The "lie-detector" test has a storied history in law enforcement, as well as a dubious place in American courts.
The idea of using physical response to determine whether a person is lying is nothing new. While the first polygraph machine was invented in 1921, the practice of listening to a defendant's heart beat during testimony dates back to ancient times.
In 1923, shortly after the invention of the "modern" polygraph, which analyzes physical responses (heart rate, blood pressure and breathing), the United States Supreme Court ruled in Frye v. United States that this new "science" would be admissible in court only if and when it was "sufficiently established to have gained general acceptance."
Well, it's been almost 100 years, and it has failed to gain that acceptance. The accuracy rate in determining truth or deception by polygraph is as low as 70 percent, according to some critics. In fact, a person can purportedly train himself or herself to beat the test by controlling certain physical responses.
Do an internet search and see for yourself.
Recently, the Oklahoma owner of Polygraph.com was convicted and sentenced to two years in federal prison after teaching people how to beat a lie detector. It is not against the law to learn how to beat a polygraph, yet Douglas Williams, a former Oklahoma City police officer, was convicted of two counts of mail fraud and three counts of witness tampering. Apparently, he knew his strategies would be used in federal matters to deceive the United States government.
Regardless, there are institutions like the American Polygraph Association (APA) which advocate for the use and accuracy of polygraph evidence. According to its aggregate data meta-analysis, the APA reported that a review of all results "produced a decision accuracy of 87% (confidence interval 80%-94%) with an inconclusive rate of 13%." However, the APA also conceded that the only studies included in the meta-analysis were those which met the requirements of the APA Standards of Practice.
Now don't get me wrong: I firmly believe that polygraph results have a place and a purpose when it comes to negotiations behind closed doors. When a seasoned prosecutor and an experienced criminal defense lawyer can sit down and discuss the mitigating (or damning) findings of a polygraph test administered during the investigation, there is a fair chance that the test result may have some serious bearing on the direction of negotiations.
But to put the results in front of a jury? No thank you.
Possibly the largest problems with the polygraph, and why it has been inadmissible in most U.S. courts, is the notion that jurors give the test too much weight, particularly in light of its noted inaccuracies. In some states, the mere mention of a polygraph test to the jury, especially without a prior stipulation, can result in mistrial or even sanctions against an attorney.
A recent ruling by the United States 10th Circuit Court, however, might lead to sweeping changes in those jurisdictions.
In United States v. Tenorio, a man convicted of sexually assaulting his 16-year-old niece claimed that his confession was coerced, and he was innocent of the charges. Tenorio alleged that when an investigator offered a "lie-detector" test, he willingly took it, hoping the test would clear his name. Instead, the investigator intensified her investigation with "confrontational questioning," accusing Tenorio of lying and telling him to "man up."
Tenorio ultimately confessed, but later argued it was only because he was distraught and did not understand why the agent was "coming at [him]" and didn't believe him. In bringing up coercion as his defense, prosecutors argued Tenorio opened the door for mention of a polygraph test he took voluntarily.
The District Court cautioned prosecutors not to bring up the actual results of the test -- only to indicate that the defendant took a polygraph test as part of the attempt to disprove coercion. The judge furthermore cautioned the jury:
Not to speculate or take into consideration anything regarding the polygraph examination or its potential results in reference to the guilt or innocence of the defendant or in reference to whether or not he did or did not commit the acts charged in the indictment.
Such an admonition shows exactly why the polygraph should not even be mentioned to the jury. The judge had to order the jurors not to take into account what every normal person would logically ask when put on notice that a test was administered: "Did he pass or not?" Taken one step further, and the inference from the lack of an answer becomes: "Well, he must have failed the innocence test if they are still arguing about this in court."
After his conviction, Tenorio's appeal was denied when the 10th Circuit ruled the test was not being used as scientific evidence of guilt or innocence, but rather as evidence to confirm or deny the alleged coercion.
And herein lies the slippery slope. The mere mention of polygraph tests in the presence of the jury begins to shift the burden onto the criminal defendant. The notion that there is a test which can prove if someone is lying puts the obligation on any innocent defendant to take, and pass, such test. If the test exists, and its being talked about in court, the failure to present the results leaves the impression that the results aren't good.
If and when the practice of discussing polygraph exams in jury trials becomes the norm, criminal defense attorneys might as well kiss their profession goodbye. While many folks might jump for joy at the idea, one has to wonder what would remain of the prosecution's burden to prove a criminal defendant guilty "beyond a reasonable doubt." That burden is the cornerstone of the United States criminal justice system and it adds an undeniably human element to the cold and calculated court proceedings. Consequently, humans are necessary to help explain and argue the facts upon which the reasonable doubt is established.
The idea of citizens contemplating "reasonable" doubts as to the guilt of the accused implies that they are allowed to use reason, which seems to be distinctively human in nature, once we consider a certain level of cognition (a dog can reason its way out of a hole under the fence but not a pair of pants). This "human element" is exactly why we are allotted more than one juror to hear our case, and why any verdict returned must be a unanimous decision (or at least a majority).
Allowing a machine, or set of tests, to establish or influence one's guilt takes the "reason" out of reasonable doubt.