Federal Turf Wars Over How to Train Explosives Detection Dogs

On August 11, 2004, Attorney General John Ashcroft sent a memo to the Directors of the FBI and the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) specifying that as �soon as practicable, all Department components that use explosives detection canines shall use only canines certified by ATF.�(1) This did not happen in the Bush administration, and seven years later in the Obama administration the FBI continues to use dogs not certified by the ATF.(2) While the turf war between the FBI and ATF has involved much more than dogs,(3) here we will focus on those aspects relevant to canine training and deployment.

From a fairly clear designation in 2004 of the ATF as the agency responsible for certification of explosives detection dogs, it now appears that both the FBI and ATF apply different philosophies with regard to training explosives detection dogs, and it is not clear that the Department of Justice still expects there to be any real amalgamation of canine functions. It may be that Justice has settled for relatively specific boundaries between the two bureaus that can at least keep their squabbling from being aired in public.(4)

An increasingly important player in the field of canine training is the Transportation Security Administration within the Department of Homeland Security. TSA�s power may become sufficiently great in the end that this agency will ultimately determine the most widely used parameters for federal explosives detection dog training.

ATF�s Odor Recognition Proficiency Standard

The Omnibus Consolidated Appropriations Act of 1997(5) had authorized the Secretary of the Treasury to establish a standard for explosives detection canines to be employed by federal agencies or other agencies providing explosives detection services at airports in the U.S. At the time, ATF was in the Department of Treasury, but was transferred to the Department of Justice in 2003. Section 653(a) of the Act provided: �The Secretary of the Treasury is authorized to establish scientific certification standards for explosives detection canines, and shall provide, on a reimbursable basis, for the certification of explosives detection canines employed by Federal agencies, or other agencies providing explosives detection services at airports in the United States.�

Treasury designated ATF with the responsibility of developing the standard.(6) In 1999, ATF noted that it had provided copies of the odor recognition standard to �all interested persons� and that it had �solicited input and recommendations from other Federal law enforcement agencies that use explosives detection canines.� On March 10, 1999, the Oak Ridge National Laboratory, managed by Lockheed Martin Energy Research Corporation for the U.S. Department of Energy, issued a final report concluding that ATF�s odor recognition proficiency standard was �valid for the measure of the proficiency for detecting explosives odors at the recognition and alerting phase of training.�(7)

There does not seem to have been a problem until ATF was transferred to the Department of Justice on January 24, 2003, at which point it became apparent to ATF officials that the FBI was not accepting ATF�s mandate regarding explosives detection dogs for federal agencies.

2007 Agency White Papers

In 2007, the FBI and ATF developed �white papers� that were attached to an Explosives Review Group report to Deputy Attorney General Paul McNulty. It must have been obvious to McNulty and other DOJ officials that Ashcroft�s memo had not been implemented. The FBI�s white paper argued that:

� DOJ components should use only canines meeting Scientific Working Group on Dogs and Orthogonal Detection Guidelines (SWGDOG) Certification.
� The FBI should continue its joint training initiative on peroxide explosives scent training.

ATF argued that:

� All DOJ explosives detection dogs should be procured, trained, and certified by ATF.(8)
� Any training related to explosives detection canines sponsored, coordinated, or presented by DOJ components should be coordinated through ATF.

ATF was willing to agree that if an ATF-certified canine was not available, a DOJ component could, in the interim, use a dog trained and certified to standards set by SWGDOG.(9) The photograph is from ATF�s website, showing an ATF dog in training.

2009 Inspector General Audit

An audit report of the Department of Justice Inspector General released in October 2009 (�2009 audit report�) concluded that the FBI and ATF had failed to implement Ashcroft�s 2004 directive, and that the two agencies continued to disagree on the guidelines for training explosives detection dogs.(10) The report determined that �the FBI generally uses non-ATF certified canines.� A survey of FBI explosives specialists determined that more than 80% rarely or never used ATF-certified dogs, relying instead primarily on explosives detection dogs provided by state and local agencies that are often not ATF-certified. The FBI�s uniformed police use ATF-certified dogs, but FBI field divisions generally rely on state and local dogs or the explosives detection dogs actually owned by the FBI (numbering only four in 2009).

�Although the FBI and ATF agree that DOJ should have a single certification standard for canines, they disagree on the how the standard should be established. ATF developed the National Odor Recognition Training and Testing (NORT) program as a standardized method for assessing a canine�s ability to recognize explosives odors. NORT is a test administered by ATF forensic chemists to federal, state, and local canine teams.

�The FBI believes NORT may not actually assess the operational capabilities of the canine and that the certification standards should be determined by the Scientific Working Group on Dog and Orthogonal detector Guidelines (SWGDOG). SWGDOG is composed of members from federal, state, and local agencies, including both ATF and the FBI.(11) While SWGDOG is a non-certifying body that provides best practice guidelines, it anticipates that these best practices will be incorporated into participating organizations� certification standards."(12)

The Inspector General described the FBI and ATF as having �developed separate and often conflicting approaches to explosives investigations and related activities such as explosives training, information sharing, and forensic analysis.� The Inspector General warned that this continuing friction meant that the agencies might not meet the requirements of Homeland Security Presidential Directive (HSPD)-19 (February 12, 2007), �which requires a united, multi-layered strategy to mitigate the threat and prevent the use of explosives by terrorists.�

The Inspector General noted that despite ATF being responsible for certification of all DOJ explosives detection canines, �the FBI continues to disagree with ATF on canine certification standards. We also found that DOJ sent conflicting signals to the components, directing that one standard for training canines be administered through ATF, but also funding a working group seeking to adopt a different standard through the FBI and DOJ�s National Institute of Justice.� The working group is a reference to SWGDOG.

The FBI told the Inspector General that the HSPD-19 Implementation Joint Program Office (JPO) was designed to resolve issues not previously resolved through other mechanisms. Concerning this, the 2009 audit report stated:

"We found that the JPO was not designed to function as the deciding authority on roles and responsibilities for the FBI and ATF in handling explosives incidents, but instead was conceived to be a formalized, interagency discussion forum. Unless there is consensus among the agencies involved the JPO cannot force settlement between components. Therefore, while the JPO and its members may be used to assess and reflect community opinion and advise on priorities, individual agencies will continue to make programmatic and budgetary decisions independently."

The Inspector General found �no evidence that the FBI and ATF worked together to establish DOJ explosives training priorities, and the two agencies reached no consensus on the use of explosives detection canines.�(13) This was true despite the fact that the two agencies train dogs for similar, or at least overlapping, purposes.

"[E]xplosives specialists from both ATF and FBI are providing peroxide-based explosives detection training to state and local bomb squad canine teams and both components disagree about the standards that should be used to certify explosives-detection canines."

Other agencies, such as the Transportation Security Administration, work with both the FBI and ATF on canine issues:

"According to ATF, it began training explosives detection canines on peroxide-based explosives in 2002 after the ATF Laboratory worked with British authorities to develop an effective method of producing explosives used for training purposes. In 2006, the National Explosives Detection Canine Training Program, which is part of the Transportation Security Administration of the Department of Homeland Security, worked with the FBI to train canine teams to detect various peroxide explosives."

The Inspector General then detailed specific canine training issues on which the FBI and ATF disagreed:

�FBI explosives specialists told us they provide bulk (15-30 grams) samples of peroxide-based explosives to state and local canine handlers to sensitize their animals to the peroxide scent.(14) ATF also conducts similar training for state and local canine handlers with trace (5 milligrams) amounts of peroxide-based explosives.(15) ATF officials argued that the FBI should not be providing such training, saying that ATF�s method is superior because training with trace amounts of peroxides enhances the canines� ability to detect explosives. For example, because these explosives would generally be sealed in containers, the canine must be able to alert based on recognizing a trace amount of explosives left on a container lid or its scent on the potential bomber.(16)

�An FBI Explosives Unit official noted in a published FBI Bomb Data Center Investigators Bulletin that trace amounts, like those used by ATF, can be utilized to conduct training if suitable precautions are taken [citing FBI Bomb Data Center Investigators Bulletin 2006-3 entitled K-9 Detection of Peroxides]. However, the official noted that any time trace amounts of material are utilized, they are susceptible to contamination. For example, if the handler using these aids handles any other type of explosive or has an explosive residue near these aids, it is possible to introduce interfering odors. In addition, the FBI official contended that trace amounts of peroxides dissipate rapidly, and once exposed, have a very short shelf life.

�Despite these differing opinions, the differences between the FBI and ATF�s peroxide-based explosives training programs do not appear to be irreconcilable, and consolidation of the training standards should be possible. Therefore, we recommend the FBI and ATF consolidate the training for peroxide-based explosives.�

This means that in 2009, the Inspector General thought the canine training approaches of the FBI and ATF could at least be reconciled. It was not clear that this reconciliation would come about through any specific mechanism, but it appears the FBI, at least, had faith in the National Explosives Detection Canine Advisory Board (NEDCAB). Concerning this group, the Inspector General stated:

�[T]he FBI noted that the Department of Homeland Security (DHS) and DOJ will co-lead an interagency advisory board responsible for developing uniform standards for explosives-detection canine teams, including annual certification and recurring proficiency training. DHS and DOJ, building on the previous ATF National Canine Advisory Board, created the National Explosives Detection Canine Advisory Board [NEDCAB], which includes participants from major professional canine associations. The FBI believes that as a result of the creation of this advisory board for the first time, there is consensus across the explosives-detection canine community that national training and performance standards are needed.�

The Inspector General was not so sure that NEDCAB would be a final answer.

�As part of the JPO, the Department of Homeland Security (DHS) and DOJ are co-leading an interagency advisory board responsible for developing uniform standards for explosives-detection canine teams, including annual certification and recurring proficiency training. DHS and DOJ, building on the previous ATF National Canine Advisory Board, created the National Explosives Detection Canine Advisory Board, which includes participants from the four major professional canine associations. The FBI believes that as a result of the creation of this advisory board for the first time, there is consensus across the explosives-detection canine community that national training and performance standards are needed. We recommend that DOJ select and enforce a single standard for the use of certified canines for DOJ components, consistent with the requirements of HSPD-19.�

Upon release of the Inspector General�s report, FBI Assistant Director Michael Kortan and ATF Assistant Director W. Larry Ford issued a joint statement describing how the ATF and FBI have cooperated, including recommending for prosecution 192 explosive-related cases involving 397 defendants. Not surprisingly the joint statement said the two agencies are working �to resolve the identified issues to improve coordination and response to explosive incidents.�

GAO Reports

In a 2010 report,(17) GAO noted some differences between SWGDOG and ATF training, and also stated that the Transportation Security Administration�s dogs are following certification standards of ATF.

�While the mechanism of how canines detect explosives through their sense of smell is not well understood, there are several certification programs to validate the canines� ability to detect explosives, which include specifying standards for explosives detection. These standards vary based on which entity is certifying the canine. A guiding document on the training of canines is the Scientific Working Group on Dog and Orthogonal Detectors Guidelines that specifies recommended best practices for canine explosives detection. These standards call for an EDC to detect explosives a certain percent of the time and a probability of false alarms less than a certain rate. Certifying entities, however, may have more stringent standards. For example, ATF requires that its canines detect all explosives that are presented to them, and have limited false alarms in its tests. TSA requires that their certified canines find a specified percent of explosives in a variety of scenarios, such as onboard an aircraft, mass transit rail, and mass transit buses. Homeland Security Presidential Directive-19 tasks the Attorney General, in coordination with DHS and other agencies, with assessing the effectiveness of, and, as necessary, making recommendations for improving federal government training and education initiatives related to explosive attack detection, including canine training and performance standards. According to ATF officials, TSA, in coordination with ATF, is developing standards for EDCs, which are nearly complete and are similar to the standards that ATF uses�.

�Canines have a history of being trained to detect items and in recent years have been trained to detect, among other things, explosives, fire accelerants used in arson investigations, and drugs. While training methods differ among canine training schools, these methods typically train canines by rewarding them for locating certain items. Rewards include toys, a food treat, or the canine�s food itself. In turn, these canines are trained to alert their handlers if they detect an item of interest, usually by sitting down next to the item.�(18) (emphasis added)

Yet another GAO report stated that �[p]articipants in TSA�s Transit Security Grant Program and DHS�s Homeland Security Grant Program are required to maintain data to document compliance with guidelines for their explosives detection canine teams. These guidelines were developed by a scientific working group that included officials from DHS.�(19) This is another reference to SWGDOG, the standards of which could only with some stretch be said to be �similar to the standards that ATF uses.�(20)

2010 Protocol

In 2010, Gary G. Grindler, Acting Deputy Attorney General, released a protocol superseding all prior guidance on ATF-FBI explosives coordination, including, it would appear, Ashcroft�s 2004 memo. Grindler noted the respective capacities of the two agencies:

� ATF has 2,593 Special Agents trained in post-blast and fire scene investigations.
� ATF has 244 Certified Explosives Specialists who cover much of the country.
� The FBI has 143 Special Agent Bomb Technicians (SABTs).
� The FBI has 100 Joint Terrorism Task Forces (JTTFs) around the country staffed by federal, state, local, and tribal law enforcement agencies.

Despite these areas of competence, Grindler states that �the current situation�as chronicled by the Office of Inspector General�must be remedied.� Grindler then outlines guidance, beginning with domestic terrorism explosives investigations:

�I have concluded � that the FBI should ultimately be considered the lead agency for domestic terrorism explosives investigations�. Unless the matter of lead agency has otherwise been addressed between the ATF and FBI in a particular matter, the ATF will be the lead agency for explosives investigations where there is no credible nexus to international or domestic terrorism�.�

Grindler acknowledges that lead agency jurisdiction may shift during an investigation.

As to explosives training, Grindler notes that the FBI opened its Hazardous Devices School at the Redstone Arsenal in Huntsville, Alabama, while ATF has the National Center for Explosives Training and Research in Fort A.P. Hill, Virginia, and the Canine Training and Operations Support Center in Front Royal, Virginia. Grindler does not side with either agency on training, but directs �ATF and FBI to develop a joint plan for consolidated explosives training of ATF and FBI agents and technicians, and state and local law enforcement, that is consistent with this protocol.� A working group�was another working group really needed?�will be created to aid in this effort, which �should identify the cost savings that can be achieved by integration of training curricula and facilities.� Joint training should begin within six months of Grindler�s memo, i.e., by February 3, 2011.

The protocol attached to Grindler�s memo ends with an explanation of what is to happen henceforth in disputes between the FBI and ATF:

�If disputes arise relating to the operation of this protocol, the Deputy Directors from ATF and FBI shall meet in the first instance to resolve those disputes. If a matter cannot be resolved at the Deputy Director level, it should be brought to the Office of the Deputy Attorney General for resolution by the Explosives Coordination Committee.�

This sort of knocking of heads may work, but if so it would likely be because of Grindler�s direct influence, since preaching to the parties has not worked for the preceding six years.

2011 Developments

A 2011 GAO report on cost savings noted that since 2004, the Department of Justice �has taken actions intended to address duplication and overlap in the areas of explosives investigations jurisdiction, training, information sharing and use of databases, and laboratory forensic analysis.�(21) The GAO refers to the DOJ Inspector General�s 2009 report, noting how the problem has continued for years without resolution. Nevertheless, the GAO�s recommendation is no more than that the Department of Justice should continue to monitor the situation.(22) The GAO made no specific mention of canine training in this report.

An ATF Fact Sheet issued September 2011 reiterated ATF�s faith in its National Odor Recognition Testing Standard (NORT) and provided the following statistics:

"There are 29 ATF�trained explosives detection canine teams with ATF special agent canine handlers. Also, there are currently 122 ATF�trained explosives detection canine teams deployed throughout the United States with local, state or other federal agencies. In addition, there are 61 ATF�trained accelerant detection canine teams currently active in the United States and one in Canada�.

"Since 1991, ATF has trained 708 explosives detection canines and 157 accelerant detection canines. These dogs and their ATF�trained handlers are located throughout the United States in local police and fire departments, fire marshal offices and federal and state law enforcement agencies. Teams are also located in 21 foreign countries�. The canines are capable of detecting 19,000 explosives compounds."

Needless to say, ATF did not mention the doubts expressed by the FBI to the DOJ Inspector General regarding the NORT program in 2009.

Enter the Department of Homeland Security

Wind the clock back to 2006 when Congress was grappling with ways to turn the 9/11 Commission Report, issued in 2004, from policy perspectives and general recommendations into legislation. In the Implementing Recommendations of the 9/11 Commission Act of 2007,(23) Congress authorized the Department of Homeland Security to create the National Explosives Detection Canine Team Training Program, and provided that this authority would extend well beyond TSA:

�Based on the feasibility in meeting the ongoing demand for quality explosives detection canine teams, the Secretary shall establish criteria, including canine training curricula, performance standards, and other requirements approved by the Transportation Security Administration necessary to ensure that explosives detection canine teams trained by nonprofit organizations, universities, and private sector entities are adequately trained and maintained.� (6 U.S.C. 1116(c)(1))

The Secretary of Homeland Security is to �coordinate with key stakeholders, including international, Federal, State, and local officials, and private sector and academic entities to develop best practice guidelines for such a standardized program, as appropriate.�

The legislative history provides limited perspective on what Congress was thinking about with regard to the TSA program:

�The Conferees recognize that explosives detection canines are not trained to additionally detect chemical, nuclear or biological weapons and that, at present, such teams cannot detect radiological materials. Further, the Conferees recognize that canines are trained to detect specific threats and cannot, at this time, effectively be crossed-trained to identify multiple threats. In requiring the TSA to develop canine training curriculum and performance standards under this section, the Conferees expect TSA to do so for those threats within the definition that are currently applicable to canine team detection. However, the Conferees trust that TSA will explore opportunities to train and/or acquire canines that are able to detect new and emerging threats, such as chemical, radiological, nuclear and biological weapons. To that end, the Conferees expect that prior to developing and distributing canine training curriculum and performance standards under this section, TSA will fully vet any ongoing training, whether domestic or international, that has a proven method to successfully detect those additional threats that may not currently be applicable to TSA-trained canines.�(24)

The legislative history is largely restricted to transportation issues, but does say that the Secretary of Homeland Security �may use the canine teams on a more limited basis to support other homeland security missions, as determined appropriate.� Had Congress considered that it had previously designated ATF with overall responsibility for explosives detection canine training in the 1997 Appropriations Act, this would have been the place to mention it, because the 9/11 legislation effectively meant that two federal agencies had been blessed by Congress with significant authority over explosives detection work with dogs.

The 9/11 Act led to no regulatory releases, though the Federal Register in 2009 referred to the National Explosives Detection Canine Team Program, which noted that survey data was being �collected electronically through the NEDCTP secure Canine Web site (accessible by authorized personnel only)� to provide feedback to the Chief of NEDCTP, �staff and supervisors on how the training material was presented and received.�(25) The TSA described NEDCTP as follows in March 2009:

�The National Explosives Detection Canine Team Program (NEDCTP) is a partnership between TSA, airports, and local law enforcement. The NEDCTP supports TSA�s mission by preparing law enforcement canine-handlers, who are not Federal employees, and canines to serve on the front lines of America�s War on Terror. These canine teams (handler and canine) are trained to quickly locate and identify dangerous materials that may present a threat to transportation systems.�(26)

In July 2011, the Department of Homeland Security published a report indicating the Department�s intent regarding canine training and deployment:

�Establish DHS as a center of excellence for canine training and deployment. Canines serve essential roles in homeland security. Specially-trained canines and their handlers are essential elements of terrorism prevention efforts at the Federal, State, local, tribal, and territorial levels and in the private sector. Canines also serve essential roles in detecting narcotics at the air, sea, and land ports of entry, and in search and rescue activities following disasters. DHS will increase specialized breeding activities for canines, enhance its training and certification of canines and handlers, and become a center of excellence for employment of canines across the homeland security missions.�(27)(emphasis added)

Thus, the ultimate authority on the optimal methods of U.S. explosives detection canine training could now be the Department of Homeland Security, not an agency in the Department of Justice. Acting Deputy Attorney General Grindler�s resolution of the issue might not have been final, as DHS dogs were certainly beyond his jurisdiction.

A 2011 thesis by John P. Joyce at the Naval Postgraduate School in Monterey, California,(28) describes the TSA as having �partnered with the Federal Bureau of Investigation and National Institute of Justice to sponsor the Scientific Working Group on Dog and Orthogonal Detection Guidelines (SWGDOG) to enhance the performance of the EDCTs [Explosive Detection Canine Teams] �. The SWGDOG was established in January 2005 in an effort to develop consensus-based guidelines that can be shared across all groups involved in canine detection work�.�(29) Given DHS�s interest in expanding the impact of its canine training, this may mean that by 2011 TSA was siding more with the FBI on canine issues than with the ATF.(30) It may be argued (and certainly would be by anyone in the press office of either agency) that DHS and the FBI (or DOJ) will see no need to carry on the same sort of turf war as has been the case between ATF and the FBI, but government agencies seem to have an inevitable instinct for defending their authority.

Website Claims

The ATF webpage devoted to accelerant and explosives detection canines states the following:

�As the Federal Government explosives and postblast experts, ATF offers certified explosives detection canine to other Federal, State, local and foreign law enforcement agencies. ATF�s uses a food and praise reward training methodology that exposes canines to five basic explosives groups, including chemical compounds used in an estimated 19,000 explosives formulas. It is believed by ATF that exposing canines to various explosives from the basic explosive families will give the dog the ability to detect the widest range of commercial or improvised explosives possible when working in field. Successful detection of an explosive or firearm earn the canine a food and praise reward, which encourages repetition.(31) To earn ATF certification, all dogs must pass a blind test wherein they must successfully detect 20 different explosives odors, two of which they were never exposed to during training. The scientific methodology, and the training and testing protocols are certified by the ATF National Laboratory, and produce an extremely versatile, mobile, and accurate explosives detection tool.�

This does not read as though the agency has ceded authority to any other agency either within the Department of Justice or without.

The Transportation Security Administration is equally confident in the webpages devoted to its National Explosives Detection Canine Team (NEDCT) program:

�TSA's National Explosives Detection Canine Team Program prepares dogs and handlers to serve on the front lines of America's War on Terror. These very effective, mobile teams can quickly locate and identify dangerous materials that may present a threat to transportation systems. Just as important, they can quickly rule out the presence of dangerous materials in unattended packages, structures or vehicles, allowing the free and efficient flow of commerce.

�Law enforcement officers from all over the country travel to TSA's Explosives Detection Canine Handler Course at Lackland Air Force Base in San Antonio, Texas where they are paired with one of TSA's canine teammates. These dogs are bred specifically for the program by TSA's puppy program, also at Lackland AFB. German Shepherds, Belgian Malanoises, Vizslas and other types of dogs are used in the program because of their keen noses and affinity for this type of work. In addition to providing a highly trained dog and handler training, we provide partial funding for handler salaries, care and feeding of the canines, veterinary and other costs associated with the dog once the teams return to their hometowns.� (emphasis added)

The ability to provide �partial funding� is certainly proof of power. Lackland Air Force Base, sometimes called �Dog School,� is the largest canine training facility in the country, and the location where many military working dogs are trained. Thus, DHS is also likely using MWD training methods for domestic explosives detection dogs.

The FBI keeps a relatively low profile as to canines on its website, emphasizing its scientific approach to the use of canines. The website, on the other hand, has seven webpages that refer to SWGDOG (on November 22, 2011, in any case), more than any other federal website.

SWGDOG�s website states that the organization �is a collaboratively funded effort of the DHS, FBI, NIJ [National Institute of Justice], and TWSG [Technical Support Working Group].� Although the TWSG does have ATF connections, these appear to involve non-canine operations. SWGDOG has one representative from ATF, so it is not clear if the list of collaborative funders means merely that ATF does not now provide funding, or something more political.

If SWGDOG is to be granted such broad influence over federal canine policies, comments on proposed guidelines should be made public, and the organization should hold a public hearing before making any guideline final. Representatives of federal agencies intending to expect compliance with the guideline for purposes of contracting with private entities should be present and willing to answer questions regarding their interpretation of the proposal.

Is Uniformity Essential?

Divergence in training approaches is not necessarily a bad thing, despite the continuing concerns of the Department of Justice. Although the investigative authority of the agencies overlap, and any investigation could shift from one side to the other as new evidence concerning the perpetrators and targets develops, having dogs under different regimens, with different volumes of target samples, may to some extent make it difficult for criminals and terrorists to know how to hide their bombs. On the other hand, any approach that cannot be shown to be reliable should be jettisoned.

Conclusion

ATF has trained hundreds of dogs for its own agents and for other law enforcement units around the country. ATF has also partnered with a number of certifying organizations (USPCA, NAPWDA, IPWDA and NPCA) to achieve a consensus for canine explosives detection capabilities as part of the National Canine Initiative Explosives Detection Canine Certification. ATF has earned the loyalty of many explosives detection handlers around the country by having the agency�s chemists send highly sensitive peroxide-based explosives to events sponsored by such organizations.

The FBI relies on others to train its dogs (including the ATF, according to the ATF website), but has supported the development of SWGDOG, which seeks to become a standards authority for training and testing law enforcement canines in the United States (though many certifying organizations do not see SWGDOG as a resource, but more as a competitor).

TSA has control of Lackland Air Force Base, the largest canine training facility in the country, and has also supported SWGDOG. DHS also uses dogs in other subagencies, including U.S. Customs and Border Protection, the U.S. Coast Guard, FEMA, and the U.S. Secret Service. SWGDOG, by making DHS a fan, may have effectively won control of federal explosives detection canine standards.

It is unlikely that any of these agencies will easily give up its programs or sphere of influence, and since the canine programs of the agencies are connected with other aspects of their operations, it will be increasingly difficult to impose a policy of consolidation in training philosophies. DOJ may now be accepting a bifurcated canine philosophy in its ranks, hoping to minimize friction, or at least the public evidence of friction. DHS, through TSA, is unlikely to have to accept any DOJ approaches it does not like, both through sheer numbers and because of its robust budget.

1. In a March 4, 2004, memorandum, the Attorney General identified explosives-related training as an issue for an Explosives Review Group (ERG) to review. The report produced by the ERG had recommended that Department of Justice components should use only ATF-certified explosives detection canines.
2. The FBI had 15 dogs for its uniformed officers in 2010, while ATF had over 300 dogs worldwide. James W. Hawkins, Explosives Recognition and Awareness Training: A Psychological Approach to Pre-Blast Mitigation, Master�s thesis, Missouri University of Science and Technology, 2010).
3. Administrators inside of the government were well aware that the feud between the FBI and ATF continued after 2004. The Inspector General states that �in January 2007 the ERG [Explosives Review Group] reported to the Deputy Attorney General that training related to post-blast, canines, and render-safe procedures either had not been implemented or remained highly contested.�
4. It is appropriate to note that the issue does not have to do with the number of dogs owned by the agencies. The FBI had 15 dogs for its uniformed officers in 2010, while ATF had over 300 dogs worldwide. Hawkins (2010).
5. PL 104-208 (September 30, 1996).
6. 62 Fed. Reg. 50982 (September 29, 1997
7. 64 Fed. Reg. 41487 (July 30, 1999).
8. ATF maintains the Canine Training and Operations Support Branch in Fort Royal, Virginia. This facility develops explosives detection dogs for federal, state, and local agencies.
9. The 2009 report included information on the funding of SWGDOG, showing that it received funds from the National Institute of Justice and the FBI totaling just over $500,000 from 2004 through 2008.
10. U.S. Department of Justice, Office of the Inspector General, Audit Division, Explosives Investigation Coordination Between The Federal Bureau of Investigation, and the Bureau of Alcohol, Tobacco, Firearms and Explosives, Audit Report 10-01 (October 2009).
11. The FBI has a member of the SWGDOG Executive Board, but ATF does not (board webpage checked 11/21/2011). ATF has a regular member. The Inspector General does not mention that there are also members selected from associations and other nonprofit organizations. The 55 SWGDOG members are listed on the organization�s website under the heading �Membership.�
12. The involvement of SWGDOG members in patent applications and other profit-making activities that may correlate with proposed standards has apparently not been of concern to the federal agencies.
13. Administrators inside of the government were well aware that the feud between the FBI and ATF continued after 2004. The Inspector General states that �in January 2007 the ERG [Explosives Review Group] reported to the Deputy Attorney General that training related to post-blast, canines, and render-safe procedures either had not been implemented or remained highly contested.�
14. SWGDOG�s standard SC8-Substance Detector Dogs: Explosives Detection states that �[m]inimum weight of substance odors being tested for certification shall be 113.4 grams (1/4 lb).�
15. ATF news releases often mention that ATF trainers use trace amounts of explosives. See ATF News Release, ATF Continues to Put Bite on Explosives (August 14, 2009).
16. Relationships of the agencies and organizations discussed here with individuals entities making private profit from canine scent detection approaches is a topic to be discussed separately.
17. GAO, Technology Assessment: Explosives Detection Technologies to Protect Passenger Rail, GAO-10-898 (July 2010).
18. Id., 40-41.
19. GAO, Maritime Security: Ferry Security Measures Have Been Implemented, but Evaluating Existing Studies Could Further Enhance Security, GAO-11-207 (December 2010), p. 32, n.38.
20. See TSA�s Explosives Detection Canine Program: Status of Increasing Number of Explosives Detection Canine Teams, GAO-08-933R (July 31, 2008).
21. Government Accountability Office, Opportunities to Reduce Potential Duplication in Government Programs, Save Tax Dollars and Enhance Revenue, GAO-11-318SP (March 2011).
22. The GAO emphasized coordination in following up on the protocol released by Grindler in a 2011 report. GAO, Law Enforcement Coordination: DOJ Could Improve Its Process for Identifying Disagreements among Agents, GAO-11-314 (April 2011).
23. PL 110-53, 121 Stat. 266 (August 3, 2007).
24. Conference Report to Accompany H.R. 1.
25. 74 Fed. Reg. 55248 (October 27, 2009); 75 Fed. Reg. 4579 (January 28, 2010).
26. 74 Fed. Reg. 9621 (March 5, 2009).
27. Department of Homeland Security (July 2010). Bottom-Up Review Report.
28. Joyce, John P. (March 2011). Thesis: The Transportation Security Administration�s Four Major Security Programs for Mass Transit�How They Can Be Improved to Address the Needs of Tier II Mass Transit Agencies. Naval Postgraduate School, Monterey, California.
29. Joyce cites the 2005 statement of David Kotny, Director of the National Explosives Detection Canine Team Program, to the Subcommittee on Management, Integration, and Oversight of the House Committee on Homeland Security.
30. DHS has a member on the SWGDOG Executive Board. In his thesis, Joyce found that some mass transit authority officials noted that dogs �trained in a closed and somewhat sanitized airport environment are not conditioned to operate in a mass transit environment, and therefore are not as effective as those trained exclusively on mass transit.�
31. Whether this five-component approach will in fact allow a dog to alert to any explosive containing one of the components is far from certain. There is evidence that different ratios in a mixture will affect a dog�s ability to recognize a component in the mixture. For a recent study on this issue in narcotics detection, see Macias, M.S. and Furton, K.G. (2011). Availability of Target Odor Compounds from Seized Ecstasy Tablets for Canine Detection. Journal of Forensic Sciences, 56, 1594-1600, noting: �MDMA solutions were analyzed by liquid chromatography�mass spectrometry. Analysis of these samples showed a wide variance of MDMA (8�25%). Headspace SPME-GC/MS analysis showed that several compounds such as 3,4-methylenedioxyphenylacetone and 1-(3,4-methylenedioxyphenyl)-2-propanol are common among these MDMA samples regardless of starting compound and synthesis procedure. However, differences, such as the level of the various methylenedioxy starting compounds, were shown to affect the overall outcome of canine detection, indicating the need for more than one MDMA training aid. Combinations of compounds such as the primary odor piperonal in conjunction with a secondary compound such as MDP-2-OH or isosafrole are recommended to maximize detection of different illicit MDMA samples.�

This piece was written by John Ensminger and L.E. Papet.

The Dogs of the Conquistadors

Reading Dogs of the Conquest by John and Jeannette Varner reminded me of reading Dee Brown�s Bury My Heart at Wounded Knee almost forty years ago. After a while you just want it to stop. Every page in the Varners' book has a horror. Just one example:

�A letter of protest, dated June 4, 1516, from fourteen Dominican priests of Hispaniola [the island that now contains Haiti and the Dominican Republic] to a royal counselor related � how on one occasion some Christians [Spanish soldiers on the island] came across an Indian woman with a nursing child in her arms. Since their dog was hungry, they seized the child and tossed it to the dog, which tore it to pieces in the presence of the mother.�

One can find histories of the conquistadors with scant mention of the dogs they brought and bred in the New World. The Varners have brought the horrors perpetrated with the dogs into the light, just as Dee Brown made us see that settling of the West was also a genocide.

Dogs in Fifteenth and Sixteenth Century European Warfare

Large numbers of dogs were used in warfare from the late Middle Ages into the Renaissance. Before the age of exploration, the Spanish Christians had used dogs against the Moors (Varner & Varner, p. xvi). King Henry VIII was said to have sent hundreds of war dogs to the Emperor Charles V of Spain in a war with France, �each garnished with good yron collars� (Lloyd 1948, Weir 2002, p. 33).

An entry in the English State Papers of Elizabeth's reign for January 1599 includes a letter stating that "Essex's commission for Ireland is agreed on after many difficulties, but not signed. He is called Lieutenant, may return at pleasure, make barons, dispose of lands won from the rebels, &c.; he akes great provision for horses, and many are presented him.  They talk of taking over 200 or 300 mastiffs, to worry the Irish or rather their cattle."  An entry in the State Papers for April 29, 1598, but probably correctly 1599, states that the Earl of Essex left for Ireland with "12,000 men and 3,000 mastiff dogs" to deal with the Irish insurrection. This information may have been obtained from two captives of the Spanish.  Lloyd (p. 179) questions the historicity of this report. Taking 3,000 dogs would have been a massive undertaking.  It is possible that the captives were exaggerating, if not totally fabricating, knowing that the Spanish were accustomed to the use of war dogs and would find such numbers, if believed, quite discouraging.  

Jaques du Fouilloux, some of whose plates of hunting scenes have been reproduced in a prior blog, depicted an ancient voyage to Brittany by hunters who fled the flames of Troy, but instead of settling in Rome had gone to Bretaigne (Brittany), believing there would be good hunting there. This is the first plate above. As was true of Renaissance depictions of antiquity, the ships and clothing were contemporary, not ancient, and it is quite possible that du Fouilloux had seen Spanish, English, or French ships with war dogs pacing the decks.

War Dogs in the New World

Gonzalo Pizarro, brother of Francisco, brought as many as a thousand dogs with him in an expedition begun in Peru in 1541. This may be the largest assembly of attack dogs in history, but the Spanish had dogs they could use in battle against the natives as early as the second voyage of Columbus (Varner and Varner, p. 4, Coren, p. 74), who had brought Spanish war dogs to Hispaniola in 1493. Columbus used a dog in dispersing a hostile group of Indians that came to stop his landing in Jamaica in 1494. He used 20 dogs at the Battle of Vega Real in 1495, as shown in the plate from Herrera. (Only four dogs are visible, apparently an artistic limitation.) By 1509, dogs running loose on Hispaniola were so numerous that they were killing livestock and themselves began to be hunted.

The Historia General of the Spanish historian, Antonia de Herrera y Tordesillas (1601-15), may contain the most detailed images of dogs in battles between the Spanish and the Indians. The next plate from the cover of the first volume of the Historia General shows another scene in which the dogs are being sent ahead of the Spanish forces. One dog is still held by a leash, but the others have been loosed and are running simultaneously with the cavalry, perhaps in part to make it difficult for the Indians to choose a target. The dogs are not shown as armored in any of the scenes in Herrera's work, but this is generally the case in contemporary depictions of the conquistadors. The final plate from Herrera shows the dogs still on leashes at the center of the Spanish line, with cannons on the outside. Their release was timed for maximum effect both in terms of defensive value to the men and to add to the terror just before the contact of the front lines.

Why did the use of attack dogs reach its zenith in the New World? War dogs had been taught to attack the enemy since antiquity, but the numbers brought to and then bred in the New World far exceed what is known about any other military use of attack dogs. Why was it worth the while of the Spanish invaders to bring dogs for their explorations and subjugations? The answer probably lies in the fact that the Indians dressed differently from the Spanish, had different diets, different skin color, and at least initially a heart-stopping awe of the huge mastiff-like animals the Spaniards used against them. (Dogs were widely known in the New World, but those preferred by most Meso- and South American Indians were small dogs, such as the beautiful plate of an Aztec dog from the 16th century Florentine Codex of Friar Bernardino de Sahag�n.)

The contrast between the people who used the dogs and those against whom they were used differed from most European wars going back to antiquity where the armies dressed alike, had the same diets and skin color, and became indistinguishable in the field of battle. Like the elephants used by Hannibal and some Macedonian generals, dogs brought the risk that they could turn against those who employed them. That risk was minimal in the Americas and the advantage this provided was realized not just by the Spanish, but also by the Portuguese, French, English, and finally, during the period of slavery, the Americans.

The dogs were not always rewarded for their service. The same dogs that made the expedition of Gonzalo Pizarro so formidable gave up their lives when his army, facing starvation, ate all but two of them. (Markham, p. 17)

Care and Training

The dogs were of the late medieval hunting types�mastiffs, alaunts, and greyhounds�in Spanish, mast�ns, alanos, and lebrels. (For detailed descriptions of hunting dogs of the period, see Cummins 1988.) Varner and Varner (p. 36) note that even that dogs identified by the conquistadors as belonging to one group might well have had blood of others. Although greyhounds were often kept separate from other hunting animals in European castle life, it is not clear if such an effort was made in the New World. An Indian revolutionary known as Enrique, who was baptized by Spanish priests on Hispaniola but rebelled against his Spanish lords in 1519, was said to have kept dogs for hunting isolated and under the care of only two or three families. This might reflect an early Spanish practice of isolating dogs from natives so that they would not become friendly to those whose scent they were trained to track and attack. (Helps, vol. III, p. 106)

Aldrovandus (1522-1605) describes the training of war dogs as prescribed by Blondus:

�The dog ought to be trained up to fight from his earliest years. Accordingly some man or other is fitted out with a coat of thick skin [pelle densa], which the dog will not be able to bite through [quam canis lacerare nequeat], as a sort of dummy: the dog is then spurred on against this man, upon which the man in the skin runs away and then allows himself to be caught and, falling down on the ground in front of the dog, to be bitten [morderi patiatur]. The day following he ought to be pitted against another man protected in the same manner, and at the finish he can be trained to follow any person upon whose tracks he has been placed [super cuius vestigiis collocabitur]. After the fight the dog should be tied up, and fed while tied up, until at the finish he turns out a first-class defender of human being. Blondus is even of the opinion that from time to time it is a good thing to go for this dog with drawn swords [strictis mucronibus]; in this way, he thinks, the dog will develop his spirit and courage to the utmost; and then of course you can lead him against real enemies. And Blondus adds that such dogs are frequently to be met with in Spain of the present day [in multis Hispaniae locis hodie habentur].

Aldrovandus even says that a relative of his, a senator from Bologna, had such a trained dog for his defense. Since Aldrovondus was citing Flavius Blondus (1392-1453) for this training regimen, he could not have been describing the use of dogs in the New World. Although referring to this training as being for defensive purposes, the dogs could be used in attack situations and Blondus may have been describing dogs that were used against the Moors.

Dogs in Expeditions

In 1509, Alonso de Ojeda left Santo Domingo for Carthagena, with horses and dogs. The same year Juan de Esquivel attacked Indians that had fled to the mountains of Jamaica. The dogs he brought with him were said to be �almost as destructive as his musquetry.� (Long, p. 959) In 1513, Vasco N��ez de Balboa set out to find the �other sea,� the Pacific, with 190 men and dogs, �which were of more avail than the men.� (Helps, vol. I, p. 358) Encountering resistance from an Indian leader named Pacra, Balboa had him torn to pieces by dogs. Balboa had a dog named Leoncico supposedly able to distinguish between an �Indio de Guerra� and an �Indio de Paz.� Gonzalo Fern�ndez de Oviedo y Vald�s, the Spanish historian, gives a detailed description of the dog:

�We must not omit to mention a dog which Balboa possessed, called Leoncico [little lion], from the dog Becerrico [little bull], of the island of St Juan, and no less famous than his parent. This dog gained his master, in this and other entries, more than 2000 pesos of gold, because he received the share of a companion, in the distribution of gold and slaves. And truly the dog deserved it better than many sleeping partners. This dog�s instinct was wonderful; he could distinguish between the warlike or peaceful Indian; and when the Spaniards were taking or pursuing the Indians, on loosing this dog and, and saying, �There he is�seek him,� he would commence the chase, and had so fine a scent, that they scarcely ever escaped him. And when he had overtaken his object, if the Indian remained quietly, he would take him by the sleeve, or hand, and lead him gently, without biting or annoying him, but if he resisted, he would tear him in pieces. Ten Christians escorted by this dog, were in more security than twenty without him. I have seen this dog, for when Pedrarias came to this territory, in the year 1514, he was still alive.... He was of a red colour, had a black nose, was of a middle size, and not handsomely formed, but stout and powerful, exhibiting many wounds, which in the course of these wars he had received from the Indians. The dog was at last maliciously poisoned. Some dogs of his race were left; but nothing equal to him has been seen in these regions.� Oviedo�s General History, book xxix, chap. 8. (Hodson translation, p. 18) The wounds were most probably from arrows, which both a man and a dog can often survive.

Evidently, the teaching of the dogs to attack involved some repeated behaviors on the parts of those attacked, so that individuals not in certain postures or acting aggressively were not part of the attack command for the dogs.

In 1513, Balboa used his dogs to tear apart 50 Indians allied with the chief Torecha in the domain of Quarequa. This event led to one of the most widely produced depictions of war dogs in the New World, Theodore de Bry�s engraving of dogs attacking defenseless and nearly naked Indians. based on the account of Bartolom� de las Casas. After that, the barking of the dogs was often enough to disperse any group inclined to oppose Balboa�s advance. (Hodson, pp. 45, 54)


The dogs were not always effective. In meeting resistance from Indians at Santa Marta, now in Colombia, in June 1514, Pedro Arias de Avila, often known as Pedrarias, the dogs failed to attack the Indians and began fighting among each other (Varner and Varner, p. 47). Although hardly consistent in their opposition, the priests sometimes recognized that the dogs made their efforts at converting the natives more difficult. Father Domingo de Betanzos believed that the noise of weapons and the barking of fierce dogs �had so stunned the Indians as to render them deaf to the Christian Faith.� (Helps, vol. III, p. 312)

Mexico

Cort�s landed in Mexico with dogs and other animals, as shown above in a plate from the Florentine Codex by Friar Bernardino de Sahag�n. The Aztec account of the conquest of Mexico by Hern�n Cort�s in 1519 includes a description of the dogs running ahead of the column, their muzzles high, lifted to the wind, saliva dripping from their jaws. (Broken Spears, p. 41) Moctezuma was told that the Spaniards rode deer without horns, and he received an extensive description of the dogs:

�Their dogs are enormous, with flat ears and long, dangling tongues. The color of their eyes is a burning yellow; their eyes flash fire and shoot off sparks. Their bellies are hollow, their flanks long and narrow. They are tireless and very powerful. They bound here and there, panting, with their tongues hanging out. And they are spotted like an ocelot.�

An Aztec visual perspective on Cort�s�s dogs can be found in a mid-16th century painting on a large cloth panel, or lienzo, referred to as the Lienzo de Tlaxcala (Kranz 2007). The original lienzo is lost, but copies were made from the original, including the depiction here of soldiers marching with a dog on leash (Chavero 1892). The Aztecs were only familiar with smaller dogs (Valadez 2000).

As late as 1805, a Spanish Lieutenant, Antonio Narbona, entered the region of Arizona on a punitive expedition, with horses and dogs, an event captured in a Navajo pictograph at Standing Cow Ruin in Canyon de Chelly. Bones recovered from the area have been identified as coming from Spanish greyhounds (Dix 1980). Jett (1981) questioned whether the figures may all be of horses, displayed as smaller to give a distance perspective to the Spanish column. The smaller figures, however, appear riderless and unleashed, and the coloration could be that of Spanish dogs of the sort that in Louisiana founded the Catahoula breed. The picture of the rock art at Canyon de Chelly was taken by Cori Ryan, who maintains a website with many excellent photographs of rock art.

Note (2012). A picture of the assembly of figures contained in Thor Conway's Painted Dreams (1993) convinces me that all the figures are probably horses, though some are sufficiently eroded as to make certainty unlikely.

Peru and the Straits of Magellan

Olive trees were brought from Seville to Peru in 1560 but, as only three of 100 plants survived the journey, they were planted in a fruit garden near Lima. It was necessary to post 100 slaves and 30 dogs to guard the trees night and day. (Markham, 1864a, p. 401)

Pedro Sarmiento de Gamboa had dogs with him as he explored the Straits of Magellan in 1579. Gamboa�s brief narrative indicates he may not have expected his dogs to be excessively warlike:

�It was noteworthy that our dogs, and those of the natives, flew at each other until they came within four paces, when they turned round without touching, and we could never get them to attack again.� (Markham, 1895, p. 322)

After the Conquest

As the New World was subjugated, dogs began to take on less war-like responsibilities, and became pets of all members of the societies that emerged after the conquest. Felipe Guam�n Poma de Ayala (1535 - c. 1615) wrote an illustrated chronicle, Nueva Cor�nica y Buen Gobierno, showing daily life in the Andes. One plate shows the author walking near Lima with his son and two dogs, named Amigo and Lautaro, both from mastiff or possibly alaunt lines. The next plate below shows an Andean hunter with a falcon and two dogs. The dog behind the hunter is likely an indigenous small dog, but the one preceding him appears to have the blood of an alaunt or mastiff, probably having mixed with an indigenous line. The final plate shows a city, perhaps with hunting dogs running in the fields before the walls. (See Adorno 2000.)

Adoption of Spanish Practices in the Americas

Even after the conquest, the Spaniards were said to use their bloodhounds to track down natives for sport, or in order to train them to catch human game. By 1638, the Dutch used dogs against the Indians. (Southey, vol. III, p. 488). During the final Maroon War, which began in 1795, the Jamaican legislature sent to Cuba for large hunting dogs to seek out and destroy the blacks in their stronghold. (Long, vol. II, p. 79) The French employed Cuban dogs in a Haitian rebellion. (Childs, p. 42)

Benjamin Franklin recommended the use of dogs during the French and Indian War. In a letter to James Read of November 2, 1755, Franklin's detailed military advice includes the following observations:

"The 50 Arms now sent are all furnish�d with Staples for Sling Straps, that if the Governor should order a Troop or Company of Rangers on Horseback, the Piece may be slung at the Horseman�s Back. If Dogs are carried out with any Party, they should be large, strong and fierce; and every Dog led in a Slip-String, to prevent their tiring themselves by running out and in, and discovering the Party by Barking at Squirrels, &c. Only when the Party come near thick Woods and suspicious Places, they should turn out a Dog or two to search them. In Case of meeting a Party of the Enemy, the Dogs are then to be all turn�d loose and set on. They will be fresher and fiercer for having been previously confin�d, and will confound the Enemy a good deal, and be very serviceable. This was the Spanish Method." (emphasis added)

After the end of the war, Pennsylvania was still dealing with Indian raids, and a Une 4, 1764, letter preserved in Franklin's correspondence from Henry Bouquet to John Penn contains the following passage:

"I can not omit to Submit to your Consideration the use that might be made of Dogs against our Savage Enemies; It would be needless to expect that our Foot Soldiers can overtake an Indian in the Woods, and their audacious attempts in attacking our Troops and settlements may, in a great Measure, be ascribed to the certainty of evading our Pursuit by their flight: a few Instances of Indians Seized and worried by Dogs, would, I presume, deter them more effectualy from a War with us, than all the Troops we could raise, and as we have not in this Country the Species of those animals, which would best answer this Purpose, I beg leave to recommend it to you, to have Fifty Couples of proper Hounds imported from Great Britain, with People who understand to train and manage them. They might be kept on the Frontiers, and a few given to Every Scouting Party, to discover the Ambushes of the Enemy, and direct the Pursuit: This requires that the men intended to follow the Dogs should be well mounted."

The success of Cuban mastiffs in the Maroon War led to a recommendation to the American War Department in 1839 that they be used against the Seminoles in Florida:

�My object in this communication is principally to show how much benefit was derived in the Maroon war from the use of dogs, and to suggest to the War Department the propriety of turning its attention to similar means of prosecution of the war in Florida. I might deduce facts from the ablest writers in Ethics and International Law, to prove that the employment of these animals against the Seminoles, who have been guily of every enormity, would violate no moral duty; and the general tendency of events during the last two hundred years, goes to show that the transgressions of our red brethren, notwithstanding the declaration of enthusiasts, has attracted upon them the displeasure of the Almighty, as did the cities of the Plain and the people of Edom in days of old.� (Homans, p. 147)

Professor Matt D. Childs of the University of South Carolina notes that in the American Civil War, �Cuban dogs aided the Confederacy in its attempt at secession by pursuing runway slaves as battlefields reached the plantation zones of the South.� (Childs, p. 42) The post-Civil War use of the same tracking dogs in finding criminals is discussed in detail in Police and Military Dogs (Chapters 3 and 5).

Sources: 

1. Adorno, R. (2001). Guaman Poma and His Illustrated Chronicle from Colonial Peru. Museum Tusculanum Press.
2. Aldrovandi, Ulyssis (1645). De Quadripedibus. Translation from Ash, Edward C. (1927). Dogs: Their History and Development, vol. II. Ernest Benn Ltd., London.
3. Anderson, Fred. Crucible of War: The Seven Years' War and the Fate of Empire in British North America, 1754-1766 (indicating, at 809, that the British were familiar with the Spanish use of dogs against the Indians, where Henry Bouquet says of enemy Indians, "I wish we could make use of the Spanish methods, to hunt them with English dogs").See also note regarding Benjamin Franklin, below. 
4. Brown, D. (1970). Bury My Hart at Wounded Knee. Rinehart & Winston, New York.
5. Chavero, A. (1892). Explicacion del Linzo de Tlaxcala. In Homenaje � Crist�bal Col�n: Antig�edades mexicanas. M�xico: Oficina Tipogr�fica de la Secretar�a de Fomento.
6. Childs, Matt D. (2006). The 1812 Aponte Rebellion in Cuba and the Struggle Against Atlantic Slavery. University of North Carolina Press.
7. Conway, T. (1993). Painted Dreams: Native American Rock Art. NorthWord Press, Minocqua, Wisconsin.
8. Coren, S. (2002). The Pawprints of History. Free Press, New York.
9. Cummins, J. (1988). The Art of Medieval Hunting: The Hound and The Hawk. Weidenfeld & Nicolson. London.
10. Dix, A.S. (1980). Spanish War Dogs in Navajo Art at Canyon de Chelly, Arizona. The Kiva, 45(4), 279; (1983). More on Spanish War Dogs in Canyon de Chelly Rock Art: A Reply to Jett. The Kiva, 48(4), 323.
11. Espana, V.M., and Azua, R.V. (2003).  Los Perros de Guaman Poma de Ayala: Vision Actual del Estudio del Perro Precolombino Sudamericano. AMMVEPE, 14(2), 43-53. (This is a very important article that, unfortunately, did not come to my attention early enough, but which contains much thought that will be important for me to incorporate as I turn this blog into a chapter.)
12. Fontana, B.L. (2000). Pictorial Images of Spanish North America. Journal of the Southwest, 42(4).
13. Fouilloux, Jaques du (1502). La Venerie.
14. Franklin, B. Letter to James Read, Sunday, November 2, 1755. Correspondence, vol. 6. The Papers of Benjamin Franklin Sponsored by the American Philosophical Society and Yale University.
15. Helps, Arthur (1855). The Spanish Conquest of America. John W. Parker & Son, London.
16. Herrera, A. (1601-1615). Historia General de los Hechos de los Castellanos en las Islas y Tierra Firme del Mar Oceano. Madrid.
17. Hodson, Mrs. (1832). Lives of Vasco Nunez de Balboa, and Francisco Pizarro from the Spanish of Don Manuel Josef Quintana. William Blackwood, Edinburgh.
18. Homans, B. (ed.) (1839). Army and Navy Chronicle, vol. 9.
19. Jett, S.C. (1981). War Dogs in the Spanish Expedition Mural, Canyon del Muerto, Arizona? The Kiva, 46(4).
20. Kranz, T.B. (2007). Sixteenth-Century Tlaxcalan Pictorial Documents on the Conquest of Mexico. In Sources and Methods for the Study of Postconquest Mesoamerican Ethnohistory, ed. James Lockhart, Lisa Sousa, and Stephanie Wood (e-book). Eugene, Oregon: Wired.
21. Leon-Portilla, Miguel (1962). Broken Spears: The Aztec Account of the Conquest of Mexico. Beacon Press, Boston.
22. Lloyd, H.S. (1948). The Dog in War. In Vesey-Fitzgerald, B. (ed.). The Book of the Dog. Nicholson and Watson, London.
23. Long, Edward (1774). The History of Jamaica. London.
24. Markham, Clements R. (1864). Expeditions into the Valley of the Amazons. The Hakluyt Society; (1895). Narratives of the Voyages of Pedro Sarmiento de Gamboa to the Straits of Magellan. The Hakluyt Society; (1864a). The Travels of Pedro de Cieza de L�on, A.D. 1532-50. Hakluyt Society.
25. Morgan, L.H. (1881). Houses and House-Life of the American Aboriginees.  Washington: Government Printing Office.  (noting that the cacique, chieftain of a tribe near Ocute in present-day Georgia, gave many presents to Hernando de Soto, including "conies, partridges, bread of maize, two hens and many dogs.").
26. Prescott, William H. (1843). History of the Conquest of Mexico, vol. II. Richard Bentley, London.
27. Prescott, WIliam H. (1847). History of the Conquest of Peru, vol. I. Baudry�s European Library.
28. Southey, Robert (1810). History of Brazil. London.
29. Turbervile (1576). Booke of Hunting.
30. Valadez, R. (2000). Prehispanic Dog Types in Middle America. In Dogs Through Time: An Archaeological Perspective: Proceedings of the 1st ICAZ Symposium on the History of the Domestic Dog. BAR International Series 889.
31. Varner, John G., and Varner, Jeannette J. (1983). Dogs of the Conquest. University of Oklahoma Press, Norman.
32. Weir, A. (2002). Henry VIII: The King and His Court. Random House Digital. 
33. Worthington, B.E. (2008). An Osteometric Analysis of Southeastern Prehistoric Domestic Dogs.  Florida State University M.Sc. thesis. 

Thanks to Suzanne G. White, Richard Hawkins, and Brian Duggan for help in completing this blog. Thanks to Cori Ryan for the excellent photograph of the Arizona petroglyph.  Thanks to Professor Paul E.J. Hammer for leading me to additional sources regarding Elizabeth I's Irish War.

K9 Fraud! Essential Reading for Handlers, Lawyers, and Judges

A book that should be in every police canine handler�s library, as well as the library of every lawyer and judge who handles cases involving canine evidence, is K9 Fraud! Fraudulent Handling of Police Search Dogs, by Resi Gerritsen and Ruud Haak (Detselig Enterprises Ltd., Calgary, Canada, 2010). The authors are married and live in the Czech Republic near the Austrian border, where they are training directors for the International Red Cross Federal, the UN, the International Rescue Dog Organisation, and the F�d�ration Cynologique Internationale.

Gerritsen and Haak are also the authors of K9 Professional Tracking: A Complete Manual for Theory and Training (Detselig, 2001). With Adee Schoon, Haak wrote K9 Suspect Discrimination: Training and Practicing Scent Identification Lineups (Detselig, 2002). Haak is the editor of Onze Hond, one of the widely circulated dog magazines in Europe. In the interest of full disclosure, my book on service and therapy dogs was mentioned favorably in Onze Hond.

The subtitle of K9 Fraud! is somewhat unfortunate in its reference to police search dogs, as this might suggest to American readers that the book is about search and rescue operations. Most of the chapters actually discuss fraud in scent identification lineups and in tracking and trailing. An entire chapter is devoted to canine responsiveness to human gestures, which any lawyer, or any judge, dealing with a cueing claim should read. The publisher did not serve this important work well in failing to provide an index, and I found it necessary to use post-its to highlight pages I will want to refer to again.

Those who have read other books by these authors, and authors with whom they have associated, will find some of the material familiar, if not on occasion repetitive. The difference is that where the other books (at least the ones with which I am familiar) look at canine practice from the perspective of how to do it right, this book focuses on how it can be done wrong. I am inclined to think that the latter approach may be more necessary for lawyers and judges, particularly in the U.S. As I pointed out in recent discussions of a drug bust and a station identification, many American judges are still inclined to accept the work of canine handlers uncritically and do not adequately consider what might have been done wrong. In all fairness, I must acknowledge that some courts, such as the Florida Supreme Court, have begun to look past the mystique and require a solid foundation for the admission of canine evidence.

Gerritsen and Haak are not afraid to take on handlers by name, some of whom are prominent and spoken of with respect. Of course, these authors are writing from and probably more influential in continental Europe than in the U.S.

A Botched Murder Prosecution Leads to Blanket Exclusion of Lineups in the Netherlands
 
A murder investigation and prosecution led to the exclusion as evidence of scent lineups performed in the North and East Netherlands over a period of nine years because it was found that a critical requirement for performing a scent lineup had been ignored by police during this period. The mistake was regarded as so critical that six police dog handlers were sentenced to prison for two years. One of the cases that led to this exclusion is discussed by Gerritsen & Haak in detail.

Jacqueline Wittenberg, a rich old widow living in the Dutch town of Deventer was murdered on September 23, 1999, but was not found for several days. There were no signs of a break-in. It was known that she did not admit strangers to her home, and only admitted people she knew if she expected them. The last one to call her on the night of the murder was Ernest Louwes, her tax advisor. Louwes claimed to have been far from Deventer at the time the call was made. A telecom expert testified, however, that if Louwes had made the call from anywhere but Deventer, the call would not have been relayed through the telecom antenna in Deventer.

A knife was discovered two days later under a neighbor�s porch. The neighbor who found it picked it up with his sleeve to avoid leaving his own fingerprints on it. The knife was used in a canine scent identification lineup, resulting in the identification of Louwes as having left scent on the knife. In 2002, the Supreme Court of the Netherlands concluded that the knife had not been the murder weapon. Louwes was released in 2003.

DNA extraction from the blouse that Wittenberg was wearing when she was murdered again pointed to Louwes, however, and he was resentenced in 2004 to 12 years in prison. Suspicions regarding Wittenberg�s handyman and his girlfriend began to surface, however, and additional investigation brought to light a number of discrepancies in their earlier stories. Louwes was again released in 2009. Dutch law still considers Louwes to be the murderer, but a number of observers now doubt his guilt.

The fact that Louwes was identified in a scent lineup where a knife that was later found not to have been used in the crime has apparently never been explained.

Prosecution of Dutch Police for Scent Lineup Failures 

The paragraph following the description of the Deventer case by Gerritsen and Haak says much about why scent lineups have come into question in Europe, despite being in general much more rigorously performed there than in the United States:

�The seven police dog handlers of North and East Netherlands region were prosecuted for perjury and forgery, because it was found that they hadn�t performed the scent identification line-ups exactly in accordance to the regulations. According to these rules, the dog handler may not know the position of the scent carriers of the suspect, to avoid influencing his dog. This step was found to be omitted in many cases between September 1997 and March 2006. Even worse, was that the official reports stated that all identification line-ups were processed in compliance with the standard codes of operation. In November 2007, the seven police dog handlers were convicted to penal labor of 240 hours. Six of the seven handlers were also incarcerated for two years and were removed entirely from their positions. Only the dog handler that had shed light on situation was allowed to continue his duty.�

Whether SWGDOG guidelines should be recognized as U.S. national standards involves issues of transparency and delegation of regulatory authority. Even if such issues were overcome and there were some general recognition of SWGDOG as establishing best practices applicable in a broad range of criminal investigative contexts, it is unlikely that the organization would ever be the beneficiary of criminal enforcement for failures to implement its recommendations.

A Tool for Judges and Lawyers

In researching Police and Military Dogs, I found that early decisions from the end of the 19th and the beginning of the 20th century had a better feel for what dogs could do than more recent cases. Many judges of those times�particularly judges in the American South, where tracking largely began as a police function�were hunters and experienced with the tracking of animals. More recent decisions, in contrast, often showed undue deference to the testimony of handlers, or, alternatively, regarded canine evidence as being produced by silly animals, hardly worthy of the court�s time, dogs of the sort the judges saw in the elevators of their high-rise condo buildings. For a brief time I wondered if I could create a taxonomy of pro-dog and anti-dog judges based on their tolerance for and appreciation of canine evidence, but no clear pattern of this sort emerged as I read over 1,200 judicial decisions. What did emerge was that many judges failed to recognize the slow but steady progress that was being made in the practical applications of canine olfaction, and the decisions often lacked any real understanding of how dogs behave or the limits of their skills. Appellate decisions often avoided careful analysis of trial court failures by labeling as harmless the errors that appeared to have been made.

Another circumstance I noted as I researched police canine law was an increase in claims of ineffective assistance of counsel being made by defendants convicted in significant part by canine evidence. No law school that I am aware of offers a class in this increasingly active area of criminal law. Reading through the lines of opinions, it was often apparent that defense counsel was quite effective in raising procedural issues and in offering stiff resistance to most evidentiary aspects of the prosecution�s case, but cross-examination of police dog handlers was often anemic and missed apparent weaknesses in their testimony. Defense experts were chosen for some general canine expertise, or for superior academic credentials, but were often not versed in the areas where the prosecution�s case might have been vulnerable to a countervailing perspective. Lines of questioning often bogged down on issues that did not bolster any viable defense position, while other aspects of a handler's testimony remained unaddressed.

I should say that this lack of adequate legal attention to canine evidentiary issues also comes from those, such as the Innocence Project, who want to have police dog work labeled as �junk science� and excluded altogether. Here the weaknesses of judges and lawyers are being distilled into a knee-jerk battle cry against the acceptance of canine evidence at all.

I recommend that all these participants in the criminal justice system read K9 Fraud! before writing another brief, opinion, or press release.

Thanks to L.E. Papet and Gail McConnell for suggestions that improved this blog.

Canoe Conquests of the Western Pacific: Who Brought the Dingo Ancestors, and Why?

That Australian dingoes descended from domesticated dogs was considered a possibility by Hamilton Smith (1856) more than a century and a half ago, and the first plate is one of several paintings of dingoes that he published in his volumes on dogs.

Edward C. Ash (1927, p. 24) described the dingo as follows:

�Canis dingo is smaller than the wolf, and has somewhat long legs. It stands 24 inches at the shoulder. The tail is bushy. There is a greyish under-fur, but except in the black variety, the long hairs are generally yellow or white. In the whole family there seems to be a natural tendency for the feet and end of the tail to be white. The muzzle is very often black. It is found in the wooded districts throughout Australia, and was at one time extremely numerous. It runs unlike dogs, the head held up, and the ears erect and forward. In its habit it is far more like the fox than the wolf.�

Genome studies have in fact established that dingoes and New Guinea Singing Dogs are dogs, not wolves or some separate species of canid, and are descended from lines once domesticated though long isolated on large islands. Dingoes and Singing Dogs have formed relationships with the indigenous peoples of Australia and New Guinea, but their separation from other dog populations is much more recent than the arrival of the aboriginal peoples and these connections are unlikely to continue prior domestication behaviors.

So when did dingoes arrive in Australia? When did New Guinea Singing Dogs arrive in New Guinea? Dr. Peter Savolainen and a team addressed this question in a short paper published seven years ago, and he and another team�this one headed by Mattias Oskarsson�have now looked at the issue again. They have also looked at the genetic overlap of dingoes with other dogs in Southeast Asia and the western Pacific.

The problem is to explain not only when dogs arrived on the islands, but also who brought them there. Thus, what the DNA says must be placed in the context about what is known of the cultures that settled on the islands, or at least stopped at them. That the dogs arrived in canoes is generally assumed from the distances involved (Horridge 1995; for depictions of dogs on the canoes that met Captain Cook, see Luomala 1962), but canoe cultures have been present in the western Pacific for thousands of years and making connections between genetic studies of dingoes and cultural anthropological findings remains elusive.

Unfortunately, at least for those looking for firm answers concerning the relationships of dingoes to humans, the paper by Oskarsson et al. may actually increase the number of possibilities concerning who brought the dingoes and when.

Ancient Dogs of Southeast Asia and the Pacific

Australia and New Guinea were colonized by hunter gatherers about 50,000 years ago according to Mulvaney and Kamminga (1999), who themselves cite Jones (1979). A 2002 radiocarbon study of Australian bones by Gillespie concluded that the �oldest occupation horizons in four different regions reliably dated by defendable multi-method results are in the range 42-48,000 calendar years ago.�

Oskarsson et al. note that by 30,000 years ago, colonization �had reached as far into Near Oceania as the Bismarck Archipelago and the western-most Solomon Islands.� Islands further east were not colonized until the Neolithic, which reached western Polynesia about 3,000 years ago and eastern Polynesia only 1,400 years ago.

Neolithic culture developed in the Yangtze Valley about 8,500 years ago, reached Taiwan by 5,500 years ago and Southeast Asia between 4,500 and 3,500 years ago. A �cultural complex� known as Lapita appeared in Near Oceania about 3,500 years ago and spread across Polynesia, reaching New Zealand by 1250 AD. (See Greenhill et al. 2010, discussing Austronesian expansion at 5,200 years ago.)

The dog appears in the archeological record of Australia about 3,500 years ago, where a pre-Neolithic culture received �virtually no influence from external sources.� Oskarsson et al. note that how �the dingo, as the single item of possibly Neolithic origin, arrived to Australia is therefore an enigma.� They note that the spread of dogs in Southeast Asia �in parallel with the spread of Neolithic culture is clearly indicated.�

Dogs were probably present in Taiwan by 4,500 years ago, yet the genetic results of Oskarsson et al. do not find a connection between Taiwanese dogs and the dingoes of Australia and New Guinea Singing Dogs. Bones of domesticated dogs have been excavated in Thailand from 4,000 years ago. (Higham 1996, p. 59, noting that �by at least 2000 B.C., people with dogs domesticated from wolves were already living in the Khorat pleateau.�) Bellwood (2005) describes the distribution of dogs with Austronesian culture:

�All of this leads to the fairly astonishing observation that, between 2000 and 800 BC, assemblages with related forms of red-slipped and stamped or incised pottery, shell artifacts, stone adzes, and keeping of pigs and dogs (neither of these animals being native in most of the regions concerned) spread over an area extending almost 10,000 kilometers from the Philippines through Indonesia to the western islands of Polynesia in the central Pacific.�

Theories regarding the spread of Austronesian culture generally break down to arguments that Austronesians spread out from Taiwan in a relatively short time frame, or that more complex interactions between Southeast Asian cultures and the Pacific Islands occurred over more protracted periods that may have involved a significant infusion from Taiwan. Language studies have pointed to Taiwan, but human DNA studies have found only a minor contribution from Taiwan, pointing instead to �a largely Melanesian ancestry for the Polynesian people.� Regardless of which approach is more descriptive of cultural expansion, the dingo remains a curiosity.

The 2004 study by Savolainen et al. found that pre-European dogs from across Polynesia carried only two haplotypes, and that the Australian dingo population �was founded from a small number of dogs carrying a single mtDNA haplotype (A29).� (See also vonHoldt et al. noting lower diversity of Australian dingoes and New Guinea Singing Dogs.) Two New Guinea Singing Dogs �were shown to carry haplotypes A29 and A79.� A29 is found among East Asian dogs, but is rarely found west of the Himalayas. This earlier study had concluded that dingoes arrived in Australia from 5,000 years ago and perhaps as far back as 10,000 years ago, substantially before the archeological record can verify. (See also Klutsch and Savolainen 2011.)

Mainland Roots of Dingoes

The 2011 study concludes that South China was the probable source population for dingoes and Singing Dogs. Although genetic overlaps of haplotypes were found in mainland Southeast Asia and Indonesia, no overlap was found with samples from Taiwan and the Philippines. Unlike the previous study, the researchers in this one found the time of arrival of dingoes in Australia to be as recent as 4,640 years ago and as far back as 18,100 years ago. They find �a clear indication that Polynesian dogs as well as dingoes and NGSDs trace their ancestry back to South China through Mainland Southeast Asia and Indonesia. Thus, there is no indication that these dogs were introduced via Taiwan and the Philippines together with the expansion of the Neolithic culture and Austronesian languages, as suggested in some theories about Polynesian origins.� This has also been shown to be true of pigs.

This study thus appears to align itself with the school of thought that would argue that Austronesian languages and agriculture originate in Southeast Asia, without mediation through Taiwan. (See discussion in Piper et al. 2009.) If Polynesian peoples originated in Taiwan, the current research concludes that �their Neolithic cultural package was modified en route.� The researchers argue:

�We suggest that, with the evidence on the origins of Polynesian domestic dogs and pigs, a likely scenario for the origins of Polynesians is that farmers spread from Taiwan bringing the Neolithic culture (e.g. pottery) and Austronesian languages, but mixed extensively with local Melanesian populations, and picked up some cultural traits (e.g. the domesticated dog and pig, and the commensal Polynesian rat) � en route. Therefore, the cultural package of the Polynesians was probably formed from different sources, some parts deriving from Taiwan and others incorporated at the spread through Indonesia and Melanesia.�

The researchers speculate that the dingo may have been introduced by indigenous Australians trading with Neolithic groups, perhaps through New Guinea, or trading with pre-Neolithic groups. They argue that a common ancestry of New Guinea Singing Dogs and dingoes is suggested by a number of factors, including genetics and similarities in morphology and behavior. They may have been founded �from very few individuals from the same Indonesian population, but obtained different haplotypes because of founder bottlenecks.� Koler-Matznick et al. (2000) note that blood isoenzymes differ between dingoes and New Guinea Singing Dogs, as well as the fact that dingoes are about twice the size of the latter.

The map shows the general pattern of movement of the predecessors of dingoes from Southeast Asia as posited by Oskarsoon et. al. (Double click to enlarge.)

Reasons for Bringing Dogs in Canoes

Oskarsson et al. found that the founder haplotypes for ancient Polynesian dogs, though not that of dingoes, could also be found in southern China, mainland Southeast Asia, and Indonesia, and as with dingoes, were absent from Taiwan and the Philippines. (One Polynesian haplotype was found in 7% of New Guinea Singing Dogs.) This would seem to indicate a separate migration for the dogs that arrived in Polynesia than that of dingoes, but it may still be useful to consider why dogs would be brought on boats at all, as some of the same uses for the animals were likely to be present. (For a drawing distinguishing the morphology of dingoes from other Polynesian dogs, see Hemmer 2005, p. 33.)

Unfortunately, little can be said about what sort of relationship the predecessors of dingoes had with those who brought them to the islands. As to the current level of domestication, Carl Lumholz, in his book, Among Cannibals (English edition, 1902), states:

�On Herbert river there are rarely more than one or two dingoes in each tribe, and as a rule they are of pure blood. The natives find them as puppies in the hollow trunks of trees, and rear them with greater care than they bestow on their own children. The dingo is an important member of the family�; it sleeps in the huts and gets plenty to eat, not only of meat, but also of fruit Its master never strikes, but merely threatens it. He caresses it like a child, eats the fleas off it, and then kisses it on the snout.

"Though the dingo is treated so well it often runs away, especially in the pairing season, and at such times it never returns. Thus it never becomes perfectly domesticated, still is very useful to the natives, for it has a keen scent and traces every kind of game; it never barks, and hunts less wildly than our dogs, but very rapidly, frequently capturing the game on the run. Sometimes it refuses to go any farther, and its owner has then to carry It on his shoulders, a luxury of which it is very fond. The dingo will follow nobody else but its owner�.�

The drawing of a dingo is from Lumholz's fascinating book.

Koler-Matznick et al. (2003) reported that New Guinea Singing Dogs avoid human contact, but when captured as puppies they may be raised to assist in hunts. Some tribes also eat the dogs on occasion. Teeth may be used as ornaments. Gill (1873-4, p. 46) argued that dingoes must have spread from New Guinea to nearby smaller islands where they were sometimes domesticated, but sometimes roamed in destructive packs.

Dogs of the Maoris

The Maori people of New Zealand, who arrived from eastern Polynesia, had many uses for their dogs. Blood taken from the ear of a dog and boiled was a remedy for spear wounds, taken internally or used as a lotion. Balls used in a game played by young women were ornamented with the white hair of a dog. Birds, fish, dogs, and rats were the principal animals in their diet. Dogs were cooked. Dog-skins were used in clothing. Tregear (1904, pp. 166-7) gives the following description:

�The Maori dog (Kuri ruarangi) has now entirely disappeared and it is highly improbable that even the very earliest of the white settlers ever saw the real animal, although doubtless some of its blood was running in the veins of the mongrels that roamed around the native villages. Those seen by Cook, Forster, and others, about the time New Zealand was discovered, were small dogs, something like degenerate sheep-dogs, with large heads, sharply-pierced ears, and a short, flowing tail. It was considered a valuable article of food, being bred for its edible qualities rather than for any other purpose, and as such even appreciated as eking out the slender resources of the explorers with Captain Cook. Crozet described native dogs as looking like domesticated foxes, indeed they would destroy poultry just as foxes do, and he relates that they were fed on fish, and would not be domesticated among white men, whom they would bite on occasion. The skin was highly valued as an article of attire, and a mat of dogskins was a precious possession. The white hair (awe) of the dog's tail was also used as an ornament for the weapons of a chief; the tail of the living animal being kept regularly shaved, and the hair put away for this purpose. The flesh of the dog was not allowed to be eaten by women, and not by men except under certain restrictions.�

Accounts of old natives said that the native dog did not bark, but howled a good deal. Owners prized and petted them and gave each a name. They were sometimes castrated, and were given birds and rats to eat. They were often trained to catch ground game, such as the ground-parrot (kakapo), rails (weka), and apteryx (kiwi). Training the dog was explained by Tregear as follows:

�This was done by the master squatting down, and holding his dog, at the same time giving a cry in imitation of that of the bird, who hearing the cry would come towards the hunter. The little dog was then let go and would catch the bird and hold it or bring it to his master. The dog might get lost through its stupidity, but never ran wild.�

Dogs were taken in canoes to hunt ducks and sent overboard to catch the ducks when the canoes had gotten as close as possible. Kiwi hunting is described as follows:

�The kiwi goes along looking for worms or rather listening for the rustle of the earth-worm underground. When the bird hears the worm creeping below the soil the long beak is prodded down and finds its prey. The kiwi hunter fastened little pieces (patete) of wood to his dogs' neck, so that they would rattle or rustle, and the kiwi would stop to listen, thinking that it heard the worms creeping. Then the dogs would rush in, and the men came forward with torches which they had hitherto concealed. The bird was astounded at the sudden dazzling light, it being a nocturnal bird and not used to the light, so that it was easily killed.�

Dogs even advised when to hunt kiwi: �If a dog twitched or barked in his sleep, you would go hunting with him soon, and he would catch plenty of kiwi (apteryx) for you.�

As mentioned earlier, women could not eat dog flesh, but Tregear reports other restrictions:

�The flesh of the dog was held to be a tapu food, only to be indulged in by certain persons and under certain restrictions. A dog was always killed at the great ceremonies connected with the children of chiefs and on other important and formal occasions, but the priest ate its flesh. A dog was also killed for the tattooer, when he was operating on a chief; but anciently they were kept for sacrifice.�

Theft of a dog could lead to bloodshed. Dogs had a position in the spiritual life of the Maori:

�The spirits of dogs were supposed, like those of men, to pass to the World of Shadows (Te Reinga) but they travelled by a different path than that taken by the souls of human beings. If a dog barked in a certain way at a man it was supposed to denote the death of the person barked at; the god of evil and death (Te Nganahau) inspired the dog to give the warning. Dogs frequently became goblins (taniwha) and sometimes the guardian spirits of certain places. The sacred dog of Maahu lived under the waters of a lake named Te Rotonuiaha, and was a kind of banshee, its bark proceeding from under the water being a warning of the approaching death of a chief.�

The spirits of dogs could pass into other things on death:

�A chief of high descent and great powers had a dog that was killed by a falling tree, and thereon the chief commanded the spirit of the dog to pass into a large tree growing near, and in that tree the spirit dwelt for ages and spoke (in the dog language) to travellers who dared to address it.�

Calling someone a dog was an insult in Maori society:

�The tutelary deity of dogs was Irawaru or Owa, the husband of the sister of Maui the hero, but Irawaru offended Maui who changed him into a dog and then insulted his sister by telling her to call aloud for her husband with the cry 'Moi moi!' the usual call to a dog, and which is even to-day an insult if used to a man.�

The Maoris believed that their ancestors brought dogs on canoes:

�Kupe, one of the legendary discoverers of these islands brought his dogs with him, and not only do the Hokianga natives show some curious markings in stone as being footprints of one of these dogs, but in another place they exhibit a stone into which another of the animals was transformed.�

The plate of the canoe shows a war party that met captain Cook. A detail from the plate shows a dog up close.

Dogs of the Native Hawaiians

Bryan (1915) describes the arrival of dogs on the Hawaiian Islands:

�Just as the Polynesian people carried useful plants with them on their wanderings, they also brought with them in their canoes these two highly-prized and useful domestic animals known to them in their more ancient home�. The Hawaiian dog was fed largely on poi, and was much relished as food in old-time Hawaii. Like the hogs, they were classed according to their color, there being several well-recognized color types.�

As to eating the dog, he states:

�The poi-dog, when carefully fed and fattened on poi, was regarded as even more delicious in flavor than pork. Dogs always formed an important dish at the native feasts and on such occasions large numbers of them would be baked in earth ovens.� Dogs were wrapped in banana leaves and placed on stones that had been heated by wood in a hole. The fat of the pig and dog was used as an illuminating oil.

Hamilton Smith (1840) contains a letter describing the Poe Dog of the Hawaiian Islands (once known as the Sandwich Islands) as follows:

�At the Sandwich group, where the inhabitants have been more remarkable for the use of this animal as food, and where that custom is yet pertinaciously retained (owing probably to the scarcity of swine and spontaneous fruits of the earth), the pure breed of the Poe dog has been better protected; and although becoming yearly more scarce, examples of it are yet to be met with in all the islands, but principally as a delicacy for the use of the chiefs. As late as October 1835, I noticed, in the populous and well civilized town of Honoruru at Oahu, a skinned dog suspended at the door of a house of entertainment for natives, to denote what sumptuous fare might be obtained within.�

Dogs were paid as taxes and tribute. Dogs teeth were worn in bracelets and anklets.

Conclusion

Whereas with many cultures we cannot know if camp following was something that was actively sought by humans, in the islands of the Pacific we can be sure that devoting canoe space to a dog or two meant that they had real value to the ancient travelers. Unfortunately, two very large populations of dogs in Australia and New Guinea have so long been isolated from the cultures that carried them between islands, and the time frame of settlement is so broad, that it is uncertain who brought them or why. They are sufficiently different from the other dogs of the Austronesian expansion that this expansion must have occurred in several distinct periods and geographical areas if dingoes were part of it.

Oskarsson et al., by studying dogs, have made a significant contribution to the study of the settlement and cultures of the western Pacific. It is to be hoped that anthropologists will be able to integrate these results into further research.

The last plate is another of Hamilton Smith's dingoes.

Sources:

1. Allen, M.S. (2004). Revisiting and Revising Marquesan Culture History: New Archaeological Investigations at Anaho Bay, Nuku Hiva Island.  The Journal of Polynesian Society, 113(2) (noting dogs were brought to Marquesas Islands by Polynesian settlers perhaps as early as 150 BC).
2. Ardalan, A., Oskarsson, M.C.R., van Asch, B., Rabakonandriania, E., and Savolainen, P. (2015).  African Origin for Madagascan Dogs Revealed by mtDNA Analysis.  Royal Society Open Science, 2, 150552.  DOI: org/10.1098/rsos.140552 ("A small component of the Madagascan diversity seems to have originated from Europe, probably in the colonial era.  Other genetic contribution from Southwest Asia or India, as may be anticipated considering the long history of seatrade between the African east coast and the Persian Gulf region, and later the Indian subcontinent ... is not indicated but cannot be ruled out.").
3. Ash, Edward C. (1927). Dogs: Their History and Development, vol. 1. Ernest Benn Ltd., London.
4. Bellwood, Peter (2005). The First Farmers: The Origins of Agricultural Societies. Blackwell Publishing, Oxford.
5. Bryan, William A. (1915). Natural History of Hawaii. The Hawaiian Gazette Co., Honolulu.
6. Gill, W. Wyatt (1873-4). Three Visits to New Guinea. Proceedings of the Royal Society of London, 18(1), 31-49.
7. Gillespie, Richard (2002). Dating the First Australians. Radiocarbon, 44(2), 455-72.
8. Greenhill, Simon J., Drummond, Alexei J., and Gray, Russell D. (2010). How Accurate and Robust Are the Phylogenetic Estimates of Austronesian Language Relationships. Plos One, 5(3), e9573.
9. Hamilton Smith, C. (1856). Dogs, vol. I. Lizars, Edinburgh.
10. Hemmer, Helmut (2005). Domestication: The Decline of Environmental Appreciation. Cambridge University Press.
11. Higham, Charles (1991). The Archaeology of Mainland Southeast Asia. Cambridge University Press.
12. Horridge, Adrian (1995). The Austronesian Conquest of the Sea�Upwind. In Bellwood, P., Fox, James J., and Tryon, D. (eds.) The Austronesians: Historical and Comparative Perspectives. Australian National University, Canberra (�With the ability to carry fire, family, dogs, chickens, tuberous roots, growing shoots and seeds by sea, the Austronesians eventually occupied the Pacific Islands, travelling into Melanesia about 3500 years ago and onwards into Polynesia.�).
13. Jones, Rhys (1979). The Fifth Continent: Problems Concerning the Human Colonization of Australia. Annual Review of Anthropology, 8, 445-66.
14. Klutsch, Cornelya F.C., and Savolainen, Peter (2011). Geographical Origin of the Domestic Dog. Encyclopedia of Life Sciences. John Wiley & Sons, New York.
15. Koler-Matznick, J., Brisbin Jr., I.L., and McIntyre, J.K. (2000). The New Guinea Singing Dog: A Living Primitive Dog. In Dogs Through Time: An Archaeological Perspective: Proceedings of the 1st ICAZ Symposium on the History of the Domestic Dog. BAR International Series 889.
16. Koler-Matznick, J., Brisbin Jr., I.L., Feinstein, Mark, and Bulmer, Susan (2003). An Updated Description of the New Guinea Singing Dog (Canis hallstromi, Troughton 1957). Journal of the Zoological Society of London, 261, 109-118.
17. Koler-Matznick, J., and Stinner, M. (2011). First Report of Captive New Guinea Dingo (Canis dingo hallstromi) Den-Digging and Parental Behavior. Zoo Biology, 30(4), 445-50.
18. Lumholz, Carl (1902). Among Cannibals: An Account of Four Years' Travels in Australia and of Camp Life with the Aboriginees of Queensland. Charles Scribner's Sons, New York.
19. Luomala, Katharine (1962). Additional Eighteenth-Century Sketches of the Polynesian Native Dog, Including the Maori. Pacific Science, 16 (April).
20. Mulvaney, Derek John, and Kamminga, Johan (1999). Prehistory of Australia. Smithsonian Institution Press, Washington, D.C.
21. Oskarsson, Mattias C.R., Klutsch, Cornelya F.C., Boonyapracob, Ukadej, Wilton, Alan, Tanabe, Yuichi, and Savolainen, Peter (2011). Mitochondrial DNA Data Indicate an Introduction through Mainland Southeast Asia for Australian Dingoes and Polynesian Domestic Dogs. Proceedings of the Royal Society B (doi:10.1098/rspb.2011.1395).
22. Piper, Philip J., Hung, Hsiao-chun, Campos, Fredeliz Z., Bellwood, Peter, and Santiago, Rey (2009). A 4000 Year-Old Introduction of Domestic Pigs into the Philippine Archipelago: Implications for Understanding Routes of Human Migration through Island Southeast Asia and Wallacea. Antiquity (September 1, 2009).
23. Savolainen, P., Leitner, T. Wilton, A.N. Matisoo-Smith, E., and Lundeberg, J. (2004). A Detailed Picture of the Origin of the Australian Dingo, Obtained from the Study of Mitochondrial DNA. Proceedings of the National Academy of Sciences, 101(33), 12387-12390.
24. Tregear, Edward (1904). The Maori Race. A.D. Willis, London.
25. vonHoldt, Bridgett M., Pollinger, John P., Lohmueller, Kirk E., et al. (2010). Genome-wide SNP and Haplotype Analyses Reveal a Rich History Underying Dog Domestication, Nature, vol. 464(8).

Thanks to Richard Hawkins and Brian Duggan for suggestions.