Follow the reluctant adventures in the life of a Welsh astrophysicist sent around the world for some reason, wherein I photograph potatoes and destroy galaxies in the name of science. And don't forget about my website,

Tuesday, 11 August 2015

The Universe Is A Great Big Bloody Mess And You Won't Convince Me Otherwise

Last week, Moti "Modified Newtonian Dynamics" (MOND) Milgrom visited Prague and gave a lecture. I was fortunate to have a few minutes afterwards to talk with him, which has prompted this and the previous post. Last time I explained one small line of reasoning why I think dark matter is a more likely explanation for certain observations, but I also mentioned some arguments against it. Today I wanted to go into more details on those. But I'm not going to, because I realised it's worth a short post about the the philosophical differences (or indeed, lack thereof) between those who believe in dark matter and those who favour modifying gravity.

The Crab Nebula, the remains of a massive star that exploded about a thousand years ago. Similar explosions created the heavy elements found in life on Earth.
Let's get one thing straight. The Universe is not elegant. Beautiful, sure, but creating the elements needed for life by blowing up entire stars has all the elegance of a constipated walrus in heat. Human scientists may like to think there's some kind of underlying order, but the Universe is under no obligation to behave how we think it should.
You could say to the universe this is not fair. And the universe would say: Oh, isn’t it? Sorry. - Terry Pratchett, Interesting Times.
However beautiful, obvious, simple, or mathematical harmonious a theory is tells you precisely nothing about whether it is really true. It's useful to start with the simplest theory you can come up with to explain observations (just because simpler theories tend to be easier to test), but there's no good reason to expect the Universe to follow a rule because it's simple. Or not to follow a rule because you don't understand it.

Take gravity, for instance. We started out believing that heavy objects fall faster than light ones -  a simple and obvious idea that's utter garbage. Thousands of years later we came up with the idea that there's a force acting by different amounts to keep objects of different masses accelerating at exactly the same rate toward the ground. That's a totally inelegant, contrived solution (though the equation predicting the acceleration is very very simple), but it works much better.

Centuries later and we realised that that too is wide of the mark, because actually objects are falling on the shortest paths through curved spacetime and obviously that's a much more elegant solution.

Of course the bear is only thinking that calling this idea "elegant" is ridiculous. He's not making any comment on the validity of the theory, you understand.
But Einstein's gravity works ! It works far better than Newton's model in many situations, however ridiculous making the speed of light constant for all observers (no matter their speed) may seem, or however much overkill allowing time travel might be just to sort out why things fall at the speed they do. Few people regard Einstein's theory as the last word on gravity though, because, amongst other things, it predicts the existence of singularities - points of infinite density and gravity. And apparently infinities are just silly, inelegant things... and we wouldn't want any of them, right ?*

* String theory postulates that there are eleven dimensions as a way of avoiding these singularities. I don't know about you, but I find that prospect at least as incomprehensible as the idea that gravity is infinite.
Now don't get me wrong. I don't like singularities any more than anyone else does, and I'd prefer a theory I can understand intuitively as much as the next man. But the Universe doesn't owe me anything. It has no obligation to be any more comprehensible to a short blonde Welsh astronomer as it does to a chimpanzee or even to a daffodil.

So when it comes to galaxy rotations, we've got an uncomfortable choice. Galaxies are spinning much faster than standard gravity models predict. If we want to save the theory, we've got to suppose that there must be a heck of a lot of extra mass that we can't see - about ten times more than the luminous matter. Or, we can say that there's a really fundamental problem with the theory.

It's interesting to ask what would have happened if we'd measured the rotation speed of galaxies back in 1915, before relativity passed a series of tests with astonishing precision. Maybe significant doubts would have been raised... but we didn't, so they weren't. Since relativity works so well, postulating a large amount of invisible matter doesn't seem so unlikely. We already know with certainty that neutrinos exist (particles which barely interact with other matter at all), so the idea that there might be a particle which interacts with normal matter even less is, perhaps, not so far-fetched.

Gravitational lensing, as predicted by general relativity.
Because we know how successful Einstein's theory is in most situations (singularities aside), the existence of dark matter is (in a sense) a prediction, not a fudge to save it. If we took that view, we could end up altering every theory whenever we found a flaw, and therefore end up making no predictions at all. It's true it's not a direct prediction of relativity, because relativity is a mathematical model which (by itself) says nothing about the real Universe. Like any mathematical tool, you have to couple it with observations to understand what's going on.

Those who doubt the existence of dark matter sometimes point to the fact that no such particle was ever predicted by the standard model. Which is to say that they trust the standard model of particle physics, but not the model of gravity. For those who believe in dark matter it's the other way around. Basically everyone is arguing about which bit of physics they think is wrong - hence the uncomfortable choice.

Honestly people, y'all sounding a bit silly. A classic case, methinks, of "the competition is so fierce because the stakes are so low"

Another point people harp on about incessantly is falsifiability. If it's possible to prove that a theory doesn't work*, then that counts in is favour (it's much rarer to be able to directly determine if something is actually true). If you can't, then you might never know if the idea is good or not, and you could keep toying with it until the end of time.

* I mean in principle. If you actually disprove it, obviously your theory is just wrong and therefore no good.

Proponents of modified gravity sometimes say that their ideas are falsifiable, whereas dark matter isn't. In fact one of their theories, TeVeS, has already been falsified, as Milgrom admitted last week. The trouble is that now they're looking for other ways to modify gravity to explain away dark matter... which reminds me of the time Ireland voted the wrong way on an E.U. treaty - they were asked to vote again.

Yes, you can falsify individual theories of gravity. It's far less clear if you can falsify all theories of gravity that could explain dark matter. You could very well keep saying, "well then let's try another one", ad infinitum.

Dark matter doesn't do any better than this. Although there are direct detection experiments underway, if they don't find anything one could always say "well then the particle must be even more difficult to detect". Much worse, though, is that some people would like to define dark matter as the anomalous gravitational effects themselves, rather than any specific particle :
Even if all of this was explained by a modification of gravity rather than an unknown type of matter, it would still have to be possible to formulate this modification of gravity in a way that makes it look pretty much like a new type of matter. And we’d still call it dark matter.

Umm, no. No we wouldn't, because that is plainly ridiculous. We decided to call it "matter" for a reason. And that reason is that we didn't want to modify gravity !!!

Then again I hear an equally interesting mix of noises from the MOND group. One person is prone to saying, "and thereby dark matter is falsified", almost as a reflex action.... "Merging galaxies ? Falsify CDM cosmology. Tidal dwarf galaxies ? HAH ! They disprove CDM completely. See that dog over there ? He falsifies CDM cosmology too. This piece of cheese... ? How about my enormous hat ?" (I've thought about making a drinking game out of this). Another person insists that dark matter can't be falsified. You see my dilemma.

Just to make things even more complicated, most simulations of galaxy formation with dark matter only use dark matter, because modelling the gas and stars is much more complicated (that is beginning to change, but slowly). So DM supporters cry out, "we need baryonic [normal matter] physics !" whenever a discrepancy, however major, is found between theory and observation. In contrast MONDers don't really have much in the way of simulations yet, so we don't really know how well the theory works at all. That, however, will soon change.

The reality is that front-line research doesn't always conform to the idealised standards of the scientific method. It can't. We are, after all, talking about explaining the nature of the Universe here - we are not at the point where we even know all the predictions of the theories to test them. We have no choice but to grope in the dark, but if we don't try at all we shall most certainly fail.

Philosophy, then, hasn't been of much help. The Universe doesn't give a monkey's if your theory is more elegant or not. Arguing about whether you think particle physics or gravity is more likely to be wrong isn't likely to get you anywhere. And neither the notions of dark matter nor of modified gravity are clearly falsifiable. 

But... we do know that Einstein's theory is well-tested in controlled conditions, and we know that some particles similar to dark matter do exist. Which to my mind cautiously swings it in favour of dark matter, but not by anywhere near enough to say, "let's give up modifying gravity, that was a silly idea." Since neither idea appears to have any intrinsic advantage over the other, the only way to choose between them is to look in detail at the arguments for and against both ideas. Which is what we'll do next time.

Of course, you can always just say, "I honestly don't know", but where's the fun in that ?


  1. Hi, has there been any work done to see if the amount of dark matter (or its effects) has changed over time? Perhaps in work on the evolution of galaxies?

    1. The amount of dark matter is thought to remain constant over time, except for a very small amount which ends up in black holes. One exception would be if dark matter is also its own anti-particle, in which case the amount will be slowly decreasing as it annihilates itself over time. However, I have no idea if the amount by which it decreases is thought to be significant - and of course we still have no idea what dark matter actually is, let alone whether or no it's its own anti-particle.

      The effects of dark matter definitely do change over time (I give an example in part one of this trilogy : For example as dark matter collapses early in the Universe, it draws in normal matter and is thought to be responsible for the large-scale structures we see today - this is one of the model's best predictions.
      Later on, once most of the gas has collapsed into galaxies, it's still important but in a different way. With not much gas left it becomes less important for forming large structures, but because it's so much more massive than the gas and stars it strongly affects how galaxies interact with one another and their environment.