Does the cold nose know?
Canine "nose witnesses" may be more convincing than reliable.
Copyright Jessica Snyder Sachs, as first appeared in Popular Science
"The jury eats it up," says police K-9 handler Pat
McAlhany, a veteran of the Miami-Dade County Sheriff's Office. "From a
prosecutor's standpoint, there's nothing better than my actually bringing my
dog into the courtroom for a demonstration."
Indeed, the practice of using dogs as "nose witnesses" to finger, as
it were, the accused is rapidly growing. A rape case that passed through the
Los Angeles court system last year illustrates the procedure. A young woman
came into the police department and described a brutal attack. Two men had
dragged her from a bus stop in an industrial part of the city into an alleyway,
where they sexually assaulted her. Traumatized, she initially did not report the
attack. But two days later she found a necklace she thought one of the rapists
had ripped from her neck. It lay outside the school where she worked. Did one
or both of the attackers live in her neighborhood and know her daily habits?
She also spotted a man at a local market who reminded her of one of the
attackers. Now terrified, she went to the police.
The delay meant a rape kit could not provide DNA evidence. Not a problem. Enter
Reilly, a confident, experienced expert witness with boundless enthusiasm for
his job. Under the leash of volunteer scent-evidence K-9 handler Joseph
D'Allura, this chocolate Lab's scent-detection skills had put killers behind
bars. For the rape case, D'Allura created a "scent lineup," using a
scent transfer unit that worked like a Dustbuster, sucking the perpetrator's
odor off the snatched jewelry and onto a sterile gauze pad placed over the
vacuum's air intake. He then vacuumed a piece of clothing taken from the
neighborhood resident who had aroused the rape victim's suspicions in the local
market.
D'Allura prepared the scent lineup by placing the suspect's pad alongside three
decoys infused with scent from other individuals. Finally, he presented Reilly
with the necklace scent pad and directed him to find its match. Passing over
pad number one, the dog gave the second sample a wag and a bark--an
"alert" in dog-handling lingo. The police had a positive ID.
Or did they? Juries may love a dog show, but some experts remain skeptical.
"In all honesty, we don't know what a dog is picking up on when it
alerts," says Lawrence Myers, an Auburn University sensory and behavioral
biologist called as an expert witness for the defense in the rape case. Over
the past 21 years Myers has trained more than 50, and studied hundreds, of scent-detecting
dogs under laboratory and field conditions, mostly for federal agencies wanting
to perfect the use of dogs for finding explosives, drugs, trapped disaster
victims and hidden graves.
Myers considers the canine nose the ultimate odor-detection system in use
today. But in many ways, he says, "we're still dealing with a black
box." Scientists have yet to fully understand the process of canine odor
identification. Neither have they defined the limits of a dog's sense of smell,
nor isolated any universal "scent factor" that dogs use to
distinguish one person from another.
The widely held view is that dogs pick up on variations in the chemical makeup
of the skin flakes and perspiration residue we all continually shed. But no one
knows which aspects of this microscopic brew grab a dog's attention as it plays
a forensic game of mix and match. Might racial or gender differences skew
results? Or, for that matter, some yummy-smelling food the person ate the
previous day?
When it comes to something as complex as human scent, Myers
believes that different dogs likely tune in to different things. And no dog is
perfectly consistent. Myers's research has shown that something as mundane as
dental tartar can have a ruinous effect on a dog's powers of discrimination.
"Clean the teeth and you get an almost immediate recovery of smell,"
he says. Scent dogs also have off days, suffer allergies, get colds.
"Trouble is," he says, "you can't cross-examine a dog and ask,
'Are you sure?' "
So Myers has deep misgivings about the growing number of calls he's getting to
testify in rape and murder cases that hinge largely on a doggie's positive ID.
"It's like the floodgates have opened," he says.
Dog scent lineups aren't new, how-ever. Estimates suggest they've played a role
in more than 1,000 criminal cases in the United States, going as far back as
1923. Today, most states and the District of Columbia admit dog scent evidence
as valid identification of the accused, "provided a proper foundation is
laid."
Despite this long history, a scientific basis for that proper foundation has
never been established. Many prosecutors try to skirt that fact by arguing that
scent lineups are no more than a logical continuation of the practice of
employing dogs to track fugitives or sniff out drugs--uses that have long passed
legal muster. Some judges buy that argument, some don't.
The courts rely on handlers to demonstrate a dog's reliability by
submitting training records. But it's widely known in handler circles that many
are loathe to record a dog's mistakes, lest the errors later be used to
discredit the dog's identifications.
Research reveals that even experienced and well-trained dogs sometimes
misidentify individuals. In studies conducted in Europe (where scent lineups
have become tightly regulated and standardized), dogs' identification scores
varied from a high of 58 percent to a low of 22 percent. Researchers have
conducted such studies both on-lead (with a handler holding the dog's leash)
and off, in an effort to tease out or exclude a handler's influence.
That some handlers inadvertently prompt their dogs is a major criticism of
scent lineups. And the critics include some of the nation's most experienced
search-dog handlers. "We have a saying in dog training circles,"
explains police K-9 instructor Roger Titus, vice president of the National
Police Bloodhound Association. "Your body language goes down the
leash." A well-trained dog becomes hyperaware of its handler's every move,
Titus explains. "You lean forward, it moves. You slow down, it does too.
You walk down a lineup of six baseball caps, and all you have to do is think
you're at the right one, and the dog picks up on it." Some call it the
Clever Hans effect, after the 19th-century equine genius whose mathematical
prowess turned out to be nothing more than the horse responding to the
unconscious nod of his trainer's head.
Dog-evidence enthusiasts have reason to be cautious, Titus warns. Much of the
precedent-setting case law, including murder convictions, behind the acceptance
of scent lineups came from the now discredited work of a single handler, who
was later shown on videotape to be cuing his dog.
The expanding use of scent lineups has become a hot-button issue in the
police-dog handler community. The National Police Bloodhound Association and
the Law Enforcement Bloodhound Association have both developed guidelines for
proper lineup procedures, while their memberships remain deeply divided over
the legitimacy of the practice.
"The fact that a dog might send somebody to the electric chair places a
tremendous responsibility on the handler," admits McAlhany. "That
said, my confidence in my dog and myself is very high. I would feel comfortable
going to court and testifying that, yes, she can use a scent article and make a
positive ID."
Confidence in dog and handler aside, the scent-collecting
process itself is not above controversy. "Scent is a fragile creature to
begin with," argues Titus. "In my opinion, the use of scent machines
takes the whole idea further down the line of pipe dreams."
The scent transfer unit's inventor, Bill Tolhurst of the Niagara County (New
York) Sheriff's Department, dismisses that argument. Tolhurst says he's
successfully run dog scent lineups using evidence pads stored for more than 11
years in frozen, heat-sealed plastic bags. Tolhurst, a three-time past
president of the National Police Bloodhound Association, says he's pulled human
scent off shell casings from drive-by shootings. Among his proudest
accomplishments, he says, is a conviction in a kidnapping and attempted murder
case in which his bloodhound matched the defendant to scent vacuumed off the
seat of a car last driven by the offender.
To date, Tolhurst has sold more than 80 of his scent-collecting machines to law
enforcement agencies--including 35 to California police and sheriff's
departments and 7 to the FBI.
Lawyers will be arguing about the merits of dog-witness identifications and
procedures for years. In the rape case described earlier, the judge ruled
Reilly's scent identification inadmissible, citing, among other things,
D'Allura's lineup procedure. D'Allura stopped the lineup as soon as Reilly
alerted on the second pad, rather than allowing him to proceed down the line.
In addition, D'Allura couldn't recall when or from whom he'd made the lineup's
three decoy pads. Were they from individuals of the same race as the suspect?
D'Allura couldn't say.
Of course, having evidence dismissed is part of the legal process; win some,
lose some. D'Allura points to several cases in which Reilly's testi-mony played
a decisive role. In a major coup for scent lineup proponents, a California
appeals court recently upheld Reilly's identification of a teenager convicted
of a double homicide, as well as the young man's sentence of two consecutive
life terms.
Does the cold nose know? Criminal juries may be happy to say yes. But the
scientific jury remains out.
Jessica Snyder Sachs, a contributing
editor to Popular Science, is the author of Good Germs, Bad Germs:
Health and Survival in a Bacterial World (Hill&Wang/FSG) and Corpse: Nature,
Forensics, and the Struggle to Pinpoint Time of Death (Perseus/Basic Books).
Jump back to WEBSITE HOME.
