The "Lie Detector Machine" was popularized by William Moulton Marsden, who styled himself "the father of the polygraph". "Polygraph" is the formal name for the device. "Poly" means "many" and "graph" means "to draw a line". So a polygraph is literally any device that simultaneously draws several lines.
But we are going to restrict ourselves to the common usage of the word "polygraph". In common usage, a polygraph is a device that makes several simultaneous physiological measurements of a "subject" and charts the results with the intent of separating true statements from lies.
The polygraph has featured prominently in countless movies and TV episodes. There, a serious person, the "operator", hunches over a "chart recorder", on which several squiggly lines spool out onto a long sheet of paper. Various gadgets are connected to the subject. The output of each gadget drives the activity displayed by one of the lines.
The operator asks the subject a series of questions and periodically makes makes cryptic marks along the edge of the chart. When the session ends, he declares either "he's lying" or "he's telling the truth". He does this by interpreting the squiggles on the chart. So, what's going on?
Let's start with what's being measured. One line tracks heart rate. Another tracks blood pressure. Still another tracks the rate of respiration. Another common measurement is skin conductivity. In this latter case, the idea is that perspiration increases skin conductivity. So the intent of the measurement is to determine whether the subject is sweating or not. The operator's cryptic marks are used to indicate when various questions were asked.
Marsden didn't invent the polygraph. The invention is generally credited to James Mackenzie, a cardiologist. But many people, including Marsden, contributed to the design that eventually moved into widespread use. In 1938, for instance, he published a book called The Lie Detector Test. And it is Marsden who deserves the bulk of the credit for the device becoming a tool used commonly by law enforcement and allied disciplines.
It helped that he was a credentialed scientist. He received both a Law degree and a PhD in Psychiatry from Harvard. He went on from there to teach and do research at the college level for many years. All this preceded his efforts to introduce and popularize the Lie Detector.
So when Professor Marsden later opined on the Lie Detector, it is not surprising that people listened. Finally, he was a born showman. He had something to say and a way of saying it that attracted attention while also sounding completely credible.
All in all, he was a fascinating character. There is a very interesting movie called Professor Marsden and the Wonder Woman that is an easy way to learn more about him. Yes! He's also the guy who invented the Wonder Woman of comic book and now movie fame. And he spent most of his adult life living under the same roof with two women and their children. Like I said, fascinating.
But back to the device itself. Marsden succeeded in convincing large segments of law enforcement and the public that the polygraph was an accurate way to determine whether a person was lying or telling the truth.
But the actual device is more wish fulfilment than reality. We wish that there is an infallible way of determining whether a person is being truthful or not. So, if "experts" tell us that the polygraph can perform that feat, we want to believe them. But does it?
The potential effectiveness of the device depends on an assumption. The assumption is that your level of stress is higher when you are lying than when you are telling the truth. Telling the truth is easy, so the assumption goes, so that you will remain calm while doing so. Lying leads to agitation. Agitation leads to stress. If this assumption of a correlation between lying and stress is true then the device should work as advertised.
The device does do a pretty good job of measuring stress. And that led to a considerable amount of early success. Back then, few people knew much about the machine or what it actually did. Marsden and other proponents would spit out a bunch of authoritative sounding mumbo jumbo. Both members of law enforcement community and people suspected of a crime had every reason to believe what they were saying. And that led to a self fulfilling prophesy.
If a suspect believed that the Lie Detector worked, and why wouldn't he, then all of a sudden it becomes very dangerous to lie while hooked up to the machine. And danger is closely associated with elevated levels of stress. If follows that an attempt to lie is likely to increase the suspect's pulse rate, quicken his breathing, cause him to start sweating, etc.
Not all the indicators changed significantly every time. But enough of them did for a skilled operator to detect the changes that signaled a lie. The polygraph's ability to detects changes in the level of stress a suspect was experiencing actually did correlate reasonably well with whether he was lying or telling the truth.
But all this depends on the subject not knowing much about the machine or how it works. The widely held belief of the time that the machine performed as advertised completed the chain of causation that made the machine work much of the time.
But the device has now been around for a hundred years. Far more is now known about the machine by anyone who has looked into it. But also and more importantly far more is now known by criminals and others who have a vested interest in learning how to reliably "beat" the machine by lying without the machine catching them at it.
They now know that the machine can be beaten, and it can be beaten fairly easily. How often is it actually beaten? We don't know because the people who are in a position to keep track don't. They have a vested interest in maintaining the fiction that Lie Detectors work and are highly accurate. So they go out of their way to avoid keeping track of how often and by whom the machine has been beaten.
The basic way to beat the machine is to break the correlation between lying and stress. Even in the early days some people were able to beat the machine. Back then nobody understood how they were able to do so.
We now know that anyone who doesn't find lying stressful can easily beat the machine. Many con artists fall into this category. The very attributes that contribute to fooling potential victims contribute to fooling the machine.
Early on, the mechanism didn't matter. What mattered was that in criminal circles the news got out that some people were able to successfully and repeatedly beat the machine. That led to a general decrease in fear of the machine by the criminal element. And that led to a decline in the machine's effectiveness.
And over time the mechanisms that allowed the machine to be beat became clearer. That led people who were not natural con men to wonder if there were methods they could use to beat the machine. And it turned out that there are several.
Various forms of meditation aimed at calmness can be used. Alternatively, techniques used by actors can be used. One technique actors use is to inhabit their character. If the mind can be fooled (or trained) into believing that the actor is actually the character being portrayed then the body takes its queues from the mind and responds accordingly.
If the character believes something is true then the physical responses of the body will react accordingly. Pulse and respiration will be low and steady. Palms will be dry so skin resistance will be high. The Lie Detector will indicate that the response is true.
Early on it was hard to determine which group should be believed, the "it works" group, or the "it doesn't work" group. But since then the "it works" case has gotten progressively weaker and the "it doesn't work" case has gotten progressively stronger. In fact, no evidence to support the "it works" position has emerged since those early days. However, the evidence that it doesn't work keeps piling up higher and higher.
In fact, a quick web search will turn up numerous "how to beat a Lie Detector" courses. Trust me, if these courses didn't deliver results, they would have long since been hounded out of existence by unhappy students. And the fact that many of these courses are run by people who have retired from the communities that routinely use the machines tells you everything you need to know.
But by now the people who have long championed the routine use of Lie Detectors are heavily invested in their continued use. They really didn't want to the general public to know how often they fail. The early failure rate was low. But it kept increasing and increasing and increasing. And that just made it more and more important for the Lie Detector's supporters to keep the myth alive.
Distressingly, another group that has become heavily invested in the Lie Detector is the intelligence community. They have a desperate need to know which of their employees can be trusted and which can't. Decades ago they pinned their hopes on the Lie Detector.
Yet it is routine for people who turn out to be untrustworthy to pass multiple routine Lie Detector tests. Pretty much all of the high profile espionage cases you can think of fall into this category. But efforts to do away with it's use within the intelligence community keep failing. It still remains in routine part of employee screening.
If I was teaching a "how to beat the Lie Detector" course I would have a Lie Detector on hand. That way I could keep testing students until it became apparent to them that they could reliably beat the machine. That would be the best way to show them that they were getting value for their money. It would also have the additional benefit of making the taking of "real world" Lie Detector tests far less stressful. (If they couldn't become proficient in beating the machine I would refund their money.)
Lie Detectors are inexpensive and easy to get hold of. They consist entirely of standard electronic/medical components. You can find one that substitutes your computer for the chart recorder on Amazon for from $100 to $200, depending on how fancy you want. For instance, the "Police" version will set you back $140 plus tax and shipping.
It's bad that anyone can easily get their hands on one so that they can play around with it. It's even worse is that pretty much anybody can become a "Polygraph Examiner". All you need to do is get a job administering polygraph tests.
Many organizations that use polygraphs require no training or accreditation for the people that they hire to operate them. After all, it's not very hard. You can figure out most of what you need to know by watching a few movies or TV shows.
Oh, there is an organization, the American Polygraph Association, that provides a certification service. And many organizations do require their operators to graduate from an APA certified course. Doing so allows you to be able to call yourself a "Certified Polygraph Examiner".
But it is harder to get a license to sell real Estate than it is get your APA certification. This seems like the sort of thing a Community Collee would want to provide. But there are no colleges, community or otherwise on the APA list. The whole thing is so sketchy that even for profit colleges won't touch it.
And this is a big problem because it is relatively easy for the examiner to cook the results so that the guilty look innocent and the innocent look guilty. To make an innocent person look guilty, for instance, all the examiner has to do is to use various techniques to bias the results in that direction.
He can change his tone of voice. He can vary the timing of his questioning. He can emphasize or de-emphasize various words or phrases. Or he can just crank up the sensitivity of the machine whenever the subject is asked a key question. That will make the needles swing wildly even though the subject is completely calm. The possibilities for mischief are endless.
This sort of manipulation is forbidden by the APA guidelines. But the enforcement of the guidelines is completely ineffective. And there are often incentives present for biasing results in the direction of the outcome preferred by whoever is paying for the examination. Is conscious manipulation rampant? No one knows. I suspect the big problems are incompetence and overreliance on ambiguous results. But I don't know.
It should be no surprise that there is no scientific validation for the idea that a polygraph examination functions as an effective test for the honesty of the subject. In fact, to the extent that careful studies have been undertaken, they confirm the unreliability of the method. James "the Amazing" Randy, the late noted magician and debunker, has beat a Lie Detector under controlled circumstances. It's just not that hard to do.
So why are they still in widespread use? Follow the money. There is no money in admitting that they are bunk. There is, however, lots of money in pretending that they work. This is one of the many situations where money talks. And what it has to say makes things worse, not better.
