establishment psycho-bunk 1 —

‘lie detection’
otherwise known as complete bunkum

a briefing document

site map

advertising disclaimer

‘lie detection’

index to report on ‘lie detection’
‘Lie detection’, otherwise known as complete bunkum.

From a press agency report:
‘ “We stress that no spy has been caught yet using a polygraph,” said Kathryn Laskey, an associate professor of systems engineering and operations research at George Mason University in Virginia.’
[Reuters, 08.10.02]

The‘lie detection’ ‘report’ is full of words and music, but amounts to what is widely known among sophisticates: ‘lie detection’ is a nonsense.

There is a story of some small town miscreant arrested by police as a suspect. The sheriff took out a colander and placed it on the suspect’s head. He then led a couple of wires to a copier machine on which had been placed, unknown to the suspect, a sheet of paper with the words, “he is lying”, in large bold print.
    Each time the suspect denied the crime, the sheriff would press the copy button, causing the machine to eject another copy of the ominous message, until eventually the suspect confessed.
   Whether the suspect was actually guilty is not recorded.

False confessions are surprisingly easy to wring from foolish and uneducated people. ‘Lie detection’ only ‘works’, in as much as it does, by the fact that many foolish people believe it does.

All the machinery detects is tension, nothing else. Naturally, if you are being questioned for a serious crime you will be tense, whether you dun it or not. The 57 studies upon which the report relies are all ‘laboratory’ tests where those being tested are playing a game. They are not studies related to real-world situations.

This is all the expensive 300+ page report tells you, but ‘professionals’ can’t really be paid to sit on a committee and produce a report if they just say, “it is bunk”; hence 300+ pages. And no-one becomes a member of such a committee by telling the paymaster in advance that it is all bunk.

Such is the politics of ‘science’ and associated government waste. Such are the ‘research grants’ that enable people to do the dozens of ‘studies’ quoted in the report. Such are the vested interests who provide the ‘lie detection’ ‘services’ and those who manufacture the machines. The ‘lie detection’ industry profits well from the sale of their high-priced, high-margin paraphernalia. Even the small players do well: machines (two tins and a simple galvanometer) used by scientologists, among others, are sold at a considerable price.

In primitive tribes, the witch doctor would dole out bunches of leaves to be chewed, for the same purpose, relying upon the possibility that the guilty party would have a dry mouth from fear, and thus would not wet up the wodge.

How long does it take humans to act rationally? ‘Lie-detectors’, and all their fancy brothers, do not ‘detect lies’. They only detect tension. ‘Lie-detectors’ can no more read minds than you can.

Humans are exceedingly good at lying, they also easily become tense. They are also often very ready to believe bunkum. They are also very good at fooling themselves that they are good at things when there is no evidence to support the assumption. Humans can even fool themselves into believing they are good at ‘detecting lies’, especially if they have some bits of machinery and a few dials to help them. But, in reality, you might just as well peer into a crystal ball, or use a colander, it would be as much use.

You can even buy a copy of this report for around $40, if you haven’t found some moderately introspective pigeon to tell you the results in advance, in far fewer and shorter words, that ‘lie detection’ is simple snake oil.

But don’t tell anyone, or ‘lie detection’ might not continue to be a useful deterrent.

For those wishing to understand just how bad the situation is, go to cause, chance and Bayesian statistics, particularly see end-note 2.

What is a lie?

Quite apart from these logical difficulties with ‘detecting lies’, understanding the meaning of the word lie is extremely complex. Here is some fairly detached logic; here are some points on the unreliability of memory; and here is text which discusses relevant pragmatics associated with ‘lying’, attend particularly to practical ethics.

If you have managed to get a fair grip on the foregoing, you just may have developed sufficient scepticism to read the following with due caution:

Telling Lies: clues to deceit in the marketplace, politics, and marriage
by Paul Ekman, WW Norton & Co, 2001, 0393321886, pbk.
$11.17 [amazon.com] {advert} / £9.01 [amazon.co.uk] {advert}

There is a tremendous drive in humans to believe that they understand, or can do things, which, to put it lightly, they cannot. There is a terrible inclination to confuse an ability to guess right some of the time with an ability to ‘know’ that which they cannot know. In my view, the author of Telling Lies has such a problem. However, the book does cover fairly thoroughly the subject of relationships between dissembling, human emoting and facial expression. This is the best book I know on this area, and the author has spent a very long time studying the problems.

This book, like so much that is written in the ‘social sciences’, is appalling in its disorganisation and lack of clarity of expression, but is probably what you are stuck with in the present level of development of academic writing on this subject.

Building on the work of Ekman, though it is not fully acknowledged, a Cambridge University group under Simon Baron-Cohen have produced DVD/CD-ROMs with actors trying to express ‘emotions’. The authors have attempted to break down ‘human emotions’ into 24 sections, with 412 ‘human emotions’contained amongst those 24 sections! This DVD (or 4 CD-ROMs) will be shipped to you for about £80! The compilation may be useful for actors, or for dedicated professionals, but it is not very well done and I cannot recommend it.

Ekman has spent much time, with colleagues, attempting to learn how to make forty-three facial muscles work independently from one another, in order to assess how easy or how difficult they are to control in the expression of emotion. His work is discussed, with what I regard as naive, journalistic hyperbole in an article, the Naked Face.

dishonesty-detecting textual analysis program cited in national election campaign

Computer programmes yet again are surpassing human performance.

“According to a new computer algorithm, [Canadian] Prime Minister Paul Martin, of the Liberal Party, spins the subject matter of his speeches dramatically more than Conservative Party leader, Stephen Harper, and the New Democratic Party leader, Jack Layton.”
—
“The programme analyses the usage patterns of 88 deception-linked words within the text of recent campaign speeches from the political leaders. They then determined the frequency of these patterns in each speech, and averaged that number over all of that candidate's speeches. Martin received a ranking of 124, while Harper and Layton scored 73 and 88, respectively.
—
“Conservative parliament members, such as Jason Kenney, point to the analysis as proof of their leader's honesty. "People used to think he's boring, but now they recognise that he's a straight shooter without the spin.”
—
“The computer algorithm is based on a psychological model constructed by James Pennebaker at the University of Texas, Austin, US. While studying the lying and truth-telling of hundreds of test subjects, he uncovered patterns linked to deception, such as the decreased use of personal pronouns - such as I, we, me, us - and exception words, such as "however" and "unless". ”

Pennebaker’s work on language patterns and deception

“At the University of Texas at Austin, psychology professor James Pennebaker, PhD, and his associates have developed computer software, known as Linguistic Inquiry and Word Count (LIWC), that analyzes written content and can, with some accuracy, predict whether someone is lying. Pennebaker says deception appears to carry three primary written markers:

• Fewer first-person pronouns. Liars avoid statements of ownership, distance themselves from their stories and avoid taking responsibility for their behavior, he says.
• More negative emotion words, such as hate, worthless and sad. Liars, notes Pennebaker, are generally more anxious and sometimes feel guilty.
• Fewer exclusionary words, such as except, but or nor--words that indicate that writers distinguish what they did from what they did not do. Liars seem to have a problem with this complexity, and it shows in their writing.

“The LIWC software--published by Lawrence Erlbaum--has been significantly more effective than human judges in correctly identifying deceptive or truthful writing samples, with an average accuracy rate of 67 percent [wrong one time in three] as opposed to 52 percent [50% means chance!!].”

End notes

1. There is a great difficulty in making people in Western culture understand clearly and properly the difference between an individual instance and a statistical measure or generalisation. This confusion leads to layer upon layer of error. For more discussion see intelligence’: misuse and abuse of statistics and cause, chance and Bayesian statistics.

   This problem is discussed in some detail by one of the better writers in psychology, David P. Myers, in his book,

   Intuition, its powers and perils
   Yale UP, 2002, 0300095317, hbk.
   $17.47 [amazon.com] {advert} / £14.10 [amazon.co.uk] {advert}

Like the book by Ekman, discussed above, this book also is hardly written with the precision and clarity I would wish, but it gets by and, for most people, will be worth the time and money expended.
Intuition maybe described as ‘instinctive’ reaction, founded upon your accumulated life-experiences and your genetic drives. I do not believe that this is spelt out clearly by the author.

email email_abelard [at] abelard.org

© abelard, 2003, 28 february

all rights reserved

the address for this document is https://www.abelard.org/briefings/lie_detection.htm

1460 words
prints as 4 A4 pages (on my printer and set-up)