Comment Theory Is Nice, Scientific Validation Ain't There (Score 1) 324
Agreed that it would be handy if such a thing as a "lie detector" existed. But wishing doesn't make it so. The technology may have changed a bit since the 16th century --polygraphs and MRI scans vs. throwing the victim into the water to see if (s)he floats-- but the lack of any credible double-blind studies to test any of this stuff remains the same.
The dilemma is that it's not possible to conduct a double blind-study that replicates the only conditions where a "lie detector" would have any utility. That is:
[a] Subject has a lot to lose by being detected as lying. The hypothesis that a person has the same pyschological reaction to answering "Is the year 2006?" as to answering "Did you murder John Doe?" seems at best unproven and at worst preposterous.
[b] Subject knows that Tester cannot independently verify what the truth of the matter is. By definition, the more reliable the independent evidence as to what actually happened, the less need there is for a "lie detector".
Thus, the only questions where "lie detector" effectiveness can be tested are so trivial and unrepresentative that test results demonstrate nothing about how the technique might work in actual operation.
Further, the hypothesis that all output from the "memory" area of the brain is "true" is contradicted by many published studies about memory.
Believing that "lie detectors" work as advertised seems as irrational as believing that torture produces the "truth" (as opposed to whatever the subject thinks his interrogator wants to hear).
The dilemma is that it's not possible to conduct a double blind-study that replicates the only conditions where a "lie detector" would have any utility. That is:
[a] Subject has a lot to lose by being detected as lying. The hypothesis that a person has the same pyschological reaction to answering "Is the year 2006?" as to answering "Did you murder John Doe?" seems at best unproven and at worst preposterous.
[b] Subject knows that Tester cannot independently verify what the truth of the matter is. By definition, the more reliable the independent evidence as to what actually happened, the less need there is for a "lie detector".
Thus, the only questions where "lie detector" effectiveness can be tested are so trivial and unrepresentative that test results demonstrate nothing about how the technique might work in actual operation.
Further, the hypothesis that all output from the "memory" area of the brain is "true" is contradicted by many published studies about memory.
Believing that "lie detectors" work as advertised seems as irrational as believing that torture produces the "truth" (as opposed to whatever the subject thinks his interrogator wants to hear).
