Hacker News new | past | comments | ask | show | jobs | submit login
Gravity: We might have been getting it wrong (phys.org)
137 points by ryan_j_naughton 24 days ago | hide | past | web | favorite | 87 comments



Agh, mildly infuriating (enfrustrating?) article. A little tease about the result and then several paragraphs about how science is serendipitous and face to face interactions are better than email and video conferencing.


> mildly infuriating (enfrustrating?) article

Par for the course for phys.org.


Because like every article I tried to find was pretty much the exact same rubbish overview, here's a link to the pre-print paper: https://arxiv.org/pdf/1810.05337.pdf

As someone very not versed in physics, the paper is actually... really readable so far and short (I'm almost done). At the very least, there is more information than anything mentioned in the links/summaries I could find.


The real paper is https://arxiv.org/pdf/1810.05338.pdf

1810.05337 is a summary of results for PRL

Both are very good


Eternal plea: for those of us who want to see what the paper's about before downloading it, it's better to link to the abstract at http://arxiv.org/abs/1810.05338 .


Even the title is wrong, we know for a fact that we are getting gravity wrong.

A better article with actual content:

https://www.ipmu.jp/en/20190619-symmetry


> we know for a fact that we are getting gravity wrong

This statement is way, way too strong. We have a theory of gravity, General Relativity, which has had its predictions confirmed by countless experiments, in some cases to thirteen or fourteen decimal places. That is not "getting it wrong".

Many physicists believe that GR, as well confirmed as it is, can't be the final fundamental theory of gravity because it's not a quantum theory. But nobody has yet found a quantum theory of gravity that (a) is consistent with the rest of quantum mechanics, and (b) makes experimental predictions that have been confirmed. Unless and until both (a) and (b) are done, the claim that we are "getting it wrong", even in the restricted sense of not having a final fundamental theory of gravity, because GR isn't a quantum theory is just speculation and might itself be wrong. It should certainly not be stated with the confidence you are stating it.


> Many physicists believe that GR, as well confirmed as it is, can't be the final fundamental theory of gravity because it's not a quantum theory. But nobody has yet found a quantum theory of gravity that (a) is consistent with the rest of quantum mechanics

It was my (lay) impression that we believe general relativity cannot be the final fundemental theory of gravity because general relativity itself is not consistent with quantum mechanics. Whether general relativity is a quantum theory or not is beside the point; we want to have a theory -- any theory -- that is consistent with itself. Right now we have two separate theories (general relativity / quantum mechanics) that can't be combined because they conflict with each other.

Obviously, that in itself doesn't prove that the problem is in the theory of gravity, but it does prove the problem is either in the theory of gravity or the theory of quantum mechanics, and we have to be getting one of them wrong.


> It was my (lay) impression that we believe general relativity cannot be the final fundemental theory of gravity because general relativity itself is not consistent with quantum mechanics.

That's the same as "not a quantum theory", which is what I was saying. "Consistent with quantum mechanics" requires "is a quantum theory", according to the many physicists I was referring to.

> Right now we have two separate theories (general relativity / quantum mechanics) that can't be combined because they conflict with each other.

Most physicists believe that they conflict, yes. But not all. Freeman Dyson, for example, has speculated that gravity might just be different and not require a quantum theory the way the other interactions do.


> the problem is either in the theory of gravity or the theory of quantum mechanics, and we have to be getting one of them wrong.

Not true: we could also be getting them both wrong.


There is no way to be getting both of them wrong but not be getting one of them wrong.


Pedantically that is true, but you should still word it as "at least one of them" to avoid being misunderstood.


Now, why am I seeing this factual comment in grey?

See here for instance a popsci take on the problem: https://www.scientificamerican.com/article/is-gravity-quantu...

Everybody expects a quantum theory of gravitation, but if we were drawing it from scratch no one would make it that weak. It is a difficult problem, people work in string theory because you can get Einstein's equations from strings and the alternatives look indeed worse. I don't think that means GR is a wrong take on gravitation, it's more like an amazing one to begin with.


It's grey because someone in their infinite wisdom downvoted you.

Most likely because they'd rather downvote rather than challenge your statement. It's a worrying trend.


>(a) is consistent with the rest of quantum mechanics

i think we're on a right way in general :

QM -> GR : https://en.wikipedia.org/wiki/Quantum_gravity#Quantum_mechan...

" In particular, contrary to the popular claim that quantum mechanics and general relativity are fundamentally incompatible, one can demonstrate that the structure of general relativity essentially follows inevitably from the quantum mechanics of interacting theoretical spin-2 massless particles (called gravitons)"

GR -> QM : We did observed gravitational waves of GR. A force field with waves mediating that force interactions - that requires just a bit of math to "quantize" it. It is a bit more complicated when classic EM -> QM because of added complication of non-stationary spacetime and just vastly different scales of spacetime/energy to run the experiments to establish/verify the gravitons parameters.

For the "true deep" nature of graviton I subscribe to the view that - given gravitational force not being the "real" force and the gravitational waves being really a ripples of spacetime - graviton is more like phonon, ie. quasi-particle from the point of view of QM/SR theories.


> one can demonstrate that the structure of general relativity essentially follows inevitably from the quantum mechanics of interacting theoretical spin-2 massless particles (called gravitons)

While this is technically true, it doesn't help much, since the quantum theory described is not considered a viable candidate for a fundamental theory of gravity for a number of reasons.

> We did observed gravitational waves of GR. A force field with waves mediating that force interactions

We have observed classical gravitational waves. We are many, many orders of magnitude away from being able to measure any possible quantum properties of those waves.

> that requires just a bit of math to "quantize" it

Math (more than "a bit") can quantize it in theory. But theory is not enough. We would actually have to observe the quantum properties experimentally to confirm the theory.


>We would actually have to observe the quantum properties experimentally to confirm the theory.

While not directly quantum properties of gravitational waves yet, we've already observed quantization of gravitational potential (and that naturally suggests "gravitational quantas" at least by similar machinery as with EM->QM) - at small scales/energies when the spacetime curvature can be not paid attention to, the gravity potential of an ultra-cold neutron quantizes just nice using the Schrodinger equation in full agreement with experimental observation - https://www.physi.uni-heidelberg.de/Publications/dipl_krantz... . In my view, giving the amount of clarity in gravitation as well as in QM that experiment produced, it is worth of a Nobel.


> we've already observed quantization of gravitational potential

No, we haven't. We've observed that, under appropriate conditions, gravitational potential has to be included in the potential term in the Hamiltonian. But the potential term in the Hamiltonian is not quantized; it's not part of the quantum state and it does not exhibit quantum properties. The only things in the experiment that exhibit quantum properties are the neutrons themselves.


>gravitational potential ... it does not exhibit quantum properties. The only things in the experiment that exhibit quantum properties are the neutrons themselves.

It is like saying that discrete orbits of electron say nothing about quantization of associated EM potential.

>The only things in the experiment that exhibit quantum properties are the neutrons themselves.

And those properties are position and momentum.

They specifically chosen neutrons to avoid effect of other forces and the paper is pretty clear that position and motion of neutrons were primarily affected by gravitational field, and as result the neutron position got quantized :

"... we conclude that the measurement manifests strong evidence for quantisation of motion in the gravitational field as is expected from quantum mechanics

...

the quantum theory we derived neglecting gravity is unable to reproduce the shape of the neutron height distribution even on the largest scale.

..."

Do you think the height of observed "steps" in neutron position would be the same or different if the experiment were repeated in different gravity, say Moon?


> It is like saying that discrete orbits of electron say nothing about quantization of associated EM potential.

That's right, they don't. There is no such thing as "quantization of EM potential". The electromagnetic field is quantized, if you're using quantum field theory, but in QFT there is no "EM potential". In non-relativistic QM, the EM potential term in the Hamiltonian is not quantized and does not exhibit quantum properties. Even in models which add relativistic corrections, such as the models used in the late 1940s to predict the Lamb shift, the "quantization of the EM field" only applies to the "quantum electromagnetic field" external to the atom; it does not apply to the EM potential due to the nucleus.

> the paper is pretty clear that position and motion of neutrons were primarily affected by gravitational field, and as result the neutron position got quantized

No, the neutron positions were not quantized "as a result" of the gravitational field. They are already quantized in the absence of a gravitational field, as shown by numerous previous experiments. The only thing this experiment shows, as I've already said, is that under appropriate conditions, you have to include the gravitational potential in the potential term in the Hamiltonian.


> Do you think the height of observed "steps" in neutron position would be the same or different if the experiment were repeated in different gravity, say Moon?

Obviously since the gravitational potential would be different, the neutron position "steps" would be different. That still doesn't change the fact that the experiment is not showing "quantized gravitational potential". It's only showing that "quantized neutron position", which is already well established by other experiments, is affected by the gravitational potential in the same way as it would be by any other classical potential in the Hamiltonian.


We also have the flyby anomaly where multiple space probes flying past the earth do not exactly match the predictions of GR. So is GR therefore falsified?

https://en.wikipedia.org/wiki/Flyby_anomaly


The predictions are based on what we know is out there, and how we would expect it all to react with us simultaneously. So as long as we have it exactly understood and haven't missed anything whatsoever, then yes, it would be therefore falsified. But we can't make such a claim.

I haven't really seen science work that way, since it isn't all that often we base our expectations solely on things we can completely account for. Science necessitates a certain amount of calculated projection. Theories are adjusted as new details are discovered, and only fools entertain themselves by resisting the bigger leaps needed to make actual discoveries.


I'm not quite sure what you meant in the 2nd para. But I find it interesting that multiple probes showed the same effect. It's so compelling I wish they would send some missions specifically to test this even more precisely.

The obvious parallel was https://en.wikipedia.org/wiki/Pioneer_anomaly which was finally solved. But who know at this point?

Personally I think GR has to be "wrong", since it conflicts with QM, and if the 20th century showed us anything it's that QM alway wins.


There could be unaccounted mass (dark matter) or magnified sensor error, just off the top of my head. Lots of ways to interpret this result.


>Lots of ways to interpret this result.

rounding error in The Grand Simulation.


I meant evidence based interpretation.


GR is no more or less wrong than Newtonian physics.


IANAP but I’m reasonably confident in asserting that GR is less wrong than Newtonian gravity, based on each models’ predictive power.


> GR is no more or less wrong than Newtonian physics.

Then why does GR make more accurate predictions?


At low speeds relative to the speed of light, they have the same accuracy don't they? And in that case, the Newton model is simpler so more useful. At high speeds, it becomes inaccurate and then you need the other model.

So there is no global "right" or "wrong", just different measures of usefulness depending the problem constraints.


> At low speeds relative to the speed of light, they have the same accuracy don't they?

GR is always more accurate, but the increased accuracy is not always needed; at low speeds relative to the speed of light the difference is often too small to matter in practical terms.


> there is no global "right" or "wrong"

You are wrong, so there is.


> At low speeds relative to the speed of light, they have the same accuracy don't they?

When measuring signal (ie radio, light, gravity) from slow speed perspectives, newtonian physics is not as accurate - https://futurism.com/newtonian-physics-vs-special-realtivity - Eg we actually need to account for it between gps sattelites and we do so from the perspective of GR as well as the recent evidence of gravity propogation being restricted to localization (ie roughly the speed of light)


Yup. And when you need to measure where artillery shells land, the GR model is overkill and you'll waste less time by using a Newtonian prediction. I think this proves my point?


"Newtonion physics is a perfectly acceptable approximation in many circumstances" and "GR is less wrong than Newtonion physics" are not in conflict. Newtonion physics can absolutely be "more wrong" and "useful in many circumstances" at the same time- and in fact is!


If your point is that they're both equally wrong, then no, I think you've disproven your own point. There are no circumstances in which Newtonian math produces better results than GR. That the latter is more complex is not relevant to the discussion of whether one is more accurate for a wider range of phenomena than the other.


> I think this proves my point?

No, because "less accurate but in practical terms the difference can often be ignored" is not the same as "the same accuracy".


So GR is more right in more scenarios and that makes it no more or less wrong?


i am a big fan of newtonian mechanics


Besides the valid points of the other commenters, I have a nitpick: It is special relativity that deals with things near the speed of light, you do not need GR for it. GR is necessary to account for strong gravitational fields.


TLDR:

  [...] the holographic principle allows physicists to study gravitational systems by projecting them on a boundary

  [...] proved that symmetry is not possible in a gravitational theory if it obeys the holographic principle.

  [...] meaning that symmetry would not be possible in quantum gravity.


PBS Space Time have done a number of episodes on the holographic principle[1]

I strongly recommend anyone interested in physics check out the channel

[1] https://www.youtube.com/watch?v=klpDHn8viX8


By inference. this means we trust the holographic principle so thoroughly that we don't even consider that it could be the unbalanced part of this equation.

Is that true?


AdS/CFT fell out of string theory. It's only holographic in the sense that in some circumstances you can summarise an n=dimensional system using n-1 dimensions - not in the literal sense of using the kind of optical transforms that make visual holography possible, but in the more abstract sense that all of the information in some spaces ends up on a notional surrounding boundary.

The AdS universe is basically a mathematical toy for physicists to play with. It doesn't correspond with the physical structure of the real universe. There's a hope that it might correspond to some kind of mathematical meta-structure that defines physical laws.

This article suggests that's very unlikely.

Personally I wouldn't be surprised if AdS/CFT turned out to be a dead end - probably like supersymmetry - and the real structure of quantum gravity turned out to be completely different and unexpected, and not a Conformal Field Theory at all.

But IANAP, so I may well be wrong about that.


For people on a mobile device:

> TLDR:

> the holographic principle allows physicists to study gravitational systems by projecting them on a boundary

> proved that symmetry is not possible in a gravitational theory if it obeys the holographic principle.

> meaning that symmetry would not be possible in quantum gravity.


... if it obeys the holographic principle.


Just because quantum physicists can't get their act together over gravitation doesn't make General Relativity wrong!

When the lay person reads headlines like this, they assume you are talking about Einstein's theory.


gravity sucks. thanks for the link :)


I always thought gravity was much more attractive.


Does this account for the subject of "dark matter", which at this point is absent from all evidence?


No one would have come up with a hypothesis like dark matter without evidence pointing in that direction: https://en.wikipedia.org/wiki/Dark_matter#Observational_evid...


Dark matter was hypothesized because the universe is expanding faster than our understanding of gravity allows. The problem in that is if the universe is expanding so fast then something else must be holding galaxies together since galaxies are so incredibly vast and yet contain so very little matter.

Particle physics is currently defined according to the Standard Model. Dark matter was proposed because it plugs many of the gaps of universal expansion in contrast to the Standard Model, but not all the gaps. It would take far greater effort to invalidate the Standard Model and all of particular physics built upon it. There is also the problem of not having any other model to replace the current one. https://en.m.wikipedia.org/wiki/Standard_Model

The expansion of the universe is observed as red shifts and these observations are numerous and largely uniform. Dark matter, on the other hand, has not been directly observed.

* https://arstechnica.com/features/2014/07/dark-matter-makes-u...

* https://arstechnica.com/science/2019/11/dark-matter-link-to-...

* https://arstechnica.com/science/2019/02/more-bad-news-for-co...

* https://arstechnica.com/science/2019/04/one-night-of-telesco...

* https://arstechnica.com/science/2017/02/a-history-of-dark-ma...

* https://arstechnica.com/science/2016/07/dark-matter-still-mi...


Seems like you confuse dark energy (needed to explain the universe expansion acceleration) with dark matter (needed to correct for the galaxy rotation curve)


> Its presence is implied in a variety of astrophysical observations, including gravitational effects that cannot be explained by accepted theories of gravity unless more matter is present than can be seen.

https://en.m.wikipedia.org/wiki/Dark_matter

Even Wikipedia says dark matter is a stop gap of gravitational models. The very next sentence, first paragraph, says experts believe dark matter exists for that reason.

> Without introducing a new form of energy, there was no way to explain how an accelerating universe could be measured.

https://en.m.wikipedia.org/wiki/Dark_energy

Dark energy is an unobserved stop gap as well. Dark energy was introduced because dark matter was not enough of a stop gap to qualify the acceleration of the expansion.

The reality is that there is just as much theoretical qualifiers for the presence of dark matter as there are unexplained phenomena in the standard model that needs dark matter as a qualifier.

The best possible answer to this nonsense is that the standard model is wrong because we don’t have the evidence to form a better model. All our evidence is limited to observations from Earth surface and low orbit. Hopefully the James Webb telescope will provide better evidence.


All you're saying here is that dark matter and dark energy are evidence-based. The ideas would not exist if not for the evidence, as there wouldn't be any reason to try and think of explanations and theory modifications. This completely contradicts your original statement, so I'm not sure what point you're trying to make.


That is the opposite of what I am saying and the opposite of the various links and quotes I have provided. They do not exist due to evidence but because the standard model demands their existence.


Much more information from a previous phys.org:

https://phys.org/news/2019-05-constraints-symmetries-hologra...

Their work assumes the AdS/CFT correspondence.

"Our new paper provides a rigorous proof of this claim in the context of the AdS/CFT correspondence, where quantum gravity is defined in a mathematically precise way, and we have done so in the most general way, excluding all possible global symmetries from quantum gravity,"


Interesting ... another time holography in the context of physics has come up. Youtube recommended the Wired series where some expert explains some concept at different levels and one recent one was on Gravity [1]. In the expert section they talk briefly about holography and how a two dimensional surface seems to be able to encode all of the information of a three dimensional volume. They go on to speculate how this might relate to the event horizon of black holes. I get a bit lost since it is heavily edited and they are clearly talking about things beyond my ken. My takeaway was that there is a possibility the information we thought was getting lost within blackholes is actually still present within the surface of the event horizon.

There was also a brief mention of holography in this interview with String Theorist Jim Gates on Lex Fridman's Youtube channel [2]. I can't remember the exact context of how holography is related to strings but I believe he was talking about how it could impact some of the maths behind the multiple dimensions some string theories required.

[1] https://www.youtube.com/watch?v=QcUey-DVYjk

[2] https://www.youtube.com/watch?v=vGQ3q4dO_9s


this article is way to vague. Do they means symmetry as it relates to string theory or some other theory for quantum gravity? I was hoping for more info.


Modeling, even if incorrect, is important for at least 2 reasons. The most obvious one is that we can refine the theory as new results come out of the experiments. The second is that a successful model, even if wrong for the experiment it was devised, will eventually shed light on a similar problem elsewhere.


If you talk to physicists and they tell you we've got something right, stop talking to them.

Every good(!) one knows that we got it wrong. Just a little less wrong than in the previous model. You cannot get it right. It wouldn't be a model but reality as a whole if you got it right. So the only questions are how you can get it less wrong than the last model, and which of the available models is most useful for the problem you are currently tackling. So yes, sometimes you even want to use a more-wrong model because it fits your problem better.

edit: thought about bringing in an example, but honestly it would need too much googling to be precise enough.


> sometimes you even want to use a more-wrong model because it fits your problem better.

Or they are equal in fit, but the more wrong is easier to use.


This part is fun: "Their result has several important consequences. In particular, it predicts that the protons are unstable against decaying into other elementary particles, and that magnetic monopoles exist."


That caught my eye too. Magnetic monopoles would be very useful, if we could find them.


What would you do with a magnetic monopole? I might make this an interview question at work. Hmmm


Make an infinite energy machine by having magnets on a rotor, surrounded by monopoles on a stator (or vice versa), so the rotor is infinitely accelerated with no external input energy. Hook it up to an alternator and harness the free power.


I'm pretty sure that doesn't work for the same reason that replacing monopoles with charged objects (which we can actually build!) doesn't produce a perpetual motion machine.


Huh, interesting. We know there are (or were) some fundamental asymmetries in the universe, leading to, for example, the imbalance in matter vs anti-matter. I wonder if this research will lead to anything in terms of understanding why these kinds of asymmetries exist?


An interesting question is also why the universe has any kind of structure and is not uniform.


See [0] for the actual paper mentioned in the article.

[0] https://arxiv.org/abs/1810.05338


Huh. We said the same thing a bit over a century ago.


If we can't figure out gravity, what hope have we of doing ANY science? I mean stuff falls. Everything falls. The orbits of the planets aren't that complex, but gravity pretty much predicts them close to perfectly.


It's not that we can't model gravity at planetary-orbit scales; that's been done. The issue is that relativity (which addresses gravity) is a good model at large scales and the quantum "Standard Model" (which addresses the other three forces) is a good model at small scales, but these two models are incompatible with each other and no falsifiable theory that works at all scales has yet been found.


They cannot be "incompatible with each other," otherwise Nature would be "incompatible" with itself.


The models are incompatible with each other, which means that one or both are somehow wrong, or there is some boundary we can't fathom between the small scale and the large.


Wrong - we do know that the models are not incompatible. (There exists a consistent mathematical framework that includes both.) Me, I was referring to a simpler fact that these models are both pretty damn close to what Nature itself tells us.


Models are not nature.


We have a working model of gravity that is very accurate in an incredibly broad array of circumstances. You acknowledge this in your comment.

This is a scientific result. What are you trying to say?


I don't understand a lot about how the computer in front of me works (ok, I know more than the average person, but still I don't know a whole lot). Yet I'm still able to use it.

We understand enough of the effects gravity to describe most phenomena that are relevant for most other parts of science. From my lay understanding what physicists are missing is how to tie in gravity with all the other theories they believe are true. However the effects, whatever they are, are so small that we can in most cases just ignore them.


this pop sci article does address that: we have a good theory of gravity at macroscopic scale but none at the quantum level.


"Science is magic that works."

-K. Vonnegut


There'also Clarke's Third Law.


How does gravity impact photons and quarks? That’s the part that’s missing.


How? If I understand the question - just like everything else. (Bends light, for example.)


yess, but then the numbers that power quantum physics stop working. something about the renormalization group?

My incorrect understanding is something like: quantum models imply a uniform layer of “virtual” particle/anti-particle pairs popping in and out of existence. but if the space they’re in is curved like general relativity describes, then the pairs don’t find each and annihilate each other reliably and all your math trends towards infinity instead of towards zero. or something vaguely like that.


I'm still trying to recover from learning that gravity waves can be curved by gravity.




Applications are open for YC Summer 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: