What distinguishes a scientific theory from a myth, an ideology, or a metaphysical system? Why do we accept as knowledge what physics tells us, but not what astrology — or psychoanalysis, at least in its more speculative form — claims? These questions, simple at first glance, opened in the twentieth century one of the most consequential debates in contemporary philosophy. Its two central figures — Karl Popper and Thomas Kuhn — formulated answers that, far from reconciling, redrew the landscape of epistemology.
The starting point: the problem of demarcation
Modern philosophy of science is born of a concrete unease. At the beginning of the twentieth century, three great theories competed for attention in Vienna: Einstein’s relativity (1905, 1915), Marxism as a science of history, and Freud’s psychoanalysis. The young Karl Popper (1902–1994) perceived a qualitative difference between them. Relativity made risky predictions — in 1919, Arthur Eddington’s observation confirmed the deflection of starlight predicted by Einstein, but the prediction could have failed, and in that case the theory would have been refuted. Marxism and psychoanalysis, by contrast, seemed to confirm themselves everywhere: any fact could be interpreted in their favor.
This observation crystallized into the problem of demarcation: what separates science from non-science?
The dominant answer until then was inductivism: science distinguishes itself by being based on repeated observations that justify generalizations. This was the classical position of Francis Bacon, Stuart Mill, and the Vienna Circle (Schlick, Carnap, Neurath) — logical positivism, which reduced science to propositions verifiable by experience. Popper rejected this path head-on.
Popper: falsificationism
The critique of induction
Popper’s starting point is a rereading of the problem of induction as formulated by David Hume. However many times we have observed the sun rise, nothing logically guarantees that it will rise tomorrow. No finite number of observations justifies a universal proposition. All swans are white is a generalization that centuries of European observation seemed to confirm — and was refuted by a single black swan sighted in Australia in 1697.
The lesson is asymmetric: an infinite number of confirmations never proves a universal theory; a single refutation falsifies it. Popper rests his philosophy on this asymmetry between confirmation and refutation.
The central thesis
In The Logic of Scientific Discovery (Logik der Forschung, 1934), Popper proposes:
A theory is scientific if, and only if, it is falsifiable — that is, if it is possible to formulate an experiment or observation that, were it to occur, would refute it.
Falsifiability does not mean the theory is false. It means it takes risks: it makes predictions that can go wrong. A theory that takes no risks — that fits any fact — is not scientific; it is dogmatic.
Consequences
Conjectures and refutations (Conjectures and Refutations, 1963). Science does not proceed by accumulating truths, but by proposing bold conjectures and systematically attempting to refute them. Theories that survive testing are corroborated — not verified. Corroboration is provisional.
Approximate truth. Popper rejects radical skepticism: scientific theories, although not demonstrable, can progressively approach the truth. The notion of verisimilitude tries to account for this asymptotic approximation.
Critique of Marxism and psychoanalysis. For Popper, both are paradigmatic examples of unfalsifiable theories: any fact can be reinterpreted in their terms. This does not make them false, but excludes them from the domain of science. (The Popperian critique of psychoanalysis is today contested; in its more sophisticated form — Grünbaum — it does formulate falsifiable hypotheses.)
Open society. Popperian epistemology has a political correlate: just as in science we reject falsified theories, in politics we must build institutions that allow us to remove rulers without violence. This is the thesis of The Open Society and Its Enemies (1945), a polemic against Plato, Hegel, and Marx.
Kuhn: paradigms and scientific revolutions
The shock with history
Thomas Kuhn (1922–1996) was trained as a physicist. When, in 1947, he was tasked with preparing a course on the history of science for humanities students at Harvard, he immersed himself in Aristotle’s writings on physics. The result was traumatic: Aristotelian physics, read from the standpoint of modern physics, seemed absurd. But read on its own terms, with its own presuppositions, it revealed itself as coherent, sophisticated, and adequate to a certain slice of experience. The “irrationality” disappeared.
This hermeneutic experience underlies The Structure of Scientific Revolutions (1962), one of the most influential books of the twentieth century (over 1.4 million copies sold and translations into more than twenty languages).
Paradigm
The central concept is the paradigm. In its more technical sense (the 1969 postscript distinguishes two), a paradigm is a disciplinary matrix: an integrated set of laws, models, epistemic values, exemplars of solved problems, and practices shared by a scientific community. Learning a science is not just learning its theories, but internalizing a paradigm — often through the exemplars (model problems) that appear in textbooks.
Historical examples of paradigms: Aristotelian physics, Ptolemaic astronomy, Newtonian mechanics, Lavoisier’s chemistry, Einsteinian relativity.
The phases of scientific development
Kuhn describes a cycle:
Normal science: most of the time, scientists operate within a shared paradigm, solving “puzzles” (puzzle-solving). They do not test the paradigm — they operate inside it.
Anomalies: certain facts persistently resist explanation by the paradigm. At first, they are ignored or treated as technical problems to be resolved later.
Crisis: the accumulation of anomalies produces diffuse dissatisfaction. A plurality of ad hoc solutions weakens the paradigm.
Scientific revolution: an alternative paradigm emerges that reinterprets the problematic material in a new light. The choice between paradigms involves not only empirical reasons, but aesthetic considerations, simplicity, and heuristic promise.
New paradigm → new normal science: the community adopts the new paradigm and the cycle begins again.
Canonical examples: the Copernican revolution, Lavoisier’s chemistry replacing phlogiston, Einsteinian relativity superseding Newtonian mechanics as fundamental theory.
The incommensurability thesis
The most controversial consequence is the incommensurability between paradigms. Two rival paradigms do not share all meanings, all observations, or all criteria of evaluation. “Mass” in Newton and “mass” in Einstein do not signify exactly the same thing. There is no neutral third standpoint from which to choose between paradigms by purely logical means.
This does not mean that the choice between paradigms is irrational — Kuhn always rejected that reading. It means it mobilizes a broader range of reasons than the Popperian model allows.
The great debate: London, 1965
In July 1965, at the International Colloquium on the Philosophy of Science in London, Popper, Kuhn, Imre Lakatos, Paul Feyerabend, and others met for a direct confrontation. The debates were published in Criticism and the Growth of Knowledge (Lakatos & Musgrave, eds., 1970), a volume that became a classic.
Popper’s objections to Kuhn
For Popper, Kuhnian “normal science” — in which scientists operate within a paradigm without questioning it — describes not genuine science, but a degraded form of professional practice. Genuine science is always at the critical-revolutionary margins.
Kuhn’s objections to Popper
For Kuhn, falsificationism is normatively attractive but historically false. Scientists do not abandon theories in the face of every anomaly — and this is not irrational: isolated anomalies rarely undo a productive paradigm.
Lakatos: research programs
Imre Lakatos (1922–1974) proposed a synthesis. The unit of scientific evaluation is neither the isolated theory (Popper) nor the global paradigm (Kuhn), but the research program: a hard core of fundamental hypotheses protected by a protective belt of adjustable auxiliary hypotheses. Programs are progressive when their modifications generate new and corroborated predictions, and regressive when they merely save themselves from refutation ad hoc.
Feyerabend: methodological anarchism
Paul Feyerabend (1924–1994), in Against Method (1975), radicalized Kuhn: no single method — neither falsificationism nor any other — adequately describes the historical success of science. Galileo, Newton, Einstein all violated methodological rules supposed to be inviolable. Feyerabend’s famous motto: “anything goes” — not as apology for arbitrariness, but as recognition of the real plurality of science’s paths.
Where philosophy of science went
The Popper–Kuhn–Lakatos–Feyerabend debate was not resolved; it was superseded by new problems. Contemporary philosophy of science operates on several fronts:
- Scientific realism vs. anti-realism: do theories speak of real entities (electrons, genes) or are they merely predictive instruments? Bas van Fraassen, Ian Hacking, Stathis Psillos.
- Models and idealizations: how do scientific models — often false, simplifications — represent the world? Nancy Cartwright, Margaret Morrison.
- Philosophy of scientific practice: focus shifted from theory to experimentation, instrumentation, and the social dynamics of laboratories. Ian Hacking (Representing and Intervening, 1983), Bruno Latour.
- Social epistemology: communities of scientists as collective agents of knowledge. Helen Longino, Miriam Solomon.
- Philosophy of particular sciences: of physics, of biology, of economics — each with its own problems.
Why this debate still matters
Today, in public debate, the Popperian and Kuhnian theses appear all the time — often misused. Whoever invokes “the scientific consensus” frequently presupposes something like Kuhn’s normal science. Whoever speaks of “science as permanent doubt” and “questioning everything” presupposes a popularized Popper. Understanding the debate in its rigorous form allows us to use these categories with precision.
In a broader sense, Popper’s question — what separates serious knowledge from speculation? — and Kuhn’s question — how does knowledge change, in fact, throughout history? — are complementary. The first belongs to the logic of science; the second, to its history. A mature philosophy of science does not choose between them: it articulates them.
← Articles