There may be criminals behind bars who shouldn't have been convicted. In recent years, we've seen DNA testing release many of these convicts, often after decades of incarceration. But fingerprints can't be wrong?or can they?
Duncan Graham-Rowe writes in New Scientist that the reliability of fingerprinting is currently being tested in court. While fingerprinting is technically reliable, it suffers from human error: nobody knows how often fingerprint examiners make a mistake. Despite more than a century of use, fingerprinting has never been scientifically validated.
Brandon Mayfield, a Portland, Oregon lawyer, who happens to be a Muslim, was incorrectly identified from crime scene prints taken at one of the Madrid terrorist bombings in March 2004. Three FBI examiners plus an external expert agreed that the prints left behind were Mayfield's, even though he hadn't even been in Madrid on the day of the crime. Spanish authorities eventually matched the prints to an Algerian terrorist.
Stephan Cowans, who served six years in Massachusetts on fingerprint evidence for shooting a policeman, was released after DNA left at the crime scene wasn't a match.
The 1993 Daubert ruling by the Supreme Court outlined five rules for admitting expert testimony. One is that all forensic techniques must have a known rate of error, but the exact rate of error has never been established for fingerprinting, although it has for DNA (even if the DNA error rate is infinitesimally small).
The reliability of fingerprinting is now on trial in Massachusetts. Terry Patterson, who was convicted of murdering an off-duty policeman in 1993, is appealing is sentence on the basis of the validity fingerprinting. His prints found on a door of the victim's car. Sixteen leading fingerprint experts will testify that the prints may not really be his.
Critics say that one of the problems is that fingerprints found at a crime scene are not easy to compare to a computer database. Many are partial prints, which are always compared first to the prints of known criminals.
The FBI's Stephen Meagher compared 50,000 fingerprints with the large number of prints in the FBI computer database and concluded that the possibility of a fingerprint being mismatched was 1 in 1,097. But other fingerprint experts think the error rate is higher?between .8% and 4.4% per year. Even the lower rate would mean that almost 2,000 fingerprints are misidentified in the US each year.
Right now, when the courts are debating about whether the words "under God" should remain in the Pledge of Allegiance, William Henry reminds us that our founding fathers were NOT Christians.
Art credit: http://www.freeimages.co.uk
Where else are you going to read this kind of edge news? Nowhere else, so make sure we?ll be here for you tomorrow: subscribe today. And tell your friends all about us.
NOTE: This news story, previously published on our old site, will have any links removed.