The Structure of Scientific Revolutions

Kuhn argues that science does not progress by the steady accumulation of facts but through periods of normal science punctuated by revolutions in which one paradigm replaces another. A paradigm, in his usage, is the whole framework of shared assumptions, methods, exemplary problems and standards that defines a scientific community's work.

Normal science is puzzle-solving inside a paradigm. Researchers extend it, measure its constants and apply it to new cases, and they treat failures as their own failures rather than as evidence against the framework. Anomalies accumulate. When they become numerous or severe enough, the field enters crisis, during which practitioners begin to examine fundamentals they had previously taken for granted, and a new paradigm may emerge.

Kuhn's most contested claim is incommensurability, the argument that competing paradigms cannot be fully compared because they define their terms, their problems and their standards of evidence differently. This implies that paradigm change is not a purely rational choice determined by evidence, and Kuhn compares it to a gestalt switch and notes that older practitioners frequently never convert, so the field changes as they retire.

The book was published in 1962 and introduced paradigm and paradigm shift into general usage, where they have been diluted beyond recognition. Philosophers of science, including Popper and Lakatos, disputed the relativism the argument appears to imply, and Kuhn spent years denying that he meant paradigm choice was irrational. Historians dispute several of his case readings. The account of how a research community defends its framework against anomalies applies directly to any field where a dominant model is failing.