Everybody is reading today about the fact Brandon Mayfield, a Portland lawyer, was released after being held for two weeks as a “material witness” (i.e., without charges being filed) in the Madrid bombings. He was held after his fingerprint was allegedly found on a bag of detonators recovered after the bombings. Many of the stories also report that Spanish authorities doubted any link from the outset. There is little, if any, mention of this statement in the May 17 Newsweek: “A top U.S. counterterrorism official told NEWSWEEK that the fingerprint was an ‘absolutely incontrovertible match.'” (The article then parenthetically noted that Spanish authorities “weren’t quite as sure.”)
Maybe Newsweek just got some bum information from a bum source. Still, Reuters reports, “While the Americans found 15 points of coincidence between the print on the bag and Mayfield’s fingerprint, Spanish police found only eight.” It is more than a little disconcerting when an “absolutely incontrovertible match” becomes “oops, guess we were wrong” and our fingerprint expertise lags behind Spain’s.
While Spain may have found fewer points of coincidence, the real question which arises because of the error is this: When we cease to limit fingerprint database searches to a certain locality, such as a nation, how dramatic will the increase of false positives be? We can wax eloquently about how every snowflake is unique and how everyone is different (“I’m not!”), but when our methods for determining identity create problems such as this, it is quite apparent that improvements must be made if we are to use this identification system on a global basis.
The increased use of retinal scans in biometrics can certainly assist in identification matters when the subject needs be present, but the reliance on it alone is certainly out of the question in the forensics field.
jeremy
http://www.adamnation.net