Scrutinizing Eyewitnesses : Hired Defense Experts Cast Doubt on the Reliability of Crucial Identifications

The robbery victim shifted nervously in the witness chair. Her eyes darted around the Van Nuys courtroom, but repeatedly came to rest on the defendant in an almost sympathetic way.

The defense attorney asked the woman to describe her knife-wielding assailant, and admonished her not to look at his client.

Pointing to the jury and defense attorney, she said, “All of you are strangers to me, but not him.”

Was this a positive identification of the robber? Perhaps not, say psychologists, who maintain that despite the victim’s unequivocal testimony, such identification might be a case of mistaken identity.

To help the jury evaluate how accurately the woman recalled the Feb. 8, 1986, robbery, Van Nuys Superior Court Judge Judith M. Ashmann last week permitted a psychologist to testify about factors that affect memory and perception in stressful situations.

Increasing Incidence

Such expert testimony has become increasingly popular since the California Supreme Court ruled in November, 1984, that the testimony generally should be allowed if the prosecution lacks substantial evidence to corroborate eyewitness identification.

In issuing the ruling, the Supreme Court noted that there are “scores of studies on the pitfalls” of eyewitness testimony. But such expert testimony also has its detractors.

The trial in Van Nuys Superior Court, which is expected to conclude today, is based exclusively on eyewitness identification. The case illustrates many of the elements that researchers say influence a witness’s ability to identify the correct suspect.

Those factors include the length of time the witness observed the assailant, the degree of stress involved and the witness’s attentiveness to details. In addition, eyewitness identification is generally less reliable if a weapon is used and if the suspect and witness are of different races, psychologists say.

Testimony Heard

According to testimony, the robber visited a law office in Sepulveda on two consecutive Saturdays, requesting to see an attorney and identifying himself as Gary Simmons. He spent 10 minutes in the office on the first trip and 30 minutes on the second one, talking to office personnel, witnesses testified.

At the end of the second visit, he pulled out a knife, robbed a woman of $175 and a man of $600, then stabbed the man in the back and fled.

On the witness stand last week, the woman victim--a Hispanic--and two other eyewitnesses--a Caucasian and a Hispanic--pointed to the defendant, who is black, and positively identified him as the attacker.

Two of the three had positively identified the same man last spring in police photos and in a subsequent lineup of suspects. The third witness tentatively identified the man in those two procedures, but said he could not be certain.

The stabbing victim, who suffered minor injuries, was unable to make any identification.

Throughout the investigation, however, none of the witnesses mentioned that the robber had an ugly 12-inch scar on the inside of his left arm, although, they said, he was dressed in a shirt with short sleeves.

Prominent Features

The scar and additional blemishes on the defendant’s hands are “very prominent, salient anomalies,” and it is unlikely that witnesses would have overlooked or forgotten them, USC psychology professor Scott Fraser testified Thursday.

The defendant, Willie James Taylor, 20, of South-Central Los Angeles, will testify today that he was never in the law office and has never used the name Gary Simmons, said his attorney, Patrick Atkinson.

Psychologists and defense attorneys maintain that jurors place too much credence in the recollections of an eyewitness.

“If a case is based solely on eyewitness identification, with no corroborating evidence, I don’t think it’s worth a damn,” Atkinson said.

But prosecutors vigorously dispute that conclusion. They say that eyewitness identification is very reliable and sends few innocent people to jail. Prosecutors criticize psychologists for basing their opinions on experiments with college students, rather than on interviews with people who have actually witnessed a crime.

Imprinted in Memory

“The events are so important to a victim or witness that it imprints on the mind what the person looks like,” said Deputy Dist. Atty. Dino J. Fulgoni.

“Most of the studies are conducted in a classroom, an artificial setting, and everybody knows it’s an experiment.”

Fulgoni acknowledged that there are several celebrated examples of innocent individuals who were accused of crimes based on faulty eyewitness identifications and later were cleared when the real perpetrator confessed.

For example, in 1979 a Catholic priest was tried in Wilmington, Del., after seven eyewitnesses identified him as the “gentleman bandit” responsible for a series of armed robberies. Halfway through the trial, another man confessed to the crime, and the charges against the priest were dismissed.

But, Fulgoni said, such mistakes are “very rare.”

A handful of psychologists in the Los Angeles area, including Fraser, have qualified as courtroom experts on the subject of eyewitness identification.

Most active among them, attorneys say, is Robert William Shomer of Van Nuys, whose testimony was the subject of the 1984 state Supreme Court ruling in the case of People vs. Eddie Bobby McDonald, which set guidelines for allowing eyewitness experts to testify.

More Experts Testify

Some judges had permitted such testimony before the ruling, but it has become far more prevalent since, attorneys say.

The court went a step further in January, ruling that in certain cases judges should instruct jurors to view eyewitness testimony with caution.

“When a juror hears that eyewitness testimony has to be viewed with caution, my God, that’s like waving a red flag in front of them,” said Deputy Dist. Atty. Robert P. Imerman.

But psychologists counter that such caution is critical to a fair trial.

“If most people get honest with themselves, they’ll concede that it’s not very easy to identify someone,” Shomer said. “Yet most jurors consider it as reliable as a fingerprint.”

Shomer, a psychotherapist and former psychology professor at Harvard University and the Claremont Colleges, has testified in more than 160 cases since 1974. He now averages two court appearances a month, he said.

His fee is $75 to $100 an hour for preparation and time spent in court, a cost borne by the defense, except in the case of an indigent defendant, who may qualify for the services of a court-appointed expert at taxpayer expense.

Although he has never been hired by the prosecution, Shomer said, he does not view himself as a defense advocate.

“I really see myself as an educator, helping juries or judges to understand what we know about the process,” he said.

No Opinions on Case

Psychologists who testify about eyewitness identification are forbidden by law to offer courtroom opinions on whether they believe that a witness identified the right suspect. Instead, they generalize about the ability of individuals to remember details.

Under the very best of circumstances, Shomer maintains, eyewitness identification has an accuracy rate of only 70%. An optimal setting, he said, is when individuals know they will be viewing a series of faces, lighting is good, stress is low, length of viewing time is adequate and the identification is made shortly after seeing the faces.

Prosecutor Fulgoni said he believes that the accuracy rate is quite high.

“I would be surprised if correct identification is lower than 85% across the board, and I believe it is far more,” he said.

In their testimony, Shomer and other experts cite studies that conclude that a certain level of stress enhances both perception and memory, but that high stress inhibits them. In addition, they say, some witnesses innocently incorporate into their own memory the statements and descriptions made by others.

Focus on Weapon

In cases where a witness is threatened with a weapon, they add, the individual tends to focus more on the gun or knife than on the assailant. And witnesses generally are more accurate in identifying members of their own race than other races, they say.

“With people of our own race, we’re more used to focusing on details of the face, rather than concluding, ‘Oh, that’s a black person or a white person,”’ Shomer said.

Shomer said jurors also should be informed about certain frailties in the judicial system itself.

When witnesses are shown mug shots, they generally believe that the assailant must be among those presented to them, that police would not waste their time on a futile exercise, Shomer said. As a result, witnesses tend to identify the person who most closely resembles the perpetrator, Shomer said.

That face then becomes fixed in their minds as the face of the assailant. At subsequent live lineups or court appearances, witnesses will confidently identify the same person.

To guard against such influences, police routinely instruct witnesses that the suspect in their crime is not necessarily among the individuals or photographs being shown to them. But psychologists say that witnesses pay little attention to such admonitions.

As a case approaches trial, witnesses typically become more certain about their identification, even though memory has been shown to fade over time, Shomer said. At preliminary hearings and trials, there is only one person to choose--the individual sitting at the counsel table next to the defense attorney, psychologists say.

“You get very dramatic trial testimony: ‘That’s the person. I’ll never forget that face,’ ” Shomer said. “But jurors need to know how familiar the witness has become with that face.”

It is not that witnesses are trying to deceive, Shomer said. Most honestly believe that their identification is correct and that they would pass a lie detector test with flying colors, he said.

Prosecutors dismiss most of the expert testimony about the accuracy of eyewitnesses as nonsense.

“These abstract studies are so outlandish that I don’t think jurors put a lot of credence in them,” said Dist. Atty. Robert L. Cohen.

“We like it better, too, when we’ve got a fingerprint, or when the guy has the stuff he stole in his back pocket 20 minutes later.”

Fulgoni, who is the district attorney’s expert on the experts, called the studies on which Shomer and his colleagues base their testimony “completely unscientific.”

He said most of the research involves students who were asked to verbally describe an individual’s physical characteristics--height, weight, hair color, eyes, clothing--rather than to pick that person from a series of photographs or out of a lineup.

Most people will readily admit that they are poor judges of physical characteristics, Fulgoni said, but they can remember a face if they see it again.

Kathy Pezdek, psychology professor at the Clarement Graduate School and a courtroom expert on eyewitness identification, acknowledged that a study of people’s ability to accurately describe others is “not a good measure to use.”

But she emphasized that sufficient research on facial recognition now exists to support earlier conclusions that eyewitness identification has flaws that should be called to the attention of jurors.

Pezdek added, however, that in most of the 85 cases in which she has testified, she believed that the witness got a good enough look at the suspect to be able to recognize the face later.

Attorneys and psychologists say they know of only a handful of eyewitness experts in the Los Angeles area who are willing to testify in trials, although the subject has been studied on university campuses for more than 100 years.

Pezdek said many professors are reluctant to share their research in a courtroom because attorneys try to box them in to conclusions, asking “yes” and “no” questions that allow for little elaboration.

“Academic training teaches you to be very conservative, . . . tentative, to generalize about your data,” Pezdek said. “The courtroom requires a whole different set of skills.

“You can’t answer every question, ‘Well, it depends.’ ”