On Friday, I noted that the New York Times had dutifully repeated information from military sources who had provided them with a “classified” report (pdf) on how cultural differences between NATO troops and Afghan troops are resulting in increasingly frequent killings of coalition troops by coalition-trained Afghan troops.  On Friday morning, the Times put up a correction, noting that the Wall Street Journal had published an article about the May 12, 2011 report on June 17, 2011.

I mentioned in my Friday post that the Wall Street Journal article included a link to what was said to be a copy of the report, but that the link was now dead. It is quite curious that the Journal article would have that link, as the opening sentence mentions that the report is classified. In comments on the post, Marcy Wheeler posed the question of whether the study “was intentionally buried after the WSJ story? Maybe that’s what NYT’s claim that it is classified is about?” So, in other words, was the study retroactively classified because of the Wall Street Journal article?

With only a little searching after reading both the New York Times and Wall Street Journal articles, I found what appeared to be a complete copy (pdf) of the same report (or at least a copy with the same title and number of pages), clearly stamped “UNCLASSIFIED” at the top and bottom of each page. Several hours after my post was published, the Times added a second correction to their story:

The article also referred incompletely to the military study’s secrecy. While it was classified, as the article reported, it was first distributed in early May 2011 as unclassified and was later changed to classified. (The Times learned after publication that a version of the study has remained accessible on the Internet.)

So it turns out that Marcy’s hunch was correct. The report initially was published as unclassified and then later classified, in a clear case of retroactive classification. There is perhaps just a hair of wiggle room in the Times’ statement that “a version of the study has remained accessible on the internet”, providing for the remote possibility that there are differences between the “classified” version provided to the times and the complete version on the internet, but that seems highly unlikely. The copy on the internet is almost certainly a copy from the time period when the study clearly was unclassified.

This sequence of events also is confirmed somewhat in the Wall Street Journal article itself:

The study was originally unclassified, but military officials in Kabul said Thursday that it has been recently classified “secret” by the U.S. Central Command in Florida at the request of coalition officials in Afghanistan. On Thursday, despite its new classification, the report was available on a publicly accessible military knowledge-sharing website.

The Journal’s use of a dead link, however, would lead a current reader to believe that even the “publicly accessible” version was no longer public, making their discussion of classification difficult to parse.

The publication date of the Journal article last June is a Friday, so it seems possible that Central Command decided to classify the report in response to inquiries about it as the Journal neared publication. On Saturday, January 21, I requested comment from a press contact in Central Command with whom I have previously corresponded, but have not yet gotten a reply.

Today, Marcy included me in an email conversation with J. William Leonard, who previously served as the Director of the National Archives’ Information Security Oversight Office and before that as Deputy Assistant Secretary of Defense for Security and Information Operations. The question posed to Mr. Leonard was whether the retroactive classification of the report was properly carried out. Leonard’s response noted that since “the purpose of classification is to preclude unauthorized disclosure”, that is “a metaphysical impossibility for information whose disclosure was authorized in the first place.”

So why would the government try to retroactively classify the report? In this case, the first explanation that comes to mind is that the report is embarrassing to NATO (primarily American) troops with the litany of ANSF complaints contained in the report. In other cases, as I will note below, the government has used retroactive classification as a tool in either silencing or prosecuting whistleblowers.

Here is more of Leonard’s response on the issue of retroactive classification:

Fortunately, from a policy perspective, there are no direct provisions to retroactively classify something that was unclassified and was properly put into the public domain which is what DoD did in this case when, as the WSJ article states: “the report was available on a publicly accessible military knowledge-sharing website” which from all appearances is a DoD sponsored website.

First, to retroactively classify such a document defies common sense.  Second, the purpose of classification is to preclude unauthorized disclosure, a metaphysical impossibility for information whose disclosure was authorized in the first place.  Finally, to do so can also undermine national security because even if the information is truly sensitive, the government simply draws increased attention to the information by such ham-fisted actions.

Looking further into the general issue of retroactive classification, we see that it was used to silence Sibel Edmonds in 2004. There is also a stern letter from Henry Waxman (pdf) to Donald Rumsfeld on retroactive classification of documents earlier that same year.

In this post on the blog for the Project on Government Oversight, we see discussion of retroactive classification in the government’s prosecution of Thomas Drake:

The document that relates to one of the counts of violating the Espionage Act that Drake is charged with was not even classified when it was in his possession. “In support of its willful retention charges, including Count Two, the government alleges that “[c]lassified information had to contain markings identifying the level at which it was classified,” according toa motion filed by Drake’s lawyers to dismiss Count Two. But, but, but…

Evidence recently produced by the government reveals that the allegedly classified “Regular Meetings’ document contained clear ‘markings’ that it was an ‘unclassified” document. According to a March 22, 2010 memorandum prepared by the lead NSA investigator in this case – which was produced to the defense just three weeks ago – the allegedly classified “Regular Meetings” document was posted on the National Security Agency intranet, called “NSANet,” and it was marked “UNCLASSIFIED//FOR OFFICIAL USE ONLY” in the header and footer.

Further along in the post, POGOBlog notes that retroactive classification also was used against Franz Gayl and Robert MacLean.

In a particularly ridiculous exercise, the government also prevented Carol Rosenberg from mentioning Joshua Claus’ name when she was reporting on testimony about his crimes at Guantanamo, even though she was able to point out that his name was publicly available and tied to the events being reported.

What is particularly galling about attempting metaphysical impossibilities in retroactively classifying material is that Leonard has pointed out that improper classification is a violation that should be treated on an equal level with improper disclosure:

Classifying information that should not be kept secret can be just as harmful to the national interest as unauthorized disclosures of appropriately classified information.

In fact, the executive order governing classification treats unauthorized disclosures of classified information and inappropriate classification of information as equal violations, subjecting perpetrators to comparable administrative or other sanctions in accordance with applicable law.

Don’t hold your breath, though, if you expect a prosecution, firing, or even a reprimand for anyone attempting metaphysical impossibilities when it comes to government officials retroactively classifying information.