It’s all in the voice

Sitting with Steven Ulrich in the big front office of Sunrise Investigative Services, I get the unshakable feeling that somewhere a hidden tape recorder is capturing all the modulations in my voice. Ulrich is Sunrise’s lead investigator, and he knows he has that effect on people. He uses voice analysis techniques to determine whether people are telling the truth. He even uses it to his advantage sometimes, throwing a wrench into casual conversation by suggesting that he’s taped you for later analysis with the help of his lie-detecting software—but that’s just a little bit of investigator humor.

Let’s start with an explanation of how voice stress analysis can detect deception.

What the voice stress basically does is it measures variances in the micro-tremor muscle located in your larynx. When you speak a word, your subconscious is telling you “yes, that’s the truth. Go ahead and say that word.” And as you utter the word, it’s picked up by, in this case, a microphone hooked up to a computer. And it’s analyzed and run through at 4,096 cycles per second to digitize it. And then that is produced on a paper graph and eventually printed out. From there, you can discern if there’s undue stress in an answer to a specific question.

How do you test it for accuracy?

Well, that’s the difficult part. The person being tested has to have some type of jeopardy involved in the situation. They have to have something to lose. So, if I were to test you right now for instance. I ask you a question: “Are you a dog?” and you say “Yes,” an obvious lie. There’s a good chance there won’t be any stress showing up in that because you’re not going to lose anything by lying. If you’re going to lie about something, you’re going to lie because you’re trying to avoid something. Your subconscious knows you’re lying, your conscience is saying the lie for you, and this is where we can pick up that the subconscious is stressing the body out, saying “Say this lie. Hopefully, they’ll believe it.”

Are there people who are immune?

I believe so. I haven’t met any. I’ve been told that somebody who’s very sociopathic could get away with lying here. For instance, there was one individual who beat the system who was a child molester. He was asked if he would inappropriately touch any more small children upon his release from prison, and he said no. And it came out with no stress. He got out and immediately began molesting children. They could not understand why. This particular voice stress analyst went back into the prison and talked to the individual and asked him, “Why did you say no, that you wouldn’t inappropriately touch them?” This individual replied, “It was not inappropriate. I love children, and that’s how I show my love.”

How was voice stress developed as a technique?

Back in the ‘70s, they were having problems overseas, mostly in the Columbian area, South America, trying to ascertain if somebody, as an informant, is giving the correct information about illegal activities and the drug cartels. They didn’t have the ability to hook them up to the polygraph. It’s very intrusive. They have to hook up your fingers and put connections all over your body to measure your heart rate, your sweat level and all that. They didn’t have that always available. So this was developed through the Army intelligence and the CIA. The technology was originally brought up through them. And they found that they could, by analyzing recorded telephone conversations, measure the stress level and find out whether or not they’re actually overstressed on certain words when they’re lying.

Who’s the most interesting person you’ve ever interviewed?

I haven’t done anybody, with their knowledge, who has an esteemed position, I guess. I’ve done it with half a dozen individuals, and they were over on the criminal defense end, and they were juveniles. Some of them were lying or showing deception. Basically, they were the gang members who didn’t want to rat somebody out. They’re kind of interesting.

The most interesting one was the one I did on Gary Condit and his interview with Connie Chung. It was very interesting because everybody thought he was going to admit something. There are people all over the United States, I’ve now learned, that have done the voice stress on the same conversation. I’ve not heard anybody say he had deception on the pertinent questions. And those were the results I came up with also.

When we spoke a little bit earlier, you mentioned some other celebrities that you had run tests on. Tell me what you found out about them.

I did this in conjunction with somebody else who had done it earlier. We did it as a test program in the classroom. They had previously done O.J. Simpson. Lot of deception in his answers.

What answers were most deceptive? Do you remember?

I don’t recall. He wasn’t interviewed very often. And of course, he didn’t testify so they didn’t have anything in court that he had said, so these were mostly questions that were thrown at him from reporters about, “Did you kill Nicole?” And he would answer a lot of times using evasive words. But he’d eventually get around to it, saying, “I love Nicole. No.” So we’d pick up on just the “no” portion of it and see a lot of stress in there. And some of these showed 45 percent or 50 percent stress on, “I love Nicole.” There was a certain amount of stress even pushing those words out.

They did the Jon Benet Ramsey interviews with the parents, the ones that were available to the public, and they found a lot of stress in there. There’s deception involved in their answers. It leads to the conclusion, at least with the voice stress analysts, that they had some type of involvement with it at some point. It’s really hard to say. All we can say is that there was some deception.