How does a paradigm shift happen, according to Kuhn?

In Kuhn's account, a mature science works within a shared framework, called a paradigm, until problems that the framework cannot resolve pile up. Confidence breaks down in a crisis, a promising alternative appears, and a dominant group of leading researchers moves to it. That move is the paradigm shift, which Kuhn called a scientific revolution. [1] [2]

A field becomes a mature science when a few concrete problem solutions come to serve as models of good research. These exemplars form the basis of a paradigm, and work carried out under it is what Kuhn called normal science. Kuhn cited Ptolemy's planetary computations, Lavoisier's use of the balance and Maxwell's mathematization of the electromagnetic field as examples. Normal science is convergent: it discourages revolutionary initiatives and unexpected discoveries because they threaten the paradigm. [1] [3]

Because normal research is so detailed and focused, it inevitably turns up anomalous results. Kuhn held that every paradigm faces anomalies at all times. Scientists usually do not treat them as refuting the guiding theories. They ignore them or explain them away where possible. [1] [3]

Anomalies become a serious problem when they undermine the practice of normal science itself. One example in the Stanford Encyclopedia is an anomaly that casts doubt on a piece of equipment that much research relies on. A widespread loss of confidence is what Kuhn called a crisis. At that point serious alternatives can be entertained. If one of them attracts a dominant group of leading researchers away from the old paradigm, a revolution has occurred. The new paradigm replaces the old exemplars, which then look defective or old fashioned. [3] [1]

Kuhn compared the moment of change to a gestalt switch. He wrote that scientists often speak of scales falling from their eyes or of a lightning flash that lets a puzzle be seen in a new way. He argued that such intuitions are not an interpretation of the old experience, because they gather up large portions of it and transform it into a different bundle linked to the new paradigm but not the old. [4]

Kuhn contrasted his view with Karl Popper's. Popper held that a single reproducible anomaly logically requires rejecting a theory. Kuhn said there are no rules for weighing the significance of puzzles against one another, so neither the decision to revise nor the particular revision is rationally compelled. The revolutionary phase is therefore open to rational disagreement among competing ideas. [3]

Some sociologists and historians of science took a passing remark of Kuhn's, that factors such as the nationalities and personalities of leading figures might matter, and developed it into the thesis that outcomes are always determined by socio-political factors. Kuhn repudiated that reading. He maintained that the deciding factors, especially in modern science, are almost always internal to science, chiefly the puzzle-solving power of the rival ideas. [3]

Kuhn said revolutionary science is not cumulative. Not everything achieved in the earlier period of normal science survives, and a later period may lack an explanation for a phenomenon the earlier one thought it had explained. This is known as Kuhn-loss. After a revolution, a new period of normal science begins under the new paradigm. [3] [5]