Defining “good science”: a fetish for falsification and observation holds back science

Read the Story

Show Top Comments

If you misunderstand “falsifiability” as a willingness to discard an entire theory as soon as it fails to explain a single new phenomenon, it’s not surprising that it’s too restrictive to make science effective. Luckily that’s not really how anyone uses it.


Even Karl Popper, the father of falsifiability, revised his simplistic theory for what he referred to as sophisticated falsifiability, which allowed for the incremental theory modifications which (for instance) is described by Neptune’s discovery. To treat falsification as naively as the author does here is to throw up an aunt sally which does not reflect the reality fo falsifiability.


>Falsifiability doesn’t work as a blanket restriction in science for the simple reason that there are no genuinely falsifiable scientific theories. I can come up with a theory that makes a prediction that looks falsifiable, but when the data tell me it’s wrong, I can conjure some fresh ideas to plug the hole and save the theory. No, you changed it and now you have a different one. >The history of science is full of examples of this ex post facto intellectual engineering. Showing that falsifiability is not a sufficient condition. Certain attitudes are still unscientific, even if your claims are falsifiable. For example, if you say “The moon is made of cheese”, have it falsified to your face but don’t hear that because you put your fingers in your ears, closed your eyes and are singing loudly, well that’s not scientific. So falsifiability is not a sufficient criteria, fine. But that does nothing to support the philosopher’s claim that it would not be a necessary condition. Because if falsifiability is a necessary condition for scientific work, then that circumstance is still pretty damn relevant, which most philosophers contest. All the while failing to provide examples of unfalsifiable yet scientific claims. Just because you don’t have the tools yet to falsify a claim, doesn’t make it unfalsifiable. If you know that you can never have the tools, then yes, that makes it unfalsifiable, like string theory, and also unscientific, like string theorists are painfully aware


This is from 2018, but ok… I think one of the major mistakes of the author is 1. Failing to understand that when you change a theory, it’s a new theory. 2. We don’t go around saying “This is correct.” and then some time later “Actually this is correct”. We are just simply correcting mistakes in existing theories (hence why falsification is important) and therefore the newer theories are less wrong than the previous ones, so over time we get less and less wrong. I agree, that we need some changes in the philosophy of science to be able to address things like the quantum measurement problem, but falsification and observation are here to stay.


Not fond of this essay. A theory that isn’t falsifiable needs no observations as it will invariably be shown true. Paul Feyerabent gives a more convincing account as to why one should oppose method in the context of discovery. When it comes to the context of justification, however, observation and falsifiability are solid requirements for science.