Why modifying gravity doesn’t add up

https://bigthink.com/starts-with-a-bang/modifying-gravity/

by raattgift • 4 years ago

96 132 4 years ago

wwarner2• 4 years ago

Another great article from Ethan Siegel. I think he makes a pretty strong case for cold dark matter, but it relies pretty heavily on theoretical cosmological arguments. For the pro-mond argument, there’s a recent interview with Milgrom on Dr Brian Keating’s physics podcast Into the Impossible [0]. There is also a talk about all the ways DM does not fit observations from Pavel Kroupa (University of Bonn) on youtube [1]. In particular, Kroupa argues that when galaxies pass through one another, DM should cause them to move slower after the collision, because DM would be drawn behind each system in a turbulent wake; but no such slowdown is observed. MOND plus visible matter predicts the motion of these systems well.

[0] https://youtu.be/oQ0Cr0hxN40

[1] https://youtu.be/PVgwLWVETIM

andrewflnr6• 4 years ago

Turbulence entirely due to gravitational interaction? Does that work? I guess I should watch it, but I can't take it seriously with the title "On the non-existence of dark matter" when there's so much evidence for dark matter. That's clearly unwarranted confidence in search of attention.

That said, if true, I think it more likely indicates that both MOND and LCDM have significant truth. The bar for a theory is "as simple as possible, but no simpler." If the evidence truly demands both MOND and DM, we just have to suck it up.

wwarner2• 4 years ago

The DM is moved around by the gravity of the moving galaxy. The DM theory is that each galaxy is surrounded by a huge spherical cloud of DM that (by definition) only experiences gravitation and is invisible and insubstantial in every other way. When Voyager gained speed from passing near Jupiter, it caused Jupiter to slow down by a tiny amount. Similarly, when galaxies pass through each other, the DM should take a measurable amount of momentum from the galaxies.

And, yeah, it's a provocative title, but he believes that his measurements falsify DM theory. It's partly a tussle between experimentalists and theoreticians. Kroupa is arguing that the claims of the theory can be tested, and that it doesn't match observations. The theoreticians have a nice story for the big bang, inflation, baryonic nucleosynthesis and the variations seen in the CMB, and the amount of DM it predicts happens to match precisely the amount of DM that we observe.

throwawaymaths11• 4 years ago

Suppose we made an observation that mercury's orbit precesses by a fraction of a degree every orbit. Heretofore every observation of perturbances has correctly presaged the presence of a large mass (a planet) that was quickly found, but this time we have looked and looked but not found anything. Should we presume that it must be an invisible ball of matter that gravitates but we have no other way of detecting except through the same effect we use to infer its existence, or should we tweak Newton's laws of physics?

You can postulate an additional separate modification of gravity that comes into play on cluster scales.

Or it might be something else altogether.

Anyways, MOND has made predictions that people set out to disprove MOND went looking for negative evidence and had to turn around (e.g. external field effect), as well as straightforward predictions JWT may have confirmed (early galaxies) and explains many things that dark matter cannot, like how satellite galaxies don't get ripped apart by the dark matter halo.

tshaddox• 4 years ago

Should we presume that it must be an invisible ball of matter that gravitates but we have no other way of detecting except through the same effect we use to infer its existence, or should we tweak Newton's laws of physics?

Are these substantially different? Of course using “must” isn’t appropriate, but spouting off the best theory you have at the time seems okay to me. Hopefully no one is saying that we cannot possibly ever find another better theory the same way Einstein’s was better than Newton’s or that we shouldn’t ever look for one.

2muchcoffeeman• 4 years ago

I guess the other side effects might be different right?

E.g. we talk about relativistic mass. But when they did experiments it was found that no other attributes that come with increased mass was found. So the answer is right but there is something off with our understanding.

throwawaymaths11• 4 years ago

Hopefully no one is saying that we cannot possibly ever

The OP article might as well. The thrust of the article's argument is, "heaven forbid we add third parameter to MOND" meanwhile, draco, fornax, and Andromeda have wildly different dark matter densities so that's three parameters right there.

lmm• 4 years ago

Should we presume that it must be an invisible ball of matter that gravitates but we have no other way of detecting except through the same effect we use to infer its existence, or should we tweak Newton's laws of physics?

"we have no other way of detecting except through the same effect we use to infer its existence" is true of everything in the universe. We believe stars exist because we can see the light coming from them, but light is our only way of detecting them.

There's no single answer to this kind of question, we have to actually do science and make observations. We should ask whether it behaves like a ball of matter (e.g. does it have a point location in space? Does it orbit the way a ball of batter does?) or like a modified law of gravity (can we reproduce the same conditions)?

The likes of the Bullet Cluster makes a very compelling case that cold dark matter is real.

throwawaymaths11• 4 years ago

"We have no other way of detecting except through the same effect we use to infer its existence" is true of everything in the universe

No. For example, staying within astronomy, we detect gravitational waves and if we find an optical event that correlates with the gravitational waves, it is multimodal (we are at n=1), and if you don't solicit events (or if you statistically blind it by adding fake events in your solicitation) it's multimodal independent (currently n=0). Going outside of astro, you'll find very rich multimodal studies "I know what this molecule is because I have an IR, UV, NMR, Mass spec, etc etc etc"

nlarion• 4 years ago

The main argument of this article seems to fall into the counterfactual fallacy category. Even if MOND has problems and ends up being wrong, that doesn't make dark matter exist.

pdonis11• 4 years ago

should we tweak Newton's laws of physics?

But you can't tweak Newton's laws, because we know Newton's laws aren't the right laws of gravity. The laws of General Relativity make better predictions. So what you would have to tweak would be the laws of GR.

And the basic problem with doing that is that you can't just tweak a single parameter in GR to get the kind of modified behavior that MOND does with Newtonian dynamics. In GR, there simply isn't any such free parameter to tweak. You have to adjust the whole theory. And any such adjustment ends up looking just like dark matter: you have to invent new fields or new forms of stress-energy that we have no other evidence for. And that destroys the whole point of MOND, which was to just modify the laws of gravity without changing our assumptions about what matter and energy is present in the universe. If you're going to have to add new forms of matter and energy anyway, why not just stick to that and keep the laws of gravity the same? That's what the dark matter hypothesis amounts to, and why it's actually the simpler hypothesis once GR is taken into account.

ejolto• 4 years ago

But you can't tweak Newton's laws, because we know Newton's laws aren't the right laws of gravity. The laws of General Relativity make better predictions. So what you would have to tweak would be the laws of GR

The invisible planet OP is talking about is Vulkan[0]. We modified Newtonian gravity to explain the orbit of Mercury (instead of discovering Vulkan), the modification is GR.

OP is implying that maybe we have to modify GR like we did Newtonian physics.

[0] https://en.m.wikipedia.org/wiki/Vulcan_(hypothetical_planet)

shusaku• 4 years ago

Should we presume that it must be an invisible ball of matter that gravitates but we have no other way of detecting except through the same effect we use to infer its existence, or should we tweak Newton's laws of physics?

I mean this is literally how they discovered Neptune? Dark matter isn’t necessarily some magically invisible particle, we’re just trying to count up all the different types of objects in the universe.

russdill5• 4 years ago

I feel like the author is assuming here that the reader knows this is how Neptune was discovered and that the procession of mercury caused by relativity was long thought to be caused by a hidden planet. A very literal example of something thought to be missing matter that instead was a modified Newtonian theory.

russdill5• 4 years ago

Except the evidence of dark matter isn't just about the orbit of a single planet. It goes all the way to predicting the spectrum of the CMB.

throwawaymaths11• 4 years ago

The explanatory power of something that has as many free (statistical) parameters as lcdm is so high it's stunning how many things it doesn't explain. You'll surely get some things in there by chance.

yongjik• 4 years ago

Heretofore every observation of perturbances has correctly presaged the presence of a large mass (a planet) that was quickly found, but this time we have looked and looked but not found anything. Should we presume that it must be an invisible ball of matter that gravitates but we have no other way of detecting except through the same effect we use to infer its existence, or should we tweak Newton's laws of physics?

Note the leap of faith from "a large planet" to "an invisible ball of matter." As one could imagine, given what we knew in the 19th century, "a small, hard-to-observe planet (or a group of small bodies) orbiting very close to the sun" was a totally reasonable hypothesis. "An invisible ball of matter" wasn't (and still isn't). It was natural that physicists never seriously considered such a possibility.

On the other hand, given what we know now, "a diffuse cloud made of particles that do not interact with light" is a totally reasonable hypothesis. In fact we even know one such "invisible" particle: neutrinos. It's not crazy to assume that there are some other kinds of neutrino-like particles that fit the bill.

atoav• 4 years ago

I mean a primordial black hole with the size of a tennis ball would have 5 to 15 earth masses. It would not be invisible, but at that size it would be invisible to us even if it was in our solar vicinity.

If primordial black holes are things that exist, one could reason that they might be an explaination for masses we cannot see maybe?

naasking12• 4 years ago

On the other hand, given what we know now, "a diffuse cloud made of particles that do not interact with light" is a totally reasonable hypothesis.

Was a reasonable hypothesis. Given the evidence so far, it's now very debatable. Just like another planet was a reasonable initial hypothesis until it wasn't.

gizmo6862• 4 years ago

Notably, Mercury was not a motivation for the development of relativity. Even with the benefit of hindsight, I don't recall ever seeing anyone critisizing contemporary astronomers for looking for the missing planet. GR was developed based on theoretical concerns. The fact that it happended to solve an open issue of Mercury's orbit was a nice confirmation that it is on the right track.

The good news is that we have a lot of physicists looking for an alternative theory of gravity called quantum gravity. It would be great if they come up with something that solves the theoretical problems with GR and happens to resolve dark matter.

wumpus4• 4 years ago

It's a shame that a crackpot comment is at the top of this discussion.

Vulcan was looked for and didn't exist. Einstein won that one. I'm unaware of any MOND theory which changes the orbit of Mercury.

aeneasmackenzie• 4 years ago

Thats the exact point the comment is making. The moNd theory you’re looking for here is GR. We don’t think of it as a modified theory anymore though.

naasking12• 4 years ago

Yes, Vulcan was looked for just like DM has been looked for. The evidence for DM is largely indirect, just like the evidence for Vulcan. The OP's comment is a good summary of where we stand now. I suggest comparing all of the evidence for DM and MOND:

From Galactic Bars to the Hubble Tension: Weighing Up the Astrophysical Evidence for Milgromian Gravity, https://www.mdpi.com/2073-8994/14/7/1331/htm

There is definitely something to MOND. It would be a remarkable coincidence if MOND was able to make accurate predictions with only a single parameter, visible mass, what takes DM multiple parameters and a tuned mass distribution specific to each observable galaxy. DM definitely "explains more" in some sense because of how much more work it's received, but that's not saying much.

sylware• 4 years ago

Aren't there several MOND flavors?

Heard there is one which is very good at comisc scale.

At "our" scale, GR is accurate "enough" as far as my limited knowledge is.

hinkley2• 4 years ago

My top, crackpot theory is we live in a black hole and dark matter is the accretion disk.

My second crackpot theory is we are four dimensional beings living in a five dimensional universe. Though I also like Cixin Liu’s idea that we are living in a universe that used to be five dimensions and has (mostly) decayed to four, with bits of five dimensional debris holding on.

pryelluw3• 4 years ago

You mean there could be black holes inside other black holes?

Filligree2• 4 years ago

Mathematically, the universe looks a little bit like a black hole running backwards in time. But only a little bit.

hinkley2• 4 years ago

We don’t exactly know what the inside of a black hole looks like. It’s infinitely curved space so it’s potentially quite big. One theory is the space inside a black hole expands at the speed of light. You can’t leave because you can’t go fast enough to get out.

We know that if you have a big enough black hole that the space near the singularity no longer shreds matter. But would it matter if a sun came through intact or a sun’s worth of material came in, condensed, and went supernova?

https://www.nationalgeographic.com/science/article/140218-bl...

vyrotek2• 4 years ago

we are four dimensional beings living in a five dimensional universe

I think I prescribe to that theory. It feels like it's also a fancy way of saying we live in a simulation. Everything we experience is a puppet/shadow/extension of its true "full" self. Things could be connected in ways we can't measure in our traditional dimensions.

pyinstallwoes• 4 years ago

How do you think about yourself in such a way? That you live a limited existence of a bigger mind that experiences more things? Like you'd be living in the "left" brain of a mind and only see the "left brain things?"

If you experience a "shadow", what is the thing/being responsible for the shadow doing?

How does this help you think or rationalize about existence?

I'm asking from a purely curious place. I've come to similar perspectives but I kind of get stuck with "I exist here, and I only experience this reality."

The only thing that has really gone beyond the above line of thinking is when contemplating what the "mind" is. It seems to be the only thing truly beyond "space-time" and limits of "speed of light." Your mind can imagine any point, and travel there one way or another. It's a bit of a logical jump, but it does break the limits of how fast one can travel in some type of definition. So the mind is kind of the "4d/5d being" and everything else is the 3d/4d.

ars• 4 years ago

Something I've always wondered: Right after the Big Bang there was a massive annihilation of matter and anti-matter, with just a bit of extra matter.

But the energy resulting from that annihilation did not go away. So where is it? No matter what form it takes, energy gravitates, so it should be able to be observed.

frankchn• 4 years ago

My impression (I might well be wrong about this) is that the energy from that forms part of the cosmic microwave background radiation.

exmadscientist• 4 years ago

A lot of it just went back into producing particle-antiparticle pairs again. Some of it is the CMB, yes, but not all.

Why we didn't just end up with even numbers of particles and antiparticles repeating this dance endlessly is the single biggest open question in particle physics.

beeforpork• 4 years ago

Photons. Each annihilation event emitted a photon, and they are still around. IIRC, about 5e9 times as many photons are observed as there are baryons, and that seems to match up well with the calculated imbalance of the baryosynthesis.

debacle• 4 years ago

We are in the "epicycles" phase of physics discovery, at both the micro and macro level.

I may not live to see it, but I'm excited for when everything just "clicks" and the model becomes more simple and elegant, as it was before the 1920s.

MichaelZuo2• 4 years ago

That's an intuitive way to put it, ironic considering epicycles themselves are viewed as the opposite.

wumpus4• 4 years ago

When you run into physics/astronomy people mentioning epicycles, it's not ironic, it's saying that we're missing out on a simpler rule.

colechristensen• 4 years ago

My fun, science fiction theory to solve the gravity problem is that the universe is actually full of intelligent life mostly billions of years old civilizations and all of the extra mass we can’t see is actually the galactic version of cities built from materials that don’t interact with light much. Likewise we haven’t met any other intelligent species because the growth period before reaching this common technology is short from thousands to millions of years so there just aren’t collisions in space or time where two intelligent species would interact.

russdill5• 4 years ago

Except an important property of dark matter is they it's interacts weakly with itself. But still good for sci-fi

Beldin4• 4 years ago

If that is true, where are the dark matter black holes? And why hasn't the majority of DM collapsed into black holes?

rapjr9• 4 years ago

Maybe dark matter does have electromagnetic interactions with normal matter, but they are at an exceedingly high (or low) frequency. Science has limits in what frequencies it can probe, matter may not have the same limitations. A century long EM wave would look like background noise on a human scale. Petahertz (and beyond) vibrations could have extremely short range interactions that mostly occur in the near field.

Beldin4• 4 years ago

I've got two problems with this.

First: if theory and observation don't match up, that can be due to lacking theory or lacking observations. Both should be investigated.

Second: all candidates for dark matter i know posit an enormous number of places where this dark matter is.

It is not that "dark matter solves this problem". It is a particular distributing of mass that solves the problem - typically, a highly specific distribution of mass. So not only is this theory supposing that most material in the universe is invisible to us, but its properties other than mass are unknown to us except it allows us to fill in the blanks exactly as needed.

E.g, we're not expecting a lot of dark matter black holes, as the hypothesised stuff doesn't even interact with itself and so cannot slow down. Basically, "dark matter" is fairy dust that physicists sprinkle over everything till it works.

And you know what: maybe that is the way to go. Probably it plays at least some part in answering these questions. But until we have a clearer picture, we may as well call it "magical fairy dust". That at least reminds us how much we're fudging.

pfortuny• 4 years ago

Until it is experimentally tested, DM is essentially equivalent to the ether. A useful unexplained tool which to date seems to work.

Until another Michelson-Morley experiment comes (and MOND might have something to say, or at least they seem to be testing DM experimentally), yes: as untestable as string theory is.

DiogenesKynikos9• 4 years ago

It is a particular distributing of mass that solves the problem - typically, a highly specific distribution of mass

It's actually an extremely simple initial distribution of dark matter in the early universe. You can specify the entire distribution with one number: a spectral index. The thing is, we already know that the universe had that distribution of energy density anyways, because of the Cosmic Microwave Background.

If you take that very simple distribution and let the laws of physics run for 13 billion years, you get the required present-day distribution of dark matter. That distribution can be verified in several ways, including gravitational lensing, the rotation curves of galaxies, the motions of galaxies inside clusters, and the temperatures of hot gas inside galaxy clusters.

The fact that such a simple set of initial conditions can explain pretty much everything we can see on scales ranging from stars to the entire visible universe is why cold dark matter is overwhelmingly the favored paradigm.

bowsamic• 4 years ago

Yes but dark matter is explanatory while MOND is predictive.

We see a set of problems, we fix them by choosing our unobservable dark matter halo in order to fix those problems. That should automatically be extremely suspicious for any physicist. On the other hand, MOND predicts the correct galaxy rotation curves of galaxies we have not looked at yet.

Sure DM explains more than MOND, but it is also basically not a scientific theory, since it has far too many degrees of freedom that can explain away any problem and is extremely difficult to falsify.

rini172• 4 years ago

If you have to repeatedly modify "predictive" theory to match observation, then it was not predictive in first place. The situation with MOND appears much worse than for example Planck's theory for blackbody radiation. He revised it once and that was it.

Generally I don't get this splitting hairs. Things that we can only observe but can't reproduce in lab are not a science then?

DiogenesKynikos9• 4 years ago

Lambda CDM makes very few assumptions. It only has 6 free parameters.[0] If you start with those 6 parameters, and then let the laws of physics work, you get the universe we see today. There are a huge number of different types of phenomena that the theory has to correctly predict (e.g., the abundance of different elements, the numbers of galaxies with different masses, the rotation curves of galaxies), and it does so extremely well.

0. https://en.wikipedia.org/wiki/Lambda-CDM_model#Parameters

naasking12• 4 years ago

Lambda CDM makes very few assumptions. It only has 6 free parameters.[0] If you start with those 6 parameters, and then let the laws of physics work, you get the universe we see today.

Not including the DM distribution needed to match each galaxy's rotation curve, or external field effects, and dozens more observations:

From Galactic Bars to the Hubble Tension: Weighing Up the Astrophysical Evidence for Milgromian Gravity, https://www.mdpi.com/2073-8994/14/7/1331/htm

Contrast that with MOND's 1 parameter, visible mass.

LatteLazy• 4 years ago

I thought the problem with modifying gravity was that there was no one modification you could make to explain observations. It's plausible that 1000s of different galaxies spin differently because they each have different amount of mass of dark matter in them. But it's not really possible to explain 1000s of different data points without 1000s of different gravity equations. If there were some rule (eg on large scales gravity is actually an exponential) then someone would have found it and we would have a candidate for a new law. But they haven't...

Is that wrong?

naasking12• 4 years ago

It's plausible that 1000s of different galaxies spin differently because they each have different amount of mass of dark matter in them.

You also need a plausible explanation why all of these galaxies have such different distributions even though they have similar visible mass. If DM is only supposed to interact gravitationally then there should be a tight correspondence with little deviation, but there are galaxies entirely free of DM and distributions that differ across visible mass. There are also observations that don't even match DM (external field effects, tidal effects, and more). There's clearly a lot more to the story.

Tepix• 4 years ago

Looking back at this episode - after we figure this out - will be similar to looking back at the period before the discovery of general relativity. Right now, we're really quite clueless.

twawaaay2• 4 years ago

It is epicycles all over again.

I think modifying gravity might be useful.

Probably people who were slapping more terms to epicycles knew what they were doing is wrong but did it anyway because it allowed them to predict what they needed.

fallingknife• 4 years ago

Isn't dark matter an epicycle?

twawaaay2• 4 years ago

I am not a physicist, but I think it is not.

It seems they can account for mass in various places in universe, they can model how things should be moving (being affected by gravity), they can measure how things are actually moving and from this they can infer where this missing mass actually is placed.

This is currently all pretty early stage but the resolution is expected to increase with time.

So no, I don't think dark matter is an epicycle. It is more or less like the case of gravitational disturbance affecting planets in solar system and somebody theorising a planet placed somewhere would explain the effect, we just can't yet measure things accurately enough to tell where the planet is exactly.

Why modifying gravity doesn’t add up

Related Stories