In this section, we shall present evidence that homoeopathy is pseudoscience. To establish this, we need to show, firstly, that (at least some) homoeopaths claim scientific credentials for their doctrine, to wit, that highly diluted homoeopathic remedies can have therapeutic effects above placebo. Secondly, we need to show that when these homoeopaths defend their doctrine, they produce argumentative bullshit.
Regarding the first criterion, we should repeat what we have said in Sect. 3.2: Some homoeopaths do not think of their doctrine as scientific. Their claims should not be classed as pseudoscientific, then. There are, however, homoeopaths who undoubtedly do pretend, explicitly or implicitly, that homoeopathy is a science.
George Vithoulkas, a prominent homoeopath, does so explicitly in the title of his book The Science of Homeopathy (1980) (for another example, see Sankaran, 1988). The same is true for some outlets of studies in homoeopathy, for instance, the International Journal of Homoeopathic Sciences. The German society WissHom (“Wissenschaftliche Gesellschaft für Homöopathie”) also has the attribute “wissenschaftlich” (“scientific”) in its name, and the homoeopathy-promoting charity Homeopathy UK also wants us to believe in the scientific status of the doctrine. To this end, it produces a paper by her majesty’s late personal homoeopath, Peter Fisher, that discusses the supposed scientific evidence favouring the creed (Fisher, 2021).
Homoeopaths also claim scientific status implicitly. To do so, they mimic, for instance, conventions and organs typical of science. They show their academic credentials (titles such as MD, PhD, etc.), organise conferences, run, as we have mentioned, their own societies and research institutions, and publish books and papers with scientific publishers (the journal Homoeopathy, which is the official outlet of the UK based “Faculty of Homoeopathy”, is published by the scientific publisher Thieme). Also, some influential homoeopaths hold official posts in universities (for instance, Michael Frass and Harald Walach; August Bier had a chair of homeopathy in Berlin in the 1930s).
So, the first criterion for pseudoscience is clearly met in the case of homoeopathy.Footnote 20 In the remainder of this section, we shall provide evidence that the second criterion is also fulfilled. The argumentative moves of homoeopaths frequently contain egregious violations of important epistemic standards. However, before we turn our attention to the many examples we have gathered, let us briefly mention a few arguments we explicitly do not endorse.
How not to argue against homoeopathy
There are some forms of reasoning one should avoid when arguing that homoeopathy is pseudoscience.Footnote 21 For instance, it is often said that the origins of the doctrine date back to a time when basic medical facts, such as the germ theory of disease, had not been established. However, as Thagard (1978) points out in a discussion of astrology, the provenance of a doctrine is not, by itself, a reason to reject it as pseudoscience. Astrology, he says, “cannot be condemned simply for the magical origins of its principles” (Thagard, 1978, p. 225). The same courtesy should be extended to homoeopathy.
Similarly, the psychology of those who believe in a doctrine is, by itself, no reason to disqualify it. To be sure, belief in the theories of complementary and alternative medicine (CAM), including homoeopathy, seems to be driven, to a large extent, by an intuitive thinking style (rather than an analytical one), paranormal beliefs, and fundamental confusions about physical, biological, and mental phenomena (Browne et al., 2015; Lindeman, 2011). Nevertheless, CAM theories, including homoeopathy, may be accurate and genuinely scientific. What causes belief in a theory in laypersons may be suggestive. However, it is ultimately irrelevant to assessing its scientific credentials.
This has an interesting corollary, namely, that several well-known bullshit arguments homoeopaths frequently employ to promote their services should, in fact, not be cited as direct evidence that homoeopathy is pseudoscientific. For instance, homoeopaths often use scientifically irrelevant considerations, such as appeals to the celebrity status of their patients.Footnote 22 They also employ appeals to authority, tradition, and popularity (Ernst, 2020),Footnote 23 and they even engage the conspiratorial tendencies of would-be users.Footnote 24 These may be flawed arguments and manipulation tactics, problematic from an ethical perspective. However, we are presently not concerned with an issue in ethics but with a problem in the philosophy of science, namely, what makes homoeopathy pseudoscience. And to that issue, whether homoeopaths use bullshit for advertising their doctrine to laypeople is irrelevant as long as they do not use their flawed arguments to defend its claim to scientificity.Footnote 25
Note, however, that homoeopaths do sometimes use such arguments in professedly scientific discussions. Lionel Milgrom, a frequent contributor to the homoeopathic literature, does precisely this when he writes in a journal paper that behind the critics of homoeopathy, “like some eminence gris [sic], is the financial reach of the globalized pharmaceutical industry”Footnote 26 (Milgrom, 2008a, 590; emphasis in the original). This is bullshit, and it is used in a discussion to respond to those who criticise homoeopathy. Accordingly, Milgrom’s comment is, indeed, an indication that homoeopathy is pseudoscience.Footnote 27
Let us make another surprising statement: The fact that homoeopaths hold pseudoscientific beliefs does not conclusively prove that homoeopathy is pseudoscience. This is because these pseudoscientific beliefs may be logically unrelated to homoeopathy. Consider an analogy: Isaac Newton espoused alchemy. But this does not mean that modern physics is pseudoscientific. The case of homoeopathy is similar. Some homoeopaths, for instance, may believe in dowsing as a means for finding water, which is pseudoscientific. But this does not establish, in and of itself, that homoeopathy is also pseudoscientific. However, if homoeopaths advocate dowsing as a means for finding the correct homoeopathic remedy, this pseudoscientific belief becomes relevant to the assessment of homoeopathy as pseudoscience.
Finally, it is, of course, important not to cherry-pick examples. The fact that some homoeopaths make pseudoscientific assertions does not establish that their doctrine is a pseudoscience, even if they do use these assertions to support that doctrine. What matters is whether what they say is sufficiently representative of their community. Of course, since homoeopathy comes, as discussed in Sect. 2.2, in various forms, we face the apparent problem that the views of homoeopaths differ. Accordingly, it seems difficult to show that all homoeopaths systematically violate important epistemic standards. Recall, however, what we have said in Sect. 2.6, namely, that all homoeopaths face two problems: They have to address, firstly, the implausibility of the claim that homoeopathic remedies diluted beyond Avogadro’s limit can work and, secondly, the lack of sufficient acceptable empirical evidence in its favour. Homoeopaths, hence, face a dilemma. Either they bury their head in the sand and ignore these points, or they address them. Ignoring them is, evidently, epistemically careless. If they cannot provide a satisfactory answer to these problems, the only epistemically responsible move is to abandon their belief in homoeopathy. Therefore, all we have to make plausible is that when homoeopaths address these points, they systematically produce egregious violations of important epistemic standards. This, we believe, can be done through examples from influential homoeopaths, official institutional bodies, and publications that homoeopaths themselves recognise as valid contributors to their field.
That said, note that if our examples were, in fact, cherry-picked, this could be easily demonstrated. To this end, one would only have to cite reasonable arguments from homoeopaths that resolve the implausibility problem and provide sufficient acceptable evidence for homoeopathy.
The implausibility of homoeopathy
As discussed above, homoeopathy posits that substances diluted beyond Avogadro’s limit can have therapeutic effects above placebo. This proposition suggests that a non-existent substance can cause the body to heal itself, which contradicts basic natural science and is, hence, wildly implausible. Accordingly, homoeopaths are in the same boat as, for instance, parapsychologists, who also make claims that are hard to square with natural science (Goode, 2013).Footnote 28 When pressed to justify this lack of plausibility, homoeopaths have, as far as we can see, chosen two basic strategies.
The first strategy is to leave the naturalistic framework and make bizarre ontological claims. An example can be found in Applying Bach Flower Therapy to the Healing Profession of Homoeopathy (1993/2005)—a book by Cornelia Richardson-Boedler, who has served as the director of Bach Flower Studies of the British Institute of Homoeopathy. She writes that highly diluted homoeopathic remedies
lose their physical properties after the 12c or 24x potency, or after Avogadro’s number. In this way, the released and highly activated simple substance of the remedy is able to resonate with the highest realms of man’s simple substance or inmost identity. Nonetheless, the higher potencies act powerfully on tissues as well, just as the human soul animates the totality of being. (Richardson-Boedler, 1993/2005, p. 19; emphases added)
What Richardson-Boedler says in the passage may be in line with the thinking of Hahnemann, who also spoke of “spirit-like vital forces” animating the body. However, the suggestion that material substances suddenly lose their physical properties is incongruent with our best knowledge of the physical world that comes from the natural sciences. As such, it is an egregious violation of an important epistemic standard, namely, epistemic connectedness with other fields of knowledge (Hoyningen-Huene, 2013).
Another example comes from George Vithoulkas’ The Science of Homeopathy (1980):
It appears that some form of energy is released by this technique [i.e. the homoeopathic preparation of remedies]. The energy which is contained in a limited form in the original substance is somehow released and transmitted to the molecules of the solvent. Once the original substance is no longer present, the remaining energy in the solvent can be continually enhanced ad infinitum. The solvent molecules have taken on the dynamic energy of the original substance. (Vithoulkas, 1980, p. 104; emphasis in the original)
It is unclear which kind of energy Vithoulkas means since he does not specify it. Perhaps, he does not even know it himself. In that case, his talk of dynamic energy transmission and molecules only serves to “language it up” (Dawkins, 2003, p. 6) and dress up his ignorance in scientific lingo.Footnote 29 If he is referring to a specific kind of energy, his assertion is empirically unsupported because there is no evidence of energy transmission in the preparation of homoeopathic remedies (Ernst, 2016). Most likely, he is thinking of a new form of energy that is not part of present-day physics. In that case, he, like Richardson-Boedler and other homoeopaths, is making an ontological claim that is hard to square with a naturalistic outlook.
Some homoeopaths pursue a second strategy. They seek to show that homoeopathy is compatible with natural science, after all. This shtick is well-known from other pseudoscientific realms. Biblical creationism, for instance, was repeatedly thrown out by courts as an alternative theory to biological evolution. So, its proponents refashioned it as “intelligent design theory” and tried to pass it off as a genuinely scientific theory.Footnote 30 In the case of homoeopathy, two different approaches have been pursued.
One approach is to appeal to quantum woo-woo.Footnote 31 An example of this is found in Milgrom (2002, 2007).Footnote 32 Following Kent’s suggestion “that a medicine is only homeopathic when the patient and the practitioner are included,” he proposes “to use quantum mechanics terminology” and think of the two as “entangled” (Milgrom, 2002, p. 243). Of course, no quantum theorist would be able to make sense of such a suggestion—not least because, in physics, the notion of entanglement applies at the level of particles, not people. So, Milgrom clarifies that he uses weak quantum theory, which “explicitly allows its application beyond the narrow confines of particle physics” (Milgrom, 2002, p. 243). He also says that he intends the entanglement relation as a metaphor. It is not clear what this is supposed to accomplish. A metaphor may, of course, help us to envision how a process might work (Hofstadter & Sander, 2013). But, of course, it does not provide any evidence that it actually exists. Suffice it to say, then, that physicists whose work has been implicated in the writings of homoeopaths have distanced themselves from the doctrine.Footnote 33
Another approach is to show that the diluent used to prepare homoeopathic remedies somehow “remembers” the substances with which it has come into touch. One way to establish this would be to dissolve a substance in water, dilute it beyond Avogadro’s limit, and bring it into contact with a biological system to see how that system reacts. If the solution still has an effect characteristic of the diluted substance, this suggests the existence of water memory.
A research team around the esteemed immunologist Jacques Benveniste famously conducted this kind of experiment and published it in the venerated journal Nature (Davenas et al., 1988). They used IgE antibodies to prepare water solutions. In them, antibodies were diluted until no active biomolecules were present anymore. Then, they applied the solution to basophils, a type of white blood cell that can be activated by IgE antibodies. The basophils reportedly showed an immune response when they got in touch with the highly diluted solution, suggesting that the water “remembered” the antibodies.
The result, hailed by homoeopaths as proof of the mechanism behind homoeopathy, was met with scepticism from the start. As Nature’s editor John Maddox commented, there is “no evidence of any other kind to suggest that such behaviour may be within the bounds of possibility” (Maddox, 1988, p. 787). Subsequently, Benveniste’s team failed to reproduce the experiments under blinded conditions (Maddox et al., 1988), and other teams were not able to reproduce them consistently either (Ball, 2004).
Nevertheless, some homoeopaths still treat Benveniste’s work as suggestive of a water memory (see, for instance, Thomas, 2007), which is epistemically irresponsible. Others take the idea to new extremes. In his book The New Physics of Homeopathy (2002), homoeopath Colin B. Lessell suggests, for instance, that individual water molecules may have a memory. This idea is so far out that even other homoeopaths find it unpalatable. As Lionel Milgrom complains, Lessell makes this suggestion “without bothering to offer any sensible explanation, within the known laws of chemistry and physics, as to what that memory consists of” (Milgrom, 2003, p. 62). Bearing in mind that this comes from someone who thinks of practitioner, patient, and remedy as “quantum entangled,” this has to count for something.
We can record, then, that homoeopaths believe in a doctrine that is wildly implausible given our best knowledge of the natural world. This is, in itself, an egregious violation of a crucial epistemic standard. Moreover, when pressed to justify this, they respond with argumentative moves that constitute further violations. They draw on bizarre ontological ideas or quote debunked experimental results.
Shifting the burden of proof
The second problem for homoeopaths is that their doctrine is, to date, not backed by sufficient acceptable evidence. As we have explained in Sects. 2.4 and 2.5, the intellectually honest way of dealing with this problem is to conduct high-quality RCTs. However, when confronted, homoeopaths regularly resort to illicit argumentative tactics. One such tactic is to shift the burden of proof to one’s critics.Footnote 34 When this is done without a good reason, it is an illegitimate argumentative move characteristic of pseudoscience more generally (Pigliucci & Boudry, 2014). Homoeopathy shares this feature with other pseudosciences, such as intelligent design creationism (Pigliucci, 2010/2018) and ufology (Oberg, 1979).
There are various ways to shift the burden of proof. Some homoeopaths simply appeal to rhetorical phrases, which is evidently ludicrous and shall, hence, not be discussed here.Footnote 35 Others pursue an approach that is superficially more plausible. They rely on the principle that the absence of evidence is not evidence of absence.Footnote 36 In other words, just because we have not found evidence for homoeopathy, we have not found evidence against it.
Arguments that appeal to this principle seem initially reasonable because they merely appear to reject a fallacious way of reasoning. To see this, consider the argument from ignorance, which is widely viewed as a misstep in thinking.Footnote 37 It alleges that since we do not know that p, we know that ¬p. Now, if we replace “know” with “have evidence”, we get the negation of the above principle: Since we do not have evidence that p, we have evidence that ¬p. This seems equally fallacious, and adherents of homoeopathy seem quite right to reject it.
Indeed, arguments that rely on an absence of evidence are often unsound. The absence of evidence for a proposition, say, “that a storm is not brewing in the atmosphere of Jupiter,” is, generally speaking, not evidence for its negation, to wit, “that a storm is brewing” (Kelley, 1988/2013, p. 130; emphasis in the original). However, consider the following example, which shows that this is not generally so:
A man is sitting inside a warehouse that has a tin roof and no windows. Tin roofs are notorious for making lots of noise inside a building when it rains outside. The man in the warehouse cannot see outside, so he could not tell directly if it were raining at a given time. But he could infer indirectly, using, for example, the following argument: if it were raining now I would know it (by the noise); but I do not know it; therefore, it is not raining now. (Walton, 1996, p. 1)
The latter argument seems unobjectionable. Hence, the question arises when appeals to an absence of evidence legitimately shift the burden of proof.
Evidently, the difference between the two cases is this: If a storm were (or were not) brewing in the atmosphere of Jupiter, we would not expect to have any evidence of it because we are not looking for evidence. Accordingly, we should suspend judgement as to whether a storm is, in fact, brewing in the atmosphere of Jupiter. In the case of the man in the tin warehouse, this is different. If it were raining, he should expect to have evidence of this. He knows that whenever it rained in the past, he would hear the sound of raindrops on the roof. To be sure, he cannot be entirely certain that, this time, the rain would also produce the same sound. After all, someone may have, unbeknownst to him, installed a giant fan on the roof that blows away the raindrops before they hit the tin surface. However, the chance of that is remote. Accordingly, for the man in the warehouse, the absence of evidence for rain is, indeed, evidence of the absence of rain.Footnote 38
We should ask, then, whether the case of homoeopathy is more like the Jupiter case or more like the rain case. The answer is simple: It is much more like the latter. Homoeopathy does not fit the protoscience category, which we have discussed in Sect. 3.3. This is because the methodology of evidence-based medicine is an effective tool for picking up evidence that a remedy works if it works. After 200 years of research, the chance that we would not have found sufficient acceptable evidence for homoeopathy is tiny if it were true. In this case, the absence of evidence is also evidence of absence. So, the burden of proof lies squarely on advocates of homoeopathy and rejecting it is evidence of epistemic carelessness.
Mischaracterising the evidence
When homoeopaths are confronted with the problem of the insufficient empirical foundation of their doctrine, they often mischaracterise the evidence. Dey et al. (2021), for instance, do this in a recent paper. They examined in an RCT whether classical homoeopathy was effective in treating warts and determined that their study was “inconclusive.”
At first glance, this sounds innocent. However, it is a manipulative choice of words because the category “inconclusive” does not exist in statistical analysis. In an RCT, recall, the investigation aims to determine whether a remedy is effective or not. To this end, experimental subjects are divided randomly into two groups. Those in the verum group receive the remedy. Those in the control group get a pharmacologically inert placebo instead. Researchers then collect data and analyse whether there is a difference between the two groups, that is, a therapeutic effect. Crucially though, it is not enough to find that subjects in the verum group have improved more than subjects in the control group since such a difference can arise by chance. The difference between the groups must be statistically significant, as explained in Sect. 2.4. That is, the probability of a difference that large (or larger) being due to chance has to be 0.05 or less. Only then is the trial counted as a confirmation of the hypothesis that the remedy works. If the difference between the two groups is not statistically significant, the hypothesis counts as disconfirmed (or the null-hypothesis counts as accepted). To be sure, the convention for statistical significance is somewhat arbitrary (Ziliak and McCloskey, 2008). But it does ensure some degree of comparability across trials. Therefore, it seems dishonest not to acknowledge what, by convention, is a disconfirming finding.Footnote 39
Homoeopaths also mischaracterise other researchers’ findings. The British Homeopathy Association (BHA), a UK-based charity devoted to promoting homoeopathy, summarises the results of 104 peer-reviewed journal papers with RCTs as follows:
41% of these RCTs have reported a balance of positive evidence, 5% a balance of negative evidence, and 54% have not been conclusively positive or negative. (Homeopathy UK, 2021b)
Here, the BHA uses essentially the same trick. In statistical analysis, the category “inconclusive” does not exist. Therefore, the only adequate description of the evidence is that 41% of RCTs were positive, and 59% were negative.Footnote 40
Cherry-picking is the failure to consider all available and relevant evidence on a given issue. At its most extreme, one picks out a single case report or a small number of cases and draws substantive conclusions. This is problematic for at least three reasons. Firstly, there is no guarantee that all the relevant data was gathered and recorded without error.Footnote 41 Secondly, even if the data were recorded reliably, individual cases do not allow us to disentangle multiple possible factors that might explain the result. It is well known that many factors can explain the observation that the patient gets better after administering a remedy. Among them are, for instance, the placebo effect, the natural course of the condition, other drugs that the patient may have received, and so on. Thirdly, there is no guarantee that the cases are representative. This is why case reports are at the low end of the hierarchy of evidence in evidence-based medicine (Nissen & Wynn, 2012).
Cherry-picking the data is a common problem in pseudoscience (Boudry, 2013; Hansson, 2017; Shermer, 2013). It is also common in homoeopathy. Prominent homoeopaths have advocated drawing far-reaching conclusions from individual cases in the research literature.Footnote 42 George Vithoulkas, for instance, argues that homoeopathy journals should invite practitioners to publish more case reports. This way, he thinks, “a huge body of important evidence could be amassed of what homeopathy can or cannot do” (Vithoulkas, 2017a, p. 198).
Some practitioners follow Vithoulkas’s advice and go very far in their conclusions. For instance, Wadhwani (2015) as well as Choudhury and Khuda-Bukhsh (2020) conclude, in two separate cases, that homoeopathy cured a patient of deep vein thrombosis (DVT), and Yaseen (2020a) claims, in one case, to have cured a patient, “gently and softly,” of acute lymphoblastic leukaemia and of primary pure red cell aplasia in association with Johnson-blizzard syndrome in another (Yaseen, 2020b).
The strategic selection of convenient cases is not the only form of cherry-picking homoeopaths frequently use. When reviewing existing studies and, particularly, systematic reviews and meta-analyses, they frequently cherry-pick the statements made in them. One example of this is the review paper of Weiermayer et al. (2020). The authors discuss the evidence for the homoeopathic treatment of infections in humans and animals and select six systematic reviews they deem relevant (Cucherat et al., 2000; Kleijnen et al., 1991; Linde et al., 1997, 1999; Mathie et al., 2014; Shang et al., 2005). Not only do they claim that five of them confirmed the effects of homoeopathy, which is false (Ernst, 2015). They also ignore many eminently relevant studies (for instance, Antonelli & Donelli, 2019; Doehring & Sundrum, 2016; Hawke et al., 2018; Qutubuddin et al., 2019; Reisman et al., 2019).
So far, we have argued, in Sects. 4.4 and 4.5, that homoeopaths mischaracterise and cherry-pick the available evidence. Moreover, they do this against the background of an already distorted evidence base because they also selectively report their findings, as Gartlehner et al. (2022) found in a recent cross-sectional study and meta-analysis. The authors systematically investigated the extent of reporting bias in trials on homoeopathy. Before we summarise their findings, we need some background.
The Declaration of Helsinki is one of the cornerstone documents regulating human subjects’ experimentation. In 2008, it was amended to include an obligation on the part of researchers to preregister and publish all their trials. The rationale for this new requirement is straightforward: If we run enough trials with an ineffective remedy, we will inevitably stumble upon significant findings because these are to be expected in 5% of cases, as we have discussed in Sect. 2.5. If, in addition to this, we change the end-points of our studies after the results are in (HARKing), we increase the probability of getting positive trial outcomes further. Accordingly, to assess the evidence, it is not only necessary to know how many studies with confirmatory findings exist. It is essential to know, also, how many trials have been performed in total and what endpoints and hypotheses they sought to investigate. The requirement to preregister and publish all trials is to ensure these conditions are met.
The study by Gartlehner et al. reveals that homoeopaths poorly adhere to the preregistration and publication requirements. They found that almost 38% of homoeopathy trials remained unpublished, and 50% of published trials were not registered. In addition, 25% of the primary endpoints were changed, as a comparison with the preregistered research protocols revealed—suggesting frequent HARKing. The researchers hypothesise that, due to the lack of a preregistration requirement for homoeopathy trials, many more unregistered trials likely exist. So, the authors’ results are likely an underestimation. In conclusion, Gartlehner and colleagues state that the lack of preregistration and reliable publication of trials “likely affects the validity of the body of evidence of homeopathic literature and may substantially overestimate the true treatment effect of homeopathic remedies.”
Rejecting scientific methodology
Pseudoscientists tend to reject essential elements of established scientific methodology. Creationists, for instance, reject radiometric dating (Kitcher, 1982). Similarly, homoeopaths tend to dismiss evidence from RCTs (Mathie et al., 2014).
Ricotti and Delanty (2006) write, for instance, that “[i]ndividualized therapies such as homeopathy and reiki cannot be compared with medicines in a conventional pharmaceutical model,” and Vithoulkas (2017a, p. 197) calls RCTs of homoeopathy “a waste of time, money, and energy.” Milgrom argues “that no therapeutic modality, conventional medicine included, is ever practiced in real life according to the DBRCT’s [i.e. the double-blind randomised controlled trial’s] procedural separation of therapy and context” (Milgrom, 2008a, p. 591). This is, of course, precisely the point of these experiments. What researchers want to find out is whether the administered substance had any effect of its own, and this can only be done if other factors that could conceivably affect the patients’ outcomes—importantly: the placebo effect—are rigorously controlled. To reject this aspect of established scientific practice in evidence-based medicine is an egregious violation of epistemic standards. This gets even more obvious once we consider the alternative methodologies homoeopaths propose to have their theories “tested,” as we shall see next.
Pseudoscientists often use immunisation strategies to protect their doctrines from recalcitrant evidence (Boudry & Braeckman, 2011). George Vithoulkas proposes to build them right into the homoeopathic research methodology. In his view, homoeopaths should accept research only if it abides by the following principles:
Homeopathy does not treat diseases but only diseased individuals. Therefore, every case may need a different remedy although the individuals may be suffering from the same pathology. …
In the homeopathic treatment of serious chronic pathology, if the remedy is correct usually a strong initial aggravation takes place. Such an aggravation may last from a few hours to a few weeks and even then we may have a syndrome-shift and not the therapeutic results expected. If the measurements take place in the aggravation period, the outcome will be classified negative. … At least sufficient time should be given in the design of the trial, in order to account for the aggravation period. …
In severe chronic conditions, the homeopath may need to correctly prescribe a series of remedies before the improvement is apparent. Such a second or third prescription should take place only after evaluating the effects of the previous remedies. …
As the prognosis of a chronic condition and the length of time after which any amelioration set in may differ from one to another case, the treatment and the study-design respectively should take into consideration the length of time the disease was active and also the severity of the case. (Vithoulkas, 2017b, p. 48; emphases added)
To be sure, it is possible, as principle 1 requires, to investigate the efficacy of individualised homoeopathic treatments rigorously by examining how the individually prescribed homoeopathic remedies perform against placebo. However, the other principles, in effect, make homoeopathy immune to recalcitrant evidence.
Principles 2 and 3 ensure, in conjunction, that every empirical observation is compatible with the hypothesis that homoeopathy had an effect. Frank Cioffi, in his discussion of the pseudoscientific nature of psychoanalysis, explains the mechanism at work. It is “characteristic of a pseudoscience,” he writes,
that the hypotheses which comprise it stand in an asymmetrical relation to the expectations they generate, being permitted to guide them and be vindicated by their fulfilment but not to be discredited by their disappointment. (Cioffi, 1998, p. 118)
If homoeopathy were effective, we would expect patients to improve more than in the placebo group. Presumably, Vithoulkas would also view the fulfilment of this expectation as a vindication. At the same time, however, principle 2 allows him to avoid admitting defeat if the opposite were to be observed. If patients got worse compared to placebo, this would also be good news for homoeopathy. After all, this could be reinterpreted as an “aggravation,” which, on Vithoulkas’s principle 2, also proves that homoeopathy works.
Finally, if no improvement is found, the condition is apparently chronic, and principle 3 applies: The homoeopath has to prescribe other medications. So, the methodology Vithoulkas proposes in effect immunises homoeopathy against empirical criticisms. He hedges his bets by preparing various possible ad hoc hypotheses that he can draw upon to explain why the prescribed medicine did not make the patient better. This makes homoeopathy unfalsifiable.Footnote 43
Principle 3 has a further interesting consequence. Vithoulkas writes that “a second or third prescription should take place only after evaluating the effects of the previous remedies.” That means, of course, that the homoeopath has to know which remedy was prescribed before in order to evaluate its effects. This prevents adequate blinding and increases the risk of bias.
Finally, consider principle 4. It says that “the treatment and the study-design respectively should take into consideration the length of time the disease was active and also the severity of the case.” In and of itself, this principle may seem innocuous since it is, of course, correct that the history and severity of the disease should be taken into account when predictions are made about its future trajectory—be it with or without medication. Nevertheless, it is an essential principle of scientific research, as discussed in Sect. 2.5, to formulate a hypothesis before the results are in. The opposite, hypothesising after the results are known or HARKing, can, as discussed in Sects. 2.5 and 4.6, massively increase the chance of finding a positive result. Vithoulkas’s principle 4 is problematic as it may be interpreted as an invitation to do just this.
The latter seems especially likely in the case of homoeopathy. As discussed in Sect. 2.1 above, homoeopaths propose to identify the correct remedy based on the law of similars, that is, based on the principle that a substance known to cause particular symptoms in a healthy individual can be used to cure these symptoms in a sick individual. In other words, homoeopaths are not interested in the causal mechanisms behind a disease. To them, the disease is, essentially, a black box. This being so, they are at liberty to speculate about its future course and are unbound by the theories of science-based medicine that tell us how the body works.
We have argued that homoeopathy is pseudoscience. Our argument contained two parts: firstly, a theoretical criterion for categorising a pursuit as pseudoscience, which we introduced in Sect. 3, and, secondly, empirical evidence suggesting that homoeopathy matches that criterion, which we have provided in this section. Accordingly, one can object to our argument in two ways, namely, by rejecting our theoretical criterion or the empirical evidence we have presented.
As for the first part, we have followed other authors (Ladyman, 2013; Moberger, 2020; Mukerji & Mannino, 2022) in construing pseudoscience as a form of bullshit in conjunction with a symptomatic approach for detecting it (Boudry, 2021). This relatively new approach may, of course, be challenged, and we cannot defend it here in detail. We believe, however, that it wears its plausibility on its sleeves as it can theoretically unite many alternative criteria other authors have proposed to demarcate science from pseudoscience. Fasce (2017) has reviewed 21 demarcation criteria and found that they contained 70 individual factors. Given the limited space, we cannot discuss them all. However, a few examples should suffice to show how criticisms based on other criteria can be recast in terms of our bullshit criterion.
Fasce’s top-ranked criterion is external incongruity. It corresponds to the implausibility problem discussed in sections 2.6 and 4.2. As we have argued, homoeopaths egregiously violate important epistemic standards by advocating a doctrine that is incongruent with our best theories of the natural world. This is undoubtedly an indication of bullshit unless it is defended with very persuasive evidence.
The second-ranked criterion is deficient methodology. It is also covered by our approach. As shown in sections 4.4 through 4.8, homoeopaths mischaracterise and mishandle the evidence, reject RCTs, and immunise their theories against recalcitrant findings. Their methodology, in other words, is highly deficient, making their defence of homoeopathy bullshit.
The third-ranked criterion is lack of progress. Homoeopaths would, of course, insist that their field does progress, as is evidenced by the constant discovery of new homoeopathic remedies. However, this is not what the criterion means. For a research field to count as progressive, it must make bold new predictions that are actually confirmed by the data. As we have seen, however, homoeopathy does not do that. Instead, its proponents constantly try to explain away its implausibility and recalcitrant evidence by resorting to grotesque and epistemically irresponsible arguments. This is the mark of a degenerating research programme that does not exhibit progress (Lakatos, 1978).
As for the second part, we have shown, using examples from the literature on homoeopathy, that when homoeopaths are confronted with the two central problems of their doctrine, namely, its scientific implausibility and its lack of sufficient acceptable evidence, they produce bullshit in response. They make bizarre ontological claims or resort to quantum woo woo. They also illegitimately shift the burden of proof, mischaracterise, cherry-pick, and misreport the evidence, reject important parts of the scientific method, and seek to immunise their doctrine against recalcitrant evidence. Now, critics may object that we may have cherry-picked the evidence ourselves. Perhaps, they may argue, we have presented a few outliers that are not representative of the academic literature on homoeopathy as a whole? Let us make two brief points in response.
Firstly, the examples we used come from publications, persons, and organisations that are well-established within homoeopathic circles. Secondly, as discussed at the end of Sect. 4.1, it is clear how our empirical case could be disproven if it were incorrect. To do this, one would need to show that the implausibility problem of homoeopathy can be reasonably resolved and that sufficient acceptable evidence for it can be provided. This would be an easy feat if we had, indeed, cherry-picked the evidence and sidestepped aspects of the literature that did not support our position.