Early last September, attacks on U.S. forces in Iraq were spiking and an Army general dispatched from a military prison at Guantanamo Bay, Cuba, concluded in a classified study that the detention of Iraqis at the Abu Ghraib prison near Baghdad "does not yet set conditions for successful interrogations."
Under pressure to extract more information from the prisoners -- to "go beyond" what Army interrogation rules allowed, as an Army general later put it -- the senior U.S. military commander in Iraq sent a secret cable to his boss at U.S. Central Command on Sept. 14, outlining more aggressive interrogation methods he planned to authorize immediately.
Defense Secretary Donald H. Rumsfeld, left, confers with Lt. Gen. Ricardo S. Sanchez, head of U.S. Central Command in Iraq, on a helicopter over Baghdad.
(Pool Photo David Hume Kennerly)
The cable signed by Lt. Gen. Ricardo S. Sanchez listed several dozen strategies for extracting information, drawn partly from what officials now say was an outdated and improperly permissive Army field manual. But it added one not previously approved for use in Iraq, under the heading of Presence of Military Working Dogs: "Exploit Arab fear of dogs while maintaining security during interrogations."
Sanchez's order calling on police dog handlers to help intimidate detainees into talking -- a practice later seen in searing photographs -- was one of a handful of documents written by senior officials that Army officials now say helped sow the seeds of prison abuse in Iraq. They did so, according to an Army report released Wednesday, by lending credence to the idea that aggressive interrogation methods were sanctioned by officers going up the chain of command.
But the issue of using dogs is also an example of how the U.S. military's ad hoc and informal decision-making in Iraq created confusion and allowed these harsh methods to infiltrate from Afghanistan to Guantanamo and finally to Iraq, despite Bush administration contentions that detainees in each theater of conflict were subject to different rules and that Iraqis would receive the most protections.
The text of the Sanchez cable was not included in public copies of the Army's report, but was obtained by The Washington Post from a government official upset by what Sanchez approved.
The authors of the Army report did not accuse Sanchez of directly instigating abuse, and they did not cite the contents of his memo in the unclassified version. But Army Gen. Paul J. Kern -- who oversaw the drafting of the report -- said in an interview last week that Sanchez "wrote a policy which was not clear," and that by doing so, he allowed junior officers to conclude mistakenly that they were following an official policy as they stepped over a legal line.
This interpretation of the role senior officials played -- that they committed sins of omission, rather than commission, by writing ambiguous instructions and then failing to police the errant ways of subordinates -- is likely to be challenged in court, according to lawyers for some of the soldiers on trial in connection with the prison abuse.
No one above the military grade of the top intelligence commander at Abu Ghraib was legally "culpable" for the abuse, the Army report concluded. But a separate report on the abuse released Wednesday by a panel appointed by Defense Secretary Donald H. Rumsfeld referred to Sanchez's memo on Sept. 14 as one of several documents that led "some soldiers or contractors who committed abuse" to believe "the techniques were condoned."
Other such documents cited by officials who participated in the two probes include a December 2002 memo signed by Rumsfeld that authorized harsh interrogation methods for prisoners at Guantanamo, and a controversial Feb. 7, 2002, memo signed by President Bush that declared that fighters detained in Afghanistan were not entitled as a matter of law to the protections afforded by the Geneva Conventions.
The Rumsfeld memo included authorization for the use of dogs; the Bush memo was cited by legal advisers to Sanchez as the basis for their determination that some Iraqi detainees were not entitled to the full legal protections provided by the Geneva Conventions, according to the independent panel. This "confusion" between interrogation rules devised for use at Guantanamo and Afghanistan and the protections mandated by international law in Iraq contributed to some of the abuse, according to the Army report's executive summary.
Kern said: "We found not culpability" among senior officers such as Sanchez, but "clear responsibility" for not deterring junior officers and enlisted men from inappropriate behavior. "They didn't clarify for those young interrogators what their responsibilities were."
Several abuses in particular are highlighted by the two reports released last week: the use of dogs to frighten detainees, the repeated stripping of detainees, and the use of extended isolation and sensory deprivation. Each clearly violated Army rules and violated Geneva Conventions that protect civilians under military occupation from threats of violence, isolation from visits by the Red Cross, and humiliating and degrading treatment, the Army report said.
The issue of using military dogs illustrates how a blizzard of memos from senior officials sowed an impression of tolerance, if not approval, for aggressive interrogations. It has been a particular embarrassment to the Pentagon since photos of dogs snarling and barking in front of cowering Iraqis -- and in one case preparing to bite a detainee -- were made public in June, about six months after soldiers there recorded the images.
It also illustrates how, as the independent panel's report concluded, the migration of lists and interrogators from one theater to another resulted in "policies approved for use on al Qaeda and Taliban detainees, who were not afforded the protection of the Geneva Conventions, [being] applied to detainees who did fall under the Geneva Conventions."
Army investigators probing the abuse in Iraq traced the initial idea of using dogs -- a technique that does not appear in the service's standard field guide -- to interrogation practices followed by U.S. intelligence officials and Special Forces teams deployed in Afghanistan. Kern said the officials there concluded that Afghans feared dogs because of religious beliefs that those bitten are unhealthy or condemned, and became convinced that this fear could be exploited to compel intelligence disclosures.
The technique migrated first from Afghanistan to Guantanamo Bay, via Washington. In late 2002, aides to Rumsfeld -- responding to a request by officials at Guantanamo for approval of more aggressive interrogation methods -- canvassed officers in Afghanistan and elsewhere. On Dec. 2, Rumsfeld approved techniques for use only at that site, which included "the use of dogs to induce stress and the removal of clothing as Counter-Resistance techniques," according to the Army report.
Rumsfeld rescinded his memo the following month, after a private protest by Navy general counsel Alberto J. Mora over its sanctioning of practices in violation of international law and military regulations. The independent panel's report faulted Rumsfeld for not obtaining "a wider range of legal opinions and a more robust debate" before he approved the rules. It also said his promulgation of these guidelines -- even temporarily -- contributed "to a belief that stronger interrogation methods were needed and appropriate."
By April, after a Pentagon review, Rumsfeld approved a new list of interrogation techniques that omitted the use of dogs. But U.S. Special Operations forces in Afghanistan, meanwhile, continued to use many of the practices on Rumsfeld's Dec. 2 list, including "isolating people for long periods of time, using stress positions, exploiting fear of dogs, and implementing sleep and light deprivation," the Army report concluded.
U.S. military commanders there urged the removal of clothing on grounds that "no specific written legal prohibition existed." The Pentagon has not released details of abusive Special Forces activities in Afghanistan. But the independent panel said an unreleased Defense Department report has found "a range of abuses and causes similar in scope and magnitude" to those involving interrogators at Abu Ghraib.
In Afghanistan, these tactics were also employed by members of the Army's 519th Military Intelligence Battalion, a unit transferred to Iraq in the summer of 2003. After Maj. Gen. Geoffrey D. Miller, the top official at the Guantanamo prison, visited Abu Ghraib from Aug. 31 to Sept. 9 and called for more rigorous interrogations there, some of these tactics -- including the use of dogs -- were incorporated in a memo drafted by Sanchez's legal office on Sept. 10 and sent to prison interrogators.
Sanchez's legal advisers subsequently drew on both this guidance and the legal justifications in Bush's 2002 directive while drafting the Sept. 14 cable from Sanchez to Lt. Gen. John P. Abizaid, the head of U.S. Central Command, the independent panel's report said.
"Enclosed is the policy modeled on the one implemented for interrogation conducted at Gitmo," Sanchez said in his cable, referring to Guantanamo Bay. It authorized not only exploiting prisoners' "fear" of dogs but also the use of isolation; "sleep management"; "yelling, loud music, and light control . . . to create fear, disorient detainees and capture shock"; deception, including fake documents and reports; and "stress positions," such as forced kneeling for as many as four hours at a time.
The cable placed no restrictions on the use of dogs on "detainees" and "security internees," but said any use involving enemy prisoners of war would require Sanchez's direct approval. In fact, as Maj. Gen. George R. Fay, an intelligence official who co-wrote the Army report, said in an interview last week, the use of this narrow qualifying phrase in Sanchez's memo reflected bad "staff work" by the lawyers who drafted it for Sanchez's approval, because U.S. military forces "did not have very many enemy prisoners of war at that point."
Within one month, Sanchez's cable was rescinded on instructions from senior officials at U.S. Central Command and replaced with a more cautious memo that allowed the use of muzzled dogs during interrogations only when Sanchez gave his direct approval -- something he told investigators he was never asked to do.
His new memo was based in part on an outdated 1987 version of the Army Field Manual for interrogations, which was more permissive than the 1992 version then in effect because it allowed complete control of light, heat, food, clothing and shelter as interrogation techniques, the Army report concluded. Investigators attributed this error by Sanchez's office to the Army's failure to update a key Web site with the 1992 report.
But whatever Sanchez's intent or policy, the practice of "abusing detainees with dogs started almost immediately" after the Army, acting at Miller's urging, brought several dog teams to Abu Ghraib in November 2003.
The fact that at least three "confusing and inconsistent" interrogation directives were approved within a month-long period "contributed to the belief" that illegal interrogation techniques were condoned, the Army report stated. An absence of leadership and oversight also left room for what the Army report described as "word of mouth" techniques to be passed around and followed by interrogators deployed to Iraq.
The Army report quoted Sanchez as saying he "never approved use of dogs." Fay also said in the report that "no documentation was found" showing approval by the Combined Joint Task Force 7, headed by Sanchez, "to use dogs in interrogations."
Asked to explain the apparent conflict between language in the report and the text of Sanchez's cable, Kern said that what Sanchez meant is that he never specifically approved an interrogation plan submitted to him for review that involved the use of dogs, while Fay said that Sanchez believes he only endorsed the general presence of muzzled dogs at the time interrogations were being conducted, rather than inside prison interrogation booths -- a practice that was clearly misunderstood.
Col. Thomas M. Pappas, the senior intelligence official at Abu Ghraib, told Army investigators that Miller, in addition to Sanchez, had authorized the use of dogs to "set the stage" for productive interrogations. But the authors of the report accepted Miller's contrary contention that he only recommended using dogs for detainee custody and control at Abu Ghraib. Miller is the head of U.S. military detainee operations in Iraq.