Wednesday, September 28, 2016

California Supreme Court Brings Flawed Station Identification Procedure Back from Brink of Deserved Oblivion

On May 13, 2001, Geraldine Myers disappeared from her home in Riverside, California.  Her car was found in Las Vegas, Nevada, but her body has never been found.  Myers was 82 years old.  In the early hours of June 22, 2001, Myrna Mason, an 84-year-old woman living in the same neighborhood, was raped in her home.  Bailey Lamar Jackson was arrested the day after the rape, and evidence implicating Jackson in Myers’ earlier disappearance was developed during the investigation of the rape of Mason.  California v. Jackson, 1 Cal.5th 269 (2016).

There were many strands of evidence regarding the murder of Geraldine Myers and the rape of Myrna Mason, and the connection between the two crimes—the evidence that there was a single perpetrator—was only partially made by the use of trailing dogs.  That aspect of the case, however, will be the sole focus of this analysis.  Whether a conviction would have occurred had the trailing evidence been excluded, as the authors believe much of it should have been, is doubtful.  The Supreme Court of California, unlike some lower California courts, did not state anything to the effect that, even if the evidence would properly have been excluded, its admission was harmless error.  The judicial mantra of harmless error has been repeatedly invoked by courts not wishing to face the fact that scent identification evidence, particularly in the U.S., is often highly flawed, as it was in Jackson.  (See, e.g., People v. Mitchell, 110 Cal.App.4th 772, 2 Cal.Rptr.3d 49 (2003); People v. Schoppe-Rico, 140 Cal.App.4th 1370 (2006).)

The case is notable for the number of canine procedures used, with the California Supreme Court describing seven procedures involving at least six dogs. 

The Riverside Sheriff’s Department Canine Team

Riverside Deputy Sheriff Coby Webb began training as a bloodhound handler in 1998 and obtained Maggie, her second police dog, in 1999.  Maggie completed 20 hours of training at a Southern California Bloodhound Handler Coalition Seminar in July 1999 when she was four months old.  She also attended the Colorado Canine Training Academy in September 2000, where she was certified to trail human scent.  In April 2001, Webb and Maggie completed 40 hours of training through the Law Enforcement Bloodhound Association.  They were also certified by the National Police Bloodhound Association. 

Maggie testified that she continued to train Maggie for 10 to 15 hours a week, following trails laid by other officers or volunteers. Since many departments limit paid training time to four hours a week, either the Riverside Sheriff’s Department has a very liberal canine training policy or Maggie was training outside of normal hours.  This should have been a matter of inquiry for defense counsel to make sure that it does not include grooming and other non-training-related activities.  Maggie was given “negative trails” in which the dog was scented to an object by someone who had not been in the training area.  In such cases, Maggie was rewarded for not trailing.  Maggie was also trained with double-blind trails, where neither Webb nor the evaluator knew the correct path (and where, presumably, anyone who did know the trail was not visible to the dog while it worked).

The Trail from Mason’s House

Deputy Sheriff Webb brought Maggie Mae to Mason’s house the morning of June 22, the day of the rape.  She also brought a scent transfer unit and gathered scent from a shoe print in freshly raked dirt outside Mason’s home.  She presented the scent pad produced by this process to Maggie, who trailed across the street where a garbage can was found to contain Mason’s purse.  It is not specifically stated in the opinion that the purse was found in the garbage can by the dog handler or someone with her, though the fact that the dog was brought back to this point after it became confused may indicate this was the last point that Webb was sure the dog was trailing the scent obtained from the shoe print.  Subsequent to reaching the garbage can, Maggie led Webb to the porch of Richard Shrader, a neighbor with whom the defendant, Jackson, lifted weights, then to the back yard of Billie Harris, another woman living in the neighborhood.  Harris’s daughter, Angelina Fortson, was living in the house with Jackson, who was her boyfriend.  In the back yard, Maggie “seemed confused and appeared to have lost the trail.” 

During testimony, Webb said that the dog’s confusion in Harris's back yard may have been the result of “pooling,” which she described as occurring “when many trails of the target scent overlap at a single location,” which “can occur at the residence of the person who left the scent the dog is trailing.”  Because of the dog’s confusion, Webb brought the dog back to the garbage can to start over.   The choice of the garbage can is somewhat troubling because the standard practice, when a dog loses a track, is to go back only as far as the closest point at which the dog was known to be trailing or tracking. With the handler standing firm at this point, the dog is allowed to search at full lead length while attempting to re-acquire the track. Alternatively, a sweeping motion can be used to guide the dog into an oblique angle (300-450) intercept in the direction of travel of the last known track point. Both of these methods are sometimes referred to as “casting.”  Defense counsel should have inquired as to why this was not done, say at Schrader’s porch or just prior to entering the back yard of Billie Harris.  

This time, Maggie “trailed down the street past the Harris residence and then appeared to have lost the scent.”  No explanation was provided in the Supreme Court’s opinion as to why the dog followed a new and apparently unproductive trail from the garbage can where the purse was found. The issue should have been of particular concern to defense counsel.

Station Identification

Later the same day Webb and Maggie went to the Spruce Street police station, where Jackson was being detained as a result of other evidence concerning the rape.  Webb was asked whether Maggie could eliminate a suspect who had been in the lobby and was now somewhere else inside the station, but Webb was not told where the suspect was or what path he had taken.  “In the lobby, Webb presented Maggie with the scent pad collected from the shoe print outside Mason’s house.” Maggie first trailed to one interview room that Jackson was not in, and then to the interview room where he was being held.  It was not stated whether Jackson had been in the first interview room or not.  In the second interview room, Maggie stopped in front of Jackson but, according to the court’s summary of Webb’s testimony, “did not ‘commit herself to jump up on’ Jackson….”  Webb testified that when Maggie “gets confused, she will just stop, which tells me she did not know which subject was the possible suspect.”

Webb apparently vacillated on the significance of Maggie not jumping on Jackson.  According to the California Supreme Court:

Webb testified that during training Maggie had been taught to indicate that she had found the person she was trailing by jumping up on them. Webb later testified that it was not necessary for Maggie to jump up on someone to indicate that she had finished trailing. Instead, when Maggie would stop trailing and stand still in front of someone, it was enough for Webb to question that individual to see if “that’s the person we are looking for.”

Webb also testified that the air conditioning system in the Spruce Street station may have been confusing the dog because a target “scent is going all through the air-conditioning vents.”  Webb “had to pull Maggie out of the room, which further suggested to Webb that Maggie had located the person she was trailing.”  Since the dog did not jump up on the suspect, there was no trained final response and the authors believe the evidence of the dog’s interest in either the first or the second interview room was insufficient to justify its admission as evidence of Jackson’s identity as the perpetrator, and was sufficiently speculative and prejudicial as to be inadmissible.

Jackson’s Description of Throwing a Woman from a Car

During an interview regarding the rape, Jackson described a memory of having thrown a woman from his car.  Police began to suspect that Jackson, who stated that he took Haldol, an antipsychotic medication, and Cogentin, which is used to reduce the side effects of certain antipsychotic drugs such as Haldol, was mixing up memories of the events of Myers’ disappearance with those of Mason’s rape.  Jackson thought he could find the place where he had thrown a woman from his car, but a search of the area with a trained dog did not result in any body being found.  It is not clear if this was a cadaver dog or a tracking or suspect apprehension dog scented to one of the scent pads produced by the scent transfer unit. In any case, no body or evidence of a body was found and as already indicated, Myers’ body has not yet been found.  

Second Station Identification

A second station identification was conducted on June 25, 2001, three days after the first, at a different location, the Orange Street station.  This procedure is described as follows:

To lay the trail, Jackson was led through the station by Barnes and Johnson from the entrance, around at least two corners, and into a men’s locker room. The detectives sat Jackson on a bench, closed the door, and stayed in the room with him. They were dressed in casual clothing, while Jackson was wearing an orange jail jumpsuit.

Defense counsel should have explored whether Barnes and Johnson could have themselves have been laying scent from the crime scenes, having sat beside Jackson in a cruiser (as they likely were on June 22), or having been in contact with objects that Jackson himself may have come in contact with.  The continued presence of the officers in the room meant the procedure was not double-blind.Jackson was also the only black man in the locker room and wore handcuffs in addition to the orange jumpsuit. 

As to Webb’s preparation for the procedure, the Court states the following:

Webb testified that she asked the detectives: “'Can I please have three turns to make sure Maggie is committed to trailing, and I need to have it where there’s no air conditioner on and I don’t know where the subject is.'” Although she did not know the path that Jackson took through the basement, she did need to know “where the subject had walked into the building.... to match the trail with the scent item or not match, I needed to point [Maggie] on a trail. So I just needed to know where the person first walked into the building.”

It is not clear why Webb could not have scented the dog outside the building, walked her to each entrance, and seen where the scent was picked up by the dog, which would have verified the dog's willingness to trail outdoors.  In any case, on the sidewalk outside the station Webb presented Maggie with a crumpled manila envelope that had been found on Myers’ bed, inside of which a sterile gauze pad had been placed. The first trailing on June 22 had involved scenting Maggie to odor taken from a footprint with a scent transfer unit. It was not explained if the footprint had been deemed made by someone other than a perpetrator, perhaps because of the lack of clear identification during the first station procedure.  The fact that a second station identification was conducted after a first was unsuccessful may also indicate that, at this stage of the investigation, the sheriff's office was concerned that not enough non-canine evidence might be found to convict Jackson or force a plea deal. 

It is not clear why the gauze pad was placed inside the envelope since crinkling up an envelope as trash will leave substantially more odor of the person doing the crinkling on the outside of the envelope than the inside.  It should have also been investigated whether Barnes, Johnson, or others had touched the envelope.  The envelope had been sprayed with ninhydrin to test for fingerprints before it had been placed in a Ziploc bag with the gauze pad.This would not have been a problem with the footprint scent, and resulted in efforts to discount the effects of ninhydrin on the dog's ability to take a scent from the envelope, as will be described further below. 

It took Maggie 10 to 15 minutes to trail from the starting point [at one entrance], around both corners that Jackson had turned, to the locker room. When Webb opened the locker room door, Maggie walked past the detectives, walked down the row where Jackson was seated, and put her paws on the bench and her head next to his chest. Webb testified that this indicated that Jackson’s scent was on the manila envelope.

At a later point in the opinion, the Court's description of these events differs somewhat, saying that "Webb lost sight of Maggie when the dog turned the corner to walk down Jackson's row."  Also, instead of saying that Maggie's paws were on the bench and her head next to Jackson's chest, the court in its second description says that when Webb reached the row where Jackson was seated, "she saw Maggie jump into Jackson's lap."  The inconsistency of these descriptions was not explained.
The defense objected that “the presence of the two detectives in the locker room transformed the Orange Street trailing into a hybrid of a trailing and a scent lineup,” using the terminology employed by Professor Andrew Taslitz in his famous Hastings Law Review Article, “Does the Cold Nose Know? The Unscientific Myth of the Dog Scent Lineup.”  The California Supreme Court rejected the argument:

Although Barnes testified that there were communications personnel in another room in the basement while Maggie was trailing, Jackson does not argue that their presence transformed the trail into a scent lineup, nor does he argue that as a general matter dog trailing can only be reliably conducted in areas where other people are not present. Crucially, at no point was Maggie presented with a line of multiple people or objects and asked to distinguish among them; instead, the detectives and Jackson were waiting in different parts of the locker room. Webb testified that she could not see Jackson when she followed Maggie past the detectives and did not see him until she turned the corner to watch Maggie jump into his lap.

This defective logic misses the fact that there is an entire history of distinguishing between tracking or trailing and interacting with individuals who might be found at the end of the trail, a logic that has infused the scientific development of scent lineups and which more astute courts have seen as meaning that a dog’s interaction with an individual at the end of a trail can be influenced by factors other than an identity of scent between the individual at that point and the odor with which the trailing procedure began. Dogs, for instance, can distinguish emotional expressions of human faces, and it is likely that someone like Jackson would have a different type of expression on his face than would officers near him.  See Muller et al. (2015).  Dogs Can Discriminate Emotional Expressions of Human Faces.  Current Biology, 25(5), 601-605; Cuaya et al. (2016). OurFaces in the Dog's Brain: Functional Imaging Reveals Temporal Cortex Activation during Perception in Human Faces. PLoS/One, 11(3), e0149431.

While one cannot control many variables in a real-world trailing or tracking situation, it was the realization that variables could be defined in scent identifications, and thus controlled, that led to the scientific tradition of scent identification procedures, specifically lineups, that have been refined by Adee Schoon, Tadeusz Jezierski, and recently by Barbara Ferry.  (See Tadeusz Jezierski, Scent Lineups: Variables in Procedures and Statistical Verification, chapter 19 in Canine Olfaction Science and Law.)  

The Court was apparently not apprised of research, some of which has been discussed here previously, indicating that dogs that follow trails with high statistical success have lower success rates at the end of these trails when they are presented with a choice between several individuals and are expected to identify the individual who left the trail that they had correctly followed.  That one individual in a room is wearing a jumpsuit and is in handcuffs is an obvious invitation to cueing by the handler, which does not need to be a conscious manipulation of the leash.  (See Ensminger and Papet, Cueing and Probable Cause, article online at Michigan State University College of Law’s Animal Legal & Historical Center, as well as Police and Military Dogs, Chapter 6 at 96.)

Webb was asked—whether on direct or cross examination is not stated—whether the first trailing at the Spruce Street station may have influenced the result at the Orange Street station.  The Court quotes her reply as follows:

“No. We run numerous trails. An example, like this month I believe I’m up to 14 callouts. So I change scents, I mean, almost on a daily basis. And she—I’ve never seen her able to remember something—a trail of somebody else. I’ve never, ever seen that.”

The defense expert, Dr. Lisa Harvey, whose testimony will be discussed in more detail below, also testified that dogs trained to trail human scent will not follow a scent simply because they have been asked to follow the scent on past trails. Trainers of tracking and trailing dogs, however, when using a prior scent for testing a dog's ability to avoid cross-tracks of other scents often use prior odors as a proof.  Many dogs, particularly newer dogs, will follow a previously encountered scent rather than maintaining focus on the correct trail.


The fact that the envelope used to scent the dog at the Orange Street station procedure had been sprayed with ninhydrin raised the possibility that a dog might not be able to reliably follow a trail upon being scented on an item that had been so treated.  This possibility was subjected to field experiments by two experts for the prosecution, Maryland State Trooper Douglas Lowry and Dr. Lisa Harvey.  Lowry used one dog for his experiment, while Harvey used three.  Dr. Harvey has conducted some of the most important trailing research in the last twenty years.  Both experts concluded that the presence of ninhydrin on an envelope would not significantly alter a dog’s ability to detect a human’s scent on the envelope and follow a trail of an individual who had put the scent on the envelope.  

Apparently no research was cited regarding the effect of ninhydrin as a masking agent for odor detectible by dogs, so the need was felt to conduct field experiments to provide at least tentative (though unrefereed) evidence that ninhydrin would not interfere with a dog’s ability to trail.  The results are not surprising.  Waggoner et al. (1998 SPIE Proceedings) notes that “although the detection performance of dogs is susceptible to being perturbed by the presence of an extraneous odor, it takes a large or even very large amount of this odor in relation to the amount of target odor for this effect to be realized. The extent of this effect depends on the nature of the substances.”

Pre-Trial Motion for Kelly Hearing on Canine Evidence

Before trial, the defense moved for a hearing on the admissibility of the canine evidence under California v. Kelly, 17 Cal.3d 24, 549 P.2d 1240 (1976), which defines California’s Frye requirement with regard to scientific evidence.  Kelly requires that (1) the scientific technique involved be “generally accepted as reliable in the relevant scientific community,” (2) the witness testifying concerning the technique and its application must be “a properly qualified expert on the subject,” and (3) “the person performing the test in the particular case used correct scientific procedures.”  However, California case law provides that “Kelly/Frye only applies to the limited class of expert testimony which is based, in whole or in part, on a technique, process, or theory which is new to science and, even more so, the law.”  (Quoting California v. Stoll, 49 Cal.3d 1136, 783 P.2d 698 (1989).)  Relying particularly on California v. Craig, 86 Cal.App.3d 905, 150 Cal.Rptr. 676 (Ct.App. 1978), the Supreme Court of California concluded in the present case:

Scent trailing evidence is not so foreign to everyday experience that it would be unusually difficult for jurors to evaluate. Jurors are capable of understanding and evaluating testimony about a particular dog’s sensory perceptions, its training, its reliability, the experience and technique of its handler, and its performance in scent trailing, such as performed in this case.

The Court also argued that it was “unlikely that a juror would believe that dogs are scientifically infallible….”  This presumption that juries are largely neutral to canine evidence and participation in trials should, in the opinion of the authors, be subjected to sociological research.  (See discussion of jury influence in “Cases and Statutes on the Uses of Dogs by Witnesses While Testifying in Criminal Proceedings” on the website of the Animal Legal & Historical Center.)  The effect of dogs is known to advertisers.  See Lancendorfer, K. M., Atkin, J. L., & Reece, B. B. (2008). Animals in Advertising: Love Dogs? Love the Ad! Journal of Business Research, 61(5), 384-391.

On the issue of trained canine fallibility, the Court referred to the testimony of State Trooper Douglas Lowry, an expert witness for the prosecution. 

On cross-examination, Lowry testified that in practice, he observed dogs change their trail from an older scent to a newer scent, and that he observed this more often with newer dogs. He also testified that dogs, like people, have good days and bad days, as well as days when they simply do not want to work.

The fact that dogs can change what they are trailing is one of the reasons for a frequently mentioned trailing requirement that a dog be put on the trail during the period of its reliability.  Since there were samples that preserved the scent of the crime scene, the issue becomes whether the scent remained in sufficient strength to be used for a trailing activity.  This issue is particularly relevant to the station identifications performed by Harvey four years after the crimes. 

Harvey’s Station Identification Procedures in February 2005

After the defendant was found guilty, but before his sentence was imposed, there was a penalty phase that had to be repeated because the jury deadlocked in the first penalty trial.  During this penalty-phase retrial, Dr. Harvey testified concerning two additional station identifications she performed in February 2005.  These procedures were performed in the San Bernardino police station, where Jackson had never been before.  The purpose of the procedures was, according to the Supreme Court, “to determine whether Jackson’s scent was on a gauze pad that had been stored with the envelope recovered at Myers’s house.”  It was not stated, and perhaps not explored yet again, whether Jackson may have been brought to the station and taken to a location by officers who themselves could have been bearing the suspect's scent. 

A video of the procedures was shown to the jury and is summarized by the Court as follows:

First, Dr. Harvey scented her dog, Shelby, off the gauze pad that Traughber had placed inside the envelope found on Myers’s bed in June 2001. Shelby trailed through the sally port, smelled three doors, and indicated that she wanted to go through one of them. On the other side of the door, she trailed down a corridor and indicated she wanted to go through one of two doors. Shelby arrived in another corridor containing eight locked jail cells. Shelby smelled all eight and then stood still in between the seventh and eighth cells. The door of the eighth cell was opened; Shelby went inside, smelled the detainees, and walked out again. The seventh cell, containing Jackson, was opened; Shelby went inside, smelled Jackson, and walked back out. Dr. Harvey testified that this was not an identification “to her satisfaction.” Dr. Harvey further testified that she “felt that Shelby had a good trail” but was being “rather lazy.” Because Shelby “was not trying to trail out of that area” and “was going around and around in circles,” Dr. Harvey opined that “that smell that she was looking for was in that area, but for some reason she just refused to make an identification.”

So the dog, given a choice of eight cells, could not choose between two of them and the choice was made for her.  Then, because the dog did not leave the vicinity, it was concluded that a match for the scent she had been given was somewhere near. Harvey then repeated the procedure with a second dog:

Dr. Harvey conducted the same task with her other dog, Dakota. Dakota walked through the sally port and the same series of doors as Shelby had until she arrived on Jackson’s corridor. Dr. Harvey opened Jackson’s cell door, Dakota entered, sniffed around, and walked out again. Dr. Harvey gave Dakota the verbal command “show me,” but Dakota just stood there and whimpered. Dr. Harvey testified that Dakota’s behavior constituted an identification….

When asked what the dog’s behavior indicated, Harvey testified:

A. That indicates to me that she was still doing her job; however, she chose to do her job the way she wanted to do it, which was in direct conflict to how I would like her to do it, which is a jump-up.... [E]ach dog, I guess you could say, chooses their own identification, and the handler has to learn that identification from the dog....

Q. What does Dakota usually do?

A. Dakota does one of two things. She will either do what you saw her do, or she will do a jump-up identification.

As with Webb’s dog, a trained final response—assuming the jump-up was trained—was not performed, yet the handler claimed that the dog had made an identification apparently based on an undefined level of interest.  The Supreme Court, to its credit, realized that this had not happened:

Notably, neither Shelby nor Dakota positively identified Jackson. Even though Shelby began to trail immediately, indicating that she had found a match between the scent item and the scent trail, she did not unambiguously alert on Jackson. Instead, she went into and out of two locked rooms, smelling everyone in both rooms, including Jackson. Yet, as Dr. Harvey testified, Shelby did not “choose to make an identification” in her usual way. Likewise, according to Dr. Harvey, Dakota began to trail immediately following the same route, perhaps even a little faster than Shelby. She also sniffed several people behind the two locked doors, including Jackson. Yet she did not unambiguously identify Jackson. Dr. Harvey testified that she “wasn’t happy with [Dakota’s] identification,” stating that she would “prefer [Dakota] actually jump on them, so I can see a good identification.” Even after being directed to identify, Dakota refused and began to cry and whine. Dr. Harvey testified that she believed Dakota had made an identification, but she had “picked” her own way of making the identification, contrary to her training.

Harvey acknowledged that dogs, including her own, do make mistakes.  According to the Court:

Dr. Harvey, on both direct and cross-examination, testified that the accuracy of dogs was correlated with their age, with “veteran dogs” having a success rate of 95 percent, while novice dogs had a success rate of 60 percent. Dr. Harvey also admitted that even well-trained dogs have good days and bad days. When asked if she thought Maggie made mistakes, Dr. Harvey responded, “Oh, yes, I’m sure.” Testifying about her own dog, Shelby, she stated that she could tell “for sure” that Shelby had made mistakes in practices, even if she was not sure of mistakes made out in the field.

Despite acknowledging the possibility of mistakes, Dr. Harvey believed an identification had been made.  This belief was apparently based on the dog’s training:

“[W]e give consistent training to the dog. Then, when we take them to an unknown area trying to trail someone, as long as the dog is working in the same manner that they worked during training, we are able to say that this is consistent; therefore, we say it is reliable.”

This comes close to the frequently made argument that even if a dog alerted when it should not have, the dog was not wrong because there must have been residual odor present.  This becomes a means by which the scientific evidence for odor recognition becomes irrelevant because every alert has an explanation that is not the dog’s fault.  In the narcotics and explosives detection world, this makes dogs into “walking search warrants.” Here, if broadly accepted, it could make them into walking identification witnesses. 

Another defect of poorly conducted scent lineups and station identifications occurs when dogs may follow each other because the earlier presence of a dog at the same location has not been antiseptically removed (See Jezierski, above). The Court noted that the failure to make an unambiguous, positive identification “arguably worked to [Jackson’s] advantage.”  Nevertheless, the prejudicial effect of allowing an expert to state that the dog had made an identification should have resulted in the evidence being declared inadmissible. 

Testimony of Dr. Lawrence Myers

The defense called Dr. Lawrence Myers, who has been an expert witness in a number of scent identification cases.  Myers testified as to the complexity of human scent and how not knowing what chemicals dogs rely on in trailing humans made it, in the Court’s words, “hard to determine the reliability of an identification based on an older scent in which certain organic compounds may have degraded or otherwise changed.”  He also testified that the extent to which “various contaminations” might affect a dog’s abilities was unclear.

Dr. Myers testified that a dog can be cued visually to alert by the fact that only one person in an array is wearing a bright jumpsuit and handcuffs. He also described how a handler can cue a dog by walking at a different speed, tugging or letting up on the leash, or changing her tone of voice. He said he would be “very concerned” if the target or decoys in a scent identification task were known to the dog, although he could not say that it would “definitely” affect the outcome.

This testimony raises questions about Lowry’s use of a decoy that was known to the dog in his ninhydrin experiment. 

Myers also indicated that a scent lineup where only one of the individuals in the lineup was in a jumpsuit and handcuffs was “not a very good lineup” because the situation involved “a cue that’s hard to ignore.”  In rebuttal, the prosecution asked Webb if dogs are “colorblind,” to which Webb said they were.  This is not entirely true, though dogs have fewer color-sensitive cone photoreceptors than humans do.  See Miller, Wyman (2001). Vision in Animals-What Do Dogs and Cats See? PP. 27-8 in the 25th Annual Waltham/OSU Symposium: Small Animal Ophthalmology.  In any case, a dog would not need to distinguish ethnicity by sight, since there are odor markers, volatile organic compounds, that likely allow them to distinguish members of different races.  See Prokop-Prigge et al. (2015).  Ethnic/racial and Genetic Influences on Cerumen Odorant Profiles. Journal of Chemical Ecology, 41, 67-74; Prokop-Prigge et al. (2016). The Effect of Ethnicity on Human Axillary Odorant Production. Journal of Chemical Ecology, 42, 33-39.

Thus, there does not seem to have been any real exploration of what Myers was saying, that the circumstances may point to an individual in a group for a dog just as they may point to that individual for a person looking at the group. 


Canine procedures occurred at least seven times in this case:
  1. Trailings of Webb and Maggie May at Mason’s house the morning of June 22, 2001. Result: first trailing went to backyard of residence where suspect was living; second went elsewhere.
  2. The station identification of Webb and Maggie later the same day at the Spruce Street police station. Result: no alert, dog stopped trailing at suspect but may have been confused by air conditioning.
  3. A canine search of the area where Jackson thought he may have thrown a woman from his car. Result: no body or other evidence found.
  4. The station identification of Webb and Maggie on June 25, 2001, at the Orange Street police station. Result: Court provided two descriptions of the dog interacting with Jackson, one of which indicated an alert. Since dog was scented to envelope that had been sprayed with ninhydrin, subsequent verification of effects of ninhydrin were undertaken.
  5. Trooper Lowry’s ninhydrin-related trailing test using a dog (pre-trial but otherwise undated). Result: ninhydrin found not to interfere with dog's abilities.
  6. Dr. Harvey’s test with three ninhydrin-related trailing tests (also pre-trial), using three dogs. Result: same as 5.
  7. Dr. Harvey’s station identification of Harvey and two dogs in February 2005 (post-trial, between two sentencing trials).  Result: handler claimed identification occurred by Court's description only indicated interest. 
In sum, there were three identification procedures, two near the time of the crime and one (involving two dogs) four years after the crime.  Only one of the earlier identifications may have resulted in an alert.  In none of these procedures were there adequate scientific protocols and in only one station identification was there even arguably a trained final alert. Only in the first procedure, trailing from Mason’s house, was valid evidence produced that should have been admissible, and that evidence was of the route that the perpetrator may have followed upon leaving Mason’s house, and where the perpetrator left the purse taken from Mason.  Even in that case, a repeat of the trailing from the garbage can did not confirm the first result so arguably even these events should have been excluded as prejudicial.  

Other non-canine evidence was far more damning than anything the dogs did, including items of both victims found in Jackson’s possession, blood evidence that could be connected with Jackson, his confused memories of having thrown someone from his car, and testimony of other parties placing Jackson near the crime scenes.  The cumulative evidence, without any of the station identifications, might well have been enough to convict Jackson.  Inconclusive results from questionable canine procedures should not have been allowed to muddy the evidentiary waters in this case.

Although the California Supreme Court seems certain that jurors are not overly influenced by canine testimony and will naturally understand that dogs can make mistakes, the Court itself uncritically accepts some claims that should have been subjected to rigorous cross-examination. 

[Webb] also described challenging tasks that Maggie had performed. In one instance, Maggie followed the scent of someone who wrote a letter that was mailed to Webb and kept in the freezer for four weeks. In another instance, Maggie identified the person who had handled an explosive device by following the scent that remained after the device was detonated. Maggie performed successfully on this task four times out of six and did not falsely alert on the remaining two trials.

Although a frozen letter may still have scent that can be detected and used to scent a dog, the circumstances of the trailing were not described (at least in the Court’s opinion), including whether this was a double-blind test.  Following the trail of a person who had handled an IED sounds remarkably like a 2010 study published in Forensic Science International and the subject of a prior blog on this site.  That study did find that dogs could be very accurate in following the trail of someone who placed an IED, but were less accurate in identifying these individuals from others in a room where the “perpetrators” had gone after leaving the devices. 

Andrew Taslitz, whose brilliant skewering of the scent lineup procedure as employed by American law enforcement remains the most trenchant analysis of the weakness of canine identification evidence, contacted me after reading an unpublished article I wrote with Tadeusz Jezierski and Michael McCulloch, on which we still receive occasional suggestions for new studies to incorporate.  Professor Taslitz, sadly no longer with us, then wrote an article that appeared in Criminal Justice in 2013 (vol. 28, p. 4), "The Cold Nose Might Actually Know? Science and Scent Lineups."  The article acknowledged that scientific procedures being developed primarily in Europe might be getting closer to an acceptable level of reliability to be admissible in U.S. criminal prosecutions.  The problem is, and I acknowledged this to the professor and have done so on several occasions, scent identification procedures as implemented in the U.S., including station identifications, are not up to those standards. 

A station identification is an artificial trailing situation in which a dog’s ability to follow a scent taken from a crime scene is verified in an environment that has nothing to do with that crime. Unlike the more sophisticated European scent lineups, the dog is not given a choice between scents of numerous individuals of similar economic and ethnic backgrounds, may in fact be following a trail of someone whose scent is closer to that obtained from the crime scene than other scents in the area of the station, may only have one choice of individual to choose from at the end of the trail, and may be allowed to interact with the suspect from a distance away, thereby possibly eliciting a response that can be taken as, or mistaken for, an alert. The procedure makes no more sense than would replacing scent lineups with a procedure by which a human witness would be asked to walk through a police station and identify an individual as the perpetrator from among all the people in the station at the time, with the possibility that the suspect is behind bars and in an orange suit.

Station identifications also encourage cueing, particularly when the target individual is easily distinguishable from others in the vicinity, such as by being handcuffed or in prison attire or both as in this case. It is not double-blind to third parties since the suspect is under arrest and individuals holding him may have to be advised that the station identification being conducted, and even if they are not told about it in advance, they are likely to understand what is happening and that it is intended as an identification.  Scent identifications are also seriously flawed because they do not involve antiseptic protocols as are required in more rigorous European procedures.  The station identification procedure has never been subjected to rigorous scientific analysis precisely because it is itself not capable of rigor, and no doubt because it is largely confined to a few trainers in one state. 

Nevertheless, the California Supreme Court saw no problem with letting such evidence in and has just given life to a bogus procedure that, fortunately, at this point in judicial history, seems largely to confined to several police departments in Southern California. Hopefully the Court's poor reasoning will not infect other judges.

This blog was written by John Ensminger and L.E. Papet.