Dogs of Detention, from Abu Ghraib to Camp Bucca and Camp Delta

There is probably no image more associated with the abuse of prisoners at Abu Ghraib in 2003 and 2004 than the photo of Specialist Lynndie England holding what appears to be a dog leash attached to a naked prisoner.  Investigations into the abuse revealed that at least some of the five Military Working Dogs (two Army, three Navy) working at the prison were used to intimidate and frighten the prisoners.  Some prisoners were apparently bitten by the dogs. The 2010 court martial of a dog handler was described here in a prior blog. 

The recent release of over 100 U.S. detainee policy documents by Wikileaks might be expected to bring further light on abuse involving canines, but a search through all of the records so far posted shows no evidence that this is the case.  In fact, it appears that deployment of Military Working Dogs to detention facilities may have been cut back to such an extent that dogs are no longer available for some of the legitimate functions they might be expected to perform at such locations.    

Camp Bucca, Iraq, Standard Operating Procedure

Camp Bucca is a detention facility in southern Iraq near the town of Umm Qasr.  Its primary mission is to handle overflow from the Baghdad Central Confinement Facility (Abu Ghraib).  Among the documents posted by Wikileaks was the Camp Bucca, Iraq, Standard Operating Procedure (105 Military Police Battalion, October 30, 2004), labeled for official use only.  The Procedure goes into detail about how Military Working Dogs are to be used at a detention facility.  

Since the Abu Ghraib prosecutions usually referred to the general Army Policy (190-12) on canines, this document’s specific concerns with detention operations is worth detailed analysis.  Portions of the Camp Bucca documents appear adapted from Camp Delta (Guantanamo) documents that were not part of Assange’s release. The Procedure is described in detail in the appendix at the end of this piece. As can be seen from our parenthetical comments, the Procedure is not always well written or considered. 

Abu Ghraib

Just a small part of one deposition of a Navy canine handler at Abu Ghraib will be sufficient to show how some of the handlers were either tricked or ordered into ignoring standard procedures:

“I have a Belgium Melon [Malinois]. He is a control [patrol?] explosive dog. My dog had never been in a prison environment. The dogs didn't have to be retrained we used the same commands.

“On the night of 24 NOV 03 we were with the IRF [Internal Reactionary Force] when we received a call to search the hard site for explosives. We had heard also that a prisoner ... had been shot. All 5 dogs were there that night. The army dogs were for security while our dogs searched for explosives. We went to specific Tier, I couldn't tell you the name, but I could show you if I were in the facility. We searched the cells for explosives, none of the dogs responded. We were preparing to leave when we were told we needed to search another cell. I decided to conduct the search. I reentered the Tier and saw two individuals who were dressed in green BDU's [battle dress uniforms]. They instructed me to search the cell and as I approached the cell I heard a lot of shouting and screaming. I looked around and noticed that there was no one else up there; that no other cells were occupied.

“I [peeked] my head in the cell and noticed four individuals, three males and one female. One male was in the corner, two males were crouching, and the female was next to the rack. I was having a hard time controlling my dog because of the noise, and the actions that were happening. I finally got my dog under control when on[e] of the MI guys told one of the detainees, "If you don't tell me what I want to know I'm gonna get this dog on you", or words to that effect. I realized that it wasn't a search and they actually wanted my services for interrogation. I exited the cell, but my dog would break the plane of the cell. The two individuals and the female, which I know now to be an interpreter, reentered the cell and went back in a started yelling and screaming at this guy again.

“With all the yelling and screaming going on my dog breaks my control and charges the cell. The cell was very dark; the only illumination was a pin light. I see the female wincing, and I notice my dog on her arm. I immediately call my dog off of her, and my dog comes back to the heel. I ask the female to come out so I can check her. She comes out of the cell and I constantly ask her, ‘Are you alright, did he bite you??’ She says she is ok and that she wasn't bit. I looked at her arm no bite mark, no blood, and her DCU uniform wasn't torn. I once again ask her if she is ok, and does she need medical attention.  At that point I see SGT [redacted], and I ask what’s going on, and he responded by saying he didn’t know.  Once I realized what was going on I left, I didn’t want any part of what was going on.  As I was leaving, SGT [redacted] followed and then I heard someone say, ‘where’s the dog, where’s the dog??’”

(ACLU FOIA files DOD000679, DODDOA002964) For other accounts of incidents involving dogs, see The Torture Papers: The Road to Abu Ghraib, edited by Karen Greenberg and Joshua Dratel. 

Army vs. Navy

The deposition of the Navy canine trainer has some partially redacted language that is disturbing, suggesting that the Army and Navy handlers did not work well together.  The Navy handler giving the deposition stated that he raised questions “on what we could and could not do in this environment, but we never received straight answers.” He undertook to develop a standard operating procedure for canine operations at Abu Ghraib.  He states, “I know *** tried making the Army Handlers follow the SOP and I personally gave *** two copies of the SOP, but like I said before *** didn’t want to be part of a team…. *** set up a wall between the Army dog handlers and us.  *** stated that the Navy was basically dipping into his Kool-Aid.  I sit down with *** to try to work as one team, but *** didn’t want to work as a team.  We work for the IRF commander, and the Army dogs work for the compound.” 

The meaning of the last sentence is unclear, but may suggest that the Army handlers were more amenable to using their dogs for intimidation.  Perhaps the reason for the apparent lack of cooperation was fathomed by Major Denzil Frost in a 1990 master’s thesis prepared for the U.S. Army Command and General Staff College.  In the thesis, A Centralized Source of Information for the Military Working Dog Program, Frost argued:

“The goal of management should be to facilitate mutual support and cooperation between the various phases/sections that produce MWDs.  Effective management creates an environment that allows the organization to meet the challenges to solve its problems.  Ineffective management generates rivalry, jealousy, parochialism, and disunity.  In this age of daily change, capability lies in flexibility.  Fragmented executive authority, coupled with the absence of a centralized source of information on working dog production leads to the compounding of bureaucratic inertia that is built into the current MWD program.”

This observation was made in 1990, but it was true at Abu Ghraib, and may be true today. As we have described before, the FBI and ATF have never seen eye-to-eye on the training of explosives detection dogs, and it is not surprising that the same is true of other federal agencies. 

Riot at Camp Bucca

Camp Bucca Layout (courtesy Zone Interdite)

An incident at Camp Bucca in November 2004, which did not involve any dogs, reads like an out-take from a Kathryn Bigelow movie.  The incident is in a memo posted by Wikileaks involving an attempted escape:

“While walking around the exterior fence of Compound 4, SGT Thorne noticed several detainees running to the cooling tent in the back of compound 4 and crawling under the southwest corner.  SGT Thorne then notified SSG Fischer of the situation where he rushed over to verify.  After verification SSG Fischer and SSG Dalton entered into Compound 4 to clear all detainees out of the cooling tent.  While SSG Fischer and SGT Dalton were walking up the east side of the compound Tower 2-4 noticed detainees grabbing at the corner of the tents pulling out stakes and bamboo sticks.  Since there were Compound Control Team members inside the compound SPC Montanez yelled for the detainees to put the weapons down and when they didn’t SPC Montanez fired one point shotgun round of non lethal munitions at the crowd.  The detainees were out of range, but they did start running toward the cooling tent. 

“When SSG Fisher and SGT Dalton came around the Northeast corner of compound 4 they noticed a large group of detainees headed toward them swinging tent poles and stakes.  SSG Fischer yelled for help from the towers.  SPC Eakes in Tower 2-5 responded by shooting seven shotgun rounds of non lethal.  Four rounds of point and three crowd dispersal shotgun rounds of non lethal munitions.  Three rounds hit one detainee in the chest, but the detainee was not apprehended and did not seek medical assistance as he took off into the compound.  SSG Fischer and SGT Dalton were escorted along the perimeter of the compound by a rover team.  As SSG Fischer and SGT Dalton crossed the shower area and in front of the holding area SPC Sommers fired two shotgun rounds of non lethal at the crowd following them  One round of point and one crowd dispersal non lethal munitions in order to help the CCTs safely escape.  The crowd of detainees was throwing rocks, poles, and anything else they could get their hands on at the two CCTs in the compound. 

“While everything was going on inside compound 4, 20-30 detainees rushed the vehicle gate of compound 4 and pushed it open.  SGT Thorne called for help and the rovers came immediately to close the gate.  Roughly 20-30 detainees escaped out of the vehicle port and headed to the southern exterior perimeter fence.  As the rovers were chasing after the escapees, 10-15  detainees started turning back and heading back to compound 4.  As they arrived at compound 4 SGT Thorne and SFC Schrock organized them by the vehicle port of compound 4. 

“When SPC Heid and SPC Sommers responded to compound 4 to assist SGT Thorne in flex cuffing the group, they noticed the detainees attempting to get up and move away.  SPC Heid yelled for them to stop and sit back down.  The detainees ignored him and started moving away.  After yelling multiple times at the group of detainees to stop, SPC Heid fired four point and one crowd dispersal non lethal shot gun rounds at the moving group of detainees.  Two point rounds did hit the detainees, but the other two pint and one crowd dispersal non lethal rounds, there was not a positive report they were hit.  There was not a positive identification on this group since they turned back and forced their way back into the compound through the same vehicle gate they came out of. 

“After the gate was secured the second time SFC Shrock parked his Humvee up against the gate for extra security. During the escape attempt SSG Signorella from compound 10, chased after and caught two detainees.  As he was walking the two detainees back to compound 4, SSG Signorella noticed three detainees running alongside the west side of compound 10, which he yelled for them to stop.  The detainees kept going so SSG Signorella fired one crowd dispersal shotgun round of non lethal munitions.  One detainee was hit for sure, but the other two kept going.  Detainee 155786 was hit with four pellets.  Three in the upper thigh of the left leg and one on the tip of the nose. 

“Because of everything going on, detainee 155786 was taken to medical instead of being treated at the compound.  Also while chasing the escaping detainees SGT Grubbs fired two point rounds of non lethal shotgun munitions at two detainees climbing up the southern outer perimeter fence.  Detainee 158798 was hit by one of the point rounds, but did not need medical assistance.  The other round did not appear to hit the other detainee.”

The fact that dogs might have secured the perimeter better than non-lethal shotgun blasts, which seem to have had little deterrent effect, is worth noting.  Also, dogs would have been useful in finding the detainees who escaped.  They may have been used, but their presence was not noted.  The policy of not deploying dogs except as an alternative to lethal intervention, described in the Camp Bucca Standard Operating Procedure, is apparently being followed. 

Conclusion

Camp Bucca at Night (original source unknown)

The job description of Military Working Dogs assigned to detention facilities is perhaps closer to that of the ancient war dog than any other assignment given to modern military dogs.  They are assigned to guarding, patrol, and tracking duties, but also to crowd control.  Make crowd control more aggressive and one has the war dogs the conquistadors brought to the New World. In one of the Abu Ghraib investigations, an assessment of an incident that occurred on August 23, 2003, stated the following:

“MP dog handlers cooperated with MI interrogators under the MPs’ watch to use dogs to frighten, intimidate, and even bite detainees…. MP dog handler SG [redacted] was disrespectful and racist (he said, ‘After working at the prison for so long, the dogs came not to like Iraqi detainees.  They didn’t like the Iraqi culture, smell, sound, skin tone, hair color, or anything about them.’)”

War dogs always did best when the enemy wore different clothing, had a different skin tone, and a different diet than the members of the army they served. This is not a matter of what they like, but whom they are trained to attack. In the heat of battle they were less likely to turn on soldiers of their own forces if the enemy was distinguishable by multiple features.  The prejudiced statement of the unidentified sergeant is a projection of his or her racist perspective, but nevertheless captures the essence of why dogs could be so effectively used in detention facilities in the Middle East.

Although the training materials released by Wikileaks are not well written or correct in some respects, we want it noted that we believe that many MWD handlers receive additional information from other sources and do not fail in some of the ways that might occur if the formal procedural materials were all that governed their actions.   

We looked at the State Department documents released by Wikileaks almost two years ago. The references to dogs in those documents largely concerned State Department support of canine operations for generally friendly governments.  Some governments were found to be abusing dogs or not using them for what they had been trained to do.  Still, there was not too much for anyone to be embarrassed about.  There is even less in the latest batch of Wikileaks releases.  The military seems to have learned its lesson.  The evidence is that dogs are not now part of abusive interrogations.  The criminal penalties given to the military staff at Abu Ghraib has cut deep.  Still, it may have made some commanders too shy. 

One also has to wonder if the Army policies that are so restrictive of service dogs on military bases might in part be the result of the bad press that some commanders feel was attended on them from Abu Ghraib.  It is time to move on.  

APPENDIX: Camp Bucca Standard Operating Procedure 

The Standard Operating Procedure includes a number of general procedures for handlers.  Handlers are to “take up a short leash upon approaching anyone with their dog to allow a safe distance between their dog and the personnel in the area.” Generally handlers are to “keep their dogs on leash.”

“As a rule, there should be at least fifteen feet between dogs or between a dog and another person. When it becomes necessary to approach other dog teams, dogs must be held on short leashes. A greater distance must be allowed during break periods because this is the time for the dogs to romp and play at the end of the leash.”

A handler should not sit or lie down when accompanied by a dog “because he would be in an extremely awkward position to control the dog if it suddenly lunged.” (Actually, it’s quite easy to control a well-trained dog from a sitting position.)

The handler should not leave a dog unattended in a patrol vehicle “for an extended period of time.” (This needs to be more specific. As we noted in a prior blog, a dog can die in 20 minutes if a vehicle is sitting in the sun on a hot day.)

“The handler will not tie his dog to any object with the leash for any extended period of time since the dog may chew through the leash, break it, gain freedom and perhaps cause injury to itself or others.” Also, dogs are not to be staked out unobserved or left unobserved with a muzzle on.   “The handler will never tie his dog to a vehicle or other movable object.  The dog might receive a serious injury if the vehicle is moved.”

Physically harmful behavior to the dog is not completely prohibited.  “Personnel should never turn their backs on dogs.  Never kick, slap or hit a dog, except when directed in training and as a training aid.” 

Kennels

Standard Kennel Building Shell (AR 190-12)

The Procedure specifies that at “NO TIME will any unauthorized personnel be allowed access to the kennel facility.”  Handlers are to refrain from “running or engaging in any type of horseplay in or near the kennel area.”  The concern is that such horseplay “can create a situation wherein a dog might break out of its kennel and/or cause injury to itself, a person or another dog.” 

If a loose dog is noticed in the kennel area, “the first person to notice the animal must give the alarm ‘LOOSE DOG’ and everyone in the area, except the dog’s handler, must freeze.” The handler is to secure the dog. 

Stopping a Dog Fight

The Procedure gives detailed information about breaking up a dog fight:

“[B]reaking up a dog fight is a two-man job; no one should attempt to accomplish it alone. Fighting dogs should never be pulled apart; pulling them apart may cause a ripping or tearing of the flash and/or disable the dogs. In breaking up a dog fight. the handler should:

(a)    Keep his leash taut as he gradually works his hand toward the snap of the leash.

(b)   Firmly hold the leash one inch from the snap with his strong hand and slip his weaker hand underneath the dogs' neck.

(c)    Grab the throat of the dog with the weaker hand at a point below the lower jaw.

(d)   Choke the dog until the air supply is cut off, thus forcing the dog to release its hold.

(e)   Repeatedly command the dog "OUT!" during the entire procedure."

This seems to accept a poor level of training. A well-trained dog should not engage another dog and two dogs being handled by two handlers on lead can surely be kept apart.  While cutting off the air supply to regain control in a fight may be used in extreme circumstances, it seems a draconian response to a situation that dogs and handlers should be trained to prevent in the first place.

The Procedure assures that an “alert handler can avoid being bitten by his own dog.”  (One would hope.)

Veterinary Treatment Facilities

The handler “must always be alert and prepared to control his animal at all times while medical care is being given….  A dog must always be kept a safe distance from other dogs, handlers and vet personnel. Dogs must be muzzled and kept on a short leash at all times while in the vet clinic.” (It’s not clear how a dog that is taken to a veterinarian for medical care can always be kept a safe distance from medical personnel.  This may just be poor writing.)

Other personnel are advised not to stare at MWDs, not approach them suddenly or aggressively, avoid horseplay around them, and generally that they “are not pets and are limited in the amount of socialization they receive.” (This also assumes, and perhaps accepts, insufficient training.  Dogs should be under control and should not attack people or other dogs who look at them.)

MWD teams always have the right of way.  “If you cross paths with a MWD team ensure you give the handler at least 10 feet of distance between you and him.”  

Patrol Procedures

Handlers “must warn personnel who attempt to approach their dog that the dog will attack you without command.”  (Dogs, if trained to attack, should do so only on command, so this further indicates a lack of understanding of appropriate canine methodology.) 

“Handlers will ensure that when patrolling with the dogs in the front of the vehicle that the right side (passenger) window is rolled up sufficiently so that the dog cannot escape or jump from the window while riding in the vehicle.”

The Procedure states that the primary function of patrol dogs “is detecting unauthorized personnel and warning the handler. Once the handler has been alerted, it becomes the responsibility of the handler to cope with the situation in the most appropriate manner.”  A secondary function is “to pursue, attack and hold any escapee.  Releasing the dog constitutes the application of physical force and is governed by use of force requirements.”  (This indicates protocols accepting of the “bite and hold” approach, which may explain why this is still the primary training method in wide use by civilian law enforcement today even though “bark and hold” may be the preference for agencies, if only to reduce potential liability.)

Before a dog is released, all lesser means of force must have been attempted “when it is reasonable to do so.” The handler is “to apply the minimum amount of force necessary to prevent escape. Before releasing an MWD the handler will give the command ‘Halt’ three times.” 

The Procedure states that a MWD team is most effective on foot and that “a patrol dog used during daylight hours provides a good psychological deterrent to escape.”  Patrolling is also appropriate at night and during periods of reduced visibility, presumably e.g., sand storms. 

“The K-9 NCOIC [non-commissioned officer in charge] will ensure at a minimum there is a dog team on duty every night from 2000—0400. When both handlers are working, one team will randomly work 8 hours between 0400—2000. Handlers are encouraged to occasionally have dogs bark, and conduct aggression training within the IF [Internment Facility] area to demonstrate MWD capabilities. It is recommended that handlers do both foot and vehicle patrols within the IF area.”  In an incident at Camp Bucca described below, this procedure should have assured that a dog would be available to deal with a riot. 

Nevertheless, dogs cannot go everywhere and MWD's are to “stay out of inner perimeter of compounds unless doing riot/crowd control, extraction/protection of personnel or a command directed search of the tents for contraband.”

Guard Duty

Helicopter Seating Charts (Camp Bucca Procedures)

MWD teams may be posted around holding areas and processing centers to prevent prisoners from escaping.  They may be used to guard detainees while in transit, and assist in the search for escaped prisoners.  When taken on helicopters transporting detainees, they will be near the guards, whose seating placement is indicated in one of the documents released by Wikileaks. 

Crowd Control

Using dogs in crowd control must be determined to be “absolutely necessary by the responsible commander [and] direct confrontation with demonstrators is not recommended.”  It is to be considered “only as an alternative to the use of deadly force to gain control of the situation.”  Specific procedures are the following:

(a)    During peaceful stages of confrontation MWD teams will be held in reserve, out of sight of the crowd. As the situation worsens, dog teams may be moved forward to within sight of the crowd, but away from the front lines.

(b)   When on the front lines dogs are kept on a short leash and allowed to bite only under specific circumstances authorized by the responsible commander. Dogs will never be released into a crowd.

(c)    Other riot control personnel should be positioned approximately 10 feet from MWD teams to avoid unintentional injury.

(d)   MWD may be used to help apprehension teams catch and remove specific individuals in a group of detainees. In this role, the dog team is used only to protect members of the apprehension team.

Using short leashes in bite situations, as with riots, would seem to put handlers unnecessarily in harm’s way.  It is not difficult to see why dogs at Abu Ghraib might have been easily used for intimidation in situations that did not involve extractions.

If a handler is hurt or unconscious, another handler is to be notified immediately to take control of the MWD. 

Tracking

Specific directions are given with respect to tracking dogs:

“When the need for tracking arises, personnel who are on scene should avoid the area and keep other personnel from entering the area to reduce contamination of the area. MWD's will always precede the search party: the search party should maintain a sufficient distance from the dog team to reduce distractions.”

Dogs Not Present in Interrogations

The Procedure states that Military Working Dogs are not to be used “to coerce or menace detainees during interrogation, or any other similar circumstances.” 

This was emphasized in many other places among the Wikileaks documents since the horrors of Abu Ghraib were made known.  A policy distributed by Headquarters, Multi-National Force-Iraq, on January 27, 2005, states:

“Under no circumstances will the following interrogation techniques be approved or utilized: sleep management, stress positions, diet manipulation, environmental manipulation, removal of clothing, or sensory deprivation.  Military working dogs will not be used for, or be present during interrogation.”

Violations of this policy are to be “reported through the change of command and intelligence oversight channels … and if appropriate, referred to competent authority for criminal investigation and disposition.... Suspected or alleged violations may also be reported through other appropriate military officials such as criminal investigators, Inspectors General, Chaplains, or Judge Advocates.” 

It would appear that less egregious situations might not lead to full-blown criminal actions, and might be dispensed with less formally. 

Training

The Procedure states that “K-9 personnel train almost every day to keep up proficiency levels.  Non-handler personnel are authorized and encouraged to act as decoys.”  Realistic law enforcement based scenario training may be used.

This blog was written by John Ensminger and L.E. Papet. The authors thank Kingsbury Parker for comments.