7 ms·

One of my favorite "anomalies" is dark matter. Not much explanation for it so far. One of my favorite quips on HN put it thusly: GR only matches observation be

One of my favorite "anomalies" is dark matter. Not much explanation for it so far. One of my favorite quips on HN put it thusly: GR only matches observation because you added in 20x more stuff that is undetectable other than as a deviation from the predictions of GR. https://news.ycombinator.com/item?id=19537192 https://news.ycombinator.com/item?id=19537192

knzhou 7y ago
That's a witty quip, but it's witty because it's unfair. We don't grade theories based on what fraction of the stuff is obviously visible. If that were true, even electromagnetism would be a terrible theory because the vast majority of the spectrum we consider is invisible to human eyes. What always matters is how well the theory predicts given how simple its assumptions are, and dark matter is great at that.
- ryacko 7y ago
  Dark matter is a fudge factor.
  - knzhou 7y ago
    You can call anything a fudge factor. One person's fudge factor is another's elegant explanation.
    - jacquesm 7y ago
      Fudge factors are natures Magic Constants. There's a #define in there somewhere.
  - dogma1138 7y ago
    Isn’t the cosmological constant the OG fudge factor in GR? AFAIK we need dark matter to explain any theory of gravity, the observations don’t match what we would expect form classical mechanics either which is still the go to theory for things like galaxies. MOND was introduced initially to fix the observations mostly in regards to how we understand Newtonian mechanics, the relativistic versions of MOND can be used as an alternative to GR but these have been more or less debunked with the observations of gravitational waves since afaik all the versions of MOND have “instant gravity” just like Newtonian mechanics.
    - raattgift 7y ago
      "The relativistic versions of MOND" are just adding fields to General Relativity. Thus making them relativistic theories. Famaey and McGaugh §7. https://arxiv.org/abs/1112.3960 https://arxiv.org/abs/1112.3960 Consequently the rest of your last paragraph is incorrect. "we need dark matter to explain any theory of gravity" No. MOND is a theory of gravitation, but it's not relativistic, and it does not work at scales larger than that of galaxies (it is notably wrong with respect to the peculiar motions of galaxies in massive clusters, and a residual mass term must be added ibid. §6.6.4 : essentially, MOND still needs dark matter at galaxy cluster scales, even if it were to correctly describe all the individual galaxies in the massive cluster). Additionally, General Relativity does not require dark matter any more than Newton's F=ma requires dark matter. The issue is that General Relativity's G=T like F=ma does not tell you about the initial trajectories, you plug those in by hand. If you start with the trajectories learned by observing galaxies (as Vera Rubin did), you can work out a stress-energy tensor that satisfies those trajectories -- and the majority of it has to be electromagnetically-uncharged, interacting very weakly or even only gravitationally (i.e., they can't clump or diffuse on the scales of mere millions of years), and slow-moving. If neutrinos weren't so inclined to zip around at speeds very close to that of light, they'd fit nearly perfectly; unfortunately, we're left trying to find the microscopic details of the unknown parts of the stress-energy tensor. Milgrom's MONDian approach to Rubin's discoveries that the orbits of "surface" stars are non-Keplerian was to turn "a" in F=ma into a function that depends on the radial distance from a galaxy's core; the function was found empirically, comparable to how the stress-energy distribution was found. However, it's the dependence on a coordinate distance that makes this approach non-relativistic. (For terseness, in the paragraph above I've discarded some factors, set the constants c and G to unity, and omitted the greek-letter indices on the stress-energy and Einstein tensors (T resp. G). In the MOND context, F=ma is more appropriately written as in the first paragraph of §6 of Famaey & McGaugh.)
      - dogma1138 7y ago
        I'm not sure if I've either explained myself too simplistic or that you haven't understood my post. My points were: 1) DM isn't a fudge factor for GR or Newtonian Mechanics, (Classical) Newtonian Mechanics doesn't have one, PPR does but PPR isn't a theory. GR has a fudge factor built-in into the theory - the Cosmological Constant that gives you a variable that can adjust the predictions to match observations as it wasn't known at the time if the universe is static or not. It's not a perfect fudge factor since it can't deal with accelerated expansion easily but it's a fudge factor nonetheless. 2) MOND was based on classical mechanics and as such isn't a replacement for GR, I'm not sure if I agree with the assertion that TeVeS is an extension of GR. And yes all theories of gravity need "Dark Matter". Dark Matter isn't a fudge factor it's a placeholder for missing mass needed to align predictions with observations doesn't matter if those predictions are derived from GR or Classical Mechanics or as you've mentioned even MOND. And yes MOND still requires "Dark Matter" or to be exact some additional mass however it requires much less of it and it requires it to be concentrated in the center of galaxies which means it can be much more easily explained for through known mechanisms and forms of matter (e.g. the black holes in the center of a galaxy are more massive, higher density of interstellar medium and gas etc.), hence it can be described as a theory that "solves" the problem of Dark Matter, because it doesn't implicitly require new forms of matter or complex explanations for missing mass. However like you've mentioned MOND is a flawed theory, it doesn't even work on galaxies that well since while it can describe their movement today, it has difficulties aligning with observations of various clusters and more importantly it doesn't work well when you start to wind back the arrow of time, and you don't even have to go as far back as to the formation of galaxies (which aren't possible under vanilla MOND) rather it can fail as quickly as by rewinding the clock half a billion years in some cases.
  - gnaritas 7y ago
    No, there are other reasons[1] we believe dark matter exists, it's not just a fudge factor in the math. There are multiple lines of evidence pointing at its existence. [1] https://medium.com/starts-with-a-bang/five-reasons-we-think-dark-matter-exists-a122bd606ba8 https://medium.com/starts-with-a-bang/five-reasons-we-think-...
- K0SM0S 7y ago
  Given how GR is entrenched now, confirmed so many times, it's pretty clear we're at worst in presence of a Newton-degree of understanding: maybe not all of it, but certainly true within boundaries. Whatever lies 'after' GR can only build upon what GR got right. However in the case of dark matter, let's not forget that a model may be incredibly "right" and yet false, e.g. Bohr's atom; still taught today until you enter quantum stuff, but totally ok (valence etc) for basic chemistry. Newton is also totally ok for small non-relativistic speeds and masses, even though the model is ultimately wrong. Dark matter may also 'work' ok to represent and estimate 'basic galactic motion', but may be totally flawed as a model. We would need to test the stuff itself, not its effects, to validate any kind of theory about it.
earthboundkid 7y ago
I can believe in Dark Matter. I can believe in Dark Energy. I cannot believe in both. It’s an epicycle for sure.
- jfengel 7y ago
  Why not? They describe completely different phenomena, two different things we do not understand. There is no reason to think that they might have a single underlying solution, nor is there any theory that points to one. Science can have more than one mystery, with independent leading solutions. It would be like historians declaring that they could believe in the Sea Peoples or Jack the Ripper but not both.
  - earthboundkid 7y ago
    No, it would be like positing that the Ripper killings were done by 5 different people.

knzhou 7y ago

That's a witty quip, but it's witty because it's unfair. We don't grade theories based on what fraction of the stuff is obviously visible. If that were true, even electromagnetism would be a terrible theory because the vast majority of the spectrum we consider is invisible to human eyes. What always matters is how well the theory predicts given how simple its assumptions are, and dark matter is great at that.

ryacko 7y ago
Dark matter is a fudge factor.
- knzhou 7y ago
  You can call anything a fudge factor. One person's fudge factor is another's elegant explanation.
  - jacquesm 7y ago
    Fudge factors are natures Magic Constants. There's a #define in there somewhere.
- dogma1138 7y ago
  Isn’t the cosmological constant the OG fudge factor in GR? AFAIK we need dark matter to explain any theory of gravity, the observations don’t match what we would expect form classical mechanics either which is still the go to theory for things like galaxies. MOND was introduced initially to fix the observations mostly in regards to how we understand Newtonian mechanics, the relativistic versions of MOND can be used as an alternative to GR but these have been more or less debunked with the observations of gravitational waves since afaik all the versions of MOND have “instant gravity” just like Newtonian mechanics.
  - raattgift 7y ago
    "The relativistic versions of MOND" are just adding fields to General Relativity. Thus making them relativistic theories. Famaey and McGaugh §7. https://arxiv.org/abs/1112.3960 https://arxiv.org/abs/1112.3960 Consequently the rest of your last paragraph is incorrect. "we need dark matter to explain any theory of gravity" No. MOND is a theory of gravitation, but it's not relativistic, and it does not work at scales larger than that of galaxies (it is notably wrong with respect to the peculiar motions of galaxies in massive clusters, and a residual mass term must be added ibid. §6.6.4 : essentially, MOND still needs dark matter at galaxy cluster scales, even if it were to correctly describe all the individual galaxies in the massive cluster). Additionally, General Relativity does not require dark matter any more than Newton's F=ma requires dark matter. The issue is that General Relativity's G=T like F=ma does not tell you about the initial trajectories, you plug those in by hand. If you start with the trajectories learned by observing galaxies (as Vera Rubin did), you can work out a stress-energy tensor that satisfies those trajectories -- and the majority of it has to be electromagnetically-uncharged, interacting very weakly or even only gravitationally (i.e., they can't clump or diffuse on the scales of mere millions of years), and slow-moving. If neutrinos weren't so inclined to zip around at speeds very close to that of light, they'd fit nearly perfectly; unfortunately, we're left trying to find the microscopic details of the unknown parts of the stress-energy tensor. Milgrom's MONDian approach to Rubin's discoveries that the orbits of "surface" stars are non-Keplerian was to turn "a" in F=ma into a function that depends on the radial distance from a galaxy's core; the function was found empirically, comparable to how the stress-energy distribution was found. However, it's the dependence on a coordinate distance that makes this approach non-relativistic. (For terseness, in the paragraph above I've discarded some factors, set the constants c and G to unity, and omitted the greek-letter indices on the stress-energy and Einstein tensors (T resp. G). In the MOND context, F=ma is more appropriately written as in the first paragraph of §6 of Famaey & McGaugh.)
    - dogma1138 7y ago
      I'm not sure if I've either explained myself too simplistic or that you haven't understood my post. My points were: 1) DM isn't a fudge factor for GR or Newtonian Mechanics, (Classical) Newtonian Mechanics doesn't have one, PPR does but PPR isn't a theory. GR has a fudge factor built-in into the theory - the Cosmological Constant that gives you a variable that can adjust the predictions to match observations as it wasn't known at the time if the universe is static or not. It's not a perfect fudge factor since it can't deal with accelerated expansion easily but it's a fudge factor nonetheless. 2) MOND was based on classical mechanics and as such isn't a replacement for GR, I'm not sure if I agree with the assertion that TeVeS is an extension of GR. And yes all theories of gravity need "Dark Matter". Dark Matter isn't a fudge factor it's a placeholder for missing mass needed to align predictions with observations doesn't matter if those predictions are derived from GR or Classical Mechanics or as you've mentioned even MOND. And yes MOND still requires "Dark Matter" or to be exact some additional mass however it requires much less of it and it requires it to be concentrated in the center of galaxies which means it can be much more easily explained for through known mechanisms and forms of matter (e.g. the black holes in the center of a galaxy are more massive, higher density of interstellar medium and gas etc.), hence it can be described as a theory that "solves" the problem of Dark Matter, because it doesn't implicitly require new forms of matter or complex explanations for missing mass. However like you've mentioned MOND is a flawed theory, it doesn't even work on galaxies that well since while it can describe their movement today, it has difficulties aligning with observations of various clusters and more importantly it doesn't work well when you start to wind back the arrow of time, and you don't even have to go as far back as to the formation of galaxies (which aren't possible under vanilla MOND) rather it can fail as quickly as by rewinding the clock half a billion years in some cases.
- gnaritas 7y ago
  No, there are other reasons[1] we believe dark matter exists, it's not just a fudge factor in the math. There are multiple lines of evidence pointing at its existence. [1] https://medium.com/starts-with-a-bang/five-reasons-we-think-dark-matter-exists-a122bd606ba8 https://medium.com/starts-with-a-bang/five-reasons-we-think-...
K0SM0S 7y ago
Given how GR is entrenched now, confirmed so many times, it's pretty clear we're at worst in presence of a Newton-degree of understanding: maybe not all of it, but certainly true within boundaries. Whatever lies 'after' GR can only build upon what GR got right. However in the case of dark matter, let's not forget that a model may be incredibly "right" and yet false, e.g. Bohr's atom; still taught today until you enter quantum stuff, but totally ok (valence etc) for basic chemistry. Newton is also totally ok for small non-relativistic speeds and masses, even though the model is ultimately wrong. Dark matter may also 'work' ok to represent and estimate 'basic galactic motion', but may be totally flawed as a model. We would need to test the stuff itself, not its effects, to validate any kind of theory about it.

earthboundkid 7y ago

I can believe in Dark Matter. I can believe in Dark Energy. I cannot believe in both. It’s an epicycle for sure.

jfengel 7y ago
Why not? They describe completely different phenomena, two different things we do not understand. There is no reason to think that they might have a single underlying solution, nor is there any theory that points to one. Science can have more than one mystery, with independent leading solutions. It would be like historians declaring that they could believe in the Sea Peoples or Jack the Ripper but not both.
- earthboundkid 7y ago
  No, it would be like positing that the Ripper killings were done by 5 different people.