The Structure of Scientific Revolutions [Paperback]

In The Structure of Scientific Revolutions, Kuhn defines a "paradigm" as a set of assumptions, rules, or model problems that define what the important questions are and how to go about answering them. Without a paradigm, would-be researchers are overwhelmed by the sheer mass of data. A "paradigm shift" occurs when a group of scientists reject all or part of their existing paradigm to adopt a new one. This process not only means changing assumptions: it also means reevaluating previous conclusions to see if the old facts still fit within the new paradigm.

Kuhn uses the term "normal science" to describe the work that scientists do as they work within a given paradigm. Their shared set of assumptions, rules, and model problems fairly makes it easy to see what research remains to be done. Occasionally, anomalies will appear. These are events that cannot be explained within the existing paradigm. Normal science tends to ignore anomalies. Instead, by concentrating attention on a small range very specific questions, "the paradigm forces scientists to investigate some part of nature in a detail and depth that would otherwise be unimaginable."

As more and more research is done within a given paradigm, anomalies tend to crop up. This is because the existing paradigm makes very exact predictions about the expected results, and normal science tests those predictions in ever-finer detail. At first, when the results do not match the predictions, those results are discounted. Some researchers assume the equipment was faulty and so they don't publish results that would only seem to embarrass them. Others try to account for the results by some refinement of the existing paradigm. (The classic case of this involved the medieval astronomers, who kept adding more and more "epicycles" to their Earth-centered model of the universe to explain the results they observed.) Finally, researchers are human, and have been known to simply "fudge" the data to match what the paradigm predicts. Thus, even if every experiment produced exactly the same results, the published research in that field might show a range of results.

Eventually, as the anomalies accumulate, scientists begin to acknowledge a crisis. The results no longer fit the paradigm. According to Kuhn, however, simply abandoning the paradigm is not an option. A scientist can get so frustrated with the paradigm that he abandons it to become a priest or open a bicycle shop, but in doing so, he quits being a scientist. A scientist is not a scientist without a paradigm. The only way a scientist can abandon a paradigm and still be a scientist is to adopt a new one. Kuhn calls this a "scientific revolution."
Kuhn blamed textbooks for creating a false impression of the nature of science and of the role of discovery and invention in its advance. When Kuhn first published his book, science was generally presented as an objective advance towards truth.

According to Kuhn, textbook publishers downplayed the "revolutionary" changes that had taken place in their fields. In 1962, if a textbook covered the history of science at all, it tended to make the advances look inevitable. Kuhn argued that science textbooks present an inaccurate view of the nature of science: they make it look as if science had reached its present state by a steady process, like adding bricks to a building.

The revolution is over when one paradigm displaces another, after a period of paradigm testing. According to Kuhn, however, this is not the result proving one paradigm true and another false, however. To some degree, each paradigm is able to account for all the observations that fit within its set of assumptions and rules. The great German physicist Max Planck used to say that old scientists never change their minds: they just die. Kuhn claims this goes a little too far: instead, scientists slowly convert to the new paradigm, for a number of different reasons. Eventually, if a new paradigm is successful, only a handful of hold-outs support the earlier worldview.

Kuhn's book set off a scientific revolution in its own right. People routinely speak of "paradigm shifts" now, and historians of science (and textbook writers) are much more likely to report on the kinds of controversies that were invisible before The Structure of Scientific Revolutions was published.

Kuhn concludes with a startling claim. He argues that scientific revolutions take place in a blind evolutionary process. Paradigms compete for survival, not for truth. This contradicts the "modern" assumption that mankind is steadily advancing towards the truth through science. Given Kuhn's revolutionary impact on our view of science, this book may mark the beginning of the end of the "modernism."