Saturday, February 13, 2016

Gravitational waves, evidence of the fourth dimension of spacetime

Most of the headlines are right: gravitational waves are a long-known prediction of General Relativity, and their detection show that the theory is correct. I waited a bit to see if an important consequence of this fact will be uncovered, but it seems it did not, so let me tell you: This experiment refutes a great deal of alternatives to General Relativity proposed in the last decades. You perhaps already noticed that many physicists brag on social networks or even in online articles that the detection of gravitational waves confirmed not necessarily GR, but also the alternatives to GR they endorse. But in fact this experimental result refutes those alternative theories in which the background metric of spacetime is fixed, as well as those in which space is a three-dimensional thing that is not part of a four-dimensional spacetime, as in GR is. I will discuss first the latter. Many relativists would say that such theories were already refuted, but if you talk with a supporter of such a theory, you will hear that it is not necessarily so. The idea of a 3-dimensional space still could be defended, with the price of complicating the things. But in my opinion, LIGO just put the last nail in the coffin of such theories. Because gravitational waves are waves of spacetime, and not of space. They are waves of the Weyl curvature tensor, which simply vanishes in a space with less than four dimensions!

The number of those trying to replace GR with other theories increased very much lately. The main reasons may be that they don't know how to handle singularities, or that they don't know how to enforce to gravity the few methods we know to quantize fields, so they come up with alternative theories. While I don't think it is easy to replace GR with something that explain as much starting from as little as GR does, I agree that these alternative should be explored (by others, of course). Related to whether there is a 3-dimensional space or a 4-dimensional spacetime, you can find reasons to doubt the fourth dimension too. First, even Galilean space and time can be joined in a four-dimensional spacetime, but not as tight as in Relativity. In Relativity, indeed, Lorentz transformations mix the time and space directions, leading to length contraction and time dilation, but some think that these are sort of due to the perspective of the observer, without needing a fourth dimension. In addition, many quantities become unified in the four-dimensional spacetime, such as energy and momentum, electric and magnetic fields etc. But maybe these are all just circumstantial evidence of the fourth dimension. You can take any theory and make it satisfy some four-dimensional transformations. Especially since the evolution equations are hyperbolic, you can do this. Also, you can express any equation in Physics in curvilinear coordinates, and this doesn't mean four dimensions, neither that the invariance to diffeomorphisms means something physical. So people cooked up or even revived various alternatives to GR, in which three-dimensional space is not part of a spacetime. If such a theory does not include curvature, it will not predict gravitational waves. Also, if it admits curvature, but only of the three-dimensional space, nothing in four dimensions, it still doesn't predict gravitational waves out of this curvature. So now the proponents of alternative to GR will have to adjust their theories. Maybe some predict naturally some sort of gravitational waves, but most don't, so they will put the waves by hand. The Cotton tensor, which is somewhat analogous to the Weyl tensor in three dimensions, because its vanishing means conformal flatness, is believed sometimes to give the gravitational waves. But the Cotton tensor vanishes in vacuum, where the Ricci tensor vanishes too. So this can't give gravitational waves in three-dimensional spacetime.

What about theories with more dimensions? For instance, Kaluza theory is an extension of GR to 5 dimensions, which is able to obtain the sourceless electromagnetic field from the extra dimension. You can also obtain other gauge theories as Kaluza-type theory. Such modification predict gravitational waves too.

What about String Theory? It is said that String Theory includes GR, so it must include gravitational waves too, isn't it? But the reason why is said to include GR is because it contains closed strings, which have spin 2, and they are identified with the still hypothetical gravitons (not even predicted by GR alone) just because they have spin 2. But if your theory has spin-2 particles, even if you call them gravitons, it doesn't mean you have included GR. String Theory usually works on fixed background, which usually is flat, or with constant curvature as in the anti-de Sitter spacetime. I am not aware of a successful way to include GR in String Theory such that gravity is an effect of spacetime curvature. If this can be done, can it predict gravitational waves in a natural way? Can it even include GR in a natural way?

To my surprise, the advocates of theories which don't have dynamical background, or are based on three-dimensional space, didn't take the chance to predict that there are no gravitational waves, as their theories imply. They should have done this, and they should have waited for the confirmation of their prediction by LIGO. My guess is that maybe they doubted that GR will be refuted - nobody wants to make predictions which contradict GR in regimes that can be experimentally verified. Whenever we could test the predictions of GR, they were always confirmed, so I think not even those supporting alternative theories actually believed that it will be refuted this time. So I guess that's why they didn't say that their theory predicts no gravitational waves, and that they really think that LIGO will show there are. Instead, now you can see that some claim that gravitational waves confirm their theories too. Like for examples they are waves of space alone, and not of spacetime, which is not true, unless you put them in your theory by hand (while in GR they are just there, not a mobile or replaceable part). So I expect to see a lot of papers in which it is explain that their theory was there too, along with GR, when gravitational waves were predicted.

Since the model was based on calculations made using GR applied to two colliding black holes, LIGO confirmed GR (again): it confirmed gravitational waves, and black holes (again). This does not exclude though the possibility that other modifications, alternatives or extensions of GR can work out similar predictions. So further experiments may be needed. But what I can say is that the theories that remained are modifications of GR that still explain gravity as spacetime curvature, and still make use of the four-dimensional spacetime. Theories that at purpose mimic most of GR.

Space is dead, long live spacetime!