Paradigm Shift (Kuhn)

The picture in three movements

Thomas Kuhn was a Harvard-trained physicist turned historian, and the central insight of Structure came to him in 1947 while preparing a lecture on Aristotelian mechanics. Reading Aristotle as a physicist, Kuhn found him incompetent. Reading him as Aristotle would have been read in 350 BCE, Kuhn found a coherent, sophisticated theory. The shock was the realization that the same observations could be organized in radically different ways depending on which conceptual scheme — which paradigm — the observer carried into the lab. Aristotle wasn't bad Newton; he was good Aristotle. The task of the historian was to recover the paradigm in which old work made sense.

From this Kuhn built his three-movement picture of how science develops. Most of the time, scientists practise normal science: they accept a paradigm — a shared set of theories, instruments, problems, and exemplary solutions — and use it to solve detailed puzzles. Newtonian astronomers calculate orbits; Lavoisier-trained chemists track combustion mass balances. Anomalies are noticed but mostly tolerated, postponed, or blamed on apparatus. The paradigm works; one does not abandon a working tool because it is imperfect.

Eventually, however, anomalies accumulate or sharpen. Some prove unusually stubborn. The community enters a state of crisis: experimental results no longer slot into the paradigm cleanly, ad hoc fixes proliferate, and bright young researchers start exploring alternatives. A successful alternative emerges; the community — usually after the older generation retires — switches allegiance. The new paradigm reorganizes the field: terms shift meaning, problems disappear or reappear, exemplars are replaced. This is the revolution. After it, normal science resumes, this time within the new framework.

Five canonical revolutions

• Copernican (1543–1687). Geocentric → heliocentric astronomy. The Earth moved from the centre of the universe to a planet among others; "up" and "down" lost their cosmic meaning; the sublunary/superlunary distinction collapsed.
• Newtonian (1687). A unified mechanics explained both falling apples and orbiting planets, dissolving the Aristotelian split between terrestrial and celestial physics.
• Chemical (1772–1789). Lavoisier replaced phlogiston with oxygen. Combustion became combination-with-oxygen rather than release-of-phlogiston; mass conservation became central; the elements were redefined.
• Darwinian (1859). Species ceased to be fixed essences; common descent, natural selection, and deep time replaced static creation. Biology acquired a historical dimension.
• Einsteinian (1905–1915). Absolute simultaneity disappeared; mass and energy became interconvertible; gravity was reconceived as spacetime geometry. Newton survived as a low-velocity, low-curvature limit, but its picture of space and time did not.

Worked example: the chemical revolution

Before 1770, chemists explained burning and rusting through phlogiston theory. A combustible body contained phlogiston; burning released it into the air. The theory unified diverse phenomena and made testable predictions: metals should weigh more after burning's reverse (calcination)... no wait, less, because phlogiston had escaped. But careful weighings by Joseph Black and others showed metals gained weight when calcined. A patch was added: phlogiston had negative weight. The patch worked locally but felt unprincipled.

Antoine Lavoisier, working in Paris in the 1770s and 1780s, ran balance-sheet experiments: weigh inputs and outputs of every reaction, including gases captured by inverted jars over water. He showed combustion combined the burning material with a component of air — Priestley's "dephlogisticated air", which Lavoisier renamed oxygène. Mass was conserved. Phlogiston was unnecessary. By 1789 Lavoisier had rewritten chemistry's nomenclature, redefined elements, and produced the Traité élémentaire de chimie. Phlogiston theorists like Priestley and Stahl never converted; they died defending the older view. The revolution was completed by the next generation. This pattern — older scientists holding out, change carried by students — is what Kuhn meant when he quoted Max Planck: "a new scientific truth does not triumph by convincing its opponents... but rather because its opponents eventually die".

Normal science vs revolution

	Normal science	Revolutionary science
Goal	Articulate the paradigm; solve puzzles	Replace the paradigm
Attitude to anomaly	Tolerate, postpone, attribute to apparatus	Treat as decisive
Tools	Established instruments, exemplars	New instruments, new questions
Community structure	Convergent; orthodoxy enforced by training and journals	Divergent; competing schools
Standard of success	Puzzle-solving precision	Promise of broader scope; aesthetic appeal
Duration	Decades to centuries	One scientific generation, typically
Risk profile	Conservative; safe career path	Career-defining gamble

Incommensurability — what it is and isn't

The most explosive claim in Structure is incommensurability. Kuhn argued that successive paradigms are not just inconsistent but incommensurable: their key terms refer differently, their problems are not the same problems, and they share no neutral observation language. When a Newtonian and an Einsteinian both say "mass", they mean subtly different things — the Newtonian's mass is invariant, the Einsteinian's depends on velocity (or, in modern formulation, includes rest mass and relativistic mass as separate concepts). Translation between paradigms is possible but never trans­parent; it is more like translating between two natural languages than mapping a single coordinate system.

Critics seized on this. Hilary Putnam in Reason, Truth and History (1981) and Donald Davidson in "On the Very Idea of a Conceptual Scheme" (1974) argued that thoroughgoing incommensurability is self-undermining: if we can describe Newton and Einstein as different, we already share enough common ground to render them commensurable. Kuhn's later work — especially the 1969 Postscript and his unfinished posthumous book — softened the thesis to "local incommensurability": only a handful of interrelated terms shift meaning at a revolution, leaving the rest of the language stable enough for translation. This is now the mainstream position even among Kuhn's defenders.

Counterargument: Lakatos and Laudan

Imre Lakatos accepted that scientists do not falsify theories instantly but rejected Kuhn's image of revolutions as gestalt switches. His methodology of scientific research programmes (1970) replaced "paradigm" with a more articulated structure: a hard core of inviolable commitments, surrounded by a protective belt of revisable auxiliary hypotheses. Programmes are progressive when their belt absorbs anomalies and predicts novel phenomena; degenerative when the belt offers only ad hoc rescues. Choice between programmes is rational, not a leap of faith — it is grounded in track record over time.

Larry Laudan went further in Progress and Its Problems (1977). He denied that Kuhnian paradigms even exist as the holistic, all-or-nothing entities Kuhn described. Real scientists, Laudan argued, change their theoretical, methodological, and value commitments piecemeal; a "revolution" looks revolutionary only because we focus on a particular axis at a particular moment. The Copernican revolution rolled out over a century and a half, and at any given decade most working astronomers were doing some kind of hybrid normal science.

A different objection comes from sociologists of science influenced by David Bloor's strong programme: Kuhn was insufficiently radical, treating science as eventually rational despite revolutionary breaks. Steve Fuller in Thomas Kuhn: A Philosophical History for Our Times (2000) argued that Structure's legacy was conservative — by depicting normal science as a healthy default, Kuhn licensed precisely the puzzle-solving conformism his book had described.

Variants and successors

• Disciplinary matrix — Kuhn's preferred technical replacement for "paradigm" in his 1969 Postscript: the bundle of symbolic generalizations, metaphysical commitments, values, and exemplars that holds a community together.
• Research programmes (Lakatos) — a hard core plus a protective belt; assessed as progressive or degenerative over time.
• Research traditions (Laudan) — looser than paradigms, allowing piecemeal change.
• Themata (Holton) — recurrent metaphysical and aesthetic motifs that thread through revolutions (continuum vs discreteness, simplicity, symmetry).
• Episteme (Foucault) — a related but more sweeping notion in The Order of Things (1966): the overall epistemological framework of an era, including but not limited to its science.
• Punctuated equilibrium (Eldredge and Gould 1972) — a Kuhn-influenced model of biological evolution: long stasis broken by rapid speciation.

Common confusions

• "Paradigm shift" is not just a big idea. A new product, a clever marketing slogan, or a generational fashion is not a Kuhnian paradigm shift. The term targets shared community-binding frameworks, not isolated breakthroughs.
• Normal science is not bad science. Kuhn meant it descriptively, not pejoratively. Normal science is what produces most of the actual knowledge in physics, chemistry, and biology.
• Revolutions are not irrational. Kuhn argued that paradigm choice involves more than evidence — it involves values like simplicity, scope, and fruitfulness — but these are themselves rational considerations, not whims.
• Incommensurability is not untranslatability. Kuhn always allowed translation; he insisted only that it is hard, partial, and historically situated, like translating Homer.
• Not every field is paradigmatic. Kuhn explicitly thought the social sciences were still pre-paradigmatic — riven by competing schools without a unifying framework. Whether psychology or economics has yet matured into normal science is still debated.