Imre Lakatos
Imre Lakatos developed the methodology of scientific research programmes as an attempt to mediate between Popper's falsificationism and Kuhn's paradigms. His framework proposed that science consists of competing research programmes, each organized around a hard core of theoretical commitments protected by a belt of auxiliary hypotheses. Programmes are progressive when they make novel predictions that turn out to be confirmed; degenerating when they modify their auxiliary hypotheses only to absorb disconfirming evidence without generating new predictions. The framework influenced Laudan heavily but was ultimately superseded by the problem-solving model, which rejected Lakatos's emphasis on predictive novelty in favor of a broader account of problem-solving capacity.
In the AI Story
Lakatos's work was itself a response to Kuhn, attempting to preserve the rationality of theory-choice that Kuhn's framework seemed to threaten. Where Kuhn spoke of paradigms as total frameworks, Lakatos spoke of research programmes that could be modified internally. Where Kuhn made paradigm shifts quasi-sociological events, Lakatos proposed criteria by which the progress of programmes could be rationally evaluated.
The criterion of progress Lakatos proposed — predictive novelty — was narrower than Laudan found defensible. Many scientifically important developments do not produce dramatic novel predictions; they resolve conceptual tensions, integrate previously separate domains, or provide better explanations of already-known phenomena. Laudan's problem-solving model captured these forms of progress in a way Lakatos's novelty criterion could not.
Lakatos's distinction between progressive and degenerating research programmes was nevertheless adopted and refined by Laudan. The pattern of a tradition that protects its core by generating ad hoc modifications — Ptolemy's epicycles, psychoanalysis absorbing any clinical observation — was what both thinkers identified as the signature of degeneration. Laudan's progressive versus degenerative traditions distinction is a direct descendant of Lakatos's analysis, refined by the broader problem-solving framework.
Lakatos died young, at 51, before he could fully develop his framework in response to the critiques Laudan and others eventually brought. The trajectory of post-Kuhnian philosophy of science in the 1970s and 1980s can be read as an extended conversation between the Lakatosian and Laudanian frameworks, with the broader problem-solving model eventually gaining wider acceptance.
Key Ideas
Research programmes. Scientific traditions organized around a hard core of commitments and a flexible belt of auxiliary hypotheses.
Progressive vs degenerating. Programmes progress when they generate novel predictions; degenerate when they absorb disconfirming evidence without new predictions.
Novelty as criterion. The specific criterion Laudan found too narrow.
Ad hoc modification as degeneration. The pattern of protecting the core by modifying the belt without generating new content.
Appears in the Orange Pill Cycle
Further reading
- Imre Lakatos, The Methodology of Scientific Research Programmes (Cambridge University Press, 1978).
- Imre Lakatos and Alan Musgrave, eds., Criticism and the Growth of Knowledge (1970).
- Brendan Larvor, Lakatos: An Introduction (1998).