the eDiscovery Daily Blog), we should worry that the new rules’ explicit emphasis on proportionality will further encourage judicial officials to dramatically limit discovery about data. The irony is, of course, that discovery about data sources, repositories, and locations facilitates proportionality. You are right that good “discovery-about-discovery” makes sure we go after, target and find the likely important data with the right tools. Without “discovery-about-discovery” we wander haplessly on the Data Great Plains.
David Tobin said:
Pingback: No Perfection Standard In E-Discovery? | The Researching Paralegal
Pingback: Is There a Right to Fail in E-Discovery? | @ComplexD