This site is not maintained. Click here for the new website of Richard Dawkins.

← The trouble with homeopathy

The trouble with homeopathy - Comments

Duff's Avatar Comment 1 by Duff

The Prince, when he hears of this suggestion, will become "unavailable for comment". There is no way any homeopath is going to submit this money printing fraud to a valid test.

Wed, 24 Feb 2010 11:45:00 UTC | #443705

InYourFaceNewYorker's Avatar Comment 2 by InYourFaceNewYorker

Homeopathy. Mind-body therapy. Call it what you will, it's the same candy in a different wrapper, with each wrapper given a mystical-boneheaded-sounding name.

Wed, 24 Feb 2010 12:05:00 UTC | #443708

SaintStephen's Avatar Comment 3 by SaintStephen


We can worry about molecular memories and things like that afterwards, if the experiment gives a positive result.
In homeopathic hangover remedies, the water molecules actually arrange themselves into tiny Bloody Marys, complete with celery sticks and cocktail napkins.

This was actually a very funny piece by the Good Professor. And educational!

Wed, 24 Feb 2010 12:16:00 UTC | #443709

pberry's Avatar Comment 4 by pberry

Richard, in the article you mentioned: "How would you set about administering a dummy, or placebo needle-prick."

This very experiment has been tried, with specially-created dummy needles whose point telescopes up into the shaft. It's detailed in "Trick or Treatment" by Simon Singh and Edzard Ernst.

Needles(s) to say, the results were inconclusive.

Wed, 24 Feb 2010 12:32:00 UTC | #443712

emactan's Avatar Comment 5 by emactan

"in the case of acupuncture it is difficult to imagine how patients can be shielded from knowing whether they are in the experimental or control group"

That's actually the easy part and has been done by simply blindfolding the subjects (study by Lixing Lao). The participants were questioned thereafter and it seems that the fake acupuncture had been fairly successful in tricking the participants. The hard part is blinding those who are administering the sham procedure. They would obviously know what they're doing.

Wed, 24 Feb 2010 12:34:00 UTC | #443713

Newswede's Avatar Comment 6 by Newswede

It's far too easy for the experimenters to cheat. All they have to do is use a real medicine for the "homeopathic" doses and omit it for the control doses, and then NOT dilute the doses very much.

Wed, 24 Feb 2010 12:37:00 UTC | #443714

Pyrion's Avatar Comment 7 by Pyrion

How about an open letter to any homeopathy society or journal with Richards suggestion? I wonder how they would rationalize not to do such an experiment. It's very improbable that they would do it.

Wed, 24 Feb 2010 12:41:00 UTC | #443716

emactan's Avatar Comment 8 by emactan

If money is no object why not add a second control group, one that is given pure water that has not been succussed?

Wed, 24 Feb 2010 12:44:00 UTC | #443717

Richard Dawkins's Avatar Comment 9 by Richard Dawkins

It's far too easy for the experimenters to cheat. All they have to do is use a real medicine for the "homeopathic" doses and omit it for the control doses, and then NOT dilute the doses very much.
Obviously any experiment, in any branch of science, is invalidated if the experimenters deliberately cheat. It goes without saying that, in a contentious experiment like this, policing against deliberate cheating would have to be more than usually scrupulous. The point of double blind trials is to guard against not deliberate cheating but inadvertent self-deception. Inadvertent self-deception is what lies behind almost all the alleged 'evidence' in favour of 'alternative' therapies, and other dubious procedures such as dowsing and telepathy. That is what is ruled out by a good double blind design. Deliberate cheating obviously has to be ruled out by other means, in any experiment.


Wed, 24 Feb 2010 12:45:00 UTC | #443718

Jumped Up Chimpanzee's Avatar Comment 10 by Jumped Up Chimpanzee

Never mind that we have drunk a molecule of water that has passed through the bladder of Oliver Cromwell, we have probably all drunk a molecule of water that has passed through someone who has already taken a homeopathic remedy. In fact, we've probably all already had every homeopathic treatment ever dreamt up! I wondered why I felt so great!

One thing I don't understand, though, about alternative medicines is how do they qualify as "alternative", so that they are not required to go through the very rigorous and expensive tests that "conventional" medicines have to go through. If both forms of medicine are available on the NHS, what determines the difference?

Wed, 24 Feb 2010 12:49:00 UTC | #443720

katja z's Avatar Comment 11 by katja z

I wish I could write to the British MPs! In my country homeopathy is not officially recognized, but the homeopaths' pet argument is "it is part of the NHS in Britain, why not here?" (Surely, surely, the British can't have got it wrong, can they? ;) I'm guessing it's the same in at least some other European countries.

It's far too easy for the experimenters to cheat.
Don't tell them that! ;)

Wed, 24 Feb 2010 12:53:00 UTC | #443721

Hamilton Jacobi's Avatar Comment 12 by Hamilton Jacobi

This experiment was already done some years ago on BBC's Horizon program. The result was a severe embarrassment for the homeopaths.

Wed, 24 Feb 2010 12:57:00 UTC | #443722

jinmane's Avatar Comment 13 by jinmane

I'm probably being really dumb here, but I don't quite understand why the control samples need to be shaken. Surely if, due to the placebo effect, both groups report some positive effect, then that may be attributed to the shaking. However, if you use plain unprepared tap water, any positive result in the control group could be confidently dismissed as placebo. Then only if there's a greater positive result in the experimental group than in the control group would you consider that there may be something in the homeopathic method that works, although you wouldn't know what, but that wouldn't matter because we're trying to show that it doesn't work at all.

Wed, 24 Feb 2010 13:00:00 UTC | #443724

pipsy's Avatar Comment 15 by pipsy

The homeopath gets money for nothing.

The patient gets nothing for money.

I don't buy it.

Wed, 24 Feb 2010 13:16:00 UTC | #443726

Mark Jones's Avatar Comment 14 by Mark Jones

Excellent! I've written to Mancis Fraud, er, Francis Maude, my MP.

Wed, 24 Feb 2010 13:16:00 UTC | #443725

hungarianelephant's Avatar Comment 16 by hungarianelephant

But until homeopathy is demonstrated to work (which it almost certainly will never be) it should not be supported by the NHS.

I agree, but the problem is that this experiment won't either demonstrate that homeopathy works or disprove it in the sense that matters for this question. It will show that the effect of homeopathy (if any) is or is not placebo, which is not the same question at all.

Even some of the adherents of homeopathy say that it's likely that some positive effects come from an extended consultation. Most people turning up at the GP's surgery don't have very much organically wrong with them, and a nice chat might well help them to feel better. Since the average GP consultation is 12 minutes, up from 8 minutes, and since there seems to be an in-built expectation that a prescription will be written in the course of that, and since most patients have already been there fairly recently, there's definitely some scope for some kind of alternative approach. Homeopaths are a lot cheaper than GPs. It just might be cost-effective.

But what is essential is that there is a level playing field. It should be up to the homeopaths to show that they have a method of treatment which is effective and cost-effective, not for robust science to be constantly pushing down bubbles under the wallpaper as the claims subtly shift.

This could not be done (fully) with a placebo-controlled double blind trial, but it could be done with a large study in which a larger number of patients are randomised into homeopathy group and conventional medicine group. (It would be possible to have placebo controls within those groups, which would make for a more interesting study.) The symptoms of the various groups could be assessed at the end of the study, together with data on repeat visits, in order to work out an average cost and average effect for the two groups.

Then we'd have a basis for deciding whether the NHS should fund. At the moment, we don't.

Wed, 24 Feb 2010 13:24:00 UTC | #443727

Mark Jones's Avatar Comment 17 by Mark Jones

Comment #463521 by Hamilton Jacobi

I think you'll find that experiment is invalidated by having James Randi nearby. He has that effect, don't you know? :-)

Wed, 24 Feb 2010 13:26:00 UTC | #443728

ennui's Avatar Comment 18 by ennui

How does homeopathy work

Wed, 24 Feb 2010 13:29:00 UTC | #443729

SaganTheCat's Avatar Comment 19 by SaganTheCat

I will eat my shirt.

please, not the hawaiian one!

Comment #13

as I understand the experiment, it is to draw a line under the claim of "water memory".

of course the meta-analysis of existing trial results tells us all we need to know

Wed, 24 Feb 2010 13:30:00 UTC | #443730

Jumped Up Chimpanzee's Avatar Comment 20 by Jumped Up Chimpanzee

jinmane asks:

"I'm probably being really dumb here, but I don't quite understand why the control samples need to be shaken."

I understand it is standard in any kind of double-blind test as RD suggests here, that both groups must be handled in exactly the same way. Those taking part in the various stages on the experiment will not be aware which is the control group (that's why it is double-blind). The object is to eliminate any possibility that those running the experiment may give away which group is which, even subconsciously by body language or voice intonation.

Wed, 24 Feb 2010 13:34:00 UTC | #443731

Truth_seeker's Avatar Comment 21 by Truth_seeker

Why is homeopathy on the NHS? Did it get there accidentally, or was it purposely put on it?

1. Does it work for some people?

2. Why is it that something that does not work have any success for anybody?

3. If we take it off the NHS does that stop them from setting up privately owned practises?

4. If people pay for something that does not work can the governement collect taxes on it.

5. If the government wants to collect taxes on it then does that mean they have to allow it to survive?

6. Is that how it came unto the NHS in the fist place?

Wed, 24 Feb 2010 13:38:00 UTC | #443732

Peacebeuponme's Avatar Comment 22 by Peacebeuponme


there's definitely some scope for some kind of alternative approach.
I don't see this from what you have said at all.

There are two elements to a homeopathic consult:

1) Genuine beneficial intervention, such as the style of the consult itself, the nice chat, the power of the placebo effect etc; and
2) Unscientific mumbo jumbo, being the remedy itself and the theory behind it.

Conventional medicine should seek to extract 1) from the whole and use it ethically and ditch 2). There would be no need to label this an 'alternative' approach, but it would retain the cost-effectiveness and efficiacy criteria you mention.

Your suggested double-blind trial would not be fair unless it appropriately distinguished the effects of 1) from the non-effects of 2).

Wed, 24 Feb 2010 13:40:00 UTC | #443734

Luke_B's Avatar Comment 23 by Luke_B

I'm probably being really dumb here, but I don't quite understand why the control samples need to be shaken.

As I understand it part of this is so that the samples are treated in exactly the same way so as to remove any other unknown but possible factors and also so that those administering the treatment are unaware as to whether it's the control water or the homeopathic water.

Wed, 24 Feb 2010 13:40:00 UTC | #443735

Tyler Durden's Avatar Comment 24 by Tyler Durden

13. Comment #463523 by jinmane

I'm probably being really dumb here, but I don't quite understand why the control samples need to be shaken.
All conditions in the experiment must be replicated across the board to rule out bias, and to isolate exactly what effect the Independent Variable (IV) is having (or not) on the Dependent Variable(s) (DV).

In Richard's example above, the IV = the homeopathic remedy itself; the DV = the participant(s) in both groups.

The variable which is being manipulated by the researcher is therefore called the Independent Variable and the Dependent Variable is the change in behaviour measured by the researcher.

In a True Experiment, we are testing the deliberate manipulation of one variable, while trying to keep all other variables constant. If not, we cannot know for sure why the change, if any, occurs. Homeopathic remedies are shaken x times as part of their "effectiveness", so must the control substance.

We're not trying to show "that it doesn't work at all", we're looking for no difference between the two groups.

Wed, 24 Feb 2010 13:41:00 UTC | #443736

JonathanWest's Avatar Comment 25 by JonathanWest


Tests such as you have described have already been done, innumerable times. They have even been done sufficiently often that the Cochrane Collaboration has been able to perform a meta-analysis of the trial results. Unsurprisingly, it shows that homeopathy works no better than placebo.

Ben Goldacre has described all this in the homeopathy chapter of his book Bad Science. That chapter is in turn based on a Guardian article he wrote in 2007. Highly recommended reading.

Wed, 24 Feb 2010 13:50:00 UTC | #443738

Tyler Durden's Avatar Comment 26 by Tyler Durden

18. Comment #463528 by ennui

Superb! :)

Wed, 24 Feb 2010 14:06:00 UTC | #443739

hungarianelephant's Avatar Comment 27 by hungarianelephant

22. Comment #463533 by Peacebeuponme

Conventional medicine should seek to extract 1) from the whole and use it ethically and ditch 2). There would be no need to label this an 'alternative' approach, but it would retain the cost-effectiveness and efficiacy criteria you mention.

You are right. I was using "alternative" in the sense of an alternative to what we currently do, rather than in the sense of woo. Poor choice of word, sorry. If anything different which came from homoeopathy were adopted by conventional medicine, it would cease to be "alternative".
Your suggested double-blind trial would not be fair unless it appropriately distinguished the effects of 1) from the non-effects of 2).

My first suggested trial would not distinguish between the two (and would not be double blind), but it would establish whether homeopathy taken as a whole is effective and cost-effective. That's "fair", at least in the sense that it is the same test which has to be applied to conventional treatments in order to get funding. I agree that the optimum would be to add a double-blind procedure for the homeopathy group, in line with Richard's suggestion, as this would then distinguish between the two groups and work out which parts are effective / cost-effective.

Wed, 24 Feb 2010 14:07:00 UTC | #443740

spellingmastake's Avatar Comment 28 by spellingmastake

Ok then i have to get this off my chest.

My experience with homeopathy: I have a friend who is really into it. If you have something wrong with you he will be the first to say you should try homeopathy.

About a year ago i was getting symptoms of an overactive thyroid. I felt hungry all the time no matter how much i ate, i sweated too much, i was shaking badly. I finally went to the doctors and he tested me for an overactive thyroid.

While i was waiting for the test results back my friend talked me into trying the remedy for overactive thyroid. I thought i may as well. The remedy was called Iodum.

I wasn't really expecting anything to happen as i had tried homeopathy for other things including hay fever without any luck. I didn't even really know what homeopathy was. I just thought it was like herbal medicine.

Within seconds after putting a single drop of Iodum on my tongue my symptoms improved drastically. The swelling i could feel in my neck just went. And the constant hunger stopped. It was instant. Then i just left it for a couple of days and I improved even more.

Then it was time to go back to the doctors. The doctor said my tests had come back positive for a overactive thyroid and he wanted to begin treatment right away. I told him about the homeopathy, he acted in the way that you would imagine a dr would act when someone tells him he is putting off proper treatment for homeopathy.

anyway i will cut it short as this is getting too long. He ended up testing me again and my results had improved. I no longer had an overactive thyroid.

So now sometimes it comes back. When i feel the symptoms come back i just take another drop. In my eyes homeopathy MUST work.

So when i read articles by scientists saying it can't work or it's absurd, I just think well it does work so you must be missing something. Something pretty damn significant.

So why hasn't it been proved to work yet? I have some theories.

1. Maybe most of the remedies are crap and it's only a very small percent that can actually help specific conditions. So the theory of homeopathy is correct but over the years they were too keen to add more and more remedies. I think i've heard of one remedy that is just water that is left in the light of Venus. This is obviously crap. And i've never understood how Onion could work for treating runny eyes. I mean you get runny eyes when you sniff an onion not eat it.

2. When homeopaths conduct the tests they often come back in favour of homeopathy. Scientists claim this is because they are being biased even if it is unintentionally biased. I think the same thing is happening when scientists test it. When a scientist is working on something that he wants to prove right if the results are inconclusive he will keep at it in hope he will eventually find a way to get the results he expects. Only after a lot of trying and failing will he give up on an idea. If the scientist is testing homeopathy it's like one test, the results were inconclusive so it must be crap.

3. If someone were to do your experiment and see how many people improved it might not be that high. Because there are a lot of remedies it may take them 10 times to get the right remedy for the condition. Not to mention how everyone responds differently to different potencies. I doubt a scientific experiment would allow the homeopath 10 times to get it right. They would have one chance if they don't get it right homeopathy doesn't work.

As i have had luck with Iodum for overactive thyroid i would love to see that tested. Although the problem is overactive thyroid is dangerous so it would be unethical.

Anyway i don't' mind articles like this one. As it explains why you think homeopathy doesn't work in a sensible way. It's the childish ones that mock the people who use it that bother me.

With all this talk of stopping the NHS from using homeopathy and stopping boots from selling it, it makes me worry that one day you will be trying to get it banned completely. Then all the people who benefit from it will suffer. Remember even if you think we are all deluded and it's all in our heads we would still suffer. Because if someone thinks homeopathy helps them beat depression then it helps them beat depression. You take that away from them then the depression will come back.

ps if anyone is thinking of starting the argument that the person suffering from depression should be receiving proper medical care. Just remember that medical care would involve giving them SSRI drugs that have side effects including causing self harm or even suicide.

Wed, 24 Feb 2010 14:23:00 UTC | #443744

the PC apeman's Avatar Comment 29 by the PC apeman

Thank you for taking the time and effort to respond to this truly important issue. What you choose to leave unaddressed speaks volumes as well.


Wed, 24 Feb 2010 14:26:00 UTC | #443745

Witticism's Avatar Comment 30 by Witticism

What you choose to leave unaddressed speaks volumes as well.

Its the little bits of nothing that really make it something.

Kind of like Homeopathy itself really.

Wed, 24 Feb 2010 14:30:00 UTC | #443747