Sure, a few prosecutors still want to argue over bite mark evidence, in bizarre denial of the conclusive science that it’s forensic nonsense. Their reason is understandable. It enables convictions where the evidence otherwise would fail. And if they believe they’ve got the right defendant, the bad dude, why wouldn’t they want to use whatever they can get away with to convict? That it might not be legit, or the defendant might not be guilty, isn’t their concern.
But historically reliable evidence is entirely different in the minds of jurors. Television shows make it all appear so serious, so legitimate, so very conclusive, that we take comfort knowing that science is there to convict the guilty and acquit the innocent. We don’t quite get how it works, but if trusted scientists say it’s the real deal, we are thrilled to believe. It’s one less fear to check off the list, the fear of wrongful conviction.
Except when it isn’t.
The American Association for the Advancement of Science is asking the Justice Department to revise newly announced uniform language on what fingerprint examiners can use in court testimony and written reports, saying statements that “rest on speculation, rather than scientific evidence” should not be allowed.
Fingerprints? But they’ve been admissible for decades, and as long as we forget about Brandon Mayfield, we believe they’re as conclusive a means of establishing the identity of a perp as it gets, except for DNA.
“There is no scientific basis for estimating the number of individuals who might have a particular pattern of features; therefore, there is no scientific basis on which an examiner might form an expectation of whether an arrangement comes from the same source,” said Holt. “The proposed language fails to acknowledge the uncertainty that exists regarding the rarity of particular fingerprint patterns. Any expectations that an examiner asserts necessarily rest on speculation, rather than scientific evidence.”
We’ve long believed fingerprints are like snowflakes, each one unique. This belief has allowed examiners to testify with a reasonable degree of scientific certainty that their conclusion (pointing at the defendant), “his prints were on the gun,” was true. Sure, there are the problems with human examiners seeing what they believed when determining whether the prints matched, because human error can always make science look bad, but the fingerprints themselves?
And sure, the number of matches, of “loops, arches and whorls,” which started at seven so many years ago and has since grown, and grown, and grown again, certainly suggested that examiners may have assumed too much in the early days, even though they testified then as now that their conclusions were both scientific and irrefutable. But now? Well, now they’re really conclusive. Don’t worry your silly non-scientific head about it.
The one constant in the use of fingerprint identification is the belief that prints were unique, that no two people would have the same “loops, arches and whorls” as another person, and so the examiner could authoritatively state that “his prints were on the gun.” Not maybe or sorta, but definitively.
And, indeed, recognition is slowly creeping in, not from the judicial side of the equation despite the putative role of judges as gatekeepers, but from the science side where the disgrace of constantly realizing that serious science is reduced to a joke by courtroom exaggeration of forensic “experts.”
While the Uniform Language forbids an examiner from making “the unsupportable claim that the pattern of features in two prints come from the same source to the exclusion of all others,” Holt said in his letter to the Justice Department, “it does allow examiners to say they ‘would not expect to see that same arrangement of features repeated in an impression that came from a different source.’”
See the cool trick played there? A bit of rhetoerical legerdemain and a false statement is wiggled into a quasi-accurate statement that is artfully designed to create the same false impression. They “would not expect” it? Crafty wording there.
But before believing that this call to cast facts before swine will turn trial from games to the “search for the truth” we keep pretending we can achieve, consider what the AAAS argues should be the accepted testimony from fingerprint examiners.
The Justice Department’s guidelines, Holt proposed, should instead instruct examiners to adopt more tempered language, which is backed by scientific evidence and steers clear of unsupportable claims.
Such language would allow examiners to note when two fingerprints display “a great deal of detail with no differences,” Holt proposed. Yet, such an observation would have to be accompanied with the admission that, “There is no way to determine how many other people might have a finger with a corresponding set of ridge features, but it is my opinion that this set of features would be unusual.”
Nuanced. More accurate, provided you squint a bit to gloss over the vagary of “a great deal of detail,” which defies precise definition. But to the ears of jurors who have been weaned on a belief that fingerprints are golden, if not platinum, will it make any difference in their understanding of the fact that “unusual” isn’t the same as conclusive?
And what of the gatekeepers, our judges, who have similarly been weaned on the admissibility of prints when they weren’t accurate means of identification but were allowed to be offered anyway? Will they revisit historical decisions to reject expert testimony that is calculated to prejudice the jury with junk science and the false certainty of its practitioners? After all, what juror is equipped to reject the testimony of a fingerprint examiner who testifies that it would be “unusual” rather than to a reasonable degree of scientific certainty? They’re still admitted as experts, and the jurors are not.
What we keep learning as science advances is how many of the things in which we believe turn out to be not quite as believable as we thought. Yet, we invariably ignore this constant under the belief that at this stage in scientific development, we finally have a conclusive answer, just as we believed that a decade ago, 50 years ago, 100 years ago. Of course, we were wrong then, but surely we’re right this time.