As Kuhn was finishing up his Ph.D. in physics at Harvard in the late 1940s, he worked with James Conant, then the president of Harvard, on a general education course that taught science to undergraduates via case histories, a course that examined episodes that had altered the course of science. While preparing a case study on mechanics, Kuhn read Aristotle's writing on physical science for the first time.

At first, he was surprised at how Aristotle "appeared not only ignorant of mechanics, but a dreadfully bad physical scientist as well." But as he continued to read, Kuhn did something unusual for a scientist. He wondered whether Aristotle might not be simply ignorant of physics, but whether he, Thomas Kuhn, had not properly understood--that is, whether Aristotle could have meant something other than his own first first guess:

I was sitting at my desk with the text of Aristotle's Physics open in front of me and with a four-colored pencil in my hand. Looking up, I gazed abstractedly out of the window of my room--the visual image is one I can still recall. Suddenly the fragments in my head sorted themselves out in a new way, and fell into place together. My jaw dropped, for all at once Aristotle seemed a very good physicist indeed, but of a sort I'd never dreamed possible. Now I could see why he had said what he'd said, and why he had been believed. Statements that I had previously taken for egregious mistakes now seemed to me, at worst, near misses within a powerful and generally successful tradition.

This passage is from "Regaining the Past," the first of three lectures in The Presence of Past Science, a lecture series Kuhn delivered in 1987. Kuhn discussed there how Aristotle, Volta, and Planck presented difficulties for contemporary historians, due to the incommensurability of their language with ours. Aristotle's conception of motion, Kuhn argued, cannot be directly translated into Newtonian terms. The historian must become bilingual, speaking with both Aristotle and Newton on their own language.

The Structure of Scientific Revolutions

Critics might cite the Damascus-road quality of Kuhn's Aristotelian conversion testimony as evidence of his irrationality. Yet for supporters, this was a powerful example of his historical insight. Kuhn had hit upon a real issue in scientific history. As he would write in The Structure of Scientific Revolutions, a mature scientific community is, "like the typical character of Orwell's 1984, the victim of a history rewritten by the powers that be."

In Structure, Kuhn distinguished between periods of normal science and periods of crisis, during which revolutionary reforms occur. Every scientific era hosts its share of empirical anomalies, those observations that have not yet been folded into the accepted theoretical framework, the framework that Kuhn famously (if inexactly) dubbed a "paradigm."

Normal science treats anomalies as puzzles to be solved using standard tools. But during periods of crisis, the buildup of anomalies may lead to a revolution. As a physicist, Kuhn frequently reached for the Copernican revolution--the subject of his first book-length history--and the quantum revolution--the subject of his last--as his two most familiar examples. After such revolutions, scientists working in their new paradigms reweave the web of history, forgetting old practices and erasing old objections so completely that, in retrospect, all changes appear to have been inevitable.

Latter-day critics of Kuhn have charged that his description of scientific progress is self-undermining.