Does it pass the smell test?
By Leslie Wilber - firstname.lastname@example.org
July 12, 2009 at 2:12 a.m.
Updated Sept. 25, 2009 at 4:25 a.m.
A tall, soft-spoken black man robbed one elderly woman of her life's savings and sexually assaulted another. He attacked the women in their homes, early in the morning, as they left for church.
Shops in the small town of Yoakum sold out of Mace. Women took self-defense classes.
As pressure mounted on law enforcement to make an arrest, an unnamed source told investigator Collin Lee Campbell Jr. that Calvin Lee Miller was unemployed and "had been buying a lot of cocaine with cash."
Miller, a 43-year-old Yoakum native, is a former high school football player of imposing bulk. He speaks quietly. Police know him as an habitual - though petty and non-violent - criminal.
To divine Miller's connection, the investigator swabbed the suspect with a gauze pad. He took that pad and scent from the victim's sheets to Deputy Keith Pikett in Fort Bend County.
Pikett is the only dog handler in the state who performs scent identification lineups. Pikett's three bloodhounds indicated Miller's scent was on the sheets.
Miller spent 62 days in jail. Finally, he was cleared by DNA evidence and the victims' failure to identify him.
"It was just like my life was taken away from me," he said.
No laws or regulations govern scent lineups, but they're admissible in courts across the nation. Only tighter oversight can keep shoddy - and sometimes shady - scent IDs from becoming key evidence, a growing number of critics say.
"This is junk science. This isn't even science. This is just junk," said Jeff Blackburn, chief counsel for the Innocence Project of Texas. The group works to free the wrongfully convicted and began investigating Pikett recently.
Miller's case and several others across the country reveal problems with scent dogs:
- During a murder investigation, the home of a former Victoria County sheriff's captain was searched after Pikett's dogs led officers there, then identified his scent in a lineup. Another man later confessed to the murder.
- A California man, accused of arson, awaited trial in jail for two years. His arrest was based largely on scent evidence collected from charred shrubs. He was acquitted.
- Two Florida men each spent more than 20 years in prison - one convicted of rape, the other of murder. The evidence against both included testimony from a dog handler the Arizona Supreme Court later called a charlatan.
How scent IDs work
Two ideas underlie scent identification: Dogs have a keen sense of smell - sometimes 10,000 times more sensitive than humans - and everyone has a unique scent. The challenge is to train dogs to overtly and accurately alert observers to matching smells.
The best-run scent lineups can provide results as accurate as witness identification lineups, said Kenneth Furton, the chairman of the federally funded Scientific Working Group on Dog and Orthogonal Detector Guidelines. Furton is a chemistry and biochemistry professor at Florida International University. For a model, he points to the Netherlands, where lineups are closely regulated and happen in near-sterile rooms without human interaction.
But training for dogs in the United States is often left to lone handlers or agencies.
"There isn't really any requirement to adhere to any kind of best practice," Furton said. Lineups are not as widely used as other scent evidence - for example, drug sniffs - so the technique has glided under regulatory radars, he said.
The National Police Bloodhound Association quit endorsing the technique years ago, calling scent lineups unreliable.
"We don't even want to take a chance on that," said Doug Lowry, the group's president and a chief instructor.
Kevin Kocher, president of the National Bloodhound Training Institute, said he doesn't run lineups and finds them hard to defend.
Critics list several potential weaknesses of scent identification:
Handler or observer influence. Dogs are eager to please and can pick up subtle cues, especially if lineups are conducted on leash, said Steven Nicely, a police-dog handler-turned-defense witness. Handlers can read observers and unintentionally relay that information to the dogs.
"They learn about the pressure on the leash and the way you stand," Nicely said.
Sample contamination. Lineups typically include a suspect's scent and scents from five other people. The samples should all be fresh and about the same age because scent fades over time.
The pads should be handled carefully, to avoid contamination, and lineups should be conducted in clean rooms, without distracting smells. Human scent is best stored in glass jars at room temperature and out of direct light, Fulton said.
Handler reliability. Dogs can't talk, so handlers are their voice in the courtroom. Affidavits should be precise. Records should be detailed, showing errors and successes.
"As a dog handler, you'd better be acting as a scientist," Nicely said. "Otherwise, you're acting on myth and folklore."
Michael Buchanek's federal lawsuit in January 2008 was the first complaint about Pikett's work, said Randy Morse, an assistant Fort Bend County attorney representing the deputy. A scent lineup and trail identified Buchanek as a suspect in the high-profile murder of Sally Blackwell.
Pikett's dogs, Quincy and James Bond, walked from the site where Blackwell's body was found, along Zac Lentz Parkway, to her home more than five miles away. From there, they went to Buchanek's house nearby.
The hounds also picked the former Victoria County sheriff's captain's scent in a series of lineups. The identifications were used to get a search warrant for Buchanek's home.
His lawyer, Rex Easley, represents Calvin Miller in another civil suit that names Pikett.
In one motion, Easley wrote Pikett's lineup was "so recklessly flawed that it violated the constitutional rights of the plaintiff. First, the dogs were leashed during the lineup, which fails to exclude handler input. Second, the site, the pads and the cans were contaminated with countless other scents so as to render it unreliable and impermissible to base a warrant upon."
Easley hired Bob Coote, who led a police-dog force in the United Kingdom and worked with scent dogs guarding the British border, to review Pikett's work in Buchanek's case. The lineup was "the most primitive evidential police procedure I have ever witnessed. If it was not for the fact that this is a serious matter, I could have been watching a comedy," Coote wrote.
In California, Michael Espalin spent two years in jail awaiting an arson trial. Firefighters saw Espalin walking his dog near a Riverside, Calif., landscape fire in 2004. Fire Capt. Robert Rappaport knew bloodhound handler and community college professor Lisa Harvey.
Harvey used a method similar to scent lineup and claimed her bloodhound, Dakota, connected a scent from a matchbook wrapped in toilet paper to a series of fires. Then, she linked the fire-starting device to Espalin's home.
Espalin was arrested, although another suspect - a reported firebug - fit witness descriptions. The other man's truck looked like one seen near several fires.
"She was all fired up about this," said attorney Joseph De Clue, who represents Espalin in a civil suit filed in federal court in California. "A little knowledge is a dangerous thing. She looked at it as a new way of solving crimes."
A jury heard Harvey's testimony in January 2007 and deadlocked. Harvey did not testify in the second trial. Espalin was found not guilty.
Dog handlers and criminal justice advocates draw parallels between current scent identification cases and the work of a man discredited more than 20 years ago.
John Preston's German shepherd, Harass II, gained a super-canine reputation for his work on high-profile, violent crimes. The team's techniques included scent lineups and trailing.
The Pennsylvania state trooper and his dog were called to investigations around the country in the early 1980s.
Florida authorities used Preston in scores of investigations, estimates Seth Miller, executive director of The Innocence Project of Florida. Preston's testimony was part of the evidence that convicted William Dillon of murder and Wilton Dedge of rape.
By the mid-1980s, the Kings County District Attorney in New York gathered a list of 14 dog handlers who said Preston's claims were "highly improbable." The Arizona Supreme Court called Preston a charlatan.
Dedge and Dillon were exonerated in recent years after each spent more than 20 years in prison. The state of Florida paid Dedge a $2 million settlement. Dillon has not been compensated for his 27-year prison stay, which is as long as any exonerated inmate's, Seth Miller said.
Preston has since died.
Agencies, including the FBI, continued to use scent identification even after Preston's work was debunked, former dog handler Nicely said.
Special Agent Ann Todd with the FBI office of public affairs responded to several requests for interviews with an e-mail.
"The FBI does not wish to participate in an interview at this time. The FBI Laboratory's Forensic Canine Program does not currently use dogs to do scent ID lineups," she wrote.
Handlers, too, resist the idea of regulation.
During a pretrial hearing, Harvey said no group would be qualified to evaluate her dogs.
Although Pikett works full time for a sheriff's office, supervisors have not set guidelines for his work, his lawyer said. No one else understands what he does.
"He would just be writing regulations toward himself," his lawyer, Morse, said.
Morse advised against allowing his client to comment for this article.
How does evidence few people understand, and no one monitors, make it into court?
Some prosecutors and investigators support scent identification because it can offer leads where there were none.
San Jacinto County District Attorney Bill Burnett used Pikett as an expert witness to prosecute three co-defendants in a murder case. One was convicted of murder, another of capital murder and the third was acquitted.
"I felt like this evidence was certainly credible," Burnett said. He grew up around hunting dogs and knows they can do incredible stuff, he said.
Big investigative agencies - the FBI, the Houston Police Department and the Bureau of Alcohol, Tobacco, Firearms and Explosives - have used Pikett, Burnett said.
An outright ban on a type of evidence is rare, said Eric Ferrero, director of communications for The Innocence Project.
"It's the jobs of courts to police whatever evidence should be allowed into the courtroom," Ferrero said.
Getting binding appellate court decisions about evidence is often a difficult, inefficient process, said Blackburn of The Innocence Project of Texas.
"It takes a good lawyer who raises the issue correctly at trial," Blackburn said. Many public defenders are ill-equipped to mount meaningful criticisms of questionable evidence, he said.
Because it is unmonitored, seeing the full picture of scent ID, with all of its potential flaws, is difficult, Blackburn said.
"As far as we know, no one disallows it," Blackburn said. "That doesn't mean it's being used."
But in prosecutor-friendly states like Texas, California and Florida, he said, scent IDs can get into courtrooms unchecked.
Pikett was disqualified from testifying twice when held to standards for scientific forensic evidence.
"There is no way to adequately examine the theory behind a dog's identification of a suspect," Houston defense attorney Daphne L. Pattison wrote in a motion that successfully barred Pikett's testimony. "We have no way to interview the dog and determine what the dog is smelling, no way to examine the dog's biology to determine the mechanism employed, in short, no way to understand the science."
Pikett's credibility has been upheld by at least one Texas court of appeals. The decision relied on Pikett's testimony that his dogs were almost flawlessly accurate.
The court also noted Pikett's experience and education, including a master's degree in chemistry.
According to transcripts from a pre-trial hearing, Pikett twice told the court he earned the advanced degree.
In a deposition for the Buchanek case, Pikett said that wasn't his testimony. Instead, he has a master's in education from the United States Sports Academy.
The Rin Tin Tin factor
Heroic, loyal dogs lope through the American consciousness in folklore, television and books. Consequently, scent evidence can be compelling once it's in front of a jury, Blackburn said. Most people are raised to trust cops and dogs, former handler Nicely said, and people are reluctant to question those values.
"They don't want to step on the nose or the ears or the tails of Rin Tin Tin or Old Yeller," Nicely said.
The demand for Pikett's services remains, and he performs scent lineups daily in the parking lot of Pilgrim Journey Baptist Church in Rosenberg. Although his attorney, Morse, contends the lineups are legitimate, he said more suits could halt Pikett's work.
"If he becomes a target, I don't know if it's going to be worth it," Morse said.
Waiting for courts to flush bad science from the legal system is slow, Nicely said, so he wants a legislative solution. Detailed records of training and field work should be mandated, he said.
"We need to make it a criminal act to fail to do this," Nicely said.
The Scientific Working Group for Dog and Orthogonal Detection Guidelines is drafting scent lineup guidelines. The group will likely suggest an international board to oversee certifying agencies, chairman Furton said. The agencies would regularly vet dog and handler teams, he said.
Even with certification, Furton said, no criminal case should be built on scent lineups alone.
The lasting effects
Innocence Project volunteers could work for years, Blackburn said, and never find out how many people are wrongfully imprisoned based on what one man says his dogs smelled. The group learns of cases slowly, one at a time.
The Innocence Project of Florida is still searching for cases in which John Preston testified, more than 20 years after the fact, executive director Seth Miller said.
In South Texas, Calvin Miller is trying to pick up the life he had before his arrest.
Good news came to Miller slowly in jail. In April, his defense lawyer, Bill Caraway, called to say DNA evidence proved Miller was not the rapist. The case against Miller unraveled from there, but he was not released until May 5.
Caraway met Miller in the lobby of the Lavaca County Jail. Both men cried. They drank ice-cold, bucket-sized soft drinks in celebration.
But Miller's relief and joy have not tempered his outrage or his fear of police. He moved away from Yoakum, the town where he grew up, attended high school and lived most of his life. He visits his parents there, but never strays far from their houses.
"The place you call home, you can never go there," Miller said.
The crimes he was accused of remain unsolved.
Miller still can't understand how an innocent man can be so easily locked away.
"His question was, 'If I didn't do it, how could those dogs say I did it?'" his attorney Easley said. "And I told him dogs can't talk."