There are so many tests and techniques used by US law enforcement that are known to create false positives, are highly arbitrary (and the testers are easily swayed by the cops), or just generally proven not to work. Yet they’re still in widespread use (and being added to all the time). US cops would rather catch someone than the right person.
They used to match bullets from crimes to bullets still in suspect’s possession by looking at the exact alloy make-up. Someone showed that even if there were perfect matches (and exclusions) for same/different bullet batches (which hadn’t been convincingly demonstrated), lead from one batch could end up in bullets from different manufacturers, and consecutive bullets from one maker might have different lead, and the whole test was actually pretty worthless. Despite being discredited, I’m not sure if that test is still in use or not…
There’s a witness identification technique that’s known to have a high chance of creating false positives in identifications. Pretty much every developed nation but one stopped using it. Guess which country still widely uses it!
Also some DNA testing. They get everyone on board with the cleanest use (comparing substantial fingerprints; comparing a DNA sample from one person with one suspect), then start expanding to suspect cases where it’s down to individual interpretation (fragmentary fingerprints, mixed DNA samples) and pretend it’s all the same…
I saw a study done some years ago where they sent a mixed DNA sample to various labs along with a “suspect’s” DNA and asked if the suspect was part of the mix. They used different “suspects” in what they sent to each lab, but the labs were always happy to match the various peaks and troughs in the mix to whoever was identified as the suspect (even though they didn’t quite match), even though in no cases was their DNA in the mixed sample. It’s bad - the problem is labs are being asked to match X to Y (with the understanding that these are the two the cops want to match), and all you’d need to disprove it is to change what X is.
I think it’s pretty telling that US police murder clear-up rates when down around the time DNA testing became widespread. This suggests standard police operating procedure for murder arrests was, “create a convincing narrative around this suspect, evidence be damned,” but DNA could be used to exclude suspects, so that ruined things.
This is great news. Justice will only be served when we filter out all the junk science in the court room, and, there is a lot of it. Blood spatter analysis, bite mark analysis, polygraphs, bullet matching, handwriting analysis, burn pattern analysis, eyewitness testimony, the list goes on and on. It’s all bullshit and countless innocent people are sitting in prison for all of it.
Off the top of my head, I can’t think of a single forensic practice outside of DNA that has actually stood up to scientific study. It all collapses when looked at with proper scientific controls. This is not surprising when you look at the origins of it all. Each of those fields was made up by some guy who wrote a book, then went around selling it to police departments. “Oh I’ve looked at a lot of blood, so I know what causes which patterns” is the beginning and end of their proof of their systems. None of it except DNA originates from science.
Complicating the matter are “dull” cops who somehow make it to detective (consider the pool that detectives come from) and prosecuters who have their eyes on being DAs and, from there, mayors. They want things easy and sellable (such as firearm identification) because of dullness and ambition.
To be clear, the use of subjective examinations is not what makes it unscientific. People can be trained to do this stuff methodically, and lots of medical diagnostic techniques work that way.
That’s what makes it unscientific.
Also: if there’s no data on ballistic experts’ hit rates, that implies the qualification process doesn’t involve, you know, taking tests.
So my first question is always, why not try harder to make it scientific (which, to be fair, is sometimes done. There IS research done on these various methods of criminal science, it just generally isn’t applied to actual investigations)
I think the answer is clearer if you realize that these techniques are tools to get convictions, and not tools to discover the truth. I do think most cops and DAs think they have the right guy and so they apply every tool they can to get a conviction, but they’re wrong about who “the right guy” is so often that it ends up being injustice.
I don’t know if this is really true or not, but at the end of a book I read recently (“People Who Eat Darkness”) they said that Japanese courts have insanely high conviction rates, but that’s because they generally won’t go to court without a confession, and they apply a LOT of pressure to get those confessions.
I’m not sure how often bad implementation renders the overall process scientifically flawed; but once you get out of the wacky world of cheap field test kits there are definitely some techniques in forensic toxicology that are borrowed from real chemists and are scientifically robust. Putting the person operating the mass spectrometer under the DA or the police department on the org chart creates significant perverse incentives; but that’s not really a mass spectrometry problem.
That makes sense. After all, the tooling marks on the barrel that were incised into the bullets were ultimately made by the tools at the factory. So barrels manufactured by the same tools consecutively are going to have similar tooling marks. And they may or may not have close serial numbers since they aren’t generally serialized until later in the assembly process.
Ballistics isn’t reliable, neither are fingerprints or arson patterns. False confessions and accusations happen all the time. Even honest eye-witnessing is suspect. Video and photo fakes can be undetectable, and therefore can’t be relied upon. Digital tracking finds equipment, not necessarily the person who used it. Even DNA isn’t foolproof.
So exactly how are we going to ever convict anyone of anything? All of the above can be used in courts where the opposing councel is free to educate the jury about unreliability. Except now for bullet forensics in Maryland. If this spreads, in the future all lawyers will have is names, cause and time of death (exept coroners also rely on subjective experience, do it’s only a matter of time before it’s illegal to use in court).
Nothing is error-proof. Nothing. So whether something is error-proof or not should not be the yardstick used to qualify evidence and testimony.
I’m not sure what you’re advocating for. The use of completely bullshit “expert testimony” to convict people, because otherwise it will be difficult to convict people? Won’t someone think of the poor district attorneys??
there’s a vast difference between error proof, and error causing. a lot of the techniques people are mentioning in this thread are the latter, and yet they are represented in court as the former
i think the path from beat cop to detective is ridiculous anyway.
detectives should go to school for detecting ( criminology, sociology, forensics, … ) being a cop doesn’t seem like it would provide any of the knowledge or skills needed for investigation
This is a straw man and a false dichotomy. There’s a huge middle ground between flawless science and total bullshit. Plenty of room in there to draw a reasonable line.
Sending someone to death row because some guy wrote a book full of nonsense he made up about blood spatter (which has literally happened) is well over that line.
Climate change denialists make the same arguments for doing nothing because making things better won’t make them perfect.
If the only way we have to convict someone is nonsense about bullet markings matching barrels (and to be clear, it is nonsense. Not just low accuracy, but total made up nonsense) then we shouldn’t convict that person. That’s the whole point. If your evidence is all garbage, the right thing to do is let the person go, not convict them just because you can’t find better evidence.
In my defense, I’d argue that the point of a headline is to make someone read the article and learn more — and it sounds like that’s precisely what this accomplished.
