Bradley Manning supporters at Ft. Meade during his pre-trial hearing (photo: suemelk)

(update below)

Josh Gerstein of POLITICO has obtained a copy of the Army’s analysis of media coverage of Bradley Manning’s Article 32 or pre-trial hearing. The public affairs staff was tasked with putting together daily reports on coverage. They used a commercial service called VOCUS to “track traditional and social media coverage of Manning’s hearing.”

It was evident from the first day that the media pool was being tracked in the Media Operations Center (MOC). The military public affairs staff sat in a back room behind the screen where the court feed was broadcasting and watched Twitter (and possibly other social media sites) for any evidence that members of the pool were violating rules for the hearing. In particular, they were tasked with making certain nobody sent messages on Twitter from the hearing while court was in session.

This was my first time as a credentialed media person at an event handled by the military. I was so pleased that they had credentialed Firedoglake, a non-establishment media outlet, that I was extremely diplomatic about what I had to say if there was reason to criticize how the military was handling the media. In fact, I mostly let individuals from major networks or newspapers lay into the public affairs representatives if there was reason for criticism because the outlets they worked for meant they could challenge staff much easier than me, if necessary.

For example, one time the feed to the courtroom did not come on. We missed almost ten minutes of testimony. Members of the press were furious and demanded a transcript. Because both the prosecution and defense would have to approve releasing a transcript of the portion we missed, we could not get a transcript immediately. Media had to be briefed by the few press that had actually been in the court room and not the media operations center. This, public affairs representatives said, was “a glitch.” When “glitches” happened (as they did two or three more times), the court proceedings didn’t pause so the “glitch” could be fixed. They just kept on going.

On the first day, the social media director on the public affairs staff asked all the media if they would all agree to use one hashtag. Some were using “#Manning.” Some were using “#WikiLeaks.” (I don’t know if other tags were being used.) She actually asked us all if we were tweeting. In retrospect, it appears she may have been trying to make it easier to track media tweets.

I was nearly ejected on the second day of the hearing. Two military police officers showed up and called my name. I went up and asked about the problem. They said they had my Twitter handle and I had said I was tweeting from the court room. They had read this and gone into panic mode. They looked in the courtroom and nobody was tweeting. So, they left because they figured I might be in the MOC.

The military police officer had written down my name on his hand. When I challenged the accusation that I had ever said anything about tweeting from the courtroom, we went outside the MOC to talk this out. Every time I tried to give my side of the story he would show his hand to me where he had written my Twitter handle, as if that proved I had tweeted from the courtroom. He also said someone was running in and out of the MOC to tweet from outside, which I explained no one was doing.

I kept my cool and explained the feed had been off for almost ten minutes and the media might have been tweeting while the court was in session. The public affairs representatives stepped in and began to explain to the military police what I was saying. That was a possibility because there had been an issue with the feed. [By now, a Los Angeles Times journalist had come out to see what was the problem.]

A military police officer indicated I wasn’t going to be removed but that this was important. I needed to understand I was in the MOC, not the courtroom. While courtroom rules applied to the MOC, there was a huge difference because media were allowed to tweet from the MOC when court was in recess. Wanting to return to the courtroom, as I was missing the proceedings, I conceded I had mixed the two up (even though I had sent no such tweet). I know someone used my handle and said something like #FF @kgosztola who is tweeting the Manning hearing from the courtroom. But, they didn’t understand what I was saying when I suggested that is what caused this hubbub.

Naturally, a number in the media pool wanted to know what had just happened. One member called what had happened “weird.”

I share this to add to the discussion about how the military was monitoring press and to further confirm the fact that the military didn’t want tweets going out while court was in session. That received much media attention. And though the rule was consistent with a rule media have had to abide by during other court hearings, the rule peeved many, especially because the majority of the pool was in a media center and not the courtroom itself.

All of this points to how tightly the military was trying to control the narrative. Another example is the ejection of Lt. Dan Choi from the base on the fourth day of the hearing. This they were very concerned about.

The fourth day’s summary includes this note: “Since former Lt. Dan Choi’s disturbance outside of the courtroom there have been 118 print and online stories; and 67 social network conversations in reference to the incident.”

The military was prepared to bring someone in to brief the media. Separately, one public affairs representative who had been escorting media to and from the courtroom was there to answer questions on whether it was appropriate for Choi to have been at the hearing in his military uniform (which is why he was ejected). He even produced the exact language in the military code that dictates when it is appropriate for a veteran to wear his or her uniform.

Finally, Salon‘s Justin Elliott posts on what VOCUS is designed to do: identify influencers, cover more blog posts, track sentiment and tone, monitor Twitter in near-real time. The analysis by VOCUS showed the media coverage was “negative” though mostly “balanced and factual.” They were able to discern this because VOCUS is capable of  doing analysis, which sifts through words used to ”gauge the feeling” or tone of blog posts and tweets.

Were public affairs representatives trying to pick out who was an “influencer”? If they were watching anyone, they were following Kim Zetter of Wired magazine. She was attracting a lot of attention on Twitter and I believe someone from the public affairs staff approached Zetter once to talk about how much of a response she was getting to her messages.

Though I was getting a huge response on Twitter as well (I picked up more than 1,000 followers because I was covering the hearing), no representative ever engaged me on my coverage.

Update

Not only were tweets being monitored but what the prosecution and the investigative officer (judge) were to say was scripted. Which essentially means, as mzchief notes in the comments thread, “all communications possible were being micromanaged.”