The Lie Machine

We humans value our secrets as highly as we do our gold watches and money clips, and just as muggers wait in parks to rifle our pockets for things to pawn, spies make a living stealing and trading those secrets. They've become very damn good at it by now. They can get into our businesses and houses (it's so easy for them) or onto our telephones or even into our martinis. They can use the light of the stars to photograph us where it's too dark for an ordinary thief to see, and about two years ago they began jimmying the locks around the deep and dark parts of a man's mind where the real secrets are: lies, hatreds, jealousies, fear, love, the hidden chemistry of human behavior---which is really what they were after all the time and which they once weren't able to photograph or record or bribe out into the light. Even starlight.

Now they've invented a lie detector that knows by the tremor in your voice when you are lying. They call it the Psychological Stress Evaluator (PSE), it's designed to fit into a Samsonite briefcase, it costs about $4000, anybody can buy one, it takes three days to learn to run it, you can use it surreptitiously (all you need is a tape recording of the subject's voice---get that any way you can) and it works.

It's Pandora's youngest kid, and I would run away from such machines, but you can't run that far anymore. I said something like that over the phone to Allan Bell, one of the former spies who invented the machine, and he said, "We've talked it over"---he has a big orator's voice, and by we he meant Charles McQuiston and Bill Ford, his coinventors, also spies---"and we'll be happy to teach you to run the instrument. We were hesitant at first about training a journalist, because you need a pretty good background in interrogation techniques. But we decided that if a reporter's any good, he gets into that sort of thing all the time."

I was agreeing with him and we talked about small things and then he said, "Besides, we have nothing to hide." I stopped agreeing, because I never knew anyone like that, and then he said something holy: "We're not trying to catch all the liars in the world, we're just trying to make everybody tell the truth."

The what? "The truth." It sounded so strange. A little like "Progress is our most important product" or some other piece of high-minded corporate nonsense; but if you had held the phone away from your ear when he said it, and closed your eyes, it could have been some kind of reveille prayer coming over the loud-speaker at a sensitivity ranch in Big Sur. Spies and gurus, muggers and archbishops; I will never be used to the kinky bedfellows this age is producing. But if their promises are good enough, I don't care. "The Truth," I almost said, "keep her talking till I get there, 'cause I been scuttling after that old whore ever since I found out she hung around with Beauty and could---they promised---set me free."

• • •

Dektor Counterintelligence and Security is on Port Royal Road in Springfield, Virginia, about 20 miles from Washington, in a dumpy light-industrial plaza with an asphalt parking lot around it; two stories, glass-and-steel panels, square and gray in feeling, as if it had come off the architect's drawing board without gaining a third dimension. Dektor has part of the building, a 7-11 Foodstore headquarters has another part, and the two of them share the parking lot with an unnamed Air Force project across the way. The three establishments are at quiet war over parking spaces (they block one another's cars and nick fenders) and it is bizarre to think what the supermart people and even the Air Force might learn about the current state of the spook arts if the dispute ever heated up. For Dektor is a complete spy shop. Besides the PSE, they design and make telephone bugs and anti-bugs. You can hire them to sweep a room clean of bugs, or for $1000 they will teach you what they know about surreptitious entry, which is lock picking. Or you can hire their whole act and all its electronic baggage as consultants (which could mean anything) to guarantee your parking spot secure against the greedy traffic of the times.

My six classmates and I met in the parking lot on the morning the course was to begin. I didn't know much about the group, but I'd watched them eat breakfast (ham and eggs, coffee and cigarettes) at a round table across the coffee shop of the Charterhouse: two private eyes from Los Angeles, another from Denver, a police lieutenant from New Jersey, the head of security for a Midwest food-store chain and a shrink. Gatherers of men's secrets, all of them, and I had decided, as I sat at my own table eavesdropping on the eavesdroppers, spying on the spies, postponing my own introduction, that a journalist isn't a bad fit in such a group even if he wants to be. Many kinds of people believe that the proper study of mankind is man, that the truth rots when it is hidden, that secrets strain to be told. Reporters use notebooks and sit at tables that have good earshot, and they watch people's eyes and listen to their voices to tell the truth from the half-truth and the lie. These people, laughing now at the wa-wa alarm we had tripped going through Dektor's front door, simply believe in using no-nonsense machines to get the job done.

Inside, Bell greeted us, shook hands, took introductions and said, "Well, a little," when one of the group asked if he ever got tired of being told that he looks exactly like Fred Astaire. He does, and after he had offered us all a cup of bad office-urn coffee, he took us to a long folding table that had been set up behind some partitions in the middle of the shop. It had a slide screen at one end and a carrousel projector at the other. Bell said that McQuiston ("Mac" to everybody) would be out to start the class in a minute, and he was: a manatee of a man, with a chest-up military strut, in a white technician's coat and a crewcut that looked fresh that morning. "Let's get this show on the road," he said and then began pacing as he explained the general outline of the course, said it wasn't a picnic or a party and that it was going to take our best efforts to learn to use this new tool effectively and with confidence. Almost everything he says is wrapped in military catch phrases and clichés. He's a lieutenant colonel who worked as a spy for the Army for 26 years on every continent except Africa; and though he has been retired for over a year now, there are American eagles in his voice, his humor is Army rough, doesn't wait for the laugh and hides when there are women in the room.

He is a graduate of the Fort Gordon, Georgia, School of Polygraphy (perhaps the best lie-detector school anywhere) and he wears his diploma in the form of a little round patch sewn to his lab coat. The badge has the word Veritas above a seismographic squiggle and the word poly below. Men who run lie machines have always been scarce (no one really knows how many there are; estimates range from 1500 to 2000) and their clients fly them all over the world to pop liars out of groups of honest men; they think of what they do as an art. For no lie detector yet invented can work without the skillful interpretation of a human operator.

"The instrument," McQuiston said, still pacing, "whether it's a polygraph or a PSE, is only a tool. You, my fine feathered friends, are the lie detector, and unless you know how to interrogate, brother, you are going to have a very expensive piece of useless equipment when you walk out of here." He is now using a pointer that looks like a car antenna. "And you are going to have to let your ego go to do it. You're gonna have to become an actor par excellence, a surgical tool, if you're gonna crack the subject's shell. You're gonna have to remain dominant every minute you're with this guy ... hold him at bay with a panatela ... manipulate him ... tell him you're gonna be his friend right up to the moment he tells you his first lie ... and you tell him that when he does lie ... Gang Busters ... he's on his own ... the instrument is gonna nail him."

He is telling the class about the technique and he is demonstrating it. He's got the pointer instead of a panatela, he's prowling the front of the room and his eyes and his bluster are going after the people around the table, one at a time. Sometimes his voice is almost quiet and sometimes he raises it to a rubber-hose pitch that has one of the private detectives chain-smoking into a big glass ashtray in the middle of the table.

"And when that cat walks into the room, the vibes start from the first handshake ... and you have to remain in control ... you have to get inside his hang-ups ... inside his set. He may be the dirtiest, slimiest bastard you ever saw and you're still gonna have to get inside his moral code and maneuver him until you've created a threat so effective, so particular, that when the instrument charts it, you'll be able to say it isn't fear, it isn't hate, it isn't love, it isn't sex, it is a lie."

Now his manner rests and he tells a classic polygraph anecdote: Examiner and subject together in a room with a polygraph machine and the examiner says (in a fatherly tone): "You know, this damn machine is superhuman ... I mean it ... I can't always tell a lie from the truth, you can fool me ... but this thing is so sensitive ... these three pens here, see? ... they're uncanny: You absolutely can't fool them." Then the examiner leaves the room and watches through a one-way mirror while the subject sneaks to the machine and bends all three pens off register and then sits down quickly. Examiner, back in the room now, goes straight to the machine, puts his ear close to the pens, lets a horrified look take his face and then says, "He did what to you?"

"And that guy," McQuiston said, "spills his guts right there, I mean a perfect confession."

The cop from New Jersey asks a question about the accuracy of the polygraph. "Well, as you know, there are many problems with the standard polygraph," comes the answer. He has taken this as his cue to begin an attack on the old method of lie detection, an attack that will go on for the whole three days, in the spirit of an ex-priest knocking the ceremonies of his used-to-be church. "First of all, you get the guy hooked up like he was four mules ready to pull---he's got that big bellows around his chest, the blood-pressure cuff on his arm, finger caps that are wired and make him think he's going to be electrocuted and---Has everybody here seen a polygraph work?"

Everyone had and they all raised their hands except me. "Well," he said, "I'd like to take our friend from Chicago here and tighten him down with a polygraph till his arm turns blue and all the capillaries are about to burst ... then he'd know." (By the look on his face, I was sure I would.)

"That's the reason we get so many inconclusives off poly," he said, "because when you get a guy wired and tied like that, any little thing---a car backfiring outside, a telephone ringing in the next room---anything can blow the whole test. Altogether, I'd say that the polygraph is the most effective piece of 19th Century equipment still in use. And that's why we invented the PSE: We think it's a much easier and more effective way to chart stress in the human body, and that, in lie detection, is the name of the game."

Turns out it always has been, even in most of the old folk methods of lie detection, because a good witch doctor never needed a machine or an education in human physiology to notice that a man under stress goes through a lot of bodily changes he has no control over. He breathes more rapidly, his blood races, his mouth goes dry, he becomes pale, he fidgets, his eyes become unsteady and something small seems to happen to his voice.

The Greeks were aware that a man's pulse quickens when he lies and certain physicians used this knowledge, according to reports, mostly to detect infidelity in wives of prominent citizens. The Chinese and later the Spanish at the Inquisition had the suspect chew a mouthful of rice or dry bread: If it went down easily, he was said to be telling the truth, but if his dry mouth made the rice stick or go down hard, he was lying. The Arabs used a hot dagger drawn across the tongue: The moisture in an innocent man's mouth was (continued on page 102)Lie Machine(continued from page 94) expected to save him, and for the guilty a badly burned tongue was the first of his punishments. In Africa, witch doctors literally smelled suspects, and in another kind of ordeal, all tribe members were made to plunge their arms into a pot of cold water, then into a boiling pot. Whoever had blisters on his arms the next day was called guilty. And there were other methods, all loosely based on some psychophysiological reaction, and each as unreliable as the next. The object of most of them was to obtain a confession, by provoking first fear and then perhaps hysteria. And, in fact, the modern polygraph and even the PSE are used mostly in the hope that the guilty man will confess when faced with a machine he believes is going to catch him.

The modern polygraph measures cardiovascular response through a cuff that can be used on the arm or the leg, respiration by a bellows that goes around the chest and galvanic skin change (electrical resistance that changes with sweating) by means of finger clips. Over the last part of the 19th Century, several researchers began finding relationships between physiological changes and fear. Before 1895, a man named Lambroso began working with blood-pressure and pulse measurements to detect deception. Then, in 1927, Leonard Keeler put together the components to measure pulse, blood pressure and respiration and his machine was the prototype for the modern polygraph. Since his time there has been little significant change in the theory or design of the machine. It has been miniaturized and in some cases they have added two pens (to measure fidgeting and head movement), but the basic principles and drawbacks have remained the same for 45 years: The subject must sit perfectly still, he is in pain, he fears electrocution, the smallest distraction can destroy the results and heart patients, young children, drunks and people taking some drugs cannot usually be tested. And, of course, like most of the folk methods that went before it, the standard polygraph test cannot be administered without the knowledge and consent of the subject.

• • •

After the first morning session, Allan Bell sat in his office and told me, "When you have a problem, you have to put all your preconceptions away and then think to yourself, 'What is the simplest way to solve it?' " He doesn't gesture much when he talks, this 20-year spy, but puts the finger tips of his right hand against the finger tips of his left, as if he were completing some bodily circuit that helps him think like a knife. There are deep lines in his face from the years of running in a world of high-caliber handguns and cover stories that have to be held onto through everything as if they were his real and only history. There is a college of secrets in him and you can almost watch the muscles that keep them inside at work when he talks: Even when they steady each other, his hands shake.

"Essentially, the problem as we saw it was how to measure the stress a man is under without having to hook him up to do it. Once we narrowed it down, the simplest solutions seemed to be either body smell or voice, and voice seemed more promising."

First they offered their idea to the Army (they were all on active duty then), but the military mentality and the arrogance of a few superior officers turned them down, they say. So they left the Army and began working in the basement and family room of Bell's house: McQuiston the polygraph veteran, Ford (just back from spook duties in Vietnam) the electronics designer, Bell the master problem solver and Bell's wife, Clare, to keep the answer from pertaining to males only.

They worked together, looking first for stress reactions in the voice, then isolating them, then building components that would make them graphic, finally looking for lies in the sharp up and-down lines their machine was making when a tape-recorded voice was fed into it.

When they began to recognize promising patterns, they started getting together in front of the TV program To Tell the Truth. The three contestants would appear on the screen, each would say, "My name is -----," and the spies would tape it. Then they would run the tape recording of that first statement only through their prototype PSE and before the panel could do it, they would call the liars. Out of 75 experiments, they say they got it right 71 times. That's almost 95 percent (a phenomenal average for any kind of lie-detection experiment) and it was around this time that they became a company.

• • •

Back in class, McQuiston began using a remote-control switch to flash cutaway drawings of the human body onto the screen. "The physiology of this thing is important, so pay attention," he said, and then for two hours he explained in painful detail and with many charts all the things that happen to a body under stress. He talked about the central nervous system and the autonomic nervous system in both its sympathetic and its parasympathetic divisions and finally, for a very confused tableful of students, he zeroed in on the tiny part of the system that they think the PSE uses to identify stress. (They are not sure.) And it is relatively simple.

The timbre, or total quality of the human voice, is made up of a base frequency of from 50 to 350 cycles per second and something called formant frequency, which varies with the size of a man's mouth, sinus and nose cavities, length of vocal cords, etc. Also in the voice (as well as other parts of the body) is a phenomenon called micromuscle tremor, which creates an FM frequency wave that is present in the normal voice. This FM signal is controlled by the central nervous system, and when you put a man under stress, his autonomic nervous system overrides the central system and erases micromuscle tremor and its wave. The PSE reacts to the absence of the m.m.t. wave, producing a blocked pattern that can be read as stress.

"When a wave line that starts out looking like mountains and valleys turns into something resembling railroad ties---brother, look out," McQuiston told us. "This cat is under some kind of heavy stress."

Then he said he could run a monkey on PSE, that he'd already run a parrot named Speedy ("That damn bird couldn't stand me---he blocked up every time I walked into the room"). He told us he'd run a deaf-mute because he could grunt and that he hoped to work with dolphins and whales someday. "They all have micromuscle tremor," he said, "and when they are under stress, it disappears and the PSE tells on them. The human subject, by the way, can be drunk, or walking around the room, or on heroin---we had one of those---or have a bad cold; once we ran a woman who was nursing her baby through the test. And none of it makes any significant difference."

"Do you know a way to beat the machine?" I asked him.

He came back like a drill sergeant. "Don't call it a machine. It's an instrument and there is a difference and I want you to get it straight."

We did a little word wrestling that ended only because I finally said, "I give---instrument. Do you know countermeasures that can be employed to fool the instrument?"

"Yes," he said.

"Do you think you could spot somebody who was using them?"

"Oh, yeah," he said.

"What are the countermeasures?"

"Never mind," he said.

We spent the rest of the afternoon in pairs, in little rooms, learning which buttons to push on the PSE and its companion tape recorder. We ran a training tape and then compared our squiggles with those in a paper-bound manual we had been given.

"Don't worry too much if your patterns don't look exactly like the ones (continued on page 164)Lie Machine(Continued from page 102) in the book," McQuiston told us. "Today I just want you guys to build your manual dexterity.... I want you to be able to play that thing like a piano."

None of us felt much like virtuosos when we gathered around the table about two hours later. But we kept getting assurances from Mac that it would all come to us in time and that if we understood the. basic moves, that was enough for now.

It had been a long day full of bad coffee and excruciating science and I had the kind of headache that is really just the brain feeling like an out-of-shape and overworked muscle.

"OK," he said, "tomorrow we'll learn how to spot a lie out of all the other kinds of stress a man can be under. Go on back to your hotel and relax ... but be here at nine hundred sharp tomorrow morning."

• • •

It's hard not to play the liberal at a tableful of cops who are relaxing with drinks and trying to get to know one another by trading on-the-job nightmares. Two drinks before dinner, at a big round table in the Charterhouse again, then dinner, then two drinks after dinner and no one got drunk. The stories they told---about brutal killings, child murders, massacres, police ambushes, the way dead bodies look when you don't find them right away---were all designed to rattle and then convert tender sensibilities that care too much about head trips like the rights of privacy or presumed innocence. For every intellectual notion about law and order that I pleaded, they would tell me something more about the way bullets go into flesh, or about the kind of insane animals that men sometimes become. And I would say yes, I know (but they kept saying I didn't), and when I felt enough like Aristotle at Attila's table, I shut up. There is no such thing as the "police mentality," I decided, only the police experience, which is perverse, and reinforced for them every day, and which is almost unrelated to anything that happens in the cool world of judicial minds.

When we talked about the PSE, there was excitement at the table---a sense that another safe harbor had been mined against the flight of guilty men. Some of them even talked about clever criminals they had known who might have been trapped by such a weapon. But the thing that seemed to excite each of them most was that they were learning a skill whose power could get them and keep them out of the street fighting. Because that's success for a cop: a desk, a civilian suit and interrogation rooms, where there is no shooting and no dead bodies. Lie detection, the art, could keep a man from goon duties and make him a technician-priest in the fight of civilization against its wreckers. A polygraph operator can charge from $150 up for an examination and there are never enough of them to do the work. And though it takes from six weeks to six months to become a certified polygraph operator, it takes only three days to be trained on the PSE, and presumably you can use it to do the same job for the same price. Only 16 states have sorted through the complicated process of making laws on who is qualified to run a lie detector and who is not. The game is open in 34 states, and although Florida has begun to consider it, no other state has thought about how the law should apply to PSE.

The psychiatrist had been quiet through most of the evening. He was shy anyway and it was clear that his fix on the machine had nothing to do with guilt or innocence. But the PSE promises as much for him, or maybe more, as it does for the criminal catchers. An instrument that measures stress in even its smallest degrees is like an X-ray machine for a behavioral scientist, a finger with which to take a man's emotional pulse.

One psychiatrist who bought a machine uses it on himself when he is overtaken by the vague feeling of frustration and discontent that comes to all of us now and then. He makes a list of the 25 things that might be the cause of his funk and then asks himself about each one. He answers himself on tape and then runs it through the PSE. He says it helps.

There is an acting teacher in Southern California who uses the PSE on his students. Those who are comfortable in their parts show less stress than those who are not. He says acting is a form of lying, anyway. There's a drug company that has had one of its men trained on PSE so that he can use it to chart the effects of tranquilizers on people; and a doctor is working with the machine to log the long-term effects of emphysema. Dektor has trained businessmen who use the machine to test good faith in negotiating situations, personnel managers who need to catch job applicants who might lie about their arrest records or educational backgrounds. In the recent history of polygraphy, something like 80 percent of all tests have been given to employees at the request of industry. There is even a man who bought a PSE and used it on his wife for the same reasons he might have hired a private detective a few years ago. (He nailed her.)

But it is the catching of liars, the undoing of secrets that inspired this machine, and whatever other uses it will be put to will be discovered and explored by someone other than the Dektor people. They are businessmen now and anxious for all the potential PSE uses to be experimented with, but they are police first, military men, spies, in a world that has only two kinds of people---the guilty and the innocent---and what the people at my table that night liked best about the PSE was that it can tell the difference. When the evening ended, I went off to the rocky kind of sleep that lets only small parts of you rest.

• • •

The morning session was technical again: The five-o'clock headache of the day before would be there by noon. McQuiston was in his same galloping form, however, and he began using the slide projector and his telescopic pointer again, this time to teach us what to do with a man after you've got him pinned in the corner with your cigar.

He took us through all the simple and sophisticated methods of structuring a test that polygraphers have developed and practiced over the past few years: the peak-of-tension test, the general-question test, the zone-of-comparison test. "Remember," he told us in a moment of professional propriety, "we are not concerned with the guilt or innocence of a suspect, only in whether or not he seems to be lying. He's either D.I. or N.D.I.---deception indicated or no deception indicated. That's all you're there to judge ... got it?"

We learned (or at least we watched him try to teach us) how to set up a structured series of questions that would determine with reasonable surety how much of the subject's stress was general nervous tension, how much might be from worries that had nothing to do with the test, how much might be from guilty knowledge and how much would indicate direct deception. It's all a system of check-and-balance questions, really, things called relevants and semirelevants and outside-involvement questions all leading to a point of direct jeopardy ("Did you take the money?") where the stress reading would be high and clear if the man were D.I.

Then we looked at the paper tapes again and McQuiston tried to show us the subtleties, the echoes, the shadows of stress patterns that could, if we were not careful and savvy, make a liar look an honest man and an honest man a liar. That was spooky, because most of us were getting it dead wrong on the tricky ones, and almost all of them seemed pretty tricky, McQuiston kept telling us that we would get it and at one point the cop from New Jersey said, "I see what you mean about the lie detector being me." We all chuckled as if there weren't a man among us who had more than half the circuits for the job.

"Well, you're gonna make mistakes now," he told us. "I want you to make mistakes now. You're not going to have any confidence until you crack your first case, anyway. But it takes only one."

Someone new, in a suit, had taken a chair quietly at the end of the table and McQuiston had noticed him. It was Mike Kradz., a Howard County, Maryland, police lieutenant who had at that point run 150 actual cases on the PSE and was a true believer in its powers.

"Isn't that right, Mike?" McQuiston asked. "I can sit up here and tell you guys about cases I cracked, or Mike can say I did this or I did that, but you're gonna have to run your first case and bust it, and then, brother, your confidence level will be right up here." He had his hand about an inch above his crewcut.

When he dismissed the class for lunch, Kradz and I sat down to talk. Below a receding hairline, he has strong deepblue eyes that are nearly motionless when he talks to you. He has been a cop for 22 years, an experienced polygraph operator, an evidence technician, the guy who looks for Holmesian clues after the deed is done: One of the first things he told me was that he has the largest collection of shoe-heel prints in the world.

I asked him if he'd ever read Sherlock Holmes and he laughed. "Sure, I guess I did, I liked mysteries and it's true that nothing on God's earth is ever repeated exactly the same way. You can take anything you like---fingerprints, two cigarettes, two broken car aerials, handwriting, torn sheets, two PSE charts---and when you learn to work with this stuff, you'll find it's just like a neon sign pointing the way."

When Kradz heard about the PSE, he wrote to some of his colleagues in the American Polygraphers Association to ask them what they knew about it. Most of them didn't bother to answer and the ones who did said either that it was untested or that it was unlikely. For the most part, the A. P. A. has tried to ignore the PSE. When it hasn't been able to do that, it has disparaged the machine in ways that betray the established leader's fear of being unseated. And although there are signs of conciliation now (sooner or later, they will meet each other), the A. P. A. is worried, too, that, because the PSE can be used surreptitiously, it may bring the increasing outrage of Congress down on the whole lucrative lie-detection industry.

But Kradz saw it as a new tool that might help him as his other skills do in building chains of evidence that finally take a man to trial or let him free. And because the administrative and technical division of the Howard County Police Department (which he runs) has a limited budget, he bought the PSE with his own money and began experimenting with it.

"I wasn't looking for something to replace the polygraph," he told me in his very gentle voice. "I was looking for a companion that I could run with poly as a check, another evaluation. I ran the two machines side by side for 48 cases and I found that the PSE did exactly what Dektor said it would do. Plus, I found it easier to read and interpret. It put the subject more at ease and I could deviate from the old method of interrogation, ask more questions, allow the person to explain or rationalize if he wanted to. It loosened the whole process and gave me what I think is a more objective test. And I can use it on people who can't take a polygraph test because they are too fat or jittery or have back trouble that keeps them from sitting still for as long as it takes. There are other problems with poly, too, you know. I had a guy come to me and say, 'I know what you're going to tell me, because I've had these damn things before, and I don't want to take it.' I asked him why he was so shook up and he said, 'The last time I had one of these things, the guy put the cuff on my leg instead of my arm and when he got through with the test, he told me I lied. I said, "Show me where I lied," and he said, "Right here." I said, "What's that?" and he said, "It's off that cuff on your thigh," and I said, "You son of a bitch, I lost that leg in Korea, and if you can get a reading off that wooden thing, you're better than me." '

"And that's where the polygraph examiner got tripped up in his own black arts. That's the trouble with poly. They haven't been able to cut it scientifically, so they go in for this mumbo jumbo and cats' whiskers and that crap and sometimes they get caught. But with PSE, I let the subjects run the test. I tell them what we are looking for, and then when the tape comes off the instrument, I let them see it and then I ask them about it. The general rule with polygraph is that if you don't get a clear reaction, you go into hostile interrogation and see if that won't get something. I don't ever have to do that with the PSE. If an indication of deception is there, I just show it to them and then we talk about it. I'm very comfortable with it."

Kradz is convinced that he has had no failures among all the cases he has run and the one he likes best proved a man innocent, not guilty.

"It sounds corny," he told me, "but my job is not to determine guilt or innocence. I'm a detective and detectives are just supposed to turn in a report based on the evidence they've gathered, and they're supposed to make it as objective as possible. So I had this boy who had been in jail for four months, accused of armed robbery. He'd had several polygraph examinations, all inconclusive, and he was scared and looking for someone to help him. 'I was out walking the dog,' he says to me, 'it was a rainy night, nobody saw me, the dog can't talk.' There were three eyewitnesses who said they saw this guy do it, so he had a problem. Well, all of us have problems sometimes, so I told him, 'Look, if you turn up clean on this test, I don't care how far we have to go, all the way to the Supreme Court, anywhere, I'll stand up beside you.' Well, he took the test and it indicated he was innocent, and the judge and the prosecutor and, of course, the defense attorney stipulated that the PSE results could be admitted as evidence. They let him go, paid him for his time in jail, and it made me feel awfully good. I'm not a priest, but I get very close to these people."

Since then, Kradz has had three other PSE tests admitted into court to prove innocence. It has never been admitted to prove a man guilty.

"I think it's as valuable a tool as has ever come into the evidence technician's hands," he told me finally. "But I do understand why people are afraid of it. Just after I bought it, the officers who came into my office started writing notes instead of talking to me. Really. Even I've started to think before I talk. Used to be I'd say whatever, and then take my chances on poly if it ever came to that. But not with PSE. In unscrupulous hands it could be a frightening thing, but I think its value far outweighs its potential dangers."

Kradz had to leave to go back to work and after he left, I thought that if it is the man who is the lie detector, it is not surprising that so many people end up confessing to those eyes. Unlike most of the other interrogators I met, there is nothing of the bully about Kradz. Every man wears his shoe leather away at a different angle, according to the load he carries and how much he is made to shuffle. Kradz collects the heel prints because each is different and because the world is full of surprises even for a cop who has watched it through a microscope for 20 years.

Through the rest of that afternoon and the morning of the next day, we ran actual cases. McQuiston used his remote button to flash PSE tapes on the screen. Most of the cases had been run by Kradz: rape, robbery, murder, theft, arson. McQuiston would read the questions that had been asked in the interrogation and we would look at the stress patterns on the screen.

"D.I. or N.D.I.? Do we burn this guy or let him go?" he would ask and then remind us of this or that subtlety of interpretation. "Watch for the sharp leading edge ... is there a plateau? ... Watch for the relief pattern after the lie ... don't forget to check the outside-involvement question."

Then we would guess, hesitantly, tentatively, looking for support from the others in the class and usually with a qualifier like, "Well, I'm not sure, but...."

And then, without saying yes or no, McQuiston would push a button and a hand-drawn figure would appear: a stick man with a halo for N.D.I., a cat-o'-ninetails for D.I./guilty knowledge and, for plain straight D.I., either a hangman's noose or an electric chair. On McQuiston's carrousel projector, the penalty for lying is death.

We went through perhaps 20 cases (with significant mistakes and one or two hopeful successes) and took a break. Almost everyone got coffee and drifted into the pegboard laboratory in the next room, where Bill Ford and another technician named Pete Preston do their tinkering. Ford wasn't there, but Preston---in his 50s, wearing a sport shirt and horn-rims---was chatty and eager to explain the theory and practice of what he calls information gathering.

He told us that there are about 50 ways to attack a telephone and showed us a test bug he is working on that shuts itself off when detection devices are brought into the room. He showed us a roll of rubber molding, the kind used in building walls, that has a secret wire running through its core. And he showed us "The Cloak," a black box with buttons on it that sits on the desktops of paranoid executives and (Preston said) foils several methods by which their phones could be tapped.

He has a headful of anecdotes that he salts into his explanations and demonstrations. Gathering secrets isn't always laboratory clean, he told us. During World War Two, he saw men following the retreating German army, picking up pages from training manuals that the Germans had wiped their asses on: The Third Reich was out of toilet paper and leaving valuable information behind as they crapped and ran.

He showed us some of the anti-bugs he's designed and said that if there are 50 ways to get at a telephone, the foxes' part of the spy's game is knowing 51 ways to undo a tap. He said a talented spook never releases a new method of bugging until he has designed the antidote.

When we got back to the table, McQuiston was waiting. He had a salesman's sample case in front of him and when we sat he said, "I thought you might be interested in a preview of our course in methods of surreptitious entry."

Everyone was and he began unloading from the case a leather kit about the size of a wallet and eight or ten padlocks.

"This lock," he said, holding up a heavy silver one, "is the best you can buy. They use it on the gun lockers in the Army and it's guaranteed by the manufacturer to be impossible to pick or cut in less than twenty minutes."

With no other words, he fished two hairpin picks out of the little leather packet, fiddled for less than 15 seconds, popped the lock and then held it up like the magician he is. It was like watching an armory being looted by a ghost. Then he violated the other locks, each as easily as the first, until he got to the last one, a small cheap lock that you might trust to keep your garden tools safe, but nothing more. It took him nearly a full minute and three picks to crack it. Then he showed us a long flat piece of metal that he said fits into the window well and gets you into almost any car without breaking anything in less than 30 seconds.

"I'll be giving the course in about a month---costs a grand." he said and then held up his own wallet and pointed to a little card inside. "By the way, you need one of these. Bonded-locksmith I.D. If they catch you with this little set of tools and without this card, it's five to ten years." He was smiling. Then he put his kit back together, locked it (for what reason I couldn't imagine) and we began running cases again until all eyes except his were glassed over.

We ran more cases, flogged some half liars and hung the rest, until noon the next day. By that time, plane reservations were being discussed and worried over. And because no one's confidence level was very high, McQuiston delivered the final guarantee that goes with the machine.

"If you've got a case that you're not sure of, that you want consultation on, just phone it in and we'll look at it. You all have a telephone hookup with your machines and we can make our own tapes over the phone and then run them."

Everybody liked that. There was a general feeling of excitement and even power as the machines were checked and readied for travel. The PSE worked. That seemed clear from all we had seen and heard. But whether or not it really was a machine built by geniuses to be used by idiots (a phrase we'd heard often) was on its way to some kind of acid test.

I saw Allan Bell near the door on my way out. I had a tape recorder with a small carton of accessories hanging from a rope below it in one hand. The PSE briefcase was in the other. I looked like Huck Finn, but I felt like James Bond and I asked Bell, "Whose idea was it to fit this thing into a briefcase? It's so sexy and secret."

He laughed. "All our people have been in the field," he told me. "So they design equipment to be used out there. It's vastly different to test something in a lab and then to go out into an operational environment where it's dark, for instance, and you can't see what you're doing, you're apt to be discovered, your hands are sweaty, your stomach is tight, the skin on your back is crawling and every sound is like a pistol shot. It's a matter of having been there."

"Jesus," I said, "I don't think I'm going to be able to give this thing anywhere near its real test. I don't even want to."

He laughed and told me good luck and when I had said goodbye to my classmates and McQuiston. I left: Out of the spies' nest into a world of liars and secret keepers, into the field, into the parking lot. As I was loading the gear into the trunk of my car, one of the 7-11 people slammed his car door. It didn't sound much like a pistol shot, but I made the connection.

• • •

I had the PSE for about five months and putting it to any kind of ultimate test was farther beyond me than I had imagined. I kept a notebook that I was going to call a journal, but after a few months of riding with me, it looked like a shoe that had gone 10,000 miles. There were holes in it, some parts were literally rubbed away and it was weak with the thought processes of a novice in the field. But with a little polish and some afterthought additions, it looks like this:

Washington National Airport: The thing hefts like a briefcase full of booze and standing in line waiting to go through the metal detector and then past two Customs agents, my arm is asleep. The rest of me is waiting to be searched and hassled, questioned and maybe detained when they open it for inspection. They smile at me as I go through. Nothing. Over the five months, I'll ride a half-dozen planes and go through anti-hijack inspection at Dulles, O'Hare, Los Angeles and San Francisco International, and no one will ask me to open the case.

Chicago: People who know where I've been and what for ask me where the lie machine is. When I hold up the briefcase, they usually don't believe me. Once they do, they always ask if it's on.

The choice now is between tapping phones and otherwise wiring people to get tape recordings where they talk in earnest about their jobs, marriages, money, morals, fears and wildest ambitions---or designing a game that will demonstrate the machine in a benign and playful way.

There are legal problems if I use the machine for real. Illinois is one of the 16 states that have a law prohibiting anyone but a licensed polygraph operator from using any kind of lie-detection equipment. And Congress has before it right now several bills that will attempt to curtail even the licensed use of polygraph for situations like job screening and internal theft at industrial plants. (The unions lobby hard against that sort of snooping.)

Dektor is unsure how current and proposed legislation will affect the PSE, but it isn't very worried about it. Industry has always found its way around laws that concern polygraph and has been using it more and more. What gets whispered in the lie-detection community now (Nixon's landslide) is that the general political climate is unlikely to produce major reform or repression of lie-detector activity. (I'll bet they're right.) And if the going ever gets really tough over legal definitions of what PSE is and when it can be used, the Dektor people can always say that its principal purpose is not lie detection but stress analysis. That makes it an invaluable research tool for doctors, psychiatrists and other behavioral scientists, and prohibiting its use becomes foolish. And it's true enough that the Dektor people expect their biggest market to be in the scientific community.

So the laws are up for grabs (as they always are for spies), but I decided on the game anyway. Mostly because there are some things I don't want to know for sure about the people around me.

The game: Subject takes from one to ten dollar bills and puts them into his pocket (secretly). I then make a tape that begins with two or three irrelevant questions (name, home town, etc.) to get a general nervous-tension level, and then I ask, "Did you put one dollar ... two dollars ... three dollars ... etc. into your pocket?" Subject answers no to all ten amounts asked. Then I run the tape through the PSE and nail the filthy liar (in theory) and if I do, he gives me the money in his pocket. If I don't, I give him a matching amount.

Out of 12 subjects I nailed only six, which was a little embarrassing because of the pre-interrogation bravado that I had employed to psych them. (I do not smoke cigars, and it's hard to hold a man at bay with a cigarette, but I was trying.) The six I caught all had textbook-perfect stress patterns on the relevant-amount question and they were easy.

Of the six I didn't get, one was because I misinterpreted the lines in front of me and two beat the machine (and me) cold by trying. One of these was a psychologist who said he simply thought of very stressful things (falling over a cliff, getting busted, losing loves) all the way through the test, and I lost several dollars to him.

The other, a man with a huge voice, broke the machine---or at least it broke while I was running his test. The pen went crazy and began making patterns that looked like freeway maps. I sent the machine back to Dektor and they said they had encountered this problem with one or two machines before, then they fixed it and sent it back, I ran the tape again and although I was very sure he had put eight dollars into his pocket, it turned out to be six dollars. After I gave him the money, he revealed his simple and devious method: When I asked him if he put one dollar into his pocket, he rephrased the question, silently, to himself, substituting the actual amount (six dollars) and then answered. He did the same thing on each question and the result confused the machine successfully.

The three others who beat the machine did so probably because they were not in enough real jeopardy under the circumstances. That's the trouble with all games on the PSE: It measures stress, and if the subject doesn't worry enough about the consequences of being caught, he may not react enough for the machine to chart it. Or else those three were moral idiots, morons, psychopathic liars or one of the similar types who defy all lie detection because they do not have a moral code that they are aware of violating. But I know these three people, and it was probably lack of jeopardy that kept them safe.

I decided that 50 percent is a very good average for the PSE on such games.

California: I am at the point of dexterity with the machine where I play it like a piano and a diploma has arrived in the mail that says I'm a certified PSE examiner. All that means is that Dektor trained me and that I passed. They have flunked only one man, and he passed later. I imagined my classmates, out in the field, and up to here in confidence, laminating their diplomas and busting cases like crazy.

A real offer: A man from a large international company calls to say that he has heard I have the machine. Would I be interested in running some of his employees who he suspects are getting into the till pretty good?

I'm vaguely tempted but quote the California law, which is one of the 16 and similar to the Illinois law. Later I find out that he has bought one and taken the course himself.

The election: I have been waiting in front of the television for a month, waiting for the campaign to heat up. The Dektor people had told me about making tapes off the TV of the Howard Hughes / Clifford Irving press conferences. After they ran the Hughes phone call through the machine, they say they knew that Irving was doing the lying. That was nearly two weeks before the rest of us knew, and so I am waiting for either Nixon or McGovern to say something that I can reasonably call a lie. Nothing. Nixon never comes on television. It's as if he knows about the PSE. And McGovern answers direct questions so generally and with such deft vagueness that his stress levels have more to do with camera lights than they do with the substance of anything he says.

The Eagleton crisis begins. I'm fumbling with the tape recorder in front of John Chancellor's six-o'clock news when McGovern comes on and says, "I'm behind Tom Eagleton a thousand percent." I get it, just barely. Then I look up at the screen: McGovern sweating, fidgeting, voice out of control, eyes watery, and it comes to me that I'm using a bunch of electricity and recording tape and $4000 worth of equipment to try to find out if this man is under stress. It was architect Max Frisch who said that technology would keep us from experiencing the world if we weren't careful. I put down in my grungy little notebook that Max Frisch was a wise man. Then Eagleton appeared, looking like a ping-pong ball that had been stepped on, and although I taped him, too, I never ran either of the recordings on the PSE. It seemed too much like going grave to grave through a cemetery with a stethoscope, just to be sure.

The Watergate circus begins. Nixon is still nowhere and even the most remote of his spokesmen are appearing only for seconds at a time to read short general disclaimers. You can't get a reliable PSE result on a man who is reading something, because he isn't necessarily listening to himself down around his central nervous system. It's as if everybody in the Nixon Administration knows about the PSE. Nothing.

In all the five months I had the machine (including what should have been the hottest part of the campaign), Nixon never put himself into even one tapeable situation. I decided that if the Watergate spies represent his offense, he doesn't have much. His defense, however, is dynamite.

Some old tapes: I've kept all the tape recordings I've made for stories over the past four years. Perhaps 50 hours of short interviews, mostly; the moments in an assignment when you finally get whatever guru, or millionaire, or junkie, or movie star or revolutionary to sit and answer some direct questions into a microphone. Almost every subject will draw back a little when the tape starts rolling. He'll become careful and sometimes he'll lie, because telling your secrets to someone who's going to retell them to a crowd is comforting only if you're on your deathbed.

All the stories for which the tapes were made are written now (or else given up for dead) and all decisions about how much was the truth I made a long time ago by the witch powers we all use on each other. I believe in those powers, and of all the tapes, there was only one I wanted to check with the machine. The subject had lied to me at least as much as she had told the truth. I'd thrown out the lies I was sure of---or had called them lies---and left the things that seemed true or even probably true. But I had left out one particular fact because I couldn't, by any of my powers, prove it true or false. It had seemed important then; now I ran it through the PSE.

I took the parts of the tape I knew to be true and checked their patterns against what I knew to be false and then compared. Then I ran the answer I wasn't sure of: It was a stone lie.

And I thought as I stood with the tape in my hand that they had me. I believed in the PSE enough to use it seriously, and I believed in the results enough to feel bad that I didn't have that information when I wrote the story. The tyranny of new-age machines doesn't come as rape; it is seduction. Speed, accuracy, objectivity, "the truth," whatever it is a new machine promises, man must like it, or need it, or think he needs it. No one buys a machine that promises to make him more ignorant than he is. And we have never been very afraid of our machines, anyway. Most of that show has been a sham. Finally we all know that the only animal it is really dangerous to turn your back on is man.

I wondered later why the hell she had lied to me about that.

• • •

The parking-lot situation at Dektor hadn't gotten any better by the time I arrived to return the machine to them. Bell and McQuiston are driving brandnew Lincoln Continental Mark IVs now, and they take up more room than the cars they drove five months before.

The staff has grown, too. They've hired accountants, business managers, secretaries, salesmen and a woman to share training duties with McQuiston. Mike Kradz, the police lieutenant from Howard County, works full time for Dektor now, and they have hired another spook. His name is Mark Hanson and he has perfect gray hair and eyes as steady and beautiful as Kradz's. He was a full bird colonel in Army counterintelligence and spent 26 years trading favors and secrets with the CIA. He is said to be legendary around Foggy Bottom and other secret places. For the past three years, he has been working as an investigator for New York's Knapp Commission, the group that has been exposing police corruption for that city (looking for needles in a stack of needles).

McQuiston wasn't there. He was in Lebanon---there is now something called Dektor Middle East---conferring, dealing, selling the PSE. Many airlines have considered putting money into the development of a smaller, quicker PSE that would be used to screen boarding passengers for hijackers. McQuiston was in the Middle East working up a profile of 1000 passengers to establish the normal stress response to questions about their name and destination for Middle East Airlines. When he talked to some American airlines about the same project, he told them that the PSE might hold 10 or 20 smugglers, nervous fliers or people on their way to forbidden rendezvous off a boarding 747. The airlines said great but hesitated because they want the Federal Government to pay for such work.

Bell seemed pleased to see me and we sat in his office again while he told me that they've sold 178 PSEs by now (the rate is up to one per working day), that they have new and excited private investors and that they are thinking about taking the company public. Then he said that the problems businessmen face are more wearying than those of a spy.

"We're trying not to get carried away with the magic of this thing," he told me.

We talked for a while about what I had done with the machine and he asked me if I believed it worked. I told him yes and then we talked about the notion that even if the machine didn't work, it would be enough that people believed it did. I asked him if he ever felt like Edison. It was an awkward question and it sounded like it came from a young newspaperman who was carried away with notions of historic moment, but he handled it gracefully.

"No," he said, "I'm looking for the next thing, something new to work out. I'll be happy if we can use the PSE to finance our other hobbies. I need to keep moving. We're all like that, me, Bill, Mac."

Then he told me about some consulting jobs that Dektor has been asked to do. One of them is staggering in its proportions, spooky and full of international implications, and before he told me about it he said, "This is off the record, Craig." In fact, he had said that to me many times over the five months. He would start stories about the PSE and some of Dektor's specific customers and jobs and when they became juicy, revealing, scary, he would swear me to the kind of secrecy that a reporter must sometimes agree to if he is to move freely through a complicated world.

"You're not trying to make everybody tell the truth," I said to him. "You've got as many secrets as anybody I know. For Chrissakes, I'm helping you keep some of them."

"There's a difference between a confidence and a secret," he said.

"Only among lexicographers," I told him, but he never quite agreed to that. Once a spy, always a spy, I thought to myself: He was wearing a little jeweled black-widow spider for a tie tack. And he is impressive; Power is gathering around him the way cigarette smoke does around a writer at work and worried over the telling of other people's secrets.

We said goodbye and on my way out I met Ford near the door. He was carrying something that looked like a metal cane with a box attached to the handle and a two-pronged device at the end. When he pushed the button on the box, the thing made a sound like a bicycle horn. "Well," he told me when I asked, "you hook it up on an oscilloscope, plug the end into a wall socket and then push the button. It tells you if there are any bugs along the line. We just came up with it and it ought to save a lot of time sweeping a room."

I almost asked him to show me the anti-PSE machine. I don't know if it exists yet, or even if they're working on it. But it will, and probably in this lab first, but if not here, then somewhere else where spies gather. That's the foxes' part of the game, after all, and Bell needs his secrets at least as much as I do.