≡ Menu

Is There a Flaw in General Relativity?

By Larry Klaes

We have much to do as we scramble to explain the universe’s continuing acceleration. Dark energy seems to be demanded by the data, but there are holdouts who argue for a reinterpretation of General Relativity. Tau Zero journalist Larry Klaes looks at one proponent of a revised GR who sees exceptions to the rule in a far earlier era.

Albert Einstein’s work created one of the biggest revolutions in the history of science and radically changed our perceptions of the Cosmos. One of his later breakthrough ideas is the Theory of General Relativity, or GR for short. Really massive objects such as the Sun literally warp space and time around them as they move through the heavens. Since Einstein first published his ideas on GR in 1915, scientists have been able to use the theory to understand the behaviors of even more massive celestial bodies and the very beginning of the Universe itself.


Image: This key Einstein paper included the effect of gravitation on the shape of space and the flow of time, proposing that matter causes space to curve. Credit: NCSA/UIUC.

The Need for a New Look

Now an assistant professor of cosmology at Cornell University thinks she may have found a deviation from the predictions of GR at the very largest cosmological scales. If her finding turns out to be true, it would be a major shift in our understanding of how our Universe works. Rachel Bean, whose paper on the subject has just appeared on the arXiv site, explains that while gravity is very well tested on Solar System scales, when it comes to the distances of the farthest galaxies, which are billions of light years from our own Milky Way galaxy, astronomers still have much to understand.

“Einstein’s theory of General Relativity tells us how gravity determines the relationship between the matter in the Universe and the Universe’s size,” says Bean. “If matter in the Universe is attracted by gravity, then the Universe’s size should be expanding at an ever slowing rate (decelerating). What observations show, though, is that in the last 6 billion years (out of the Universe’s 13.7 billion year existence), the expansion has been speeding up. So we need new physics, labeled ‘dark energy’, to explain this apparent disparity.”

Dark energy is a hypothetical force which makes up almost three-quarters of the mass-energy of the known Cosmos. Cosmologists currently use dark energy to explain why the Universe is expanding when it should otherwise stop moving or even collapse back upon itself.

Scientists do not yet know what dark energy is composed of. Adds Bean:

“There are two broad possibilities for what dark energy is. Either it is a strange, new type of matter that is not attracted by gravity, or Einstein’s Theory of General Relativity needs to be modified on cosmic scales. Part of my research is to try and understand the origins of dark energy and how we can use current and future astrophysical surveys to measure its properties.”


Image: The bending of starlight by matter and energy is well attested. Is there any way General Relativity may have differed in earlier eras? Credit: Ethan Siegel/Lewis & Clark College, OR.

Dark Energy and Cosmic Structure

Bean examined and compared the data collected from the recent Cosmic Evolution Survey, or COSMOS, which involved a number of telescopes both on the ground and in Earth orbit, including the Hubble Space Telescope (HST). The survey examined over two million galaxies, along with distant exploding suns called supernovae and the relic radiation created just 400,000 years after the formation of our Universe in the Big Bang, otherwise known as the Cosmic Microwave Background.

The Cornell scientist says that in order to distinguish between the two possible types of dark energy, astronomers need to include a second type of observations in addition to the distance measurements, namely how galaxies and clusters of galaxies, or large scale structures, evolve over time under the influence of gravity. Bean did this by examining the data on distant “lensed” galaxies whose light from their billions of suns is distorted by other galaxies between themselves and our Milky Way.

“My paper uses a combination of large scale structure and distance measurements to test gravity on cosmic scales to see whether it is consistent with GR or not,” declares Bean. “What I find is that there is a better fit to one time slice in the weak lensing data from the COSMOS survey using the HST, between 3 and 7 billion years after the Big Bang, if gravity is allowed to deviate from GR.”

A Modification to Gravity

If her idea proves to be correct, the results could “indicate that dark energy is related to a modification to gravity rather than a new type of matter.”

While Bean recognizes and admits there may be systematic errors in the data she studied, she also sees her work as a method for using these large scale structure datasets to test theories on dark energy:

“Future astronomical surveys coming in the next six years, namely the Dark Energy Survey (DES) and the Large Synoptic Survey Telescope (LSST), of which Cornell has just become an institutional member, will have much improved control over errors in the analysis and will be able to constrain the properties of gravity with far better precision.”

When it comes to the Universe, Bean says she was always interested in astronomy, mathematics, and physics as a child.

“My mum said that I used to go into bookshops even as a small child and buy math books rather than reading books. However, I didn’t get into astrophysics until my mid-20s after a few years working as an analyst at a Management Consultancy. Since then I’ve never looked back….”

The paper is Bean, “A weak lensing detection of a deviation from General Relativity on cosmic scales,” available online.


Comments on this entry are closed.

  • andy November 3, 2009, 15:49

    There’s already been some discussion of this at Cosmic Variance. There’s a lot of interesting discussion in the associated comment thread, in particular, note the comment from Alexie Leauthaud at #6 is worth bearing in mind. We’ll see whether this result holds up under the new reduction of the data that’s going on…

  • Steven R. Martin November 3, 2009, 21:51

    I worked on a project that hoped to measure the frame-dragging effect stipulated by GR. It was called Gravity Probe B, and we are still awaiting the final verdict from the Stanford University team who is analyzing the data. One has to wonder, based on Ms. Bean’s paper, whether a new theory needs to augment GR for distances of the grandest scales….

  • ljk November 3, 2009, 23:27

    Title: Melia, Fulvio: Cracking the Einstein Code

    Publisher: University of Chicago Press

    Albert Einstein’s theory of general relativity describes the effect of gravitation on the shape of space and the flow of time. But for decades after its publication it remained little more than a curiosity; however accurate it seemed, Einstein’s mathematical code – represented by six interlocking equations – proved extremely difficult to crack. That is, until Roy Kerr solved the great riddle in 1963.

    Fulvio Melia here offers an eyewitness account of the events leading up to Kerr’s great discovery. For more information, see the book synopsis at:


  • ljk November 4, 2009, 0:50

    General Relativity and Quantum Cosmology

    Title: Discovering the Kerr and Kerr-Schild metrics

    Authors: Roy P. Kerr

    (Submitted on 8 Jun 2007 (v1), last revised 14 Jan 2008 (this version, v2))

    Abstract: An historical account of the reasoning that led to the discovery of the Kerr and Kerr-Schild metrics in 1963-1964, and their physical interpretation as rotating black holes, is presented.

    Comments: 34 pages, 2 figures, cupbook.cls. To appear in “The Kerr Spacetime”, Eds D. L. Wiltshire, M. Visser and S. M. Scott, Cambridge Univ. Press

    Subjects: General Relativity and Quantum Cosmology (gr-qc)

    Cite as: arXiv:0706.1109v2 [gr-qc]

    Submission history

    From: Roy Kerr [view email]

    [v1] Fri, 8 Jun 2007 04:21:18 GMT (209kb)
    [v2] Mon, 14 Jan 2008 22:52:40 GMT (226kb)



  • Ron S November 4, 2009, 1:28

    In the article’s title, the word ‘flaw’ grates a bit. I prefer Steven’s term: “augment GR”. General relativity is not flawed, but rather has its domain of known validity. The theory itself gives strong hints where its validity is likely to fail. Figuring out precisely how and where it departs from reality is the challenge, and an opportunity to gain a deeper understanding. That makes Bean’s result interesting.

  • Adam November 4, 2009, 17:59

    What Ron said. Just like Newtonian gravity is the approximation at low field strengths, so too GR would be the approximation at subgalactic/cosmic scales if the suggested modifications are adopted.

  • Ron S November 4, 2009, 22:07

    Adam, a quibble with what you said. Not modification, but augment. GR itself is fine for now, although we may need a wider-ranging theory (like QG) and/or a more accurate metric to plug into GR at cosmological scales. The application of GR at the cosmological scale requires finding a metric that applies to the large scale structure of the cosmos, whereas GR is at its foundations a local theory (infinitesimal scale).

  • Athena Andreadis November 5, 2009, 1:47

    I agree with Ron and Adam — particularly because there have been significant changes to some of the input data that Dr. Bean used for her calculations. So there are three possibilities at least: newer, more accurate numbers bring the results in line with prevailing GR predictions; GR as postulated by Einstein et al undergoes a qualitative refinement; or GR is discovered to be a boundary case of an even larger theory.

  • Erik Anderson November 5, 2009, 11:48

    I don’t see this finding necessitating modified gravity, but it may constitute yet another demonstration that current formulations of lensing theory are unworkable (the other examples being the lensing profiles of individual galaxies and galaxy clusters, which portend, perhaps falsely, the existence of more mass than seen).

    Time will tell if the analysis holds up. Who knows — revising the data may put their value for eta even further away from 1.

  • James M. Essig November 5, 2009, 15:27

    Hi Folks;

    The notion that General Relativity might need a cosmic scale correction would be good new to me since I enjoy the whimsical nature of mysteries in physics and cosmology.

    Note, however, that conventional General Relativity can still be very usefull in developing exotic space time topology and kinematical distortion technology such as any workable forms of warp drive, somehow stabilized macroscopic wormholes, and in general, Einstein Rosen Bridges.

    The analogy is simmilar to the use of proton-neutron models of the atomic nucleus as it pertains to advanced nuclear engineering research and research into high energy density materials such as the Hafnium m2 nuclear isomeric state.

    Another obvious example is the usefullness of classical electromagnetic theory, for instance, in the study of the phenomenon of magnetic field lines reconnection and negative electromagnetic refractive index meta materials.

    Classical General Relativity as any approximation to modified General Relativity will I hope be instrumental in getting us to the stars sooner rather than later.

  • forrest noble November 5, 2009, 21:00

    I think she is right, GR needs to be modified. Not for the reason she proposed, however, but instead I think the reason involves the hypothesis of dark matter rather than dark energy. Since spiral galaxy rotation rates and the rotation rates of galaxies in a cluster do not match the predictions of General Relativity unless there is a very large additional amount of unseen mater, hence the dark matter hypothesis was proposed. The formulations of GR, it would seem, could readily be altered without great difficulty by ad hoc adjustments, to provide more correct predictions of galactic rotation and eliminate the need for dark matter. The main problem with this however is that Einstein’s Cosmological Equations which is the foundation physics of the Big Bang model, are also founded upon the equations of General Relatively. To drastically change the Einstein’s Cosmological Equations would result not only in a different concept of gravity, it would result in a different cosmological model which might be very different and contrary to the Standard Model in most ways.

    To eliminate the need for dark energy the easiest explanation is that the Hubble Formula is wrong rather than General Relativity for dark energy, as she suggests. This also can be readily reformulated but the result again would probably result in a different cosmological model. To change GR to account for both dark matter and dark energy, I believe, would result in an unworkable and illogical cosmological model.

    In my opinion all three need to be changed: GR, the Hubble Formula, along with the standard model of cosmology. Make a clean sweep.

  • forrest noble November 5, 2009, 21:13

    Ron S and Adam,

    I agree with your two comments.

  • Peter November 8, 2009, 9:12

    Ron S comments
    “In the article’s title, the word ‘flaw’ grates a bit. I prefer Steven’s term: “augment GR”. General relativity is not flawed, but rather has its domain of known validity.”

    This “augment GR” argument is similar the argument that the Scholastics used to deal their Ptolemaic geocentric theory that failed to predict the wandering of the planets. It seems to me that the time is ripe for our revered and sacrosanct Scientists to consider that R. Bean’s recent observations, the flat rotation curves and cosmic acceleration represent “serious anomalies” similar to the photoelectric effect and the black body spectrum that call call for a far-reaching paradigm shift.

    Since there just happens to be a curious similarity between the Ptolemaic theory and Einstein’s General Relativity, it time to come to come to grips with the problem that that the amateur Copernicus faced and dealt with. The 1000 year-old idea that the earth has some mysterious unspecified ability to make all the objects in the sky revolve around it in a 24 hour period has a lot of similarity to the 300 year-old idea that mass has some mysterious unspecified ability to either attract mass or warp space. And guess what: this mysterious problem with mass has a very simple and solution to the mysterious powers of the earth that the Scholastics should have faced.

    The Stephan-Boltzmann law indicates that if mass has temperature it has heat in the form of radiation leaving it. Given the the Tully-Fisher law and the Mass-luminosity relation it is reasonable to consider that all along it has not been mass that mediates the gravitational force but the heat leaving it that has mediates it.

    I have five experiments which show that the weight of a test mass will either increase or decrease at the 2-9% level depending on the direction which heat transfers through that mass. For these experiments and a heat-based gravity theory go here: http://vixra.org/abs/0907.0018.

  • Ron S November 8, 2009, 12:20

    Peter, life is short. I do not waste it on cranks. Look elsewhere.

  • Peter Fred November 8, 2009, 20:09

    Ron S

    I would be interested what your thoughts might be about my experimental results. Do you think I made them up? My experiments are easy to replicate. Just because they results run counter to what you have read in the text book or what you believe, does that make automatically spurious? If they are valid results, they raise some questions about the equivalence principle and Einstein’s interpretation of E=mc^2. So if I were you, I would spend some time worrying about my experimental setup.

  • Chris December 5, 2009, 5:17

    Great care should be taken when trying to modify GR. If it is found that a correction needs to be made at larger scales than the solar system, then how large? Is there some “magic” point to where traditional GR gives way to the modified version? Is it a gradual transition? Trying to accurately modify the work of the greatest mind of the 20th century? Seems like a very slippery slope.