When we talk about something being ‘pseudoscience’ what we tend to mean is that it’s ‘not true’, and we reach that conclusion because what we mean by pseudoscience is something that is ‘not scientific’, and we associate the sciences with truth. Yet the alternative to truth is not automatically falsehood; there is also ambiguity, indeterminacy, and incompleteness to consider. What’s more, if we call things scientific only if they are true, we are admitting that we don’t actually know what is or isn’t scientific until some future time when the arguments about some given topic are finally resolved. There is a confusion here worth examining closely.
Ask someone to explain how the sciences work and chances are they will tell you about the scientific method:
1. Observe a situation
2. Come up with a hypothesis (an untested theory) to explain a phenomenon
3. Devise an experiment to test whether the hypothesis is valid
4. If the experiment is successful, the hypothesis becomes a theory. Congratulations you’ve discovered scientific truth!
This description is so far from adequate that it is a wonder that so many university students are taught it! Quite apart from the way it sets aside the most difficult aspect of scientific practice (the interrelationships of existing knowledge on any subject) it fancifully imagines that scientists determine truth simply by performing just one experiment, as if scientific truth were as simple as revealing a scratch card – three microscopes, we have a winner! Rather than an adequate description of how contemporary scientific processes operate, this is more akin to a catechism recited in order to bolster faith in the ability of the sciences to reveal truth – and as such, it obfuscates the complexity of the relationships between experiments, theories, and truth, and prescribes a method almost certain to lead to error every time.
If a hypothesis and experiments are indeed the necessary elements of a claim that a certain activity is ‘scientific’, then anthropology, economics, almost all of the evolutionary sciences, and a fair amount of biology and medicine are all doomed to be ‘unscientific’. These kinds of accusation are indeed sometimes advanced – a furore occurred in 2010 when the American Anthropological Association decided to removed the word ‘science’ from its mission statement, despite many of its members feeling this was a consequence of a narrow and reductionist description of the sciences. There are also questions here about concluded research programmes: no-one has needed to perform further experiments in optics, for instance... has it ceased to be scientific? Or did it earn its place in scientific heaven by being a good research field while it was still alive...?
Tied up with this confusion is the idea that the sciences are ‘value free’, i.e. that scientific research is inherently unbiased. This is a naive mistake to make, and on two counts. Firstly, as Nietzsche warned back in 1882, we are “still pious” when it comes to scientific truth – all scientific research rests on a core moral value, a commitment to the pursuit of truth. Without this, the endeavours we undertake in the name of science make no sense; ‘valueless science’ is entirely implausible. Secondly, and even more importantly, scientists are still human, and as such they have their own values. The attempt to purge the sciences of values is nonsensical and indeed impossible! No matter how much you try to present scientific research as a purely rational, emotionless, valueless activity, scientists will continue to pursue research motivated by their own moral values (to save lives or to save the planet, to advance knowledge or technology, to win fame or wealth etc.). The idea that having these values is somehow unscientific is to doom all the sciences to oblivion! The values and the facts are intimately related or, as Hilary Putnam described it, entangled. The idea of a science without values is pure nonsense.
At this point, you have a choice in how you respond to this critique of ‘scientific method’, and this in itself may be illuminating. On the one hand (and especially if you’ve spent any time at all thinking about philosophy of science), you can happily cast off this quite ridiculous dogma and still maintain a viable understanding of the sciences without it. That’s the easy way... but it still has some hard consequences. Or alternatively you can dig in your heels and try to cast out the demons of those that don’t follow ‘the method’, attempting to purify research of pseudoscience, meaning in this case ‘not following the scientific method’, but usually playing out by simply deriding counter-claims against whatever dogmatic position has been adopted on any given point. That path is so misguided it’s a wonder that plenty of otherwise intelligent people seem to fall for it.
As it happens, the sciences themselves show us why this purported ‘scientific method’ is unworkable. Psychology – which has been staunchly dedicated to ‘the method’ yet still gets cast out as ‘soft science’ – has provided a lot of neat titles for the various kinds of human bias. Defenders of ‘the method’ like to evoke hindsight bias to defend the need for hypotheses – “if you don’t make a hypothesis, you’ll just end up seeming to expect the result you get!” But these cognitive biases cut both ways: if you do make a hypothesis, you are now prone to confirmation bias – cherry picking your data and references to support the position you have chosen. This is why medical sciences insist on good quality evidence from randomized trials where even the experimenters don’t know what’s going on until all the data is in. We know from bitter experience that when you set out to prove some specific claim, you are more likely to find (and report) the evidence that supports what you have chosen. In other words, not having a hypothesis condemns you to bias, and having a hypothesis condemns you to bias! What makes something legitimately scientific cannot be the elimination of bias, or else nothing could ever be sufficiently purified of values to qualify. There has to be another way of conceptualising the difference between ‘science’ and ‘pseudoscience’ if either is going to have any legitimate meaning.
Ghosts of Science Past
The celebrated historian of science, Thomas Kuhn, lays out the question of pseudoscience at the very outset of his project to understand the nature of scientific change. The problem as he presents it is that if we judge the historical precedents to our scientific practices as pseudoscientific (he talks of them being ‘myths’), then we have to acknowledge that pseudoscience can be pursued and justified by the same methods and reasons we now use to defend science against its alternatives. Yet if we call these artefacts of older research ‘science’, then we have to accept that the sciences were beset by wild beliefs that today we would find unthinkable (even abominable). He argues very persuasively that from a historical perspective we have no choice but to accept that “out-of-date theories are not in principle unscientific because they have been discarded.”
Kuhn’s position is widely accepted today – yet it runs directly contrary to the view of Sir Karl Popper that the boundary of legitimate science is falsification – the ability to have a theory proven false. Amazingly, this viewpoint is also widely accepted today, even though the two approaches are essentially incompatible, and indeed were the basis for an unresolved despite between the two academics. Kuhn saw Popper’s falsification as applying solely to those rare periods of scientific upheaval (paradigm shifts) where one way of thinking replaces another. His view was that ‘normal science’ never dabbles in big theoretical changes at all, but is always about solving problems using the current theoretical apparatus. Again, these two viewpoints are entirely incompatible, yet both are widely supported views on the sciences.
Popper suggested that Kuhn’s approach committed him to saying that astrology is a science because it entails problem solving within its own paradigm. Kuhn denied this, and argued that in the context of astrology “particular failures did not give rise to research puzzles” and thus astrology was never a science. Both men died without resolving their disagreement; I think it clear, however, that both are wrong about astrology. We cannot – as Kuhn himself warns – back-project our current scientific judgements upon prior practices that were claimed as sciences at earlier times without distorting what we are trying to assert. To do so is to deny the very capacity for scientific revolutions that Kuhn’s account provides. The suffix ‘-ology’ by itself is a clue that the practices of astrology had at one point in its history a claim to knowledge, and the question of whether astrology was ever a science in Kuhn’s terms is a historical investigation requiring far more application to the task than either Popper or Kuhn were willing to commit. As such, this question is in fact still very much open to debate! But nobody wants to do so, because everybody with any skin in this game wants to show that astrology isn’t a science and never was – thus again preempting any possible research except that which will prove this one tenuous point.
If Kuhn’s historical theory (albeit not Kuhn himself) is able to defend against Popper’s attack, Popper’s falsification criteria has no equivalent defence against Kuhn’s criticisms. Indeed, Kuhn expressly doubted that falsifying experiences ever really happen. He did not need the psychologist’s label ‘confirmation bias’ to realise that giving up a scientific paradigm is a major conversion for anyone (comparison with religious conversion is quite justified here), made all the less likely by the problem that if every failure of a theory in the face of contradictory evidence were sufficient grounds for rejecting it, all theories ought to be rejected at all times! That’s because the very reason that Kuhn’s ‘normal science’ has problems to solve is precisely that no theory is capable of fitting all the observations it seeks to explain. As the French science studies scholar Bruno Latour puts it, the theories are all under-determined with respect to the evidence – and this conclusion is unavoidable if you spend time examining what scientists actually do rather than merely reciting the catechism.
But this does not mean there is no way of distinguishing science from pseudoscience, even though we have to accept a certain amount of historical contingency after Kuhn (or Foucault – he gets to the same place via a different route). What we might reasonably suggest as a provisional criteria for calling something ‘pseudoscience’ is a combination of Popper and Kuhn’s claims: when even the possibility of falsification is removed, or when the investigative practices cease to produce further enquiries in response to the questions the previous research implies, the claim to be scientific evaporates. As chemist-turned-philosopher Isabelle Stengers attests, successful experiments in the sciences give rise to new research questions. When they do not produce any more, it is because the field has managed a complete description of its subject matter (as with optics). The difference here is that such ‘completed’ fields have produced theories capable of making unfailing predictions. And such cases are vanishingly rare.
The Condition of Pseudoscience
What tied us up in conceptual knots here, and kept Popper and Kuhn from reaching an accord, is that we want to level the accusation ‘pseudoscience’ at fields like astrology or phrenology. But understanding the sciences as an ecology of practices, as Stengers has brilliantly discussed, shows that this is not the only way we might identify a breakdown of Kuhn’s ‘normal science’. We could (indeed must) give up the idea that ‘pseudoscience’ is a way of trashing any theory, research, or evidential claims we don’t agree with. On the contrary, I propose that the clearest way of understanding pseudoscience is as a condition within a scientific discourse that undermines or destroys its power to investigate.
Thus, to continue with phrenology’s original models of mental function after animal experiments began to show that its suggested brain regions did not hold up to scrutiny would have been to enter into a condition of pseudoscience, because its practices could not produce viable new research questions in the light of this new evidence. It would, however, be wildly unfair to it to suggest it was always in this condition: it is from phrenology, after all, that the idea of the brain being the organ of the mind originated, and while most of its specific claims did not pan out, it remains an important part of the backstory of neuroscience. If phrenology had not become spread around as working class ‘popular science’ (thus earning the enmity of Victorian cultural elites), we might well have kept the name ‘phrenology’ (science of the mind) rather than renaming brain research ‘neurobiology’. It’s not at all clear to me that phrenology was ever in the condition of pseudoscience, except perhaps at the very end – although anyone practicing it today would be behaving very oddly indeed.
Pseudoscience is thus akin to an ailment afflicting scientific practices that have become shorn from the logic of legitimacy provided by their current paradigm. The sign that a field has fallen into pseudoscience is not the truth or falsehood of its claims as such. Indeed, these will frequently not be in any way settled, forcing us into highly suspect retrospective accusations, such as that levelled routinely at phrenology. Rather, you can see the condition of pseudoscience occurring whenever scientists give up the values that motivate their enquiry - when they purposefully falsify data, or conceal it ‘to defend the truth’, or give up experiments and data gathering entirely in order to maintain a status quo based upon whatever happens to have been previously claimed. And once we see this, we are forced into the realisation that we are currently in the condition of pseudoscience in several entirely legitimate research fields, and over the last year we have had the audacity to defend the breakdown in the medical discourses that has put us into a state of collective pseudoscience as “following the science”!
The truth is, we cannot ‘follow the science’, it is the science that must follow us. For the values of science are those of discovery and verification, and this only has a purpose in so much as it serves to resolve those questions our other values compel us into exploring. Thus, while medicine commits to ‘first, do no harm’ as a supreme value governing its own practice, that particular principle sets no positive goal at all. The medical practitioners and the cybernetic networks supporting them take on the objectives that we have collectively given to them. If the circumstances that follow from that pursuit make falsification of a medical claim impossible, or provide no means to reliably answer the relevant medical questions, those medical practitioners affected (and anyone trusting their judgements) enter into the condition of pseudoscience, a (temporary) renunciation of the values of scientific practice, capable of precisely the great harm doctors are sworn to avoid. For the collective medical power we exercise cybernetically always causes some degree of harm along with the pursuit of its goals – requiring medical practitioners, on pain of becoming (temporary) pseudodoctors, to commit to studying the impact of any procedure or intervention attempted or else risk violating all the values of contemporary medical science. This is an extreme example, but it is also an extremely important one.
Now whether the values of discovery and verification have always conditioned the work of scientists, and whether they always will isn’t the point, for they are our moral requirements for the sciences now and on this point we quite miraculously do not disagree. In so much as pseudoscience is a phenomenon, it is merely a consequence of recognising that scientists are human, and what makes them seem otherwise is the remarkable power that they bring to bear when cybernetically linked into singular networks, working together – not just by co-operating but just as importantly by disagreeing, refining the research questions by honing the essential ambiguities into points sharp enough to penetrate our ignorance by pursuing further investigations and experiments. Pseudoscience prevents that dialogue from happening, and breaks up the network connections, making research harder or preventing it entirely, setting bias against bias and thus blocking the communication essential to verification, which is necessarily a distributed activity.
When verification stops, pseudoscience has begun... it goes away when we can go back to listening to those objections that our human bias prevented us from hearing. The ugly truth of it all is that fear, anger, and self-righteousness spread pseudoscience all too easily, yet banishing it is as easy – or as impossible – as going back and listening to the objections in order to work out where in the maze of ambiguity, indeterminacy, and incompleteness the truth of each disagreement can be found.
More philosophy of science soon.