Part of the Solution

Idealistic musings about eDiscovery

5 responses to “Stripping Application Metadata: The Counter-Argument

  1. craigball August 29, 2011 at 10:32 pm

    “But Gary,” I’m NOT arguing [here] that different technology makes for different requirements. My post wasn’t about the duty to preserve or the consequences (if any) of shredding drafts when you are not under a duty to preserve. In that instance, shred your heart out and turn track changes on, off or sideways.

    My point is that, if the collaborative content *was* saved and is relevant/responsive to a production request, producing parties can’t just surreptitiouly strip out the collaborative content because they don’t want to deal with it. Using your paper analogy, if I shred a draft before I have cause to recognize it as evidence (“no anticipation of litigation”), then no harm, no foul for my actions. But, if I retain a copy of the paper document that a collaborator has used to write notes to me, I can’t just summarily exclude that marginalia from my review anymore than I can summarily ignore other relevant communications (e.g., e-mail about the document or Post-It notes to myself, if they may be relevant). Even privileged responsive content must be logged for the other side, right?

    My point is that if the document and the communications about it are relevant, then the fact that a collaborative feature was used as the means of communication doesn’t serve to exclude it from the scope of discovery. Yet, out of a fear that such communications might be disadvantageous or ~shudder~ privileged, some lawyers just ignore the content or create copies of the document that strip the content away without reviewing same for relevance. That’s what I’m railing against. Calling a communication “versioning” doesn’t give it some special immunity from discovery, if it is relevant ESI (or integral to relevant ESI).

    Put another way, I play a lot of Words with Friends. WWF contains a messaging tool that stores the messages to/from your oppoent with the game. Suppose those messages were relevant to a lawsuit (“Monopoly! Great play. reminds me, wanna fix prices on our lysine products?”). I wouldn’t get to pretend they didn’t exist simply because I used a Scrabble app as the means to transmit them. Sorry, Marshall MacLuhan, it’s the message, not the medium.

    If it’s a relevant communication that I’ve retained when the production duty arises, I either need to search it for relevant content or I must identify it to my opponent as a potentially relevant source that I am not going to search. Do you disagree?

    • Gary Wiener August 30, 2011 at 6:20 am


      That’ll teach me to knock out a blog post after a long day of flying … these posts can always use yet another round of editing and revision.

      I do agree with you that, if the version information exists, it should be subject to review and production. To say that it shouldn’t be is analogous to saying that because a criminal suspect was stupid enough to leave a smoking pistol out in plain sight, it should be ignored as potential evidence; I have no problem with using a party’s ignorance against them.

      However, I’m concerned about the potential backlash of what you have proposed. Most (non-eDiscovery) attorneys don’t know enough about metadata, or (let’s be honest) the current state of the ethics debate over scrubbing it for production, to accept your suggestion at face value. I’m cynical enough to believe that the knee-jerk reaction for many counsel will be to scrub more metadata out of their collections, on the theory that (as you say) they don’t even want to know what might be hidden in that file wrapper.

      I think the fundamental problem is that our legel brethren (and sistren) have not embraced modern eDiscovery with the fullness and fairness that we technophiles had hoped to see. The McDermott malpractice case is a good example (whether it’s true in McDermott’s case or not) of just how undisciplined current document review methods are. Now we propose to make ESI review more challenging by requiring review of previous embedded versions, when most document review platforms don’t easily support such review. It’s not always as easy as turning on the “Final Showing Markup” option in Word; the trend in modern doc review is to get away from reviewing most native files directly.

      I don’t disagree at all with your ideal of how eDiscovery metadata SHOULD be managed; far from it. I’m afraid, though, that until litigation counsel have learned the wisdom of managing ESI properly for review in the first place, your proposal might overwhelm less-savvy lawyers to the point that they may tend to strip out metadata more frequently, just so (to their thinking) they can get that work off their plate altogether.

      Of course, my angle is purely an academic one and may not be practically valid at all. However, I think you said it best in your own post regarding the current state of version history production: “Embedded content must be captured to a load file, without any context to give it meaning, Long comments tend to get truncated, and colors greyed out. Reviewers employing key word searches may get hits on embedded text the reviewer can’t see onscreen. It can be a real mess.”

  2. craigball August 30, 2011 at 10:44 am

    I understand and share your aptly-stated concerns. There’s a divide between benign edits in a document’s history and revealing communications between collaborators. The former tends to be merely noise in review where the latter gravitates to smoking gun status. Bathwater : Baby.

    Like much in e-discovery, the principle is simple and the execution challenging. But, if we don’t say it matters–and if courts don’t demand its review–then the bar will go right on acting like this pool of potentially significant evidence doesn’t exist. Isn’t this how e-discovery evolves; i.e., yesterday’s undue burden is today’s best practice?

    I’m working on a post about it, but I want to make clear that I am *not* advocating that native applications (e.g., Microsoft Word or PowerPoint) be employed as review tools in native productions. That’s a very poor practice indeed. Native productions have many advantages, but none is, “I can review the files using my own copy of Microsoft Office.” I get that pushback often: “Hey Craig, yer always tellin’ people to get them native files. Well, I think it’s more expensive cause it takes so durn long to load each file into Word, and doin’ it changes the evidence.”

    An acceptable workflow demands that reviewers employ review tools enabling them to see and search embedded commentary in common productivity files without altering the evidence. Review tools that can’t are as obsolete as floppy disks.

    Let’s say it simply: If it’s data in the file, if it’s potentially probative of issues in my case, I want to see it.

  3. parapadakis September 3, 2011 at 4:43 pm

    Craig and Gary,

    Interesting perspectives from all the post above. I can’t comment on the pure legal side, but I can share some anecdotal conversations from eons ago (over 12 years) when I was still designing Document Management systems for Pharmaceutical companies.

    The actors: The Legal & Compliance Team, the Research & Development department and the solution designer (me!). The subject: “Should interim draft versions of compliant documents (i.e. records) be retained or purged when the final version is approved?”. Now, it’s worth remembering that a “document” in this instance, is potentially a 300,000 page marketing application for a new drug, where some of the component sections are in production (author, review, revise, approval) for up to 7 years. As you can imagine, some of these see a lot of versions, revisions, and a lot of comments before a final version is “approved” and signed off.

    The R&D position was quite simple: Draft versions and interim comments are valid business documents. They document the thought process that led to the conclusion and provide additional background information and thoughts, findings, methods that may very well prove invaluable in the research process of another document or the next drug, but may be excluded for brevity from the final version. Why, when faced with the same decisions, should they duplicate the effort only to reach the same conclusion? Why waste all that expensive knowledge or have to expend even more effort extracting it and keeping it separately, potentially out of context.

    The Legal team had a very different perspective: What if something was to go wrong with the drug at a later day, and the company was sued? All draft versions would be also discoverable. If a bio-statistician ad added comments in one of the draft versions alluding to a possible negative side-effect, that were dismissed in the review process, the company would be liable. Therefore, as soon as a final version of a document had been formally approved, all interim versions and comments should be purged as soon as possible since they could become a liability to the company.

    To this day, I see no way to settle this argument. I can see validity in both views.

    The argument was never settled at the time and it was left to the discretion of the authors (on an entirely arbitrary basis…) to decide which of the interim versions could be thrown away and which should be kept. Needless to say, human nature being what it is and the authors being busy scientists, they had better things to do than retrospectively read multiple draft versions to “clean up”, so in most cases draft versions were retained.

    As a footnote: From a technology perspective it is becoming a lot more difficult these days to distinguish between document “versions” (i.e. multiple historical instances of the same document) “revisions” (commentary and changes which tend to be kept inside each document), “internal metadata” (also typically contained within the authored document) and “external metadata” (which is kept in a document or content management system, alongside the document). There are tools to preserve or dispose of all of the above, and it is an Information Governance issue for every organisation to negotiate the fine line between business value and potential liability.


  4. craigball September 3, 2011 at 6:09 pm

    Thank you for the perspective, George. Very valuable. My view is simple, if not simplistic: If there is no regulatory, business or legal obligation to preserve drafts, delete them or not as the custodian chooses. But once there is a duty to preserve potentially responsive data and the drafts qualify as potentially responsive, they must be preserved.

    You wrote, “If a bio-statistician had added comments in one of the draft versions alluding to a possible negative side-effect, that were dismissed in the review process, the company would be liable.” If those side effects were the basis for the claim and the allusions to same were made in anticipation of litigation, they are precisely the sort of information that companies should be obliged to keep.

    If the communication occurred in a discrete e-mail, there would be little doubt that it must be retained, reviewed and produced (if relevant and not privileged). What difference should it make that the communication occured in an embedded comment?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: