Fingerprints Addendum – Stuff I forgot about in my huffing and puffing about studies
In 1995 there was a big, and to my knowledge one of the only publicly known, proficiency study of fingerprint examiners! They tested 156 examiners by showing them seven “real” latent criminal prints and gave them four 10-print cards. Five of the latents were known matches to the cards furnished. I’m gonna just outright steal the numbers from the study here: 44% of them noted all five matches and the two eliminations. That means 56% provided wrong answers. 6 examiners couldn’t ID any of the latents. 58 examiners made false positives…twenty two percent. More than 1/5th. Just saying….
Age Determinations (of the latent fingerprint)
You can’t do it. It’s not possible. No way to know when a print was left. At best, you can say “It was definitely left here after the last time this thing was thoroughly and rigorously cleaned!” Sometimes experts will try to say a print is “fresh,” whatever that means. No one knows; it’s intentionally ambiguous and designed to create false impression with the jury.
The surface itself, the environmental conditions, the thing the print was made in/with all factor in to how “old” a print might appear. We don’t know the statistics on these factors so there’s really no way to know. Certain visualizing techniques work on the oils left by fingers, certain ones work on the salts, others on amino acids. Theoretically, the oil ones reveal “fresher” prints but environmental conditions may preserve oils or disperse them prematurely. Also not every fingerprint you leave contains all the age signifiers. Again, environment!
There have actually been some promising studies on trying to suss out the factors involved but they always ended in fuzzy conclusions and “further studies needed.” This was in 1986. No further studies were done. Looks bad for the concept!
Examiners have sometimes claimed they can determine if a fingerprint was left by a man or a woman. This isn’t done through some chemical analysis but just by looking at the print. Sound fishy? You’re right! Men have, allegedly, “coarser” ridge detail and women have “finer” ridge detail. Sound even fishier now? You’re RIGHT! There were significant overlaps in density between men and women and the author called for “replicate studies” because he doubted the repeatability of his own findings. I gotta say he’s a pretty honest dude to say that in the field of forensics, especially fingerprints. So good for you!
I raged against computer matching because they don’t match. They do narrow the field marvelously. What would have been impossible, a “cold match” – a latent matched abstractly to the fingerprint database – is now extremely common! A neat news article from July 10, 1984 in the Christian Science Monitor, Fingerprints: With New Technology, a Better Crime-Buster by some guy named Armstrong said that 19% of arrests in San Francisco were made with cold hits. In Houston it was like 30 arrests a month. That’s a lot of likely unsolved crimes now bearing fruit. Either that or its freeing up investigators to review tougher cases. Good thing either way probably.
Context Context Context!
We’ve discussed this in almost every other technique and this is no different. The context of the fingerprint is hugely important! Finding a fingerprint on a wine bottle placed on a store shelf is not damning at all as a defendant may have touched it while browsing. Finding a fingerprint on a wine bottle held in a private cellar of Sir Charles Henry Thurgood Pierce is an entirely different matter. Unless you can pull a Law & Order and show that Sir Pierce was having a hot gay affair with his alleged burglar then you are pretty fucked evidence wise. In the same way, finding a fingerprint in the victims blood is pretty damning, but finding a fingerprint on a murder weapon that was commonly held by dozens is far less decisive.
Where was the print found? What were the circumstances surrounding the print? What alternative explanations are there besides CRIMINAL? These are crucial questions you have to ask when someone says fingerprint evidence. Any sucker can accidentally or casually pick something up. Finding prints on the shell casings still contained in the revolver used to shoot Mrs. Peacock are entirely different matters.
The Grain of Salt
I personally think fingerprints are probably reliable evidence. I believe they are likely unique and permanent. I believe they can sometimes transfer that uniqueness to a surface and that uniqueness can be read by a trained examiner. I believe that a trained examiner can sometimes accurately match a latent to an exemplar. However, I don’t know to a statistical certainty. I know there is an error rate. It could be huge; it could be tiny. I don’t know. The problem is, no one does. That is also generally the standard for admissibility (except for fingerprints).
We don’t know how unique prints are, how often or under what circumstances they transfer, how circumstances affect the uniqueness of the transfer, how circumstances affect the visualizing and collection of the print, and we sure as shit don’t know how much training you need to be able to accurately and reliably ID fingerprints. These are important questions! The answer is probably “pretty often and pretty reliably” to most of these questions, but how often, how reliably! 50/50? 60/40? 90/10? Fingerprint examiners will assert that there is a 100% accuracy rate to their IDs. That’s right, they openly assert perfection. Furthermore, they will claim the field as a whole has a 0% error rate. These kinds of claims cannot be allowed to stand! People have a right to know the actual numbers behind science being used to convict them. The fact that these numbers have been intentionally skewed, obfuscated, and sometimes fabricated leads me and many others to believe that perhaps the statistics would be highly unfavorable to prosecutors. Although when you claim perfection any accusation of normalcy is inherently dangerous.