It might be ‘fringe’, but I doubt it will ever truly die…
Forgive me if I appear to beating a dead horse, but I feel vindicated in posting this last instalment by this recent article. It contains the usual anecdotes, and absolutely nothing to suggest that the same old mix of ideomotor effect and educated guesswork isn’t also at work here. See the JREF forum for some discussion. Suffice to say that there is just no evidence to support the dowser’s claims to be able to precisely locate grave cuts.
Anyway, I promised in Part II (Part I is here) to dissect the one readily-accessible ‘study’ of archaeological dowsing cited in the recent pro-dowsing article in ‘Time and Mind’ journal. So here we go…
I tend to draw a parallel with another deluded group of people that think they can divine special information – ‘psychics’. In this case it seems to me that, like a psychic with semi-conscious prior knowledge of a client, the dowsers here knew roughly what types (and sizes) of structure they were looking for, and approximately where they ought to be. They also started from visible features above ground. Despite these advantages, the results remain far from impressive. The final dowsing survey, by the author’s own admission, varies considerably from the true layout of the structures. How is it possible to divine roughly the right shape, yet have it be metres out of place? Do the rods need calibrating?
If you’re trying to make sense of the plot yourself, you should note all the many instances in which red lines seem to overlie existing features on the map – this is because surface features not discovered via dowsing are for some reason still included in the ‘dowsing’ overlay, along with other features known archaeologically – like the cellar and tower base – also not surveyed by dowsing. For comparison of dowsing results with reality, this basically leaves the chapel – which is nevertheless known to exist on the site. Further, it is a type of structure well known (archaeologically and colloquially) to appear as a simple rectangular structure aligned east-west, often with an apse at the east end. Nor can we by any means discount prior knowledge of this particular structre. The approximate size of the building appears in the historical record, and the excavations that uncovered the chapel were done many years beforehand, qualifying the claim that the dowser’s “..work all came before the excavations..”, which must refer to the most recent round of digging. The site of the building is even marked on Ordnance Survey maps, including a Victorian map that the author shows in his own report. Even without this information to hand to refer to, the surface features surrounding the area really don’t give much room for maneouvre, making the chances of placing this east-west oriented rectangle of known size somewhat accurately into this constrained area actually fairly high. Despite all of these advantages, by the author’s own admission the accuracy of the divined features ‘is poor’. Rather than the technique perhaps being at fault, this is blamed upon the subsequent mapping process. As for the ‘correct disposition’ claimed – we’ve already seen how the chapel plot might resemble sub-surface remains – but as no dimensions are given for the dowsed plot, we can’t even be sure that compare the level of precision in the plotting of the walls. Finally, the chapel itself contains a number of gravecuts, as one would expect – but are they really that close a match for the excavated graves? There are so many that hits are almost inevitable – and why the discrepancies (e.g. no tightly-spaced second row of graves, some out of place, others not even plotted)? And what of the distinctive apse plotted by the dowser? Bzzzt, wrong. There’s a pile of debris that could once have been an apse, but why would the dowser show up an intact apse rather than a pile of rubble?
I should stress that there is no question here of an intent to deceive – most of this information is presented by the author of the report himself. In fact, some of the worst failures in accuracy, or even missing some pretty substantial features entirely, are included in the report (and further undermine it). The author is simply placing too much weight on the results that can be rationalised as hits – presumably because of wishful thinking. But these hits are highly subjective in the interpretation. For example – how close a match do you think this culvert plot actually is? What of the features revealed by geophysics that don’t line up? Once again – pointing out ‘hits’ whilst ignoring the misses, also known as ‘observational selection‘ just like a psychic reading or an observation that the full moon affects violent behaviour.
The same report also includes a ‘test survey‘ that one might hope would be a controlled, blinded scientific study, but actually amounts to a subjective survey of the experience of dowsing, with no quantification or qualification of any successes had. The ‘positive results’ amount to an aggregated and near-random scatter of points – and again a good deal of faith is required to match any clusters of points with real features. If you were to draw a new plot of imaginary features and overlay that on top of the same scatter, you would still find ‘correlations’. It is nothing more than a ‘join the dots‘ game. And at the risk of poisoning the well somewhat, despite the citing of this study by the Time and Mind article, note the inclusion of Test 5, intended to “locate their own body field which is normally at about 450mm”. There is no evidence whatever for (and a few laws of physics against) the existence of such an ‘aura‘. Hard science, this ain’t.
As an aside however, I would like to give full credit to the author of that report for debunking what is a myth even within dowsing circles – that dowsing is depicted in the cave paintings at Tassili nAjjer in the Sahara. It isn’t – the figures are patently archers.