Dowsing: Water Witchcraft

It seems as though we might be seeing more and more Skeptoid episodes about water far into the future, as global warming drives our annual weather extremes to crazier and crazier levels. Areas with flooding can expect to see more and worse floods throughout this century, areas with drought can expect to see even less rainfall and drier weather, and people will be driven to more remote limits in search of solutions. One example we're seeing right now is increasing reliance on the paranormal in the search for water, as some in drought stricken areas are turning in desperation to water dowsers. Holding their divining rods in front of them as they pace carefully across your property, these practitioners of the ancient occult ritual have persuaded paying customers — and in many cases, themselves as well — that they are truly channeling some supernatural influence. Today we're going to look into the real success rate and the claims (including by some that it's science based) of the water witches.

Let us first get one thing off the table, the thing that experienced skeptics probably suspected this episode would be all about, and that's the conventional explanation that dowsing is explained by the ideomotor effect. There is a complete Skeptoid episode on the ideomotor effect if you'd like a true deep dive, but the short explanation is that it's our bodies' tendency to make tiny involuntary movements in response to some stimulus, often (but not always) with our being unaware of it. We all do it, as it's a normal body function. Sports fans might make small muscle movements in harmony with the athletes they're watching. Players of the Ouija Board (the ones who aren't cheating and moving the planchette deliberately) might see it creep toward some expected letter on the board. We might tap our feet to music even when we're trying to be quiet audience members. When we're sitting at a table with someone we respect, we might unconsciously mimic their body language or gestures or the way they sit. These are all various manifestations of the ideomotor effect. But the most famous example, given in all the textbooks, is that of dowsing.

Dowsers contend with a multitude of stimuli, all competing to influence those delicately balanced sticks. They're usually walking and so have to counteract that body movement. There might be wind or uncertain footing. But most of all, there are expectations. The dowser knows that anyone else present has expectations of the sticks moving or not, triggering a cacophony of conflicting impulses within the dowser's brain. But mostly they have their own expectations, and these come at many levels. They often have knowledge of where water is and is not found in that area, and experience with how deep wells go there. They have some experience with geography in general that tells them where water is more or less likely located. Sometimes they may even have specific foreknowledge. And, as human beings, they're buffeted by nonspecific intuition. Somehow the cognitive centers in the dowser's frontal lobe has to manage all of these stimuli, and the manifestation of the ideomotor effect is a question of when, not if.

There hasn't been any scientific doubt for many decades that the ideomotor effect is responsible for the entirety of movements made by dowsing rods during a dowsing session. But while that's interesting and educational, it doesn't explain this resurgent tendency of intelligent adults in the 21st century to place so much faith on a practice that's usually presented openly as occult magic.

The other thing you've probably heard about dowsing and that you expected this episode would be all about is that dowsers always fail in controlled tests of their abilities. This is quite true. Dowsers have been put to the test many times, both by credulous experimenters and by skeptical experimenters, and the results are always the same: the better the experimental controls, the worse the results. For some examples, the Wikipedia page for dowsing includes a long list of such tests. Nevertheless, believers in dowsing continue to tout the positive results shown in poorly controlled tests, and/or misrepresent negative results in better controlled tests as positive results.

Perhaps the most famous of these latter instances took place in Munich in the late 1980s and was led by Hans-Dieter Betz, a lifelong believer in divination and a now-retired professor of experimental physics. He has devoted the latter half of his life to his belief that not only people but also objects radiate with mystical energy fields — beyond the knowledge of science — that can be detected using dowsing techniques; thus he was hardly an unbiased researcher for this study. This dowsing experiment was funded by a DM 400,000 grant from the German Federal Ministry of Education and Research, and was preceded by a public statement from the experimenters that dowsing is probably real, and they were setting out to prove it. A two-story barn was used. The dowsers performed their feat up on the second story, while on the ground floor below, a cart holding a pipe of flowing water was placed somewhere along a 10-meter center line, offset to the left or right of that line by a distance determined each time by a random number generator. To the credit of Betz and his colleagues, the experiment was double blinded, and they even had a professional magician on hand with the right skill set to look for opportunities for the dowsers to cheat.

500 dowsers participated in a preliminary round, but their results were hopelessly random and showed no dowsing ability at all — even though blinding was not yet employed in this round, and the dowsers were given feedback with each guess to help them. The experimenters were not deterred. They studied the data and selected 43 dowsers whose results were the best — of course, necessarily, in every set of random data there is going to be 43 who are closest to the mark.

These 43 then underwent 843 individual tests with the full double blinding protocol in place. Again, the results were consistent with random chance, with no patterns of hits at all. But the experimenters were still not deterred, and they did the same thing they did at the end of the previous round: selected the top six, whose random results happened to be closest to correct, for a third round of testing.

Unfortunately, in this third round, all six dowsers produced results that were actually worse than random chance. Later analysis showed that if they'd all simply guessed the pipe was running straight down the middle of the floor (which it hadn't been), all six would have done better.

Nevertheless, the experimenters declared those six to be proof that dowsing works. In their paper, published in the Ministry's journal in 1990, they stated "Dowser phenomena can be regarded as empirically proven."

Funded by a grant from the National Science Foundation, professor of physiology and data analyst Dr. Jim Enright published a response, in which he savaged not the experimental protocol, which was reasonably solid, but this absurdly invalid interpretation of the results. By his analysis of the data, even these best six results were nothing other than empirical disproof of dowsing. His central finding was that the allegedly positive result was purely a Texas Sharpshooter fallacy — imagine someone firing a machine gun randomly at the side of a barn, and then afterwards drawing a target around the three bullet holes that are most closely grouped. It's essentially what the Munich experimenters did. They had no criteria for what would constitute a successful result for any given dowser. Instead, all they did was collect a huge amount of random data, draw a circle around the ones they liked the best, and then characterize those as representative of the phenomenon's accuracy.

This triggered a back-and-forth in the German literature that lasted for years. The experimenters responded to the response, touting the statistical significance of the six results, and Enright countered back again that using only a preferred subset of the total results does not represent those results. Neither party is likely to persuade the other.

To date, the Munich study remains the only large-scale published test of dowsing which any number of legitimate academics has endorsed as successful — even though those academics professed their belief and their intended outcome before the test even began, and then used highly questionable analysis to support their preordained result.

Nevertheless, public interest in dowsing remains, and the global warming caused drought in the western United States — and many other places around the world — drives further interest in outside-the-box methods of finding water. It's to the point that the US Geological Survey actually publishes a pamphlet titled simply Water Dowsing to clear the air. But get this, it was originally published in 1977, and hasn't had to be updated yet. Why not? Because there's nothing new to say on the subject. Dowsing isn't science, it's magical divination; and divination isn't exactly something where new discoveries are published, peer reviewed, and experimentally replicated. So much so, in fact, that this pamphlet refers the reader who wants to learn more to another USGS publication, The Divining Rod, A History of Water Witching, published in 1917!

No doubt, the world is full of reports that contracted water witches successfully found water on someone's property. The USGS pamphlet explains that accordingly:

In a region of adequate rainfall and favorable geology, it is difficult not to drill and find water! Some water exists under the Earth's surface almost everywhere. This explains why many dowsers appear to be successful. To locate ground water accurately, however, as to depth, quantity, and quality, a number of techniques must be used. Hydrologic, geologic, and geophysical knowledge is needed to determine the depths and extent of the different water-bearing strata and the quantity and quality of water found in each.

Combine this with the fact that most professional dowsers work in certain geographic locales where their experience has made them familiar with where wells will generally succeed and how deep water is generally found, and there's no need to introduce any magical elements into those many stimuli discussed earlier for dowsing to appear to work.

And just to be clear, because they are the USGS, they also provide plenty of information and resources on how you can actually locate ground water as a functional, intelligent 21st century adult who is not — for real — investing money and neuron cycles in magic.

It goes without saying that just because no remotely compelling evidence of the effectiveness or reality of dowsing has been found yet, that it won't be found tomorrow. But some would say the allowance for potential future evidence is an equivocation. It's based on literally prehistoric beliefs in the occult and witchcraft, and we don't need a Skeptoid episode to tell us that's part of the human experience with its roots in aspiration, not in reality. If you need water and it's no longer the year 1200, please: call a hydrologist, not a witch.

Correction: An earlier version of this referred to dowsing as necromancy, which is not accurate. It is more properly described as witchcraft. —BD

By Brian Dunning
Follow @BrianDunning