Something caught my eye in James Taranto’s Best of the Web column from yesterday about a CNN interview with Hillary Clinton’s press secretary. Here it is:
This column has been noting for some time that the “55,000 pages†Mrs. Clinton initially turned over to the State Department were not actual emails but printouts of their content—lacking metadata and not electronically searchable.
Here’s my question for someone more knowledgeable than I. Is the metadata that was attached to emails sent from the private server maintained by Sec. Clinton in which Sec. Clinton was carrying on official business government property or not? Last time I checked destruction of federal government property was a felony (18 U.S.C. § 1361).
I don’t know. At the last hearing before Judge Sullivan, the Judicial Watch attorney argued that because the server was set up by Clinton, the server was a “government server,” the person who set up and managed the server was either a “government contractor or employee,” and the whole system was part of “agency system of records.” The State Department lawyer was aghast and characterized this situation as one in which an employee uses a private e-mail account to perform government business and the State Department’s obligation is merely to request copies be turned over.
The Judge did not resolve this issue, but I found it interesting that Judicial Watch appeared to be buying into the possibility that Hillary Clinton had authority to do what she did, though there may be further implications. Or maybe JW wanted to force the State Department to disown her actions. Who knows?
The Judge simply does not know enough to use those terms, but what is metadata. I don’t think it’s a record, it’s probably part of the server, and its probably party of a “system of records.”
I do note that the more recent State Dept. letters requesting public records from former employees specifically ask records be
provided “in original/native electronic format with the associated metadata.”
A little confused by the “printouts . . . lacking metadata and not electronically searchable.” Those printouts could certainly be scanned into a searchable format. Is it a time/cost issue, or is there a substantial loss of accuracy? Frankly, I assume people want access to the metadata for other reasons than superior search functions.
The email metadata (also referred to as the email header) includes information like the dates and times the email was sent and received, the IP addresses of the sender and receiver, whether it had attachments or not and whether it was encrypted or not.
It’s also likely to include something called the message ID which can be pretty useful in tracing the message through the several servers that may have handled it in the course of its delivery. There’s also hop information that can be used for the same purpose.
As to digitizing printed emails, yes it can be done. Whatever you may have read while the character recognition rate of optical character recognition may approach 99% the message recognition rate is around zero. Think of it this way. In a 100 character email the odds are you’ll have one error which means that the message was not recognized. As soon as you introduce human readers you introduce other problems which I presume would be used to attack evidence in a court of law.