Homeopathy: Rustum Roy drops the bomb

[wilsontown]

Judging by what was written in a previous thread discussing the powerpoints, the figures in the paper are the same ones, showing UV-vis and Raman spectra. So, apart from all the things you folks have already pointed out about how rubbish this is, it seems they got their graphs mixed up as well.

As for review...well, it doesn't seem as though this stuff was reviewed by anyone with much of an understanding of science generally, never mind the methods used. It was presumably reviewed by homeopaths. I wouldn't expect a reviewer to play around with Corel Draw, to be fair, but there are plenty more problems that could be found.

As for the 'envelope of difference', well, who knows what it is. The thing with the two graphs is that in figure 1, the one of the curves is supposed to be Nux Vom and one Nat Mur, but in figure 2 they're both supposed to be Nat Mur. Who knows which of them, if either, is the correctly labelled one?

And no, I'm not Johnnie Scobie...

[Pipirr]

Originally Posted by wilsontown

As for the 'envelope of difference', well, who knows what it is. The thing with the two graphs is that in figure 1, the one of the curves is supposed to be Nux Vom and one Nat Mur, but in figure 2 they're both supposed to be Nat Mur. Who knows which of them, if either, is the correctly labelled one?

Good point. Hence, very embarrassing that it's the same graph in Figure 1 and 2.... maybe it's quantum?

[wilsontown]

I wonder if there's any point in sending a comment to Homeopathy, pointing out the basic errors in this paper? Surely they would have to at least issue an erratum, or even withdraw the paper?

[Pipirr]

Send in a comment? Not a bad idea.

I wrote to the lead author a few weeks back asking for clarification on the ethanol experiment, but never got a reply. Heck, maybe I shouldn't have indicated that I was but a lowly PhD student.

Anyway, reusing the same graph for two different things should at least merit an erratum.

Worth a try.

[vidiviciveni]

Over the past 7 months I and 3 colleagues have complied a definitive document which debunks all Homeopathy totally - amounting to some 100,000 words it is a comprehensive tome of the very lasted scientific proofs that finally lay rest to this nonsense.

In order that these charlatans fully understand just how much I and my colleagues deplore their blatant disregard for pharmaceutical solutions we have decided to put our full documented claim to them in the strongest form at our disposal....

This is what we sent them.......

.


You may be shocked that we sent . which is a 10c dilution, however we felt a 4c paragraph, or 6c letter was not sufficient to state our case! We alas were not able to find a 0.0000000001 font size for a 20c tirade which was our initial aim.



(apologies if this has been done before, I'm a noob)

[Rolfe]

Figure 2 is slide 29 in the PowerPoint. That is captioned "Nat Mur (open circles) and Nux Vomica (closed circles)". This makes no sense to me as the lefthand set of graphs is labelled Nat Mur and the righthand set Nux vomica. I thought each set of axes was one preparation as described, with the open and closed circles representing the two most extreme spectra, but that doesn't really explain why the lines cross in some of the sets, as already discussed.

What I can't understand is the obsessive interest in measuring so many points right along the spectrum for every preparation. The shape of the spectra might be interesting if you were going to assign bond energies to the peaks or anything so rational. However, if you're just going to say, look at the difference, the way to do it would be to pick the wavelength at which the difference was greatest and do lots of measurements at that wavelength.

Geni estimated the lambda-max at about 324nm, which was a reasonable match for a possible contamination by acetone if I remember correctly! Anyway, if 324nm is the interesting wavelength, and it does appear that wherever there is a hint of a peak that's where it is, then just do lots of measurements of different preparations, with lots of repeats of each preparation, and do the statistics. Either there is a significant difference in absorbance at this wavelength or there isn't. If in fact there isn't due to poor precision, then there isn't a snowball's chance in hell you'll find it at any other wavelength.

It just beggars belief that anyone who knows anything about UV spectroscopy could accidentally produce such sloppy work. I know Roy is 84, but that doesn't mean he's ga-ga and if he's ga-ga then he shouldn't be publishing. Anyone at all involved with this who has the remotest idea about the subject whould be able to tell him this isn't the way to go about it.

I think he's run relatively few spectra, just with lots and lots of different wavelength-points for each run, and is then waving these spectra at us with no statistical treatment, as "different". Wilsontown seems to have spotted that he's run rather fewer spectra than one might at first sight assume. I just don't wholly understand the reference to the "series of 10" preparations, and if he's really done that, why aren't we seeing all the data for that. You know, means and standard deviations for each point and so on.

How much of this is incompetence and how much deliberate obfuscation, deliberately intended to provide an impressive-looking paper to be referred to be homoeopaths, realising that most people won't even bother to look to see if it's garbage or not? I have trouble going with the incompetence theory, given that actual possession of a spectrophotometer rather implies some degree of competence somewhere among the personnel. I have to suspect deliberate obfuscation.

But having said that, I'm rabbiting, and I haven't really looked at the thing in great detail yet.

Rolfe.

[Rolfe]

Originally Posted by wilsontown

I wonder if there's any point in sending a comment to Homeopathy, pointing out the basic errors in this paper? Surely they would have to at least issue an erratum, or even withdraw the paper?

You jest! These guys take even the slightest critical look at what they've printed? I don't see it happening.

(Sorry about the "Johnnie Scobie" joke, just the local version of "For we're no awa tae bide awa" - "As I cam doon by Wilsontown, I met wee Johnnie Scobie....". It's the only thing my tone-deaf father could sing, and I got to hear it often as a kid.)

Rolfe.

[Rolfe]

Originally Posted by vidiviciveni

Over the past 7 months I and 3 colleagues have complied a definitive document which debunks all Homeopathy totally - amounting to some 100,000 words it is a comprehensive tome of the very lasted scientific proofs that finally lay rest to this nonsense.

In order that these charlatans fully understand just how much I and my colleagues deplore their blatant disregard for pharmaceutical solutions we have decided to put our full documented claim to them in the strongest form at our disposal....

This is what we sent them.......

.


You may be shocked that we sent . which is a 10c dilution, however we felt a 4c paragraph, or 6c letter was not sufficient to state our case! We alas were not able to find a 0.0000000001 font size for a 20c tirade which was our initial aim.



(apologies if this has been done before, I'm a noob)

Someone from Manchester uni (I think) actually did this. They sent a load of blank pages to Peter Fisher, with suitable covering letter and so on. They put all this on the internet, including the po-faced, offended reply they got. It was so funny. But the last time I looked for it, the link was broken.

Rolfe.

[Pipirr]

Originally Posted by Rolfe

You jest! These guys take even the slightest critical look at what they've printed? I don't see it happening.

Do you think Elsevier would have an interest in the credibility of it's journals? I'd like to think so.

But having said that, they don't make any claims to provide quality research, not that I could find.

[Rolfe]

Trouble is, you'd have to contact the authors and/or the journal first, and get the brush-off from them, before Elsevier might be interested. And as I'm pretty sure there's deliberate obfuscation going on here, it might be difficult to get any reply.

The criticisms go deeper than mixed-up graphs anyway. I wonder if it's worth taking our time about this, putting it all together - ethanol spectrum, methodological problems (lack of statistics, repeatability and data for the ten repetitions and so on) plus the duplicated graphs, and presenting it all wrapped up in a nice parcel with a bow round it?

Rolfe.

[Pipirr]

Originally Posted by Rolfe

How much of this is incompetence and how much deliberate obfuscation, deliberately intended to provide an impressive-looking paper to be referred to be homoeopaths, realising that most people won't even bother to look to see if it's garbage or not? I have trouble going with the incompetence theory, given that actual possession of a spectrophotometer rather implies some degree of competence somewhere among the personnel. I have to suspect deliberate obfuscation.

This does not look like the publication list of an incompetent individual; the lady does appear to have done some good work in materials science.

[Rolfe]

Originally Posted by Pipirr

This does not look like the publication list of an incompetent individual; the lady does appear to have done some good work in materials science.

Right. And Roy has a pedigree too (pointed out to me after I queried his credentials on the back of the obvious incompetence of the slide show).

So what the hell are they up to?

Rolfe.

[Pipirr]

What indeed. Pushing a belief despite the lack of supporting evidence. It looks like an attempt at obfuscation, and nobody called them on it.

[fls]

Originally Posted by Rolfe

Right. And Roy has a pedigree too (pointed out to me after I queried his credentials on the back of the obvious incompetence of the slide show).

So what the hell are they up to?

Rolfe.

This reminds me of an article published in Skeptic magazine by Marshall Deutsch called "The Truth About Cholesterol", that was so misguided that I seriously expected it to be revealed as a joke. I really appreciate that those of you with specific knowledge have taken the time to uncover this mystery. I suspect that this would benefit from being taken to a wider audience (wrapped up with a bow, as you put it) so that it comes to the attention of the peers of those involved. Being forced to answer or lose face may clear up just what they thought they were doing. Perhaps they were hoping that those in the know are unlikely to be regular readers of Homeopathy?

Linda

[Rolfe]

I've been scheming. And trying to read the paper!

I think a reasonably short letter to the editor (Peter Fisher himself!) BUT copied to Elsevier (and maybe to the authors) might be constructive. Or at least instructive. Signed (real names, pack drill the lot) by whoever understands spectrophotometry enough to feel able to, and has enough alphabet soup after their names to confer suitable gravitas. I'm certainly prepared to do it, but I think it will have more weight if there are several signatories.

I'm off sailing for the weekend, but I can start to put something together. My idea is to make it very factual, just pointing out the basic data problems, nothing contentious about why are you supporting this rubbish, not even any criticism of the pro-homoeopathy rhetoric, and maybe just finish it with a comment that if the authors are interested in addressing the plausibility of homoeopathy they should perhaps look to the plausibility of their own results in the first instance.

We've got the bogus ethanol spectrum, and the lack of any provenance for the source of the ethanol, and the lack of any data on repetitions, particularly the lack of any statistical treatment, maybe the nonsense-sentence where repeatability and precision are declared to be different things, and the duplicated graphs. I personally wouldn't touch the Raman spectra with a bargepole, but I know enough about UV spectroscopy to make it both technically convincing and right.

Wilsontown, can you pretty, pretty please post a list of which of these UV graphs are identical to which? It would make it so much easier.

Rolfe.

[Pipirr]

Originally Posted by Rolfe

I think a reasonably short letter to the editor (Peter Fisher himself!) BUT copied to Elsevier (and maybe to the authors) might be constructive. Or at least instructive. Signed (real names, pack drill the lot) by whoever understands spectrophotometry enough to feel able to, and has enough alphabet soup after their names to confer suitable gravitas. I'm certainly prepared to do it, but I think it will have more weight if there are several signatories.

Good idea. I'm sure there's a chemistry professor around here somewhere who can add some weight to a letter (although the letters MPH! carry weight with these guys).

Just a PhD student myself. I'll pm you my email to the author so you can see what didn't work.

[Rolfe]

Given these guys' level, I think a first degree in a science subject will look quite good enough. I've got a BSc (in biochemistry) and a PhD and I'm FIBiol too. So there's enough gravitas there to head up a reasonable list. Also, I'm used to refereeing papers.

I need Wilsontown to tell me which traces are identical though.

Rolfe.

[JJM]

For the masochists among you- the text of all the articles in the "special" issue of Homeopathy are available, free http://www.badscience.net/?p=490 .

[pgwenthold]

Before any letter gets sent, I strongly recommend finding a published spectrum of ethanol to compare it to this one. Either that or getting your own spectrum of ethanol. If you want to question the spectrum (and I think you can legitimately do that) there should be more basis for it.

If you can say, "Roy's spectrum of ethanol does not match that published by XXX). How do the authors account for the discrepancy?" it will add significant weight.

[JJM]

Originally Posted by Pipirr

Do you think Elsevier would have an interest in the credibility of it's journals? I'd like to think so. {snip}

Elsevier prints and markets good journals and fanzines, alike, if they can make money doing so. The quality of a publication lies in the hands of the sponsoring organization, in this case "The Faculty of Homeopathy" which is recognized by the British Parliament.

The quality of Roy's publication is in line with the quality of regular articles in this fanzine. Therefore, there is no point in complaining.

[Pipirr]

Originally Posted by pgwenthold

Before any letter gets sent, I strongly recommend finding a published spectrum of ethanol to compare it to this one. Either that or getting your own spectrum of ethanol. If you want to question the spectrum (and I think you can legitimately do that) there should be more basis for it.

If you can say, "Roy's spectrum of ethanol does not match that published by XXX). How do the authors account for the discrepancy?" it will add significant weight.

Harder than you might think! I scoured my university library in vain for such a thing.

The fact that ethanol does not signficantly absorb after what, 280nm, is why ethanol can be used as a solvent for things that do absorb in that range.

That may be why even basic spectroscopy primers don't carry full wavelength scans of ethanol. None that I could find...

As for getting a spectra of our own, that could be useful. It would be even better if the authors provided the product number for the ethanol that they used, so at least one could check with Sigma or whatever company provided it. The company may be able to provide the absorbance values. Sigma does, for some of it's ethanol grades.

[JJM]

Originally Posted by pgwenthold

Before any letter gets sent, I strongly recommend finding a published spectrum of ethanol to compare it to this one. Either that or getting your own spectrum of ethanol. If you want to question the spectrum (and I think you can legitimately do that) there should be more basis for it.

Done it. Roy's "standard" is highly contaminated (possibly- acetone, a standard denaturant). His homeopreps look like ordinary (i.e., slightly contaminated), potable, 95% alcohol.

Originally Posted by pgwenthold

If you can say, "Roy's spectrum of ethanol does not match that published by XXX). How do the authors account for the discrepancy?" it will add significant weight.

This is nice, in theory; but you are dealing with the "Faculty of Homeopathy," not scientists. The sloppiness in Roy's article is within the norm for those people.

If one wants to know how the authors account for spectral differences, they should have explained how, given a coded sample of a remedy, they could use their data to identify the "remedy" and its "potency." All they say is "look, they are different."

As for those wondering about Figure 2, we don't have enough data to speculate on it. Oh, heck, I speculate that they have one bottle of each prep (NV, NM). They prepared 10 samples from each for the UV-Vis. Then, they plotted the highest and lowest trace from each set of 10 to show ... just how poor their sample handling technique is. That is, the "envelope" (or, range) of traces which different samples of the same prep can generate. (2A- 12c and 6c- almost look like they had someone competent working that day.)

[JJM]

Originally Posted by wilsontown

{snip} When you superimpose the part of figure 2 that supposedly represents the 'envelope of difference' for 30C Nat Mur, onto the first graph in figure 1, which shows a comparison between 30C Nat Mur and 30C Nux Vom...the graphs are identical. So the same graph is supposedly representing two different things.

When I started posting this afternoon, I did not realize how much I had missed.

There is an old saying "Never attribute to malice what can be explained by incompetence." A simple explanation for this is that all we are seeing are selected spectra of variously mis-handled ethanol samples. It is not surprising if two different sets of spectra correspond. However, your point is well-taken.

Originally Posted by wilsontown

Apart from that, the curves in figure 1 are all from figure 2 (the envelope of differences figure), suggesting they are comparing extreme examples rather than 'representative' ones. {snip}

Thanks, I did not see that.

[pgwenthold]

Originally Posted by JJM

Done it. Roy's "standard" is highly contaminated (possibly- acetone, a standard denaturant). His homeopreps look like ordinary (i.e., slightly contaminated), potable, 95% alcohol.

When you said "done it" does that mean that you did get a spectrum? If so, what did you see in that 320-330 region?

Could you reproduce the feature by adding a touch of acetone?

[pgwenthold]

Originally Posted by Pipirr

Harder than you might think! I scoured my university library in vain for such a thing.

It's going to be in the VERY OLD literature. I think I found a reference from the 50s, but from a Korean journal that our library didn't have. I would suspect you might even have to go older than that. Like in Chem Ber in the 30s or so.

Quote:

The fact that ethanol does not signficantly absorb after what, 280nm, is why ethanol can be used as a solvent for things that do absorb in that range.

But they don't call it a significant absorption, so that doesn't really help. Just because a solvent doesn't have significant absorption past some wavelength does not require that it have NO absorption out there. Very weak transitions are still measurable spectroscopically, and don't get in the way of most chromophores.

I wouldn't be surprised if Aldrich even has a spectrum past 300 nm. No one bothers.

[JJM]

(Sigma) Aldrich has a large variety of "ethanols" for sale http://www.sigmaaldrich.com/catalog/...ltsPage/Expand . Most have a description of its UV-Vis absorption (click on a number; e.g., 493456 for potable alcohol).

[pgwenthold]

But no "description" of a UV spectrum is going to mention a minor bump at 320 nm, even if it is there. Note that this supposed feature is claimed to be minimally detectable, which means that the extinction coefficient is probably less than 1/1000th of that at lambda max, if that. You can't find this in a general description of the spectrum (just as a description of the vibrational spectrum won't mention every single bump, either)

You'd really need to see the spectrum to examine it properly.

[JJM]

Originally Posted by pgwenthold

But no "description" of a UV spectrum is going to mention a minor bump at 320 nm, even if it is there. {snip}

Where did you get this idea? Look at the "standard" spectrum in Roy's paper, it is not a minor bump at 324 nm. And the absorbance below that is way above the reported spectrum. You need considerable experience with spectra before you can argue with us about them. If you know what you are doing, you can get a lot of information from the published descriptions of spectra; otherwise it would be pointless to publish them.

Also, comparing your own spectra to his is a meaningless exercise since you don't know what impurities are causing his results. We know he is not using "spectral grade" ethanol. All you could say is yours look like his, or not.

[JJM]

Originally Posted by JJM

Where did you get this idea? Look at the "standard" spectrum in Roy's paper, it is not a minor bump at 324 nm. And the absorbance below that is way above the reported spectrum. {snip}

The [newly] italicized part was not clear- the absorbance in Roy's published spectrum in the 300-250nm range is much greater than in the spectrum described by Aldrich.

[wilsontown]

Quote:

I need Wilsontown to tell me which traces are identical though.

Righto...

As we know, the top graph of figure 1 is identical to the top graph in figure 2a, although they purport to represent different things.

In the middle graph of figure 1, the 12c NM curve matches the curve picked out by filled circles in the middle graph of figure 2a. The 12c NV curve matches the curve picked out by filled circles in the middle graph of figure 2b.

In the bottom graph of figure 2, the 6c NM curve matches the curve picked out by filled circles in the bottom graph of figure 2a. The 6c NV curve matches the curve picked out by filled circles in the bottom graph of figure 2b.

Hope that's all intelligible...

I'd be happy to sign any letter. I'm BSc and PhD, though in an unrelated field (geology).

[wilsontown]

Oh, missed one...

In the top graph of figure 2b, the curve picked out with filled circles (30c NV) is the same as the curve picked out with open circles (30c NM) in the top graph of figure 2a, which is in turn the same as the 30C NV curve in the top graph of figure 1.

What a mess. I only hope it all makes sense...

[pgwenthold]

Originally Posted by JJM

Where did you get this idea? Look at the "standard" spectrum in Roy's paper, it is not a minor bump at 324 nm. And the absorbance below that is way above the reported spectrum.

Here is the Aldrich description of the spectrum of spectro grade ethanol

UV absorption λ: 205 nm Amax: 1.00
λ: 210 nm Amax: 0.65
λ: 220 nm Amax: 0.35
λ: 255 nm Amax: 0.04
λ: 300-400 nm Amax: 0.01

The stuff from 300 - 400 nm is 1% of the intensity at the 205 max. That IS a minor bump in the spectum.

[JJM]

Originally Posted by pgwenthold

{snip} The stuff from 300 - 400 nm is 1% of the intensity at the 205 max. That IS a minor bump in the spectum.

That is exactly the point! We went all through this a while back. Roy's spectrum for his standard, pure ethanol has a large, local maximum at 324nm (A=.65), and strong absorbance below that (i.e., towards shorter wavelengths, A>.60). Clearly, Roy's homeopathic preps are different from his "ethanol standard" because his "standard" is not pure ethanol.


Wilsontown- Your observations are useful; however they fit in with my notion, which is that his homeoprep samples are all just USP/NF 95% ethanol (slightly dirty, compared to spectral grade). Thus, one would expect a lot of identical spectra from "nominally" different preps. On the other hand, if they use a collection of various unmatched cuvettes, some with fingerprints, dirty pipettes to transfer samples, different spectrometers, etc.; that could explain the range of spectral traces they recorded.

[pgwenthold]

Originally Posted by JJM

That is exactly the point! We went all through this a while back. Roy's spectrum for his standard, pure ethanol has a large, local maximum at 324nm (A=.65),

That is NOT a "large" local maximum, at least not compared to the other peaks. A relative absorption of .01 is not large by any stretch.

In fact, without actual extinction coefficients, you can't determine how intense the peak actually is. Based on the information from aldrich, though, we know that the region from 300 - 400 is about 1% of the max at 205.

Put this spectrum so that the 205 max is on full scale, and you basically can't see the peak at 324, especially on top of the background in that region.

[JJM]

Originally Posted by pgwenthold

Yada yada yada {snip}

I am out of patience. I can understand a person grasping at straws hoping to be exonerated; but, can someone else explain this to a sophomore? Apparently, I cannot.

[pgwenthold]

Apparently we have different views about what constitutes a "large" peak. I tend to consider peaks that are 1% of the intensity of the maximum to be minor.

I'll wait for you to explain to me how you determine that it is a "large maximum."

Oddly, I've tried to bring the level of this conversation up. I keep wondering about the absorption intensity, but you never address it. What is the cross section for that peak? In general, I usually consider "intense" absorptions to have extinction coeffeficients in the 10s of thousands. If the intensity for ethanol at lambda max is 1e4, then this peak might be 100. That's not very intense at all.

I don't believe any of Roy's work, but your objection, "The peak isn't listed in the Aldrich description of UV spectra" is probably the lamest criticism I've heard. Read what Rolfe wrote about the problems with the spectra, including purification of the solvent, and the issues of scaling are completely mind-boggling (in fact, if you look at the spectra for the "homeo" solutions, it's not just in the 300 - 400 region where they absorb less, they don't absorb the same at 200 nm, either)

In fact, if you go back and read what I wrote at the beginning of the thread, _I_ was the one who brought up the issue of transition that 324 peak refers to.

If I want to know what a detailed spectrum looks like, I am going to have to look at the spectrum itself. I won't rely on someone's description of it, because you never know what they consider sufficient enough to mention.

Since you brought it up, I will mention, I'm pretty sure I have seen a lot more spectra in my life than you have. I look at and interpret spectra pretty much every day, and have been doing it for more than 15 years, so if you think this is a "sophomore" then you are mistaken.

[JJM]

Originally Posted by pgwenthold

{snip} Since you brought it up, I will mention, I'm pretty sure I have seen a lot more spectra in my life than you have. I look at and interpret spectra pretty much every day, and have been doing it for more than 15 years, so if you think this is a "sophomore" then you are mistaken.

Against my better judgment ... I am underwhelmed ... Fifteen years (!) ... even a lab tech should know more about the subject.

I have been a chemist more than 35 years, and I am now retired as a chemistry professor. I doubt you have seen more of anything in your career ... certainly, not spectra. If you have your old spectroscopy books, re-read them; the answer is out there.

[wollery]

Watch out, they'll be comparing the size of their burettes in a minute!

[Cuddles]

Originally Posted by pgwenthold

That is NOT a "large" local maximum, at least not compared to the other peaks. A relative absorption of .01 is not large by any stretch.

Which is the whole point. There is not a large peak in the absorption of ethanol. There is one in the data Roy presents. Therefore he was not using pure ethanol.

[Rolfe]

I'm just signing in here - I haven't read all the new posts, but grateful thanks to Wilsontown for the detailed information on duplications.

Spent all weekend on a friend's yacht, and the weather was unexpectedly great (quote, "if you get home without disseminated melanoma, you'll be lucky"), so although the paper came too it remained in my handbag in the locker.

I think we've way enough ammunition to do a letter, and way enough people prepared to sign I believe. I'm used to scrutineering medical/biological papers for journals, so if you like I'll draft something, once I get caught up.

This isn't about objecting to the publication of woo. This is about clear incongruities and inconsistencies in the data presented, no different from similar incongruities if they were found in a paper proposing something quite uncontentious and mainstream. On that basis, even Peter Fisher, as editor of the journal, ought to take some notice.

Rolfe.

PS. My address is in Britain, so my suggestion is that I write on my own headed paper, with address, then it doesn't matter where any other signatories come from, they can just append names and qualifications alone.

I'm half-tempted to leave off my veterinary qualifications and just include my biochemistry qualifications, to head of the "oh she's only a vet" comments, but probably it won't make any difference.