The Pendulum of Truth

In Physics by Brian Koberlein17 Comments

In an earlier post I talked about the Titius-Bode relation, and how it had been applied to exoplanetary systems with questionable success. It’s an example of trying something just to see if it works. But sometimes an experiment produces results that are strange and unexplainable. It’s even more frustrating when further results are frustratingly inconclusive. Take, for example, the phenomenon known as the Allais effect.

Allais' initial pendulum results. Credit: Enterprise Mission

Allais’ initial pendulum results. Credit: Enterprise Mission

Maurice Allais was an economist who was awarded the Nobel Prize in Economics in 1988. He also had an interest in physics, particularly some alternative approaches to gravity and electromagnetism. In 1954 he performed an experiment during a total solar eclipse, where he measured the rate at which a Foucault pendulum shifted over time. According to known physics, there shouldn’t be any effect at all, but Allais observed an abrupt change in the precession rate during the height of the eclipse.It came to be known as the Allais effect, or Allais anomaly. Given the unexpectedness of the result, the obvious weak point is to look at Allais’ methodology. After all, he wasn’t an experimentalist, so perhaps the result isn’t valid. But tracking the precession of a pendulum isn’t particularly difficult. Allais repeated the experiment in 1959 and again saw an anomalous result.

The results were interesting enough that others researchers have made similar experiments in the 1960s and 1970s. Some had an anomalous result, and some didn’t. Variations of the experiment have been done with a torsion pendulum (basically a horizontal bar suspended on a wire), and some verified the result and others didn’t. In 1991 a precise torsion experiment was done, and found no effect. Because of this the common view is that the effect isn’t real, but there are still experiments that claim to confirm the Allais effect. Since the effect requires a total eclipse, you can’t do the experiment very often, and you need to have a setup portable enough to do on site. So getting good, consistent results is difficult at best.

The debate over the Allais effect still lingers. Some argue that it isn’t a real effect, some argue that it’s a real effect, but is due to external factors such atmospheric changes of temperature, pressure and humidity which can occur during a total eclipse. Others argue that it’s a real effect, and is due to “new physics.” This latter view has become popular among supporters of alternative gravity models. Allais himself claimed that the effect was the result of new physics, though never proposed a clear mechanism. As a result, the experiment has become “tainted” by fringe science to the point that mainstream scientists don’t really do the experiment any more. The 1991 result is pretty clear, and Allais’ results are likely due to experimental error.

In 2017, when a total eclipse will cut across the entirety of the U.S., you will likely see a flurry of new experiments being done. Perhaps these new experiments might verify the result, but I wouldn’t count on it. In the end, it’s the experimental results that matter, and if there does happen to be clear results from careful experiments that find an anomalous result, then that’s the way the pendulum of truth will swing in the end.

Paper:  Maurice Allais. Should the Laws of Gravitation be Reconsidered?, Aero/Space Engineering 9, 46–55 (1959)

Paper: Kuusela T. Effect of the solar eclipse on the period of a torsion pendulum, Phys. Rev. D 43, 2041–2043 (1991)


  1. Seems odd that it would only happen during a total eclipse. I’d guess that partial eclipses or even regular new moons would produce similar but perhaps lesser effects if the effect is real.

  2. I find this very odd, very much the doubter here. A pendulum depends of gravity for the the swing frequency and I find it difficult to find a relationship between the pendulum and the gravitational pull of the very distant Moon and the Sun.

    Now if for some reason the gravitational alignment of the Moon and the Sun augments its effect with the Earth, including its surface, and its own gravity, that is beyond my ability to sum up. You’d think that any anomalous result would have to do with gravitational alignment because that is indeed happening as in the additional pull of a solar tide added to the lunar tide.

    Would there be a difference in anomalous results depending on whether it was an eclipse during when the Earth is around closest or most distant from the Sun? After writing the above I found this following article on solar tide

    1. Author

      There are a few of these strange anomalous experiments that make the rounds every now and then. The difficulty is that they are claimed to occur only at certain times, which makes them difficult to repeat. Personally, I don’t put a give of credence to them. Some of the results are interesting, but not very compelling.

  3. Yes, thanks, that is the difficulty I perceive. Conditions at different eclipses can be very unlike, but I’m sure that accounting for those differences in the Solar System environment and knowing the maths it would be possible to prove or disprove any purported results attributed.

  4. Brian, I don’t believe in so-called Allais effect. If such effect exists this means that at the moment of solar eclipse the attraction between Sun an Earth changes. But this must changes the Earth’s orbit. The duration of the year must changes, too. If it is real then astronomers will see these changes. But they have never seen such anomalies. I am sorry for the clumsy English. It isn’t my native language. Best wishes.

  5. What is this? Why did you not mention the major NASA-fronted study of a 1999 eclipse which used many different types of devices and pendulums in many locations across the eclipse path…. ? Where are the results and data from this major study? They are not anywhere. There is plenty of before-the-test info and then a mention of “we are just going to double-check stuff before releasing the results” and then…. nothing. There is something very unusual or unsurprising about this if you will google it.

    1. Author

      The 1999 eclipse led to more of the same. Negative results in agreement with standard physics, a couple of positive results that were then criticized for their methodology. There are claims among some conspiracy types that after the 1999 NASA “shut down” work on the Allais effect, but there’s no real credibility to this story. Mainly I wanted to focus on an interesting aspect of inconclusive data rather than start fueling fringe debates.

  6. Its odd though that the 1999 results are not on the net… well I cant find them anyway (?). While there are several other eclipses tested before and after 1999 that have nice graphs which clearly show the results.

    I only heard of this amazing, enigmatic effect 3 days ago but looking at various sites, Allais’ experiments were carefully checked by authorities and were of execellent standard and that no explanation has been found. It seems all positive results are criticized for methodology. The only “explanations” due to error I can find sound more like hearsay and have no actual instances or details. Changes in temperature pressure and humidity… cannot account for it. Also, while the 1991 Finland was negative, it was a similar set-up to one done at Harvard in 1970 which was positive.

    I get the impression it is not a yes/no question but likely one of more than one factor affecting the result so that it could, for example, cancel out the effect. As you know, all new paradigms begin as fringe ideas and old paradigms will fight maintain the appearance of relevance despite being obsolete.

    1. Author

      Not that odd. Most of the 1999 effort was haphazard, and peer review is pretty stringent. Since 1999 there have been other experiments published, and they don’t support the claims of new physics. Allais’ experiments were critiqued by “authorities” and there has been criticism of his method. Claiming that old paradigms are fighting this new discovery is completely unfounded. There is no paradigm or testable model proposed to explain the effect, and stringent observations fail to confirm the effect at all. So it’s mildly interesting, but not amazing nor enigmatic.

      1. Old paradigms always fight against all new discoveries, but this is entirely proper, and does not usually amount to a conspiracy. That’s how science is supposed to work, up to a point. Sometimes the old paradigms take it too far. Remember what Max Planck said: “Science advances one funeral at a time”.

        The fact that “there is no paradigm or testable model proposed to explain the effect” is not decisive (although it would be nice to have one). For example, there is no testable model proposed as to why a mass causes the geometry of spacetime to be changed in the vicinity of the mass. It’s just something you have to accept as being a foundation of GR.

        It is not true that, in the last twenty years for example, “stringent observations fail to confirm the effect at all”. There have been both positive and negative claims.

        1. Author

          No, GR is the testable model for spacetime curvature. GR proposes it as an explanation for gravity, but you don’t “just have to accept it.” GR has been rigorously tested, and it keeps passing the tests, so we have evidence to support it. The point about the Alais effect having no proposed explanation is that science depends upon a confluence of evidence. Different experiments and observations should come together. If something has no clear explanation, then the results need to be extraordinarily clear, because it’s validity will require a re-examination of a range of theories. If the result isn’t clear, the most likely explanation is that it was an experimental issue.

          The effect hasn’t been confirmed. Having a mix of positive and negative effects makes the effect inconclusive, thus not confirmed.

          1. You misunderstand my comment. Yes of course, as an explanation of gravity, GR has been well tested and has passed the tests. My point is that the foundations of GR have not been further clarified – for example, we must accept that local mass/energy entails change of the local spacetime geometry, and no model is proposed for that. In fact, there is no settled opinion as to whether “gravity has gravity”. Moreover GR cannot be other than an approximate theory, because it assumes continuity of matter.

            I absolutely agree that the claimed Allais effect is currently upon shaky ground, because physics, or at least fundamental physics, is supposed to be a unity, and therefore if something doesn’t fit in with other things that thing is necessarily very suspect. However, observations cannot always be simply dismissed in a cavalier fashion. Of course it depends upon the strength and consistency and repeatability of the observations. I think we are on the same page; the difference is only one of emphasis.

  7. If the 1999 results were negative or inconclusive, shouldn’t they have been published anyway? Especially after NASA made such a big deal in advance about the experiment? There IS something odd about that. How often does a government agency sponsor a major study and then not publish any results and say nothing about it henceforth?

  8. Yes, but it was a long time ago. This is 2016, and we should be considering how we can do similar experiments better during the 2017 eclipse.

Leave a Reply