So I've looked over multiple research papers and books and still can't grasp the idea of what the difference between Popper and Kuhn is based on their view of how science works and progresses?

Based on what I've read and managed to come up with is that Popper says scientist always question scientific methods and is made up of unfalsified theories. While Kuhn says science is socialized into a paradigm and they never question the theory.

5 Answers 5


Regarding progress specifically, it might be useful to start with Peirce. Peirce proposed a pragmatist conception of truth as the limit point of the process of empirical investigation and critique — the scientific community gradually approaches a consensus, and Peirce either defines that consensus view as the truth (the standard reading) or suggests that processes of empirical investigation and critique will ensure that the consensus view is true (some Peirce scholars prefer this reading). In short, Peirce sees truth as the end-state of scientific progress.

Popper agrees with that last statement, but has a more specific mechanism in mind that the vague "process of empirical investigation and critique." Popper proposes that scientific progress is the elimination of false theories. This produces Peircean convergence, as proponents of false theories gradually have their preferred theories eliminated and are rationally compelled to accept the consensus view.

With Kuhn, it's important to keep clear whether we're talking about progress within normal science (inside a given paradigm) or progress in the transition from one paradigm to its successor. Within a paradigm, a field of normal science will have goals (applying the conceptual framework to novel phenomena or interesting cases, refining functional forms or the values of constants, confirming important predictions, etc.), and progress within the field can be measured in terms of the achievement of those goals. Between two rival paradigms, partisans will often argue that their paradigm can resolve anomalies for the other side; GTR resolves the anomalous perihelion of Mercury and the lack of an ether wind, that kind of thing. So here you can get a notion of progress in terms of the resolution of anomalies.

So Kuhn as these two notions of progress, within and between paradigms. The critical difference from Popper is that both of Kuhn's notions of progress are local. As Peirce and Popper see it, you should be able to tell a grand historical narrative in which the scientific community — thousands of individuals distributed over time and space — gradually converges on the truth. On Kuhn's view, the transition between paradigms doesn't preserve conceptual frameworks or notions of which cases are interesting or which predictions are important. The important puzzles of paradigm A aren't necessarily the important puzzles of paradigm B, and might be totally different from the important puzzles of paradigm C. Similarly, "resolves the anomalies of" isn't a transitive relation. B might resolve the anomalies of A, and so B might be regarded as progress over A. Later, C might resolve the anomalies of B, and so C might be regarded as progress over B. But C does not necessarily resolve the anomalies of A. (Think of the way non-locality has waxed and waned as a major problem with physics.) So we can't necessarily say that C is progress over A.

On Kuhn's picture, there's local progress, but it doesn't necessarily add up to a grand historical narrative of convergence on the truth. It looks much more like just one damn paradigm after another.

  • Kuhn does, however realize that science does, on an overall and complete scale, make progress. By looking at history we can make sure that paradigms do progress over time by avoiding past failure. Some parts of this 'grand arc' will no longer make any sense to us, as they should were science a single process, but we will not simply go back and forth or around and around. He actively resists attempts by others to make his notion of incommensurability suggest cycling or absolute relativism.
    – user9166
    Commented Oct 17, 2016 at 15:24
  • This captures the idea that Alchemy feels like nonsense without suggesting we will ever have a substance-describing paradigm that is equally weak. We are not headed back to an Aristotelian level of science by sheer drift.
    – user9166
    Commented Oct 17, 2016 at 15:29
  • That sounds somewhat more like Ernan McMullin (cf this piece, 41ff) than Kuhn, at least on standard readings. Here (or here) is a good example of how contemporary philosophers of science read Kuhn on progress.
    – Dan Hicks
    Commented Oct 18, 2016 at 21:40
  • They do so over the protestations of Kuhn himself. Modern folks can insist there is no explanation for the notion of progress in Kuhn, and that he is sorely at odds with Popper, but he does not agree. You can see this in his own pieces in Criticism and the Growth of Science or The Essential Tension
    – user9166
    Commented Oct 19, 2016 at 3:11

As Kuhn keeps insisting, there is not that much difference between what Kuhn and Popper suggest for most science as we see it function. What Kuhn labels 'Normal science' includes most of what Popper identifies as science at all. Where they differ, is upon what happens at major theoretical shifting points. Take Darwinism as an example.

Popper sees what Kuhn is calling a paradigm as simply a large and audaciously risky single question. For instance, you can think of Darwin's entire theory of evolution as a single proposition on species creation. To meet Popper's demarcation criterion, then, you have to imagine what would be necessary to make it falsifiable, given that what is and is not a species is somewhat subjective. Since it originates well before DNA and the other parts of the fossil record that we ultimately tested it against, that leaves Popper confused as to how it succeeds as a theory.

For a significant period of Popper's own career, he did not believe that Darwinism met his demarcation criterion, and declared that it must, instead, be considered a mere philosophy attempting to contribute to science, instead of a scientific theory.

It was not falsifiable as proposed, but it became so, largely on the basis of theories that sprung up after it. So when did it become 'real' science? Why would anyone proceed to study the biology that was not yet 'really' scientific long enough to get it to the point it was testable? So, as in this case, Popper's demarcation criterion sometimes has problems putting cause and effect in the proper temporal order.

Kuhn's way of looking at those periods is to see that what is shaping development is more basic than simply being a single high-risk hypothesis. Darwinism is an entire way of shaping how one looks at biology. For those who found it compelling, Darwinism was not simply a theory, but a new paradigm. Finding data that fit into it, or that tested it became a focus that directed study. And they were not open to easily abandoning it. When it was mature, it displaced the preceding attempts at paradigms and simply took over the discipline as a whole.

Until then, it was contending with other established systems in terms of which it was simply wrong, starting from the assumption that species were individually created and held relationship to one another in terms of an overall design, meant to fit nicely together into niche structures for the best use of the environments available. If one simply took the data at hand and insisted upon a traditional interpretation, every newly discovered species fit nicely into the existing cladism, with no intermediary forms, so Darwin was immediately falsified.

If, instead, you cut the new theory any slack at all, it was hard to see where to stop, and at what point it could be interpreted as having been adequately challenged, and not coddled out of animosity to the reigning sense of predetermined order. There is no logical place to make that cutoff. But the call got made. Darwinism passed the modified tests. So it seems that the tests themselves got chosen on some basis other than sheer falsifiability and experimental challenge.

Kuhn looks at the history of a science in terms of the succession of paradigms, each focusing development in a given way, instead of Popper's alternately smooth and rocky succession of simple theories with different levels of risk and implication, that are not 'layered' in any way, but any one of which faces all the others on equal terms.

It is very hard to really tell these two apart in practice, and Kuhn's theory is layered upon that of his mentor, so they also do not really conflict obviously. Kuhn, in his successive publications, keeps alternately emphasizing this ultimate lack of conflict, and pointing out how much more he favors his own ideas, which, in the end, does not improve our clarity. But the notion of paradigm does seem, to many, to fit the pattern of historical periods of peace and strife within a discipline, as periods where paradigms had been chosen and times when they were in contention with one another.

Kuhn's whole theory seems to have trouble identifying the edges between disciplines that would allow one to identify the given paradigm. E.g. Is Durkheim's study of suicide a sociological, psychological or an anthropological theory. Sociologists claim him, and the other two acceded, but on what grounds? His test criteria are against sociological factors -- national wealth, social class expectations, family structure etc. But the theory is ultimately about an individual, and not a group action, so it is psychology. And the economic, tribal and class factors themselves make sense only in the overall cultural history of Europe, not generalizing to the rest of the world, so it is anthropology.

Given that, one may need to see the layering of paradigms on a number of scales at once, and see Kuhn's own theory through a sort of fractal prism of super-, and sub-disciplines. One can respond to this by falling back onto Popper's unlayered uniformity, or by seeing the layering of networks of assumptions as a more basic characteristic that determines Popper's "level of risk" according to the size of the part of the paradigm that would be reconfigured by finding a conflict.

I propose the latter approach as a general theory of knowledge evolution that applies not just to science, but to social convention, logical foundations, and individual perception.


Strictly speaking, Popper wasn't concerned at all with how science progresses, only with the demarcation problem, that is how to tell the difference between science and pseudo-science.

Kuhn on the other hand, was concerned with the history and progress of science, but ended up discussing the demarcation problem as a result of his investigations into the history of science. In fact that was one of his key insights: That it was impossible to separate the philosophical demarcation problem from the history of science and scientific progress.

Philosophically, the main difference between the two lies in their assessment of the problem of auxiliary hypotheses that faces falsificationism.

Consider the problem of falsifying Newtonian mechanics based on measurements of planets orbits. You measure the orbit of a planet, and if it corresponds to the predictions of Newton's laws, then Newtonian mechanics is confirmed. However if the measurement doesn't confirm with Newton's laws, you are faced with two possibilities:

  • Either Newtonian mechanics (your main hypothesis) is falsified, and we need a new theory of motion.
  • Or, you were wrong about the number of planets in the solar system (this is the auxiliary hypothesis), and an unknown additional planet is causing your calculations to be mistaken. If you accept that there might be an additional planet, than you can still "save" Newtonian mechanics from falsification.

Popper thought that this problem could be solved: It should be possible to tell the difference between (a) reasonable auxiliary hypotheses which help confirm a theory, and (b) unreasonable auxiliary hypotheses which are simply desperate attempts to save a theory which is false.

Others, however, believed that it was impossible to completely separate the auxiliary hypotheses from the main hypothesis being tested (See Duhem-Quine thesis). In the above example, one might respond that we are already using Newtonian mechanics to figure out the number of planets in the solar system, so treating the number of planets as an independently testable hypothesis is wrong, and it can't be examined separately from Newton's laws.

Kuhn was one of those who saw auxiliary hypotheses as inseparable from the main thesis, and they were all bound together in paradigms (For example Newtonian mechanics is one paradigm, Quantum Mechanics is another, etc...). He argued that these paradigms were driven as much by sociological and historical reasons, as they were by falsification and purely empirical evidence.


Imre Lakatos' "Falsification and the methodology of scientific research programmes" (1978) offers a deeply insightful middle ground between Kuhn and Popper by integrating the best parts of each.


The difference between Popper and Kuhn on scientific progress is large and important.

Kuhn follows a long and dishonest tradition in philosophy that uses the following method. Pick examples of people acting like morons and then insinuate that this is a necessary feature of reality without saying it explicitly. So Kuhn produces examples where people who like to call themselves scientists act like complete dumbasses, e.g. - refusing to answer criticisms. He then insinuates that everyone is a dumbass and so that scientific progress is impossible. Now, since he never sez this explicitly, people deny that's what he meant and you can have endless discussions about it without resolving anything. And even Kuhn may sometimes have denied that's what he meant.This makes philosophers look clever to people who don't know any better. Those who do know better have described the sorts of bullshit many philosophers get up to. For a discussion of linguistic philosophers that applies to many other philosophers see Ernest Gellner's book "Words and Things" chapter VI.

Popper pointed out explicitly that just looking at what scientists do is a complete non-starter for any serious discussion of philosophy of science, see "Logic of Scientific Discovery" Chapter 1 Section 2. You can't identify somebody as a scientist just cuz he sez he's a scientist. Creationists and evolutionary biologists both call themselves scientists and accuse the other group of not being scientists. They can't both be right unless science vs non-science is an invalid distinction. So then you have to explain some standard of scientific conduct or stop using the word as anything other than an expression of approval. Popper explains what the standard is: scientific practice involves seeking criticism of theories, including looking for experimental results that contradict the theory in question.

More broadly, Popper pointed out that the main precondition for progress is to take the position "I may be wrong, you may be right and by an effort we may get closer to the truth." Even people with very different ideas may be able to make progress if they have a genuine critical discussion. See the title essay in Popper's book "The Myth of the Framework". Scientists often don't live up to this standard. There are reasons for this. One problem is that our culture has a lot of anti-intellectual ideas some of which are peddled by philosophers. Other problems include the fact that scientists suck up to government to get money.

For a detailed discussion of Kuhn, including his intellectual dishonesty in insinuating Popper is a naive falsificationist, see the introduction to Popper's book 'Realism and the Aim of Science'. Less accessible but also interesting are Popper's replies to Kuhn in "Criticism and the growth of knowledge" edited by Lakatos and Musgrave and "The Philosophy of Karl Popper" 2 volumes edited by Schilpp.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .