HOUSTON — Houston. - The Thomas hearings featured that notorious device, the polygraph, which emerged, alas, with its image more battered than ever. Anita Hill volunteered for an examination, which indicated she was truthful; Judge Thomas refused one. Senate (and popular) opinion in effect discredited the results.
Three groups are implacably opposed to polygraphs: those who resent having their word questioned, those who regard it as an invasion of privacy -- and those with something to hide. Opponents have been numerous enough to bar its use in court cases or for job applications.
This is a pity, for the polygraph, in properly trained hands, is an extraordinarily accurate tool for separating sheep from goats. Only two groups are unsuitable subjects; psychopathic personalities (who even if they can distinguish right from wrong, simply don't give a damn), and the extremely aged.
The polygraph's heaviest cross is the silly label ''lie detector'' hung on it decades ago; most regard it as some sort of voodoo box flashing green for truthful answers and red for lies, and think it's easy to beat. It isn't. It doesn't make any difference what you answer, or even if you keep your mouth shut, and attempts to ''beat'' it are glaringly apparent to the operator.
The device measures autonomous (uncontrollable) reactions -- pulse, blood pressure, perspiration, breathing -- to questions; verbal answers are immaterial. It simply discloses ''guilty knowledge,'' and has probably cleared 100 suspects for every guilty party it ever fingered.
A simple demonstration convinces most. They write down a number from 1 to 10 and hide it, remaining silent while the operators asks ''Is the number 1? 2? 3?'' They will invariably react to the correct number.
Assume John Doe has been found with a dagger in his heart, a fact not released to the media. There are three suspects. They aren't asked ''Did you kill John Doe?''' -- they're asked ''Did you kill John Doe with a pistol? With an ice-pick? With a dagger?'' Innocent suspects think ''My God, they're convinced I did it!'' and will react equally strongly to all three questions. The guilty party will react quite differently to ''dagger'' than to the other questions. And reactions to all questions, by all suspects, will not vary whether they answer ''Yes'' or ''No'' -- or remain silent.
In properly conducted examinations, groups of ten carefully prepared questions -- nine irrelevant, one significant -- are given twice; on the second run guilty parties know where the significant question is, and react even less to irrelevant questions, heightening reaction to the significant one. No competent operator simply fires away, as in verbal cross-examination, devising new questions as he goes, trying to ''trick'' the subject.
Another technique is to question suspects not about the actual incident, but about another -- notional -- incident. Instead of asking about the John Doe murder, questions concern James Smith. Innocent suspects think ''I had nothing to do with this Doe case, and now they're trying to pin every case in the book on me.'' They react identically to questions about the two cases. The guilty suspect thinks, ''I may have killed Doe, but they can't pin this Smith case on me,'' and reacts markedly differently to questions about the two cases.
Unless the examination leads to immediate confession (which it frequently does), exclusion from court cases is justified -- not because the device is ''inaccurate,'' but because (for guilty subjects) it constitutes self-incrimination. It is simply an investigative tool. Most criminal cases start with an initial pool of suspects; police work starts with eliminating innocent parties from that pool -- which even if it doesn't immediately finger the guilty party, vastly reduces the work required to solve the case.
For intelligence services, it is axiomatic that a witting contact with an opposition service can't be hidden from a polygraph. All U.S. intelligence personnel are polygraphed at the outset of their careers, as are all agents; subsequent tests (desirable after tours in denied areas) are often delayed by the case load.
Intelligence service operators can usually predict how a run will turn out. Subjects are always informed in advance they are to be polygraphed (''boxed,'' or ''fluttered,'' in the jargon). Innocent subjects are always early for the appointment; guilty subjects are invariably late -- or to try to postpone the appointment. Told to ''relax,'' innocent subjects go well-nigh catatonic in their effort to cooperate; guilty parties continually tense up, twist and writhe, or clench their fists or jaws, complaining the attachments are ''uncomfortable.'' (They're less restricting than a wristwatch or a necktie.) And in nine of ten cases, they'll make a derogatory remark -- something like ''The courts don't accept this, do they?'' And the use of drugs to mask responses is ludicrously apparent.
In the 1950s, the KGB flooded a neighboring Mideast nation with a stream of low-level agents, mostly trained by an officer known as ''Nikolai.'' He told all his agents ''The Americans gave the local police a silly device, but it doesn't work; don't worry about it. If they take you to a chair with a funny box with wires on the table, just make a joke -- 'So this in the famous American ''hot seat,'' eh?' ''
The local police chief learned of this from a defector. He led all border-crossers to a room with a spotlight on an old suitcase on a table, with extension cords hanging out. Any immigrant uttering the suggested witticism was immediately pounced on: ''Aha, and how is my old friend Nikolai?'' He was amazingly successful until, to his regret, Nikolai was finally transferred.
Donald R. Morris, a free-lance columnist, formerly worked in intelligence.