Why Science Only Looks Cumulative: Kuhn, Textbooks, and the Problems That Simply Disappear — Epoche B2
Open almost any science textbook and you will find the same comforting story: knowledge grows brick by brick, with each generation adding established facts on top of earlier ones. This picture feels obviously true. Yet Thomas Kuhn, in The Structure of Scientific Revolutions (1962), argued that it is largely an artefact of how the story is told — and told, crucially, by the winning side. Following the Continental habit of thinking in thesis, antithesis and synthesis, let us test this claim dialectically. Since the method is a tool and not a decoration, a word on what it means: first state the strongest honest case for the received view (the thesis), then the strongest case against it (the antithesis), and finally keep whatever survives the collision (the synthesis). The exercise is only useful if each side gets its best weapons; refuting a weak version of an opponent proves nothing. Thesis: science accumulates The common view has real strength. Techniques, instruments and predictive power plainly do accumulate. We can calculate planetary orbits better than Newton could — and "better" here is measurable. By 1859 the French astronomer Urbain Le Verrier had shown that the slow rotation of Mercury's elliptical orbit outruns the Newtonian prediction by a small residue, about 43 seconds of arc per century (a second of arc is 1/3600 of a degree, so the discrepancy is roughly a hundredth of a degree accumulated over a hundred years). Einstein's general relativity derived exactly that residue in 1915, with no adjustable constant. Notice what the example shows: the new theory did not discard the old predictions; it reproduced them and then accounted for the leftover. Nor did we lose the ability to build telescopes when relativity replaced classical mechanics. Karl Popper turned this observation into a philosophy of growth. His starting point was a rejection of induction — the everyday habit of arguing from repeated experience to a general law, as in "every swan observed so far has been white, therefore all swans are white". No number of past confirmations can logically guarantee the next case, so, Popper concluded, science cannot rest on induction at all. What logic does license is the reverse inference: if a theory implies a prediction and the prediction fails, the theory is false — the logicians' modus tollens , which is deductively watertight. Science therefore proceeds, for Popper, by conjectures and refutations: bold guesses followed by severe attempts to prove them wrong. Growth happens when a refuted theory is replaced by a better conjecture, and "better" had a definite meaning for him: the successor must explain everything its predecessor got right, forbid more, and survive tests the old theory failed — it must exceed the old theory in what he called truth content, the stock of true consequences flowing from it. Relativity qualifies because Newtonian mechanics drops out of it as a limiting case: at speeds small compared with light, the relativistic equations reduce to Newton's, which is precisely why the old successes are preserved. Popper thus agreed that growth proceeds through revolutionary overthrow rather than simple addition; but on his picture each overthrow leaves us objectively closer to the truth. Antithesis: paradigms set the questions Kuhn's objection cuts deeper than the usual complaint that scientists sometimes make mistakes. A paradigm — his term for the shared framework a scientific community absorbs in its training — supplies not only answers but the very standards for what counts as a legitimate question. A paradigm includes worked example problems, approved instruments, mathematical techniques, and unstated assumptions about what kinds of explanation are acceptable at all. Most research, which Kuhn calls normal science, is puzzle-solving inside that framework; the framework itself is no more on trial during ordinary research than the rules of chess are on trial during a game. Consider gravity, and let us unpack the example slowly. In the physics descended from René Descartes, dominant on the Continent through the late seventeenth century, the universe is a plenum : space is completely full of matter, with no void anywhere, and all influence is by contact — one body pushing another, as billiard balls do. On this view the planets are carried round the Sun by vortices, vast whirlpools of fine invisible matter, much as straws are carried round the drain of an emptying basin. The demand behind the picture was explicit. "Action at a distance" — one body affecting another across empty space with nothing passing between — seemed no better than magic, so any respectable theory of gravity had to supply a contact mechanism, a story about pushes. Newton's Principia (1687) supplied something else entirely: a law. Every two masses attract with a force $$F = G\,\frac{m_1 m_2}{r^2},$$ where $m_1$ and $m_2$ are the masses, $r$ is the distance between them, and $G$ is a universal constant. The law states exactly how strong the attraction is and stays silent about what carries it. From it Newton derived the elliptical orbits of the planets, the tides and the fall of the apple in one stroke — a predictive unification without precedent. Pressed for the mechanism, he answered in the General Scholium added to the 1713 edition: hypotheses non fingo , "I feign no hypotheses" about the cause. To Cartesian eyes this was not modesty but scandal. Leibniz, in his correspondence of 1715–16 with Newton's defender Samuel Clarke, charged that attraction without a mechanism revived the "occult qualities" of medieval philosophy — the scholastic label for a hidden power that merely renames what it should explain, as in Molière's doctor who explains why opium causes sleep by citing its "dormitive virtue". By the standards of the reigning paradigm, the criticism was exactly right. What happened next is Kuhn's point. The demand was never met: no contact mechanism for gravity was ever found. Instead, within two