A sleeper: a book that at times makes you want to laugh, and at times weep, at the combination of human naivetÃ‰ and cruelty. Lykken, professor of both psychology and psychiatry at University of Minnesota Medical School, got interested in the gimmicks and gadgetry of lie detection some time ago, and is clearly an expert. He reviews the 50-odd years of the use of the polygraph--and makes it clear that they are indeed odd. Police and public alike have been all too easily conned into believing that the machines are infallible; if the pen quivers on the crucial question, You Are It. Fortunately the courts have been more conservative, but Lykken points out that many states allow expert witnesses--polygraphers--to testify; and the defense may soon be permitted to report that a client has ""passed"" a lie-detector test. Perhaps more ominous is Lykken's report of the growing use of polygraphs, or ""truth"" questionnaires, in employment-screening or on-the-job tests of loyalty. Throughout, his approach is calm and reasonable. He explains the variant methods and machines in use, pointing out that in virtually no case have there been tests of reliability or validity. What often happens is that the test-giver becomes a clinical tester--judging a person's veracity by an intuitive assessment of behavior. Lykken doesn't fault all approaches to lie detection; indeed, he concludes, somewhat surprisingly, with his own researches into what he calls the Guilty Knowledge Test (GKT). A suspect is shown photos of the crime scene, say, in which an accurate picture is mixed with doctored photos; presumably only the Guilty will flip over the True scene. To amplify the point, Lykken includes a marvelous grade-B mystery scenario illustrating how GKT might be used. Fortunately he is quick to acknowledge that all this is experimental and conjectural, but still. . . . Anyway, a surprisingly entertaining and enlightening text by a psychologist whose pen is swift and sure.