A theoretical particle physics question. (Is it dumb?)

If gravity is a particle, how do these gravity particles escape a black hole to affect another object?
Is this part of Hawking Radiation? or do the gravitons need not travel from one object to another to cause them to attract?
or (more likely) is it something else I haven’t even considered?

If the graviton exists it is expected to be massless so has no issue escaping a black hole.

Um, photons are massless…

I’ll leave answering the main question to those better equipped than myself.

This is actually a fairly deep question, and there are a couple of possible answers, some of which are easier to understand than others.

The answer that your typical particle physicist might give is the following: when we say that a force between particle A and particle B is “mediated” by some particle, we mean that so-called “virtual particles” are exchanged between A and B. These virtual particles can travel faster than light; they can’t transmit information, since they’re virtual, but they allow for the possibility of the influence of gravity extending outside of the black hole.

The answer that your typical general relativist might give is a little more complicated; but since I’m a relativist myself, I feel obliged to give it. When we say “graviton”, we mean “a quantized disturbance of flat spacetime.” This is all very well and good if your disturbances are small, but there’s no real way in which a black hole can be viewed as a disturbance of flat spacetime. You can also try to quantize disturbances of a black hole in the same way, but you’ll end up with a rather different theory that, on its surface, doesn’t seem to have anything to do with the one you started with. Many physicists think that a theory of quantum gravity, once we find it, should be “background independent”, which is to say that it shouldn’t depend on starting with some “reference” spacetime (i.e. flat spacetime or a black hole); and there’s at least one serious ongoing effort (loop quantum gravity) to obtain a quantum theory of gravity in this way. But nobody seems to have a full handle on to write down a background-independent quantum theory… at least not yet.

From a thermodynamics standpoint (?), the question isn’t terribly complicated if one assumes a static black hole, since the gravity field does not actively generate power. However, consider the case of two black holes rotating around each other. As these supermassive entities swirl around in space, they generate an ever-changing gravitational field due to their acceleration. As this emanates from the black holes it can cause motion in faraway objects.

The energy to do this has to come from somewhere. In some way or another it is coming from the black hole. Perhaps the energy entirely comes from the slowing down of the black holes which inexorably spiral toward each other. Even if this is the case, however, in a sense the energy did come from within the black hole, since its weight will be less when it is slower.

But perhaps somehow some of the gravity waves are caused by actual rest mass from within the black hole “converting” to energy? I dunno, me not know enough about physics to say.

Good point…

And it occurred to me later that my answer was wrong (or misleading at least) but I was thinking along a different track.

Whatever happened to curved spacetime? Why the need for a graviton? If a black hole massivley curves spacetime around it then there is no need to “radiate” gravitons or anything else.

You’re right that the exterior of a black hole is perfectly well described by “classical” general relativity, where spacetime curves in the way prescribed by Einstein’s equation. However, classical GR is pretty much universally believed to be incomplete, since it’s incompatible with quantum mechanics. Naive attempts to make a quantum theory of gravity predict that gravity is caused by the exchange of virtual particles called “gravitons”, just as electromagnetism is mediated by photons.

The only problem is that these naive attempts are spectacularly ill-behaved when you try to compute anything but the simplest interactions between gravitons & matter; so you have to be a little more creative. String theory claims to fix this ill-behavedness, although they still haven’t addressed what is to my mind the deeper issue of background independence (which I tried to explain in my previous post.) Loop quantum gravity does away with this notion of the graviton entirely, and attempts to quantum mechanically describe spacetime qua spacetime instead of as a bunch of particles running around. This is somewhat more background-independent, but to the best of my knowledge LQG still requires some amount of background structure (a global time coordinate on the spacetime, to be precise), and it only has the vaguest of notions on how gravity interacts with matter. Obviously, more research is needed.

There are actually several different (though related) things one might mean when one refers to the “mass” of a black hole. Of interest here, though, is what’s called the “irreducible mass”. The IM is closely connected with the surface area of a black hole, and, as the name implies, it cannot be reduced (classically, at least: Let’s leave Hawking radiation out of this for the moment). When a set of black holes does something (two black holes colliding, a black hole eating some normal matter, whatever), the irreducible mass is certain not to decrease, and will probably increase. Likewise for the total area of all of the event horizons: In fact, the area of a black hole’s event horizon can be regarded as its entropy (and we all know that entropy never decreases).

Getting back to the OP, gravitons are somewhat tricky to deal with, since they’re a feature of quantum gravity, and nobody really knows yet how to do quantum gravity. The electromagnetic interaction, on the other hand, is quite well-understood, so we have no problem talking about photons. So let’s talk about photons. One of the few properties black holes are allowed to have is electric charge. It’s expected that black holes with a significant charge would be rare in nature, but there’s no rule against them existing: If you take an existing black hole, and feed it a steady diet of electrons without protons (or vice-versa, or any other charged particle), it’ll build up a charge. You’ll now have a black hole exerting a Coulomb force on things around it (the Coulomb force could even be greater than the gravitational force), and it’s exerting that force by interchange of virtual photons. These photons are allowed to be out and about because they’re virtual, and virtual particles can do things that real particles aren’t supposed to (as long as they don’t get caught at it, but if they got caught, they wouldn’t be virtual any more, and the process of getting caught would generally put them back into the realm allowed to real particles).

:smack: Just as I thought!

Gravitons are better understood if you think in term of quantized fields rather than classical quantum mechanics. Basically, a particle is a “basic excitation” of one universal field.

As a very simple example of a classical field, consider a stretched string between two points. It can vibrate, and any vibration is a combination of a number of basic vibrations which each look like a sine wave with varying numbers of crests and troughs between the two endpoints. If we call the vibration mode with n crests and troughs (together, not n of each) v[sub]n[/sub], then overall the state is

Sum[sub]i=1[/sub][sup]infinity[/sup] A[sub]i[/sub]v[sub]i[/sub]

where A[sub]i[/sub] is the amplitude of that component of the total vibration.

When we quantize this system, it turns out that the amplitude of a given vibration type can only be a multiple of some basic amplitude. a[sub]i[/sub] = kA[sub]i[/sub]. So we’ll just absorb that basic amplitude into the definition of the basic vibration modes and get

Sum[sub]i=1[/sub][sup]infinity[/sup] k[sub]i[/sub]v[sub]i[/sub]

meaning a vibration consisting of k[sub]i[/sub] vibrations of type i for each i. Each basic vibration is a fundamental building block of any kind of vibration, and so we may refer to them as if they were objects – say, “stringons”.

So, back to gravity. There is one gravitational field in the universe, which measures the local curving of spacetime. There are a few idealized solutions, but most can’t be easily written down. Still, say we start with a nice solution and then tweak it a bit (like plucking a string). That tweak propagates out across the fabric of spacetime like a wave, and (in theory) those waves behave in certain similar ways to the vibrations of a stretched string. In particular, the fundamental building blocks of those waves are gravitons.

What are virtual particles?

Are they just mathematical constructs to make the math easy?

OK, well…

Thanks for the responses so far.
From what I’ve read here, I don’t like the idea of gravitons: too many problems.
(And, of course, my opinion should be the end of the matter: you can go ahead and shut down all those particle accelerators, no need anymore…)
Thinking about these rule-ignoring virtual photons, however…
Could it be said that gravitons are always virtual? It seems this would tie up some loose ends.

A related new question:
Does quantum mechanics even need gravity or gravitons?
The relativity spacetime warping/condensing seems straightforward at a macro level, why muck it up?

Then there’s something I’ve heard about gravity being a by-product of a holographic universe model, where numerous dimensions are condensed into the four we can perceive, yet obeying the same laws. Care to explain (or dismiss) this?

Well, they’re certainly mathematical constructs. As to whether they’re just mathematical constructs, well, we dunno. It’s by definition impossible to detect a virtual particle directly, so that avenue of inquiry is out. But the way particle physics works sure makes it seem like there are virtual particles at work down there.

garygnu, quantum mechanics doesn’t need gravity, in the sense that you can construct a perfectly self-consistent model of quantum physics which doesn’t have gravity in it. The real Universe, however, is not quite so obliging: Our Universe does, indeed, have gravity. And in every situation we’ve ever observed, General Relativity is quite sufficient to explain gravity, but the situations we have observed must inevitably lead to situations where GR would not be sufficient. For example, if you were to let a black hole sit for a long time (a very, very long time), it would evaporate away. There is nothing in GR alone to prevent it from evaporating away to nothing. But before it reached 0, it would have to reach the Planck mass, and a black hole at the Planck mass would most certainly behave in a quantum mechanical manner. So we need quantum gravity to either say what a very small black hole would do, or to tell us why it wouldn’t ever get that small to begin with.

Whoops; forgot to answer this part. There is (it would seem) such a thing as gravitational radiation, and there are several projects currently in the works to detect it. A gravitational wave can be regarded as a stream of real gravitons, in much the same way that a light wave can be regarded as a stream of real photons. There’s no real hope of ever detecting individual real gravitons, but there is significant hope (though it’s still difficult) of detecting a stream of many of them, and in fact it’ll probably happen in the next few decades.

I came to a similar conclusion while walking my dogs today.
Quantum Mechanics (QM) doesn’t need gravity, but the Real World (RW) has it, so must QM, because QM is the cause of all things in the RW.

So this is why quantum physics is so fun.

But why must this be? Is it because of the big muddy hole in Texas?

Are there any direct observations that lead to the postulation of the existence of gravitons? What do scientists use as evidence for supposing their existence? Is this all in the theoretical? What are some of the derivations of the notion?

Nah, the Big Muddy Hole in Texas[sup]TM[/sup] was never meant to study gravity.

Individual gravitons are well-nigh impossible to detect because gravity interacts so weakly with matter compared to, say, how photons interact with matter. So not only do you need to have a lot of mass moving around to produce a strong gravitational source, you also have a dickens of a time making your detector respond to it. What’s more, many of the standard tricks involved in demonstrating the quantum nature of light require some material that is transparent to light; but there’s no such thing as a material that is transparent to gravity. (Roughly speaking, this is because there are positive and negative electrical charges, but there’s only positive gravitational charge.)

Analogy with every other field theory, mostly. Various properties have been predicted by noting properties of gravitational waves (how disturbances from a background gravitational field propagate) and seeing what sort of particle would have give rise to that sort of field. For instance, the graviton is predicted to have spin 2 because the first nonvanishing term of gravitational waves is the quadrupole moment.

I’m pretty sure you meant to say “opaque” there, not “transparent”. Everything’s transparent to gravity.

And the weak interactions with other matter is one of the big problems with detecting individual gravitons, but not the only one. There’s also the problem that gravitational waves typically have very low frequencies (there are not expected to be any gravitational waves in the Universe with frequencies higher than 10[sup]4[/sup] Hz, or so, and most would be much less than that). Gravitons would have the same relationship between energy and frequency as do photons, so an individual graviton in one of those waves would have an energy of around 10[sup]-11[/sup] eV. I don’t think we could detect even a photon, at those kinds of energies.

It’s still theoretically possible to detect individual gravitons, of course, but it’s well beyond anything humans could possibly hope for, even if scientific research got 100% of the federal budget. If I absolutely had to design an experiment to do it, I’d use an evaporating primordial black hole for the source and a hundred-billion-year-old neutron star for shielding, and I’d still probably have to repeat the experiment billions of trillions of times (with a new black hole and neutron star each time) to get a good statistical chance for an unambiguous result. So, yeah, it’s a bit impractical.