E-Mail This Article to a Friend Blind spots A spate of wrongful convictions has convinced Suffolk County DA Dan Conley and Boston Police commissioner Kathleen O’Toole to reform how the police use eyewitness evidence. While they’re at it, they should reopen these three cases. BY DAVID S. BERNSTEIN

IF THERE’S ONE THING we all should have learned from the string of wrongful convictions recently come to light, it’s that you can’t always believe someone else’s eyes. Questionable or flat-out-wrong eyewitness-identification testimony led to Anthony Powell’s 1992 conviction for rape, and the murder convictions of Shawn Drumgold in 1989, Donnell Johnson in 1996, and Marlon Passley in 1995. All have been released — with apologies from Suffolk County district attorney Dan Conley in the first two cases and former Suffolk County DA Ralph Martin in the others — in the last five years.

Not every wrongful conviction is the fault of eyewitnesses, of course. A jury found Stephan Cowans, freed this January, guilty of a 1997 police-officer shooting largely on the basis of an erroneous fingerprint match. But witnesses remain the most common problem, in Boston and elsewhere. The New York–based Innocence Project, a nonprofit legal clinic at Benjamin N. Cardozo School of Law that takes on prisoner appeals in which DNA or other scientific tests can prove innocence, says that more than 80 percent of prisoners exonerated by DNA evidence since 1992 were originally convicted on mistaken eyewitness testimony. A University of Michigan study released Monday found that almost 90 percent of wrongful rape convictions and half of wrongful homicide convictions since 1989 included at least one eyewitness misidentification. The results of the study suggest that thousands of those incarcerated today may, in fact, be innocent.

All this has prompted calls for change in Suffolk County. After Powell, who’d been identified by the victim from a photo array, was exonerated by DNA evidence and freed in March, Conley and new Boston Police Department (BPD) commissioner Kathleen O’Toole announced the establishment of an eyewitness-identification working group. The group includes three defense attorneys, a prosecutor, a BPD superintendent, and an academic researcher specializing in witness IDs.

The group has taken as its guide previous reports by the US Department of Justice in 2001 and by the states of New Jersey, North Carolina, and Illinois, according to task-force member James Doyle, an attorney with the Boston law firm Carney & Bassil and author of several studies on eyewitness identification. Doyle, who also served on the working group that issued the Justice Department’s report, says that the BPD uses outmoded techniques for eyewitness identification. For example, photo arrays or line-ups, which display several faces at once, encourage the witness to select the one "most similar" to their memory; sequential arrays, showing one photo at a time, more reliably result in a correct identification. "Double-blind" identification, in which the presenter does not know who the suspect is, eliminates the subtle — and sometimes not so subtle — ways that a police officer may influence the witness to choose the "correct" person.

The working group’s charge is to make recommendations like these to improve the reliability of eyewitness identification. Which means that the group will not concern itself with old convictions. "There are two kinds of innocence issues," Doyle says. "One is reopening specific cases. The other is creating a template for doing things in a systematic way in the future."

But clearly, Suffolk County needs a second group to study old eyewitness-based convictions. In too many instances, questionable eyewitness testimony has resulted in guilty verdicts for people who may well be innocent. Such convictions are not hard to find; just ask criminal-defense attorneys if they have any cases that still gnaw at them. In conversations with more than a dozen attorneys who defend homicide suspects in Boston, the Phoenix found three cases that continue to haunt not only the lawyers, but also the families and friends of the young men who are in prison because of eyewitness identifications. All three involve homicides committed between the fall of 1999 and summer of 2000. No significant physical evidence tied the defendants to the crime in any of these cases. All three young men named below are in prison today; their sentences range from eight years to life.

Keyon Sprinkle

Convicted of first-degree murder in 2002; sentenced to life.

On November 16, 1999, Charles Taylor was shot six times, in broad daylight, on a Roxbury street corner. Keyon Sprinkle, who was then 17 years old, told police that he was in his house three blocks away when he heard the gunshots. But eyewitnesses told a different story during his trial, and on March 27, 2002, Sprinkle was convicted of first-degree murder and sentenced to life in prison.

"This case eats me up, because I’m as certain as I can be that Keyon is innocent," says Peter Parker, the North End attorney who defended him. "I think about this case all the time."

Sprinkle, just 5’3" and 120 pounds, was an unemployed high-school dropout living with his grandmother at the time of his arrest. His only previous legal trouble was a marijuana-possession charge that was never prosecuted. He has a daughter with his girlfriend Jayda, and wanted to put together a life with them, says his cousin Loreen Jones, a Boston Public School teacher living in Randolph. "He had some plans," Jones says. "After the baby came, he knew he had to buckle up." But he never got that chance.