Responding to criticism

Bad Archaeology logo

By and large, it’s something I avoid. Many of the criticisms levelled against either this blog or the main Bad Archaeology website are trivial, vapid or misinformed. I tend to give a short reply to the original comment and move on: there isn’t usually anything substantial in the criticism that warrants revision of the original post or page.

That is especially true of the blog. I view blogs pretty much as opinion pieces, like the editorial in a newspaper. If I want authoritative facts, I’ll go elsewhere. This guides my writing: opinion pieces get posted here, while more factually-based pieces go on the main site, which I hope is used more as a work of reference than the blog. Blogs are effectively entertainment.

A little while ago, I switched the main site over to a content management based system, using WordPress as the software to run it (it’s simple, it does what I need and it is hugely customisable. And did I mention that it’s free?). This allows users to comment on those pages where I permit them to do so (which, actually, is just about every page on the site), in addition to which there’s a contact form. Since doing this, there has been a slow but steady stream of comments.

I am now faced with a quandary, though: most comments are from people who have not registered as users and therefore need approval, as do pingbacks from external websites. By and large, I approve all comments, no matter how ill informed, and the last thing I want to do is to censor dissent, so I will allow through comments that just disagree with what I’ve written (or what James has written). So far, so good. But what happens when I get a pingback from the forum of an extremist organisation? I won’t mention them, as I haven’t approved the two pingbacks from them, but one of their regular users has tried to rubbish what I write on the ridiculous grounds that I’m gay. Because my sexuality doesn’t fall in line with their very narrow definition of what a citizen of their country should be (and it shouldn’t be non-white, Jewish, Moslem, Roman Catholic, gay, socialist… you get the picture), the user thinks that my opinions and my handling of data are worthless. But this ad hominem attack was brought up as a reply to another user of the forum who had linked approvingly to one of the pages on the main site. Do I approve the pingback or do I delete it? I certainly don’t want to send traffic to a hate-filled extremist website.

Then there is a commenter who regularly posts largely incomprehensible statements. I’ve approved all of their comments so far, but it’s getting tedious. Their comments add nothing to the page in question. Do I block the comments on those grounds? I have been hoping that the commenter will eventually get bored and give up, but it’s been going on for some weeks now; it’s not the usual internet troll, out to pick an argument, because the comments are so far off-the-wall that there’s nothing to respond to. I really don’t know the answer.

And finally, there’s a recent criticism of the page devoted to The Turin Shroud. Rather than comment, the person who disagrees with what I wrote, a blogger called Dan Porter, has written an entire blog post, Bad Archaeology at Bad Archaeology (how I wish I could have used that title!). In his comment on Bad Archaeology, he calls it a “comprehensive response”, but it’s far from comprehensive. It cherry picks elements of the page for specific criticisms, but I found that I had to delete only two errors of fact. What Dan Porter has done has been to use the very dubious claims of Ray Rogers that the linen samples used for radiocarbon dating were contaminated, to press on with the silly notion that the image on the Shroud encodes three-dimensional data (an inexplicable miracle!) and generally disagree with what I wrote.

Albert Einstein

A miracle of relativity (praise be!) or a trick of image manipulation?

What his criticism did allow me to do was to test the claims about the encoding of three-dimensional data in images. I took a well known facial image and processed it with results that look fairly similar to those obtained from the Shroud. It even rendered unevennesses in the photographic print as three-dimensional! Another miracle!

In the end, I want to be reasonable. I really don’t want to upset people, but if they have wrong, silly or downright objectionable beliefs, I can’t stand by passively and let them persist. If they have raised valid objections, I’ll happily change what I’ve written: that’s how real science works. If they are wittering on incomprehensibly, I’ll tolerate them. If, on the other hand, they want to push an ideology that is hate filled, I feel that I must ignore them and not give them the dubious benefit of a link from my site.


Dowsing in archaeology (part 2)

Bad Archaeology logo

The bank where the west wall fo the Cheese Warehouse was located

The bank where the west wall of the Cheese Warehouse was located

I’ve described my direct interaction with dowsing in a previous post. The semi-serious hunt for the eighteenth-century Cheese Warehouse on the bank of the River Dee in Chester yielded equivocal results: we “identified” a rectangular “anomaly” that most of us agreed upon. The problem was that the one wall we identified that lay on a line predicted by dowsing also lay under a bank that may have been a visual prompt for the responses we got. What was surprising was that I had not believed that the wall of the warehouse lay so close to the river bank. The subsequent location of the wall in this trench (actually a metre or so east of the line indicated by dowsing and also east of the flat top of the bank, which carries a footpath) occasioned some surprise and was seen by some of the team of volunteers (not least the member who has brought the equipment to site) as a confirmation of the reality of the technique

Now, I don’t think that it’s unnecessarily cynical of me to suggest that something other than the detection of some buried drystone masonry by means of dowsing was going on here. We had dug a linear trench at right angles to the line of the western wall of the Cheese Warehouse–wherever that may have lain–and would have hit it at some point along its line. The fact that we did has more to do with what we already knew about the location of the building from historic maps than from any use of bent coathangers swivelling in empty ballpoint pen tubes. Of course, it was difficult to persuade the rest of the team that they had not necessarily been witness to a confirmation of the reality of archaeological dowsing. It didn’t seem to matter that dowsing failed to locate the south-eastern corner of the Warehouse (the trench dug over the suggested position turned out to be well inside it) or that the wall we did locate was off the line by around a metre. No, the willingness to believe outweighed the evidence of excavation. I’m not suggesting that the team of diggers from the Chester Archaeological Society was especially credulous; no, they were simply prone to the usual human fallibility of confirmation bias.

Stapleton’s Field henge and the involvement of a well known dowser

Moving on ten years, I had changed jobs. The Cheese Warehouse was a distant (and still, to my shame, unpublished) memory and I had returned to the part of the world where I grew up: North Hertfordshire. A local group of enthusiasts – Norton Community Archaeology Group – had been formed in 2007 to investigate the heritage of one of the three historic parishes that make up Letchworth Garden City. I was asked to provide the Group with a wish-list of ten sites I considered worth investigating. One of them was actually a landscape that appeared to consist of a series of Bronze Age monuments in a field known as Stapleton’s Field (a recent name: its historic name seems to be unknown). They included a group of ring ditches (all that is left after round barrows have been ploughed flat), a possible trackway, an enclosure and a series of probable field ditches. This was exciting, as it appeared to be a nearly complete landscape from around 4000 years ago. The more I looked at aerial photographs and a geophysical survey of one of the ring ditches, described as a “double ring ditch” by the Historic Environment Record for this specific monument, the more convinced I became that something was wrong with the description. Rather than an unusually complex burial mound, I thought it looked like a henge.


Your humble author and his brother, doing something that can no longer be done at Stonehenge (it was a long time ago and I was very young!)

A henge isn’t necessarily what you might think it to be. On hearing the word, most people think immediately of Stonehenge, a unique monument that is one of the most instantly recognisable sites anywhere in the world. There is a henge at Stonehenge, but it’s not the stones: it consists of the circular bank and internal ditch that forms the defining edge of the monument. To an archaeologist, this is what makes a henge. While some may contain stone circles, the majority do not; in some, the stone circles are a secondary addition. The henge I suspected might exist in Stapleton’s Field is one of those that did not have a stone circle, largely because the local chalk bedrock is quite unsuitable for use in megalithic construction.

To test my ideas, we dug a trench across the centre of the monument in 2010 as well as two others across anomalies seen in the geophysical survey that I thought belonged to the Bronze Age landscape I had hypothesised. The enclosure turned out to be Romano-British, as did the field ditches; the potential henge turned out to be Neolithic, although we did not find conclusive evidence for its interpretation as such. However, there was enough to go public with the idea that it was likely to be a henge, as we had found Grooved Ware pottery in the centre of the monument.

It was shortly after we had a number of stories on the radio, television, the press and the Group’s blog that the Chairman was approached by Paul Daw, a dowser who has made a study principally of stone circles, but who also has an interest in Neolithic monuments including henges and causewayed enclosures. He said that he had dowsed the site and could outline the henge; he was also willing to give demonstrations of the technique to the Group and to teach members how to dowse for themselves. He gave a talk to the Group on 16 March 2011, which I attended, followed by a practical session on 4 June, which I did not.

The talk given by Paul Daw was curious. He showed a lot of slides of scanned newspaper articles about his discoveries as well as some plans of the results of his dowsing. He focused largely on East Anglia (he is based in Cambridge) and on the Neolithic, particularly causewayed enclosures. However, at no point did he present any evidence that the sites he had discovered by dowsing had been confirmed by other techniques. In particular, there was no presentation of data derived from excavation. Allowing for his dowsing to have discovered buried anomalies, we were given only his assurance that they were of Neolithic date. I did not find this good enough to convince me and there were certainly others in the audience who felt the same way. Despite what readers may think, I had gone to the talk with an open mind and was prepared to be convinced. As an exercise in presenting a case for the reality of the technique, the talk was a failure.

Changes in the perception of dowsing

Changes in the perception of dowsing following training

I am less able to comment on the practical session, as I did not attend it. However, one of those who did managed to collect information about how the participants in the exercise perceived dowsing. They were asked before the fieldwork to rate how strongly they believed that dowsing could detect buried archaeological anomalies on a scale of 0 to 10, with 0 for complete disbelief and 10 for complete belief. They were then asked again, after the fieldwork, using the same scale. In every case, the perception of dowsing improved after taking part. The average pre-fieldwork response was a rating of 2; after the fieldwork, it had risen to 8. This is an impressive improvement. Although no plans were made of the anomalies detected, blue flags were left in the ground to mark the positions of what was dowsed. There were two main groups of flags: a circular area, supposed to correspond with the position of the henge, and a linear ‘anomaly’ that I was told was very strong. The flags were still there when we began the 2011 season of excavation on 27 July.

A flag marking a dowsed 'anomaly'

A flag marking a dowsed ‘anomaly’ (arrowed)

This is where I can vouch for what was dowsed. There was a circle of flags in roughly the right place, although it was perhaps five metres too far to the south-east: it looked as if it had been put there by someone who knew roughly where the monument was located and roughly how big it was but not the precise location or size. This may be an unfair judgement on my part. However, when we opened up the trenches, it became even less clear what the flags were supposed to be marking: was it the inner ditch, the chalk bank or the outer ditch? The circle of flags corresponded with none of the archaeological features we excavated. Of course, one could always argue that as we haven’t yet excavated down to bedrock, the dowsing has detected a first phase that has not yet shown up. This would be special pleading and is not supported by the results of the geophysical survey or aerial photography.

The linear anomaly was even less convincing as an archaeological anomaly. It lined up perfectly on the tower of Baldock Church, just 910 m away to the east-south-east. For this reason, somebody suggested that it was a ley line. Well, ley lines don’t exist, despite the intuitive certainties of New Agers, so we can rule out that explanation! One thing that I did wonder was whether or not a twentieth-century ditch located in the 2010 excavation might have been the basis for this anomaly. The alignment was right, although its position was once again wrong, being about 15 m off the line of the archaeological feature. Nevertheless, nothing in the excavation corresponded with this anomaly.

If we treat the excavation as a test of the reliability of the dowsing, then the dowsing definitely failed. One of the real issues over the results that were obtained is that they were obtained with foreknowledge of what exists in this part of the field. I first published a plan of my suggested interpretation of the site as a henge in 2009 and there has been a page on the Group’s website giving details since February 2009. This means that anyone has access to information about the site, should they choose to seek it out; it is also the case that everyone who attended the dowsing session on 4 June had seen the site under excavation and had participated in the first season of work there. I could, uncharitably, argue that the dowsing was little more than a test of the memories of those taking part: the circular shape of the monument is known from a variety of sources, while the twentieth-century ditch, which ran roughly parallel with the footpath crossing the field, may have provided the “inspiration” for the linear anomaly. The details of the monument, which only became clear following the 2011 season on the site, were not picked up by the dowsing that preceded it. I wonder why.

The St Michael (Ley) Line

Places supposed to lie on the St Michael (Ley) Line

As a postscript to the dowsing of Stapleton’s Field henge, I was informed that it lies on the St Michael’s (Ley) Line, a notorious line said to run from St Michael’s Mount in Cornwall through to Hopton on the coast of Norfolk. It is supposed to be marked by a large number of churches dedicated to St Michael; it is also said to run through Royston Cave. Unfortunately, it doesn’t pass through any part of Stapleton’s Field, nor does it pass through Royston Cave, missing them both by about 3 km. This isn’t minor quibbling: 3 km is a long way off a line that’s supposed to be dead straight and accurate across hundreds of kilometres. Either sites that are supposed to provide evidence for it (such as Royston Cave) lie exactly on the line or they lie off it and must be discounted as evidence: you simply can’t have it both ways!

Dowsing as a technique

As the reader will have gathered by now, I am far from impressed by my encounters with dowsing on archaeological sites. On two occasions, I have seen it used in an attempt to locate archaeological sites whose existence was already known and on both those occasions, it failed to locate the sites with any accuracy. I may have been unlucky; I may have gained an accurate impression. But there is one more instance of a site I know that has been dowsed for information that I have deliberately held back from describing. It’s another site investigated by the Norton Community Archaeology Group, this time in 2007.

This attempt to dowse a site was very different. Based around the earthworks of part of the village that had been deserted during the Middle Ages, the dowser used a pendulum in an attempt to locate structures and date their abandonment. He also pointed to the locations of human burials, again supposed to be of medieval date. At the same time, a soil resistivity survey of the site was carried out. Although the geophysics was inconclusive, the dowser pointed to a number of buildings and graves and gave the dates (to the nearest year) of their demolition or burial.

This is the sort of technique that Tom Lethbridge believed could be used to identify different materials, date sites and even recognise abstractions. It is a long way from the use of a hazel twig or bent coathangers to locate buried anomalies, however they might be detected. Instead, the dowser has a more mystical role, tapping into data that simply cannot be encoded in a purely physical form. This is the realm of ‘subtle energies’ of which conventional science is ignorant. This sort of thing is removed from scientific testing: the basic principles on which it supposed to rely involve things that defy measurement. The nature of this type of dowsing is what Robert Sheaffer has described as a “jealous phenomenon”: one that disappears before conclusive evidence for its existence can be gathered. The phenomenon does not manifest itself, so the believers’ argument goes, in the presence of sceptics. This is the very essence of pseudoscientific thinking.

For this reason, I have no truck with the use of pendula on maps. There is nothing that can be tested. However, I am more open to the idea that dowsing might have some basis in reality. Might it be possible that the dowser is sensitive to gravitational or magnetic gradients in the landscape, such as might be produced by holes in the ground? Some dowsers have claimed that this is how the phenomenon works. That being the case, dowsers relying on changes in the background magnetism should be able to detect hearths, kilns, fired clay and ironwork; indeed, the magnetic signals should be so strong that they could swamp other signals. Yet dowsers generally seem to ignore such things. Why, if magnetism is the source of signals being picked up by the dowser, would the effects of these highly magnetic materials remain hidden? Then, if the dowser relies on an ability to recognise changes in the gravitational background, there ought to be a correlation between the size of a buried feature and the prominence given to in the results of dowsing. At Stapleton’s Field, the outer ditch of the henge – which geophysics indicates is at least 3.5 m wide – ought to be the most prominent ‘anomaly’ to be recognised in dowsing. Why, then, were the strongest responses received from a linear ‘anomaly’ that aligned on the (perfectly visible) tower of Baldock church yet did not have a buried correlate?

You can see where I’m going with this. Suggest a mechanism known to science that might explain how dowsers can get the results they claim, and there will always be something that doesn’t fit. If dowsers wish to explain the phenomenon using forces known to science, they then need to explain how the individual dowser can select from among the responses received to locate only those things that the person using the dowser wishes to find. Once they start to invoke forces unknown to science, we are in the realm of pseudoscience. The Bullshit Historian has done an extensive analysis of dowsing claims in archaeology and finds them wanting. So do I.

2011 in review

The stats helper monkeys prepared a 2011 annual report for this blog.

Here’s an excerpt:

The Louvre Museum has 8.5 million visitors per year. This blog was viewed about 90,000 times in 2011. If it were an exhibit at the Louvre Museum, it would take about 4 days for that many people to see it.

Click here to see the complete report.

Sticks, wires and pendula: dowsing in archaeology

Bad Archaeology logo


Dowsing for an eighteenth-century warehouse: Phil Miles in Chester, May 2000

This is another of those posts I’ve been meaning to write for some time without knowing quite where to start. I’ve been given a kick start by a twitterer (Marcus Smith) and by a recent BritArch announcement of a dowsing ‘experiment’ (you must be a subscriber of BritArch to see the link!). The problem is that it’s such a huge topic, it’s difficult to know where to begin.

Personal experience

I’ll start of with a bit of personal history. In my early teens, I devoured every book in Letchworth library on Egyptology, followed by every book on archaeology in general. I was also interested in fringe archaeology, even then, so I was drawn to the Dewey Decimal System’s notorious class 001.9, ‘controversies’. It’s basically a catch-all for busy librarians who don’t know where to put pseudoscience but who know better than to lump it in with the genuine science it mimics. So, I would turn left on entering the library, past the card indices of authors, titles and subjects (and, in those days, it really was a physical card index in wooden drawers), and into the realms of Isis Unveiled, Chariots of the Gods and the like.

Thomas Charles Lethbridge (1901-1971)

Thomas Charles Lethbridge (3 March 1901 - 30 September 1971)

On one of those visits, a whole raft of books by T C Lethbridge had appeared. I was vaguely aware of the name, as he was an archaeologist who had excavated at a well-known Romano-British cemetery site in Guilden Morden, close to my home. These books – Ghost and ghoul (1961), Witches: investigating an ancient religion (1962), Ghost and divining rod (1963) and ESP: beyond time and distance (1965) – were about various matters on the occult side of things. Like many teenagers, I found the occult fascinating: perhaps there was secret knowledge that the Establishment was either unaware of or was keeping from the rest of us. This sort of fascination, I now understand, is all part of growing up, of learning how to be an individual, of discovering that there are no real authorities to whom we can turn for the answer to everything.

Yet here was a proper archaeologist discussing such matters. Some of it made intuitive sense to me and I was swept up in the rest of it. One of Lethbridge’s great discoveries was that he could dowse using a pendulum. Like a proper scientist, he conducted experiments. He found that the pendulum reacted to different materials if he varied the length of its string; he then found that as well as materials, the pendulum could be made to react to concepts (such as male, female, age and so on). He discovered that he could use the pendulum to determine the precise age of something (a site, an artefact, anything that he wanted to date). He also found – and this is where my teenage credulity was stretched to its limits – that it was possible to dowse for archaeological sites using a map: one did not have to travel to places to discover new sites (and, of course, date them with a precision that radiocarbon will never achieve).

Of course, I tried the technique. It didn’t work. I rationalised this (can one really rationalise the irrational?) as “it doesn’t work for me“: it clearly worked for Tom Lethbridge. After all, he was a retired archaeologist and I was determined that I would become one, too. Archaeologists were people I admired and trusted. The curator of my local museum, John Moss-Eccardt, was an archaeologist who ran evening classes and a museum club that I attended. Archaeologists were serious people who knew their stuff, so Tom Lethbridge just had to be on to something.

The view from the mainstream

Professional archaeologists have always been a bit ambivalent about dowsing. Here’s the entry from Warwick Bray and David Trump’s (1970) Dictionary of archaeology:

dowsing A technique for discovering buried features or materials by the use of a Y-shaped hazel wands, bimetal strip or the like. The scientific principle behind it is not understood and indeed by many people its validity, at least for archaeological prospecting, is doubtful.

Scientific principle“? Do the authors believe it or not? There seems to be some fence-sitting going on here. A little over a decade later, Lesley Adkins and Roy Adkins’s A thesaurus of British archaeology (1982) has this to say:

Dowsing is the same procedure as water-divining and can be used to located buried archaeological features. The success or failure of the method depends on the talent and skill of the dowser, who usually uses some form of simple instrument such as a Y-shaped piece of wood which is held in the hands and whose movements indicate the position of the features as the dowser walks over them. Once the position of a site has been located, it can also be surveyed by dowsing. A grid is laid out over which the dowser walks, so that the results of the survey can be plotted on to a scale plan of the area.

So, it works then? It’s all to do with the “talent and skill of the dowser“, which is an assertion that my teenage self was happy to contemplate. What about more recent texts? Colin Renfrew and Paul Bahn’s Archaeology: theories, methods and practice is currently the most popular undergraduate textbook of archaeology: they must say something about dowsing. And indeed they do:

In concluding this section on subsurface detection, we may refer in passing to a controversial technique that has a few followers. Dowsing (in the U.S. witching) – the location of subsurface features by holding out a twig, copper rod, coathanger, pendulum, or some such instrument and waiting for it to move – has been applied to archaeological problems for at least 50 years, but without being taken seriously by most archaeologists.

Aha! A note of scepticism. The writers go on to describe an experiment carried out in Northumberland, in which sceptical archaeologists took part in the survey of buried church foundations and were convinced by its results. Some of the walls detected by dowsing were located, while other predicted walls were not located. Another experiment described by Renfrew and Bahn involved an attempt to locate a Romano-British pottery kiln, for which the results obtained by dowsing did not match those obtained by magnetometry: they do not say whether excavation was carried out to test the two techniques. Nevertheless, they conclude:

For the moment… until overwhelming proof of the validity of dowsing and other unconventional methods is forthcoming, archaeologists should continue to put their faith in the ever-growing number of tried-and-tested scientific techniques for obtaining data about site layout without excavation.

The greatest authority on archaeology for students nevertheless leaves open the possibility that “overwhelming proof” of dowsing’s ability to detect archaeological remains may one day be found.

More personal experiences

The Port of Chester

The Cheese Warehouse, from an eighteenth-century painting of the Port of Chester

In 1999, I worked with members of the Chester Archaeological Society on a project to locate a warehouse on the banks of the River Dee that had been used during the eighteenth century to store Cheshire cheese before being shipped to London, which was becoming an important market for the product. The warehouse burned down in the nineteenth century and no trace of it survives on the site today; it had been the inspiration for Lewis Carroll’s grinning Cheshire Cat, which is said to have described the happy cats who sat on the quayside of the Cheese Warehouse, waiting for the mice who, attracted from the ships by the smell of cheese, would run up the tethering ropes of the ships into the cats’ ready paws.

We began the work by undertaking a resistivity survey in February/March 1999; this was inconclusive, but showed up some foundations that we could identify as part of a house known as Copfield House, which is first seen on a map of 1874 and which was demolished c 1979×81. A photograph of the back of the house taken in the 1960s bore a slight resemblance to a depiction of the Cheese Warehouse on an eighteenth-century painting of the port. Excavation in May of the same year revealed the foundations of the nineteenth-century house but nothing that could be recognised as part of the Cheese Warehouse. We returned to the site in 2000 to carry out more excavations in different areas, still with little success. One of the society members is a keen dowser, so he got out his rods and everyone soon began dowsing. Several people – myself included – got reactions on a bank on the western edge of the site, so we decided to dig there. Very soon, a stone wall was located. Another trench placed over an area where dowsing had suggested another wall should be located failed to locate it; it did, however, locate some small stone pillars on which the wooden joists for the raised floor of the warehouse were held.

The south-west wall of the Cheese Warehouse

The south-west wall of Chester's Cheese Warehouse: not quite where dowsing had located it

So, what did the dowsing achieve? In my opinion, nothing. The one trench where a length of wall was found ran across a bank (which had apparently not existed in 1972), where the dowsing suggested that there ought to be a wall. I suspect that the bank misled us into identifying it as something piled up over the demolished warehouse wall. In fact, the top of the bank did not coincide with the position of the wall; nor was it on the line suggested by dowsing. The other walls “located” in this way proved not to exist.

To be concluded

I am currently working on a fascinating site with the Norton Community Archaeology Group, where I have identified what appears to be an early (‘formative’) henge of Neolithic date. The group has worked with a well known dowser on the site, and the next instalment of this blog will deal with their discoveries.