Originally published at: https://boingboing.net/2018/10/01/ducking-stools.html
…
While I think it would be great if there was a technology to formally interview a person without the possibility of deception. The technology does not exist. Even assigning some reproducible estimate of risk that has any statistical meaning does not exist.
Short of actual mind-reading technology, I doubt a real-world alethiometer can exist even in principle. Any such technology would depend on analyses of physiological responses just like polygraphs do, and those responses are shown to be tied to emotional state, not honesty.
In principle a device with sufficiently high resolution imaging of brain waves would record information about whether the subject believes what they say, but we have absolutely no idea how to interpret detailed information encoded in the brain, and proposals for such imaging technology are still highly speculative and well beyond the theoretical limits of exocranial nuclear-magnetic resonance scanners. Algorithmic learning might be able to suss out the meaning of brain activity in some human-readable way, but we’d never be able to follow it’s internal logic for doing so.
Funnily enough, I just finished reading Snow Crash. I think the bits about the Feds having to repeatedly undergo polygraphs was meant to be viewed as satire?
Its not that polygraphs don’t work - they do. Its that what they measure isn’t lying. I’ve used polygraphs in electrophysiology research (although they are now outdated) and they can be great instruments. There are a lot of things that trigger anxiety responses other than lying, and lots of people who can lie freely and stress free.
I once interviewed for a job with a follow-up to take an on-line “personality test”. I took the test but fudged the answers because I knew what they were looking for. The prospective employer then asked me to take a polygraph. I refused, telling them polygraphs to determine truthiness was akin to witchcraft. They hired me, I worked there for years and handled transactions worth millions of dollars. The End.
I’m lying right now.
I’m detecting your lie right now.
They also work in a very real way if the test subject believes the test is real. It’s like the placebo effect of criminology. There’s even an old story about some cops who put a collander on a suspct’s head. Then they had the copier spit out pages saying “LYING” until the suspect broke down and confessed.
I can’t recall if the confession was admissible.
If the results of polygraph ain’t admissible as evidence in a court of law, then I’m never taking one. Period.
Why?
Would you also require the interviewers/employers to be similarly tested?
No, you’re not. In order to lie, there has to be something to lie about.
Yes and no.
They don’t “work” under those circumstances because they detect lies.
They “work” in that its a useful interrogation technique. And they work as one for the same reason that they’re bullshit. The subjects believe they’re real. Which makes it easy to create stess responses (all they can detect). Then you point at that stress response and acuse them of lying about something you already know or suspect they’re with holding on. Its a trick, intended to get people to talk. It does rely on the subject believing in it. But it only works because the results are so easy to manipulate. And “interpreting” them is a bit like an ink blot.
Its also why PD applicants have to undergo a polygraph. If handled properly its nothing to do with discovering disqualifying lies. But part of the psych evaluation to see how some one responds to stress. A subject the interogators to interogation sort of thing. And from what I’ve been told its common to tell applicants that failed the psych evaluation that they failed the polygraph. After getting them to admin lying about something minor. Like smoking pot or underaged drinking. Welp! Lied about something, out on a technicality!
They’re not even particularly good at that. The polygraphs commonly used in law enforcement, alien abduction “research” and job interviews these days aren’t the same sort of machine that used to be used. Tend to be a laptop with some proprietary software package from various lie detecting “expert” companies. Its a whole weird industry. As much as the old school ones were a reasonable, now outdated, way to record certain physiological markers. The ones in common use now aren’t even as good at it as the vintage ones are.
Hey kids most of the people in spooky jobs live and die by these tests. Just say’n.
So you are saying he is lying about lying? So he’s telling the truth then. And the truth is he is lying.
No, it happens all the time, that someone believes they are making a falsifiable statement, but they are in fact not. He’s not lying, he’s simply mistaken. He’s not even wrong.
I see there is a Kripke fan in the house.
It’s all Security Scientology.
And I’m not even kidding. Polygraph examiners fall into the same niche as Scientology auditors, where they’ve got a pseudo-scientific tool that they’re convinced works, with nonsense folklore about how to interpret the readings.
“If it reads on the meter, it’s true.”
I think I’ve mentioned this here before, but I like this story. When I was 16 I applied to work at Tom Thumb (now mostly Randall’s) and I had to take a polygraph test. To work as a fucking bag boy. Now I was and still am nearly phobic about that goddam pressure cuff used in taking blood pressure. This is used CONSTANTLY during a lie detector test which can last forever and a day. I was soon practically hallucinating and realized that I could say any damned thing I wanted because I was about to pass out from fear. And I did. And I passed. Whee. Got to be a bag boy. I will never, never, NEVER take one of those again! I still cannot stand to have my blood pressure taken.
The weird thing is that I always test out as having really good blood pressure.
That they’re convinced will pay the bills. A not insignificant portion of both groups know full well its bullshit. But man do they need to renovate the vacation house.