Lynch is a psychologist (Co-Director of the Psychophysiological Clinic and Laboratories at the U. of Maryland School of Medicine)--and an Irishman with a confessed tendency to blush. (""No pumpkin in all its resplendent autumnal glory. . ."") He can be a bit garrulous, but the observation is not inappropriate. Lynch argues that people with high blood pressure (hypertension) are the obverse of blushers: they conceal their feelings from the world, and ultimately from themselves. Notwithstanding all that has been said about stress and type A personalities, Lynch finds an overlooked language dimension. With automated computerized equipment, he has been able to demonstrate that (with rare exceptions) everyone's heart rate and blood pressure go up when talking. (Even just talking aloud; even talking to one's best friend--but especially when talking to people in authority, etc.) In hypertensives, these increases are even more dramatic, and the fluctuations more extreme. Lynch arrived at this thesis via Pavlov and a dog's changed physiological responses to an observer, and his own teacher W. Horsely Gantt, who had studied with Pavlov. Much of the book consists of a review of past literature and in-depth discussion of a few cases (in which subjects reinforce their habitual modes of speech and of distancing themselves behind smiling visages). Lynch also has some evidence of noteworthy exceptions to the rule--among schizophrenics (a disorder characterized by dysfunctional communication) and among some monks, prisoners, and other isolated populations. Migraine sufferers, too, show little body change or emotional reaction. To treat these kinds of patients, Lynch uses a carefully stepped approach that involves breathing and relaxation, moving toward greater self-awareness of feeling. One can question the universality of Lynch's profile, its significance in relation to other risk factors, and the wisdom of adding to a patient's stress by positing some emotional deficit. But Lynch himself is sensitive to the ""blame the victim potential""--and the implications of the evidence are at least worth consideration.