April 7, 2001
By ANDY NEWMAN 

In the long history of forensic science, prosecutors have found few weapons more powerful than the fingerprint. The whorls, arches, ridges and loops left on a surface by the skin's oil have long been considered virtually unassailable evidence tying a person to a crime. 

But now, the reliability of crime-scene fingerprint identification is being challenged. In courts around the nation, defense lawyers are using evidence of fingerprinting's fallibility to try to get it declared inadmissible under standards set by the Supreme Court to keep unproven "junk science" out of courtrooms. 

The accuracy of making identifications from dusted or latent prints, which are often smudged, distorted or fragmentary, has never been scientifically tested. And while fingerprint examiners are trained to testify only to "absolute certainty" about their work, defense lawyers point out that examiners do make mistakes, that training standards vary widely and that most examiners have either failed or never taken the main certification test. 

Trial judges have rejected the dozen challenges filed since 1999, holding that fingerprinting, which has been accepted since 1911, has proved its reliability in the courtroom. And few prosecutors are even aware of the challenges. 

But the government is taking the issue seriously enough to solicit the first studies to validate crime-scene print identification and set standard procedures for examinations. 

Defense lawyers who have brought challenges said they had in some cases secured favorable plea deals or prompted prosecutors to withdraw fingerprint evidence. 

Edward J. Imwinkelried, a leading expert on forensic science who has worked with prosecutors and defense lawyers, said there was a "very good possibility" that the challenges would lead judges to instruct juries that a fingerprint analyst was not a scientist offering exact conclusions but an expert giving an opinion. 

That, said Mr. Imwinkelried, a law professor at the University of California at Davis, "could conceivably be an important weapon in the hands of defense counsel, because you've got a widespread public perception that fingerprint testimony is infallible." 

The door to scrutiny of fingerprinting was opened by two United States Supreme Court decisions that changed the rules governing expert testimony. In two product liability suits — Daubert v. Merrell Dow Pharmaceuticals in 1993 and Kumho Tire Company v. Carmichael in 1999 — the Supreme Court declared that federal judges must determine the reliability of expert testimony before admitting it. About two dozen state court systems have followed suit. Judges have already limited the use of handwriting analysis after reliability challenges. 

In 1999, Robert Epstein, a federal public defender in Philadelphia, made the first effort to have fingerprint identification declared inadmissible under the Daubert standards. His pretrial motion, in a case involving a man accused of driving 
the getaway car in a robbery, was denied by Judge J. Curtis Joyner of Federal District Court. But Mr. Epstein's tactic has nevertheless been widely imitated. 

Last year, in a burglary case in Montgomery County, Md., where fingerprints were the only evidence, prosecutors offered a 6-year plea agreement on theft charges, rather than the 10 years the defendant was facing for burglary, after the public defender challenged the admissibility of fingerprinting. 

"We decided that rather than go through the trouble of doing the motion we would agree to a plea," the prosecutor, Michael Banks, said. 

Law enforcement officials recognized soon after the Daubert ruling that fingerprinting could be vulnerable to challenges. In early 1999, the Justice Department's research arm, the National Institute of Justice, started putting together a call for studies to come up with standardized, statistically tested procedures for comparing fingerprints that "produce correct results with acceptable error rates." 

In a sign of further concern, the Federal Bureau of Investigation asked the institute in September 1999 "to withhold releasing the fingerprint solicitation until after the Philadelphia trial had ended," according to institute memorandums obtained by Mr. Epstein. The F.B.I. declined to comment on the matter. 

The solicitation was released in March 2000, a few weeks after Mr. Epstein's client was convicted. The institute received four proposals but rejected all of them and will start the solicitation process over. 

Despite all the publicity DNA testing has received, for now, fingerprints are more useful because they are easier to collect than DNA, forensic experts say. But critics say the profession of fingerprint analysis is not as rigorous as generally believed. On a 1995 proficiency test of 156 examiners conducted with the approval of the International Association of Identification, the profession's certifying organization, one in five examiners made at least one "false positive" identification — linking a mock crime-scene print to the wrong person. Fingerprint experts point out that the error rate was lower on subsequent tests. 

The challenges have also attacked the variability in training methods for examiners, pointing out that agencies like the F.B.I. have tougher standards than smaller police departments. 

And while the International Association of Identification has a rigorous certifying test, about half the current or would-be examiners who take it fail, without apparent career consequences. 

"There's very few employers who will terminate an employee for not passing the test," said Ken Smith, the association's certification chairman. Mr. Smith added that most of the 5,000 examiners in the country have never taken the test. 

While fingerprint misidentifications are rarely discovered, they do happen. Richard Jackson was cleared of a murder conviction in Philadelphia in 1999 because three examiners had erroneously matched his prints to those found at the scene. A similar reversal occurred in 1983 in Minnesota. In that case, both the prosecution and defense fingerprint experts mismatched a print to the defendant. 

Paul Sarmousakis, the assistant United States attorney who prosecuted Mr. Epstein's client, said that the occasional human error did not invalidate fingerprinting. "Because a doctor misdiagnoses someone, does that make the science of medicine invalid?" Mr. Sarmousakis asked. 

But Simon A. Cole, a science historian and the author of a forthcoming book on fingerprinting, said print examiners undermined their legitimacy by claiming absolute certainty, which the International Association of Identification's bylaws require. 

"If they want to go in and testify, `I think it's his print and 1 percent of the time I'm wrong,' then that would be more reasonable," Mr. Cole said. 

Mr. Epstein said a test conducted by the F.B.I. in his first challenge showed the lack of rigor. After he filed the challenge but before it was heard by the court, the F.B.I. sent the defendant's official prints and the crime-scene prints to 53 law enforcement agencies. 

But 8 of the 34 laboratories that responded were unable to find a match for at least one of the two latent prints. 

The bureau sent the prints out again, with bigger photographs and red dots marking where it thought the crime-scene prints matched those of the defendant. This time, all the laboratories declared the prints a match. 

Mr. Epstein has moved for a new trial. However his motion fares, challenges to fingerprinting are likely to continue. 

"Every time the state has a fingerprint that's going to be used as evidence against one of my clients, I'm going to do the same thing over and over again," said B. Michael Mears, the chief public defender for capital cases in Georgia, who brought a fingerprint challenge last year. "And I am going to keep doing it until we win it."