Hearts are in the news. Old ones. New ones. Unapproved ones. Somebody speaks of a broken heart these days and we probably think first of a mechanical one. It is the epitome of modern medicine: If you have a broken part, it probably can be fixed.

The body as a machine is a view that has dominated medical science for centuries. Even human relationships have been factored in as simply other stimuli that act upon the body positively or negatively. Lower your intake of cholesterol, maintain your proper weight, exercise moderately and avoid loneliness for proper maintainence of cardiovascular health.

Enter James J. Lynch, professor of psychology and codirector of the Psychophysiological Clinic and Laboratories at the University of Maryland School of Medicine. His earlier book, "The Broken Heart" (1977), explored the fact that human loneliness is among the most important causes of premature death in modern America. You name it -- cancer, cirrhosis, suicide, automobile accident -- the death rates are higher among people living alone.

One statistic especially piqued Lynch's curiosity: The single, widowed and divorced were two to four times more likely than married people to die prematurely from hypertension, stroke and coronary heart disease. How could a feeling state, loneliness, affect the cardiovascular system?

His search for the answer to this question is the basis for this new book, which documents how the cardiovascular system does far more than change in response to internal and external demands. It communicates.

Using new computer technology that permits continuous monitoring of blood pressure and other vital signs, Lynch and his colleagues investigated the impact of human dialogue on thousands of individuals. Their sample included newborn babies crying in their cribs, children in school reading from textbooks, nursing and medical students describing their daily routine, hypertensive patients in their clinic, schizophrenics in psychiatric wards, people about to undergo bypass surgery and even monks in monasteries. In every case the link between language and the heart was "clear and undeniable."

Talking, even when devoid of emotional content, elevates blood pressure; listening reduces it.

The research showed that the entire body, even down to the microscopic levels of circulation and exchange of blood gases in individual tissues is involved in human dialogue. Says Lynch: "We speak literally millions upon millions of words in the course of our lives . . . we spend every living hour attending to or defending against the living world outside our skin. Surely the cumulative effects of these experiences have to affect human cardiovascular health and may even determine it."

The evidence presents a compelling case for the effects of human interrelatedness on health. Presently an overlooked factor in medicine, it was not always so. Two thousand years ago, Aristotle defined emotions as Logos, "words expressed in human flesh," and dialogue as this language of emotion shared between people.

Things got muddied when Descartes, in the 17th century, said, "I think, therefore I am" instead of "I feel, therefore I am." In one bold stroke, he made the human body irrelevent to human dialogue and, according to Lynch, marked the beginning of human loneliness. Emotions became seen as a scientific problem, the body as a machine, and modern medicine was born.

Wrong, says Lynch. The reality of a language of the heart is as definitive as the cardiovascular changes monitored on a computer screen during dialogue. Aristotle was on to something.

Most intriguing, the heart speaks even when the subject is totally oblivious to its message. He cites numerous cases of "dysfunctional communication" -- the hypertensive who speaks calmy and evenly while his blood pressure soars; the migraine patient who smiles serenely while speaking, totally unaware that her heart is racing and her hands are icy.

This lack of congruence between their words and the activity of their hearts results from a lack of awareness of their own feelings. Such dissonance, says Lynch, is dangerous to health and could be deadly. He describes an interactive therapy developed at the clinic where such patients are helped through dailogue to get in touch with their emotions and to interpret cardiovascular changes previously undetected.

This book is heavy reading. In its attempt to be an exhaustive medical, philosophical and psychological treatise, it occasionally becomes ponderous, esoteric and confusing. The thesis, though well-supported, is bound to be controversial. But let the experts argue. Lynch's vision is creative and stimulating. Stick with it. It is worth the effort.