Apart from providing faithful companionship, dogs have a variety of useful skills that they gladly contribute to humans. But are dogs reliable expert witnesses? They might possess relevant expert knowledge, but they are handicapped by an inability to express their thoughts in human language. And since no human can read a dog’s mind, the human filter through which a dog’s knowledge must pass is inherently suspect.

The Florida Court of Appeals recently considered a murder conviction that was based almost entirely on dog handlers’ opinions that their cadaver dogs alerted on the defendant’s vehicle. Although no body was ever found, the court affirmed the admission of the expert testimony and thus affirmed the conviction.

Facts of the Case

Cid and Vilet Torrez were married. They separated in September 2011 after Torrez abused Vilet. Vilet remained in the home with their children while Cid moved to an apartment.

The home was in a gated neighborhood. Surveillance footage showed Vilet driving her car through the gate in the early morning. She had returned home after spending the night with a co-worker. The children were with Torrez.

Vilet was not seen again. Days later, Torrez called 911 to report that he had not heard from Vilet and believed her to be missing.

Police officers found Vilet’s car parked on the street. A search of the home found small blood stains at various locations. The police found no evidence that Vilet traveled or made telephone calls after the morning when she was last seen.

Suspecting that a homicide had occurred, the police summoned a cadaver dog named Jewel to the scene, along with her handler, Officer Gregory Strickland. Jewel alerted to several spots on the lawn near the front door. Strickland interpreted the alerts to mean that Jewel detected the odor of a dead body in those locations. No physical evidence supported that interpretation.

Five months later, having made no progress in determining Vilet’s location, the police did a “line-up” of cars in their parking lot. Torrez’ vehicle was one of the cars. Jewel sniffed each car and, according to Strickland, alerted on the trunk and back seat of Torrez’ car. Strickland interpreted the alert to mean that Jewel detected the scent of a dead body.

The local police then asked for the assistance of Palm Beach Sheriff’s Detective Juliana Martinez and her dog Piper, who was also trained as a cadaver dog. Martinez had Piper sniff Torrez’ vehicle and interpreted Piper’s response as an alert to the odor of human remains in the trunk and back seat.

Vilet’s body has never been found. On the strength of human interpretations of “alerts” given by two dogs, Florida charged Torrez with murder. A jury convicted him. Torrez appealed, challenging the expert testimony given by the dogs’ handlers.

Challenge to Dog Handling Expert

Most people are familiar with the story of the horse that could do arithmetic. When asked “What is two plus three?” the horse would stamp its hoof five times. It was eventually determined that the horse could only perform the trick when its owner was present. The horse was reacting to visual cues from its owner, not to any understanding of numbers.

During the car “line-up,” Strickland claims not to have known which vehicle belonged to Torrez. Whether Jewel was able to see other officers in the parking lot who did know which vehicle belonged to Torrez is unclear. Perhaps both dogs alerted to Torrez’ car because they were responding to visual cues rather than scents.

Prior to trial, Torrez challenged the admissibility of the expert testimony that Strickland and Martinez proposed to give. Strickland testified that Jewel had hundreds of hours of training in the detection of human remains. She was certified as a cadaver dog by a police association that certifies police dogs. Strickland testified that he was only aware of one instance in which Jewel alerted in the absence of human remains. On that occasion, she apparently detected the odor of a bucket of shrimp.

Martinez and Piper’s trainer testified about the training and certifications that Piper received. They agreed that Piper is a reliable cadaver dog. Martinez explained that when Piper isolates the source of the odor of human remains, she “snaps her head, sometimes closes her mouth, sniffs certain areas, slows down and then sits as a final response.”

Kenneth Furton, a professor of chemistry, testified about the scent molecules that cadaver dogs are trained to detect. He contended that scent molecules can linger for “a very long period,” particularly in an enclosed area. He opined that Piper and Jewel, in combination with their handlers, were reliable teams because of their certifications.

Furton did not believe that the absence of a body in Torrez’ back seat or trunk invalidated the alerts. He speculated that the dogs may have alerted to bodily fluids that leaked into materials and were not detected by other means, or they may have alerted to residual odors that remained after the body was removed. Furton admitted that dogs, like humans, can make mistakes.

The trial court purported to apply Florida’s newly established Daubert standard. It determined that the dogs were trained and accurate in detecting human remains and that their handlers were qualified to interpret the dogs’ alerts. Vilet’s disappearance was circumstantial evidence of her death that, in the court’s view, corroborated the handlers’ conclusions that her body had been transported in Torres’ car.

Appellate Analysis

The Florida Court of Appeals noted that the United States Supreme Court has approved determinations of probable cause based on alerts given by properly trained drug dogs. But probable cause to search for evidence is not itself evidence. The question on appeal is whether a handler’s expert interpretation of a dog’s alerts is sufficient to satisfy the Daubert standard for the admissibility of expert evidence in a trial.

The court nevertheless concluded that the opinions of the cadaver dogs’ handlers satisfied the Daubert standard. The court held that the reliability of dog sniff evidence can be based on the handler’s experience with the dog. The court must be satisfied that the handler is “qualified to work with the dog and to interpret its responses.” But since no handler can read a dog’s mind, it is difficult to imagine any circumstances that qualify a handler to “interpret” a dog’s actions reliably.

The dog must also be “proved successful and reliable” and be “sufficiently trained.” There must also be evidence that corroborates the dog’s opinion as interpreted by its trainer. Finding an actual body would presumably be corroborative, but “corroborative evidence need not be evidence which, standing alone, links the defendant to the crime.” That holding is unfortunate for Torrez, given that there was precious little evidence linking him to the crime beyond the handlers’ opinions that their dogs were smelling evidence that a body had been in Torrez’ car and trunk almost five months earlier.

Remarkably, the appellate court held that “courts need not consider the science underlying testimony relating to cadaver dog evidence.” In civil cases, Daubert hearings are almost entirely devoted to the adequacy of the underlying science that supports an expert’s opinion. Why should a lesser standard be applied in a criminal case, where the evidence may lead to a deprivation of liberty?

The court held that it is common knowledge that dogs can distinguish different kinds of odors. But it isn’t common knowledge that dogs can distinguish the scent of a decaying body several months after the body could have been present in the location that the dog sniffed, or that they can reliably explain what they smelled to a human.

This case cries out for scientific evidence, but the court pointed to no peer-reviewed studies suggesting that scent molecules can be detected by a cadaver dog almost five months after the cadaver was no longer present. Nor did the court point to peer-reviewed studies suggesting that the scent molecules associated with cadavers can be reliably distinguished from the scent molecules associated with shrimp or other substances.

Finally, the court held that challenges “to an expert’s measurements, methods and determinations do not render inadmissible an expert opinion based on them but goes to the weight of the evidence, raising factual questions to be determined by the jury.” The court cited only pre-Daubert criminal cases for that proposition. After Daubert, expert opinions have been routinely excluded because an expert’s methods were unreasonable and because an expert’s determinations were not founded on the application of a reasonable methodology to adequate facts.

The appellate court said that it was applying Florida’s new Daubert standard to dog sniff evidence, but its shoddy opinion rests entirely on pre-Daubert understandings of whether expert evidence is admissible. The unfortunate result for Torrez is that his conviction was affirmed based largely on the opinions of police officers about what their dogs might have smelled in his car.