The American legal history of tracking goes back to the dark days of slavery, as discussed by us elsewhere. That tradition produced a case law, and principles of evidence, very different from what began to arise in the 1970s for narcotics and explosives detection dogs. Tracking dogs were often trained as an economic activity, and well-known dogs could be brought over state lines�sometimes over several state lines�to track in notorious criminal cases. There was also a long tradition of prisoners training dogs, dogs that would be used to follow the trails of fellow prisoners who tried to escape. Since these dogs were often not owned by law enforcement agencies, legend was far more important than recordkeeping. A prison-trained tracking dog was involved in a recent case in Louisiana.
A Robbery in Natchitoches
The Tobacco Warehouse Convenience Store in Natchitoches, Louisiana, was robbed at gunpoint by two men. Two brothers, Corey W. and Andre Oliphant, were convicted in January, 2007. The appeal described here involves only one of the brothers, Corey, who was sentenced to forty years at hard labor without benefit of parole. The appeal came about after the federal district court for the Western District of Louisiana, on a habeas petition, essentially directed Louisiana to allow an appeal in the case.
The evidence at trial is summarized as follows.
On April 23, 2005, a man with a pistol entered the Tobacco Warehouse and took approximately $700 from store employees. The robber was described as black, with a hood pulled over his head. He also had a piece of cloth across his chin. At approximately the same time, a retired detention officer, living near the strip mall where the Tobacco Warehouse was located, saw a man with a hood over his head run through his yard, jump a fence, and get into the passenger side of a white, older-model Lincoln. He noticed the molding below the bottom of a door was missing.
Two hours later a police officer saw a Lincoln Town Car run a stop sign. The officer put on his lights but had to chase the car until it stopped. The car was being driven by Nicholas Oliphant and was registered to Odell Oliphant, respectively another brother of the defendant and his father. Knowing of the robbery, the officer asked Nicholas if he had a gun in the car, which he admitted he did. The officer put Nicholas in cuffs and took a nickel-plated, snub-nosed, 22 caliber revolver with a blued cylinder from the Lincoln.
Detention Center Tracking Dogs
Nicholas said that his brother, Corey, could verify that he had been at home all day. Corey was brought to the police station and gave an inconsistent account of his own whereabouts earlier in the day. A determination was made to use tracking dogs in the investigation and Corey was asked to provide a sock, with which he complied. A tracking team was brought to the Tobacco Warehouse. The sock was given to Officer Roy Gallien, a dog handler assigned to a nearby detention center.
Regarding the dogs, the Louisiana Court of Appeals states:
�According to Officer Gallien, the Detention Center has six tracking dogs including �Bo� and �Trusty,� the dogs used in this search, as well as a number of puppies in training. He testified that �the mamma dog� had come from Angola State Prison and that she had been bred to dogs from two other correctional facilities around the state. The dogs are not certified in any capacity, and all of their tracking expertise has arisen from use at the Detention Center. While Officer Gallien testified concerning at least two situations where the bloodhounds were useful in a search, he provided no information concerning the expertise of anyone involved in the training, including himself. Additionally, he acknowledged that the Detention Center keeps no records concerning the dogs� use.�
Officer Gallien also indicated that the dogs were trained by trustees of the Detention Center.
At the Tobacco Warehouse, Officer Gallien let Bo and Trusty sniff the sock from Corey Oliphant at about 7:15, five hours after the robbery. The dogs tracked separately �along a route similar to what� Officer Gallien would later learn was described by the retired detention officer. In a later part of the discussion, however, the court states that the dogs �neither tracked the exact same trail and both terminated their tracking efforts at different locations.� It is not detailed how far apart these two tracking endpoints were.
Corey Oliphant was arrested for armed robbery. The Lincoln was searched, producing two items made of panty-hose or stocking-type material. Searches of the car and the house, however, produced no clothing resembling that worn by the robber, and no money that might have been taken from the Tobacco Warehouse. Witnesses did not pick pictures of Nicholas or Corey Oliphant from a photographic lineup. One witness, pressed to pick the photograph closest to the robber, picked someone other than the brothers. The retired detention officer did not identify anyone. Nor did anyone identify the gun found in the Lincoln as the weapon used in the robbery.
Appellate Court�s Consideration of Tracking Evidence
The defendant made five assignments of error, one of which was the admission of the bloodhound testimony. The appellate court noted that there was no evidence of certification of the two dogs but added:
�[T]he only evidence of training was Officer Gallien's testimony to the effect that much of the dogs� training was accomplished by inmates at the Detention Center. Officer Gallien was not offered as an expert in the handling of bloodhounds, and the record contains no evidence of his training or the training of anyone else associated with the dogs. While Officer Gallien did testify to incidences of successful use of the bloodhounds, he acknowledged that no records were kept concerning the success or failure rate of operations involving the bloodhounds. When the defendant objected to Officer Gallien's testimony and lack of expertise and documentation, the trial court overruled the objection, noting that the objection would go to the weight of the evidence instead of its admissibility.�
The appellate court referred to the admission of bloodhound evidence as �an uncommon occurrence,� in contrast to narcotics detection dog evidence. While this is true, it perhaps overlooks the fact that the earliest criminal cases where dogs were part of the investigation, going back to the nineteenth century, involved tracking dogs. The court noted that where tracking dogs have been used, issues with regard to them have not often been raised on appeal. This is also true, and understandable. As with cadaver dogs, the discovery of a body often makes the question of how it was discovered irrelevant. If, say, a gun from a robbery is found by a dog along the path the robber used to flee the scene of the crime, there are commonly other ways to connect the gun with the robber.
The court then reviewed older Louisiana case law regarding bloodhounds, but found the best statement of a proper foundation for bloodhound evidence in a 1980 Tennessee case, Tennessee v. Barger, 612 S.W.2d 485 (Tenn.Crim.App. 1980), which stated that the handler had to be available for examination and the dog must:
- Be a purebred and of a type characterized by acuteness of scent and power of discrimination.
- Be accustomed and trained to track human (as opposed to animal) scents.
- Be shown by experience in actual cases to be reliable in tracking.
- Have been placed on the trail at a spot where the suspect was known to have been or on a track which circumstances indicate he made.
- Have been placed on the trail within a period of efficiency, i.e., before a rainstorm or lengthy passage of time.
Applying these factors to the case before it, the Louisiana Court of Appeals said that �there is insufficient evidence to establish the qualifications of the bloodhounds used to track the defendant�s exit from the Tobacco Warehouse.� The court emphasized the lack of certification, and aside from a few examples of success with the dogs, noted that Officer Gallien �could not assert any degree of overall tracking reliability because of a complete lack of records.�
In any case, the appellate court concluded that �the trial court erred in allowing the bloodhound evidence to be introduced to the jury.�
The court might have also noted research indicating that tracking dogs are less able to distinguish the scents of people closely related than of strangers. The fact that a number of brothers and a father were involved raises the possibility that the dog might have been tracking someone other than Corey. The court did consider that the tracking, which occurred five hours after the crime, might not have been in the dog�s period of efficiency. Many dogs would be considered efficient up to 24 hours, however, or even beyond. (For a discussion of the five elements listed here for admission of tracking evidence and their evolution over more than a century, see Police and Military Dogs, Chapters 3 and 5. Some of those elements were considered by courts during the early stages of case law on narcotics detection dogs. Thus, some courts considered whether drug dogs, like tracking dogs, had to be purebred. That requirement was put aside more quickly in drug dog jurisprudence, yet still receives mention in tracking dog cases.)
Certification of Tracking Dogs
The appellate court emphasized that the dogs were not certified in any capacity, that there was �no evidence of certification of the two dogs involved in the tracking,� and that there was �a lack of Louisiana jurisprudence on the subject of bloodhound certification.� This, the court felt, was to be contrasted with the importance of certification for drug-detection dogs, for which the court cited numerous state and federal cases, including the Supreme Court�s decision in U.S. v. Place, 462 U.S. 696 (1983). (There was no reference to the more recent decision in Florida v. Harris.)
Although the court felt the lack of certification was important, it apparently would not have excluded the bloodhound evidence had the five elements of Barger been satisfied. Those elements do not mention certification. Two of the five elements specifically relate to the facts of any case before a court�where and when the dog was placed on the trail. The other three relate to the dog and its background�its breed (that it be purebred), that it was �accustomed and trained to track humans,� and it be �shown by experience in actual cases to be reliable in tracking.� Thus, tracking cases often depend more on a dog�s actual production in the field than any certification. Why this is so�and that it should be so�is worth additional analysis.
Complexity of Tracking
In Florida v. Harris, 133 S.Ct. 1050 (2013), Justice Kagan compared field records to training and testing environments, stating:
�There, the designers of an assessment know where drugs are hidden and where they are not�and so where a dog should alert and where he should not. The better measure of a dog�s reliability thus comes away from the field, in controlled testing environments.�
The Justice observed that in the field it could be impossible to determine whether a dog�s alert, where no drugs were found, was genuinely false, or was rather an alert to a residual odor or an otherwise undetectable amount. We will argue in a forthcoming law review article that training and certification environments often contain residual odors and trace amounts of target drugs, but for purposes of analyzing this case, and tracking in general, it is to be noted that the situation in which the dog is placed and the circumstances of the assignment are very different.
A narcotics detection dog, when deployed to sniff a vehicle or other location, produces one of three results: no alert, interest without an alert, or an alert. For the most part, this is a binary analysis, alert or no alert. The Supreme Court, in Florida v. Harris, said that field records �in most cases have relatively limited import,� adding that �[e]rrors may abound in such records.� It is important to understand that, for any court analyzing a tracking case, this holding should not apply to tracking situations, where the dog�s choices cannot be described as binary. In fact, the dog�s potential actions can cover a broad set of patterns, some of which are indicated in the following table.
Where tracking begins: | Dog�s actions: | Evidentiary significance: |
Scene of crime | No trail found | None or minimal |
Trail followed but soon lost | May indicate direction perpetrator took on leaving scene | |
Trail followed leading to location where perpetrator may have entered vehicle and driven off | May be useful if evidence connects suspect with location or vehicle | |
Dog finds item connected with the crime | Significant in verifying path perpetrator took after crime; item may have independent value | |
Dog leads to house or building perpetrator may have entered | May lead to warrant for search | |
Dog leads to individual who is potential perpetrator | Sometimes admitted as identification evidence (but see Curran et al. finding that tracking may be accurate without identification at end of trail being accurate) | |
Dog loses trail but then resumes trail, perhaps even after handler has discontinued search | May excluded under some state requirements that tracking be continuous | |
Dog loses field trail but then is encouraged to resume work at entrance to police station, leading to suspect already under arrest | Sometimes admitted as station identification (improperly in our opinion) | |
Location where perpetrator or suspect was seen | Dog leads back to scene of crime | May be admitted concerning path perpetrator took after crime (a reverse track), though may also indicate perpetrator returned to scene of crime |
It is evident that many of these actions are far from two-choice situations such as are faced by drug dogs, that a dog following a trail is making countless assessments of the source and direction of an odor, sometimes producing significant evidence that can stand in its own right. This is not equivalent to the possibility of a dog alerting to residual odor in a narcotics case. Finding items associated with a crime, such as a weapon used in the commission of the crime or items taken from victims of the crime, should obviously not be evidence excluded merely because the dog involved was not certified. (Finding such items may also occur during an article search.)
Further, tracking in the field is probably more controlled, in a scientific sense, than a training or certification test environment. In a typical field deployment, the handler does not know at the beginning of the trail which way the dog will go. Nor does he or she know at any point along the trail when the dog will change course. It is possible that the handler in a training or certification situation will have some idea of these matters. The dog may be tracking in an area where other dogs have tracked before. The handler may have seen the instructor or tester place objects with the scent the dog is going to pursue. A field assignment is, thus, often much more of a double-blind situation than a testing or certification course. Consequently, field records of a tracking dog can be useful in assessing the dog�s reliability.
Also, because the tracking situation is not binary, there has seldom been an argument to correlate the results of tracking dogs to exclude those dogs whose deployments do not produce evidence. A drug dog that alerts more than others in a department may be suspected of being cued, and the Supreme Court in Harris may have wanted to avoid having full-blown trials of dogs on this basis. Tracking situations are too diverse for such a simplistic analysis and there is no need to analogize tracking evidence to drug dog evidence on the issue of certification.
Conclusion
The Supreme Court�s decision in Harris should not become a means of excluding valid field evidence in every type of police dog work. Even less should it become a reason for not keeping complete field records, or for destroying them before they may have to be made available to defense counsel.
Several months ago we analyzed a currency forfeiture action in which we discussed recent research on quantification of currency contamination by cocaine. Certain levels of currency contamination are expected on U.S. bills, not from the fact that most bills are handled by criminals at one point or another, but probably from mechanical currency counters distributing cocaine residue across bills that enter the machine. When a bill has above a certain level of contamination, however, it has a significant probability of having been associated with a drug trafficker. (See Jourdan et al., 2013.) If a dog alerts to it, the prevailing theory is that the dog is alerting to methyl benzoate, an unstable compound that disintegrates rapidly, so the alert of the dog indicates that the currency was probably in contact with cocaine in the previous two days. This creates some forensic formulas:
High residue + Alert = Drug enterprise contamination within 48 hours
High residue + No alert = Drug enterprise contamination older than several days
Normal residue + Alert = Recent contamination, perhaps innocent as from currency counter
These formulas oversimplify the possibilities, but do indicate why this kind of field alert evidence should be maintained, and should fit within the exception that Justice Kagan acknowledged in Harris where a �dog�s (or handler�s) history in the field � may sometimes be relevant�.�
For any court tempted to apply the training and certification logic of Harris to a tracking dog, these considerations should be taken into account before there is any knee-jerk dismissal of field evidence or a refusal to allow discovery of such evidence. The reliability of a tracking dog may be best established by field records, in contrast to the general devaluation of such records for drug dogs by the Supreme Court in Harris. The failure to maintain field records in this case was a sufficient reason to reject the tracking evidence. Records should be kept and even prison-trained dogs should have records kept of their deployments, though this case demonstrates that this has not yet happened in Louisiana. The Louisiana appellate court is right to say that it should.
Louisiana v. Oliphant, 2013 La. App. LEXIS 2386 (Ct.App. 2013)
This blog was written by John Ensminger and L.E. Papet.
