From Abracadabra to Zombies
reader comments: polygraph
22 Nov 1999
I'm sure you don't like hearing this stuff, but I did take a lie detector test once (back in my less idealistic days, when I was applying for a summer internship at the National Security Agency) and they caught me lying in response to one of my interview questions. Before that, I was convinced that lie detectors couldn't possibly work reliably, and I was completely blindsided when they said they knew I had been lying.
(The question was about illegal activity, and at the time I was 19, and had gone to visit a 17-year-old girl that I had met on the Internet, and we, um, went to "second base". Due to the Christian church's powerful influence over laws governing such matters in the U.S., I wasn't sure if that was legal, although I have since found out that it probably was.)
So I agree that polygraphs are not *reliable*, because they don't work on some people and others can learn how to fool them, but I would disagree with your statement that they are "no better than hypnosis".
I did see a demonstration in a science museum once where one
participant would pick a number from 1 to 10 and the other participant
would ask, "Is it 1?", "Is it 2?", etc. and the first
participant would always answer "No". It didn't work on
everybody, but it did better than random guessing. That seems like a
pretty well-controlled experiment for a science museum, in keeping with
the principles of "Seeing is believing" that most skeptics would
(name withheld by request)
reply: If there were not experiences like yours, I doubt if anyone would believe in the reliability of the polygraph. It is easy to find confirmation for the polygraph thesis, but the polygraph's failures demonstrate the reasonableness of not allowing polygraph results as evidence in our courtrooms. The courts are equally reasonable in not allowing as evidence any testimony first gained by using hypnosis, which has also been shown to be unreliable.
17 Sep 1999
I have worked for a law enforcement agency for over 25 yrs. Based on my observations, I would never voluntarily submit to a polygraph examination. I've seen too many cases where the suspect showed "guilty knowledge" (while denying being involved in the crime) on the poly, and was subsequently cleared through further investigation.
That said, I must admit that there have been positive results from polygraph examinations. During the pre-examination interview (actually a form of interrogation), while the examiner is attempting to get the suspect to confirm in clear, unambiguous language what their story (alibi, etc) is. If the examiner is persuasive enough to convince them that the machine will detect untruths in their statement, many suspects will end up confessing to the crime without the test. In the same manner, during the post examination interviews other suspects, confronted with "evidence" of their untruthfulness, will confess to the examiner. While my attitude may be incorrect, I see little problem with prosecuting folks who voluntarily confess during a voluntary polygraph exam which cannot be used in court.
Perhaps this is a function of our Western Judeo-Christian culture ("Confession is good for the soul," and "Always tell the truth") that is the cause (providing a blame-free method) of the confessions. I'm not familiar (anecdotally) with any of the confessions gathered this way being found to have been false. I also wonder if someone raised in another culture that doesn't value truth and openness (e.g. USSR or possibly Japan) would show any of the reported physiological signs of lying.
In that I am still employed by the police dept (even though my thoughts are very [probably too] well known), I'd appreciate it if my name is withheld.
15 Sep 1999
Have you seen the article by Tim Beardsley entitled "Truth or Consequences" in the October Scientific American? It seems that the FBI wants to search for spies by giving "lie detector" tests to scientists at Los Alamos.
At the top of page 24, we read
... the most notorious spy of recent years, Aldrich Ames, passed routine polygraph exams as an employee of the Central Intelligence Agency, as did another former CIA employee and convicted spy, Harold J. Nicholson.
That turns out not to be evidence of polygraph inaccuracy, however, because near the end of the article, former FBI counterintelligence official, Edward Curran,
... hotly denies that the polygraph failed to raise suspicions about Ames: the polygrapher in that case made errors, Curran maintains, because subsequent examination of Ames's polygraph charts shows evidence of deceptiveness.
Yeah, right; and Dixon predicted Kennedy's assassination, too. This is clearly an ad hoc hypothesis, especially in light of another thing that the article points out:
... when polygraphs are used to test suspects for involvement in specific crimes, the tests rate as deceptive more than 40 percent of subjects who are later positively cleared. Because examiners know they cannot fail 40 percent of those being screened for a sensitive post, they set the hurdles for deceptiveness higher.
It sounds to me like they "set the hurdles for deceptiveness" to get the results that they're looking for.
reply: I read an interview with Ames. He claims he was able to beat the machine by doing such things as squeezing his toes in tight. The unreliability of the device is well-known.
The Associated Press reported on a Sept. 14th meeting between more than a dozen researchers at Lawrence Livermore National Laboratory and a panel from the Energy Department. The scientists were said to have "vented their anger" at the government's plan to force top-security researchers to take lie detector tests in an effort to catch spies. Those who thought that what Bill Clinton lacks in moral character he makes up for in intelligence may have to reassess their opinion. This attempt to catch spies is inane.
The scientists noted that the results of a polygraph are not accepted in court and their value has been debated from the moment of their inception. Thomas Thomson, who has been designing weapons for more than 30 years, said:
The question is not about the veracity of polygraph screening--they are well known to be useless for that purpose. The question is, what dissenting views are they now afraid of? It is a truism that institutions only worry about heresy when they have begun to rot from within. Just what rot is this administration trying to hide from the Congress and the American people?
Another scientist warned that the tests will make it harder to recruit and retain talented scientists.
In case you haven't heard, the Clinton Administration has ordered the polygraph tests, which will pose these four questions:
- Have you committed espionage?
- Have you committed sabotage or terrorism?
- Have you illegally disclosed classified information?
- Have you had any unauthorized contact with any foreign intelligence service?
If the polygrapher thinks the subject has lied about any one of these questions, the subject will have to take "a more complex text" and be investigated by the FBI.
This sounds swell and is likely to attract all the best minds of our generation to want to work for the government. Guilty until proven innocent! The inmates have taken over the asylum.
I imagine Doug Williams will sell a few more copies of his manual on How to Sting the Polygraph.
11 Sep 1999
Your page on the polygraph is outstanding! I have been researching the polygraph for about a year and every issue you addressed about lacking credibility is well supported.
I have my own web page at www.stopolygraph.com
with my research.
03 Sep 1999
While polygraph must never be absolutely relied upon for exactly the reasons outlined in your article, I have found it to be a useful tool in criminal investigation where objective questions can be formulated about the subject's own actions in a specific event. As with all tools, its limitations must be recognized. Aside from the most common use as a demonstration to the actor that he is detected, let me briefly relate an interesting case.
An adjacent row of shops found burglarized. Some clear theft (large quantity of expensive clothing) and some vandalism (beer bottle thrown into a fish tank) of a kind not normally associated with common burglary.
Ex-convict on parole for burglary identified as being parked behind the center at about 4:00 a.m. the night of the burglary. He left town (violating his parole conditions) when he learned inquiries were being made. Two young men who had been playing pool late in a shed behind the center were identified. They agreed to polygraph which showed one lied when he said he did not enter the shops in question that night but was truthful in denying that he took anything. One lied when he said he did not know who entered the shops or who took anything from them but was truthful about not entering himself. ("Lied" and "was truthful" refers to polygraph indications. Neither confessed to anything.
The ex-convict was induced to return after he decided it looked bad to run. He agreed to polygraph which showed he was truthful about knowing nothing of the burglary. (He stated he had parked behind the center to urinate.)
Months later, a very active traveling professional burglary was caught in another city and confessed to this one also (and many others) to get concurrent sentencing. He was a thorough professional who stated he kicked at a business door across town to set off the alarm to distract patrol officers, a fact that had not been known to be connected with the crime. He also never indulged in vandalism during a crime and was solely engaged in stealing the dresses.
It is reasonably certain that the pro did the burglary, and the pool player, finding the back door open, entered (drunk) and did random damage, his friend waiting outside out of fear of discovery.
My sample of polygraph experience over 12 years is not large enough to establish much more than its usefulness as a tool. But is was undeniably useful, even though results are not admissible evidence in Texas. The operators, state police employees, made no rash claims and were scrupulous and clinical in their approach. Much different from the pre-employment sort of private examiners I have known.
It was also interesting to note that a significant number of true pedophiles who were known beyond any reasonable doubt to have engaged in intercourse with young children were able to lie about their acts without effect on polygraph, while their response to the directed control lie was normal for a lie. I have never been given a satisfactory explanation of the phenomenon. They are certainly not deluded to the point that they believe they are truthful when they answer "no" the question, "Did your penis contact any part of ****'s body?" or "Did you cause anything to penetrate ****'s vagina?"
Terrific web site. Keep the faith (or perhaps, Keep the Doubt
05 Aug 1999
Visiting scenic if foggy Carmel-by-the-Sea on business, I found an article in the Carmel Pine Cone about the use of a voice stress analyzer by the Monterey County Sheriff's Office, in one of their criminal investigations. The article claimed that the device is a type of lie detector that is capable of 95% accuracy, evaluating yes-no answers, which may be used without the subject's knowledge (over the phone, for example). The article was completely uncritical of the claims.
Searching the Internet, I found some very interesting research conducted by the Defense Department Polygraph institute. In at least three scientific studies, the device yielded only about 50% accuracy (i.e. no better than random chance), both for detecting lying and truthfulness. Also, a lengthy article in the San Diego Union-Tribune (12/1/98) described abuses of the device by the San Diego police department, the dubious marketing practices of its manufacturer, and the shady past of the company's owner.
If you would like to see all the information I have on this matter, please let me know!
Thanks for all the good work,
15 Dec 1997
After reading your article, I felt you had left some very important points as to "WHY" polygraphs don't work:
1. People don't want to believe that someone or something can tell when they are lying!
2. People don't want to be found out and exposed when they commit a crime!
3. People don't want to be found out when they lie to their parents, friends, employer, the police, the government, the IRS, the courts and many others!
4. People don't want to be found out when they steal from their family, friends, employer, their fellow Americans by cheating on their income tax and don't pay their fair share!
5. People don't want to be found out when they commit crimes against fellow human beings or our children!!!!
6. People, politicians, the Senate, the Congress, the rest of the government, and many others don't want to face their wrongdoing by being told by a simple instrument that they are lying!!!!
reply: If these are the reasons the polygraph doesn't work, I can only guess that the reason they work is because polygraph examiners want to believe they work so they can keep their jobs.
Let face it, the only reason the polygraph was banned in 1988 was because President Reagan had suggested that many of our politicians (some in congress) should take polygraphs. It wasn't very long afterwards, that congress voted to ban the use of polygraph.
reply: You don't think it had anything to do with their being unreliable?
Their are bad and good people in this world. We certainly don't ban all humans from this earth because there are a few bad apples. Their are some good cars and some bad ones, but we certainly don't ban all vehicles. However, when it come to the word polygraph, the lie detector machine, well.......then it's another matter. So, to avoid having to tell the truth or be found out, we then created other lies to protect the lies we don't want people to know. So, we say that the polygraph can't tell a lie and doesn't work (God forbid it worked). We claim it invades our privacy and violates our rights. And, you are right... it invades our privacy because now everyone knows we have lied. It violates our rights because someone has established that we have lied and we can be held accountable.
reply: Actually, we know that sometimes the polygraph works and sometimes it doesn't. Unfortunately, we don't have any way to tell the difference when it really matters. Truth machines are as valid as truth serums: both are based on wishful thinking not evidence that they are reliable.
If the question is whether or not there are good or bad polygraph examiners...then I
say yes. Let do something to weed out the bad ones with laws and professional guidelines
that will ensure good polygraph results. But to ban or outlaw a procedure that it's main
purpose is to establish the truth.....no, then I think there's sometime wrong with those
who choose to be anti-polygraph. These types of people may be the ones that have a lot to
hide and wouldn't want to be found out. If I was one of them, I wouldn't want polygraphs
reply: I think the question of competent and incompetent examiners is one we will have more success at than reliable and unreliable polygraph tests. Good intentions are not sufficient to warrant using the polygraph. People who rush to take a polygraph to prove their innocence are fools. Passing the test doesn't prove one's innocence. If it did, we could do away with trials altogether. People who refuse to take a polygraph may not be afraid of being caught in a lie; they may be afraid of having some machine say they are lying when they know they are telling the truth. Frankly, if I were running a business, I would not sleep any better knowing all my employees passed a polygraph test.
08 Apr 1997
My husband who has a seizure disorder has never passed a lie detector test. He worked in several jewelry stores that required all employees take periodic tests or be fired. His test results were always the same -- he told lies. What happened was the stress being tested was so great that he would have an absent seizure (nodding out) or go into a fugue state, where he was 'not here'. (His medicines can only control 80 percent of his seizures.) However, in his current job, he has a U.S. government clearance of 'Secret'. (The F.B.I. did the field work to detect his honesty.) He and I agree that lie detector tests are no measure of honesty.
28 Oct 1996
I thought this treatment of polygraph machines as "lie detectors" was pretty good. Let's face it: this is Gilligan's Island science, not real science.
You left out one interesting issue, though. How come Phil Klass, considered (by "skeptics") as "the leading UFO skeptic," puts such a pathetic credence in the things? If someone makes a claim Klass doesn't like, he wants the guy to take a "lie detector" test. Why don't the other "skeptics" wise up their colleague so he'll stop making such a fool of himself--and proving, by the way, to those who want to believe it, that the "skeptics" have no real arguments for their position? Or does he just keep doing it because he wants to cow people into submission, and not say they've had experiences he "knows" they couldn't?
reply: I agree with you. Lie detector tests would not be of much value in testing alleged UFO sitings or alien abductions. I suppose the threat of such a test might scare some fabricators into owning up, but other than that I think such tests are of little interest.