A paradigm shift is a change of framework rather than a change of belief within one. The questions worth asking, the methods that count, and the standard for a good answer all move together.

Where it comes from

Kuhn introduced the term in 1962. A paradigm is the set of shared commitments, exemplary problems and accepted techniques that organizes a mature field. Most research is normal science: puzzle-solving inside the frame, which is productive precisely because the frame is not in question.

Anomalies accumulate anyway. When enough of them resist solution, confidence in the framework itself weakens, alternatives get a hearing, and eventually one replaces it.

Why it is not simple progress

Kuhn's disruptive claim is that the new framework does not simply contain the old one plus more. Terms shift meaning across the boundary, some problems that mattered stop being problems, and data that counted may stop counting. This is incommensurability, and it is why he wrote that afterwards scientists work in a different world.

His examples are the shifts from Ptolemaic to Copernican astronomy, from phlogiston to oxygen, and from Newtonian to relativistic physics.

The phrase now means any significant change, which drains it. Three things go missing.

  • It is not gradual. Kuhn's point is that the shift is a break, not an accumulation, and the crisis preceding it is part of the structure.
  • It is not simply an advance. Something is usually lost, which is what incommensurability means.
  • It requires a paradigm. Kuhn doubted that fields without a settled framework — much of social science, in his view — have paradigms to shift.

The dispute

Popper held that science advances by attempted refutation. Kuhn replied that this is not what scientists do, and a discipline behaving that way would abandon frameworks before they became useful.

Critics read incommensurability as making theory choice irrational. Kuhn denied it, arguing that good reasons operate in the absence of a rule that compels agreement — which may be true of most serious judgment. See first principles.