"SCIENCE IS NOT the dispassionate analysis of

impartial data," writes Philip Hilts in his fine new book. "It is the human, and thus passionate, exercise of skill and sense on such data." The exceptional sensibilities of three skillful men are portrayed in Scientific Temperaments -- physicist Robert Wilson, biogeneticist Mark Ptashne, and computer whiz John McCarthy -- men whose work is at the crux of the uneasy alliance of science and society.

Hilts chose wisely. The three are at the top of their fields, and their fields are elementary: Wilson, former director of Fermilab, explores the atom; Ptashne, a Harvard professor, unlocks secrets to human genes, and McCarthy, a Stanford professor, wrestles with the promise of artificial intelligence. Each must be philosopher, theologian and politician, for their discoveries strike at the heart of what's most basic -- even sacred -- in our lives.

The author, a staff writer for The Washington Post, wants to bring the personal approach of the "new journalism" to the laboratory to create a single picture of the workings and workers of science. That's not an easy task. Explaining three vastly different disciplines alone is arduous; giving us intimate, three-dimensional landscapes of the men's lives -- relations with colleagues and family, habits of work and play, and political and professional challenges -- is an ambitious project. The result is uneven: finding the balance between portraying the scientist and the science itself is troublesome. But when it's carefully weighted, Scientific Temperaments is a most gratifying book.

The sketch of Robert Wilson blends all the ingredients expertly. Hilts contrasts the physicist's twin roles, the researcher probing the subatomic world and the innovative taskmaster leading Fermilab -- the giant research center in Illinois -- "with a cowboy's ornery confidence." From his early days with Ernest Lawrence at Berkeley, to Princeton, and on to work with Enrico Fermi and J. Robert Oppenheimer on the first atomic bomb, Wilson was at the core of the midcentury's scientific action. Hilts' appreciation for the work is often eloquent: "For the past forty years, the foremost occupation of physics has been to read, like a star map, the arcs and angles and trajectories which emerge from the flash of energy in particle accelerators. . . . Alongside the great machine of physics, the physicists move in daily routines like priests repeating rituals of transubstantiation."

The rendering of Wilson's work--especially the fascinating arcana of quarks, neutrinos, et alia--is superb. Hilts also reports the story of how a boy from a Wyoming ranch becomes a renowned physicist. Wilson recalls the moment of reading Sinclair Lewis' Arrowsmith: "My job, when I wasn't in school, was riding the range for cattle. It, too, was an isolated life, but one of delightful independence. I felt I had something in common with and could empathize with Arrowsmith on his lonely research frontier." Later, at Princeton, Wilson believed he solved a major snag in the bomb project; when the inspiration came, he thought, " 'my God, I am the man who knows how to make the bomb! I, a man of about twenty-five yars old, would almost by myself win the war . . ." But now, nearing 70, he still feels "a sense of sin" about Hiroshima.

When Hilts weaves such anecdotes, piecing together the men's lives and the sources of their patience and ambition, the makings of successful science are palpable. But we get fewer insights in the profiles of Pstahne and McCarthy.

Hilts quickly relays Ptashne's fears of publicity, the sensible fear that journalists focus on hot political issues and ignore the science. Respecting Ptashne's wishes, Hilts explains the science at a length that is, perhaps, too respectful. Though the research is impressive -- indeed, groundbreaking -- the practical issues of this revolutionary work, issues with which Ptashne is deeply involved, are too briefly considered. We do get a colorful glimpse of the hip young scientist. But Ptashne's feelings about public fears of recombinant DNA and the awkwardness of his related business dealings elude us. It's a good portrait -- clever and informative -- but it lacks the symmetry of Wilson's.

The profile of John McCarthy is the least engaging. The computer scientist is a bit idiosyncratic, constantly playing little mind games that, apparently, are useful to his craft. Hilts never really gets the measure of the man nor the qualities of artificial intelligence -- indeed, we learn far less of McCarthy's science than we had of physics and biology, perhaps because of its relative infancy. McCarthy, too, hasn't the array of interests the others do -- Wilson's accelerator designs and sculpting, Ptashne's obsession with a Stradivarius. As a result, this last portrait seems pale.

There's also some scattershot political commentary near the end about society's reaction to technology, but it ends abruptly as a mini-tirade. If considered at all, the political topics raised -- public funding of research, the ethics and business of genetic engineering, and computerization -- deserve a more thoughtful treatment.

But Hilts' aim is the scientist, and, for the most part, he scores well. The book is elegantly written, with a sure sense of how much science the general reader can absorb. It's also endearing, feisty, amusing, even spiritual -- all the things we too often believe science is not.