There is a story going round the press and the internet that militant training camp facilitator Abu Zubaida was waterboarded 83 times by the CIA in August 2002. However, this story is not true, as I will show.
The story is based on 30 May 2005 memo by Steven Bradbury, who at that time headed the Office of Legal Counsel at the US Justice Department. The relevant section of the memo reads:
The CIA used the waterboard “at least 83 times during August 2002” in the interrogation of Abu Zubaida, IG Report at 90, and 183 times during March 2003 in the interrogation of KSM, see Id. at 91.
The key phrase here is “at least.” After the blogger emptywheel broke the story, the “at least” was soon dropped and all sorts of media began reporting 83 as the final number of the waterboardings.
When we turn to the relevant sections of the “IG Report” (the CIA inspector general’s report on the agency’s detention and interrogation programme, which was released later), we find how the number of “at least 83” was derived:
OIG reviewed the videotapes, logs, and cables [redacted] in May 2003. OIG identified 83 waterboard applications, most of which lasted less than 10 seconds. [Redacted] OIG found 11 interrogation videotapes to be blank. Two others were blank except for one or two minutes of recording. Two others were broken and could not be reviewed. OIG compared the videotapes to [redacted] logs and cables and identified a 21-hour period of time, which included two waterboard sessions, that was not captured on the videotapes.
So, the number of 83 waterboardings was derived from a review of videotapes that showed some, but not all, of his interrogations. Specifically, they omitted two sessions of waterboarding. Therefore, the real number is “83 plus however many times he was waterboarded in those two sessions.”
So how many times would he have been waterboarded in those two sessions?
The Bradbury memo is confusing. These are the rules it set out:
(1) Waterboarding may be used in two sessions on one day, no session lasting more than two hours;
(2) During a session, it may be applied “up to six times for ten seconds or longer (but never more than 40 seconds);”
(3) Water application in a 24-hour period may not exceed 12 minutes; and
(4) A detainee can only be waterboarded five times in 30 days.
Presumably, this means the agency could waterboard a detainee more than six times in a session, but only six times could the application of water last for over ten seconds. Otherwise, the maths doesn’t add up.
The figure for alleged 9/11 mastermind Khalid Shaikh Mohammed (KSM) is 183 applications of the waterboard technique, although a 2007 International Committee of the Red Cross (ICRC) report said that KSM was only waterboarded at five sessions. This seems hard to believe, as it gives a figure in excess of 36 applications per session (which only lasted about an hour according to the ICRC report) and does not reflect the rules set out in the Bradbury memo, which says the CIA was authorised to use double sessions on five days.
It we propose 183 applications on KSM in 10 sessions on five days, this seems much more credible, and complies with the Bradbury memo methodology. This would mean the ICRC report merely confused five sessions with sessions on five days.
It also gives us an average 18 waterboardings per session. So 83 waterboardings for Abu Zubaida plus two sessions of 18 applications each: 83 + 18 + 18 = 119.
Interestingly, the CIA IG reports comments (page 45):
The Attorney General was informed that the waterboard had been used 119 times on a single individual.
One final point, the Bradbury memo points out that headquarters officials observed the final waterboarding of Abu Zubaida:
… in the Abu Zubaida example, CIA Headquarters dispatched officials to observe the last waterboarding session. These officials reported that enhanced techniques were no longer needed.
I can’t help wondering if that’s the 21-hour period with waterboarding that’s missing from the tapes.