NEW YORK -- She definitely saw the man stride down Madison
Avenue, abruptly stop, pivot and raise his arm, holding a brick
aloft like a football. Then he threw it with "immense force," she
recalled, hitting a woman on the side of the head. The assailant had
distinct features, including wide-set eyes and "nappy hair."
Why not believe the identification from Laura Weiner? She is an
articulate lawyer who was on the street, 15 feet away by her own
estimate, when someone smashed a six-pound paving stone into a
pedestrian's head, nearly killing her.
A jury did believe her, and her testimony led, in large part, to
the conviction of Paris Drake, a panhandler. Drake' s lawyer says he
is an innocent victim of mistaken identity.
The attack, a random, daylight assault against Nicole Barrett,
28, had all the ingredients of a legal quagmire: an intense police
manhunt, heavy news media coverage, conflicting witness statements,
a changing police sketch, multiple suspects, jailhouse informants
and a victim with no memory of the attack when she awoke days later.
Twice in two days, the divided and fractious jury announced a
deadlock, obviously uncertain that Drake was the assailant. In part,
Weiner's certainty finally convinced them, jurors said. But there
will never be any scientific way to prove he was there, because, as
happens in many other crimes, the attacker left behind no DNA.
Eyewitness identification has never been as controversial or
vulnerable as it is now, in part because it has never been so
clearly rebuttable. A growing school of research -- fueled by newly
available DNA evidence -- shows that eyewitnesses as certain and
credible as Weiner are very often wrong, and that there is little
correlation between their certainty and their accuracy.
In fact, 65 of the 77 wrongful convictions overturned by DNA
evidence nationwide in the last decade resulted from eyewitness
errors, according to statistics compiled by the Innocence Project at
the Benjamin N. Cardozo School of Law at Yeshiva University in
Manhattan. And in cases like Drake's, where heavy news coverage can
expose witnesses to all sorts of new impressions, from sketches and
photographs to theories -- the chances for mistakes climb.
Recent studies on how memory works show that a clear view of a
crime does not work like a cattle brand, searing an image indelibly
into the mind, but that impressions formed after the event, through
suggestive police procedures or news media exposure, for example,
can lead a witness to feel certain.
And there is almost nothing more convincing to a jury than a
confident eyewitness, even, according to one study, when jurors know
that the eyewitness is legally blind.
The nation's legal experts have increasingly focused on a
troubling collision: Eyewitness testimony can be quite unreliable
and very persuasive at the same time.
"It is the conjunction of those two properties that should
concern us," said Gary L. Wells, a psychology professor at Iowa
State University and an expert on eyewitness identification.
Before the early 1900s, with the advent of psychology as a
science, the accuracy of eyewitness identifications was seldom