Imre Lakatos (9 November 1922 – 2 February 1974) was a Hungarian-born philosopher of science who greatly contributed to the problem of demarcation and theory choice in science. A protege of Karl Popper, Lakatos attempted to respond to problems posed by the work of Popper and Kuhn.12 His Methodology of Scientific Research Programmes (MSRP) offers a holistic approach to theory choice which extends beyond Popper's falsificationism. It assesses a particular research program as progressive or degenerative, depending on its overall record of predictive and explanatory successes and failures. Lakatos later entered into a correspondence with Paul Feyerabend, with the goal of addressing Feyerabend’s objections to the MSRP. He met an untimely death due to a heart attack at the age of 51. Some of Feyerabend’s objections remain challenging to this day.
Much of Lakatos's work was a response to the problems of Karl Popper’s falsificationism, and was expressed in a series of publications between 1935 and the early 1970's. Lakatos rejected the idea that a false prediction was alone grounds for rejecting a theory. Most theories, he pointed out, are born in an “ocean of anomalies” and are therefore falsified from the moment of their inception. For example, Copernican heliocentric astronomy predicts that the stars should change in apparent position as the Earth revolves around the sun, but for three centuries after Copernicus proposed his theory, all attempts to detect this stellar parallax failed. Astronomers nevertheless accepted the theory on other grounds. The failure of Newtonian mechanics to account for the motions of the planet Mercury was known for many decades, during which the theory also wasn't rejected.1 A well known criticism of falsificationism, the Duhem-Quine thesis,345 which Lakatos championed, was that the failure of a prediction could be due to a problem anywhere in the network of theories and auxiliary assumptions responsible for that prediction. Lakatos thus argued that Popper's theory was overly restrictive and inconsistent with much of scientific practice. In scientific practice, Lakatos observed that if a theory is the best available of its kind, it is typically allowed to undergo modifications to account for all data and not rejected.
Lakatos also responded to Thomas Kuhn’s Structure of Scientific Revolutions, published in 1962.6 He was troubled by Kuhn's incommensurability thesis, which asserts that theories with different taxonomies cannot be rationally compared. Lakatos accused Kuhn of depicting the process of scientific change as completely irrational. If there truly existed a problem of incommensurability in science, then there would be no method to demarcate between science and pseudoscience, and no way of measuring scientific progress.7 At the same time, Lakatos and Kuhn's views of science have important points in common. Both rejected the positivist, inductivist accounts of science popular in the early twentieth century, and both emphasized the importance of theory over observation. Both agreed that any theory of how science works must make sense of the actual history of science.2
Lakatos on Theory Choice
Following the Duhem-Quine thesis, Lakatos recognized that scientific theories could not be appraised individually. Rather, all of the theoretical assumptions bearing on an experimental finding had to be assessed holistically, as parts of what he called a research program.8 While Kuhn supposed that, in a mature scientific discipline, only one paradigm generally existed at a time, Lakatos argued that it was generally the case that more than one research program existed in a field at any given time, and that large-scale processes of scientific change should be understood as competition between research programs.9 Within a research program, not all theoretical assumptions are treated equally. The indispensable central theoretical assumptions of a research program are its hard core. Any modification of the hard core constitutes the abandonment of the research program and the creation of a new one. 28
Auxiliary propositions that are relevant to the hard core, but are not part of it form a protective belt. Adherents of a research program attempt to explain an increasingly wide range of relevant natural phenomena in terms of the core. In so doing, they add to the protective belt of auxiliary propositions. This expansion of the range of applicability of the program constitutes its positive heuristic. Scientists committed to a research program defend the hard core against change by using their ingenuity as needed to make alterations to the protective belt of auxiliary propositions to explain phenomena and avoid falsification of the core. This protection of the hard core is a research program's negative heuristic.8
For example, the hard core of the Newtonian physics research program would consist of Newton's three laws of motion and Law of Universal Gravitation. The protective belt would include propositions such as "the Earth is an oblate spheroid" or "Neptune is 17 times more massive than Earth". In the nineteenth century, astronomers could not explain the movements of the planet Uranus using Newton's theory and known gravitational influences. Rather than modifying the theory itself, which would have obviated the Newtonian research program, they modified the protective belt by positing the existence of a new planet, whose Newtonian gravitational influence was affecting Uranus. The prediction was a stunning success, as the new planet, to be named Neptune, was discovered in 1846.28
If any evidence is found against a theory, and if the theory otherwise possesses both greater heuristic and explanatory powers than known alternatives, Lakatos supposed that falsification should be averted by modifying the research program's protective belt. There thus can be no 'crucial experiments'; a research program cannot be instantly overthrown by a single experimental finding taken in isolation.
Lakatos held that a research program should be evaluated in terms of both its explanatory power; its ability to explain known phenomena, and its heuristic power; its ability to successfully explain newly discovered phenomena or to predict their existence.
Lakatos stipulated that a modification is progressive if all of the following conditions are met:8
- the modification has some excess empirical content, i.e. it increases the overall empirical content of a research programme (by making novel predictions or increasing their precision and accuracy);
- some of this excess empirical content has been corroborated in experiments and observations;
- the modification is in organic unity with the rest of the programme.
Thus, according to Lakatos, there are three types of regressive (ad hoc) modifications. A modification is regressive if at least one of the following obtains:10
- it does not increase the empirical content of the programme, i.e. it doesn't make novel predictions or increase their precision/accuracy (ad hoc1);
- it introduces excess empirical content, but fails to corroborate any of this excess content empirically (ad hoc2);
- it is not in organic unity with the rest of the programme (ad hoc3).
The term organic unity is intended to mean that modifications should be contiguous with the rest of the program. For instance, if the research program is Darwin's theory of natural selection, a modification which adds the proposition "extra-terrestrial beings intervened with human evolution" would not be contiguous – not in organic unity – with the rest of the research program, and therefore regressive.
Given any modification to a research programme’s protective belt, any research programme P1 would subsequently become P2. In this way we can track changes to research programme P from P1 to Pn and retrospectively ascertain if the modifications made have been progressive or degenerative. However, the degeneration of a research program doesn't necessitate its dismissal. Rather, given research programmes A and B, where programme A has been degenerating and programme B has been progressing, Lakatos suggests that the scientific community should invest most of its resources into A. The community should not invest all resources into A because there have been instances where a degenerative programme has become progressive, such as heliocentrism and atomism. According to Lakatos, working on a degenerative programme is not prohibited, but it is irrational given that it has ceased to bear fruit.
Lakatos on Demarcation Criteria
The demarcation between "progressive" and "degenerative" research programmes also serves to demarcate between science and pseudoscience. A scientific theory should not only explain past and present phenomena; it should also have the ability to be applied to and posit the existence of future phenomena. For example, proving that an object falls in an experiment does not make the Newtonian research programme scientific. However, the Newtonian research programme predicted that comets move in either hyperbolas, parabolas or ellipses (contrary to the contemporary theory that they move in straights lines). Using this hypothesis, Edmond Halley successfully predicted the return of Halley’s comet to the minute. Such predictions affirm that the Newtonian research programme was progressive and, therefore, scientific. On the other hand, a research programme such as astrology, which merely provides post hoc explanations and is subject to ad hoc modifications, is considered pseudoscientific.
Paul Feyerabend pointed out that there exists a serious problem in how one can justify working on a degenerating programme. Lakatos’ response to Feyerabend's criticism was ambiguous. While it is irrational to work on a degenerating programme, Lakatos held that it was not prohibited. Just because a research programme is degenerating does not mean that it should be rejected. Feyerabend's objection, however, remains open because Lakatos failed to provide anything more than stipulation.
Feyerabend also argued that a problem exists with the notion of a time limit. If a research programme has been degenerating for some time, how do we know when to abandon it? Heliocentrism and atomism had degenerated for well over a millennium before being reinvigorated and subsequently accepted. On the other hand, fields such as homeopathy or psychoanalysis, which are nascent in comparison to the aforementioned theories, are easily ascribed the title of pseudoscience. Lakatos’ response was that there is no discernable time limit; but once again, he failed to address the actual problem.
Here are the works of Lakatos included in the bibliographic records of this encyclopedia:
- Goodstein and Lakatos (1962): Goodstein, Reuben Lewis and Lakatos, Imre. (1962) Symposium: The Foundations of Mathematics. Proceedings of the Aristotelian Society 36, 145-184.
- Lakatos (1961): Lakatos, Imre. (1961) Essays in the Logic of Mathematical Discovery. University of Cambridge.
- Lakatos (1962): Lakatos, Imre. (1962) Infinite Regress and Foundations of Mathematics. In Lakatos (1978b), 3-23.
- Lakatos (1963a): Lakatos, Imre. (1963) Proofs and Refutations (I). The British Journal of Philosophy of Science 14 (53), 1-25.
- Lakatos (1963b): Lakatos, Imre. (1963) Newton's Effects on Scientific Standards. In Lakatos (1978a), 193-222.
- Lakatos (1967): Lakatos, Imre. (1967) A Renaissance of Empiricism in the Recent Philosophy of Mathematics? In Lakatos (1978b), 24-42.
- Lakatos (1968a): Lakatos, Imre. (1968) Changes in the Problem of Inductive Logic. In Lakatos (1978b), 128-200.
- Lakatos (1968b): Lakatos, Imre. (1968) Criticism and the Methodology of Scientific Research Programmes. Proceedings of the Aristotelian Society 69, 149-186.
- Lakatos (1970): Lakatos, Imre. (1970) Falsification and the Methodology of Scientific Research Programmes. In Lakatos (1978a), 8-93.
- Lakatos (1971a): Lakatos, Imre. (1971) History of Science and Its Rational Reconstructions. In Lakatos (1978a), 102-138.
- Lakatos (1971b): Lakatos, Imre. (1971) Replies to Critics. Boston Studies in the Philosophy of Science 8, 174-182.
- Lakatos (1973): Lakatos, Imre. (1973) The Problem of Appraising Scientific Theories: Three Approaches. In Lakatos (1978b), 107-120.
- Lakatos (1974a): Lakatos, Imre. (1974) Popper on Demarcation Criteria and Induction. In Lakatos (1978a), 139-167.
- Lakatos (1974b): Lakatos, Imre. (1974) The Role of Crucial Experiments in Science. Studies in History and Philosophy of Science Part A 4 (4), 309-325.
- Lakatos (1974c): Lakatos, Imre. (1974) Anomalies Versus "Crucial Experiments". In Lakatos (1978b), 211-223.
- Lakatos (1976a): Lakatos, Imre. (1976) Proofs and Refutations: The Logic of Mathematical Discovery. Cambridge University Press.
- Lakatos (1976b): Lakatos, Imre. (1976) Understanding Toulmin. In Lakatos (1978b), 224-245.
- Lakatos (1978a): Lakatos, Imre. (1978) Philosophical Papers: Volume 1. The Methodology of Scientific Research Programmes. Cambridge University Press.
- Lakatos (1978b): Lakatos, Imre. (1978) Philosophical Papers: Volume 2. Mathematics, Science and Epistemology. Cambridge University Press.
- Lakatos (Ed.) (1967): Lakatos, Imre. (Ed.). (1967) Problems in the Philosophy of Mathematics. Amsterdam: North Holland.
- Lakatos (Ed.) (1968): Lakatos, Imre. (Ed.). (1968) The Problem of Inductive Logic. North Holland Publishing Company.
- Lakatos and Musgrave (Eds.) (1968): Lakatos, Imre and Musgrave, Alan. (Eds.). (1968) Problems in the Philosophy of Science. North Holland Pub. Co..
- Lakatos and Musgrave (Eds.) (1970): Lakatos, Imre and Musgrave, Alan. (Eds.). (1970) Criticism and the Growth of Knowledge. Cambridge University Press.
- Lakatos and Zahar (1976): Lakatos, Imre and Zahar, Elie. (1976) Why Did Copernicus's Research Programme Supersede Ptolemy's? In Lakatos (1978a), 168-192.
To add a bibliographic record by this author, enter the citation key below:
Citation keys normally include author names followed by the publication year in brackets. E.g. Aristotle (1984), Einstein, Podolsky, Rosen (1935), Musgrave and Pigden (2016), Kuhn (1970a), Lakatos and Musgrave (Eds.) (1970). If a record with that citation key already exists, you will be sent to a form to edit that page.
- Musgrave, Alan and Pigden, Charles. (2016) Imre Lakatos. In Zalta (Ed.) (2016). Retrieved from http://plato.stanford.edu/archives/sum2016/entries/lakatos/.
- Chalmers, Alan. (2013) What is This Thing Called Science? University of Queensland Press.
- Stanford, Kyle. (2016) Underdetermination of Scientific Theory. In Zalta (Ed.) (2016). Retrieved from http://plato.stanford.edu/entries/scientific-underdetermination/.
- Duhem, Pierre. (1962) The Aim and Structure of Physical Theory. Atheneum.
- Quine, Willard van Orman. (1951) Two Dogmas of Empricism. In Quine (1953), 20-46.
- Kuhn, Thomas. (1962) The Structure of Scientific Revolutions. University of Chicago Press.
- Lakatos, Imre. (1978) Philosophical Papers: Volume 1. The Methodology of Scientific Research Programmes. Cambridge University Press.
- Lakatos, Imre. (1970) Falsification and the Methodology of Scientific Research Programmes. In Lakatos (1978a), 8-93.
- Godfrey-Smith, Peter. (2003) Theory and Reality. University of Chicago Press.
- Lakatos, Imre. (1971) History of Science and Its Rational Reconstructions. In Lakatos (1978a), 102-138.