Scientific research programme
과학적 연구 프로그램
The unit of scientific methodology put forward by the Hungarian-born philosopher Imre Lakatos to resolve the conflict between Popper's falsificationism and Kuhn's account of scientific revolutions. It holds that science advances not through single theories but through competition between research programmes, each made up of a "hard core" that cannot be given up and a "protective belt" of auxiliary hypotheses modified to defend it. A programme that predicts new facts which are then confirmed is called "progressive"; one that merely fends off refutations after the event is "degenerating". By this standard Lakatos named twentieth-century Soviet Marxism as an example of pseudoscience.
In depth
Must a refuted theory be dropped at once?
Suppose a theory's prediction clashes with observation. Popper's falsificationism was widely read to mean that the theory must then be abandoned. Yet in the history of science as Kuhn drew it, scientists achieved most when they held on to "normal" theories despite known anomalies. One side seemed to say what science ought to be like, the other what science had actually been like. Could both be right?
Imre Lakatos tried to answer this question. His model of the "scientific research programme" was an attempt to combine Popper's adherence to empirical validity with Kuhn's appreciation of conventional consistency. He saw himself as merely extending Popper's ideas. In his 1968 article "Criticism and the Methodology of Scientific Research Programmes" he distinguished the "naive falsificationist" Popper, who demanded unconditional rejection of a theory in the face of any anomaly, from the "sophisticated methodological falsificationist" Popper whom he took to be the correct extension of Popper's ideas. It is therefore very hard to say whose ideas are whose.
What path brought Lakatos to this problem?
Lakatos was born Imre Lipsitz in 1922 to a Jewish family in Debrecen, Hungary. When Germany occupied Hungary in 1944 he formed a Marxist resistance group; his mother and grandmother were murdered in Auschwitz. After the war he worked from 1947 as an official of the Ministry of Education, attended György Lukács's private Wednesday seminars, and in 1949 studied at Moscow University under Sofya Yanovskaya. English Wikipedia writes that he was a hardline Stalinist who between 1945 and 1950 played an important role in building up Communist rule in Hungary, especially in cultural life and academia.
But he ended up on the losing side of arguments within the party, was arrested in 1950 on charges of revisionism and held until 1953 in the Recsk labour camp. After his release he worked at the mathematics institute of the Hungarian Academy of Sciences and translated the mathematician George Pólya's How to Solve It into Hungarian. Hungarian Wikipedia writes that in these years the thought of Lukács, Pólya and others had a great influence on him. When the Hungarian Revolution of 1956 was crushed by Soviet troops he emigrated to England by way of Austria.
In England he took a PhD at the University of Cambridge in 1961 with a thesis on the logic of mathematical discovery, which appeared after his death as Proofs and Refutations. German Wikipedia writes that his work on mathematical proof was influenced by Hegelian and Marxian dialectics. From 1960 he worked with Popper at the London School of Economics (LSE), and he edited the proceedings of the 1965 London colloquium on Kuhn's book, published in 1970 as Criticism and the Growth of Knowledge. The theory of research programmes was worked out in the middle of this debate.
How do the hard core and the auxiliary hypotheses work?
A Lakatosian research programme rests on a "hard core": theoretical assumptions that cannot be abandoned or altered without abandoning the whole programme. Around it are more modest and specific theories formulated to explain evidence that threatens the core: the auxiliary hypotheses. One English Wikipedia article sums them up as a "protective belt" of malleable theories surrounding the core. The auxiliary hypotheses are expendable; they are altered or abandoned as empirical discoveries require, in order to "protect" the core.
Lakatos's main example is the programme founded by Newton, whose three laws of motion formed the hard core. When a planet moved otherwise than predicted, the Newtonians did not drop the laws of motion but amended the auxiliary hypotheses, for example by supposing an unknown planet. Here Lakatos followed the French physicist Pierre Duhem's idea that a cherished theory can always be protected by redirecting criticism to other theories or other parts. So Lakatos changed Popper's picture: it is not that we propose a theory and nature may shout "no"; rather we propose a maze of theories, and nature may shout "inconsistent".
A programme also contains rules of method. Lakatos called the research methods to avoid its "negative heuristic" and the methods to prefer its "positive heuristic". The negative heuristic protects the hard core; the positive heuristic directs in which direction the core and the auxiliary hypotheses are to be modified. Several research programmes thus coexist and compete to be the most progressive. In this respect a research programme resembles Kuhn's paradigm, but Lakatos held that it is guided by Popper's objective logic of discovery rather than Kuhn's irrational "psychology of discovery".
How are progressive and degenerating programmes told apart?
For Lakatos not every change to the auxiliary hypotheses (a "problem shift") is equally legitimate. If the change raises explanatory or predictive power, above all if it predicts new facts that are then confirmed, the programme is "progressive". If the change is made simply because some answer must be offered in the face of new and troublesome evidence, so that it achieves nothing beyond maintaining the core, the programme is "degenerating".
What matters is that even a degenerating programme is not dropped at once. Abandoning the present programme before a better one has been devised and agreed upon would only weaken our explanatory power further. According to Lakatos a theory has not been rightfully "falsified" until it is superseded by a more progressive research programme. He held that this is what happens in the periods Kuhn calls revolutions, and that it makes them rational rather than leaps of faith or bouts of deranged social psychology.
The standard was also a criterion for separating science from pseudoscience. A theory that fails to predict any novel phenomena, or whose predictions are mostly refuted, becomes pseudoscience. Lakatos's own examples of pseudoscience were Ptolemaic astronomy, Immanuel Velikovsky's planetary cosmogony, Freudian psychoanalysis, twentieth-century Soviet Marxism, Lysenko's biology, Niels Bohr's quantum mechanics after 1924, astrology and psychiatry. In a 1973 lecture he said that nobody had yet found a demarcation criterion by which Darwin could be called scientific.
How did Lakatos judge Marxism?
Popper had charged that Marx's followers, instead of accepting falsified predictions, reinterpreted both theory and evidence. Lakatos restated the charge in his own terms. In The Methodology of Scientific Research Programmes, published posthumously in 1978, he declared that Marxism had never once successfully predicted a stunning novel fact. The failed predictions he named were the absolute impoverishment of the working class and the prediction that the first socialist revolution would take place in the industrially most developed society. Marxists explained all their failures, but in Lakatos's view their auxiliary hypotheses were all cooked up after the event to protect the theory from the facts. He concluded: "The Newtonian programme led to novel facts; the Marxian lagged behind the facts and has been running fast to catch up with them."
The Newtonian programme led to novel facts; the Marxian lagged behind the facts and has been running fast to catch up with them.
This verdict does not cut Marxism off from science from the start; it treats it as a programme that began as a science and degenerated. One essay on the debate sums up that Popper and Lakatos both claimed Marxism made specific predictions (as a science should) but handled falsifications unscientifically, and so degenerated into a pseudoscience. The same essay notes that both men had a personal history with Marxism, strictly with Marxism-Leninism and Stalinism. Lakatos himself had once been in the front line of building Hungary's communist regime, and had known its prisons.
How do history and philosophy test each other?
Lakatos opened his 1970 article "History of Science and Its Rational Reconstructions" with the dictum: "Philosophy of science without history of science is empty; history of science without philosophy of science is blind."
Philosophy of science without history of science is empty; history of science without philosophy of science is blind
He proposed that rival methodologies of science be compared by how well they explain the actual history of science. But neither he nor his collaborators ever showed that in any scientific revolution the great majority of the scientific community converted just when his criterion was satisfied. In his 1968 article he had himself admitted that he would not go on seriously to the stage of comparing rational reconstructions with actual history.
The sharpest criticism came from his friend Paul Feyerabend. Feyerabend pointed out that if one consistently follows Lakatos's demand that research programmes be given "breathing space", one can never decide when to stop pursuing a degenerating programme, however long it has been degenerating. His example was Boltzmann's atomism, which seemed to degenerate in the nineteenth century but was vindicated in the early twentieth with Einstein's work in statistical mechanics. Lakatos's rules, then, are empty because they forbid no behaviour, and Lakatos is an "anarchist in disguise". Lakatos planned to write For Method in reply to Feyerabend's Against Method, but died in 1974 before he could. From 1971 until his death he edited the British Journal for the Philosophy of Science.