There's a fine New York Times Magazine article on Larry Summers, the economist and Harvard president. The writer seems perplexed by Summers' rough-edged personality (I can think of hundreds of economists with personalities like Summers', though almost none with his intellect) and spends more time there than on what I found his best point: Summers wants students to be more scientific.

The fundamental reason Summers wants to change the undergraduate curriculum is that, as he explains, the nature of knowledge has changed so radically. Summers often says that one of the two most important phenomena of the last quarter-century is the revolution in the biological sciences. And yet, as he also often says, while it is socially unacceptable at an elite university to admit that you haven't read a Shakespeare play, no stigma at all attaches to not knowing the difference between a gene and a chromosome or the meaning of exponential growth. Summers compares this ignorance to the provinciality of never having traveled abroad. He wants every student to live in science for a while and not just to do some sightseeing in a course designed to help you ''think like a biologist.'' Summers is not categorically opposed to the ''ways of thinking'' approach. ''The hard question,'' he said, ''is the line between learning a lot of science in one area and surveying more broadly but less deeply and thus less close to the genuine professional enterprise.''

But the intellectual revolution that Summers says he hopes to capture in the new curriculum is not limited to science itself. ''More and more areas of thought have become susceptible to progress,'' he said, ''susceptible to the posing of questions, the looking at the world and trying to find answers, the coming to views that represent closer approximations of the truth.'' Tools of measurement have become ubiquitous, as well as extraordinarily refined. Archaeology, Summers observed, ''was at one stage kind of a 'Raiders of the Lost Ark' operation. Now we're hiring a chemist who can figure out diet from fingernail clippings.''

Political scientists are using computer modeling to make comparative studies; mathematicians analyze the pattern of change in the AIDS virus to explain why the interval between infection and sickness is so long. The great universities have traditionally defined themselves as humanistic rather than scientific institutions. Summers's point is not so much that the balance should shift as that the distinctions between these modes of understanding have blurred, though clearly in a way that favors the analytic domains -- the soft has become harder, rather than the other way around.

Of course this is worrying to some members of the campus; when political theorist Michael Sandel says ''By training and temperament, economists are intellectual imperialists,'' he's quite right. It is not so much rational choice (which Sandel seems to oppose) that is the source of the imperialism, but that economists are taught to test theories and cast aside those that don't measure up. Summers himself says, ''The idea that we should be open to all ideas is very different from the supposition that all ideas are equally valid.'' I suspect Summers uses the word "Truth" with the capital T.