Neuroskepticism - Neurotechnology in National Security and Defense - page 185

reinforced by the “lie detector” monicker. Documents obtained by M. Gregg Bloche
and me pursuant to Freedom of Information Act (FOIA) requests reveal that between
August 2004 and October 2006, the U.S. Air Force Polygraph Program conducted
768 polygraphs in Iraq. 2 According to an internal summary, 47% of the polygraph
tests indicated no deception, while 46% purported to indicate some form of deception.
This was interpreted by the drafter of the summary in the following way: “Detainee
personnel are just as likely to have committed the suspected act as not.” (Of course,
this might equally have been interpreted to mean: “Detainee personnel are just as
likely not to have committed the suspected act.”) But further reading suggests that the
attribution of guilt—defined as “involvement in multiple acts of anti-coalition force
activities”—on the basis of these results requires an unjustifiable leap of faith. The
summary itself offers one important reason why the report's conclusion is unwar-
ranted. It states (without acknowledging the implications of this) that “only 10% of
requests for polygraph support contain sufficiently detailed information for specific
issue exams.” The remainder of the polygraph tests were (according to the summary)
“by definition screening examinations wherein the examiner is called to resolve
numerous and divergent issues based on extremely generic, anonymous and per-
ishable reporting.” In polygraphers' feedback forms accompanying this summary,
many respondents complained that the polygraph technology was either “over uti-
lized or not utilized properly.” Two described the use of a failed polygraph test as
“a hammer to be used against the detainee.” One said this never resulted in anything
positive, while another said that, having participated in 240 polygraph examinations
in Iraq, on only one occasion did he witness this approach produce “anything of
value.” Despite this, detainees were “regularly” hammered with polygraph results—
even when deception was not indicated (i.e., even when they had “passed” the poly-
graph test). In such cases, the only clear evidence of dishonesty was on the part of
the interrogators.
Not surprisingly, many polygraphers complained about the way their services had
been deployed. One noted that interrogators “did not fully understand how to use
our services despite multiple briefings and pretest coordination discussions.” The
polygraph was often used as a “crutch” to avoid unnecessary interrogations, one
polygrapher claimed. Another complained about the use of what s/he considered to
be worthless questions and estimated that in 70% of cases interrogators asked: “Have
you ever been involved in attacking coalition forces?” Others described larger issues
that the military failed to address. Most notably, one concluded:
'I encountered nothing but difficulties with the exams and have no reason to have any
confidence the results were valid. I attribute these problems to a host of reasons: bad
environment, problems with interpreters [who were used in most interrogations], and
cultural differences.'
Even in its traditional use in the United States, it is clear that the polygraph does not
merit the moniker, “lie detector.” The NRC (2003) has noted that while the technol-
ogy performs better than chance, it is far from perfect. But in the national security
context, where most interrogations are mediated through an interpreter and cultural
issues are often ignored, the label is surely even more problematic.