Monday, 1 February 2010
We've got no poison; we've got a "remedy"
On Sat 30th Jan 2010, I led the small but perfectly formed Brighton 1023 : Homoeopathic "overdose" event. We only had a few people as I only had a couple of weeks to put it together while the other groups had been planning for months. But we made the video and had a few laughs before and in the pub after.
Before that, at the previously unheard of hour of 8 O'clock on a Saturday morning, I had a brief appearance on BBC Radio Sussex. In the interests of the much vaunted BBC "balance", they had arranged for a practising water-wizard (and graduate microbiologist!) Trevor Gunn to come and oppose me.
Here's the link on Listen Again if you're interested, although I don't know how long those things stay up. I'm on at around 2:14.45
Firstly let me say that Trevor seemed like a very nice guy, as many homoeopaths are; a pleasant persona really helps the placebo effect along nicely. Plus they usually genuinely care about the health and wellbeing of the people under their misguided care. So I'm not intending this as a character assassination on Trevor, but some criticism is inevitable. Also props to him for keeping his cool since I was essentially suggesting that his whole profession is built on lies and/or idiocy.
One point he made, which had also been stated by several other homoeopaths in the press this week, was that the "overdose" wouldn't prove anything because homoeopathy is a very individual thing and it was likely that none of us would be susceptible, or somesuch trite weaselly nonsense. Trevor made an analogy with teasing a room full of people with a peanut allergy by eating handfuls of peanuts and saying "look it doesn't do anything to me!". Unfortunately this analogy is flawed and the argument as a whole even flies in the face of one of the things that homoeopaths themselves claim about the mechanism of action for their "remedies"
When Samuel Hahnemann pulled the idea for homoeopathy out of his arse in the late 18th century, he did so on the basis that taking a substance that would cause a symptom in a healthy subject, would also cure that symptom in an ill subject, and vice versa. This is of course completely unproven and without any basis in science, but nonetheless it's what they claim. In fact, the very mechanism that these dilution-druids use to establish what each remedy is useful for is based on this principle. Hahnemann came to this "realisation" after he consumed doses of of cinchona bark (a source of quinine, and treatment for malaria) and then experienced some of the symptoms of malaria. Here we see our old favourite fallacy "post hoc ergo propter hoc" come visiting again, along with a spectacular "Illicit Conversion". Coincidentally, the symptoms Hahnemann experienced after taking his cinchona could also be interpreted as the symptoms of stress and anxiety; draw from that what you will.
The inaptly named "provings" (from the German "Prüfung" meaning "test" or "exam") is a method where a homoeopathic "remedy" is given to healthy subjects, Hahnemann having had a rare moment of clarity and realised that symptoms of other diseases could confound the results, and their health is monitored to note any symptoms that might occur. Hahnemann recommended the use of a 30c* "potency" in "provings", the same "potency" used by everyone taking part in the 1023 event, and these ultra-dilutions are still most commonly used in "provings" today. Interestingly "provings" do often include a control group, and sometimes today the tests are even "double-blinded". Blinding is where the person concerned is unaware of whether the treatment being given is a placebo or not. In single blinding, the patient does not know; in double-blinding the therapists do not know either, thus preventing them from giving clues, unconscious or otherwise, to the patient which may cause them to respond differently to the treatment. So this at least is a step in the right direction, but of course with homoeopathy both groups are actually receiving only placebo.
Unfortunately there remains one problem with this method, and it is that the symptoms under examination are decided upon after the testing is complete. Note that this is not just that statistical analysis is performed after the results are in, as in genuine clinical trials, but that what is being looked for isn't decided until after the data is available. In genuine clinical trials it is regarded as suspect if a proposed medicine is tested for a particular condition and after the fact the statistics show that it seemed to have an effect on some other condition that was not the original focus of the trial. In these circumstances there is a good chance that the apparent pattern is simply down to chance or over-interpretation of the results. As is beautifully demonstrated by the Bible Code nonsense (check this out, for Simon Singh's amusing debunking of the Bible Codes), if you look at a large enough dataset for long enough, seemingly meaningful, but ultimately meaningless, patterns seem to emerge. Erroneous results of this kind have caused some physicians to call for all trials to declare their aims and describe their protocol before the trial begins. This would also assist with eliminating another confounding factor in medical trials called Publication Bias. It is true that some medicines have been found to be effective for other conditions because of these apparent anomalies, but the effectiveness for that condition can only be determined after further studies looking specifically for that effect, thus eliminating the possibility that the original result was down to chance alone. Essentially what happens here is that whatever symptoms the treatment group coincidentally experience in greater amounts than the control group is assumed to be what that treatment will be effective for. Of course which statistics the experimenters pick from the data will be guided by all the cognitive biases that the genuine clinical trials are designed to guard against.
Bottom line is, "provings" are bunk-science, spectacular examples of Post Hoc reasoning, and have about about as much relation to real science as the mock landing-strips of the cargo cultists have to real military airfields. This exercise in hydromancy proves nothing. Homoeopaths also claim that dilution and succussion** increase the potency of the substance. If provings were any kind of proof, then you would expect a significant portion of last Saturday's "Overdosers" to fall ill with the symptoms of arsenic poisoning. It's been two days now and as far as I know everyone is still fine.
P.S. this first post in a year or so, resurrecting this blog marks a slight change in direction away from criticism of religion alone, and a turn towards my thoughts on all kinds of bullshit. Religion is unlikely to get off lightly, but it's no longer going to be the sole focus of the blog.
*A 30c "potency" let's not forget is about the same a one molecule of active ingredient in a sphere of water with the same diameter as the distance from the earth to the sun. There is a statistically negligible chance that there is even a single molecule of the active ingredient in a given "remedy". Homoeopaths claim that the water somehow retains a memory of the substance it once contained and that this memory is enough to effect a cure. There is no credible evidence for either of these leaps of faulty reasoning. It's also worth remembering that with the pillules as sold by Boots, a drop of this water is applied to some sugar pills, and then allowed to dry out, so the water is not even present anymore. So, now we are relying on the power of sugar to remember the power of water to remember some substance that causes the same symptoms we are trying to cure. I would say that you just can't make this shit up; except of course that someone did.
**Banging the bottles of water on a pad made of leather and horsehair to simulate the "effect" of Hahnemann's transportation of his remedies on horseback. I shit you not!