Have You Thought About the Discovery that Word Processing Might Produce?

April 30, 2013
By Hanging Out with Carl Gunn



  • There may be prior, intermediate drafts of reports on a word processing system in addition to the original notes and the final report.
  • Less “polished” drafts may have valuable information that’s taken out of the final draft.
  • Intermediate drafts ought to be discoverable for the same reasons notes are.



In my post last week, I talked about wanting to get the agent’s notes of an interrogation in addition to the agent’s final report because of the tendency of reports to get polished and leave out some of the things that are helpful to the defense. This reminded me of an experience I had that suggests a related but somewhat different type of discovery to ask for in today’s modern computer word processing age. We know that we try to polish briefs and such other documents to make them as persuasive as possible, and we have the opportunity with modern word processing to easily write multiple drafts in that polishing process. So what might agents do when they write their reports? Maybe polish them in multiple drafts on the word processor as well?

This raises the question of what happens to those multiple drafts that we – or the agents – write. Usually, of course, they just get saved over. But don’t the old drafts sometimes end up not getting saved over? Maybe you’re not so sure which version to go with, so you preserve both the old version and the new version under different names, and you never delete the one you don’t end up using? Or you inadvertently save the final draft – or some other draft along the way – under a different name? Or an earlier draft’s on a computer belonging to someone else to whom you circulated it for comment? So every now and then, one of your earlier drafts still exists on your computer or someone else’s?

Well, what if this happens every now and then with an agent’s report of your client’s statement? Going on ten years ago, this actually happened in a case I had. The initial report of the investigation that was provided to me – which described the whole investigation including the interrogation of my client and was presumably the “final” version at the time it was provided – raised some issues about the voluntariness of a confession by my client, but really not very strong ones. It reflected that my client had denied guilt at first and then broken and admitted guilt, but described nothing that could be characterized very convincingly as coercion.

I then got a second version of the investigation report, however, which was an updated one to which the agent had added paragraphs regarding further investigation. The paragraphs describing my client’s interrogation in this report included an important piece of additional information, namely, that my client – who had never been in trouble before and seemed to have some ambition – had stopped denying guilt and admitted it only after being warned about “the severity of the crime committed and how this crime may impact his life ambitions.” This gave me a much better, if not absolutely convincing, argument that the confession was a product of coercion and so involuntary.

Related to the issue I’m suggesting in this post – about the possibility of different drafts saved on the computer – it turned out the inclusion of this additional information in the later report was a word processing “error.” The agent explained that he had deleted this information from the earlier version of the report I was initially given and it was included in the later report only inadvertently, because he’d mistakenly worked off an earlier draft rather than the later, “final” draft that I had first been given. In other words, the sentences describing what was said to my client to get him to confess had been deleted and weren’t supposed to be in the “final” report to be disclosed in discovery. The agent explained that it had been suggested this additional information could be deleted because “it is not really a relevant sentence.”

Putting aside the questionable ethics of deleting this information from the “draft” report and not disclosing it, this illustrates how there may be earlier drafts of a report on an agency’s word processing system that differ from the “final” report in significant ways. For the same reason we want to push for the original notes, we want to push for these earlier “drafts.” And the earlier “drafts” should be discoverable for the same reasons the notes are discoverable, which is discussed in my post last week.

Have a lot of fun with this one.