by apgaylard (from A canna' change the laws of physics)
In the third part of my series examining an attempted refutation of the critics of homeopathy (Milgrom, 2009) I look at the claim that homeopathy has a serious scientific foundation.
This part of the essay starts by outlining a common criticism levelled at the most common form of homeopathy practised in the US and UK. This calls homeopathy unscientific because:
“[…] in many homeopathic remedies, the original substance has been diluted out of molecular existence, detractors claim belief in homeopathy has no basis in science as ‘nothing cannot do something’.”
So, can apologists for homeopathy point to serious scientific work which shows that nothing can do something? Milgrom’s approach is to cite recent claims invoking concepts from materials science and physical chemistry to suggest that:
“[…] homeopathy’s method of remedy preparation leads to modifications in the dynamic long-range supra-molecular ordering of solvent molecules; an effect called the ‘memory of water’”*.
Before examining the evidence Milgrom marshals, it is worth reflecting on what we should expect to see if it is really science. This is a complex topic and there are many different descriptions of what, in practise, science is.
However, there is practically universal agreement that science is based on the formulation and testing of hypotheses. This means that to be scientific an idea must be testable. It also puts careful experimental practise at the core of science.
Also, as we all tend to become attached to our own ideas, even when there is evidence against them, scientific methods include strong precautions against scientists fooling themselves. This is vital, as Feynman (1974) remarked, “The first principle is that you must not fool yourself – and you are the easiest person to fool. So you have to be very careful about that.”
Central to this is the honest search for reasons why an idea might be wrong. Feynman (1974) described the process as:
“[…] a kind of scientific integrity, a principle of scientific thought that corresponds to a kind of utter honesty – a kind of leaning over backwards. For example, if you’re doing an experiment, you should report everything that you think might make it invalid – not only what you think is right about it: other causes that could possibly explain your results; and things you thought of that you’ve eliminated by some other experiment, and how they worked – to make sure the other fellow can tell they have been eliminated.”
This includes testing hypotheses in ways that might break them. It also means carefully comparing new ideas and results against previous work and established theory.
Scientific investigation also includes methods for limiting the influence of personal biases. The advent of automated measurement systems means that the mistake of Blondlot – imagining that he was seeing scintillations from non-existent N-Rays, because he was so personally invested in his ‘discovery’ – can be avoided.
Where an experimenter or subject can subconsciously influence the result – medical trials being a good example – a real scientist takes proper precautions. For instance, the investigators and subjects not knowing if they are receiving a new drug or a indistinguishable dummy (blinding) prevents them from being tempted to see ‘expected’ improvement where there is none. Making sure that if two groups are being compared they are as similar as possible, through assigning participants to them at random (randomization) is another precaution against being misled.
Where measurements can vary because of factors outside of the scientist’s control, repeated measurements are made and statistically tested to see if two samples, for instance, are really different or if a measured difference is just happenstance.
Finally, the use of controls – like making the same measurements on pure solvent from the same bottle used to make test solutions, for example – helps ensure that scientists are not fooled by the vagaries of the real world.
Many experiments produce anomalous results. In real science they are carefully examined in well-controlled experiments. Detailed measurements are made and possible causes of error are progressively excluded. Anomalies that survive this scrutiny may well go on to challenge current theories, but most melt away. So real scientific investigators don’t leap to wild conclusions; they carefully seek the simplest explanation.
So if Milgrom is really offering up examples of scientific investigations that support the basic plausibility of homeopathy, they will demonstrate these basic scientific virtues. I expect to see careful, well documented experiments; possible flaws pointed out and explored; evidence of blinding and randomization, where appropriate; the presentation of statistical data where the experimental system produces varying results. I also would not expect to see wild interpretations made of anomalous results; particularly if simpler explanations are possible.
The evidence Milgrom provides on the topic of materials science relies on two publications involving Rustum Roy, an eminent materials scientist with an enviable publications record. The only problem is that latterly he appears to have hitched his wagon to the alternative health movement. This seems to have compromised his objectivity.
In Roy et al. (2005) we see a review of the many amazing and occasionally anomalous properties of water. This work argues that the physical properties of water support the idea that homeopathic preparations can ‘remember’ what solute was originally added and diluted out of existence and, through structural changes, communicate this to patients.
The problem is that the gap between what can be measured and that which is merely asserted to justify homeopathy is never closed. The exercise is a lengthy non sequitur: just because water has some anomalous properties doesn’t mean that it can remember what used to be dissolved in it.
In reality, this paper is a collection of straws that are desperately clutched at. Nowhere is this clearer than in the discussion of the potential for contamination in experiments purporting to study high-dilution remedies:
“Obviously chemical contamination from the container material could itself serve as a “remedy”.
This is both desperate and not obvious. If contamination from containers could be the homeopaths actual remedies, then the remedies are uncontrolled and bear no resemblance to their claimed ingredients, or their supposed therapeutic effects. The ‘remedies’ will be different each time and will vary between homeopaths. A medical practise based on accidental contamination cannot be in any sense plausible, let alone ’scientific’.
The biggest problem with Roy et al. (2005) is that it confuses real measurable physical phenomena (electric and magnetic fields, for instance) with the immeasurable “subtle energies” of the CAM practitioner. This beggars any claim to be a real scientific publication. Citing silly papers that claim to be able to show the measurable effect of “human intention” and “qi” on chemical systems provides the final nail in the coffin. This is not science. It may have the appearance of science, but it lacks proper content. A real scientific paper would critically examine paranormal claims, not just accept them at face value.
So, how is it that it appears to have been published in a scientific journal? Actually, it’s not. It’s published in Materials Research Innovations , Rustum Roy’s own journal; a publication that rejects peer review of papers in favour of reviewing the authors. One consequence appears to be that if you have published some good work in the past, as Roy has, then you can publish any old nonsense in the future; as Roy and his co-workers demonstrate.
Next, the essay references Rao et al (2007). This deeply flawed paper, published in the homeopathic vanity press, which claims to show that homeopathy is plausible because they came up with some spectrographic measurements that appeared to show differences between homeopathic remedies and their solvents.
As Kerr et al (2008) pointed out: the spectrum contained in the paper purporting to be ‘pure ethanol’ does not look like ethanol of any recognised degree of purity. Further, from the paper, there is no way to know whether the reported differences between the spectra were the result of using solvent containing different levels of impurities.
Worryingly, it contains no statistical information, so no conclusions can be drawn as to whether the remedies were actually different. Finally, one graph was reproduced twice, with the authors claiming that it showed different things each time.
The author’s reply (Rao, 2008) failed to address any of these serious concerns. Sticking your head in the sand when serious flaws are identified in your work is not doing real science. There again, this was published in the journal Homeopathy, which is clearly not a real science journal.
Let’s get physical
So, no real science so far: just an enthusiasm for sloppy work and the paranormal. What about the evidence that Milgrom sees coming from physical chemistry?
Samal and Geckler (2001) is cited because this reports evidence that water molecules ‘clump’ around solutes; tending to form bigger clumps at lower solute concentrations. How does this help the argument that there is a scientific rationale supporting the effect of solutions where the solute is highly unlikely to have survived successive dilutions? Quite simply: it does not. Any structures formed around solute ions will be finite in number and diluted out of solution: just as any finite solute concentration of anything will be. At least this paper is real science; but it says something that it definitely does not support the claims of homeopathy.
The next anomaly that is cited to show the scientific plausibility of homeopathy is provided by Rey (2003). In this study a technique called Thermoluminescence was used to study samples of frozen heavy water (D2O); some of which were the result of diluting preparations of lithium chloride and sodium chloride beyond the point where any molecules of these salts could be expected to be found in the solution. This process replicated a homeopathic method of ‘remedy’ preparation: successive 1:100 dilutions with vigorous mechanical shaking (succussion) at each step.
Thermoluminescence is the, “emission of light from some minerals and certain other crystalline materials” as their temperature is raised. The energy of this emission is, “derived from electron displacements within the crystal lattice of such a substance caused by previous exposure to high-energy radiation.” Heating the material, “enables the trapped electrons to return to their normal positions, resulting in the release of energy.” (Encyclopædia Britannica, 2009)
This is a proper scientific technique, generally used to date archaeological artefacts and minerals. Rey appears to be pioneering its application to the study of frozen solutions – assuming that structures in the liquid phase will be preserved by freezing.
Although the Rey (2003) claims that, “despite their dilution beyond the Avogadro number, the emitted light was specific of the original salts dissolved initially” it provides few details for a paper advancing such a radical hypothesis. The investigation of the ‘highly dilute’ samples (C15 LiCl, C15 NaCl and C15 D2O) did not include an obvious control: the unsuccussed solvent (D2O) making it impossible to separate the putative influence of the (non-existent) salt ions from changes made by shaking the samples.
Neither did Rey provide any statistical information, so there is no way of telling if the differences measured were real, or just down to chance. As a contributor on this blog has noted:
“[…] the scientific basis of the emissions recorded – “what do these kind of readings tell us?” […] is actually pretty obscure. The technique relies on freezing the sample, irradiating it in one of a number of ways, and then watching the thermoluminescence emissions while the sample re-warms. The general message is that the emissions depend upon “structural irregularities” in the crystal lattice of the frozen sample, but the details, to repeat, are poorly understood.”
If there really are differences between the C15 ‘dilutions’ then what might be the cause? Rather than grasping at implausible explanations, a real scientific approach eliminates the mundane. For instance, could it be that shaking the solutions has changed their physical characteristics? Rey (2007) has gone on to explore the possibility that vigorous shaking causes the formation of ‘nanobubbles’ in the solutions, and that these, when frozen, provide the structures associated with the thermoluminescence spectra. The investigation looked at the spectra generated by samples “succussed” under the standard laboratory atmosphere, pure oxygen and vacuum. The spectra appear to be different but, again, no statistical data was included to allow a reader to be sure.
Construing this work as evidence of a scientific basis for the claims of homeopaths is unwarranted. Yes, it is science of a sort; it’s not without flaws and obvious biases. Neither is it very convincingly reported. However, all it provides is evidence for the presence of not very well understood structures in heavy water ice. There is some evidence that they are associated with the shaking of the solutions, but the missing control (unsuccussed D2O) limits the conclusions that can be drawn. In fact, I would say that Rey’s contention that the spectra were specific to the salts goes too far, as he was not in a position to separate the implausible influence of the absent chemicals from that of the process by which his solutions were made.
One of the key values in science is that of replication. If independent scientists, working in other places, can get the same result from the same experiment then this helps separate knowledge from happenstance. Milgrom claims that the findings in Rey (2003) were replicated by van Wijk et al. (2006). They were not.
This work did try and replicate and extend the findings of Rey (2003). Using Lithium Chloride (LiCl) as the solute and heavy water as the solvent (D2O) they also explored the possible influence of, “time between preparation of substance and time of experimentation, and […] time between irradiation and thermoluminescence recording”.
Unlike Rey (2003) van Wijk et al. (2006) presented a statistical analysis of their measurements. Their like-for-like attempted replication failed:
“We report here differences in thermoluminescence between C15 D2O and C15 LiCl, which correspond with the observations reported by Rey (2003). However, the difference from all of these recordings of these substances was not statistically significant.”
In the replication experiment (A) the difference between C15 LiCl and C15 D2O did not reach statistical significance (p = 0.059, ANOVA t-test). In the experiments that looked at the influence of the time between sample preparation and freezing, or the time between irradiation and thermoluminescence measurement (B and C) There wasn’t even a hint of a significant difference (p = 0.72, and p = 0.63, respectively, ANOVA t-tests).
However, they did report some statistically significant differences. When the data were processed differently the result suggested that LiCl C15 differed significantly from C15 D2O in experiment A (p = 0.0128); but not in experiments B and C. (p = 0.60 and 0.73, respectively).
The best evidence of a difference between samples was seen in the comparison between the succussed (C15 D2O) and unsuccussed (D2O) solvent for experiment C. This used the maximum time between sample preparation and freezing (12 weeks), and between irradiation and thermoluminescence measurement (3 weeks). The result was statistically significant under both data processing methods (p<0.0001, and p<0.0004, respectively).
It’s clear that van Wijk et al. (2006) did not replicate the findings of Rey (2003), as Milgrom and, indeed, Rey (2007) has claimed. There might be an interesting anomaly to pursue here, but there is no real evidence of absent solutes being the cause. Van Wijk et al. (2006) does show that the experimental system is a ‘noisy’ one and that careful statistical analysis is required.
“The point must be stressed that we obtained a very good qualitative reproducibility of the thermoluminescence pattern, but the quantitative reproducibility was rather poor, and p-values should be interpreted in the sense of descriptive statistics.”
This emphasises that the lack of statistical data in Rey (2003) is a serious flaw: most of van Wijk et al.’s comparisons did not show statistically significant differences.
The differences reported between the succussed and unsuccussed solvents also strengthen the impression that any differences are more likely the work of nanobubbles and not a watery memory of long gone solutes. Again, this highlights Rey’s oversight in not using unsuccussed solvent as a control. In a delightful irony even Roy et al. (2005) show they understand the potential importance of this omission:
“It is important to emphasize that the proper control solutions include not only untreated, unsuccussed solvent, but also succussed solvent without the initial addition of any remedy source materials to address possible artifacts generated by the shaking of the liquid per se within the test container itself.”
Next the essay sees scientific support in the work of Elia et al. (2006). This is another attempt to find physically measureable differences between homeopathic solutions that do not contain any of the original solutes and their solvents. It’s also based on very different physical principles to thermoluminescence. This adds to the impression that this is just chasing after anomalies. I don’t have access to this paper, so I’ll not comment further. However, in 2007, the same author (Elia, 2007) published a review of the evidence they had accumulated.
It’s not very impressive: no statistical data are provided to help the reader understand if any differences are significant or not. Neither is there any indication of how many times (if at all) measurements were repeated. It also contains an odd confession:
“It is important to emphasise that, from the studies so far conducted, we cannot derive reproducible information concerning the influence of the different degrees of homeopathic dilution or the nature of the active principle (solute) on the measured physicochemical parameters.”
If different concentrations of homeopathic preparations cannot be distinguished, then it casts serious doubt on any claims to be able to differentiate between homeopathic preparations. This is really clutching at straws**.
The attempt to show that homeopathy is grounded in science peters out from here. An irrelevant theoretical speculation on Quantum Electrodynamics (QED) is thrown in to the mix (Arani et al., 1995). Martin Chaplin’s fascinating website on the properties of water is also referenced.
Milgrom also resorts to an inappropriate analogy:
“Just as two physically contrasting substances, such as diamond and graphite, are composed of exactly the same carbon atoms arranged into different molecular structures, so it is not the composition of an ultra-diluted homeopathically-prepared solution that is different from plain diluted solvent, but its dynamic supra-molecular structure.”
Well, diamond and graphite both have structure – being solids – liquid water does not (Teixeira, 2007). Clearly, there is no “just as” about it! The rest is just unsupported opinion.
The same can be said of the reference to Hankey (2004) who provides evidence and data-free hand waving of a distinctly unscientific variety. Here is a sample:
“In this model, all vibrational medicines are quantized fluctuations, of mineral, vegetable, animal, mental, psychic, or spiritual origin. Succussion and dilution potentize the first; correct formulation of phytomedicines, the second; while the last four are all involved in various levels of healing. For example, in Maharishi Vedic Vibration Technology (Nader et al., 2001), use of a mantra develops the specific healing vibration within the technician’s nervous system, for transferal to the patient.”
The appeal to some kind of quantum theory is bogus***. The rest either has no meaning or is paranormal. This is not science and by citing it Milgrom eloquently debunks his own argument.
A poor memory
And that is it†; Milgrom presents this as a refutation of the claim that homeopathy has no scientific basis. At best his argument rests on a few anomalous experimental results (Rey, 2003; Elia et al, 2006), which are likely to be explained by very ordinary causes: bubbles causing by shaking, chance readings in noisy experimental systems, contaminated samples, etc.
The better work he refers to doesn’t help either. Van Wijk et al. (2006) fails to provide the replication of Rey (2003) that Milgrom (and Rey, 2007) claims. Samal and Geckeler (2001) does not provide a way for homeopaths to cheat Avogadro.
Much of the rest actually provides an elegant confirmation of the critics’ accusation by ignoring scientific values and asserting the reality of imaginary ‘energies’. (Roy et al., 2005; Hankey, 2004)
It is clear that Milgrom believes that water has a ‘memory’. Unfortunately this is not just unsupported by scientific evidence; it is contradicted by it (Teixeira, 2007). The quality of the evidence Milgrom has marshalled here bears witness to that fact.
If Milgrom really wants to turn to science, then he needs to rediscover its essential integrity:
“[…] it’s this type of integrity, this kind of care not to fool yourself, that is missing to a large extent in much of the research in Cargo Cult Science.” (Feynman, 1974)
Of course, this would entail leaving the Cargo Cult Science of the homeopathic apologist behind. From the evidence on display here, I don’t think that’s very likely.
Next, I’ll have a look at what Milgrom has to say about homeopaths, the pharmaceutical industry and money.
Also in this seriesA homeopathic refutation – part one – evidence. (on this blog)
I try to make sure that what I write is both accurate and fair. If you think that I have got anything wrong please let me know. If you are right I will happily change what I have written.
* See here for my summary of the “Memory of Water” issue of the pseudo-journal Homeopathy.
**Interestingly Elia et al. (2008) published a conductivity study purporting to show the effect of aging on homeopathically prepared solutions. This was strongly criticised by Corti (2008) who asserted that: the equipment used is not capable of taking measurements of the sensitivity reported; samples were stored over time in brown glass bottles, which are known to leak conductive ions such as Fe and Ni; the shifts in conductivity over time were cyclical over the period of a year and best explained by annual shifts in temperature. The criticisms were strongly rejected by Elia (2008).
*** See here for a demolition of Hankey’s “physics”.
†For completeness, the final reference in this part of the essay  is to Collins JC: Water: The Vital Force of Life. Molecular Presentations. New York, Kinderhook, 2000. This is out of print, but Amazon carries some details.
“thermoluminescence.” Encyclopædia Britannica. 2009. Encyclopædia Britannica Online. 25 Sep. 2009 Available from: http://www.britannica.com/EBchecked/topic/591643/thermoluminescence
Arani R, Bono I, Del Giudice E, Preparata G. QED COHERENCE AND THE THERMODYNAMICS OF WATER. International Journal of Modern Physics B. 1995;9(15):1813–1841. Available from: http://dx.doi.org/10.1142/S0217979295000744
Corti H. Comments on “New Physico-Chemical Properties of Extremely Dilute Solutions. A Conductivity Study at 25 °C in Relation to Ageing”. Journal of Solution Chemistry. 2008 December;37(12):1819–1824. Available from: http://dx.doi.org/10.1007/s10953-008-9335-6.
Elia V, Elia L, Cacace P, Napoli E, Niccoli M, Savarese F. ’Extremely diluted solutions’ as multi-variable systems. Journal of Thermal Analysis and Calorimetry. 2006 May;84(2):317–323. Available from: http://dx.doi.org/10.1007/s10973-005-7266-7.
Elia V, Napoli E, Germano R. The ’Memory of Water’: an almost deciphered enigma. Dissipative structures in extremely dilute aqueous solutions. Homeopathy. 2007 July;96(3):163–169. Available from: http://dx.doi.org/10.1016/j.homp.2007.05.007.
Elia V, Napoli E, Niccoli M, Marchettini N, Tiezzi E. New Physico-Chemical Properties of Extremely Dilute Solutions. A Conductivity Study at 25°C in Relation to Ageing. Journal of Solution Chemistry. 2008 January;37(1):85–96. Available from: http://dx.doi.org/10.1007/s10953-007-9215-5.
Elia V. Response to Comments on a “New Physico-Chemical Properties of Extremely Dilute Solutions. A Conductivity Study at 25°C in Relation to Ageing” by Horacio R. Corti. Journal of Solution Chemistry. 2008 December;37(12):1825–1826. Available from: http://dx.doi.org/10.1007/s10953-008-9345-4
Feynman RP. Cargo Cult Science. Engineering and Science. 1974 June; pp. 10–13. Available from: http://calteches.library.caltech.edu/51/2/CargoCult.pdf.
Hankey A. Are We Close to a Theory of Energy Medicine? Journal of Alternative and Complementary Medicine. 2004 February;10(1):83–86. Available from: http://dx.doi.org/10.1089/107555304322848995
Kerr M, Magrath J, Wilson P, Hebbern C. Comment on “The defining role of structure (including epitaxy) in the plausibility of homeopathy”. Homeopathy : the journal of the Faculty of Homeopathy. 2008 January;97(1). Available from: http://dx.doi.org/10.1016/j.homp.2007.10.004
Milgrom LR. Under Pressure: Homeopathy UK and Its Detractors. Forsch Komplementmed. 2009 September;16(4):256–261. Available from: http://dx.doi.org/10.1159/000228916
Rao M, Roy R, Bell I, Hoover R. The defining role of structure (including epitaxy) in the plausibility of homeopathy. Homeopathy. 2007 July;96(3):175–182. Available from: http://dx.doi.org/10.1016/j.homp.2007.03.009.
Rao M. Authors’ reply to Kerr et al. Homeopathy. 2008 January;97(1):45–46. Available from: http://dx.doi.org/10.1016/j.homp.2007.11.011
Rey L. Thermoluminescence of ultra-high dilutions of lithium chloride and sodium chloride. Physica A: Statistical Mechanics and its Applications. 2003 May;323:67–74. Available from: http://dx.doi.org/10.1016/S0378-4371(03)00047-5.
Rey L. Can low-temperature thermoluminescence cast light on the nature of ultra-high dilutions? Homeopathy. 2007 July;96(3):170–174. Available from: http://dx.doi.org/10.1016/j.homp.2007.05.004.
Roy R, Tiller WA, Bell I, Hoover MR. The Structure Of Liquid Water; Novel Insights From Materials Research; Potential Relevance To Homeopathy. Materials Research Innovations. 2005;9(4):577–608. Available from: http://www.rustumroy.com/Roy_Structure%20of%20Water.pdf
Samal S, Geckeler KE. Unexpected solute aggregation in water on dilution. Chemical Communications (Cambridge, England). 2001 November;(21):2224–2225. Available from: http://dx.doi.org/10.1039/b105399j.
Teixeira J. Can water possibly have a memory? A sceptical view. Homeopathy. 2007 July;96(3):158–162. Available from: http://dx.doi.org/10.1016/j.homp.2007.05.001.
van Wijk R, Bosman S, van Wijk EP. Thermoluminescence in ultra-high dilution research. Journal of Alternative and Complementary Medicine (New York, NY). 2006 June;12(5):437–443. Available from: http://dx.doi.org/10.1089/acm.2006.12.437.
dvnutrix for pointing me at this nonsense.
DrAust for his insightful comments on thermoluminescence.
gnu and Acleron for their comments on Elia et al. (2007) during the Homeopathy journal club run at Badscience.net.