Falsification - the foundation of good science and also good philosophy?


Karl Popper was born in Vienna in 1902.  His early contemporary influences were Marxism and the psychoanalytical theories of Adler, whom he met and worked with briefly on a social project. Popper though takes Astrology as his starting point for his criticism of Marxism and other similar nebulous systems of thought.

Predicting peoples’ futures has been around for ever, with people gulling others by making predictions about their love-life and such like. It then became more formalised by relating their predictions to people’s so-called star signs, so bringing into being the idea of Astrology. This gave a veneer of objectivity to what was being said, because most people believed that stars really did have an influence on their lives.

However, in the newly emerging scientific community, there was also a backlash against it: such was the concern of Galileo not to be taken in by its irrationality, that he rejected the idea that the gravitational pull of the moon caused the tides - it seemed just too much like astrology.  Newton, for the same reason, in developing his theory of gravity, accepted the moon’s influence, but with a lot of hesitation.

As Popper has said in his book ‘Conjectures & Refutations’: “Astrologers have always been greatly impressed, and misled, by what they believed to be confirming evidence - so much so that they were quite unimpressed by any unfavourable evidence. Moreover, by making their interpretations and prophecies sufficiently vague they were able to explain away anything that might have been a refutation of the theory, had the theory and the prophecies been more precise. In order to escape falsification they destroyed the testability of their theory. It is a typical soothsayer's trick to predict things so vaguely that the predictions can hardly fail: that they become irrefutable.”.

He found similar fault with the psychoanalytic theories of Freud and Adler. He rejected them, having realised that: “… every conceivable case could be interpreted in the light of Adler's theory, or equally of Freud's. I may illustrate this by two very different examples of human behaviour: that of a man who pushes a child into the water with the intention of drowning it; and that of a man who sacrifices his life in an attempt to save the child. Each of these two cases can be explained with equal ease in Freudian and in Adlerian terms. According to Freud the first man suffered from repression (say, of some component of his Oedipus complex), while the second man had achieved sublimation. According to Adler the first man suffered from feelings of inferiority (producing perhaps the need to prove to himself that he dared to commit some crime), and so did the second man (whose need was to prove to himself that he dared to rescue the child). I could not think of any human behaviour which could not be interpreted in terms of either theory. It was precisely this fact - that they always fitted - which in the eyes of their admirers constituted the strongest argument in favour of these theories. It began to dawn on me that this apparent strength was in fact their weakness.”
Marxism was also put forward as a scientific theory, making predictions as to how the world would develop. As Popper goes on to say: “The Marxist theory of history, in spite of the serious efforts of some of its founders and followers, ultimately adopted this soothsaying practice. In some of its earlier formulations (for example in Marx's analysis of the character of the 'coming social revolution') their predictions were testable, and in fact falsified. What was predicted did not happen”.

Even now, some 150 years on, the predicted downfall of capitalism is still somewhat overdue. “Yet instead of accepting the refutations, the followers of Marx re-interpreted both the theory and the evidence in order to make them agree. In this way they rescued the theory from refutation; but they did so at the price of making it irrefutable. Typically, they introduce “some ad hoc auxiliary assumption, or re-interpret the theory ad hoc in such a way that it escapes refutation.”. They removed the specific predictions and retained only generalisations, none of which are testable in an objective way. Unsurprisingly, Popper became disillusioned with Marxism and abandoned it fairly early on.

But it left him wishing to find a way of distinguishing between theories he saw as unconvincing, because they could never be disproved on the one hand, and something like Einstein’s theory of relativity which had very specific predictions built in. These have since been tested with increasing precision and have not (yet) shown to be wanting.

And so the possibility of refutation of a conjecture, its falsifiability, became his test of what should, in effect, be taken seriously as science and what should be regarded as metaphysics. The process of science, according to Popper, is conjecture (creating a hypothesis) and an attempt at its refutation – you have to attempt to falsify the hypothesis. You need to set up an experiment to try to prove your hypothesis wrong. If it is disproved, you have to reject it. Simples.

Popper came to this conclusion at a time when the general view was that the scientific method was all about verification. Whilst verification intuitively sounds like the best way to go, it really isn’t. As Popper saw with Freudian and Adlerian psychoanalysis, Marxism and astrology, a wish to provide verification generally meant that people look only for evidence to support their view of things, rather than looking for inconsistencies.

Now you might think that verification and falsification are two sides of the same coin, but they’re not. In fact verification was an example of trying to impose standard philosophical methods on science without really understanding its role. Verification is fine if you have a logical puzzle: you can check for errors in logical reasoning because it follows pre-defined rules. If there are no flaws, then the proof is accepted. It is verified.

Unfortunately, the world about which science is trying to inform us, doesn’t work like that. It’s completely different because we don’t start off by knowing what the rules are. We start instead from a set of observations of what is going on and then try to infer from our observations the underlying rules which govern what is happening.

If logic alone, rather than experiment, had actually been sufficient to give us the structure of the atom and otherwise sort out all the problems of science, then the ancient Greeks would have had mobile phones and would have been talking about the Big Bang rather Zeus and his bolts of lightning.

Instead, we have been on a voyage of discovery over very many millennia in trying to understand the rules governing what happens under what circumstances. And we’ve gone down many blind alleys. Thanks to the thinkers of the Renaissance, we did eventually decide not simply to adopt those rules which were thought to fulfil a requirement for ‘beauty’, or for consistency with how the ancient Greeks believed things worked, or for their supposed alignment with religious teaching.

Instead, we took speculations – hypotheses – and conducted experiments to see whether the hypotheses, those assumed rules, really did explain our physical reality. And often they didn’t. A theory might, even for many centuries, seem to have been ‘verified’ by experience or even a whole raft of experiments, but nature has a way of coming up with exceptions to the rule when you least expect them. Perhaps the most famous case of falsification is the discovery that Newton’s ‘Laws’ of Motion do not actually apply at velocities close to the speed of light. Thank you Einstein. Which takes us back to Popper and the counter-intuitive need for falsification.

Popper held that a theory which cannot be tested and so cannot ever be shown to be false, is not worthy to be called a scientific theory. It may be a metaphysical speculation, but it is not science. Popper though was speaking at a time when certain very problematic scientific ideas had not yet come into peoples’ heads. I’m thinking here of things like the multiverse, string theory, dark matter and energy and so on. These are far beyond our ability to conduct experiments and so are not in practice falsifiable. They are though possibly in principle capable of being falsified, unlike, say, the alleged existence of invisible fairies at the bottom of my garden. It is just that they are on a scale such that we do not have the physical means at the moment to do the research. But then much of our present day experimenting would not have been possible even 20 years ago. Think of the Higgs particle.

But whether we call these (currently) untestable hypotheses scientific theories or metaphysical speculation is not really of great importance. Either way, they will have to remain on the back-burner for the time-being.

And then Popper also held it to be very important that we should not try to change a theory by putting in extra variables retrospectively to make the experimental data fit the hypothesis. In this, said Popper, lies the great distinction between science and pseudoscience: the latter will try to protect itself from disproof by ‘massaging’ its theory in the light of conflicting facts. He could see the obvious truth that pseudosciences, such as astrology, have a certain ‘plasticity’ which enable them to be adjusted to fit whatever inconvenient facts are thrown at it. Today, we have the anti-vax movement which, every time one of their silly ideas is shown to be wrong, will come up with another equally incoherent idea as to why the vaccine is dangerous or the virus is a hoax and all part of a great conspiracy. In real science, Popper said, you can’t adjust the hypothesis to fit the data.

But in the form, at least initially put forward by Popper, his great theory has always been susceptible to criticism, in fact, some would say to being falsified. Various arguments have been made against it:

1.    Some would say that it is not right to suggest that the hypothesis should never be tweaked after the results come in. The results may genuinely suggest an additional factor of relevance. However, when this factor is then taken into account, this means that we are in fact trying to show that the new resulting hypothesis is false and not the original version. And it goes without saying that it must itself be falsifiable

2.    We have to be very cautious about upending well-established laws of physics – rather, they would say, our experiment results should be regarded with deep suspicion. Which is true, but not a reason for abandoning the idea of falsification.

3.    A more serious argument was put forward by Marian Talbot in her article 'Critical reasoning' (issue 106 of Philosophy Now). She said: “Karl Popper argued that we never inductively confirm a theory, and must instead be content with falsifying it. The only thing we can justifiably claim to know, he argued, is that a theory is false, never that it is true. However, in claiming this, Popper was tacitly relying on induction: after all, what makes us think that a theory that has been falsified on one occasion will be falsified on the next occasion, if it isn’t the inductive belief that the future will be like the past?“.

Of course, a theory found to be false by experiment one day may not be shown to be worthy of that ignominy the next day. This could be simply because the experiment was not conducted properly the first time and so was not in fact a falsification of the theory. Such a possibility cannot be ruled out. In the absence of error, however, if we accept that we live in a world governed by rules rather than magic, then the only other possibility for a change in the outcome of an experiment is that the laws of physics have changed.

Scientists routinely assume the invariance of the laws of physics in time and space, but also accept that such an assumption may not be justified. And if it is not the case, then experimental results may indeed change from time to time or from place to place. In fact, this possibility is quite often mentioned, particularly in the world of astrophysics.

You may say that this possibility justifies the accusation that Karl Popper is encouraging us to rely on inductive logic - because we cannot be sure that the future will be like the past. This though is wrong. The possibility that the laws of physics are not invariant is not a confounding factor which we ignore until it comes along and bites us, but is an assumption implicitly factored into the premises. Nobody bothers to say so, because it is unnecessary.  It would be like a philosopher adding before every proof the extra premise: “the rules of logic are invariant in time and space”. 

In the case of scientific theories, we could explicitly refer each time to this additional premise thus – “if we assume that the world is not based on magic and that the laws of physics are invariant in time and space and if theory x predicts y in circumstances z and if experiment shows that y does not occur in circumstances z, then the theory is falsified.”.  And so we are in fact using only deductive logic to arrive at our conclusion. Necessarily, that conclusion does not tell us categorically that the theory is in fact false, but that it is false given all the premises, including the premise that the laws of physics do not in fact change.

But what these criticisms of Popper’s idea really tell us is that those philosophers who are arguing over the finer points of theory, often ignore the way in which scientists actually do their work: scientists take falsification as a basis and then pragmatically adapt it to the real world.

Their practical answer to the philosophers’ criticisms about the unreliability of experimentation is that, obviously, great care has to be taken in establishing the conditions of the experiment. If not, then the result is hardly worth having. It is also good practice not to rely on doing the experiment only once. Most scientists would do the same experiment over and over again precisely in order to try to eliminate error in the experimental set-up. Peer review is an integral, if imperfect, part of the process.

This is also why if an experiment produces an apparent contradiction of our fundamental understanding of how things work then, rather than immediately accepting the need for a revolution in our thinking, other scientists will attempt to replicate the experiment in their own labs in order to make sure that there wasn’t an error in the way the experiment was conducted.  Others will dream up other ways of testing the hypothesis. A good example of the reaction of the scientific community was what happened when the startling claim was made by Pons and Fleischmann that they had achieved nuclear fusion at room temperature – cold fusion.

Ultimately, however, if no error can be shown in the experimental methodology, then the inconsistency of the hypothesis with the experimental results will act as a very large nail in its coffin. The hypothesis cannot be said to have been falsified beyond all shadow of a doubt, because it might, after all, just have been experimental error. As far as the scientific community is concerned, however, it will have been falsified beyond reasonable doubt. A pragmatic position to take, if not as clear cut as the philosophers - Popper included - would like. Scientists have to live with doubt, knowing that all of science is tentative. Philosophers insist on certainty, even where there can be none. If scientists had insisted upon philosophical certainty before doing anything, then we would not have seen the amazing advances which have been made.

In contrast to the philosophical world, however, some industries and interest groups have very deliberately taken Popper’s philosophy to heart - and used it to resist changes which would have damaged their economic interests.

For decades, the tobacco industry said that the theory that tobacco smoke caused lung cancer could not, of course, be verified – Popper tells us that this is impossible - but neither was it capable of being falsified. There simply wasn’t the data because no randomised controlled double-blind trials could possibly be done in order to see whether whether there was the predicted difference in life-span between smokers and non-smokers.

We have seen and still see a similar line of reasoning with climate change. Here, vested interests are attacking the science, saying that there can be no conclusive evidence that the climate will change in the way the scientific consensus predicts. They say that the modelling cannot currently be falsified because its predictions are about a future world. Then they say that the disaster scenario produced by the modelling has changed over the years and so might change again in a different, more benign, direction. In the meantime, of course data consistent with the modellers’ predictions is coming in thick and fast and showing even faster deterioration in the situation than predicted.

I would not claim that Popper’s idea of falsification is as clear-cut in its application as Popper himself clearly thought at the outset. We do not often have a nice neat theory easily capable of being shown to be wrong. I would though argue that Popper’s insight in turning away from verification to falsification was a rare example of a philosophical idea which managed to escape the bounds, the strictures, of philosophy: in so doing, it became the foundation of modern science. The adoption of what we might call ‘pragmatic falsification’ as the scientist’s tool of choice has enabled the astonishing progress made by science.

Paul Buckingham

22 August 2021

Home      A Point of View     Philosophy     Who am I?      Links     Photos of Annecy