Heston’s response is classic and profound (not paraphrasing):

“The only place where police work is easy is in a police state!”

“I suspect there’s some profound misunderstanding on your part between the initial investigation into a physical correlate to disease or lying and the implementation of a practical differentiation tool.”

I suspect that there’s some profound misunderstanding on your part between the initial investigation into a physical correlate to disease or lying and the implementation of a practical differentiation tool too.

i.e. it is very easy to find activation differences between two conditions in a study using fMRI due to the dangers of small sample sizes versus lots of signal noise. One of the most heavily trailed studies did 6vs5 or something equally ridiculous. You need to have shown that this pattern is the same in multiple experiments, and even then, the true test really is creating something that works in practice. Recent attempts to characterise malignancy based on gene expression profiles have proved not to work in practice, even though initial experiments showed that cell malignancy could be discriminated based on these expression patterns – why did they fail? Because they assayed too many genes, with too much variation to be sure that the expression patterns between the two groups weren’t simply due to chance – that is they overfit the data. The next step is to take your newly discovered patterns and see how well they classify a whole new set of cells, might turn out they don’t work well on a new sample at all, and ou’ll have to go back to the drawing board. The same with these fMRI studies, you need to show that you can distinguish lying reliably, and blind. They haven’t done this yet, so their technique is just as reliable as a brain scan for schizophrenia – i.e. it isn’t.

That would be funny, but, of course, we’re talking about functional MRI, where changes in brain activity are measured.

There’s no question that there are neural correlates of lying. Still, any modality that can accurately measure those correlates is dependent on an interpreter because the neural feature of interest has variability in and between subjects.

There is no gold standard to lie detection. It’s all witch-doctoring. Yet, any modality that can measure neural correlates normally invisible to the observer, even an expert observer, is going to be useful in lie detection.

“I haven’t followed the literature closely, but it was always my impression that the brain scanning lie detection stuff was all post hoc – much like a lot of these gene arrays for detecting diseases…

I suspect there’s some profound misunderstanding on your part between the initial investigation into a physical correlate to disease or lying and the implementation of a practical differentiation tool. Or, perhaps you’re addressing the fact that any such tool will pick up artifact and that there isn’t anything that is 100% efficient, in which case, I agree.

I doubt MRI’d be much use, structural brain changes associated with lying ;-)

I haven’t followed the literature closely, but it was always my impression that the brain scanning lie detection stuff was all post hoc – much like a lot of these gene arrays for detecting diseases – they show that you can divide the two groups up by characteristics of the brain scan, but this is partly a result of the surfeit of data (i.e. they’re overfitting) – but that they haven’t really shown that these characteristics are actually generalisable.

Yes, that’s definitely possible. But the polygraph could also help in eliciting true confessions, before it is even used. If you are guilty and not too bright, and believe your interrogators are in the possession of a Magic Truth Machine, you may conclude that lying is futile.

And if the subject does not believe in the power of polygraphs, but believes that their interrogator does or might believe in their power, the use of the polygraph is likely to make the subject nervous and anxious, and the anxiety itself may help to trip them up as they seek to remember their lie and make it sound credible. A polygraph can thus be a useful tool of interrogation, even if it is at bottom a crock. the same is true of hot lights, alleged “psychic” investigators, ouija boards and juju beans.

I think the situation with polygraphs is similar to the use of hypnosis in psychotherapy. Some patients actually believe in the power of hypnosis, and can be persuaded that they are in a “hypnotic trance”. This can be useful in getting them to reveal things they are hiding. Putative hypnosis sessions can also be useful with subjects who don’t believe in the power of hypnosis. You may have a situation in which the patient has some secret they would like to share with the therapist, but is too embarrassed to impart it. So the therapist and patient participate in a little confession game based on the following unspoken understanding on the part of the patient: “I want to tell you something, but I’m embarrassed. So you pretend to hypnotize me, and pretend you believe in hypnotic trances. I will also pretend to believe in hypnosis and pretend to be in such a trance. Then I can tell you my secret with a distant, faraway stare, and you can pretend to believe I am speaking from the depths of my trance. We thereby avoid the discomfort of direct face-to-face confession. We will both know that we are both dissembling, and know we know it – but we can tactfully avoid articulating our knowledge. I can also pretend not to remember what I just told you after I’ve finished telling you, and you can pretend to believe me.

Players at a Romanian first division soccer club will be forced to take lie detector tests after losing two matches.

Last week National Bucharest surprisingly lost 1-0 at home against Vaslui and 2-0 at Jiul Petrosani to dent their chances of playing in European competition next season.

“It’s about a moral audit which will allow our players to show they played honestly against Vaslui and Petrosani,” the club’s main shareholder Constantin Iacov said.

“I think it’s normal to check your employees from time to time,” he added.

“I’ll be the first to undergo the polygraph test and the team’s coaches soon after me.”

-Reuters