Wednesday, September 22, 2010

Dog scent lineups discredited at TX Court of Criminal Appeals

Outstanding news for those concerned about the use of shoddy forensics in criminal courtrooms: As of today, dog-scent lineups are no longer adequate to secure a criminal conviction in Texas.

In a move that long-time court-watchers may find shocking, the Texas Court of Criminal Appeals today reversed a murder conviction which had resulted in a 75-year sentence and ordered the defendants acquittal based on the unreliability of dog-scent lineups by Deputy Keith Pikett from Fort Bend County, whose bizarre and unreliable practices are well-known to long-time readers of this blog. (See initial coverage from AP and the Austin Statesman) Richard Winfrey, Sr. was was represented on appeal by Dallas attorney Shirley Baccus-Lobel (who in the interest of full disclosure happens to be a boardmember at one of my former employers, the Innocence Project of Texas). The ruling was unanimous, with the exception of Judge Meyers who did not participate in deliberations.

The entire case against Winfrey was based on two pieces of particularly shoddy evidence: A dog-scent lineup and a jailhouse snitch. (See the court's opinion, authored by Judge Barbara Hervey.) However the attorney at trial had not challenged the scent lineups, Baccus-Lobel emphasized when I spoke to her this afternoon, so the CCA could only rule on the legal sufficiency of the evidence, which they found inadequate. Here's how Judge Hervey described the case:
In August 2004, Murray Wayne Burr was found murdered in his home. Evidence at trial indicated that the victim had been stabbed twenty-eight times and had received multiple blunt-force injuries, including a broken right-eye orbit and a broken jaw. There was no evidence of forced entry into the victim's home. The evidence indicated that the victim was dragged from his living room to his bedroom where his body was found. Family members reported that the only item missing from the victim's home was a Bible.

Investigators collected a variety of forensic evidence from the crime scene including: a partial bloody fingerprint, a bloody shoe print, and several hair samples. Neither the prints nor the hair samples matched appellant. Investigators were able to obtain a DNA profile from evidence at the crime scene, however, the profile excluded appellant and his family members. (1) Appellant's children, Megan (then 16 years of age) and Richard Winfrey Jr. (then 17 years of age), became persons of interest in the murder investigation. Texas Rangers interviewed appellant approximately two weeks after the murder. Appellant was not considered a suspect at this time.
Then in 2006 a jailhouse informant came forward to claim Winfrey, Sr. had told him information he had "heard" about the murder, most of which turned out to be false. The informant said Winfrey Sr. told him a gun and knife collection had been stolen (the victim's family said only a Bible was missing) and that Burr's penis had been mutilated (untrue). At this point:
To assist in the investigation, Texas Ranger Grover Huff contacted Deputy Keith Pikett, a dog handler with the Fort Bend County Sheriff's office. Deputy Pikett testified about a "scent lineup" that he conducted nearly three years after the murder in August 2007. He used his three bloodhounds, Quincy, James Bond, and Clue. This involved obtaining scent samples from clothing that the victim was wearing at the time of his death and from six white males, including appellant. The dogs were "pre-scented" on the scent samples obtained from the victim's clothing. The dogs then walked a line of paint cans containing the scent samples of the six white males. All three dogs alerted on the can containing appellant's scent sample.

Based on this, Deputy Pikett concluded that appellant's scent was on the victim's clothing.
That was enough for the jury, who convicted Winfrey and sentenced him to 75 years. Winfrey's son, Richard Jr., was also accused but was acquitted at trial. However, Winfrey's daughter, Megan, who was 16-years old when the murder occurred and knew the victim, who worked at the school she attended, was certified to stand trial as an adult and convicted of capital murder and conspiracy to commit murder based on essentially similar evidence. She received a life sentence on the capital murder and 45 years on the conspiracy charge. Her direct appeal is presently pending, according to her lawyer Scott Pawgan, who told me he plans to file a supplemental brief in the wake of her father's acquittal. So it's possible this may not be the last reversed conviction stemming from this case before all is said and done.

Deputy Pikett, whose magic dogs provided the main testimony (Woof woof!) against Winfrey, retired earlier this year after the Innocence Project of Texas published a highly critical report (pdf) about his dog-scent lineup practices, revealing among other things that he'd exaggerated his professional credentials in the precedent-setting court case that established his dogs' testimony as admissible forensics. Judge Cathy Cochran wrote a short concurrence emphasizing that the Court could not consider the admissibility of the dogs' testimony because it hadn't been challenged at trial, but this ruling sets the stage for them to do so in the future when the appropriate case finally reaches them. Judge Hervey's ruling already will substantially limit scent-lineups use in the future, as evidenced in this notable excerpt:
We note, however, that the science underlying canine-scent lineups has been questioned; thus, we think it proper to briefly address the issue. Law-enforcement personnel have long utilized canines in crime management. For example, dogs have been employed for detecting narcotics and explosives, for tracking trails, in search-and-rescue operations, for locating cadavers, and for discriminating between scents for identification purposes. In thousands of cases, canines and their handlers have performed with distinction. Despite this success, we acknowledge the invariable truth espoused by Justice Souter that "[t]he infallible dog, however, is a creature of legal fiction."

This case pertains to canines used to discriminate among human scents in order to identify a specific person in a lineup. This process is often referred to as human-scent discrimination. Some courts, including the Fourteenth Court of Appeals, have determined that for purposes of admissibility, "there is little distinction between a scent lineup and a situation where a dog is required to track an individual's scent over an area traversed by multiple persons." Other courts, such as the Supreme Court of Florida, have distinguished scent lineups from dog tracking. ("[I]t is important to recognize that using a dog to track a human or to detect the presence of drugs or explosives is distinctive from using a dog to directly identify a specific human from items in a lineup.").

Cases involving the use of dogs, usually bloodhounds, to track humans are abundant and the law is well settled in regards to admissibility of such evidence with only a minority of courts outright rejecting bloodhound evidence. Fewer courts have addressed the question of whether dog evidence is sufficient to sustain a conviction when it is the only evidence. However, as early as 1913, our colleagues at the Supreme Court of Mississippi held that dog tracking evidence, alone and unsupported, to be insufficient to affirm a conviction. And as recently as 1983, the Supreme Court of Washington agreed. In fact, our research suggests the courts that have passed on this issue have concluded that dog-scent evidence, when admissible, is insufficient, standing alone, to sustain a conviction.

Like our sister courts across the country, we now hold that scent-discrimination lineups, whether conducted with individuals or inanimate objects, to be separate and distinct from dog-scent tracking evidence. "Even the briefest review of the scientific principles underlying dog scenting reveals that, contrary to the conclusions of many courts, there are significant scientific differences among the various uses of scenting: tracking, narcotics detection, and scent lineups." Andrew E. Taslitz, Does the Cold Nose Know? The Unscientific Myth of Dog Scenting, 42 Hastings L.J. 15, 42 (1990) (explaining that drug detection canines need only determine whether a specific scent is present. Tracking dogs, on the other hand, have the benefit of using both vegetative scents and human scent, while canines performing scent lineups must find one specific scent among many competing, similar scents). The FBI agrees, noting that tracking canines use human scent and environmental cues to locate the track of an individual. Allison M. Curran, et al., Analysis of the Uniqueness and Persistence of Human Scent, 7 Forensic Sci. Comm. 2 (2005). Accordingly, we conclude that scent-discrimination lineups, when used alone or as primary evidence, are legally insufficient to support a conviction. Like the Supreme Court of Washington, we believe that "[t]he dangers inherent in the use of dog tracking evidence can only be alleviated by the presence of corroborating evidence." To the extent that lower-court opinions suggest otherwise, we overrule them and expressly hold that when inculpatory evidence is obtained from a dog-scent lineup, its role in the court room is merely supportive. (Legal citations omitted.)
IPOT's Chief Staff Attorney Natalie Roetzel wrote this afternoon in an email, "Needless to say, we are very excited about this development today.  We have a good handful of dog-scent lineup cases in the hopper, and we are optimistic that there may be some more exoneration stories in the works."

Congrats to Baccus-Lobel and Mr. Winfrey, and for that matter to the Court of Criminal Appeals for taking what's for them a highly unusual step. Now what's needed is for somebody, somehow, by some means, to vet other old cases involving Deputy Dawg Pikett's so-called forensics. It's a virtual certainty these aren't the only cases where Pikett offered the main or only evidence on which Texas prosecutors obtained convictions.

See prior, related Grits posts:


Anonymous said...

One of the best ways to defend those falsely accused is to discredit the testimony of witnesses. Evidence produced by canines must also be discredited.

Prison Doc said...

I'm a big fan of dogs as workers, trackers, and pets...but holding them out as experts beyond the scope of their proven abilities does no good for either dogs or justice....

Anonymous said...

I've never met a bloodhound that could follow a scent.

Anonymous said...

Accepting a dog scent is like accepting the testimony of a TYC administrator - slimy and stinky.

escalante blogger said...

But sometimes this canines have a proof when scent are present in the crime scene.

Gritsforbreakfast said...

escalante, read the IPOT report before claiming that. Canines don't offer "proof," because dogs can't say what they smell.

This technique was developed under totalitarian regimes in East Germany and Cuba and has never withstood rigorous testing. For that matter, even narcotics dogs are wrong about half the time. The idea that their "testimony" can be proof beyond reasonable doubt is foolish, at best.

8:58, I've hunted with tracking dogs (admittedly decades ago in my youth) so I know there are dogs who can follow a scent, but they can also lose it or get off on the wrong trail.

9:24, leave the TYC bashing to the TYC strings. I'd like to be able to discuss other topics on this blog without every conversation heading to the same place.

roy said...

Dog-scent lineups beg to be abused. Check Wikipedia for "Clever Hans", a man's horse that could do arithmetic, or at least read people's expectations.

Anonymous said...

Your snide criticism of Pikett's testimony generally ignores the fact that Pikett himself has said that dog scent evidence should not be used by itself to convict. That statement was made in this very case. This opinion does not "discredit" dog scent evidence at all, it merely points out what Pikett himself says: that a person's scent on an object only proves a connection--not necessarily direct contact--between the person and the object. Some other evidence must be offered to corroborate the connection as being proof of involvement in the crime, as opposed to an innocent transfer of the scent.

The IPOT "Report" you constantly refer to is a joke. Anyone who actually reads it will see that it is utterly conclusory without any research to back it up. If you had done some of your own, you would see that the foremost research in dog scent evidence doesn't come from totalitarian regimes, but from the Netherlands. You might look at G.A.A. Schoon's 1995 article in Applied Animal Behavior Science.

Dog scent evidence is certainly not proof-positive evidence of guilt, the way DNA or fingerprints can sometimes be. But DNA and fingerprints are also not tell-tale proof of crime, unless other evidence shows that only the criminal could have left the DNA or fingerprint. The point of the CCA opinion is just that you have to take dog scent evidence for what it is, and not try to make anything more of it.