Karl Popper's demarcation line failed; his open society did not

Karl Popper died at his home in Kenley, on the southern edge of London, on 17 September 1994. He was ninety-two, a knight, a Companion of Honour, and the only philosopher of science whose name most working scientists could produce on demand. Thirty-two years on, he is in an odd position. His most famous idea, that a theory is scientific only if something could show it to be false, is taught in secondary schools as settled, and has not been the working view of the philosophy of science for half a century. His second most famous idea, the open society, has lent its name to a foundation and to a footnote that is quoted more often than any other passage in twentieth-century political philosophy, almost always in truncated form.

The two halves of that reputation have moved in opposite directions. Falsifiability survived as a slogan and died as a criterion: Kuhn, Lakatos and the Duhem–Quine problem took it apart in the 1960s, and Popper himself gave ground on the one theory, Darwin's, that a demarcation criterion most needed to accommodate. The political Popper has aged the other way. Piecemeal reform against utopian blueprints, democracy as a procedure for removing rulers rather than a doctrine about who should rule, the paradox of tolerance: these have become the ordinary furniture of liberal argument, so ordinary that hardly anyone checks what he wrote. The case for reading him again is that the part everyone thinks secure is not, and the part everyone thinks a cliché is more exact than the cliché.

The year 1919 gave him both problems at once#

Popper was born in Vienna on 28 July 1902, into a family of assimilated Jews who had converted to Lutheranism before his birth. His father, Simon Siegmund Carl Popper, was a lawyer with a doctorate from the university and a private library of between twelve and fourteen thousand volumes. The empire collapsed when the boy was sixteen. In 1919, in the chaos that followed, he became for a few months a Marxist; the break came on 15 June that year, when police in the Hörlgasse fired on unarmed young socialists during a disturbance provoked by the communists, killing several, some of them Popper's friends. What repelled him, on his own later account, was less the shooting than the readiness of the theory to absorb it. A doctrine that could explain the deaths of its own supporters as a necessary stage in history could explain anything, and he came to think that was exactly its defect.

That same year gave him the contrast that organised the rest of his thinking. In a lecture at Peterhouse, Cambridge, in the summer of 1953, later printed as "Science: Conjectures and Refutations", he recalled that his friends who admired Marx, Freud and Adler "were impressed by a number of points common to these theories, and especially by their apparent explanatory power". He told the story of bringing Adler a case that did not seem to fit individual psychology, and of Adler explaining it without difficulty. Asked how he could be so sure, Adler answered, "Because of my thousandfold experience"; to which Popper, by his own report, could not help replying, "And with this new case, I suppose, your experience has become thousand-and-one-fold." Einstein, by contrast, had said in advance what observation would sink general relativity, and in 1919 the eclipse expeditions went to look. Popper's conclusion was the sentence generations of students have copied out: "Irrefutability is not a virtue of a theory (as people often think) but a vice."

Its systematic version appeared in Logik der Forschung, printed in Vienna by Julius Springer at the end of 1934 with 1935 on the title page, a wrinkle that has produced two dates for the book ever since. Popper rewrote rather than translated it as The Logic of Scientific Discovery in 1959. Against the Vienna Circle, of which he was never a member, he denied that science proceeds by inductive confirmation at all. No number of white swans verifies "all swans are white"; a single black one refutes it. Science advances by conjecture and attempted refutation, and the theories we keep are not the ones we have proved but the ones we have so far failed to break. He was candid about what this leaves standing. "Science does not rest upon solid bedrock," he wrote. "The bold structure of its theories rises, as it were, above a swamp. It is like a building erected on piles. The piles are driven down from above into the swamp, but not down to any natural or 'given' base; and if we stop driving the piles deeper, it is not because we have reached firm ground."

Why the criterion did not survive contact with science#

The trouble with falsifiability as a line between science and non-science was pointed out before Popper drew it. Pierre Duhem had argued in 1906 that a physicist never tests a single hypothesis in isolation; a prediction follows from the hypothesis together with auxiliary assumptions about instruments, initial conditions and background theory, and when the prediction fails, logic alone does not say which member of the bundle is to blame. Quine generalised the point in "Two Dogmas of Empiricism" in 1951. If any theory can be saved from any observation by adjusting something else, then no theory is falsifiable in the crisp logical sense, and the criterion either rules out everything or needs supplementary rules about which adjustments are permissible. Popper supplied such rules, against "ad hoc" rescues, and with each edition the rules grew longer.

Then came the historians. Thomas Kuhn's The Structure of Scientific Revolutions (1962) argued that this is not what scientists do, and, more damagingly, that it is not what they should do. Normal science treats anomalies as puzzles for the practitioner, not refutations of the paradigm; a field that abandoned its framework at the first recalcitrant result would never get anything done. In 1965 Imre Lakatos, Popper's former student, organised a colloquium at Bedford College in London on Kuhn's book; Popper, newly knighted, attended. The proceedings appeared five years later as Criticism and the Growth of Knowledge, and Lakatos's own contribution, "Falsification and the Methodology of Scientific Research Programmes", was the moment the Popperian school conceded the main point while trying to keep the flag. Lakatos accused Kuhn of making theory-change a matter of "mob psychology", but his remedy was to give up the assessment of single theories altogether. What can be judged is a research programme, a hard core protected by a belt of auxiliary hypotheses, and the judgement is retrospective: a programme is degenerating when its adjustments stop predicting anything new. That is a useful idea. It is not falsifiability. Paul Feyerabend, another product of the LSE seminar, drew the further conclusion in Against Method (1975) that no methodological rule survives the history of science intact.

Popper's own retreat was on Darwin. In his intellectual autobiography, published in the Schilpp volume in 1974 and separately as Unended Quest in 1976, he wrote that "Darwinism is not a testable scientific theory but a metaphysical research programme", while insisting the programme was invaluable. The sentence was seized on by creationists, and in a 1978 lecture printed in Dialectica he made what he called his recantation: "I have changed my mind about the testability and the logical status of the theory of natural selection." He now thought natural selection could be tested in particular cases, industrial melanism among them. The one theory every modern person takes to be paradigmatically scientific had to be let in by the back door.

None of this dislodged the slogan, and there is a reason it should not. Falsifiability failed as a definition of science and survived as a norm for scientists: state what would show you wrong, and go and look. Every pre-registered trial is Popper as a professional ethic rather than as logic. That is a smaller thing than he claimed, and the part that turned out to be true.

A war effort written at the far end of the world#

Popper left Austria in 1937 for a lectureship at Canterbury University College in Christchurch, New Zealand, about as far from Europe as it was possible to be. "The final decision to write it," he said in the preface to The Open Society and Its Enemies, "was made in March 1938, on the day I received the news of the invasion of Austria." He later described that book and its companion, The Poverty of Historicism, as his "war effort". Friedrich Hayek, then at the LSE, helped get Routledge to publish it, which they did in two volumes in 1945; it was also Hayek who showed the manuscript to the School's decision-makers and suggested converting a part-time post into a full readership. Popper arrived in London on 5 January 1946 and began lecturing the following week. He was made professor of logic and scientific method in 1949 and retired in 1969.

The argument of The Open Society is the argument of Logik der Forschung moved into politics. If no theory is ever more than provisionally corroborated, then no political doctrine can claim to know where history is going, and any doctrine that does claim it, which Popper called historicism, licenses the sacrifice of present people to a future it cannot foresee. The book traces that licence from Plato's Republic, through Hegel, to Marx. Volume one, on Plato, is the volume that made the book notorious. Classicists replied that Popper had read the Republic ahistorically, on tendentious translations, and without an ear for its irony; Ronald Levinson's In Defense of Plato (1953) is a book-length reply. Walter Kaufmann's "The Hegel Myth and Its Method" (1951) argued that the Hegel chapter attacked a figure who had never existed. Popper conceded very little on either front.

What lasts is what needs no Greek. The first is the distinction between utopian and piecemeal social engineering: the reformer who works from a blueprint of the good society must suppress the dissent his errors generate, because the blueprint cannot be wrong; the reformer who attacks "the greatest and most urgent evils of society" one at a time can see when a measure has failed and reverse it. That is falsifiability again, applied to legislation, and it is the ancestor of every argument for pilot schemes, sunset clauses and evaluation before scaling. The second is his reframing of the question of politics. Since Plato, he argued, political philosophy had asked who should rule, and every answer invites tyranny the moment the wrong person is installed. The better question is how institutions can be arranged so that bad or incompetent rulers do the least damage and can be removed. Democracy on this account is not the rule of the people, a phrase Popper thought nearly meaningless, but the type of government of which the governed can rid themselves without bloodshed. The definition is modest, and its modesty is the point.

The most misquoted footnote in political philosophy#

The paradox of tolerance is not in the text of The Open Society. It is note 4 to chapter 7 of volume one, one of several paradoxes Popper lists there to explain why unlimited freedom defeats itself. Since it is quoted everywhere and read almost nowhere, here it is in full:

Less well known is the paradox of tolerance: unlimited tolerance must lead to the disappearance of tolerance. If we extend unlimited tolerance even to those who are intolerant, if we are not prepared to defend a tolerant society against the onslaught of the intolerant, then the tolerant will be destroyed, and tolerance with them. In this formulation, I do not imply, for instance, that we should always suppress the utterance of intolerant philosophies; as long as we can counter them by rational argument and keep them in check by public opinion, suppression would certainly be most unwise. But we should claim the right to suppress them if necessary even by force; for it may easily turn out that they are not prepared to meet us on the level of rational argument, but begin by denouncing all argument; they may forbid their followers to listen to rational argument, because it is deceptive, and teach them to answer arguments by the use of their fists or pistols. We should therefore claim, in the name of tolerance, the right not to tolerate the intolerant.

He adds that any movement preaching intolerance places itself outside the law, and that incitement to intolerance and persecution should be treated as criminal, like incitement to murder or kidnapping.

The passage is now deployed as a two-line warrant for silencing whoever the speaker regards as intolerant, and the two lines chosen are always the first and the last. Read whole, the footnote is more careful. The default is not suppression but argument and public opinion, and suppression while those are working is "most unwise". The trigger is not offensive doctrine but the refusal of argument as such, the turn to "fists or pistols". And the instrument, in the closing sentence, is the criminal law applied to incitement, a standard liberal jurisdictions already recognised in 1945. Yascha Mounk, who has made this point at length, argues that the modern use of the footnote reverses Popper's purpose: a man whose whole project was to limit state power over opinion is invoked to hand the state, or the platform, the authority to decide which opinions count as intolerant. That is a fair reading. The other side has something to say too. Popper wrote the note with Vienna in 1934 and 1938 in mind, and his own biography was evidence that movements which begin by denouncing argument do not wait for the courts. The footnote is a rule of thumb from a man who distrusted rules of thumb. What it is not is a licence, and the people who quote it as one have generally not read the sentence in the middle.

The poker, and the temperament that produced it#

On 25 October 1946, nine months into his LSE post, Popper travelled to Cambridge to read a paper to the Moral Sciences Club, of which Wittgenstein was chairman. The meeting was in Richard Braithwaite's rooms, H3 in the Gibbs Building at King's College. Popper's title, "Are There Philosophical Problems?", took direct issue with a host who held that there were only puzzles of language. Bertrand Russell was present; it was the only occasion on which the three men were in a room together. At some point Wittgenstein picked up the poker from the fire. On Popper's account in Unended Quest, Wittgenstein demanded an example of a moral rule and Popper replied, "Not to threaten visiting lecturers with pokers", at which Wittgenstein threw the poker down and left. Other witnesses remembered the sequence differently. David Edmonds and John Eidinow interviewed the survivors for Wittgenstein's Poker (2001) and concluded that the accounts could not be reconciled. The episode has become a parable about two temperaments, one turned inward towards the limits of language, the other outward towards the problems of the world, which is roughly the caricature Wittgenstein's readers keep having to correct. What it shows more certainly is that Popper was a combative man, and that he knew it.

His late work moved into metaphysics. Objective Knowledge (1972) set out the theory of three worlds: the physical, the mental, and a third world of theories, arguments and problems, made by human beings but not reducible to anyone's mental state, since a theory can have consequences no one has yet noticed. The Self and Its Brain (1977), written with the neurophysiologist John Eccles, defended a mind–body dualism that placed Popper well outside the philosophy of mind of his era. He was knighted in 1965 and made a Companion of Honour in 1982. Among his LSE students had been a Hungarian refugee named George Soros, who later named a foundation after the book.

What is left#

The honest accounting is uneven. As a logic of science, Popper's system has the status of Newtonian mechanics in a relativistic age: worth teaching, wrong in ways that matter, superseded by accounts that keep what was right in it. As a political thinker he is in better shape than most of the people who cite him. The reason is the same in both cases. Popper's real subject was never the structure of theories but the conduct of the people who hold them: whether they say in advance what would change their mind, whether they treat error as information or as betrayal, whether they can be removed. On the first question the philosophy of science found his answer too simple. On the second, the century since the Hörlgasse has not produced a better one.