Published September 30, 2002

                   Our minds don't work like video recorders, and yet the
                   moment we put an eyewitness to a crime on the stand
                   suddenly we treat his memory like truth from the
                   mountaintop.

                   Thanks to a generation's worth of research, we know an
                   awful lot about the fickleness of memory. We know
                   eyewitness accounts of crimes are fragmented and
                   suggestible. We know they're apt to go from shaky to
                   confident between the time a cop confirms a witness'
                   lineup choice ("You picked the right guy") and the start of
                   trial.

                   The live lineup is a staple of modern-day police work in
                   many station houses, Chicago in particular. Anyone who
                   has watched police dramas on film or television knows
                   how it works. And yet erroneous eyewitness testimony is
                   the single greatest contributor to wrongful convictions in
                   the United States.

                   Since the U.S. Supreme Court restored capital
                   punishment, 86 Death Row inmates across the nation
                   have been exonerated based on claims of innocence.
                   The convictions in more than half of those cases
                   depended at least in part on eyewitnesses, according to
                   a 2001 study by the Center on Wrongful Convictions at
                   Northwestern University School of Law. In 33 of the cases, eyewitness testimony was
                   the only evidence used against the accused.

                   Take the experience of Chicagoan James Newsome. His was not a capital case, but
                   the lessons apply to all felony prosecutions. Newsome was driving with a friend on
                   Halloween in 1979 when police stopped them, guns drawn, as possible suspects in
                   the robbery of a prostitute.

                   One of the officers thought Newsome resembled a composite sketch of a suspect in a
                   murder that occurred a day earlier. Next thing he knew, Newsome was being marched
                   into a live lineup at the police station, where three eyewitnesses fingered him as the
                   man who gunned down South Side convenience store owner Mickey Cohen.

                   Newsome, who had never before been arrested, spent 15 years in a rat-infested
                   prison serving a life sentence. That's how long it took for fingerprint technology to
                   develop that would prove the prints left at the scene by the killer weren't Newsome's.
                   For five years police knew, but didn't divulge, that the prints belonged to Dennis
                   Emerson, a career criminal.

                   Newsome and Emerson didn't even look alike.

                   Newsome was nearly three inches taller.

                   Newsome had a mole on his nose. Emerson didn't.

                   Newsome had short-cropped hair. Emerson had more of an Afro.

                   Decide for yourself from the photos included in this editorial.

                   Last year, a federal jury awarded Newsome $15 million for being framed by Chicago
                   police for the murder.

                   False identifications don't simply represent the failure of a single witness. It's not
                   uncommon for multiple witnesses to make the same bad ID. In Newsome's case, at
                   least one of the witnesses was shown a photograph of Newsome before he viewed
                   the live lineup, and one was instructed by police to "take another look at No. 3" when
                   he twice picked someone else in the lineup.

                   Of course, most police departments don't follow such egregiously bad practices. But
                   there are myriad ways officers conducting photo spreads or live lineups can send
                   subtle clues to steer eyewitnesses to identify the person who police believe is the real
                   suspect.

                   Gary Wells, a professor of psychology at Iowa State University, has spent more than
                   20 years studying police "sixpack" lineups and photo spreads, as well as the memory
                   of eyewitnesses to a crime. Through repeated experiments with staged crimes in front
                   of college students, he has found that the practice of lining a bunch of suspects up at
                   once has a higher error rate than if suspects are shown to witnesses individually, in
                   sequence. The same is true with photo spreads.

                   Why? Because when all suspects are viewed at once, the eyewitness is tempted to
                   make a relative judgment (Who most resembles the suspect among those present?)
                   rather than an absolute one (Is this the person who committed the crime?).

                   When the real perpetrator is not in the sequential lineup, witnesses tend not to pick
                   anyone. In group lineups, witnesses are more likely to pick somebody in the interest of
                   being helpful.

                   "I would see the need for expert testimony [on problems with eyewitness identification]
                   declining if everybody accepted sequential lineups," Wells said. "So it might turn out to
                   be a net money-saving thing."

                   To make these techniques work, however, sequential lineups have to be conducted by
                   an individual who has no idea who the actual suspect is. This precaution ensures that
                   officers don't give even inadvertent signals--through eye contact, lifted eyebrows or any
                   other kind of unintentional body language--that would steer an eyewitness.

                   Years of research shows that even the most well-intentioned, conscientious
                   individuals who conduct lineups can still send unintentional body cues to
                   eyewitnesses when the administrator knows the identity of the actual suspect.

                   Sequential lineups are now urged by the U.S. Department of Justice in its Eyewitness
                   Evidence Guidelines. Still, the practice draws plenty of resistance from police and
                   prosecutors who worry that suburban or rural police departments with small staffs will
                   have trouble finding someone who is unfamiliar with a case to conduct a lineup.

                   There are two answers to this.

                   One is that most neighboring jurisdictions are no more than 20 minutes away; it
                   wouldn't be that hard to find someone who didn't know about suspects in a case. The
                   second is that there are automated ways of administering such lineups. A Canadian
                   researcher has created a software program that administers lineups via laptops,
                   asking the witness at the end for a statement about his or her level of certainty. One
                   New Jersey assistant prosecutor designed a wooden board with six openings, each
                   of which can slide open, one at a time. Drug forfeiture funds were used to pay a
                   carpenter to make one for each of Union County's 23 police agencies.

                   Another complaint is that this technique is untested, based on university research
                   rather than real police practice.

                   "It all struck me as theoretical," said Tom Needham, former chief of staff to the
                   Chicago Police Department, who served on the Governor's Commission on Capital
                   Punishment, which included sequential lineups among its 85 recommendations.
                   "Why don't we wait and see how it goes out there in New Jersey."

                   New Jersey started doing sequential lineups a year ago, by order of the state attorney
                   general, with little fanfare and, so far, little complaint.

                   "I'm pleased to tell you we're not experiencing any difficulties," said Richard Rodbart,
                   Union County's deputy first assistant prosecutor. Rodbart said it's too early to tell
                   whether the new procedure has increased the credibility of eyewitness testimony in
                   court.

                   Michael Mastronardy, chief of the Dover Township, N.J., police department, said the
                   new procedures haven't cramped the style of his 150 officers. His only concern is
                   about the state providing sufficient funding to train them. "In my 10 years as chief and
                   28 years on the force, I've experienced situations with identifications where people are
                   sure, though all the evidence says, no, that's not the person," Mastronardy said. "I'd
                   rather have fewer overall identifications than a false identification."

                   That is the expectation. Sequential lineups don't necessarily heighten the probability of
                   positive identifications, but they reduce the chance of false ones.

                   Of course there will be instances where a formal police lineup is not possible. If a
                   lineup has to be conducted, say, outside the station house. But there are ways state
                   legislation could accommodate those exceptions.

                   The governor's commission has recommended one more protection, that lineup
                   procedures be videotaped, including a witness' confidence statement. At the very
                   least, they should be audiotaped.

                   Statements made by a witness during the course of a lineup can be essential
                   information to demonstrate later the confidence level of the witness at the time of
                   identification. Too often, witnesses who were somewhat uncertain about their lineup
                   selection suddenly become confident once trial rolls around because their decision
                   has been confirmed by investigators.

                   Given the cost--virtually nothing--these changes should be easy for Illinois lawmakers
                   to adopt this fall during the veto session, despite the inevitable thunder and lightning
                   they will hear from some law enforcement groups in opposition.

                   "Police are just resistant to change," said Sgt. Paul Carroll, a retired former Chicago
                   police sergeant who estimates he administered more than 1,000 photo and live
                   lineups, and who now travels the world as a consultant and trainer on sequential
                   lineups. "But cops know as well as anyone, it does no good to clear a crime and have
                   the wrong guy in jail."

                   Copyright © 2002, Chicago Tribune