Crop circles are nowadays generally considered to be made by human pranksters or artists. There are still some people who think they are made by other means than planks and ropes. In the nineties, there were even some people who researched this phenomenon scientifically. Or at least they tried. Many theories which were proposed assume the involvement of some kind of electromagnetic radiation during the circles’ formation. Some were even published in a peer-reviewed scientific journal. Although criticised before, a fresh look provides a new argument why this idea should have been dismissed straight away.
tl;dr the upper nodes of plants in a field do not lie in a horizontal plane, therefore the BOL model for crop circles is flawed.
In 1999 Eltjo Haselhoff, a Dutch physicist with a PhD, proposed an improved model for this, which implies that ‘balls of light’ hovering above the fields might be the source. He published some articles on this matter. The most recent one is titled “An Experimental Study for Reproduction of Biological Anomalies Reported in the Hoeven 1999 Crop Circle”( Journal of Scientific Exploration.Vol. 28 No 1. 2014). The following quote from the introduction gives a good summary what this is all about:
In the summer of 1999, Dutchman Robbert van den Broeke reported that he saw a luminescent sphere hovering above a farm field while a crop circle was apparently forming underneath (Haselhoff 2001a, 1999; http://www. robbertvandenbroeke.com). This happened in the village of Hoeven, The Netherlands, and since then the Hoeven 1999 circle has become a famous and controversial case in crop circle history. It is famous because biophysical studies of plants sampled from the circle, performed independently by researchers Eltjo Haselhoff and William Levengood, revealed biological anomalies (Haselhoff 1999, Levengood 2001). These anomalies varied over the circle’s area, with a symmetry similar to the radiation intensity distribution of an electromagnetic point source. These findings enticed crop circle “believers,” who could finally reference a scientific argument that “crop circles were made by balls of light,” and infuriated crop circle skeptics, who stated that the research methods applied by Haselhoff and Levengood were flawed and that their findings had natural explanations.
Van den Broeke is an infamous Dutch medium who ‘discovers’ almost all crop circles in The Netherlands and he is also known for his fake spirit photography. He sees in the research by Haselhoff proof that ‘his’ crop circles are a genuine phenomenon and not something that one or two guys could make during a night just to get some attention from ‘believers’.
The whole issue centres around node lengthening, which is a phenomenon that occurs in the stems of certain grain varieties. You can see the nodes (pulvini) as sort of a joint, which allows the plant to grow in a slightly different direction than it was doing. Now, these nodes will lengthen under certain conditions, for instance when the stem is bent towards the ground. What Levengood found is that this lengthening was more pronounced in the flattened crops in crop circles that some consider genuine than in crop circles that are known to be made by humans. In fact, Levengood thought this was a good test to discern the real crop circles from those made by hoaxers.
Levengood also proposed a cause for this difference, namely that in the creation of the genuine crop circles an electromagnetic field had altered the nodes in such a way that in the following days the difference would become apparent in a different pace of elongation. In 1999 Levengood managed to get an article on this theory published in a peer-reviewed journal, Physiologa Plantarum, together with Nancy Talbott. A lot can be said about the scientific credentials of Levengood and Talbott and their work, both of which have shown to be quite dubious, but you can read information on that elsewhere.
Haselhoff was at that time chairman of the Dutch Centre for Crop Circle Studies and in 2000 he wrote a comment to the article in the same journal, it was published in 2001. In this comment, he proposes his BOL model which should be considered a better explanation for the found node lengthening (Levengood thougth of a wave or vortex). BOL stands for the ‘balls of light’, which have been mentioned to be present at the time of creation of the crop circles by some people who claim to have witnessed that.
Haselhoff suggests that if you consider the origin of the electromagnetic radiation to be a point source hanging above the field, there is a very high correlation between the node lengthening and the amount of radiation that would reach that node, considering that the intensity of the radiation diminishes with the square root of the distance from that point source. This is because the energy from a point source is dispersed equally in all directions in 3D-space. Therefore you can say the energy of ‘a single moment’ is spread over the surface of a sphere, and the surface of a sphere is proportional to the square of its radius). You can also read Haselhoff’s explanation in his book The Deepening Complexity of Crop Circles (2001), the relevant pages (78-81) are available in the Google Books preview.
To find out how much radiation hits a node in the field from the point source in the BOL model, you’ll need several numbers. The distance to the centre of the circle (‘d’ in the figure left) is easily obtained. However, the heigth ‘h’ is not known, as well as a proportionality constant which depends on factors as absorption of the radiation under the specific atmospheric conditions at the time of the formation of the crop circle. But if you have several samples from nodes you can use mathematical techniques to find the best fit for these variables.
According to Haselhoff the fit to a spherical distribution is very good in crop circles which are considered genuine by ‘crop circle experts’ and far less in known hoaxes. From the conclusion of his comment:
The experimental data published in Levengood and Talbott (1999) suggest that pulvinus length expansion in crop circles is a thermo-mechanic effect, possibly induced by a kind of electromagnetic point source. Data obtained from a simple hand-made formation did not reveal the same characteristics. By no means does the author pretend to present a ‘lithmus test’ for distinction between a ‘genuine’ crop formation, whatever it may be, and a hand-flattened area of crop.
Although here he refrains from claiming that his BOL model can be used to find out whether a crop circle is genuine or man-made, in other writings he shows fewer constraints.
Hoeven crop circle, perfect fit
In his article in Physiologa Plantarum, Haselhoff used the measurements of crop circles which were given by Levengood and Talbott. In his book and later article in the Journal of Scientific Exploration (JSE) he gives the example of a crop circle he sampled himself. This is a crop circle ‘discovered’ by Robbert van den Broeke. The sampling looks pretty systematically as can be seen in this figure:
Also the way Haselhoff determined how much the nodes lengthened looks quite meticulously. He used a computer program which enabled him to measure the node lengths from photographs of the samples which were made under the same conditions. These measurements have to be done with accuracy because it is all about differences of millimetres (!).
The result of all this work must have pleased Haselhof as they almost perfectly fit his BOL model. Look for instance at sampling line B. Note that a node length of 100% means that these nodes on average show elongations similar to those of control samples some distance from the crop circle.
Here you can see that in the centre of the circle the node lengthening is the highest and that the effect diminishes towards the edge of the circle until it can not be distinguished from the lengthening of the control samples.
The 2014 JSE article describes an attempt to refute the BOL theory by making a crop circle by hand (using planks and ropes) under conditions which were as similar as possible to those under which the Hoeven 1999 circle formed. Almost needless to say that in this replication no anomalous node lengthening distribution was found and that therefore the Hoeven 1999 findings still stand as an unexplained anomaly.
A couple of Italian skeptics from CICAP, Francesco Grassi, Claudio Cocheo and Paolo Russo, were not that impressed by the BOL model. They discussed it with Haselhoff via e-mail, asked him for his raw data and performed their own analysis. This can all be found on a dedicated section on the CICAP website. They tried to get their commentary article published in Physiologa Plantarum as well, but the editor didn’t want to continue the discussion while admitting that he regretted that the papers of Levengood, Talbott and Haselhoff had been published. In 2005 they got their article published in the Journal of Scientific Exploration.
The Italians argue that the BOL model is not nearly as good as Haselhoff suggests on several grounds. The main argument is on Haselhoff’s statistical analysis. For the examples he gives in the article, other three-dimensional distributions would fit equally well as the sphere which is implied by the BOL model. This is not surprising as only a few samples were considered in those calculations and this will never give strong evidence for a model in which several variables need to be determined. Haselhoff has contested the conclusions of the Italians, you can go into the details yourself, but the statistical arguments against the BOL model would not hold quite so good when you look at the Hoeven 1999 crop circle anyway, I think.
BOL model lacks physical realism
Apart from their main criticism on the BOL model, which is purely mathematical, Grassi et al. also point out other problems. For instance, the stems may shield radiation coming from the point source so that stems in their shadow will receive less energy than the BOL model predicts:
From a physical point of view it should be pointed out that the BOL model is not realistic. A hypothetical BOL model should be much more complex, because the striking energy will depend on the incidence angle of the radiation on the stem nodes and the energy absorption will depend on the path length of the radiation inside the plants and therefore on their actual transparency. A nontransparent stem partially shields the node, so Haselhoff’s model is only valid if we assume that the plants are completely transparent to the striking radiation and so could not absorb energy at all.
While this is a valid point (and they mention more), I think there is a far more obvious physical problem with the BOL model, which is somehow missed completely by Haselhoff and his Italian critics.
The nodes are not in the same horizontal plane!
When you look at the pictures of the BOL model used by Haselhoff and the calculations involved we see that he assumes that the influence of the radiation works via points which lie in the same horizontal plane. You can say that he ‘fixes’ the average node length of a bundle of sampled stems from a spot in the crop circle on that plane. Does he implicitly assume that all these nodes are on the same plane? Probably he just forgot to take into account that it matters on which height the nodes were. But by doing so, he does away with a whole dimension!
But why is this important? As mentioned before, the differences in node lengthening measure in millimetres, but you can easily check that the differences in the vertical position of the nodes in a field (either before the flattening or afterwards) measure in centimetres. These differences overshadow the minimal differences in node lengthening completely. Even if you would like to fix the model by assuming that the influence happens via the point where the stems stick out the ground, you will find that these positions differ more in height than a few millimetres.
OK, you do have to take into account that the node lengthening can be twice as high in the centre of a crop circle if we take Haselhoff’s measurements for granted, and the fact that the difference in height is not likely to be that huge. But even if that would only be as much as 20% it would induce extra variance which the model should account for. I do not think we would see a significant fit if we could redo the analysis with the actual heights of the nodes properly taken into account, but it is not something we can check as this information was not gathered in the sampling.
You could argue that I have not actually shown that the nodes are not (approximately) in the same horizontal plane in areas where genuine crop circles are formed, but I think that it would be an even more mysterious phenomenon than a difference in node lengthening if those nodes would be found to lie in the same plane. Anyway, it is obvious that Haselhoff (as well as the reviewers of the journals and the CICAP members) did not take the vertical position of the nodes into account, which makes the whole idea look baseless.
Of course, this is not an explanation for the apparent anomalies itself, but it shows that BOL model, which gives sort of a physical explanation, does simply not suffice. And the only other support for the ‘balls of light’ was already pretty shady, because of the reputation of Van den Broeke, the only person Haselhoff brings forward to have seen that these alleged balls were involved in the formation of the crop circles.
It is also fair to mention the criticism of Rob Nanninga, editor in chief of Skepter, the magazine of the Dutch Skeptics Foundation. Nanninga pointed out that the measurements of the node lengths should ideally have been done blindly, without the person doing the measurements knowing where the stems come from (i.e. a ‘genuine’ crop circle, a known hoax, or another control area). Haselhoff did not see this as a huge problem for his working method. I think he is simply wrong on this issue, it is one of the most likely ways bias could have influenced the results.
Also, it is unclear why Haselhoff in his 2014 article in JSE doesn’t mention that he had sampled two crop circles in Hoeven in 1999, which formed separately within a couple of days. The other circle did not show the anomalous node lengthening, but leaving this out in his book and JSE article is cherry picking.
Proper sciencific conduct vs overselling results
In the discussion with CICAP Haselhoff repeatedly states that he is just presenting a hypothesis and seems to think that his article is not treated fairly because it is just a comment to the articles by Levengood and Talbott:
Anyone who reads my paper will agree that this was a mere comment to the work of the BLT team, suggesting some model adaptations and carrying ahead their hypotheses with a modified version, only to stimulate further study. In my opinion the style of Grassi’s comment, as well as the propaganda related to it that he currently carries out over the internet and beyond is way out of proportion, and casts a dark shadow over his true intentions.
To me, it looks like that Haselhoff is applying double standards here. When his ideas are cracked based on scientific arguments he is downplaying how significant he thinks his model is, or whether he thinks the proof is solid. But have a look at the end of this fragment of a documentary on crop circles (which features the Oliver Castle crop circle hoax video):
And it turns out, that the node lengthening as measured in the field, corresponds perfectly to the distribution of a small electromagnetic source hanging above the field.
Scientifically seen this statement is nonsense or at least a pretty unfair representation of what Haselhoff had shown in his research. Even taking into account that at the time of recording the criticism of CICAP probably had not yet been published. That the CICAP article was published must have been a setback for Haselhoff, because in his book The Deepening Complexity of Crop Circles (2001) he wrote the following about his comment in Physiologia Plantarum:
I find this a hilarious view on how science works. If a hypothesis is shown to be false it obviously shouldn’t matter where it is published. Stating that you will only consider such proof when it is published in a journal, shows that it is not the truth that is your main interest, but that only the ‘scientific status’ matters to you. This is just a blog and definitely not peer-reviewed, so I don’t think that we can expect to see a reaction from Haselhoff on my remarks that the BOL model is flawed because of its basic assumptions, aside from the other issues with sampling and statistics, but he is welcome in the commentary section.
It is even worse
When I first read some of Haselhoff’s articles on this matter, the question had come up whether he performed his regression analysis correctly. The problem with finding out whether he messed this up as well, was that the data of the individual stems was not given and the 2014 JSE article is not very clear on how Haselhoff performed the regression analysis on the ‘best’ example, the Hoeven 1999 crop circle. From the report Haselhoff published online in 1999 about these cases, we can, however, learn a few things.
To me, it is now clear that Haselhoff performed the regression separately for each of the ‘traces’ of sampling spots (A, B, and C). For each trace, he calculates the best fit, but the heights do not match up! So based on set A Haselhoff finds a best fit with height 3.1m, for set B it is 4.1m and for set C 6.6m. Of course, this is silly, the point source can only be in one place. This was also noted by Russo in his analysis of the Hoeven case (I don’t quite understand why he did not give this more attention in his article, as it is a direct kill for the approach of Haselhoff).
In his 2014 JSE article, which is the first ‘peer-reviewed’ article in which the Hoeven case is presented, Haselhoff mentions about the results he found: “The best correlation was found for a value of h = 4.1 meters, with a Pearson correlation coefficient R = 0.99 for the B-trace”. This is true, but this omitting of values for the traces A and C, and failing to mention that the different heights found this way pose a huge problem for the BOL model, makes me wonder whether Haselhoff was just deliberately setting up a smoke screen, or simply didn’t know what he was doing. I doubt that he missed the remarks Russo made, but maybe Haselhoff is just consistent in ignoring all comments that have not been published in peer-reviewed journals …
Also, the ‘perfect’ correlation of the node lengthening in trace B to the values predicted by the BOL model is not as good as his ‘R=0.99’ suggests. Haselhoff has used the average node lengths from each sample as input for the regression instead of the individual stems. This can be done, but then you have to take into account that these are averages coming from a sample with a variance. I’m pretty sure that Haselhoff has taken the averages as single values (the CICAP skeptics thought so too, it appears) This highly overestimates the goodness of fit as the variance per sample is quite high as can be seen in the screenshot of the program Haselhoff used for measuring the nodes lengths.
What a mess!
PS To test the BOL model correctly you would have to run the regression on all individual nodes sampled in the circle together, using the correct distance ‘r’ to the alleged point source and of course taking into account the vertical position as wel.
The most interesting articles discussing the BOL model can be found on the CICAP website, but I found that the index page does not link in a very helpful way to all the articles, so I decide to list them here and also give the links to the other articles that play a role.
- Balls Of Light: A scientific demonstration for “genuine” crop circles? Grassi et al. (October 2003)
- Balls of lights (2): The Questionable Science of Crop Circles. Reaction on the article by Francesco Grassi et al. Haselhoff.(June 2005)
- Balls of lights (3): The Questionable Science of Crop Circles. Grassi et al. reply to the internet rapid reaction by Eltjo Haselhoff on the JSE article. Grassi et al.(July 2006)
- Balls of Light (4): The Questionable Science of Crop Circles. Eltjo Haselhoff writes a letter to the editor of the Journal of Scientific Exploration; Grassi et al. reply (October 2007)
- Balls of Light (5): The Questionable Science of Crop Circles. Eltjo Haselhoff anticipates our exchange in the Journal of Scientific Exploration. Haselhoff. (August 2007)
- Balls of Light (6): The Questionable Science of Crop Circles. Reply to Eltjo Haselhoff’s open letter. Grassi et al. (March 2008)
- Balls of Light at Hoeven? Russo (October 2005, June 2006)
- Balls of Light at Hoeven? (2) Reaction to Paolo Russo’s article. Haselhoff (July 2007)
- Balls of Light at Hoeven? (3) Russo replies to Haselhoff’s open letter. Russo (March 2008)
Journal of Scientific Exploration
- Grassi et al. Balls of Light: The Questionable Science of Crop Circles (Vol. 19, No. 2. 2005) [pdf]
- JSE editor Henry H. Bauer – Haselhoff Responds to ‘‘Balls of Light: The Questionable Science of Crop Circles’’ – Grassi, Cocheo, and Russo’s Reply (Vol. 21, No.3. 2007) [pdf]
- Haselhoff et al. An Experimental Study for Reproduction of Biological Anomalies Reported in the Hoeven 1999 Crop Circle (Vol. 28, No. 1. 2014) [pdf] [alternative]
- Levengood. Anatomical anomalies in crop formation plants. (1994) [from BLT-research]
- Levengood and Talbott. Dispersion Of Energies In Worldwide Crop Formations. (1999) [from BLT-research]
- Haselhoff. Opinions and comments on Levengood WC, Talbott NP (1999). Dispersion of energies in worldwide crop formations. (2001) [pdf]
- Haselhoff. Report on Pulvinus Length Measurements. Hoeven Crop Circles (1999)
- Haselhoff. The Deepening Complexity of Crop Circles (2001). Preview via Google Books.
(title image by Daz Smith | Flickr)