Karl Popper: Theory of Falsification

Theory of Falsification

Karl Popper is prescriptive: he describes what science should do, not how it actually behaves. Popper was a rationalist. He held that the central question in the philosophy of science was distinguishing science from non-science.

Replacing Induction with Falsification

In The Logic of Scientific Discovery, Popper emerged as a major critic of inductivism, an approach he saw as fundamentally outdated. Popper replaced the classical observationalist-inductivist account of the scientific method with falsification, that is, deductive logic, as the criterion for distinguishing scientific theory from non-science.

He argued that science progresses best through deductive reasoning. He called this approach critical rationalism.

The Problem of Induction

All inductive evidence is limited: we do not observe the universe at all times and in all places. We are not justified, therefore, in making a general rule from this observation of particulars.

This is the problem of induction.

David Hume sharpened it in his 1748 Enquiry Concerning Human Understanding: no number of past instances can logically guarantee the next one will follow the same pattern.

According to Popper, a scientific theory must make testable predictions. It should be rejected if those predictions turn out wrong.

Popper gives the following example:

Europeans, for thousands of years had observed millions of white swans. Using inductive evidence, we could come up with the theory that all swans are white.

However, exploration of Australasia introduced Europeans to black swans. Popper’s point is this: no matter how many observations confirm a theory, a future observation could always refute it. Induction cannot yield certainty.

This led Popper to propose falsification as an alternative scientific method: many confirming instances can exist for a theory, but it only takes one counter-observation to falsify it. Science progresses when a theory is shown wrong and a new one better explains the phenomena.

Theory-Laden Observation

Karl Popper was also critical of the naive empiricist view that we objectively observe the world. He argued that all observation happens from a point of view.

It is shaped by the understanding we bring to it. The world appears to us in the context of theories we already hold: it is ‘theory-laden.’

For Popper, the scientist should try to disprove a theory rather than continually try to prove it. Science can help us progressively approach the truth, Popper thought, but we can never be certain we have the final explanation.

Popper’s Critique of Psychoanalysis and Marxism

Popper’s most famous application of falsification was his verdict on Freudian psychoanalysis, Adlerian psychology and Marxist theories of history. He argued that none of them counted as science, not because they were false, but because no observation could ever count against them (Popper, 1963).

Freud, Adler and Marx as Pseudoscience

Popper’s touchstone case of a theory exposing itself to genuine risk was Einstein’s general theory of relativity, tested against the sky itself during the 1919 solar eclipse. The stakes were real.

Aim: General relativity predicted a precise bending of starlight near the Sun, larger than Newtonian gravitation predicted for the same configuration. The theory would stand refuted if the observed shift did not match it.

The test was severe.

Method: During the total eclipse of 29 May 1919, Royal Astronomical Society and Royal Greenwich Observatory teams photographed stars near the darkened Sun from Sobral, Brazil, and Príncipe, West Africa. They compared the positions against photographs of the same stars taken months earlier (Dyson et al., 1920).

Findings: The measured deflection matched Einstein’s value far more closely than Newton’s, decisively at Príncipe and, within a larger margin of error, at Sobral (Dyson et al., 1920).

Conclusion: The theory made a specific, risky prediction in advance. An observation could have refuted it. It did not (Popper, 1963).

Freud, Adler and Marx impressed their followers for the opposite reason: their theories explained everything.

A Freudian could read any behavior as evidence of an unconscious fixation. Reaction formation, the idea that a fixation can surface as its own opposite, explained away whichever way it pointed.

Adler and Marx worked the same trick.

An Adlerian could read any act as compensation for feelings of inferiority, and a Marxist could reinterpret any historical event as confirming the coming revolution (Popper, 1963).

For Popper, this apparent strength was the real weakness. A theory that fits every observation forbids nothing, and a theory that forbids nothing says nothing. Hans Eysenck later carried the same argument into clinical psychology, arguing psychoanalysis had never been shown to outperform no treatment at all (Eysenck, 1985).

The Grünbaum Challenge

Not everyone accepts Popper’s verdict. The philosopher Adolf Grünbaum argued in The Foundations of Psychoanalysis: A Philosophical Critique that Popper was factually wrong about Freud (Grünbaum, 1984).

Freud, Grünbaum showed, made specific testable predictions about the causes of neurosis, the formation of dreams and the effects of therapy. He accepted that contrary evidence would refute them.

Grünbaum’s real objection was different: psychoanalysis is falsifiable, but its specific predictions mostly fail when tested against clinical evidence. Whether Freud’s claims are testable in principle and whether they are well supported are two separate questions, and Popper’s argument answered only the first.

The debate has never fully settled. Popper’s charge remains the standard exam-board criticism of psychodynamic theories.

Some psychologists respond that only some psychoanalytic claims are untestable and that others, including how early experience shapes adult behavior, are open to empirical study. A contemporary research programme called neuropsychoanalysis continues that project by testing Freudian concepts against neuroscience.

Critical Evaluation

Popper’s first major contribution to philosophy was his novel solution to the problem of demarcating science from non-science.

Science, on the time-honored view, is distinguished by its inductive method: it relies on observation and experiment rather than logical analysis alone. But no run of favorable data, however long and unbroken, is logically sufficient to establish the truth of an unrestricted generalization.

Popper’s formulations helped rein in excessive inductive speculation. They also strengthened the conceptual foundation for today’s peer review process.

Yet history tells a messier story.

The history of science gives little indication of having actually followed a strict methodological falsificationist approach. Scientists, past and present, often resist giving up a theory even after it counts as falsified in the strict sense. Looking back, they were frequently right to hold on.

Early Newtonian gravitational theory, for example, appeared falsified by observations of the moon’s orbit, yet the anomaly was eventually explained without abandoning the theory. One observation rarely settles the matter: the experiment may have been badly designed, or the data itself could be wrong.

Quine made the point systematically. A theory is never a single statement; it is a complex network of statements.

You might falsify one strand in that network (e.g., “all swans are white”) without needing to reject the whole complex theory.

The Kuhn and Lakatos Challenge

Thomas Kuhn mounted the most influential challenge to falsification. In The Structure of Scientific Revolutions, Kuhn argued that scientists rarely abandon a theory after one failed prediction (Kuhn, 1962). Instead, they work inside a shared paradigm, treating anomalies as puzzles to be solved rather than refutations to accept.

Imre Lakatos offered a middle path. His methodology of scientific research programmes described scientists defending a theory’s hard core while adjusting a protective belt of assumptions around it (Lakatos, 1970). A programme counts as healthy only while those adjustments keep generating successful new predictions, not just excuses for old failures.

Paul Feyerabend pushed further still, arguing that no single method, Popperian or otherwise, actually describes how science advances (Feyerabend, 1975).

Together, Kuhn, Lakatos and Feyerabend forced philosophers to treat falsification as one useful norm among several, not the whole story of scientific method.

Contemporary Research

Paul Meehl’s 1978 paper on soft psychology previewed today’s replication debate decades early. Meehl argued that psychology’s null-hypothesis tests set the bar for confirmation far too low: a prediction could succeed even when the underlying theory was false (Meehl, 1978). The test came three decades later.

Testing Meehl’s Warning

Aim: The Open Science Collaboration (2015) set out to give a direct, large-scale estimate of how often published psychological findings replicate.

Method: A crowdsourced team closely replicated 100 studies from three major 2008 psychology journals, comparing each replication’s effect size and significance to the original.

The result was stark.

Results: Ninety-seven percent of the original studies had reported a significant effect; only thirty-six percent of the replications did, with effect sizes roughly halved.

Conclusion: Researchers read the shortfall as exactly the failure of severe testing that Popper’s language of bold, risky conjecture had anticipated (Open Science Collaboration, 2015).

The Reform Response

Marcus Munafò and a coalition of methodologists responded with a manifesto for reforming the field’s incentives (Munafò et al., 2017). They traced low replication rates to a system that rewards novel, positive findings over accurate ones. Their fixes echo Popper’s own ideal.

Pre-register hypotheses before seeing the data, publish registered reports that review the design rather than the result, and treat replication as a normal, rewarded part of the process.

Who Was Karl Popper?

Karl Raimund Popper (1902–1994) was an Austrian-born philosopher who spent most of his career in Britain. Before developing his theory of falsification, he trained in mathematics, physics and psychology in Vienna and briefly taught in its state schools.

Vienna and the Vienna Circle

Popper was born in Vienna on 28 July 1902, into an assimilated Jewish family that had already converted to Lutheran Christianity. His father, a lawyer and bibliophile, kept a private library of some ten thousand volumes on philosophy, law and social science.

Books and music filled the house. His musical mother left Popper a lifelong amateur pianist and composer.

He left school without finishing his final exams. He apprenticed briefly as a cabinet-maker and taught at street schools for underprivileged children in the reforming climate of postwar ‘Red Vienna’.

He then studied mathematics, physics, psychology and philosophy at the University of Vienna, mostly as an unregistered visitor. He completed a PhD there in 1928 under the Gestalt psychologist Karl Bühler, with a dissertation on method in the psychology of thinking.

Vienna was intellectually crowded.

The city exposed him to rival movements: the Vienna Circle’s logical positivists, Freudian psychoanalysis, Adlerian individual psychology and a vigorous Marxist workers’ movement. He briefly considered himself a Marxist as a teenager and volunteered in one of Alfred Adler’s child-guidance clinics.

That experience stayed with him.

He attended Vienna Circle meetings without ever being admitted as a member, and later called himself ‘the official opposition’ to logical positivism. His disillusionment with both movements, their apparent ability to explain any evidence and its opposite, became the seed of his falsifiability criterion.

Emigration, LSE and Later Life

As Austrian politics darkened, Popper emigrated with his wife Josefine ‘Hennie’ Henninger to New Zealand in 1937, months before the Anschluss. He took a philosophy lectureship at Canterbury University College.

The years there were personally isolating, much of his extended family later perished in the Holocaust, but philosophically productive. It was in Christchurch that he wrote The Open Society and Its Enemies (1945), his defence of piecemeal social reform against Plato, Hegel and Marx.

He also wrote The Poverty of Historicism (1957) there.

The latter was his attack on the idea that history unfolds by inexorable laws. In 1946 Friedrich Hayek recruited him to the London School of Economics, where he became Professor of Logic and Scientific Method in 1949 and remained until his retirement in 1969.

His doctoral students there, including Imre Lakatos and Paul Feyerabend, carried his ideas into a genuine school of thought.

LSE became his home.

He was elected a Fellow of the Royal Society in 1976, knighted by Queen Elizabeth II in 1965, and appointed a Companion of Honour in 1982.

He died in Croydon, south London, on 17 September 1994, aged 92; his autobiography, Unended Quest, remains the fullest first-person account of his Viennese formation.

References

Dyson, F. W., Eddington, A. S., & Davidson, C. (1920). IX. A determination of the deflection of light by the sun’s gravitational field, from observations made at the total eclipse of May 29, 1919. Philosophical Transactions of the Royal Society of London. Series A, 220, 291–333. https://doi.org/10.1098/rsta.1920.0009

Eysenck, H. J. (1985). Decline and fall of the Freudian empire. Viking.

Feyerabend, P. K. (1975). Against method: Outline of an anarchistic theory of knowledge. New Left Books.

Grünbaum, A. (1984). The foundations of psychoanalysis: A philosophical critique. University of California Press.

Hume, D. (1748). Philosophical essays concerning human understanding. A. Millar.

Kuhn, T. S. (1962). The structure of scientific revolutions. University of Chicago Press.

Lakatos, I. (1970). Falsification and the methodology of scientific research programmes. In I. Lakatos & A. Musgrave (Eds.), Criticism and the growth of knowledge (pp. 91–196). Cambridge University Press.

Meehl, P. E. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft psychology. Journal of Consulting and Clinical Psychology, 46(4), 806–834. https://doi.org/10.1037/0022-006X.46.4.806

Munafò, M. R., Nosek, B. A., Bishop, D. V. M., Button, K. S., Chambers, C. D., Percie du Sert, N., Simonsohn, U., Wagenmakers, E.-J., Ware, J. J., & Ioannidis, J. P. A. (2017). A manifesto for reproducible science. Nature Human Behaviour, 1(1), Article 0021. https://doi.org/10.1038/s41562-016-0021

Open Science Collaboration. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), Article aac4716. https://doi.org/10.1126/science.aac4716

Popper, K. R. (1959). The logic of scientific discovery. Hutchinson. (Original work published as Logik der Forschung, 1935)

Popper, K. R. (1963). Conjectures and refutations: The growth of scientific knowledge. Routledge & Kegan Paul.