Imre Lakatos’ Approach: Bridging Popper and Kuhn in Philosophy of Science

In the mid-twentieth century, philosophy of science found itself caught between two powerful but opposing views. Karl Popper argued that science advances through bold conjectures and ruthless refutations – a theory must be falsifiable, or it is not science at all. Thomas Kuhn countered that real science rarely works this way; instead, it clings to established paradigms until a dramatic, almost irrational revolution replaces one framework with another. Enter Imre Lakatos, a Hungarian-born philosopher who proposed something more subtle: the concept of Scientific Research Programmes (SRPs). Rather than picking a side, Lakatos absorbed the strengths of both Popper and Kuhn while discarding their weaknesses, offering a view of science that is both historically accurate and rationally grounded.

Who was Imre Lakatos? 🔗

Imre Lakatos (1922-1974) was born in Debrecen, Hungary. He studied mathematics, physics, and philosophy before fleeing to the West after the Soviet suppression of the 1956 Hungarian Uprising. Settling in England, he earned a PhD from the University of Cambridge and eventually rose to the position of Professor of Logic at the London School of Economics (LSE), where he worked in the same department as Karl Popper. Lakatos is best known for two major contributions: his dialogue Proofs and Refutations in the philosophy of mathematics, and his Methodology of Scientific Research Programmes (MSRP) in the philosophy of science. His career was cut short by his early death at the age of 51, yet his influence has been enormous – extending far beyond academic philosophy into fields like educational theory, political science, and economics.

The problem with Popper: naive falsificationism 🔗

To understand why Lakatos’s work matters, we need to grasp the limitations he identified in his predecessors. Popper’s central idea was straightforward: a scientific theory must make predictions that can, in principle, be proven wrong. If a prediction fails, the theory is falsified and should be discarded. This falsifiability criterion served as Popper’s demarcation line between science and non-science.

The trouble is, as Lakatos and others pointed out, scientists almost never abandon a theory the moment they encounter contradictory evidence. Consider Newton’s laws of motion and gravitation. When astronomers observed that Mercury’s orbit didn’t quite match Newtonian predictions, they didn’t throw out Newton’s entire framework. Instead, they proposed auxiliary explanations – perhaps an undiscovered planet was pulling on Mercury. This is perfectly normal scientific behaviour, yet Popper’s strict falsificationism would seem to call it irrational.

Lakatos distinguished between what he playfully called Popper₀ (the “naive falsificationist” who demands immediate rejection of any theory facing an anomaly), Popper₁ (the more nuanced, historically situated Popper), and Popper₂ (the “sophisticated methodological falsificationist” who, Lakatos claimed, was essentially Lakatos himself). The key critique was that naive falsificationism fails as a description of actual science. Scientists routinely hold on to productive theories despite anomalies, and they are often right to do so.

The problem with Kuhn: irrationality in scientific change 🔗

While Lakatos agreed with Kuhn that scientists do persist with theories in the face of anomalies, he was deeply uncomfortable with where Kuhn’s argument led. In Kuhn’s model, science operates within a paradigm – a set of shared assumptions, methods, and exemplary solutions that define “normal science.” Anomalies accumulate, a crisis sets in, and eventually a revolutionary paradigm shift occurs. But for Kuhn, this shift is not purely rational. Since the old and new paradigms may be “incommensurable” – that is, they use different concepts and standards of evaluation – the transition resembles something closer to a conversion experience than a logical argument.

Lakatos found this deeply troubling. He famously characterised Kuhn’s view of scientific revolution as irrational, driven by mob psychology rather than logic. If there is no rational basis for choosing between competing paradigms, then scientific progress becomes a matter of sociological fashion rather than the growth of knowledge. Lakatos wanted to preserve Kuhn’s insight that scientists tolerate anomalies without accepting the implication that science is ultimately irrational.

Scientific research programmes: the Lakatosian synthesis 🔗

Lakatos’s answer was the concept of the Scientific Research Programme (SRP). An SRP is not a single theory but a succession of related theories that share a common foundation. This was a crucial shift: instead of evaluating individual theories in isolation (as Popper suggested), Lakatos argued that the proper unit of scientific evaluation is the entire research programme over time.

Each SRP has a specific internal structure built around four key components.

The hard core 🔗

The hard core consists of the fundamental assumptions that define the programme. These are the central tenets that practitioners treat as non-negotiable – they will not be abandoned or modified without abandoning the programme altogether. For example, in Newtonian mechanics, the hard core comprises the three laws of motion and the law of universal gravitation. These are the foundational axioms around which all research within the programme revolves.

The protective belt 🔗

Surrounding the hard core is a protective belt of auxiliary hypotheses. These are the adjustable, expendable elements that can be modified or replaced when the programme encounters anomalies. When observations don’t match predictions, the scientist doesn’t abandon the hard core. Instead, they tinker with the protective belt – adjusting initial conditions, refining measurement techniques, or introducing new subsidiary hypotheses. In the Newtonian case, theories of atmospheric refraction or geometrical optics formed part of this belt and could be tweaked to accommodate new data without touching Newton’s fundamental laws.

The negative heuristic 🔗

The negative heuristic is the methodological rule that tells scientists what not to do: do not direct refutations at the hard core. No matter what anomalies arise, the hard core is off-limits. All the refutational pressure must be absorbed by the protective belt. This might sound dogmatic, but Lakatos argued it is a rational strategy – it gives the programme time to develop its full potential before being judged.

The positive heuristic 🔗

The positive heuristic is the constructive counterpart to the negative heuristic. It provides a set of suggestions and guidelines for how to develop and refine the protective belt over time. Rather than just reacting defensively to anomalies, scientists following the positive heuristic are proactively building out the programme, making it increasingly sophisticated. As scholars have noted, the positive heuristic guides researchers toward fruitful avenues of investigation and away from dead ends.

Progressive vs. degenerating research programmes 🔗

With this structure in place, Lakatos introduced the crucial evaluative distinction that sets his framework apart: the difference between progressive and degenerating research programmes.

A research programme is considered progressive if it meets two conditions. First, it must be theoretically progressive – each new version of the theory must predict novel facts that were not predicted by its predecessor. Second, it must be at least intermittently empirically progressive – some of those novel predictions must actually be confirmed by observation. In a progressive SRP, changes to the protective belt lead to genuine growth in predictive power.

A programme is degenerating when it fails to satisfy these conditions. If new modifications to the protective belt only explain away existing anomalies without predicting anything new – what Lakatos called ad hoc adjustments – the programme is losing its scientific vitality. It is no longer driving discovery; it is merely playing catch-up with the facts.

Take the historical example that Lakatos frequently used. The Newtonian research programme was spectacularly progressive for over two centuries. It made bold predictions – such as the existence of the planet Neptune, inferred from gravitational anomalies in the orbit of Uranus – and those predictions were confirmed. In contrast, Lakatos argued that twentieth-century Marxism had become a degenerating programme: its adherents constantly modified auxiliary hypotheses to accommodate events (Berlin 1953, Budapest 1956, Prague 1968) after the fact, without ever predicting novel phenomena in advance.

How Lakatos bridges Popper and Kuhn 🔗

Lakatos’s framework draws specific elements from each of his predecessors while correcting their flaws.

What he takes from Popper 🔗

From Popper, Lakatos retains the commitment to empirical testability as a criterion for good science. Falsifiability still plays a role – individual theories within a research programme must make testable predictions. The emphasis on prediction and potential refutation keeps science anchored to the empirical world. Science remains a rational enterprise governed by logic and evidence, not mere social convention.

What he takes from Kuhn 🔗

From Kuhn, Lakatos accepts the historical reality that scientists tolerate anomalies and do not discard productive theories at the first sign of trouble. He also borrows the structural insight that science operates within broad frameworks – Kuhn’s paradigms map roughly onto Lakatos’s research programmes. The hard core of an SRP plays a role similar to the central commitments of a Kuhnian paradigm.

What he rejects from both 🔗

Lakatos rejects Popper’s naive insistence that any falsification should lead to immediate theory rejection. He also rejects Kuhn’s suggestion that paradigm shifts are fundamentally irrational. In Lakatos’s model, the replacement of one research programme by another is a rational process: a degenerating programme is overtaken by a progressive one that predicts and confirms more novel facts. There is no need for “mob psychology” – the shift follows from the comparative track records of competing programmes.

Coexistence and competition of research programmes 🔗

One of the most important differences between Lakatos and Kuhn is that Lakatos explicitly allows for multiple research programmes to coexist and compete within the same scientific field at the same time. Kuhn’s model suggests that normal science operates under a single dominant paradigm, with competition only arising during rare revolutionary crises. Lakatos found this too monopolistic. In practice, scientific fields often harbour several rival programmes simultaneously, each with its own hard core and protective belt, each vying to demonstrate that it is the most progressive.

This pluralism has a significant implication: scientific progress is not about a single theory standing or falling. It is about the comparative evaluation of rival programmes over time. A degenerating programme should not be abandoned unless a more progressive alternative exists. Lakatos insisted that it is rational to continue working within a programme that is currently struggling, provided there is no better option available – and also because degenerating programmes can sometimes stage a comeback.

Criticisms and limitations of Lakatos’s approach 🔗

Lakatos’s framework, while widely admired, has not escaped criticism. Perhaps the most pointed objection came from his contemporary Paul Feyerabend, who argued that Lakatos’s methodology amounted to little more than rhetoric – it offered no concrete rules for when a scientist should actually abandon a degenerating programme. Since a degenerating programme might always recover, Feyerabend contended, the distinction between progressive and degenerating is never decisive in real time. It only becomes clear in hindsight.

This connects to a deeper problem. Lakatos’s framework evaluates research programmes retrospectively – it tells us which programme was progressive and which was degenerating after the fact. But it offers little guidance to the working scientist who must decide now whether to stick with a faltering programme or jump ship. The individual scientist still faces what is essentially a subjective judgement call, which was precisely the kind of irrationality that Lakatos accused Kuhn of endorsing.

Additionally, Lakatos himself acknowledged that his famous dictum – “philosophy of science without history of science is empty; history of science without philosophy of science is blind” – had never been fully demonstrated. He and his collaborators did not conclusively show that historical scientific revolutions consistently occurred when his criterion (one programme predicting novel facts while its rival degenerates) was satisfied.

Why Lakatos still matters 🔗

Despite these criticisms, Lakatos’s concept of the Scientific Research Programme remains one of the most influential frameworks in the philosophy of science. It provides a realistic and structured way to think about how science actually develops – not as a series of dramatic falsifications or irrational conversions, but as a gradual process of competitive refinement.

The framework has found applications well beyond the natural sciences. Researchers in political science, economics, educational theory, and psychology have used Lakatos’s ideas to evaluate the health and productivity of their own research traditions. The distinction between progressive and degenerating programmes continues to serve as a practical diagnostic tool for assessing whether a line of inquiry is generating genuine insight or merely spinning its wheels.

Lakatos also gave us a vocabulary that captures something essential about the practice of science: the idea that there is a rational middle ground between discarding theories too quickly and clinging to them too stubbornly. In an era when debates about the scientific status of fields like string theory, certain branches of social science, and even climate modelling continue to generate controversy, Lakatos’s framework offers a measured, historically informed standard for evaluating scientific progress.

What do you think? Can a research programme that has been degenerating for decades ever truly make a comeback, or should scientists eventually draw a line and move on? And does the fact that Lakatos’s framework works best in retrospect undermine its usefulness as a guide for practicing scientists?

References

1. https://plato.stanford.edu/entries/lakatos/
2. https://en.wikipedia.org/wiki/Imre_Lakatos
3. https://aeon.co/essays/imre-lakatos-and-the-philosophy-of-bad-science
4. https://link.springer.com/chapter/10.1007/978-3-031-88213-5_14
5. https://en.wikipedia.org/wiki/Research_program
6. https://www.qualityresearchinternational.com/socialresearch/methodologyofscientificresearchprogrammes.htm
7. https://pmc.ncbi.nlm.nih.gov/articles/PMC9643948/
8. https://antimatter.ie/2011/02/11/kuhn-vs-popper-the-philosophy-of-lakatos/
9. https://www.cambridge.org/core/journals/perspectives-on-politics/article/abs/perils-of-paradigm-mentalities-revisiting-kuhn-lakatos-and-popper/CB86F968C82C4A3587EDB27126BD5508