Paradigms and Revolutions

Thomas Kuhn was a physicist preparing a general education course when he sat down with Aristotle's Physics and found it not merely wrong but unintelligible. This was strange. Aristotle was an extraordinarily careful observer of everything else, and here he appeared to be saying things about motion that were not even reasonable errors. Then, as Kuhn described it, the text suddenly reorganised itself, and what had looked like a failed version of Newtonian mechanics turned out to be a coherent framework in which the questions themselves were different, and Aristotle's answers were mostly right about them.

That experience (a gestalt switch, in which the same material reassembles into a different picture) became the centre of The Structure of Scientific Revolutions, published in 1962 in, of all places, the logical positivists' own encyclopedia series.

The book's first move is descriptive and deflationary. What do scientists actually spend their time doing? Not testing their fundamental theories. A mature field operates inside a paradigm: not merely a theory but a package (exemplary solved problems that students learn as models, standard instruments, agreed standards of what counts as a good answer, and a shared sense of which questions are worth asking). Within that package, work consists of puzzle-solving: extending the theory to new cases, measuring constants more precisely, tidying up discrepancies. And crucially, when a puzzle resists solution, the failure is attributed to the scientist rather than to the paradigm.

This is precisely the opposite of what Popper says scientists should do, and Kuhn's claim is not that they are failing. It is that the dogmatism is productive. A discipline that permanently questioned its own foundations could never develop the precision required to detect a real anomaly. You have to take a great deal for granted in order to be surprised by anything.

Anomalies accumulate anyway. Every paradigm has them, and normal science lives with them for decades. The anomalous perihelion of Mercury sat in Newtonian astronomy for fifty years. A crisis begins only when anomalies are persistent, central and multiplying, and when the repairs required to accommodate them start to look like propping up rather than using. Even then nothing happens without a rival available, because scientists do not abandon a framework for none. The Ptolemaic system was not refuted; it was eventually outcompeted by something with a different shape.

And when the shift comes, Kuhn says, it is not a step-by-step transition forced by logic. It resembles a conversion, and generational replacement does part of the work. He quotes Planck's remark that a new scientific truth triumphs not by convincing its opponents but because they eventually die, and he quotes it approvingly.

The claim that caused the most trouble is incommensurability: rival paradigms lack a fully shared standard by which to judge between them. They disagree not only about answers but about which problems matter, which methods count, and what the shared terms mean. 'mass' in Newton is not the concept 'mass' in Einstein, even though the equations converge at low speeds. There is no neutral observation language to appeal to, because what counts as an observation is shaped by the framework. Critics took this to mean theory choice is non-rational, essentially a matter of persuasion and politics, and Kuhn spent the next thirty years insisting that was not what he meant. His considered position is that scientists share values (accuracy, scope, simplicity, fruitfulness) which genuinely constrain choice, but which can be weighted differently by reasonable people, so there is no algorithm.

He also did not say science makes no progress, though he is endlessly accused of it. He said it progresses and that the progress is not towards anything. His analogy is evolution: development away from primitive beginnings, not development towards a predetermined goal. Later paradigms really do solve more puzzles, more precisely. What he denies is convergence on a mind-independent account of how things really are, and he regarded the demand for such a goal as a residue of theology that the historical record does not support.

The irony he lived with is that 'paradigm shift' escaped into general usage and now means something like 'a notable new idea', which is roughly the opposite of a crisis in a mature science. He complained about it for the rest of his life, which is the fate of anyone who coins a good enough phrase.