Cosenza v. City of Worcester, 2024 WL 4589489 (1st Cir. 2024)
In the summer of 2000, Melissa Horgan awoke at 0400 to find a man standing near her bed wearing a T-shirt, underwear and a cloth wrapped around his head. He tried to climb on top of her, attacking her with a wooden object, but Horgan’s kicking and screaming caused the man to flee. Horgan immediately called 911. She described her attacker as a white male, with no hair, and she described what he was wearing. She repeatedly said she had never seen her attacker before. Several officers from the Worcester (Massachusetts) Police Department arrived at her apartment. Despite their questions, the woman was unable to provide any further description. She again repeated she had never seen the man before.
Later that day, officers canvased the woman’s apartment complex for witnesses. No one could identify the assailant, but the officers learned Natale Cosenza lived in a different building of the complex. The officers were familiar with Cosenza because of his history of drug addiction. Cosenza was on probation for drug crimes, and he was not popular with the local police. The officers agreed Cosenza was a likely suspect and, according to the court, they “reverse-engineered the investigation to implicate him.”
A day after the assault, two officers met with the victim and presented her with a photo array that included Cosenza and two men who very closely resembled him. The officers told the victim Cosenza’s name and that he lived in the same complex, implying the officers believed the suspect to be featured in the photo array. They stressed that she had to identify someone in the photo lineup for the case to proceed. The victim identified Cosenza, who (unlike the man she had described the night before) had a full head of hair. After seeing Cosenza’s photo, she provided a more detailed description which more closely matched Cosenza. Despite emphatically repeating to the officers at the scene that she had never seen her attacker before, she then claimed her attacker had a “familiar face.” She also claimed her assailant was a man who had previously knocked on her door asking for money.
Cosenza was arrested and ultimately convicted by a jury of assault and battery with a dangerous weapon and armed burglary. The prosecution’s case heavily relied on the victim’s identification — both in the photo array and in court. Though the officers had found fingerprints on the window and located the wooden dowel that was likely the assault weapon, they destroyed those items before Cosenza’s attorneys could examine or test them. The officers also found shorts with semen stains. When the DNA on the shorts proved not to be Cosenza’s, the officers “manipulated details to make it appear the shorts did not belong to the true attacker.”
Fourteen years later, a state court granted Cosenza a new trial, relying on developments in the law related to eyewitness identification procedures. The court suppressed the photo array evidence, and the prosecutor dropped the charges.
Cosenza sued the officers and the city, alleging the city had a policy of failing to adequately train officers on photo identification techniques. The trial court granted summary judgment for the city, finding no evidence of deliberate indifference to Cosenza’s constitutional rights. Cosenza appealed.
Get the Xiphos law enforcement legal update delivered to your inbox: SUBSCRIBE NOW!
Eyewitness identification, whether obtained through a photographic lineup or the far less-common live lineup, has traditionally carried a strong presumption of reliability. Advances in science and scrutiny by courts have sharply questioned the validity of eyewitness identification procedures that do not follow best practices based in scientific evidence. We tend to rate our memories as very reliable. Some research suggests most adults believe memory is about as good as a video camera. The American Psychological Association reports jurors in criminal trials tend to “over believe” eyewitness accounts, giving too much credit and not entertaining doubts an eyewitness could be well-meaning and sincere, but still mistaken.
Challenging eyewitness reliability isn’t new. As early as 1907, Harvard University psychologist Hugo Munsterberg published On the Witness Stand, raising questions about the reliability of eyewitness identification. Two-and-a-half decades later, Yale Law professor Edwin Borchard wrote Convicting the Innocent (1932). Examining 65 proven wrongful convictions, he reported eyewitness misidentification was the leading contributing factor.
In the process of encoding, storing and retrieving memories, details are unconsciously added, altered or forgotten, and are vulnerable to even subtle suggestions, such as leading questions or alternate descriptions. Memory accuracy is partially a function of time, such that the longer the span between the original observation and the retrieval of the memory, the less accurate the memory. In particular, the presence of a weapon has been found to reduce the reliability of later eyewitness identifications. Because people can only perceive a limited number of stimuli at any given time, they engage in perceptual selectivity, concentrating their attention on “the most necessary and useful details” — often leading to inaccurate perceptions.
In a traditional photo lineup, the process is administered by an investigator involved in the case who knows the identity of the suspect whose photo is mixed in with five other photos (known as “fillers”). Researchers and courts have focused on whether identification procedures should be administered “blindly” and whether photographs should be presented simultaneously, as is traditionally done, or sequentially. In blind administration, the person conducting the lineup is unaware of the suspect’s identity or placement in the photo array. If the person administering the identification process knows which photo shows the presumed suspect, there is a significant risk of the administrator cuing the witness. The cue may only be a facial tic or intake of breath. Blind administration eliminates even unconscious cueing and reduces the likelihood of erroneous identification.
Years after Cosenza’s conviction, in Commonwealth v. Thomas (68 N.E.3d 1161 (Mass. 2017)), Massachusetts promulgated statewide jury instructions that inform jurors of potential problems associated with eyewitness identifications and how administration of the process can influence the identification. Police accreditation standards in Massachusetts also mandate both blind administration and sequential presentation of photo lineups.
Whether sequential presentation is superior to simultaneous presentation remains an open question. Some research suggests the simultaneous presentation method allows witnesses to make choices by comparing the lineup members to each other, resulting in a choice that most closely resembles the perpetrator, regardless of whether the perpetrator is present in the lineup. Police officers who have ever shown a simultaneous photo array using loose photos may have actually seen this occur when a witness slides a filler, or non-suspect, photo off to one side to eliminate it from consideration or to narrow the field. When the actual perpetrator is absent from the lineup, witnesses may get to their selection by picking which photo looks most like the offender, heightening the chance of a misidentification. As a practical matter, there are other reasons the sequential presentation method can protect the purity of an identification procedure. Slight differences in the backgrounds of filler photographs or in the appearance of the fillers themselves become more apparent when the photographs are laid side-by-side.
A recommended best practice policy begins with video-recording the eyewitness identification procedure and administering important instructions to the witness prior to beginning. Blind administration of the identification procedure is a best practice and courts are tending toward requiring double-blind administration. In double-blind identification lineups, the person presenting the lineup is not involved in the investigation of the case and does not know the identity of the suspect. In addition to the cautionary instructions to the eyewitness prescribed by the Massachusetts courts, the administrator should also instruct the witness that the suspect may not appear exactly as they did on the date of the incident. The process should also include a signature line where the witness acknowledges they understand the identification procedures and instructions. The lineup documentation should also include a confidence statement, in which the witness (in their own words) describes how certain they are of the identification or non-identification. This statement should be taken at the time of the identification procedure.
The filler photos (of non-suspects) should resemble the eyewitness’s description of the perpetrator. (For example, the suspect should not be the only member of his race in the lineup, or the only one with facial hair.) Eyewitnesses should also not view multiple lineups with the same suspect. Fillers should not stand out due to different height, weight, coloring, clothes, behavior, etc. In photo arrays, there are numerous ways one picture can be subtly different: lighting, color tone, brightness, sharpness, viewing angle, background, location of face in the frame, etc. If appropriate, the administrator should tell the witness the difference in format is irrelevant.
If your agency subscribes to Lexipol’s policies, your Eyewitness Identification Policy covers all these best practices and any applicable state-specific laws. Now is a great time to review the policy and incorporate it in roll call or other training.
The appellate court found there was no evidence the city had a policy or practice of failing to train officers on photo lineups. Though the city tacitly acknowledged the officers’ actions fell far short of current investigative practices, the court credited the city’s evidence that officers had received at least some training at a police academy and some on-the-job training. Because the law at the time of Cosenza’s trial did not clearly establish the procedures Cosenza argued were required, the city was entitled to summary judgment. The court also ruled there was no evidence the city had a policy of fabricating or suppressing evidence. Claims against the individual officers were allowed to proceed to trial.
