Unrefrigerated, the clippings slowly rotted for more than a decade, until, in the wake of Nawi's arrest, prosecutor John Farrell had them tested for DNA.
When the crime lab got the evidence, in 1998, DNA analyst Alan Keel scraped all 10 nails with a single cotton swab, combined the scrapings into one tiny pile, and dropped them into a genetic-<\h>typing device. According to standard forensic procedure, each nail should've been swabbed and tested separately.
Now, Burt contends, the sample has deteriorated because of a lack of refrigeration and has been contaminated with the DNA of more than one person. "[Keel] says there are three, possibly four different individuals underneath her fingernails," the lawyer says. "He's trying to grab my client out of that mixture. There's no scientific way to do that."
Norris disagrees: "There are ways to deal with [DNA] mixtures; it's not a common problem luckily, but it's something that comes up — for example, in rape cases where there are multiple assailants. There are ways to deal with it."
I run down the scenario for Dr. Simon Ford, a Ph.D. biochemist and DNA expert who heads up San Francisco–<\d>based Lexigen Science and Law Consultants. "That's not good,” Ford tells me. “You should deal with each hand separately, at least, and probably each nail separately. I don't think combining all the nails together is a good idea."
Blinding them with science
The dispassionate examination of crime scene evidence — narcotics, fingerprints, hair and fibers, genetic material, firearms, and everything else — is a cornerstone of the American justice system. The work, which can mean the difference between life and death for a suspect, is carried out by more than 500 labs nationwide, most of them run by law enforcement agencies.
In the public imagination — as shaped by endless cops-and-<\h>lawyers TV shows — forensic science is a perfectly impartial arbiter of justice. Eyewitnesses get confused. Police may be corrupt. Lawyers can corkscrew facts. Juries, not always composed of the brightest lights, can be swayed by mob dynamics. But science doesn't lie. If the analyst says the bullet came from the suspect's gun, then it must have.
It's a comforting thought.
There's just one problem: All forensic science is performed by humans, and all people make blunders. They mislabel samples. They use malfunctioning equipment. They inadvertently drop a flake of skin in a vial of blood, thus adding their own DNA to the sample.
Subjectivity, too, plays a starring role in forensic science, much of which depends on human-<\h>made comparisons. In one case heard last year by San Francisco Superior Court Judge Robert Dondero, two DNA experts couldn't agree on the meaning of a genetic sample.
In addition to honest mistakes born of incompetence and overwork, there are continuously uncovered examples of fraud: the lab analyst, believing that the verdict justifies the means, willing to lie on the stand or fake test results.
While the scientific question of DNA accuracy has been hashed out extensively in court rooms and the media, the issue of police crime lab accuracy has gone ignored, both by press and government regulators.
Each year California cops make 1.5 million arrests. Each of the state's 19 local crime labs — run by sheriffs, prosecutors, and cops — performs thousands of analyses annually. Each of those tests, if faulty, could put an innocent person behind bars, or set a guilty soul free.
And in the wild world of forensics there are precious few safeguards against human bias and error: Crime labs are almost entirely unregulated. There are virtually no federal laws governing their operation; no law that says, "Bullet comparisons must be done using the best, most accurate techniques"; no law that says, "DNA examiners must meet these basic educational criteria"; no requirement that crime labs be audited and inspected.