How To Tell Science from Pseudoscience

(prepublication draft)

Stephen Law


What is pseudoscience? Most of us will intuitively class more or less the same phenomena together under the umbrella of 'pseudoscience'. Paradigm examples include astrology, Young Earth Creationism, Christian Science, feng shui, homeopathy, flat earthism, and Chinese medicine (though there are contested borderline cases: not everyone agrees about the status of Freud's psychoanalytic theories, for example). But while it seems most of us recognise pseudoscience when we see, providing an adequate philosophical definition of pseudoscience is not so easy. The aim of this chapter is to survey some of the suggestions that have been made, and to make a recommendation of my own.

Necessary and sufficient conditions

Asking 'What is x?' type questions is a traditional philosophical occupation. Philosophers ask: 'What is justice?', 'What is truth?', 'What is the mind?', 'What is science?' etc. Coming up with a philosophically adequate answer to such questions is often assumed to involve producing a list of necessary and sufficient conditions.

Some terms can easily be defined in terms of necessary and sufficient conditions. For example: something is a triangle if and only if (abbreviated by philosophers as iff.) it is a three-sided plane figure. Being a three sided plane figure is sufficient to qualify something as a triangle. It is also a necessary condition for something to qualify as a triangle. We can similarly define vixen (something is a vixen iff. it is a female fox) and bachelor (someone is a bachelor iff. they are both unmarried and male).

However, once we switch to traditional philosophical questions such as 'What is truth?', 'What is the mind?', and so on, the task of specifying necessary and sufficient conditions becomes much more difficult. In particular, we often quickly run up against counterexample to our proposed definitions.

In fact, this sort of difficulty can arise even when trying to pin down the necessary and sufficient conditions to qualify as something as mundane a chair. Define a chair as an object made for sitting on for example, and a philosophical critics will point to counterexamples, e.g. (i) objects not made to be sat on that are nevertheless chairs (a conveniently shaped boulder, placed next to a garden table, can become a chair, despite not having been fashioned to be sat on), and (ii) objects that are made to be sat on that are not chairs (e.g. sofas, bicycle saddles). So, our proposed definition of a chair provides neither a necessary nor a sufficient condition for something to qualify as a chair. Further refinements to the definition will likely face other further examples. For example, if, in order to deal with the sofa counterexample, we suggest that something is a chair iff. it is on object made for one person to sit on, the bicycle saddle counterexample still remains, and so does the stool (neither are chairs, but both are made for one person to sit on).

The philosophical activity we are engaged in here - that of trying to pin down the necessary and sufficient conditions for something to be (an) X - can of course be pursued with respect to pseudoscience. Definitions of pseudoscience are offered, but we again run up against counterexamples.

For example, we might be tempted to define pseudoscience as any attempt to explain that appeals to the supernatural. However, while a great deal of pseudoscience is indeed bound up with belief in the supernatural (Christian Science, Young Earth Creationism, and, on some versions, astrology all involve supernatural elements), this suggestion faces obvious counterexamples.

First, there is no reason in principle while a theory that appealed to the supernatural could not be genuinely scientific. There can in principle be good scientific evidence in support of some supernatural claims. Many supernatural claims have been scientifically tested - there have been two multi-million dollar experiments, involving well-conducted science, into whether petitionary prayers for heart patients have some positive medical effect (see Benson et al. 2006 and Krucoff et al (2005)). Both studies found prayer had no effect. However, these studies could potentially have produced good evidence for the effectiveness of petitionary prayer. True, the supernatural is often assumed to be beyond the remit of science to investigate, but this assumption is false. The mere fact that the supernatural is unobservable is no obstacle to it being scientifically investigated or the focus of a properly scientific theory. Subatomic particles are similarly unobservable, as is the distant past of this planet, yet both are the proper focus of scientific theories that are well-confirmed.

Secondly, not only is involving the supernatural is not a sufficient condition for pseudoscience, neither is it a necessary condition. Many examples of pseudoscience do not involve the supernatural (flat earthism, for example). 

Given the difficulties involved in providing a watertight definition of 'pseudoscience' in terms of necessary and sufficient conditions, some may pessimistically conclude that the term is, then, just a 'hollow phrase' - a term used by speakers to express nothing more than their disapproval of a theory. Rather than pick out some objective property or kind, the use of 'pseudoscience' is more like our use of 'weed', which does not mark any objective difference between the plants growing in our gardens, but merely reflects our own personal preference as to what we like to see growing there. The philosopher Larry Laudan (1983) famously came to just this conclusion about 'pseudoscience' in part because he noted the kind of difficulties I have outlined above in terms of providing necessary and sufficient conditions.

I am less pessimistic about the term 'pseudoscience'. Certainly, our inability to provide a watertight definition of 'truth', or 'the mind', or even 'science', in terms of necessary and sufficient conditions does not, by itself, justify as us in condemning these expressions as lacking any substantive content. So why be so quick to condemn the term 'pseudoscience' on that basis?

Family resemblance

Wittgenstein's (1958) work on family resemblance is also relevant here: Wittgenstein explains how a term can be properly contentful even if, as a matter of fact, it is impossible to define in terms of necessary and sufficient conditions. Wittgenstein remarks that some concepts, such as that of a game, have a 'family resemblance' character. The various members of a family may resemble each other to differing degrees, despite there being no one feature (the big nose, the bushy eyebrows the lopsided mouth) that they all share. Wittgenstein notes that there appears, similarly, to be no one feature that all games have in common. Some games are competitive, but some are not. Some involve balls, but some do not. And so on. So what is the one thing that all and only the games have in common? There need not be anything - just a series of overlapping similarities. Yet the concept of a game is legitimate all the same. Our inability to provide a definition of the sort that we can provide for 'triangle' or 'vixen' should not lead us to abandon our use of the term 'game'.

Of course, a critic might argue that even if there's no one thing all and only games have in common, still, there should be a precise algorithm specifiable that determines what is a game and what is not.

Such an algorithm might require, say, that for something to be a game, at least three out of a set of six characteristics must be possessed (but any three, so there need be no one characteristic all the games share). Not that such an algorithm would provide a necessary and sufficient condition for something to qualify as a so-and-so (the condition being that at least three of the six characteristics are possessed).

However, Wittgenstein thought even such an algorithm was unspecifiable when it comes to games The use of the term 'game', thought Wittgenstein, is such that it is not everywhere marked by sharp boundaries, or even by any boundary at all. Therefore, any attempt to provide a definition in terms of necessary and sufficient conditions must fail - it will involve drawing boundaries where none exist. Yet, for all that, 'game' is perfectly serviceable term.

How should we explain to someone what a game is? I imagine that we should describe games to him, and we might add: 'This and similar things are called games.' and do we know any more about it ourselves? Is it only other people whom we cannot tell exactly what a game is? ... But this is not ignorance. We do not know the boundaries because none have been drawn ... We draw a boundary for a special purpose. Does it take that to make the concept useable? Not at all! (1958, p 69)

Pseudoscience might, similarly, be a family resemblance concept. The suggestion that it is a family resemblance concept has been made by a number of philosophers including Massimo Pigliucci (2013).

Having placed the debate over how to define 'pseudoscience' in some philosophical context, we now turn to the work of Karl Popper. Popper introduced the so-called 'demarcation problem' of identifying what distinguishes pseudoscience from science. Popper's suggestion, as I explain below, is that, unlike real science, pseudoscience is unfalsifiable.

Popper and falsificationism

Popper was a philosopher of science particularly concerned with a notorious puzzle: Hume's problem of induction

Why do we suppose the sun will rise tomorrow. Well, we have seen to rise countless times before, and so we conclude that it will very probably rise tomorrow too. This is an example of inductive reasoning. We move from premises, such as:

The sun rose on Monday.

The sun rose on Tuesday.

The sun rose on Wednesday.

The sun rose on Thursday.

The sun rose on Friday.

The sun rose on Saturday.

To a conclusion, e.g. :

The sun will rise on Sunday.

The mark of an inductive argument is that its premises are supposed to support, but not logically entail, the conclusion. Note that there is no logical contradiction involved in supposing the premises of this example are true while the conclusion is false. So the premises don't logically entail the conclusion. Still, we suppose the premises support the conclusion - that they provide grounds for supposing the conclusion is true.

Arguments of this sort are required to provide us with substantive knowledge of the unobserved (for substantive claims about the unobserved are never logically entailed by just statements about what has been observed). Science, in so far as it makes substantive claims about the unobserved (which is what any scientific theory does - it may predict what will happen tomorrow, for example), must then rely on inductive reasoning.

Hume famously questions whether the premises of such an inductive argument give us any grounds at all for supposing their conclusions are true. Hume suggests that when we reason inductively, we make an assumption: that nature is uniform. We assume that the local patterns we observe are likely to continue over the horizon into the unobserved portions of reality. Without that assumption, thought Hume, such reasoning is unfounded.

So how might we justify the assumption that nature is uniform? That nature is uniform is no a priori logical truth (there's no logical contradiction involved in supposing nature is not uniform). But neither can we justify the assumption by appeal to experience, for (i) we cannot directly observe nature is uniform thought (for we can observe only a small fragment of it), and (ii) we cannot infer nature is uniform throughout on the basis of the parts we have observed (because that would itself be an as-yet-unjustified inductive argument - we would be using induction to justify induction - a hopelessly circular justification). Hume concludes that the assumption that nature is uniform is therefore entirely unjustified and that consequently inductive arguments fail to provide any justification whatsoever for their conclusions. We seem forced to accept a very radical scepticism about the unobserved, a scepticism that renders all scientific theories entirely unjustified. If Hume is right, it's as reasonable to believe the sun will rise tomorrow, given what we have observed, as it is to suppose that an enormous bowl of cherries will appear over the horizon instead. 

Popper's solution to this problem is ingenious. Rather than attempting to show that we are justified in believing our current scientific theories are true, Popper accepts that we are not justified, but that this does not matter. Science progresses by theories being put forward and then tested. For example, having noticed that this object fell when released, I may develop the theory that all objects fall when released. I can now test this theory, dropping pens, bricks, feathers and so on. If any object fails to fall when released, that logically entails my theory is false. Such an observation will falsify my theory. Popper maintains that science progresses, not by theories being inductively confirmed, but by theories being falsified. We are not justified in supposing our current, unfalsified theories are true, but it's reasonable for us to prefer those theories to those that have been falsified.

However, this is not to say that, on Popper's view, all unfalsified theories are equally preferable. Popper maintains we should prefer those unfalsified theories that are more falsifiable over those that less falsifiable.

A theory can be more falsifiable by being clearly stated, preferably in terms of the mathematically quantifiable and measureable. The theory that all adult dogs are 'heavy-ish' is a vague claim that can be easily protected from falsification by insisting that what the term means has been misunderstood ('No, Chihuahuas are heavy-ish!'). The hypothesis that all adult dogs weigh over 3kg, on the other hand, can be straightforwardly falsified with the aid of a scale.

A theory can also be more falsifiable by being wider ranging. The theory that all adult dogs weigh more than 3kg is more falsifiable than the hypothesis that all dalmations weigh more than 3kg because any observation that falsifies the former will falsify the latter whereas the reverse is not true.

On Popper's view, scientists should work with the most falsifiable unfalsified theories.

One of the most obvious difficulties facing Popper's initial falsificationist account of how science should progress is that theories can often easily be protected from being falsified. Most theories do not directly entail observation statements. In order to derive a testable prediction from a theory, various auxiliary hypotheses are often required. As a consequence, when a theory might appear to have been falsified, its defenders can insist it is not the theory that is at fault but one or more of auxiliary hypotheses. For example, from the theory that all adult dogs weigh more than 3kg, I can predict that Fido will register over 3kg on this scale only if I assume this scale is reliable. When Fido fails to register over 3kg, I can defend my theory by blaming the scale.

As Popper further developed his falsificationist theory of how science progresses, he allowed that it was indeed possible to protect a theory from being falsified by making various adjustments elsewhere - e.g. diverting the falsification onto some auxiliary hypothesis. He acknowledged that a simple, conclusive falsification is not always achievable and indeed he provided examples of how theories that might appeared to have been falsified were actually, quite legitimately, retained.

One such example involves Newton's theory of universal gravitation. That theory predicts a smooth elliptical orbit around the sun for the planet Uranus. However, when the observations were made, Uranus moved in and out of the predicted path. Was Newton's theory abandoned? No. Instead, scientists posited another, as yet undiscovered planet, moving beyond Uranus that was pulling it in and out of the predicted orbit. The scientists calculated where the mystery planet would need to be to have that effect on Uranus, looked, and discovered a new planet - Neptune. Popper considered this a great scientific step forward.

This example illustrates Popper's acceptance that theories can be protected from falsification by deflecting the falsification onto some auxiliary hypothesis. It also illustrates under what circumstances Popper thought such a move might be legitimate. Popper notes that in postulating the as yet undiscovered planet, scientists opened up the possibility of conducting new tests. They could look and see if there actually was a planet at the predicted location. Similarly, if I defend my theory that all adult dogs weigh more than 3kg by blaming the weighing scale, then the scale can then be checked.

What Popper frowned upon was the introduction of such defensive moves that lead to no new tests. So, for example, if defend my theory that all adult dogs weigh at least 3kg from being falsified when Fido is placed on the scale, not by maintaining there scale is faulty, but rather by insisting that a mischievous invisible, intangible gremlin had at that moment tampered with the scale, it's hard to see how my claim might be further tested. Such a defensive move is, to use Popper's terminology, 'ad hoc'.

However, placing an absolute prohibition on the use of ad hoc moves in defence of what otherwise appears to be a falsified scientific theory is surely too strong a requirement. Consider the following example.

When the heliocentric model of the universe was first proposed, it appeared to be straightforwardly falsified. In particular, it was noted that the Earth was moving around the sun rather than vice verse, then the stars should shift back and forth across our celestial field of vision over the course of a year. That's to say, there should be observed 'parallax' (just as there should be if, while looking north, you walk around a lamp post - the houses across the street should move back and forth across your field of view). No parallax was detected. So the heliocentric theory appeared to be falsified. In response, defenders of the heliocentric theory maintained that the stars must just be too far away for the effect to be observable. In fact, this was true - there is parallax, but the effect is so minute it was not observable at that time. So this defensive move was ad hoc. Yet the heliocentric theory was not discarded given the lack of observable parallax, and, in retrospect, it seems it would have been premature to have discarded the theory on that basis. So, placing an absolute prohibition on the use of such ad hoc moves is too strong a requirement. 

Popper on pseudoscience

Let's turn now to Popper's proposed demarcation criterion. Popper suggests that what distinguishes science from pseudo science is the fact that the former is falsifiable, whereas the latter is not.

Popper illustrates his criterion in action using the examples from Marx, Freud, and Adler. Popper considered both Marx’s theory of history and the psychoanalytic theories of Freud and Adler unfalsifiable, but for different reasons.

Popper thought the problem with Freud and Adler’s psychoanalytic theories is that, whatever human behaviour is observed, it can always be interpreted in such a way as 'fit' either theory. Popper, who knew Adler, remarked:

As for Adler, I was much impressed by a personal experience. Once, in 1919, I reported to him a case which to me did not seem particularly Adlerian, but which he found no difficulty in analyzing in terms of his theory of inferiority feelings, although he had not even seen the child 2002, 46). 

Popper believed the same was true of Freud’s theories. They too seemed to 'fit' the evidence, and thus be supported by the evidence, no matter what evidence might turn up. Popper illustrated by considering two hypothetical situations – one in which a man pushes a child into water with the intention of drowning it, the other in which a man sacrifices himself to save a child. Popper claimed each events can easily be explained in either Freudian and 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). (2002, 46)

Indeed, Popper found he couldn’t think of any human behaviour that couldn't be made to fit either theory.

It was precisely this fact—that they always fitted, that they were always confirmed—which in the eyes of their admirers constituted the strongest argument in favor of these theories. It began to dawn on me that this apparent strength was in fact their weakness. (2002, 46-47)

Popper concluded that both Freud's and Adler's theories were unfalsifiable, and for the same reason. Popper believed that Marx's theory of history was also unfalsifiable, though for a different reason. On Popper's view, Marx’s theory - unlike Freud's and Adler's - started out as a falsifiable theory. Indeed, it made some risky predictions about how history would unfold. It predicted a revolution would happen in an industrially advanced society such as Britain. However, Marx's prediction turned out to be incorrect (there was a revolution, but not in the way Marx predicted - e.g. it occurred in industrially backward Russia). Marx’s theory was therefore falsified. However, rather than accept this, Marx’s followers adopted an immunizing strategy, re-interpreting theory and evidence so that the theory continued to fit the evidence after all.

…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 adopting a device which made it irrefutable… [by] this stratagem they destroyed its much advertised claim to scientific status. (2002, 49)

So, Popper frowns on unfalsifiability, considering it sufficient to qualify supposedly 'scientific' theories such as those of Marx, Adler, and Freud as pseudo science. However, he thought that a theory's failure to be falsifiable might be achieved in different ways. Marx's theory started out falsifiable, was falsified, but was then subsequently rendered unfalsifiable by the immunising strategy adopted by its adherents. Freud's and Adler's theories, on the other hand, started out as unfalsifiable - something about the way these theories were initially constructed and/or applied ensured that observation could never count against them.

Popper notes that to those who are wedded to such unfalsifiable belief systems, their truth can appear manifest. Indeed, so obvious is that truth that, the adherents suppose, those who fail to recognise it must be suffering from something akin to a perceptual defect. Popper says:

I found that those of my friends who were admirers of Marx, Freud, and Adler, were impressed by a number of points common to these theories, and especially by their apparent explanatory power. These theories appear to be able to explain practically everything that happened within the fields to which they referred. The study of any of them seemed to have the effect of an intellectual conversion or revelation, open your eyes to a new truth hidden from those not yet initiated. Once your eyes were thus opened you saw confirmed instances everywhere: the world was full of verifications of the theory. Whatever happened always confirmed it. Thus its truth appeared manifest; and unbelievers were clearly people who did not want to see the manifest truth; who refuse to see it, either because it was against their class interest, or because of their repressions which were still 'un-analyzed' and crying aloud for treatment. (2002, 45)

The suggestion that the unbelievers are somehow blinded to the manifest truth crops up in other pseudoscientific thinking. For example, Young Earth Creationist Ken Ham says about those who reject the Bible-literalist account of creation:

Why can’t the humanists, the evolutionists, see that all the evidence supports exactly what the Bible says? It is because they do not want to see it. It is not because the evidence is not there. They refuse to allow the evidence to be correctly interpreted in the light of biblical teaching. (2012, 76)

Criticism of Popper's account

I think there is considerable insight in Popper's thinking about pseudoscience. However, it is flawed. Below are a couple of criticisms.

First, falsificationism appears incorrect as an account of how science progresses.  Here's one example. What really counts in favour of a scientific theory is not necessarily that it is highly testable and has indeed survived rigorous testing - that it makes surprising predictions that have turned out to be true. A theory could be well confirmed - to the extent that we can reasonably suppose that it is true - even if it doesn't predict much.

To illustrate, consider the theory of evolution on which new species evolve over time. While that theory does actually predict a great deal, one key piece of evidence in its favour did not arise from a prediction. Whales are occasionally discovered with vestigial limbs. That's because whales are mammals that evolved from earlier land dwelling creatures. The existence of such limbs was not predicted by the theory of evolution: the theory does not say such limbs are likely. Still, the discovery of vestigial limbs on whales very strongly confirmed the theory of evolution - and also very strongly disconfirmed the Bible-literalist view - because, while such limbs aren't likely on the theory of evolution, they are nevertheless far, far more likely on that theory than they are on a Bible-literalist alternative. Yet on Popper's account, not only can theories never be confirmed to the extent we can reasonably suppose they are true (he accepts Hume's radical skepticism about the unobserved), it's only by way of our deriving predictions from our theories that scientific progress can be made. That last claim is incorrect.

Secondly, Laudan notes that levelling the charge of unfalsifiablity against a doctrine like Young Earth Creationism 'egregiously confuses doctrines with the proponents of those doctrines' (1982, 17). The theory that the universe is just a few thousand years old is, in fact, falsified. The fact that followers of the theory remain unwavering in their commitment to it may reveal something about the psychology of its devotees, but it's a fact about them that stands quite independently of whether or not the theory itself is falsifiable/falsified, which it is.

In short, on Laudan's view, in trying to demarcate pseudoscience, Popper mistakenly focuses on the product (on the theories), whereas the real fault lies with the producers - those who promote and defend those theories.

Actually, even Popper acknowledges that sometimes the fault is not with the theory per se but rather with the manner in which its proponents defend it. As we noted already, Popper thought Marx's theory of history was falsifiable, and was indeed falsified. What turned Marx's theory into pseudoscience, on Popper's view, was the manner in which it was subsequently defended.

So, in trying to demarcate pseudoscience from science, where should our focus be? On the content of the theories, or on the manner in which those theories are defended and/or supported by their proponents? Even Popper acknowledges our focus should sometimes be on the latter.

However, as Maarten Boudry (2013, 91) points out, it's often difficult to tell where a theory ends and the obfuscations of its defenders begin. In the case of Young Earth Creationism, the fault does appear to lie largely with its defenders. However, in other cases - such as Christian Science (see below) - the methodology that renders the theory pseudoscientific appears to be integral to the theory itself. Christian Science just is, in part, a dodgy methodology.

To finish, I want briefly to sketch out a couple of examples in order to identify certain criteria that I think should probably be included - even only if in a family resemblance way - in any adequate explanation of what constitutes pseudoscience. 

Example 1: Young Earth Creationism

Young Earth Creationists are Bible literalists who believe the entire universe is around six thousand years old. They also believe God made all species in the week of creation, and that no new species can or has spontaneously evolve. This theory is, of course, false, and faces a mountain of counter-evidence. This includes, for example:

The fossil record. The reveals that new species have evolved.

The light from distant stars. Given the speed at which light travels, the universe would have to be much older than six thousand years for the light to have reached us.

The white cliffs of Dover. The chalk cliffs are made from the calcium-rich shells of tiny organisms that lived in the sea. It would take much longer than six thousand years for that depth of material to accumulate.

How do Young Earth Creationists typically respond to such evidence? They explain it away. The fossil record, for example, is usually explained as a result of the Biblical flood. The sedimentary layers we see in the ground containing fossils were laid down very recently, mostly during the flood, which also drowned many creatures and trapped their bodies in the layers. The ordering of the fossils within the layers is explained as a result of different ecological zones being flooded at different times. Humans appear in only the top most layers because they managed, through their intelligence, to avoid being drowning until late in the deluge. The light from distant stars has been explained by appeal to a 'time dilation'. The white cliffs of Dover have been explained as a result of Noah's Flood producing vast blooms of microorganisms. Of course, each of these explanations faces its own evidential challenges, and so further explanations can be and are developed ad nauseum to deal with them. 

The general strategy employed above exploits the more general point that any theory, no matter how outlandish can, with sufficient ingenuity, always be made consistent with - be made to 'fit' - the available evidence.

Suppose, for example, that I believe dogs are spies from the planet Venus planning an imminent invasion. There is a mountain of evidence against my belief - evidence that dogs aren't smart, lack language, that Venus is uninhabitable by dogs, etc. However, that evidence can be explained away. I might maintain dogs hide their intelligence and linguistic abilities from us, and that they live on Venus in deep underground bunkers that protect them from the harsh Venusian atmosphere, etc. By endlessly explaining away such evidence, 'fit' can always be achieved.

Now achieving 'fit' in this manner can be misleadingly packaged as doing genuine science. For isn't science all about developing theories that 'fit' the evidence? And hasn't our Creationist shown that their theory can be made to 'fit' the evidence? The emphasis on achieving 'fit' leads proponents of Young Earth Creationism to conclude it is, indeed, scientific.

Increasing numbers of scientists are realizing that when you take the Bible as your basis and build your models of science and history upon it, all the evidence from the living animals and plants, the fossils, and the cultures fits. This confirms that the Bible really is the Word of God and can be trusted totally. (My italics)[i]

According to Ham, Young Earth Creationists and evolutionists do the same thing: they take the evidence, and then look for ways to make it fit the axioms of the framework theory to which they have already committed themselves:

Evolutionists have their own framework …into which they try to fit the data. (My italics)[ii]

This strategy, which I have previously dubbed 'But it Fits!' (Law, 2011), often crops up in pseudoscientific thinking. One of the obvious problems with it, of course, is that it conflates achieving consistency with the evidence with being confirmed by that evidence. Any theory, no matter how absurd - even the theory that dogs are Venusian spies - can be made consistent with the evidence. That's not to say it's confirmed by that evidence.

So, am I suggesting that the systematic adoption of the 'But it fits!' strategy in response to counter evidence is a necessary and/or sufficient condition for a theory, as defended by certain followers, to qualify as pseudoscience?

No. Adoption of the 'But it Fits!' strategy is not sufficient condition. Note that much the same strategy is also employed by other suspect belief systems that are not usually considered pseudoscientific. For example, evidence against conspiracy theories is often explained away by proponents by just expanding the scope of the conspiracy to take care of it (e.g. experts challenging the conspiracy theory may then be thought to be implicated in it). Theological responses to the problem of evil often also employ the 'But it fits!' immunising strategy. Abundant evidence that if the universe has an omnipotent creator, he is not supremely benevolent is ingeniously explained away by theologians by means of an assortment of theodicies that appeal to free will, character-building, and so on. Yet this sort of theology is not considered pseudoscience.

Nor, as we are about to see, is the adoption of the 'But it Fits!' strategy a necessary condition of pseudoscience. Consider Christian Science, outlined below.

Example 2: Christian Science

Christian Science - a religious movement started in 19th Century New England by Mary Baker Eddy - holds that disease is an illusion: a product of the mind. Practitioners believe illness can be healed through prayer using the exact method and means by which Jesus healed.

How do Christian Scientists know their method works? Because of the supposedly 'scientific' evidence they have amassed in its support over many decades. Tens of thousands of testimonies have been published by Christian Scientists of cases in which their methods have been applied and people have subsequently recovered.

Christian Science is pseudoscience but differs from the version of Young Earth Creationism outlined above in that very little attention is given to explaining away the evidence against it.

Instead of focussing on the 'misses' - cases where Christian Science was applied and failed - Christian Scientists focus almost entirely on recording the 'hits' - of cases where the method was applied and the subject recovered. This approach, which I have elsewhere dubbed 'Piling Up The Anecdotes' (Law, 2011) is deemed 'scientific' because, in the minds of Christian Science's followers, a huge amount of data has been built up in support of their theory. 

Christian Science is pseudoscience, and indeed, like Young Earth Creationism, it apes the methods of genuine science. However, it makes little use of the 'But it Fits!' strategy. So, reliance on 'But it fits!' is not a necessary condition of pseudoscience.

Science-like features

Is a heavy reliance on Piling Up The Anecdotes sufficient to qualify any belief as pseudoscientific? I am not sure it is. Many absurd beliefs similarly rely heavily on the use of anecdotal evidence. The widespread belief that some people are psychic, for example, is almost always justified by appeal to anecdotal evidence. So too is the belief that ghosts are real. Yet I would be disinclined to class everyone who believes in ghosts or psychic powers as wedded to pseudoscience.

However, belief in ghosts certainly can become pseudoscientific once its adherents start to use ghosts-detecting devices, start more systematically amassing anecdotal evidence and calling it 'scientific', and start developing a kind of ghost detecting gizmos, a ghost mechanics (involving e.g. ectoplasm, orbs, mysterious 'energy fields', etc.). So, perhaps what is also required for a belief or theory to qualify as pseudoscientific is that it exhibit certain further science-like features. I will return to this suggestion shortly.

Incidentally, I don't claim that defenders of pseudoscience must employ one or both of 'But it Fits!' and/or Piling Up The Anecdotes. These are two strategies that may be employed in defence and/or support of dubious belief systems. However, other unreliable strategies available.

Note, for example, that defenders of pseudoscience often suggest their theory provides the best available explanation of what is observed. Argument to the best explanation is quite properly used in science (we suppose it is reasonable to believe in some unobserved phenomena, such as the Big bang, or electrons, because it provides the best available explanation of what is observed). However, argument to the best explanation is also regularly abused by believers in woo. Those who believe in alien visitation and abduction, miracles, fairies, and even the Resurrection, will often challenge skeptics by saying, 'Explain that!', and then note how their opponents struggle to provide an explanation. They conclude their belief provides the best available explanation of what's observed.

Notice that anything can be neatly and easily explained by positing a hidden agent with extraordinary powers and a desire to bring about what's observed. Can't explain how your keys - which you are sure were on the sofa - ended up on the table? I can: there are mischievous gremlins in your house who enjoy playing such tricks. Can't explain why the flowers grow? I can: fairies imbued with magical powers force the flowers to bloom each spring. Can't explain Jesus's empty tomb, the testimony of eyewitnesses to a risen Christ? I can: God raised Jesus from the dead. Can't explain why the Twin Towers came down vertically like that, with little puffs of smoke appearing just below the collapsing structure? I can: 9/11 was a controlled demolition by secret, Government-backed forces. These aren't sound applications of argument to the best explanation, but they can appear so, particularly in the eyes of believers. We might call this unreliable argumentative strategy 'Explain THAT!'

What these three approaches to justifying beliefs have in common is of course that they are unreliable methods so far as arriving at true beliefs is concerned. 'But it Fits!', Piling Up The Anecdotes, and 'Explain THAT!' are not truth-conducive in the way that genuinely rational approaches to justifying beliefs are. 

Pseudoscience and bullshit

Here is my proposed characterisation of pseudoscience. First, I suggest that proponents of pseudoscientific belief systems exhibit certain unreliable approaches to supporting and/or defending the theory, approaches such as 'But it Fits!' and Piling Up The Anecdotes, approaches that nevertheless appear reliable in the eyes of their proponents. However, as we have seen, it's not just proponents of pseudoscience that do this - 9/11 Truthers and ordinary believers in ghosts and psychic powers do the same thing.

Second, in addition to relying on strategies such as 'But it Fits!', Piling Up The Anecdotes, and/or 'Explain THAT!', pseudoscience also typically involves one or more of the following:

(i) a claimed scientific or science-like methodology that is, in reality, highly suspect (both Young Earth Creationism and Christian Science claim to be using the scientific method, though their actual methods are highly suspect). Proponents of pseudoscience don't just employ unreliable argumentative methods in supporting and/or defending their belief. They maintain they are systematically employing a methodology that is rigorous and perhaps even properly scientific.

(ii) a dubious form of science-like mechanics - positing various forces, entities, fields, stuffs, and/or channels, which can be blocked, enhanced, and so on (Feng Shui and Chinese Medicine do this for example, and so do the Marxist theory of history and the psychoanalytic theories).

I don't claim this characterisation is perfect. You may be able to think of counter-examples. But I suggest it provides a pretty good characterisation of pseudoscience.

My characterisation explains why non-scientific disciplines such as history are not pseudoscience: historians don't (usually) claim to be employing a scientific or science-like method, and what methods they do employ are for the most part, fairly reliable.

It also explains why fraudulent science is not (or is not necessarily) pseudoscience. A scientist who has falsified data to get the result he wants is not practising pseudoscience - his professed methods are genuinely scientific. He has just fiddled the figures. 

My suggested characterisation explains why other dubious belief systems - such as 9/11 conspiracy theories, mainstream belief in ghosts and psychic powers, theodicies - aren't pseudoscience, despite their shared reliance on the use of strategies such as  'But it Fits!', Piling Up The Anecdotes, and 'Explain THAT!'. For these other belief systems don't usually involve any claimed scientific or science-like methodology or invoke any dubious science-like mechanics.

Moreover, in so far as those belief systems do start to involve (i) and (ii), they then do start to look like pseudoscience. Ghost-hunters who call their anecdotal evidence 'scientific', who build ghost-detecting gizmos designed to detect ectoplasm and mysterious 'energy fields' are engaged in pseudoscience.

My account also explains why some forms of climate change denialism are pseudoscience while others are not. Climate change deniers who justify their denial by appeal to a bunch of anecdotes about places that are colder than average are not pseudoscientists, but fools. Climate change deniers that fiddle the data to get the result they want are guilty of science fraud, not pseudoscience. However, climate change deniers who claim to be employing a scientific methodology but who are actually employing a dodgy methodology are engaged in pseudoscience.

Let's call belief systems that make heavy use of strategies such as 'But it Fits!', Piling Up The Anecdotes, and 'Explain THAT!' bullshit belief systems. My suggestion is that pseudoscientific belief systems are a variety of bullshit belief system. What further distinguishes those bullshit belief systems that are pseudoscientific is the fact that they exhibit further additional science-like features. So pseudoscience is a subcategory of bullshit belief systems more generally and indeed gradually shades into other varieties of bullshit.

I am not the first philosopher to link pseudoscience and bullshit. Philosopher Harry Frankfurt (2005) famously defined bullshitters as unconcerned with the truth of what they claim. Bullshitters are not liars, deliberately asserting untruths. Rather, they just don't care whether what they say is true or not. James Ladyman suggests that pseudoscience is a variety of bullshit:

As a first approximation, we may say that pseudoscience is science fraud as bullshit is to lies....This is only a first approximation because we usually assume that bullshitters know what they are doing whereas, as pointed out above, many pseudoscientists are apparently genuinely seeking the truth. Just because one's first-order representations are that one is sincerely seeking truth, it may be argued that, in a deeper sense, one does not care about it because one does not heed to the evidence. A certain amount of self-deception on the part of its advocates explains how pseudoscience is often disconnected from a search for the truth, even though its adherents think otherwise. This is important because it means that what makes an activity connected or disconnected to the truth depends on more than the individual intentions of its practitioners. (2002, 52-53)

Ladyman also suggests that pseudoscience 'involves some kind of emulation of science or some of its characteristics or appearance' (2013, 52).

Clearly, I follow Ladyman in supposing that pseudoscience is a variety of bullshit belief, and also in suggesting that what further distinguishes pseudoscience is further science-like features. However, I do not endorse Ladyman's suggestion that what bullshit and pseudoscience tend to have in common is the fact that adherents collectively, at some deep level, don't care about the truth. 

I don't endorse Frankfurt's characterization of bullshit. I think Frankfurt accurately captures a certain sort of bullshitter - e.g. the person who, unconcerned with whether or not something is true, nevertheless says it for effect, to self-aggrandise, or persuade, or whatever.

However, a great deal of what goes by the title 'bullshit' - in particular, bullshit belief systems - are promoted and defended by folk who are, both individually and collectively, desperately concerned with the truth.

Religious and cult belief systems are often bullshit, but to say about most religious and cult communities that, at some deep level, they don't care whether what they believe is true is quite a stretch. What makes such belief systems bullshit is a heavy reliance on non-truth-conducive methods. It's unnecessary to speculate on the psychology lying behind the use of such methods - to further insist that the folk using such methods do so because, at some deep and collective level, they don't care about truth.

Adherents of bullshit belief systems will sometimes stake both their own lives and the lives of those they love on the truth of their beliefs. Consider those cultists who commit mass suicide (those that drank the Kool-Aid at Jonestown, for example). Consider those Christian Scientists who have killed their sick children by rejecting conventional medicine and relying on prayer instead. Consider antivaxxer parents. To describe such belief communities as being, at any level, unconcerned with whether what they believe is true strikes me as, at best, dubious.

To reply that 'in a sense' such belief communities don't care about the truth because they don't heed to the evidence is to give insufficient weight to the fact that, so far as they are concerned, they do heed to the evidence.

Admittedly, if it were true that such belief communities did not, at some level, care about whether what they believe is true, that would at least partly explain why they fail to employ non-truth-conducive methods. However, there are other explanations on offer for their failure to employ such methods. Other explanations include: (i) such communities are deeply ignorant and/or deluded about what truth-conducive methods actually are, and/or (ii) such communities employ non-truth-conducive methods precisely because they do care about the truth of their beliefs - they want their beliefs to be true, and so end up employing non-truth-conducive methods in order to try justify them.

So, I suggest the case for saying that, at some level, such communities don't care about the truth of what they believe is not well made. We would do better to define bullshit, and also pseudoscience, just in terms of a heavy reliance on methods of justification that are not in fact truth-conducive but can appear so, irrespective of whether or not those employing such methods happen, at any level, and either individually or collectively, to care about whether what they believe is true.


It seems to me that pseudoscience is a legitimate and indeed useful concept. Terms like 'pseudoscience' and 'bullshit' are helpful because they allow us to class together beliefs and belief systems that are distinguished by particularly seductive and common forms of irrationality - forms of irrationality that we need to be on our guard against.

I have here provided a ball-park characterisation of pseudoscience that stresses (i) that pseudoscience is a subcategory of bullshit belief systems more generally, (ii) provided an account of what makes them bullshit belief systems (and rejected Ladyman's and Frankfurt's suggestion), and (iii) provided an account of what distinguishes pseudoscience from other varieties of bullshit.


Boudry, Maarten (2013) 'Loki's Wager and Laudan's Error' in Philosophy of Pseudoscience, M. Pigliucci and M. Boudry (eds.). Chicago: University of Chicago Press. 79-98.

Frankfurt, Harry (2005) On Bullshit. Princeton NJ: Princeton University Press.

Ham, Ken (2102) The Lie: Evolution/Millions of Years. Green Forest AR.: New Leaf Publishing.

Krucoff, M.W. et all 'Music, Imagery, Touch, and Prayer as Adjuncts to Interventional Cardiac Care: The Monitoring and Actualization of Noetic Trainings (MANTRA) II Randomized Study'. Lancet 366. 211-17.

Ladyman, James (2013), 'Toward a Demarcation of Science from Pseudoscience', in M. Pigliucci and M. Boudry (eds.) Philosophy of Pseudoscience. Chicago: University of Chicago Press. 45-59

Laudan, Larry (1983) 'The Demise of The Demarcation Problem' in Physics, Philosophy, and Psycholanalysis. Edited by R.S. Cohen and L. Laudan. Dordrecht: D. Reidel. 111-27

Law, Stephen (2011) Believing Bullshit. Amherst NY: Prometheus.

Pigliucci, Massimo, (2013) 'The Demarcation Problem', in M. Pigliucci and M. Boudry (eds.) Philosophy of Pseudoscience. Chicago: University of Chicago Press. 9-28

Popper, Karl (2002) Conjectures and Refutations 2nd Edition. London: Routledge Classics.

Wittgenstein, Ludwig(1958) Philosophical Investigations. Oxford UK: Blackwell.


[i] Answers in Genesis website

[ii] Answers in Genesis website

Become a patron to

Listen anywhere
Connect via private message