Modifications to General Relativity?

By JoAnne Hewett | August 10, 2005 4:29 pm

That was the subject of our contest at this year’s SLAC Summer Institute. It has become a tradition to have the students participate in a contest to make the best prediction of what will happen in our physics future. Ok, so the tradition is not that old, it is the brainchild of Chris Quigg from Fermilab who started it last year. It’s great fun and will no doubt continue long in future Summer Institutes. This year’s official question was:

What will be the first evidence to demonstrate that Einstein’s theory of General Relativity must be revised and when will that be found?

We decided not to ask anything provocative… We had alot of entries, some of which were even serious. Contestents came from near and far in order to vie for the prize – bottle of California’s best sparkling wine. None other than George Bush submitted two entries:

GR is perfect, no evidence will ever be found to contradict it. – George Bush.

2013: Blackholes from extra dimensions will be observed at the LHC with long lifetimes, suppressed cross sections and a high mass threshold. These properties will lead to necessary modifications to GR from HEP measurements. – George Bush.

Theory has it that W flew his Secretary of Energy, Samuel Bodman, out to SLAC just to submit these entries into the contest box. Secretary Bodman, who was observed hovering around the box, covered up the real mission behind his visit by giving a speech to the lab employees. My personal theory is that Bodman submitted the second entry by himself (note it shows more knowledge than the first), but felt compelled to sign his bosses’ name.

We also had some humorous entries from the students. One compared Hell to a blackhole, since it is well-known that upon entering Hell, it is impossible to escape. He then calculated the Schwarzschild radius of Hell in his Hell-Freezes-Over Theory.

The entries were judged by a panel of distinguished experts:


who decided upon 2 honorable mentions, 1 runner-up, and 1 winner. The first honorable mention noted that recent cosmological measurements have justified the need for a non-vanishing cosmological constant and that was already a modification in a certain sense. The panel had some sympathy for this view. Honorable mention #2 stated “I think that the most likely (first) violation for Einstein’s GR will be in terms of Lorentz violation: that is, the existence of a preferred frame in GR.” That was cute, but the principle of Lorentz violation was the winning answer to last year’s contest, so we couldn’t award it again! (Can you just imagine, every year people would be submitting answers with Lorentz violation.)

The runner-up was more succinct:

~ 6 weeks.

She is a member of the Eotvos experiment (which tests gravity at short distances) at the University of Washington and we thought maybe she might have some inside information…

For the grand prize, the winning entry:

The Cassini spacecraft will also exhibit anomalous acceleration towards the sun (“Pioneer anomaly”). A specifically designed follow-up mission will then confirm this and find oscillating anomalous accelerations out to 5 times Pluto’s orbital radius. Taking into account time scales for landing space missions and reaching the outer solar system, this will probably happen ~2050.

The winner was from Zurich and I hope he enjoys his bottle of California’s finest!

  • Steinn Sigurdsson

    Hm, the UWa rumour is intriguing, but would a deviation from 1/r^2 at short distances really count as a GR violation? Both new forces and extra-dimensions are generally compatible with GR, though specific ones need not be.

  • Clifford

    Steinn: I would say so, in keeping with the spirit of the question. Einstein had in mind four dimensions when he formulated GR, and I wouldn’t say it was a too-strict interpretation to not call gravity in five or higher dimensions GR anymore, (in this context!) even if the physics followed from a (higher dimensional) Einstein-Hilbert action.


  • Aaron Bergman

    Besides, the rumors are that the force of gravity gets weaker at shorter distances. Tough to pull that off with extra dimensions or extra (attractive) forces.

  • Ben Lillie

    Last year there was a question about why the laws of physics are selected according to an action principle. I think an appropriate question for next year is: “What extremal principle was used to select the photos of the panel?” (Note that it is self-evident that an extremal principle was used.)

  • Anonymous

    Perhaps a Principle of Extreme Unction is at work.

  • Quantoken

    Don’t read into the UW rumor too much. We all know gravity is at least 10^40 times weaker than electrical force. So the experiments could be easily tampered by all sorts of couloumb forces attributable to condensed matter physics. Wait until at least 100 groups re-produce the same thing and draw the same conclusion, before you trust any thing. Besiders, their experiment is at milimeter scale. You can pretty bet everything is still 4-D at that scale, any extra dimentions is supposed to be suppressed until you reach Planck Scale.

    Now, there is something that may spell trouble. Has any one look at the SHAPE of spiral galaxies and deduce something out of it? The spiral shape gives you a perception of rotation. And indeed it’s formed because the inner circle and outer circle does not rotate at the same angular velocity.

    Take a look at these spiral galaxies, look at the shapes and just count from inside out, you can get a pretty good estimate that the whole thing probably rotated no more than 2 turns since the birth of these galaxies.

    Now these estimate of rotations can be checked against another estimate, calculated by the rotation speed, measured from Doppler Shift, and distance to the center. However there is a huge discrepancy when you compare the two. For example, the solar system is approx. 26 K light years to the center of Milky Way. It’s rotating at approx. 289 km per second around the galaxy center. That figures to approximately 1.7 billion years per turn. If the galaxies are as old as the universe, that would be 14 billion years, divided by 1.7 billion, which is 8-9 turns. But the spiral shape says it’s no more than 1 to 2 turns!!!

    Any thought on what the spiral shapes can tell us?


  • Quantoken

    Sorry. It was a typo. The distance is 26 K light year, 26000. So it actually takes 170 million years to rotate one turn, not 1.7 billion. So the Milky Way would have rotated 80 turns already!!! But the shape, if it looks like most other galaxies, says it has rotated just 1 or 2 turns!

    How do you reconcile the two?


  • Astronomy Grad Student

    re: spiral galaxies

    –> density wave theory

  • citrine

    Wouldn’t the evolution of the spiral shape figure into this somewhere? I don’t know the most popular current theory of galactic evolution to elaborate on this but if the spiral arms grew outwards from the core, conservation of angular momentum would require that the rotational speed would decrease with age.


  • Jill

    I know it’s really uncool of me, but I have very little idea who those people in the pictures are, except for Sean C. Is that Cumrun Vafa hiding behind the book? The others are complete strangers to me, sorry. Help please!

  • Clifford

    Surely you recognize our very own dear JoAnne, Jill? I’ll leave it to others to play guessing game on the others. -cvj

  • Aaron Bergman

    So, there’s Kieth Dienes and Joe Lykken. Not getting the others, though.

  • Clifford

    That’s Peskin Peeking, no? -cvj

  • Mark

    Since nobody else is jumping in, and Jill really wants to know, I think the remaining two are Tom Rizzo, and Michael Peskin behind the book (The Higgs Hunter’s Guide, I believe).

  • Clifford

    Yep. I think you’re right. -cvj

  • Sean

    The finding chart is:

    Keith Dienes — Tom Rizzo — Joe Lykken
    JoAnne Hewett — Sean Carroll — Michael Peskin

    I don’t know about “distinguished,” but certainly opinionated.

    More importantly: the contest winner was Niklaus Berger, and the runner-up was Claire Cramer. Congratulations to them!

  • Pingback: Relative importance | Cosmic Variance()

  • Steinn Sigurdsson

    Ok, I’m intrigued, and clearly out of the loop – knew I should have gone to Aspen this summer, was supposed to be there in June but canceled.

    Weaker? That is not easy to do. It is even harder to do sensibly.
    Anyone have an alleged scale? I’m guessing 0.1 mm +/- because that’s basically the range I’d expect them to be probing…

  • Aaron Bergman

    Apparently it’s right in the region that they don’t trust their detector enough to claim an actual detection. Instead we just get these rumors that make those old fat graviton ideas look a lot more intriguing.

  • Sean

    In lieu of an actual post, here’s one of the original fat-graviton papers by Raman Sundrum. The effect (which is the PhD thesis of student Daniel Kapner) is about 4-sigma, I think, and can be explained either by gravity getting weaker (compared to your Newtonian expectation) or by a new repulsive force, e.g. coupled to lepton number.

  • Aaron Bergman

    4 sigma?!? How confident are they on systematics?

  • Sean

    Remember, they are not claiming a detection of anything, they’re just giving talks on the basis of a PhD thesis experiment. These guys are good, and they won’t claim to have discovered something until they have taken everything apart and chased down every imaginable systematic error. Which hasn’t happened yet, so we shouldn’t get too excited.

  • Quantoken

    When it comes to difficult-to-measure experiments, it is virtually ALWAYS the case that the systematic errors are under-estimated, and NEVER the case that it is over-estimated. There is an incentive to exagerate precision and suppress the error bar, for obvious reasons, especially when it’s virtually guaranteed that theoretists like Sean Carroll would never question it.

    Look at this, the currently presumed value of G is believed at 0.0128% precision. But German and Russian groups obtained something 0.6% and 0.7% higher, and the error bars do not overlap at all. So the original 0.0128% precision was probably exagerated by 100 times. The real precision was probably 1-2%. If my theory is correct, the correct G value should be 2% larger than the current CODATA value. Let’s wait and see.


  • Alejandro Rivero

    Hmm that paper of Sundrum is in the context of string theory, is it?

  • Simon DeDeo

    Sorting modifications of GR into Lorentz-preserving and Lorentz-breaking categories seems like a pretty good way to go. (Although you’d run into trouble when considering the possibility of models that break Lorentz symmetry spontaneously.)

    I remember doing the torsion-balance G measurement a long time ago. That definitely ranks as the most memorable experiment I ever performed. So insane! Things attract each other! Regardless of composition! Even lead balls! No joke!

    As a sidenote, I wonder if all the LIGO work will one day have a side-benefit of being able to do some sensitive G and short-range gravity measurements. I’ve heard that a source of noise for them is gravitational radiation from nearby sources. I know very very little about LIGO, however.

  • bittergradstudent

    about the evolution of the spiral shape–isn’t the important thing the gradient of the angular velocity as you move inward in the galaxy, rather than the angular velocity at the outer part of the galaxy? I can certainly imagine a scenario in which the spiral shape evolves despite a stationary outer rim–so if the inner part of the galaxy has rotated 5 times in the time that the outer part rotated 7, we’d see the 2 spiral pattern that we do.

  • Aaron Bergman

    Sundrum’s stuff, as I remember it, was an attempt to even be able to formulate an effective theory with two scales. String theory didn’t figure in at all. I might be misremembering, though.

  • Alejandro Rivero

    I see, I was misled by the abstract and introduction of hep-ph/9708329, which is one of these ones that try to attach themselves to the mainstream (strings in this case) because orthodoxy is better selling point than heterodoxy nowadays.

  • Aaron Bergman

    I’m guessing that Raman doesn’t feel obligated to “attach” himself to anything. I was thinking of the follow-up paper hep-th/0306106.

  • Vish Subramanian

    “… chased down every imaginable systematic error”

    What about unimaginable ones? I would think that people will (or at least, should) remain sceptical of sub-millimetre deviations until

    i) A compelling theoretical framework appears which explains it.
    ii) An unrelated experiment/observation confirms it.

    As was the case with Michelson-Morley, or the cosmological constant.

  • Mark

    At least in the case of the cosmological constant, I would say that we don’t have a compelling theoretical framework that explains it.

  • Vish Subramanian

    I should say
    i) A compelling theoretical framework appears which explains it.


    ii) An unrelated experiment/observation confirms it.

    Sorry – it seemed clear when I wrote it but not as I read it.

  • Ben Lillie

    It’s not so clear to me that point (i) is very important for whether I’d believe the sub-millimeter deviation. Even if we had a compelling theoretical framework the effect would still need independent confirmation.

  • Vish Subramanian

    I think point (i) comes simply from my reading of the history of special relativity. Not too many people put much stock in one hard-to-perform, subtle experiment (Michelson Morley) until Einstein. Not that people disbelieved the experiment (and of course Lorentz et al came up with theories to explain it), but they just sort of cautiously waited around until someone told them what it meant.

  • Shantanu

    Sean and/or Joanne did anyone in the competition mention
    about the anomolous precession of DI Hercules which if
    I recall correctly does not agree with GR, (though it’s been a while
    since I heard anything about this system and the discrepancy
    may have gone away)

  • Matt McIrvin

    I have my doubts that you could measure the Pioneer anomaly with Cassini, even if it’s new physics and even if Cassini stopped maneuvering. Cassini is orbiting Saturn, which would tend to obscure any anomalous acceleration toward the sun that doesn’t act on planets.

  • Quantoken


    I know about the DI Hercules data. Some Russian physicist proposed a model where there are three stars of about the same order of mass circling each other, although any one who received any decent training in classical mechanics knows any three star system (one all stars have similar masses) are unstable dynamic system.

    My opinion is these kind of observational data is extremely unreliable and the systematic errors are greatly under-estimated. So it is really not very credible whether it is pro or anti GR, that includes the famous PSR 1913.16, which was awarded the Nobel Prize.

    I have yet to find some one who is willing to express an opinion, whether positive one or negative one, regarding my assertion that the Eddington 1919 solar eclipse is simply too coarse in precision to say anything either way about GR. The predicted light bending of about 1 arcsecond or so is just too small to be observed by a 4 inch diameter telescope, not to meantion the atmospheric disturbance known as “seeing”. There is a deafening silence than no one eants to say anything about it.


  • Robert

    If you like you can count all dark matter as a modification of GR, at least until you’ve pinpointed the particle. As long as this is missing, all you are seeing is that galaxies (and other stuff) do not rotate according to Newton’s law (or the GR equivalent).

  • Alejandro Rivero

    Hmm yep Aaron, hep-th/0306106 does better lecture.

    About strings and attachments, consider than in 1997 Sundrum had not yet scored his “home runs”; from his bibliography he could be seen mainly as a young technicolored guy.

    Aside: it is an amusing coincidence if the fat graviton starts to show itself at the same scale that neutrinos are, isn’t it? Smolin did a listing of such coincidences some months ago in Lubos blog.

  • Simon DeDeo

    “if the fat graviton starts to show itself at the same scale that neutrinos are”

    It’s a grand conspiracy of scales!

  • Eric Adelberger

    Please don’t get too excited yet about rumors concerning the Eot-Wash test of the 1/r^2 law. We can exclude gravitational strength (|alpha|=1) Yukawa violations of the 1/r^2 law for lambda>80 microns at 95% confidence. It is true that we are seeing an anomaly at shorter length scales but we have to show first that the anomaly is not some experimental artifact. Then, if it holds up, we have to check if the anomaly is due to new fundamental physics or to some subtle electromagnetic effect that penetrates our conducting shield. We are now checking for experimental artifacts by making a small change to our apparatus that causes a big change in the Newtonian signal but should have essentially no effect on a short-range anomaly. Then we will replace our molybdenum detector ring with an aluminum one. This will reduce any signal from interactions coupled to mass, but will have little effect on subtle electromagnetic backgrounds. These experiments are tricky and measure very small forces. It takes time to get them right. We will not be able to say anything definite about the anomaly for several months at least.

  • Eugene

    As a rational theorist, I will now calm myself down after reading Eric Adelberger’s post.

    (Arrrghghghghgh arghganomalyhgh arrrrrghghghgh).

    Okay, I feel better now.

  • Mark

    Thanks for taking time to give us an update Eric. It’s very valuable.

  • Gordon Chalmers

    Maybe Eric you need a SAR apparatus, with 10^30 detectors.

  • Pingback: Rumors of new forces | Cosmic Variance()

  • Gordon Chalmers

    With the new ideas in condensed matter theory that are arriving, a slab of some meters could it, even with time polyphasing, not your usual STAP.

  • Ivan Alexander

    What if Newton’s G ain’t a “universal” constant, but merely a “variable” constant?

    I know it sounds strange, but think that we computed all astrophysics using a flat G, so if it is variable, we can’t see it. For example, why is Jupiter’s core only 2-3 Earth masses, but it can retain a giant atmosphere? Or the Pioneers Anomaly, at a constant rate of ~8E-8 cm/s^2, which is within range of delta G growing at the rate of 1 G per 1 AU? Or why outer fringe of galaxy acting as if “dark matter” gives it greater mass? Per Equivalence, greater G translates into greater inertial mass (per reference paper in the website linked) so rotation velocity is affected. Would this variable G not be a better explanation? Hypothetically, G would flatten out in space at about 10X-6 N kg^-2 s^-2, which coincides with calculations (per paper above) within range of the photoelectric effect e.m. wavelength. Strange coincidences?

    I think GR will need modifications, likely after ESA tests for gravitational anomalies in the outer system. Stay tuned!


  • Pingback: Now That’s What I’m Talking About! | Cosmic Variance()


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.

See More

Collapse bottom bar