The “Cum hoc ergo propter hoc” fallacy of homeopathic drug testing
“Cum hoc, ergo propter hoc” means “with this, therefore because of this” and, as a fallacy, is to be understood in such a way that when two events coincide in time, a causal connection between the events is also assumed – unjustifiably. Sounds complicated, but it is not. Causality is the assumed lawful connection between two successive events, one of which, earlier in time, is called the cause and the other, later in time, the effect. According to the principle of causality, there can be no effect without a cause.
However, there is also the possibility that a temporal coincidence of events is erroneously regarded as a cause-effect phenomenon – whereby this error could have been detected by a more precise (or any) examination of the facts.
Example: We justifiably conclude that by means of a drug, e.g. an antibiotically active substance that inhibits bacterial cell wall synthesis, certain disease-causing microorganisms are killed or prevented from multiplying and thus the disease triggered by the microorganism can be cured. However, it is not justified to assume, for example, that the full moon could have something to do with sleep disorders, because firstly, people with permanent sleep problems also have them during other moon phases, and secondly, even people who suffer from sleep disorders from time to time have spent a considerable number of full moon nights blissfully slumbering. So here, from the multitude of reasons for sleep disorders, the one that is used as an explanation is the one for which only a temporal, but no causal connection exists. (We know that this is the case from a large number of scientific studies).
So we see: one conclusion is justified, the other is not. Both are temporally connected to each other, only in the case of the first one is the causal connection also proven. What is important here, then, is the insight that conclusions we draw from events that closely follow each other in time can be correct – or not – because the events are causally connected to each other – or not.
The root of the evil – is us humans
The real problem now is that we are quite unimpressed by the many false conclusions we draw. On the contrary, we often insist that there is no legitimate doubt about the causalities we supposedly recognise – no matter how absurd they may be on closer examination. Why is that? I’ll try a casual explanation: it’s because of mothers-in-law.
We are causality junkies; the search for meaning and context is, as they say, in our genes.
We have the potential mothers-in-law of young hopeful hominids at the beginning of human history to thank for that. The mothers of the many Lucys, who hundreds of thousands of years ago were searching the East African savannah for the father of their offspring, advised their daughters: “Be clever, take the young one! He noticed that when black and yellow stripes appeared between the grasses, someone was missing from the herd a short time later. Now he always climbs up the nearest tree when he sees black and yellow stripes. Obviously it makes sense …
The fact is that black-and-yellow striped fur colouration did indeed appear in combination with elongated fangs and the desire to make a main meal out of our ancestor. But the concept of “camouflaged fur colouring” is not necessarily associated with sabre-toothed tigers, but also with gazelle or meerkats (or their ancestors), which ultimately led to Lucy’s dream man being up in the tree more often than was necessary. But this had far fewer consequences than being in the tree once too often, when the right stripers were sneaking through the tall grass. So the ability to recognise patterns was a clear evolutionary advantage. Those who could do this had a much better chance of spreading their genes. The failures that occurred in the process, on the other hand, did not matter because they did not worsen the chances of our hominid.
Our young, aspiring ancestor may have cursed sometimes when he was up a tree because of a gazelle, but he did not have a concept against his mistakes – which, unfortunately, did not change fundamentally in the next hundred thousand years. Neither did the conviction that it is always better to insist on one’s supposedly safe “experiences” change. Based on this, mankind has plenty of time to invent false causalities and justifications for practically every natural phenomenon and to develop superstitions and incantations of all kinds. Every rain dance (or whatever other everyday magic is celebrated) is based on an unjustifiably assumed causality, also called bogus causality: The rain dance is not the cause of a subsequent heavy shower, because it would have rained at some point anyway, sometimes a little earlier, sometimes a little later. If it happened sooner, the illusory correlation was almost obvious …
Now, despite the persuasive power that false pattern recognition, i.e. spurious correlation, also brings with it, there have always been people who have seen, or at least suspected, its falsity because the alleged causalities did not survive the test with a harsh reality. Every person who snores peacefully on a night of a full moon inevitably leads, if one follows logical reasoning, to doubt the assertion that the fully illuminated earth satellite can disturb the night’s rest – and can also be held responsible for all kinds of other mischief. Stupidly, humanity is often astonishingly stupid in such matters, these people, who insisted that the mainstream world view of the respective epoch contained a lot of nonsense, were disposed of in a way that was mostly unpleasant for these people. This led to peace in the village again for a while, but in the long run this strategy did not work.
At some point in antiquity, the ancient Greeks, above all Heraclitus and Democritus, finally put a bit of order into the confusion surrounding knowledge and the causal principle on paper, so to speak: Nothing happens without a reason! About two minutes later, as was to be expected, the arguments began: Which cause is responsible for which effect? These disputes now mark the next 2,000 years of human history – and they were ultimately disputes about the justification of philosophical speculations or religious dogmas.
But the question of what holds the world together at its core did not really go anywhere, and with the increasing development of mankind’s technical ability, e.g. with the invention of optical instruments such as the microscope or the telescope, which enormously expanded the limits of our perception, alleged causality was increasingly exposed as nonsense.
It was in this environment that modern science finally developed. Its basis is empiricism (experience and observation) and logic (lawful reasoning).
Science, on the other hand, compels us to abandon faith in simple causalities precisely where everything seems so easily comprehensible and we are the fools of the eye. Said Nie tzsche. And probably meant that there are countless functional connections in our world that explain why certain events have certain effects, but that we are all too easily satisfied with false explanations, if only they are convincing enough.
Ultimately, the scientific path of knowledge with its tools provides us with the means to recognise causalities where they are given and, accordingly, the means to identify the spurious correlations with which we have made fools of ourselves.
To put it in a nutshell: We are in the comfortable position, after 200 years of enlightenment and the accompanying thoroughly laborious shaping of a tool called science, as well as 150 years of intensive research based on it, to sit down and look at how things really are, how things actually behave and then draw appropriate conclusions. In other words: We are able (certainly not in every case) to recognise when we are dealing with sense or with nonsense, with causality or with spurious correlation.
So what does this have to do with homeopathy?
Homeopathy is a collection of spurious correlations. To list them all would make the next section of this chapter superfluous, so I will limit myself here to one example – the homeopathic drug trial. Since not everyone may be familiar with this, I will briefly describe what is behind this procedure. Healthy volunteers are given test substances, i.e. substances that homeopaths expect to have a medicinal effect. These substances are potentised according to homeopathic manufacturing practice, which outside the homeopathic world simply means “diluted”. These substances are usually diluted to a concentration of D30.
After taking this test substance, the subjects record any symptoms they experience over a period of time. Symptom in this case means everything that the test persons perceive as expressions of life. No matter what it is. The test administrator now has the task of filtering out from these manifold expressions of life those that, let’s say, particularly catch his eye. If similar life expressions occur in several test persons, this is an urgent reason for the test director to assume that these are possibly the leading symptoms of the remedy taken.
According to homeopathic doctrine, these leading symptoms are transferred to the remedy picture of the test substance. If these symptoms now occur in connection with a disease, the tested remedy is considered to be a remedy for these symptoms. The background to this idea is Hahnemann’s hypothesis that test substances which are taken by healthy people trigger an artificial disease in them, which then guarantees their effectiveness against an actual disease. (Although one should start shaking one’s head vigorously at this point at the latest, we will simply accept this hypothesis).
But what we have to ask ourselves is this: Is there evidence for the causal relationship between the ingestion of the test substance and the symptoms recorded by the subjects, or is this a spurious correlation? To paraphrase Nietzsche, is this a case of appearances deceiving us into believing that a correlation does not actually exist, and is science forcing us to look at things differently?
We are familiar with appearances: Events happen after an event. But what does science say?
D30, i.e. the dilution/potency of the test substance, means that the remedy to be tested was diluted in the order of 10^30.
Now it is the case that even at a dilution of 10^24, according to the Avogadro number, statistically only 0.6 molecules of the original substance are still contained in one mole of end product; the entire remainder of the test substance is distributed in the course of the dilution procedure in the individual dilution stages with decreasing concentration. In addition, we are only at D24, and with each further dilution step in the D-scale, the probability of encountering a molecule of the original substance is reduced by a factor of 10. It is therefore many times more likely that the test persons will not come into contact with the original substance at all than that they will ingest it. And even in the case of the greatest coincidences, it is only a matter of single molecules at the most. Now, the principle of causality requires that there is a sufficient reason for a cause-effect relationship, in this case an existing test substance. However, it is virtually impossible that this is the case. So everything speaks for the fact that the symptoms or life expressions recorded by the test persons have nothing whatsoever to do with the ingestion of the test substance.
Let us now look at the matter from another perspective: we know that everything has a reason. So is there any reason to assume other causes instead of the test substance that are capable of triggering the effect, i.e. the recorded symptoms?
Let us take two symptoms as an example, which are regularly found in homeopathic remedy pictures: Chills and sweating. Imagine, dear reader, that a drug trial takes place in January. How high would you estimate the probability of finding the symptom “shivering” in several test persons? Now the scenario with the same test substance, but with test date in midsummer, on muggy August days: How often would you probably find the symptom “sweating” in the records? What would be the probable cause instead of the test substance without substance, ? Where would you see the cause for the effect? In the weather or in the test substance? How would you, dear reader, assess the legitimacy of the symptoms “shivering” or “sweating” as a result of the test substance taken?
Drug trials, in which homeopaths imitate a scientific approach, test the test substance blinded and against placebo. The subjects who belong to the placebo group are given sugar pellets that do not contain any active ingredient.Whether a test person belongs to the verum group (the verum group is the one that receives globules with the active substance diluted away) or to the placebo group, the test person knows just as little as the test supervisor should know. This is blinding.
The result of such a setting is clear and there are enough examples of such tests. The subjects in the placebo group do not produce fewer symptoms than those in the verum group – different symptoms and, what is surprising at first, also the same symptoms. In a test conducted in this way, it is not possible to determine from the symptoms whether a test person was assigned to the verum group or the placebo group. This result is not surprising, however, if we briefly recall that the test substance administered in the dilution/potentiation D30 no longer contains any active ingredient – so that basically placebo is tested against placebo. This is exactly what the results show. By the way, the placebo control group is by no means obligatory in the drug tests of homeopaths and even if it is, their results are often “left out”. More on the regulations of homeopaths on drug testing and their practice can be found here.
The conclusion to be drawn from this is that there is no causal relationship between the test substance and the recorded symptoms.
The assumed effect relationship is a classic “cum hoc ergo propter hoc” fallacy.
What conclusions are to be drawn from this?
The entire remedy pictures of homoeopathy are consistently recordings of life expressions which would also have been observed in this way if this had not taken place in connection with the remedy test.
It is possible that individual symptoms, especially the psychological effects, can be traced back to the examination situation (stress, excitement, fear, etc.).
However, it can be ruled out that the (diluted away) test substance caused the symptoms of the test persons.
We must therefore assume that all homeopathic drug tests which were carried out with test substances in the dilution/potency D24 and higher did not produce any drug-specific, but exclusively placebo symptoms – and thus spurious correlations – and do not have the slightest significance with regard to the efficacy of a homeopathic remedy.
The same applies to the false conclusion “globules taken – patient healthy”. There can be a multitude of reasons why the patient felt better after taking the globules, but not because of taking the globules.
To read on:
Homeopathy helped me though! (this website)
Why the whole homeopathy falls with the principle of similarity (Susanna doesn’t need globules – in German)
The systemic contradiction between drug trials and the “absence of side effects” of globules (this website)
Similarity principle and “homeopathic first worsening” (this website)