Alejandro Dominguez, age 16, had no tattoos or pierced ears, and he reportedly could speak only Spanish. The woman, however, said he was the attacker, and largely on the strength of her testimony, he was convicted. Not until 2002 did DNA analysis prove Dominguez was innocent.
It's a dismally familiar tale: a victim making an eyewitness identification that later turns out to be horribly mistaken. This type of mistake is universally known as the most common cause of false convictions. Yet law enforcement authorities, courts and juries continue to treat it as pure gold.
But change is on the way in New Jersey, where last week, the state Supreme Court ran out of patience with a method that puts so many innocents behind bars. It mandated new rules that will help to prevent errors while giving defendants more avenues to expose them.
The justices said that "courts must carefully consider identification evidence before it is admitted to weed out unreliable identifications" and "juries must receive thorough instructions tailored to the facts of the case to be able to evaluate the identification evidence they hear."
Like other evidence, it must be subject to careful scrutiny and challenge. The burden of disproof will still fall on the accused, but it will be easier to meet. Chances are good that, as a result, some blameless individuals will be spared.
It's the least the courts can do, and it's something the U.S. Supreme Court will get to consider this fall, hearing the first major case on the issue since 1977. Since then, the fallibility of eyewitness evidence has been confirmed by a mounting pile of data.
In one experiment, a "customer" went into a convenience store to buy a soft drink with a traveler's check, which required him to provide an ID and spend a few minutes conversing with the clerk. Later, the clerks were asked to find the person in a group of photos. Forty-one percent made a wrong pick.
Errors don't happen because crime witnesses choose to lie. Most of them sincerely believe what they say. But their memories may be addled by shock, colored by a desire to punish the villain or led astray by police suggestions.
The palpable certitude of someone who was present during a crime makes for powerful evidence to a jury. But as the New Jersey court opinion noted, "accuracy and confidence may not be related to one another at all."
Jennifer Thompson can vouch for that. In 1984, as she was being raped at knifepoint, she forced herself to study and note "every single detail on the rapist's face" so she would be able to identify him.
At the police station a few days later, Thompson found her attacker in a gallery of photos. She picked him again out of a physical lineup. She took the stand in court to point him out, and he was convicted.
But 11 years later, DNA evidence pointed to someone else. "The man I had identified so emphatically on so many occasions was absolutely innocent," she wrote later.
Thompson had many minutes to get a closeup view. Often, arrests are made on the basis of brief glimpses, sometimes in dim light or at a far remove. But even in these instances, eyewitness testimony can obliterate other evidence, as well as common sense.
In 1990, a jury convicted David Dowaliby of suburban Chicago of killing his 7-year-old daughter, largely on the basis of a witness who, on the night of the murder, saw someone in a parking lot where the body was later found.
He was 75 yards away, the lighting "wasn't that great," and the witness wasn't sure if the person was male or female, or black or white. All he knew was that the "nose structure" matched Dowaliby's. A jury voted to convict. It took an appeals court to throw out the case.
But many times, mistakes go undetected and uncorrected. We all know it's dangerous to believe everything we hear. The criminal justice system ought to acknowledge that the same holds for what we see.