Water Divining

[naomi24]

I’ve just been listening to ‘The Bottom Line’ on Radio 4 where the guests were the vice president of CH2M Hill, the CEO of Veolia Water, and the CEO of Anglian Water, who all said that water diviners are used within their respective industries. One said if he hadn’t seen it with his own eyes, he would never have believed it works. Listen to the last few minutes of the programme from about 27.14.

http://www.bbc.co.uk/programmes/b036w3b6

Your thoughts?

[naomi24]

Jim, Sagan said it – but if you don’t like that, how does Stephen Hawking grab you?

“The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge.”

Perhaps you should think on that.

//No realistic experiment can jump to the end….//

But the one you propose doesn’t even begin at the beginning! If you want to conduct a valid experiment, you must replicate the claimed procedure in the same conditions. Simulation of any part of it is not good enough. I confess my experience of scientific study amounts to only 5 years each of physics, chemistry, and biology, all three subjects being abandoned before A-Level in pursuit of other things, but I’m pretty sure that if you ask any scientist worth his salt he'll agree with me.

[jomifl]

Gness, is there anywhere in Ireland without water?

[gness]

Do you mean mains supplied water as we have here, Jom?

[jim360]

You can quote as many scientists as you like, Naomi, but if you are using what they say in a way that contradicts a basic mathematical law, then you are using what they say in the wrong way. Anyway, that wasn't Stephen Hawking, but Daniel Boorstin (first appeared in "The Discovers", 1983).

Let us take what 2sp etc says to be true. Actually, I haven't really said that she wasn't telling the truth. Then, if I replicate her husband's actions in the same conditions, I might well end up finding water using this method -- if not on the first try, then eventually, after years of practice. Ultimately however, this will tell us nothing new. Dowsing "works", somehow, it seems, for a few people. The issue is how. An experiment that doesn't explore how isn't worth conducting. There are a number of possible explanations, some more reasonable than others -- but an experiment has to be able to distinguish between those explanations.

[naomi24]

Jim, I think you’ll find that Boorstin said "The greatest obstacle to discovering the shape of the earth, the continents, and the oceans was not ignorance but the illusion of knowledge." Hawking took that an important step further – and what he said contradicts no mathematical law – not that I’ve said anything that does that either - so I really don’t know why you’re banging on about ‘mathematical law’ at all.

//An experiment that doesn't explore how isn't worth conducting.//

I disagree. An experiment that explores ‘If’ is worth conducting. Once it is established that the subject in question is not a delusion or a trick, the time then comes to explore ‘how?

[jim360]

The Sagan quote you keep using does, in the circumstances I described, contradict a mathematical law -- that of Bayes' Theorem. Or, if you prefer something simpler, there is a logical equivalence between the two statements "If A, then B", and "If not B, then not A". I think this applies to this case, as follows:

All dowsers seem to claim that when their abilities are tested, they will be successful. Hence, we can say that if dowsing works, then it will be seen to work. This is our statement that A implies B -- and confirmed not by the scientists but by the dowsers. However, in the experiments I have mentioned earlier, that took place in conditions that were agreed to by the dowsers being tested, dowsing was not seen to work. Therefore, using the rule above, dowsing does not work.

This is of course far too dogmatic a statement, but then you can fall back on Bayes' Theorem to turn it into its equivalent in probability, which is that:

-- If you have searched for evidence of some proposition, and;
-- failed to find it, and;
-- if there is reason to believe that you would have found that evidence if the proposition were true, then;
-- The probability that the proposition is true decreases.

Therefore, we have that "absence of evidence can be evidence of absence". Not proof, of course, but evidence nonetheless.

"//An experiment that doesn't explore how isn't worth conducting.//

I disagree. An experiment that explores ‘If’ is worth conducting. Once it is established that the subject in question is not a delusion or a trick, the time then comes to explore ‘how?"

I think this is a fair point, actually, but for the fact that the experiments that explore "if" have already been conducted, and keep finding the answer "no". So I suppose I properly meant that such an experiment is not worth me conducting -- because it already has been, numerous times, with the same results. Apologies for that mistake.

[naomi24]

Jim, //Therefore, we have that "absence of evidence can be evidence of absence". Not proof, of course, but evidence nonetheless. //

You appear to be climbing up your own bottom and getting lost on the way.

[jim360]

?

[Old_Geezer]

I think it depends on interpretation of meaning. Lack of evidence, after a darned good search, provides support for the conjecture that there may be no evidence, but I'd be uncomfortable stating that advances one's knowledge of whether evidence actually exists. Even given the "not proof" qualifier.

I think if the disagreement is now concentrating on the trivial maybe the main discussion is having issues ?

[naomi24]

Jim, read your posts. You’ve gone from saying if I take Sagan's words at face value then he was wrong, to I’ve misinterpreted a simple quote of just ten words, to mathematical law, and now to ‘Therefore, we have that "absence of evidence can be evidence of absence". Not proof, of course, but evidence nonetheless.’ - which appears to contradict what you’ve been saying all along.

Frankly, you’ve lost me.

OG, //I think if the disagreement is now concentrating on the trivial maybe the main discussion is having issues ?//

I agree.

[jim360]

Those posts are all meant to say the same thing, even if they don't.

See here for someone else's attempt to explain it:

http://rationalwiki.org/wiki/Absence_of_evidence

[naomi24]

Jim, //This is because we are significantly more likely to not see evidence for a hypothesis when it is false than to not see it when it's true //

If we don't know precisely what we're looking for, we may not recognise any evidence that may be there, and therefore in cases of 'trial and error', that doesn't apply.

[jim360]

Thanks for reading it -- now it's just a matter of interpretation.

But the dowsers who were tested in 1948 and 2004, and several times in between, were the ones who told us what we were looking for: success in finding where the water was, every time. That they did not surely says something?

[naomi24]

Jim, we know what the ultimate goal is – what we don’t know is how it works– and why it doesn’t work every time. It seems to me that when it works (if it works) a combination of factors must be involved – one or more of which is missing when it doesn’t work. I’m guessing of course – but I’m pretty sure the dowsers don’t know how it works either – if they did they wouldn’t agree to undergo a test that excludes one or more essential element.

Do you know precisely how the tests were set up and what was involved?

[jim360]

The 1948 test is described briefly in a book, I don't have a copy of it. Another link goes to a page in German. Do you speak German? I'm afraid I don't, so I have to rely on Google translate and a bit of educated guessing, and with that in mind I think it's better to present to you the raw link rather than my interpretation of it:

http://www.gwup.org/infos/nachrichten/566-keine-million-dollar-fuer-psi-faehigkeiten -- the 2004 test -- links at bottom of page seem to be broken.

The other sources are a book with ISBN code 978-1-57392-979-0, "Pseudoscience and the Paranormal", and see also:

http://undeceivingourselves.org/S-divi.htm

which cites several studies. Anything I say would merely be repeating the words of that site. Incidentally, the first line is worth quoting:

"There is no doubt that water diviners do successfully find water, but success could be due to factors other than divining ability."

I hope that I have stuck to this line, or words to that effect, and if/ when I did not then again I apologise.

[vetuste_ennemi]

Naomi, Jim has won this argument. Please concede graciously.

[naomi24]

V-E, Oh, come on now. Jim has won this argument? I think not. From flat denial, I have him giving me a link that concedes "There is no doubt that water diviners do successfully find water, but success could be due to factors other than divining ability." I think it’s case proven. It does work – not always – but it does work. Oh ye of little faith. Give up? Not on your nelly! And apart from that, don’t be a spoilsport. I’m enjoying this. ;o)

Jim, love the bit about the wooden leg. :o)

[pixie373]

V-e, Jim won the case by the 4th post, but somehow, i don't think that's the point ;-)

[naomi24]

pixie, what do you think the point is?

[jim360]

If you had also read that full page, you would have seen that water divining "works" no matter, and indeed worse according to some studies, than random chance, or indeed worse than more scientific surveying methods. Dowsers can find water. That doesn't mean that dowsing works in the way dowsers say it does.

Some quotes:

"A subsequent check of the records showed that the geologists' success rate was 1 in 3 but the diviners' success rate was only 1 in 12."

"All were confident of success. Overall the results were slightly worse than expected by chance. The most successful diviner (14 hits) was a beginner."

"Of 1832 divined sites, 70.4% yielded ample usable water and 14.7% yielded no water. Which may seem like convincing support for divining. But 1858 undivined sites performed even better -- 83.9% yielded ample usable water and only 7.4% yielded no water. It was not reported whether the divined and undivined sites were equally difficult. [It is reasonable to assume that on average there was no difference in "difficulty" between the two, given such a large sample size - but this is my comment only]."

If this is evidence that dowsing works... well, so does randomly shoving a pipe down into the ground without even looking, some of the time.