I know he criticized Popper's critical rationalism as limited and dogmatic, but critical rationalism isn't like that, because critical rationalism does accept ideas which at first sound illogical and out of tune with accepted knowledge. If those ideas survive scrutiny, they are incorporated into or even change the existing body of scientific knowledge?

Did Feyerabend really understand Karl Popper?

  • Hi, welcome to philosophy SE. Feyerabend's criticisms extended far beyond "limited and dogmatic", or Popper in particular, he objected to having a scientific methodology as such, Popperian or some other, and used historical examples to argue that scientists do not use such "methodologies" in practice, see Is Feyerabend confusing discovery and justification when he criticizes the scientific method?
    – Conifold
    Jan 30, 2017 at 19:06
  • 3
    "In most of his works after Against Method, Feyerabned emphasises what has come to be known as the “disunity of science”. Science, he insists, is a collage, not a system or a unified project. Science is a collection of theories, practices, research traditions and world-views whose range of application is not well-determined and whose merits vary to a great extent. All this can be summed up in his slogan: “Science is not one thing, it is many.”" Jan 30, 2017 at 20:29
  • Popper was 'normative', saying what science should be, rather than describing how people actually practice it.
    – Mitch
    Feb 3, 2017 at 18:07

2 Answers 2


I recently got thinking what would have happened if Popper had been an art critic instead of philosopher of science; how would he have applied his method to art?

Perhaps he would have said lets begin by looking at all the art in the world; when we look at every painting, he says, we see what they all have in common is that they're set in a frame; thus, I deduce that art is defined by that which can be framed; its the frame that is the key to art; it sets its limit and its boundary, anything outside of the frame is not art.

Of course, this is something of a parody - no art critic (pace Duchamp) - thinks like this; but there is something in common with this and with his characterisation of science as being limited or bordered by the falsifiable; anything outside of that - the unfalsifiable - is not science.

Feyerabend, argued for a more organic understanding of method in science; which is why his book is called against method.


Feyerabend didn't understand Popper. The standard philosophy of science, then and now, claims that there is a process called induction that somehow makes observations into scientific theories and then makes them true or more probable or something like that through more observation.

As Popper pointed out, this process as described by its advocates is impossible to follow. And you can't do something impossible, so nobody actually does induction. Rather, all knowledge arises by evolutionary processes involving variation and selection. The only role experiments can play in this process is to provide problems for theories that currently seem unproblematic. Theories are created by guessing not by any process that involves deriving them from experimental data. A person might be reminded of something by looking at experimental data and he might pursue that idea and come up with an explanation. But a dream, or eating some yoghurt or whatever might play the same role in his creative process. This doesn't mean yoghurt contains the secrets of the universe. Popper then makes methodological suggestions of things that scientists could do to improve their ideas, e.g. - avoid ad hoc theories and only propose ideas that can be tested independent of the problem they were invented to solve.

Feyerabend didn't understand any of this and didn't have anything relevant to say about Popper, or about anything else. He was just another relativist.

  • "Popper then makes methodological suggestions of things that scientists could do to improve their ideas, e.g. - avoid ad hoc theories and only propose ideas that can be tested independent of the problem they were invented to solve." -- But according to Quine, that's not possible, how would Popper respond to Quine? Feb 2, 2017 at 17:55
  • If you find a contradiction among your ideas, any of those ideas could be wrong. The conventionalist response to this is to decide by convention which idea is wrong and to say there is no way to sort out which idea is actually wrong. Popper's response is that you should guess which idea is wrong and why and what should replace that idea. For example, if you observe en event that contradicts Newtonian mechanics with a telescope, the theory may be wrong or the telescope may be faulty or something else might be wrong. Each specific suggestion will have different implications. to be continued.
    – alanf
    Feb 3, 2017 at 11:39
  • You might guess the telescope is faulty. You do other observations with the telescope, take out the lenses and test them, test mechanisms for adjusting the length of the telescope and so on. These tests might be consistent with the telescope having a fault. You fix the telescope, do the observations and the problem no longer occurs and you conclude the telescope was faulty. The only reason to be dissatisfied with this idea is that you are looking for a guarantee of truth or probable truth, which is impossible, as Popper pointed out in 'Realism and the Aim of Science' chapter I.
    – alanf
    Feb 3, 2017 at 11:40
  • Hmm Ok. What about the following historical situation: You look in your telescope and the orbit does not correspond to the predictions Newtonian mechanics. Is Newton wrong or is there a new planet we weren't aware? Feb 3, 2017 at 18:50
  • "Is Newton wrong or is there a new planet we weren't aware?" Dunno. You guess a solution to the problem and then test the guess. Newton + some unknown planet causing the disturbance places constraints on where that planet could be. And this planet might disturb other orbits. So the existence of that planet has implications beyond the problem of the wonky orbit you happened to see. You can look for where the planet should be if the theory is true, disturbances in other orbits and so on. If you can come up with an alternative to Newton it should also be testable, like general relativity.
    – alanf
    Feb 4, 2017 at 15:26

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