By D.W. MILLER

    Until a few years ago, when the U.S. Department of Justice

invited six psychologists to help reshape police procedures

for eyewitness identification, scholars had only one way to

influence criminal justice: one defendant at a time. Many have

themselves testified to educate juries about the pitfalls of

witness memory.

    Like a lot of his colleagues, Gary L. Wells, a psychologist at

Iowa State University who testifies four or five times a year,

got into that line of research in part to save innocent

defendants from false imprisonment, and to force police to

improve methods for interviewing witnesses and identifying

suspects. "There was a time 20 years ago when I was so naive

as to think that all I had to do was document the problem and

the police would change their procedures," he says. But

eventually he decided that "the courtroom was never the place

to have that kind of impact."

    "Judges are reluctant to tell police how to do their jobs," he

says. And judges tend to hew to the established view that

juries are the arbiters of witness credibility.

That has been changing slowly. In 1993, the U.S. Supreme Court

ruled in Daubert v. Merrell Dow Pharmaceuticals, Inc. that new

federal rules of evidence permitted a broader standard for

allowing expert psychological testimony. Since then, says

Solomon Fulero, a psychologist at Sinclair Community College,

in Dayton, Ohio, several convictions have been overturned

because the trial judge had not allowed such experts to

testify.

    Still, there's a limit to the broad change that scholars can

effect by testifying. According to Mr. Wells, there just

aren't that many experts: About 50 to 75 psychologists testify

in court regularly, and only about 25 of them actually do

original research in the field.

    Furthermore, their services can be pricey. While rates vary

widely, the psychologists themselves report fees of up to

$3,500 a case, although most will take some clients pro bono.

In general, the experts try to avoid challenging the

credibility of individual witnesses or the conduct of the

police officers who worked with them. "The goal of the defense

is to cast doubt on the credibility of a particular witness.

But that's not my job," says Mr. Fulero, who was invited to

join the Justice Department's eyewitness-testimony panel

because of his courtroom experience, not his scholarly vitae.

What he can testify to, he explains, is that "eyewitnesses are

not as accurate, over all, as the jurors believe them to be."

Unfortunately for defendants, that means the research doesn't

always help their cause.

    "The deep problem," says James M. Doyle, a Boston defense

lawyer who served on the panel, "is that the research is all

statistical and probabilistic, but the trial process is

clinical and diagnostic." In other words, a jury expects the

experts to say whether a witness is right or wrong, when all

an expert can really do is explain how to assess the odds.

Mr. Wells echoes many of his colleagues when he says that he's

not really in it for the money. He was among the half-dozen

scholars who helped to fashion the new Justice Department

guidelines for handling eyewitness testimony. If they are

widely adopted, he says, "we have no business in the courtroom

on this issue. My purpose is to make expert testimony

unnecessary."

    He may get his wish. According to participants, prosecutors on

the Justice Department panel were concerned that quick-witted

defense lawyers would use the new guidelines to impeach

eyewitness testimony.

    Mr. Doyle, who has co-written a lawyer's guide to the

research, Eyewitness Testimony, calls that a reasonable fear.

In the past, his colleagues have had difficulty incorporating

the science into their cross-examination techniques, because

they haven't taken the trouble to understand the research

methods, he says. Now they won't have to.

    On the other hand, he doubts that's a bad thing. "One thing

police and defense lawyers share is that we don't really want

to deal with innocent people. It's not necessarily easier or

better for me to represent innocent people. I would just as

soon the police did their jobs."