The Logic of Scientific Discovery

Summary

The Logic of Scientific Discovery is Karl Popper's account of what makes a theory scientific and how scientific knowledge grows. It first appeared in Vienna in 1934, with the imprint date 1935, as Logik der Forschung, in a series edited by Moritz Schlick and Philipp Frank of the Vienna Circle. Popper, then a secondary-school teacher, had cut it down from a much longer manuscript. The English version, translated by Popper with the help of Julius Freed and Lan Freed, was published in London by Hutchinson in 1959 with many new footnotes and appendices. The book remains under copyright, and no full text is included in the library. Popper begins with two problems. The problem of induction, posed by Hume, is that no number of observations can justify a universal statement: however many white swans we have seen, it does not follow that all swans are white. The problem of demarcation is how to distinguish the empirical sciences from mathematics, logic and metaphysics. Popper's answer to both is the same. Science does not proceed by induction but by the deductive testing of bold conjectures. A theory is empirical only if it forbids something, so that some possible observation could contradict it. Verification can never be conclusive, but a single accepted counter-instance can refute a universal law. Falsifiability, not verifiability, is therefore the criterion of demarcation, and Popper insists that it marks off science from other kinds of thought without declaring them meaningless, as some members of the Vienna Circle held. The central chapters develop this view. Methodological rules are proposals, chosen so as to keep theories exposed to refutation; a scientist who saves a theory by ad hoc assumptions or redefinitions, which Popper calls conventionalist stratagems, has left the empirical method. Observation reports, or basic statements, are themselves accepted by decision and could be tested further, so science has no absolute foundation: it rises above a swamp on piles driven in far enough to bear the structure for the time being. Theories can be compared by their degree of falsifiability, and the better theory says more and so risks more; Popper identifies simplicity with this testability. Long chapters treat probability, which he interprets in frequency terms, and the interpretation of quantum mechanics. The final chapter argues that theories are never shown to be true or even probable, but can be corroborated by surviving severe tests. The book made falsifiability one of the best-known ideas in the philosophy of science, and it has been extensively discussed and criticized by Kuhn, Lakatos, Feyerabend and others.

Plot or Argument

• Chapter I: a survey of fundamental problems

  Popper rejects inductive logic and the view that science begins with observation. He separates the psychology of how ideas arise from the logic of how they are tested, describes the deductive testing of theories, states the problem of demarcation, and proposes falsifiability as its solution. He also introduces the problem of the empirical basis and the requirement that scientific statements be intersubjectively testable.

• Chapter II: the theory of scientific method

  Methodological rules are conventions, the rules of the game of science, chosen to serve its aim. The supreme rule is that no statement should be protected from falsification. Popper contrasts his view with naturalist accounts of method and with positivist attempts to eliminate metaphysics by a criterion of meaning.

• Chapters III–IV: theories and falsifiability

  Theories are universal statements, nets cast to catch the world. Popper distinguishes strictly universal from numerical statements, explains the logic of causal explanation, and shows how a theory's falsifiability depends on the class of basic statements it prohibits. He identifies the conventionalist stratagem, by which any theory can be saved from refutation, and lays down rules against ad hoc auxiliary hypotheses.

• Chapter V: the problem of the empirical basis

  Perceptual experiences cannot justify statements; only statements can. Basic statements, which report observable events at a particular place and time, are accepted by decision of the scientific community and remain open to further test. Science therefore rests on no bedrock, but on piles driven into a swamp.

• Chapters VI–VII: degrees of testability and simplicity

  Theories can be compared by the classes of their potential falsifiers; a theory with more empirical content prohibits more and is easier to refute. Popper argues that science should prefer theories of high content and that the simplicity valued by scientists is best understood as degree of falsifiability.

• Chapters VIII–IX: probability and quantum theory

  Probability statements cannot be strictly falsified, so Popper proposes methodological rules under which they can be tested. He develops a frequency interpretation of probability and applies it to quantum mechanics, criticizing the usual reading of Heisenberg's uncertainty relations and proposing a statistical interpretation.

• Chapter X: corroboration

  A theory that survives severe tests is corroborated, but corroboration is not probability: theories of high content are improbable, and science prefers them. Our knowledge remains conjectural; the aim of science is not certainty but the bold search for truth by trial and the elimination of error.

• Appendices of 1959

  The English edition adds the original appendices and a series of new starred appendices on probability, corroboration, the propensity interpretation and related technical questions, along with starred footnotes in which Popper comments on his earlier text.

Characters

• Karl Popper, author

  Viennese schoolteacher and philosopher writing in critical dialogue with the Vienna Circle.

• David Hume, source of the problem

  Whose argument that induction cannot be logically justified Popper accepts and seeks to answer differently.

• Rudolf Carnap and Moritz Schlick, Vienna Circle interlocutors

  Logical positivists whose verifiability criterion of meaning Popper opposes with a criterion of demarcation.

• Albert Einstein, exemplar

  Whose theories Popper treats as bold, highly testable conjectures; Einstein read and praised the 1934 book.

• Werner Heisenberg, interlocutor

  Whose interpretation of the uncertainty relations Popper discusses in the chapter on quantum theory.

Lessons

• A theory must be able to fail

  Popper argues that a scientific theory says something about the world only if it forbids something, so that some observation could prove it wrong. A theory compatible with every possible event explains nothing. The willingness to be refuted is the mark of empirical science.

  • Why does Popper think that no number of white swans can prove that all swans are white, while one black swan can disprove it?
  • What does it mean to say that a theory forbids certain events?
  • Choose a claim you have heard. What observation would show it to be false?

• Do not rescue a theory by special pleading

  Any theory can be saved from refutation by adding assumptions invented only for that purpose or by changing the meaning of its terms. Popper calls these conventionalist stratagems and proposes rules against them, because a theory protected in this way no longer tells us anything about the world.

  • What is an ad hoc hypothesis, and why does Popper reject it?
  • Are there cases in which adding a new assumption to save a theory turned out to be right?
  • How might the same temptation appear in arguments outside science?

• Knowledge without bedrock

  Even observation reports are accepted by decision and could be tested further. Popper concludes that science stands on piles driven into a swamp: firm enough to carry the structure for now, but not resting on any final foundation. Knowledge is conjectural, though it can still grow.

  • Why does Popper think that experiences cannot justify statements?
  • Explain the image of the piles in the swamp in your own words.
  • Does Popper's view make science less trustworthy, or does it explain why science can correct itself?

• Prefer bold theories

  Popper argues that theories which say more, and therefore risk more, are the better candidates for science. A highly improbable theory with great content that survives severe tests is worth more than a safe theory that says little.

  • Why does Popper say that a theory with more content is less probable?
  • What does he mean by corroboration, and how does it differ from proof?
  • How might Kuhn or Lakatos answer the claim that scientists should seek to refute their theories?

Context

In the 1920s and 1930s Vienna was a centre of new philosophy of science. The Vienna Circle, gathered around Moritz Schlick and including Rudolf Carnap and Otto Neurath, held that meaningful statements must be verifiable by experience and sought to build a logic of induction. Popper, who was not a member, debated with them in person and in print. He later wrote that he had been impressed in 1919 by the contrast between Einstein's general theory, which made a risky prediction that could have failed, and other theories of the time that seemed able to accommodate any evidence; that contrast lies behind the book's criterion of demarcation.

Composition

Popper first wrote a long manuscript, Die beiden Grundprobleme der Erkenntnistheorie (The Two Fundamental Problems of the Theory of Knowledge), which was published only in 1979. At the publisher's request he shortened it drastically for Logik der Forschung, issued by Julius Springer in Vienna in the series Schriften zur wissenschaftlichen Weltauffassung. After the war he prepared an English version with the help of Julius Freed and Lan Freed, published by Hutchinson in 1959 with new footnotes and appendices; later German and English editions added further material.

Reception

The book was discussed within the Vienna Circle on its appearance, and Einstein wrote to Popper approving of it, though the German edition reached a limited audience. The English edition of 1959 made it widely influential among scientists as well as philosophers. Its central claims have been debated ever since: Thomas Kuhn argued that working scientists rarely seek to refute their basic theories, Imre Lakatos proposed a refined falsificationism of research programmes, Paul Feyerabend rejected any fixed method, and others pointed out, following Duhem and Quine, that a failed prediction refutes a whole set of assumptions rather than a single hypothesis. Defenders reply that Popper anticipated several of these objections, and falsifiability remains a widely used test of whether a claim is empirical.

Comparisons

• mill-system-of-logic-book-iii

  Mill offers the classic nineteenth-century logic of induction, grounding it on the uniformity of nature; Popper denies that induction can be justified at all and replaces it with deductive testing, though both discuss why a single instance can refute a universal claim.

• an-enquiry-concerning-human-understanding

  Hume's argument that inference from experience cannot be justified by reason is the problem Popper accepts as unanswerable on its own terms and answers by rejecting induction as the method of science.

• einstein-relativity

  Einstein's statement that the general theory would be untenable if the red shift were not found is the kind of risky prediction Popper takes as the mark of science.

• kuhn-the-structure-of-scientific-revolutions

  Kuhn's account of normal science, in which scientists work within a paradigm and do not seek to refute it, is the most influential challenge to Popper's picture; the two met in debate in London in 1965.

• popper-the-open-society-and-its-enemies

  Popper's political book applies the method of conjecture and refutation to society, preferring piecemeal reform that can be tested and corrected to comprehensive blueprints.

• novum-organum

  Bacon hoped to build science on careful induction from observations; Popper explicitly rejects this inductive ideal, though he shares Bacon's stress on the instructive force of negative instances.

Quotes

For Students

A reading companion for classical Christian high school students. See the student guide.

Reading Guide

Watch for

• The two problems Popper names at the outset: induction and demarcation.
• The asymmetry between verification and falsification, illustrated by the white swans.
• The rules against conventionalist stratagems and ad hoc hypotheses.
• The image of the swamp and the piles at the end of the chapter on the empirical basis.
• The difference between corroboration and probability in the final chapter.

Key terms and names

Problem of induction

The question, raised by Hume, of whether inferences from observed cases to universal laws can be logically justified.

Problem of demarcation

The question of how to distinguish the empirical sciences from mathematics, logic and metaphysics.

Falsifiability

The property of a theory that some possible observation statement would contradict it; Popper's criterion of empirical science.

Basic statement

A singular statement reporting an observable event at a particular place and time, used to test theories.

Conventionalist stratagem

Any move, such as an ad hoc hypothesis or a redefinition, that saves a theory from refutation at the cost of its empirical content.

Corroboration

The degree to which a theory has stood up to severe tests, which Popper distinguishes from its probability.

Seminar Questions

1. Comprehension What is the problem of induction, and why does Popper think it cannot be solved by an inductive logic?
2. Comprehension How does Popper distinguish falsifiability as a criterion of demarcation from verifiability as a criterion of meaning?
3. Comprehension What is a basic statement, and how does Popper say basic statements are accepted?
4. Interpretation Why does Popper think a theory that is consistent with every possible observation is a weak theory rather than a strong one?
5. Interpretation Popper calls methodological rules conventions. Does that make the method of science arbitrary?
6. Interpretation How does the image of the swamp change what it means to call scientific knowledge reliable?
7. Big Question Can a belief be rational if it can never be proved, only tested and not yet refuted?
8. Big Question Popper says metaphysical ideas are not science but are not meaningless. What place might such ideas, including religious beliefs, have alongside science?

Faith Connections

• 1 Thessalonians 5:21 (Scripture)

  'Prove all things; hold fast that which is good' commends testing claims before trusting them, a habit Popper builds into his account of science.

• Deuteronomy 18:21-22 (Scripture)

  The law offers a test for a prophet: if the thing he foretells does not come to pass, the Lord has not spoken it. Students may compare this ancient test of a prediction with Popper's emphasis on predictions that could fail.

• Matthew 7:24-27 (Scripture)

  Jesus contrasts the house built on rock with the house built on sand. Popper's image of science built over a swamp makes a different claim about scientific knowledge alone, and the two images can be discussed together to distinguish provisional knowledge from ultimate foundations.

• Augustine, Against the Academics (Church Fathers)

  Augustine answered the ancient sceptics by arguing that some truths can be known with certainty; his dialogue is a classic counterpoint to Popper's view that empirical knowledge remains conjectural.

Logic and Rhetoric

Central argument: No finite number of observations can verify a universal law, but a single accepted counter-instance can refute it. Therefore science cannot be justified by induction, and its method must be the deductive testing of conjectures. A theory belongs to empirical science only if it is falsifiable, that is, if it forbids some possible observations. Since even observation statements are accepted provisionally, scientific knowledge has no final foundation; theories are corroborated by surviving severe tests, and bold, highly testable theories are to be preferred.

Rhetorical device: extended metaphor. At the end of the chapter on the empirical basis Popper compares science to a building on piles driven into a swamp: not resting on any natural base, but driven deep enough to carry the structure for the time being (see quote 2).

Debate: Resolved: That a claim which cannot be shown false cannot be counted as scientific.

Writing Prompts

Essay: Explain Popper's criterion of falsifiability and test it against two examples, one you think is scientific and one you think is not. Then consider one objection from Kuhn or Lakatos.

Comparison with A System of Logic, Book III: Of Induction: Mill thinks induction can be made rigorous; Popper thinks it cannot be justified at all. Compare their treatments of the white swans and argue which better describes how scientists actually reason.

Medical Lens

The book contains no medical subject matter, but its account of testing has been invoked in discussions of method in medicine and psychology. The immunologist Peter Medawar was a prominent admirer of Popper's view of scientific method, and the falsifiability criterion was cited by the US Supreme Court in Daubert v. Merrell Dow Pharmaceuticals (1993), a case about expert testimony on a drug's alleged effects. Popper's later criticisms of psychoanalysis as untestable appear in Conjectures and Refutations rather than in this book. These connections are matters of later influence, not content of the text.

Creative Prompts

• Classroom

  Give each group a sealed box with a hidden rule (for example, which shapes may be placed in it). Groups propose conjectures and test them only by trying to find counter-instances. Afterwards, discuss whether success came from confirming or from refuting.

• Dialogue

  Write a conversation in a Vienna coffee house in 1934 between Popper and Rudolf Carnap on whether metaphysical statements are meaningless or merely unscientific, giving each his strongest case.

• Essay

  Popper says science stands on piles driven into a swamp. Explain the image and argue whether it strengthens or weakens our reasons for trusting science.

Further Reading

• Conjectures and Refutations, Karl Popper

  Essays that restate the book's ideas more accessibly, including Popper's account of how he arrived at the demarcation criterion.

• Unended Quest: An Intellectual Autobiography, Karl Popper

• Criticism and the Growth of Knowledge, Imre Lakatos and Alan Musgrave (eds.)

  Papers from the 1965 London colloquium at which Kuhn, Popper, Lakatos and others debated.

• Popper, Bryan Magee

  A short, sympathetic introduction.

• What Is This Thing Called Science?, A. F. Chalmers

  A standard introduction that sets falsificationism beside inductivism, Kuhn and Lakatos.