MIT professor Loren Graham, the United States' foremost historian of Russian and Soviet science, doesn't publish fat tomes every decade or so as many of his peers do; he writes topical, digestible books that invite his audiences along on his scholarly travels. His last short book, The Ghost of the Executed Engineer, came out in 1993 and told the appalling story of Petr Pal'chinskii, a Russian engineer repressed and ultimately executed for his humanitarian scruples. Graham's latest book is broader in scope but still manages to weigh in at less than 200 pages. To use his own phrase, it is another "small book about big questions."

The first question Graham takes up is fascinating, but it may be "big" only to those familiar with the ongoing debate in academe over the epistemological nature of science. Does it refer to an objective reality, or is it a social construction, inextricable in style and content from the culture and the times that produce it? Graham, an exceptionally clear-headed thinker in a field rife with sophistry, uses the Soviet example to show that it is both. Lysenkoism, a disastrous agricultural policy built on the Lamarckian idea that acquired characteristics can be inherited, enjoyed a 30-year reign in the Soviet Union because it accorded so well with Marxist principles, Graham explains. But facts, not philosophy, proved Lysenkoism's undoing in the 1960s, as Western farmers outperformed their Soviet counterparts and Western biologists gathered irrefutable evidence on the existence and nature of genes.