arxiv Find: Breakdown of Classical Gravity?

By Sean Carroll | May 22, 2011 12:47 pm

The single most interesting feature of attempts to replace dark matter with a modification of gravity is Milgrom’s discovery that in a wide variety of galaxies, there’s a unique place where ordinary gravity plus ordinary matter stops working: when the acceleration due to gravity (as Newton would have calculated it) drops below a fixed value a0 ≈ 10−10 m/s2. This is the basis of MOND, but the pattern itself is arguably more interesting than any current attempt to account for it. Very possibly it can be explained by the complicated dynamics of baryons and dark matter in galaxies — but in any event it should be explained somehow.

The existence of this feature gives a strong motivation for testing gravity in the regime of very tiny accelerations. Note that this isn’t even a statement that makes sense in general relativity; particles move on geodesics, and the “acceleration due to gravity” is always exactly zero. So implicitly we’re imagining some global inertial frame with respect to which such acceleration can be measured. That’s a job for a future theory to make sense of; for the moment we’re forgetting that we know GR and thinking like Newton would have.

So now Hernandez, Jimenez, and Allen have tried to test gravity in this weak-acceleration regime — and they claim it fails!

The Breakdown of Classical Gravity?
X. Hernandez, M. A. Jimenez, C. Allen

Assuming Newton’s gravity and GR to be valid at all scales, leads to the dark matter hypothesis as a forced requirement demanded by the observed dynamics and measured baryonic content at galactic and extra galactic scales. Alternatively, one can propose a contrasting scenario where gravity exhibits a change of regime at acceleration scales less than $a_{0}$, and obtain just as good a fit to observations across astrophysical scales. A critical experiment in this debate is offered by wide orbit binary stars. Since for $1 M_{odot}$ systems the acceleration drops below $a_{0}$ at scales of around 7000 AU, an statistical survey of relative velocities and binary separations reaching beyond $10^{4}$ AU should yield a conclusive answer to the above debate. By performing such a study we show Kepler’s third law to fail precisely beyond $a approx a_{0}$ scales, precisely as predicted by modified gravity theories designed not to require any dark matter at galactic scales and beyond.

Color me dubious, but interested in seeing further studies. It’s very hard to collect this kind of data, and note that it’s just a statistical survey of velocities, not a precise measurement of individual systems. In principle a statistical survey is fine; in practice, it opens up the possibility of hidden subtle systematic effects.

Still, intriguing and worth checking out. Any time you have the chance to overthrow Sir Isaac Newton, you go for it.

CATEGORIZED UNDER: arxiv, Science, Top Posts
  • Gavin Flower


    (1) Could this result relate to the notion that gravity is quantized?

    (2) Does this have anything to do with the effects of Dark Energy pushing things apart?

  • Ben

    Hmmm, color me “MOND enthusiast”, but color me very dubious too. This study is based on proper motions for which the typical errors are 10 to 20 times larger than the proper motions themselves…. so what they plot is thus most probably just some reflection of the measurement errors…. Then, at a more academic level, in MOND you have something known as the “external field effect”. Of course, this is not an argument against the observation itself, just a clarification that it’s not a prediction of MOND. In MOND and MOND-inspired theories, the strong equivalence principle must be violated (as Sean was explaining above, it’s the “global inertial frame” that matters, and the acceleration with respect to it, not the relative acceleration). So, in short, when the external gravitational field dominates the internal one, the dynamics of a given system is not “MONDian” but Newtonian with a renormalized G. The galactic gravitational field is about 1.5*a0 at the galactocentric radius of the Sun, so the prediction of MOND for such low-acceleration binaries is the same as the Newtonian prediction but with renormalized G_m = G*(2.5/1.5)=1.66*G …… Anyway, very interested in seeing further studies too.

  • Ellipsis

    Couldn’t this be tested in the laboratory (via torsion pendula)? And, conversely, don’t torsion pendulum results over the past couple decades impact this issue? (There is no mention of laboratory studies in the article.)

    (Whenever one has an actual choice of measuring something in the laboratory vs astrophysically, the former is generally preferred, due to repeatability and control of systematics.)

  • Sean

    You can’t measure this in the laboratory because you’re sitting on the Earth’s surface, where the acceleration due to gravity is much larger than the point below which the effect is supposed to kick in. If you decide you only care about mutual gravity between two objects, I’m not sure how small of an acceleration you can measure.

  • Ellipsis

    Sean — the _galactic_ gravitational potential that these widely-spaced binaries are in is typically above the a0=10^-10 cutoff. So your reply should also hold for them as well, correct…?

  • Ben Babcock

    Whenever I hear news like this, I keep expecting a hologram of Isaac Newton to appear and announce, “Congratulations, you have successfully navigated the first Newton Crisis.”

  • Sean

    Ellipsis — yes, if that’s true. I don’t actually know what the galactic acceleration is for the stars that were observed. Of course it could be that only the mutual attraction matters (although that’s against the MOND idea), and it would be worth testing independently.

  • Baby Bones

    Why is the dark matter hypothesis a “forced requirement” merely upon requiring that GR is valid at all scales? That is, the first statement of the abstract makes it seem like GR rules out some other physical force competing with it for dominance in a particular scale of the universe. Were that true, I wouldn’t be able to raise finger to type. Gravity keeps us stuck to the ground but it certainly doesn’t have the strength on Earth to overcome the electromagnetic and nuclear forces that stop matter from collapsing to form a black hole.

    I’d much prefer a competing negative force, one that accounts for galactic and, as the authors write, binary star rotations at scales at which it is comparable to gravity but also accounts for the acceleration of galaxies away from each other. At galactic scales, it competes with gravity on close to equal terms but at intergalactic scales it beats gravity.

    Sorry about being so non-geodesic.

  • Late Isaac

    Congratulations. You have successfully instigated the first Newton Crisis.

  • Ben

    The maximum distance of these binary systems is 991 pc from the Sun, as written in the paper, so the minimal gravitational field from the Galaxy would be of the order of (220km/s)^2/9kpc (placing the Sun at 8 kpc, assuming a constant circular speed of 220 km/s, and stating that the gravitational acceleration equals the centripetal acceleration). The value 10^-10 m/s^2 being equal to 3400 (km/s)^2/kpc , you find that 220^2/9/3400=1.6, so the answer to Sean (7), is “at least 1.6*a0”, so see my post above for the consequences in “usual MOND”. Of course, one could imagine that only the mutual attraction matters, and as Sean says, this is worth testing independently: if only the mutual attraction matters, then interesting tests could be performed on Earth as suggested by, for instance, Ignatiev and And indeed, from torsion pendula asd mentioned by Ellipsis, Gundlach et al. ( claim measures at a precision of the order of less than 10^-13 m/s^2, so a modification based on mutual attraction a priori sounds ruled out from lab experiment.

  • Lab Lemming

    Would this unbind the Oort cloud?

  • Sean

    Thanks for taking the time to plug in the numbers, Ben. It looks bad for making sense of this claimed measurement in a modified-gravity scheme; either what really matters is the total acceleration due to everything, in which case these systems should be (barely) in the Newtonian regime of our galaxy, or what matters is only the mutual attraction, but that’s apparently ruled out by laboratory experiments.

  • Error Bar

    Is it just me or does it seem like there could be some systematic error evident in figure 3? Why should the distance from us shift Delta v up or down? Is this explained in the article, or can anyone enlighten me?

  • fd

    The experiment by Gundlach et. al. tests Newton’s 2nd law. For a force from the pendulum fiber, does the acceleration agree with F=ma at very low acceleration? So, it doesn’t test very small mutual gravitational attractions. Would it be interesting to do the latter test in the lab?

  • blanton

    At 500 pc from the Sun, with typical errors (as quoted in the Dhital paper) of 3 mas/yr, I think the velocity errors are of order 7 km/s. That would seem to make it impossible to perform the test that they describe, which requires 1 km/s level precision. Even if the actual relative velocities were exactly on Newton’s Law, the measured relative velocities would always have a broader distribution because of the uncertainties in the proper motions (clipped by the selection effects of the catalog, presumably). This is pretty much exactly consistent with the distribution they see. The relative velocities seem to increase slightly with radius, which is odd, but may indicate an increasing fraction of interlopers. Unless I’m misunderstanding something about what they’ve done, which is entirely possible, I think what they see is just due to the measurement uncertainties.

  • Bernard Conus

    The picture below is a loaded Eotvos torson pendulum. Four masses on the left of one composition are contrasted with four masses on the right (two obscured behind the rotor) of a different composition, the whole of it being gold plated. Earth’s spin inertial acceleration is crossed by Earth’s solar orbit gravitational acceleration. If the two sides follow divergent trajecories, the pendulum is periodically torqued,

    Physics postulates all material bodies vacuum free fall identically – minimum action parallel trajectories and identical rates, the Equivalence Principle, EP. The EP is universally observed true for all *measurable* observables in all venues at all scales to 5×10^(-14) difference/average.

    Your shoes are different. Their difference cannot be measured, though it is easily observed and calculated. Do chemically identical opposite shoes violate the EP?

    If the vacuum is observed to be anisotropic toward fermionic mass distribution, Noether’s theorems no longer couple vacuum isotropy to conservation of angular momentum. Geometric parity, chirality in all directions, is an absolutely discontinuous symmetry and therefore outside Noether (i.e., continuous Lie groups). There would no contradiction if fermionic mass had a (very small!) non-zero minimum value of angular momentum. The MOND residual is sourced.

    “Shoe” is a metaphor. Crystallography builds mathematically perfect opposite shoes, emergent at atomic scale, to kilogram size as pairs of enantiomorphic space group single crystals. Paired examples are P3(1)21 | P3(2)21 alpha-quartz or P3(1) | P3(2) gamma-glycine. A chemical footnote outside physics is the test.

    The four test masses on the left would be gold-plated single crystals of geometrically left-handed alpha-quartz, 3(2) screw axes. Those on the right would be gold-plated single crystals of geometrically right-handed alpha-quartz, 3(1) screw axes. Run the geometric parity Eotvos experiment. A non-zero net output is easily validated. Run two hemiparity Eotvos experiments, each space group of alpha-quartz opposed by amorphous fused silica test masses. Left shoe, right shoe, or socks… all are silicon dioxide.

    Examine spacetime geometry with test mass geometry, on a lab bench. A left foot is not detected by socks and left shoes. Massless boson photons as tests are excluded by prior observation, arxiv:0912.5057, 0905.1929, 0706.2031. Massed fermion right shoes are the test of a vacuum left foot.

    Somebody should look. The worst it can do is succeed – validating MOND and falsifying nearly 30 years of ad hoc manually inserted symmetry breakings patching intrinsically defective physical theory.

  • Dave

    The analysis seems completely inadequate to answer the question that they pose, as is typical of analyses that ignore the error bars. As blanton points out, the error bars are much larger than the claimed signal. It is possible, in principle, to address MOND-related questions with this data with a Bayesian analysis that reproduces the distribution of measurement uncertainties. But, this may require the full data set used by Dhital et al., and it could mean a lot of work for a useless test (i.e. showing that both Newton’s law and MOND are consistent with the data).

  • ohwilleke

    In my book, Hernandez, et al. deserve kudos for even recognizing a new way to test gravitational effects in a weak field that should produce results more or less independent of those from galactic rotation curves, lensing and similar past empirical data sets.

    Recognizing wide orbit binary stars as a good place to conduct tests of gravity in the weak field, even if more money is necessary to do a really good measurement is a breakthrough. It is almost surely cheaper to get good measures of wide orbit binary stars if we put out minds to it than it is to do high precision measurements of the frame dragging, for example.

    Also, keep in mind that this experiment alone doesn’t have to quantify the variance from GR predictions very exactly to be useful. All we want to do is test the null hypothesis that this system shows no deviation from it. With effective G at 1.6 times the predicted value, one doesn’t need much better measurement to beat the error bars at a high level of statistical significiance.

    And, as Sean notes, it doesn’t really matter if this is due to MOND or dark matter or quantum gravity or some other damn thing. The point is that this additional piece of empirical data places more limits on any possible theory. For example, if we see these effects in wide orbit binary star systems than any dark matter theory that explains it can’t rely on galactic black holes to produce the necessary dark matter distribution. In effect, if you follow a dark matter paradigm, this result shifts the balance in favor of hotter rather than colder dark matter (since cold dark matter is too cuspy in its natural distribution).

  • Jack Sarfatti

    This is trivially explained in my paper in the Penrose-Hameroff issue of the Journal of Cosmology

    a0 ≈ 10−10 m/s2 ~ c^2(Cosmological Constant)^1/2

    This is the smallest acceleration possible if we are hologram images projected from a 2D hologram screen which in this case is our future de Sitter event horizon. It is a kind of round-off error in the hologram cosmic computer.

  • blanton

    ohwilleke is correct that this could in principle be an interesting test. In particular, the Gaia data will provide much smaller error bars for proper motions of stars of these magnitudes. A few years off, but an entirely plausible use of the data, which somebody should do.

  • Jack Sarfatti

    “The existence of this feature gives a strong motivation for testing gravity in the regime of very tiny accelerations. Note that this isn’t even a statement that makes sense in general relativity; particles move on geodesics, and the “acceleration due to gravity” is always exactly zero. So implicitly we’re imagining some global inertial frame with respect to which such acceleration can be measured. That’s a job for a future theory to make sense of; for the moment we’re forgetting that we know GR and thinking like Newton would have.”

    The de Sitter metric in the observer-dependent static LNIF representation has

    g00 = 1 – /r^2

    where we are at r = 0 and /^-1/2 is the scale of our future event horizon.

    a static LNIF has acceleration

    g(r) = c^2/^1/2(1 – /r^2)^-1/2

    the Unruh temperature is proportional to g(r).

    so we have

    g(0) = c^2/^1/2

  • Ben

    A problem with using a catalog like this to test the distribution of orbital velocities is that it’s a common-proper-motion selected sample, and that the errors on proper motions (hence projected velocities) are large. The Dhital et al sample is selected such that the proper motions of the two stars are statistically consistent given the errors within ~1 sigma (their Eq 6). Figure 4 of the Hernandez et al paper attempts to test the effects of this limit by color coding as red crosses the points that are close to the Dhital selection limit (ie proper motions of the two stars are close to discordant).

    Hernandez et al claim that Fig 4 shows they don’t have a problem because the red crosses aren’t all butted up against the upper envelope of points. This doesn’t seem terribly convincing, since the red crosses are all pretty close to the upper envelope.

    There are two issues here. First, one would like to generate a hypothetical sample of binaries, apply the observational errors to them (which would scatter nearly all of them up to higher relative velocity), apply the observational selection, and see what the radius-relative velocity diagram looks like. Second, to really test for an effect and the existence of an upper envelope in the relative velocity distribution, I think you also need to know the distribution of points that fell outside Dhital’s selection boundary. Dhital’s aim was to create a highly reliable sample of binaries, not a highly complete (inclusive) sample of binaries.

  • Dentin

    One thing that occurs to me in regard to the ‘relative field strength’ issue (the galactic field is stronger than the cutoff for the observed stars) is that this could be observation of quantization. If the field strength is quantized and you’re near the quantization floor, gravity may appear to be weaker than inverse square. This would make it harder for wide orbit binaries to stay coherent.

    With a quantization floor of 1e-10, you would start to see these effects once the field strength reached the 1e-9 level, and would expect them to be pretty strong by the time you hit 3e-10.

    (Also note that this kind of minimum-field-strength floor would affect gravitational wave detectors, and those detectors could be used to disqualify the hypothesis.)

    It’ll be interesting to see more observation and analysis in this regime, but I have little hope for seeing new physics here.

  • Muggs

    I was wondering… if these MOND effects only show up in very low field space, maybe it would show up in intergalactic dust distributions.

    Shouldn’t gravitational waves tend to make regions in space collect higher density of gas and others lower, just like sound on our scale? I wonder if pairs of binary stars could generate patterns in the interstellar dust around them.

    Just an idea.

  • Pingback: Dark Matter heats up « The e-Astronomer()


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Cosmic Variance

Random samplings from a universe of ideas.

About Sean Carroll

Sean Carroll is a Senior Research Associate in the Department of Physics at the California Institute of Technology. His research interests include theoretical aspects of cosmology, field theory, and gravitation. His most recent book is The Particle at the End of the Universe, about the Large Hadron Collider and the search for the Higgs boson. Here are some of his favorite blog posts, home page, and email: carroll [at] .


See More

Collapse bottom bar