Original URL: http://www.theregister.co.uk/2006/05/12/animal_detection/

Science you can sniff at?

My police dog's got no nous...

By Amber Marks

Posted in Security, 12th May 2006 02:02 GMT

Analysis Ever wondered where the expression "barking up the wrong tree" comes from? It comes from dogs, who have been getting the wrong end of the stick for donkeys' years.

Ironically, it is a dog's knack of following mistaken lines of inquiry that makes it such a reliable police surveillance tool. The police describe false alerts by dogs as a sign of their "oversensitivity".

Oversensitivity is a much hankered after characteristic in surveillance technology: the Ion Track narcotics machine was recently praised for its "efficiency" when Edwina Hart, the social justice minister in the semi-autonomous government of Wales and William Graham, a member of the Welsh Assembly, who had arranged for police to demonstrate the machine at the Assembly building, were found to have been in contact with cannabis.

The Department of Transport and British Transport Police (BTP) recently published a glossy leaflet outlining trials to be conducted on rail and tube systems to improve security following the London terrorist attacks in July 2005. The leaflet explains that they will be experimenting with different forms of electronic passenger screening - x-ray screening of bags and the use of sniffer dogs.

The Observer has labelled the prevalence of sniffer dogs on our railway and underground stations as "the public face of terror". A drug squad officer said: "The law made it so difficult for us to stop and search people, we had to find a way round it."

Is oversensitive detection equipment the latest means for getting round the legal prerequisite of reasonable suspicion?

Police dogs are trained to detect either explosives or drugs; they can't generally do both. A BTP spokesman said drug detection dogs were being used in London underground stations because "we've got more [drug] dogs", though he wouldn't disclose how many canines were being used for this purpose. "We don't want people to know otherwise they will know our capabilities. We want to keep people guessing."

Action on Rights for Children (ARCH) became aware of "the problem" of drug dogs in Leytonstone station when two young people associated with ARCH were subjected to humiliating and aggressively-conducted searches following indications from drug dogs.

One 16-year-old boy was taken outside in the cold and made to remove his jacket. The police officer told him "the dog is never wrong". When the police had finished searching him and confirmed there were no drugs on his person, they became less aggressive towards him, but said he must have the smell of cannabis on him from somewhere he had been. He explained he had been at a Department of Education and Skills meeting all day.

The man in the newsagent kiosk inside the station, who has witnessed several dog operations, told ARCH the dogs are often wrong. His friend is a chef and the dogs bark at his clothes every time he visits the kiosk. ARCH is now collecting complaints from young people with tales of unjustified police harassment.

The BTP spokesman said it's "impossible to tell" how often dogs are wrong in their indications. Apparently, they are so sensitive "they can even detect the scent on you if you have brushed past someone who is carrying drugs".

He said "an indication from a dog is not an indication that the person has done anything wrong. What it does is provide probable cause for a search".

Normally one would imply the other, but no longer, apparently.

New surveillance technology is so sensitive it can provide police with reasonable grounds to search people without necessarily indicating they have done anything wrong.

The only empirical research into the public deployment of drug dogs was conducted in New South Wales, Australia, and revealed that over a 12 month period 73 per cent of people searched as a result of dog indications had no drugs about their person.

Despite their "oversensitivity", dogs are apparently not sensitive enough to detect target substances. A recent investigation into Chicago Metra Train system found that a bomb-sniffing dog that "protects" commuter railways failed to detect a pound of explosives concealed in luggage less than six feet from its nose and, in another case, a dog and handler did not discover explosives in luggage placed in front of the dog for 30 seconds.

Unlike other novel detection equipment, such as scanners, consent is not requested before a person is sniffed. According to BTP, this is because the use of a scanner amounts to a search, and a sniff by a dog doesn't. The BTP's reasoning is that a person has to walk though a scanner, whereas they don't have to walk past a dog.

In the Supreme Court case Kyllo v United States, the police use of a thermal imaging device was found to amount to a search because it revealed information that would previously have only been obtainable by physical intrusion. The Supreme Court reasoned that if the laws applicable to physical searches were not applied to searches made with sense enhancing technologies, the values enshrined in the law would cease to be protected with the advance of technology.

The Supreme Court failed to apply this reasoning to the use of the drug dog on the assumption that a drug dog was an infallible detection tool that picked up the scent of drugs alone, and that regulating its use would only shield information unworthy of protection. In his dissenting judgment, Justice Souter pointed out that: "The infallible dog...is a creature of legal fiction." He said "if constitutional scrutiny is in order for the imager, it is in order for the dog".

Are the police relying on the myth that "a dog never lies" to take advantage of false positives and obtain personal details of innocent citizens?

The BTP spokesman said dogs "are being used more and more because there are a growing number available and they are very useful. On being indicated by the dog, persons are stopped and searched and a record made containing their personal details and a physical description. Later, if there is a report of a robbery, the police can go through those records to see if it matches any of the persons known to have been in the area".

One US website describes this practice disparagingly as "pigs using dogs to fish humans". It is certainly a novel means of creating a database of potential suspects.

The use of dog indications to justify stop and searches has only recently come under legal scrutiny, and principally in the United States. In 2003, the District Court of Appeal in Florida (Matheson v State of Florida) found that the false response rate of dogs meant that an indication from a dog could not by itself provide the police with reasonable grounds to conduct a stop and search.

This decision threw the use of dogs as detection tools into disarray and in the ensuing controversy it emerged that dogs were regularly relied upon in courts to provide evidence of identification. Scent line-ups are conducted in which the dog is asked to match a scent from the crime scene to the scent of a suspect.

"People have been convicted of robbery, rape, and even murder when the primary evidence against them is, effectively, a bark."

According to Dr Jozef Wojcikiewicz, professor of Forensic Science, Department of Criminalistics, Jagiellonian University and Institute of Forensic Research, Poland, "canine identification of human scent does not yet have a proper scientific foundation...the method has been introduced into trial proceedings too early, by overly hasty police practitioners which have caused miscarriages of justice".

Since the court’s decision in Matheson v State of Florida, the research staff at the International Forensic Research Institute at Florida International University have been busily trying to provide "concise scientific validation of canine detection".

The institute has been generously funded over the years. On 17 February, 2006, it was awarded $246,634 by the US government "to design, develop and test a method to improve the performance and scientific defensibility of dog teams used for human scent identification".

"Scientific defensibility" is an interesting choice of words, presumably reserved for techniques with no scientific foundation. ®

Amber Marks is a barrister. She is undertaking doctoral research into olfactory surveillance with the Law Department and Forensic Science and Drug Monitoring Unit at King's College, London. Email your comments or experiences of sniffer dogs to her at amber.marks@gmail.com >