“American laws and American policy view the content of communications as the most private and the most valuable, but that is backwards today,” said Marc Rotenberg, the executive director of the Electronic Privacy Information Center, a Washington group. “The information associated with communications today is often more significant than the communications itself, and the people who do the data mining know that.”
How the U.S. Uses Technology to Mine More Data More Quickly, New York Times, June 8, 2013
Marc Rotenberg was commenting on the recent revelation that the U.S. National Security Agency gathers a staggering volume of information about domestic and international telephone calls. When he states, “The information associated with communications today is often more significant than the communications itself…,” he doesn’t expressly label that “more significant” information as being “metadata,” but that’s what it is.
Rotenberg’s right: metadata matters.
At the just-completed, weeklong Georgetown E-Discovery Training Academy, we do something not done at continuing education programs: we test the attendees coming in and test them heading out. Most of our attendees haven’t sat for an exam since the bar.
The tests include questions like this:
Which statement, if any, below is true:
A) System metadata is probative of some issue in every case.
B) Some files have more metadata than data.
C) Databases are unique in their ability to dispense with metadata.
D) None, all of the statements are false.
The answer sought is B.
It’s worth adding that some metadata is more probative than the contents of the file to which it pertains.
Consider an e-mail that says, simply, “OK.” That two letter message—a mere 14 bits of ASCII-encoded text—climbs aboard the internet express toting a big satchel of metadata, then picks up considerably more metadata as it traverses networks and servers. The “who,” “when,” “where” and “how” of the message’s metadata comprises considerably more information than the “what” of its content and often conveys more of what we need to know.
The telephone call data gathered by the NSA may not hold the words exchanged—indeed many calls may not have been answered—but the metadata about numbers called, numbers calling and the time and location of attempts speaks volumes and is, as Mr. Rotenberg aptly puts it, “often more significant than the communication itself.”
No one is up in arms about the NSA’s actions on the ground that it wastes government resources to collect metadata. It’s tacitly acknowledged on all sides of the debate that the information gathered is significantly revealing and probative. It’s evidence, and the government grabs and uses it.
Yet, metadata has enormous utility apart from its innate value as evidence. Metadata enables us organize, manage and make sense of digital evidence. Metadata is the “glue” that holds certain evidence together and the labels on the cans that keep us from grabbing dog food when we want soup. This metadata isn’t evidence, but it’s the electronic equivalent of the evidence bag, the date on the evidence label or the entry in the police property room log that helps us find the evidence and assure ourselves of its integrity.
Say what one will about issues of privacy and governmental intrusion, there should be no debate that metadata is powerfully important information. Unfortunately, a different arm of government, the Judicial Conference of the United States’ Standing Committee on Rules of Practice and Procedure, doesn’t give metadata its due.
By way of background, many believe that the American system of civil discovery is broken. E-discovery makes CEOs worry, and we don’t pay CEOs enough to worry. To help CEOs relax, companies throw gobs of misdirected money at e-discovery. Accordingly, e-discovery is too expensive.
One way to rein in the cost of e-discovery is by narrowing the scope of discovery. Under existing Federal Rule of Civil Procedure 26(b), parties “may obtain discovery regarding any nonprivileged matter that is relevant to any party’s claim or defense—including the existence, description, nature, custody, condition, and location of any documents or other tangible things and the identity and location of persons who know of any discoverable matter. … Relevant information need not be admissible at the trial if the discovery appears reasonably calculated to lead to the discovery of admissible evidence.”
The proposed Rules amendments now wending their way through the public comment process would eliminate all of the language after “claim or defense” and graft proportionality language that’s already a part of the rule onto the first sentence. The latter change accommodates lawyers whose attention deficits or narcolepsy prevent them from reading past the first few lines of the current rule. 😉
Whether you support the proposed amendment or not, you should be concerned that it makes no provision for discovery of metadata or for discovery about information systems. Such efforts are not always easy to characterize as “relevant to any party’s claim or defense.” Yet, discovery of metadata and information systems is essential in any case involving ESI (i.e., in each and every case).
Digital is Different
Discovery of ESI presents a need for contextual information rarely encountered in paper discovery. There is an aspect to ESI that reflects both a layer and a penumbra of information that, while not going to the substance of claims or defenses, bears mightily on the integrity and utility of the evidence. These are application and system metadata. Courts must condition their thinking to appreciate that, in the world of ESI, metadata are as important as dates, page numbers and CC:s were for paper correspondence. If we continue to treat metadata evidence as something apart and optional with respect to the evidence it describes, the proposed change in scope may prove problematic.
Accordingly, discovery of metadata supporting the utility and integrity of ESI requires discrete recognition in the Rules, even though such data may not directly relate to claims or defenses. The Advisory Committee should make clear in the Committee Notes that they do not seek to limit the discovery of meta-information that is either a part of the evidence or which materially bears on its integrity or utility.
Reasonable Discovery about Discovery Needs Protection
In a similar vein, the proposed amendment seeks to restrict discovery to matters relevant to any party’s claim or defense” at a moment in history when a growing focus of discovery is the exploration of where and how such information can be found in an efficient, effective and affordable manner. No one likes the notion of “discovery about discovery;” however, measured and managed meta-discovery about ESI is a necessity to reining in overbroad discovery and for attacking the guessing game attendant to e-discovery today. The gatekeepers to evidence are no longer lawyers. The gatekeepers are end users and IT. Asking IT about how and where they store data may not be “relevant to a party’s claim or defense,” but it’s highly relevant to a just, speedy and inexpensive process.
The proponents of these amendments may protest that they expect such meta-discovery to continue unabated; but, weigh such protestations against the language they seek to eliminate: “including the existence, description, nature, custody, condition, and location of any documents or other tangible things and the identity and location of persons who know of any discoverable matter.” The language sought to be struck defines meta-information as relevant. Absent clarification in the Committee Notes or further amendment, cutting such language from the Rule will doubtlessly be raised as a bar to discovery about ESI forms and sources.
Proposed Changes to Rule 30 Shouldn’t Limit Meta-Discovery
There’s also a pending proposal to amend FRCP Rule 30 to halve to five the number of depositions that may be taken without leave of court and to limit each deposition to six hours. I don;t know if five depositions will suffice to shed light on claims and defenses; but, I’m sure five won’t be enough if you have to use them to elicit essential information about information systems and databases from IT personnel and administrators.
To be more efficient in e-discovery and better able to narrowly craft discovery, there must be disclosure of meta-information about information systems and ESI. These are disclosures which many parties refuse to provide unless compelled to do so through discovery. Consequently, the proposed amended Rule 30 limits should not apply to meta-discovery, such as depositions of IT personnel who have no knowledge of claims and defenses but possess crucial knowledge about the storage and forms of ESI. If a company has a different administrator for each database and a party must learn about each database to carefully tailor discovery to each, a party shouldn’t have to choose between merits discovery and meta-discovery. Certainly, parties can seek leave for additional discovery; but, requiring same serves to compound the cost and delay of gaining access to information that forsters efficiency and cost-savings.
We should tailor rules to support the needs of the many diligent litigants who seek discovery in good faith; even when such rules could be exploited by a few who might abuse them. Judges are adept at dealing with abuses, and wield considerable power to rectify actions taken in bad faith. We should trust and encourage judges to do so, just as we should trust counsel not to abuse discovery of meta-information until and unless they have shown they cannot be trusted.
The proposed amendments have much to commend them, and the Committee that drafted them deserves our thanks for fine work. They need our input as well. As I write this it is not clear how one comments on the proposed amendments at this moment (beware of the Rules Committee website as none who venture in emerge unscathed). Though more than one public hearing is likely, the one currently scheduled will take place in Washington, D.C. in November 2013. In the interim, you might send comments by e-mail to Rules_Comments@ao.uscourts.gov and by snail mail to:Committee on Rules of Practice and Procedure
Administrative Office of the United States Courts
One Columbus Circle, NE
Washington, D.C. 20544
Robert Brownstone said:
Excellent article. I too have found the Rules part of the new version of the federal judiciary website hard to navigate. Do you know if 12/1/14 is the target date for implementation of the final version of the proposed amendments?
Thanks for the thought. I understand the proposed amendments will go through a six month public comment period beginning August 15th, 2013. If they traverse the entire approval process, my understanding is that the amendments would be effective in December of 2015, not 2014. But I defer to anyone who knows better.
Pingback: E-Discovery & Evidence » Blog Archive » New Federal Rules Amendments Proposals Published for Comment
Pingback: Subpoenas seeking identifying information and login data associated with email addresses did not violate first amendment or privacy rights | Technology Law Source
Aaron Taylor said:
One descriptive phrase that I have not noticed being used regarding metadata is that it, at least to me, appears to be “hard” data, as opposed to content which can be considered hard or soft data. That is, metadata is factual – it cannot be made up or cloaked in fiction, as far as I can imagine. In a recent Guardian article, the author asked readers to speculate on his identity based on a collection of his messages. Everyone was excited to try to identify him based on what could easily be fictitious comments, people, places as stated. No-one said, “Let me see the metadata.” That would tell them far more about the person, and far more accurately…is he truthful, is what he said accurate and dependable, on and on. Very similar for information, as Mr. Ball mentions, the who/what/where/when and so forth is there to be captured, is accurate and dependable…and useful.
I would be remiss not to note that metadata, being data, can be manipulated, too. But undetectable metadata manipulation is more challenging for those not steeped in digital forensics. Thus, metadata can be made up and “cloaked in fiction;” you just don’t encounter metadata manipulation quite as often or as successfully as you encounter users manipulating the contents of active data items. Thanks for your comment, Mr. Taylor.