{10}
By Bob Forrest
There are a lot of things I like about this book, & I wish that it had been on the market 10 years ago when I first became interested in leys. Had I read it then, I think I would have been much more sceptical of the ley hunters’ ‘evolved site’ argument, with its telescoping of history into a scheme quite unacceptable to the orthodox archaeologist. But not being an archaeologist, & with no-one to ‘set me right’, I found it relatively easy to accept this argument. I shall be interested to hear the ley hunters’ response to Williamson & Bellamy on this point, for they must be answered if leys are to retain real credibility.
The following are brief outlines of points I thought were well made in the book. Some of them clearly merit a reply from the ley hunters.
1. (p. 32) The Durham Minster example is a good reply to the Bords’ “why did they do it …?” approach.
2. (p. 59–60) That Borneo tribes today erect megalithic tombs etc without reference to ‘earth energy’ is as worthy of proper consideration by ley hunters as Tony Morrison’s ley-like ceques etc. I note, too, Williamson & Bellamy’s point that (p. 62) Morrison’s lines “do not seem to have brought any marvelous fecundity etc”, and that (p. 60) ‘mysterious’ structures can all too easily be the result of human vanity rather than occult science!
3. (p. 77–8) True, leys rely heavily on churches, and ley hunters do readily accept that pre-reformation implies the incorporation of a former pagan site.
4. (p. 85–6) Some good points here about ‘mark stones’: glacial erratical ploughed-up boulders and cattle-rubbing posts can indeed find their way onto leys – witness the fine example in the photo published in Stonehenge Viewpoint issue 59, p. 7, this being the one mentioned in my piece in Ley Hunter no. 79, p. 17, as having been dumped ‘on a ley’ by the farmer. These, like modern sites which fall on leys (Williamson & Bellamy p. 76), & which are sometimes accepted by ley hunters on a ‘subconscious siting’ argument’ demonstrate nothing more than the power of coincidence. Ley hunters may well have fallen foul of the circular argument by a which a stone or site is accepted as ancient because it is on an alignment, at the same time as accepting that because it is ancient, therefore it demonstrates the deliberateness of the alignment.
As Williamson & Bellamy note on their p. 102ff, these issues are particularly relevant to John Michell’s study of Lands End, as I myself hesitantly (not being an archaeologist) put forward in my “At Lands End” piece in Ley Hunter 79.
I also wholeheartedly agree with Williamson & Bellamy’s comments on John Michell’s Old Stones of Lands End as regards the selectivity inherent in Michell’s stones. That it can, and probably has, led to distortion in the statistical evaluation of the Lands End lines by Gadsby & Hutton-Squire etc, I have shown in my article in Stonehenge Viewpoint no. 55.
5. (p. 92) “Ley lines are alignments of largely medieval & modern features …” so why are there no mentions of ley lines in medieval literature. Good point.
6. (p. 126) Re the spurious nature of the ghost/UFO tie-up, see my article in The News no. 13; also my debate with F.W. Holiday in The News 14 & Fortean Times (formerly The News) 16.
7. (p .128–9) Lethbridge & Underwood both dowsed for all sorts of things, & yet neither of them ever detected ley-lines of force. Good point.
8. (p. 130) Few place names date from before the 5th century ect: good point. Place names are easy to play with – cf. my UlroBob Forrest ran a newsletter called The Ulro Chronicle and Bugle spoof “The Lore of Place Names in & around the County of Herefordshire”.
9. (p. 144) a) Devereux & Thomson turn a miss (prehistoric) into a sort
of hit (medieval). Fair point.
b) Ley W5: good point.
10. (p. 146) A field-work simulation – a good idea, this – I once suggested it myself in a letter in Journal of Geomancy 1.3, but never actually carried it through. I’m pleased to see that it has been done at last.
11. (p. 152f) Landscape Geometry – I agree with Williamson & Bellamy in consigning virtually all LG to the scrap heap, & have not seen a single convincing example of it yet. See my articles in Ley Hunter issues 87 & 88. Incidentally, I think that Michael Behrend has now largely, if not totally, abandoned his “Landscape Geometry of Southern Britain” – undoubtedly the finest example of its kind.
12. (p. 154) Metrology – see my articles in Journal of Geomancy 1.4 & 2.3: it is easy to juggle with metrological units, & to date I personally have seen no really convincing example of this type of study.
13. (p. 162f) Terrestrial Zodiacs: I have no patience for this sport, I’m afraid – hence my Ulro spoof, “The Terrestrial Zodiac of Nowhere-in-the-Dale”. In slightly more serious vein, see my piece in TZ Newsletter no. 8.
14. (p. 169-70) Winnie the Pooh: compare Michael Behrend’s Ulro spoof
“Winnie the Pooh as an Embodiment of the Canon”. Could it be that
there’s more to dear old Win than meets the non-geomantic eye?
My dissatisfaction with Ley Lines in Question comes with Williamson & Bellamy’s poor handling of the statistical issue. There are two main complaints:
1. (p. 95) Williamson & Bellamy seem to think that my statistical evaluations of ley lines involve comparison with what happens on an ‘average’ map containing 200 sites. This is simply not true, as Williamson & Bellamy ought to have realised from note 1 (p. 13–14) of my article in Journal of Geomancy 1.1, an article which they actually refer to in their book! The evaluation of chance expectations always uses the actual number (n) of ley sites on the map, and does not assume an n = 200 ‘average’. To drive the point home, I refer Williamson & Bellamy to my article “The Linear Dream” in Undercurrents no. 18, and to any and all of my ley analyses stored away in TLH archives. I would also put in a special plea that if their book runs to a second edition, then they correct this particular part of it.
The “200 point map” was merely an example devised by Furness to illustrate his formula: nowhere, so far as I know, did he claim this as an ‘average’ map which would act as a yardstick for the chance alignment potential of any and every area of Britain, & indeed, no statistician who knows his stuff would dream of making such an assumption. The Furness Formula (actually probably Menzel’s orthoteny formula – see Flying Saucer Review, July/Aug.1965 p. 26-8) quite clearly operates with definite parameters, one of which is n, the total number of sites on the given map.
The only person I know of who can be accused of making the blunder that Williamson & Bellamy so unjustifiably impute to Furness & myself is Francis Hitching (Earth Magic p. 99). But Hitching is no statistician, as he demonstrates in one error-laden paragraph. He seems to have misunderstood the implications of Furness, as quoted by ScreetonPaul Screeton, Quicksilver Heritage, pp. 56–58, and as a result, it would appear, misled Williamson & Bellamy into tarring us all with the same brush. The situation is much as if I had criticised Mr Williamson’s own work on the basis of someone else’s muddled account of it. I am sure Mr Williamson wouldn’t like that to happen, & nor do I.
Incidentally, the “200 point map” simulation quoted in my Journal of Geomancy 1.1 piece was designed to test the Furness Formula, using the numerical example which Furness himself had chosen. It showed, as I suspected, that there was a flaw in the formula, though that is neither here nor there now, since the Furness formula has been superseded by the more realistic Behrend formula. The simulation was not done to establish a ‘standard map’ by which other maps, containing 300 or 400 or more sites, could be judged. As I stated in my Journal of Geomancy 1.1 piece, the figures for one map cannot usually be applied to any other map, unless, of course, the various parameters for the two maps (one of which is n) agree. No-one has claimed – or should claim – that the 200 site simulation proves anything one way or the other about the 6 & 7 point leys in Devereux & Thomson’s book, yet Williamson and Bellamy set up such an aunt-sally on their p. 95.
2. (p. 97–8) Williamson & Bellamy seem not to appreciate the fact that allowing a ley-width (x in the Furness formula; 2c in Behrend’s) is mathematically equivalent to allowing for the fact that sites are areas rather than points (hence my term ‘random blob’ in Journal of Geomancy 1.1). The Furness formula was not ideal, because it assumed that the line of best fit went through the centres of two of the sites involved in an alignment, even though it was allowed to miss the centres of the other sites. Behrend’s formula, however, is more realistic in that it allows the line to pass through any part of any of the sites involved (equivalently, is allowed to pass within ±c of any of the site centres.) Behrend thus (mathematically) imitates exactly what the ley hunter does with his ruler & map. In other words, ley statisticians are fully aware of, and allow for, the fact that ley sites are small areas and not points.
Williamson & Bellamy can accuse Furness of adopting a ley width which is too low (x = 0.01 miles on ground) for many sites, but they cannot level it at me, for I have always tried to use site-related ley widths (eg on a 1:50,000 OS map I would take 2c in Behrend’s formula to be 1 mm on the map, this being the diameter of most ley-site map symbols like churches & tumuli.) I have also always recognised the poor evidential value of large earthworks (as per the diagram on Williamson & Bellamy p. 100.)
A few other points on statistical issues:
1. (p. 96) I am well aware that in practice ley-sites are not randomly distributed, and that more clustering occurs in practice than a random distribution would allow, though in practice the effects seem not to be as extreme as in the hypothetical example cited by Williamson & Bellamy. However, it should be pointed out that no-one really knows whether not clustering helps the ley hunter to ‘beat the statisticians’. This is one of the things we hope to find out from the mammoth Coldrum simulation, currently in progress, & commissioned by the BBC for their forthcoming ley programme.
2. (p. 96–7) ‘predisposed to alignment’ – good point – eg Kilmartin cemetery? – though I am uneasy that Williamson & Bellamy should dismiss a straight thoroughfare alignment quite so easily: looked at in a certain light, they are saying that ley hunters can’t count these because they are ‘obviously’ deliberately aligned! Of course, I know what they mean, but I think they could have phrased this rather more carefully.
3. (p. 142) Beware of the trap of drawing an ‘arbitrary’ rectangle round an alignment, & then doing a statistical analysis on the basis of what that rectangle contains. You have chosen a rectangle which you know to contain a good alignment, therefore the rectangle isn’t arbitrary, & therefore you will inevitably introduce bias into your evaluation. The only fair, & truly arbitrary, context for statistical analysis is the map area in which the line was first discovered.
4. (p. 159) Williamson & Bellamy’s phrase that I have “recently been looking more kindly on leys” seems to give the impression that they think I am cracking up. I am not. I am still skeptical of most aspects of ley hunting, & I am still convinced that the majority of ley hunting is the search for chance alignments. But I have come across some good statistical results, & it would be quite wrong to sweep these under the carpet just because so much else in ley hunting is faulty. Which brings me to —
5. (p. 196) I am not convinced by Williamson & Bellamy’s arm-waving & hurried sweeping under the carpet of the half-dozen or so statistically good leys that have turned up. None of these lines are as simply explained away as the Oxford ley (Williamson & Bellamy p. 194), I don’t think – indeed, if any of them were that ‘obvious’, I wouldn’t have placed much reliance on them. I could take Williamson & Bellamy to task for their rather desperate appeal to vaguenesses like “the result of the complex interplay of many factors” of which we have “a lack of knowledge”, but I won’t, save to say that, in the absence of any concrete examples, Williamson & Bellamy have postulated something as intangible as ley energy itself. Let’s be honest: Williamson & Bellamy’s “complex interplay” plus “lack of knowledge” means essentially that they can’t really think of a good way of explaining above-chance lines, & they don’t want to admit that the best explanation (whether it be true or not) is the obvious one – namely, that the lines are deliberate. So far as I can see, unless I am doing Williamson & Bellamy an injustice, & they can drum up a good, concrete example of “a complex interplay”, the only likely argument against these favourable statistical results is the Patience Argument (outlined in my article with Paul Devereux in New Scientist vol. 96, p. 826), & perhaps, rather more complex, the effects of non-random clustering on the validity of our tables of expected chance alignment lengths.
Before I (statistically) accept leys with real conviction, I certainly want to see the Patience Argument answered, and I want to know more about the effects of non-random clustering on chance alignment expected frequencies and chance alignment length distributions. In the meantime I am not about to dismiss interesting statistical results on something as flimsy as a complex interplay that we know next to nothing about!
Bob Forrest. Nov. ’84.