Truth sought with, and about, voice test Device checks speech for signs of lying, but method has critics

December 25, 1998|By Eric Siegel | Eric Siegel,SUN STAFF

From the beginning, Baltimore homicide detectives had doubts about Richard Allen Nicolas' story of an unidentified man firing a bullet into his car and killing his 2-year-old daughter, Aja. One clue was that his muffler was cold, indicating the car had been parked near a Northeast industrial park a lot longer than he said it had.

But it was not until a computer voice stress analyzer -- a device purported to detect whether a person is telling the truth by measuring tiny speech tremors -- indicated Nicolas was lying that detectives focused on him.

"We had the suspicion, but with the test, we had more conclusiveness," said Detective Raynard Jones. "Then we had to go out and put everything together."

Last year, after being presented with evidence police developed, a Circuit Court jury found Nicolas guilty of first-degree murder. He is serving life in prison without parole.

Similar cases are making believers in computer voice stress analyzers among a growing number of police agencies around the country. The machines, relatively cheap and simple to use, are supplementing and, in some cases supplanting, the more well-established polygraph as a lie-detection device in investigations.

But with its increased use, controversy over computer voice stress analyzers is growing, with critics complaining that the evidence supporting them is anecdotal, not scientific.

Legal challenges have been raised. A man who had sexual assault charges thrown out of court in Nevada has sued police and the analyzer's manufacturer in Florida.

Baltimore's police union challenged the device's use in an internal affairs investigation.

Many point to scientific studies conducted in the mid-1990s by the U.S. Department of Defense Polygraph Institute, which trains all federal polygraphers and researches technologies for detecting deception at Fort McClellan, Ala. The institute's conclusion: "We have found no credible evidence that voice stress analysis is an effective investigative tool for determining deception."

Several states, including Texas and Virginia, ban the independent use of voice stress analyzers.

"There's nothing on [the computer voice stress analyzer] to prove it works," said Frank DiTucci, executive director of the Texas Polygraph Examiners Board.

'Does not have validity'

Lie-detector operators are also stepping up criticism.

"It's convenient. It's cheap. But it does not have validity," said Donald A. Weinstein, president of the American Polygraph Association.

L Some members of the criminal defense bar also are concerned.

"That's why the thing becomes pernicious -- it can be used so easily," said Domenic R. Iamele, president of the Maryland Criminal Defense Attorneys Association. "Even if a suspect doesn't confess, he can destroy his defense."

Other defense lawyers are less critical of voice stress analyzers, provided the results continue to be inadmissible in court.

"As long as it is used as a law enforcement technique and not a prosecution technique, I have no problem with it," said Howard L. Cardin, one of Nicolas' lawyers.

Arguments over their merits aside, the polygraph and the computer voice stress analyzer operate on the same principle: that subtle physiological changes in the body are triggered by the stress of deception.

Developed in the 1920s, the polygraph measures several of these physiological changes, including heart rate, breathing and perspiration.

In the early 1970s, a device was developed to measure inaudible changes in voice vibrations caused by stress. The computer voice stress analyzer -- about the size of a laptop computer and sold by the National Institute for Truth Verification in West Palm Beach, Fla. -- is a modern-day version of that machine. Its $7,500 cost is less than half that of a polygraph machine, and it takes only a week of training compared with two to three months for a polygraph.

Some proponents of polygraphs acknowledge that the voice stress analyzer theoretically has several advantages, chief among them that the subject wears a clip-on microphone.

Criticism of voice stress analyzers hasn't halted growth in their use.

In the past three years, the number of law enforcement agencies nationwide using the device has doubled to about 600, said David A. Hughes, executive director of the company that makes the machines. "Police departments are finding out voice stress works. They don't need a study," he said.

In the 3 1/2 years since the Baltimore Police Department got its first machine, nine Maryland agencies have bought computer voice stress analyzers, Hughes said. Among them: Baltimore County and Montgomery County police and the Harford and Washington county sheriff's departments. State police are reviewing whether to use them.

To agencies that have used voice stress tests in hundreds of cases, whether the analyzer is as reliable as the polygraph is academic -- and irrelevant.

Baltimore Sun Articles
Please note the green-lined linked article text has been applied commercially without any involvement from our newsroom editors, reporters or any other editorial staff.