# Who Needs Gravitons?

*Most recent answer: 05/26/2013*

- Andy Kelsall (age 46)

Melbourne, Australia

That's a very deep question. The answer is more subtle than most of the ones we post here.

Although general relativity (GR) treats gravity as a geometrical effect of space time, it's still a sort of field. there can be waves in it, moving around very much like any classical wave. In fact, strong indirect evidence of these waves exists in the slowing of binary pulsars. So far, I still haven't given any need for gravitons.

The structure of quantum mechanics leads logically to certain uncertainty relations, by which no quantum state can have sharply defined values of each of some pair of variables, such as (position/momentum), (event time/ energy change), (x-spin/y-spin, for a spinor), etc. Now if there were any classical-like objects, whether particle-like or wave-like, lacking these uncertainties, then one could in principle make some sort of "microscope" to view the quantum state with those classical objects, unraveling the uncertainty relations, and revealing the true values of the apparently fuzzy variables. However, a series of experiments showing that certain relations, the Bell Inequalities (search this site and elsewhere), that must be obeyed if those true values exist are not in fact obeyed by our world. So we'd be left with a logical contradiction. There' seems to be no way out except for spacetime geometry to also have quantum properties, such as shot-noise in gravitational waves, like the shot-noise in light waves, arising from randomness in the graviton number analogous to randomness in the photon number. So far the quantum theory of gravity is a work in progress.*

Mike W. (posted without checking until Lee returns)

*The BICEP2 collaboration claimed to have seen what looks like the leftover effects of the quantum zero-point spread of gravitational waves, left as tiny "B-mode" ripples in the polarization of the cosmic microwave background radiation. It now looks like they really saw somethig more local and less interesting. If one of the more sensitive versions of this experiments works, then it will be fair to say that a real quantum gravity effect will have been seen. /Mike W.

*(published on 05/26/2013)*

## Follow-Up #1: quantum gravity waves

- Andy Kelsall (age 46)

Melbourne, Australia

Andy- you're on the right track, but my previous answer was too compressed.

1. The GR equations predict the necessary existence of gravity waves. Those pulsars are just the experimental case in which there's good evidence (slowing of the rotation rate) confirming that particular prediction. A wide variety of other observations confirm the accuracy of GR, which means that the waves have to exist,

2. I think you're getting at the right idea, but I never use the phrase "wave-particle duality", It's widely-used, but a relic of the early days of quantum mechanics. In most modern interpretations, nothing exists but the wave-like quantum state. Under some circumstances it has a fairly localized distribution and in other cases it's more spread out. The quantum roughness of the wave, as if particles were being counted, has a more technical description*, but no stepping outside the ordinary wave-like state.

It's extremely awkward to try to cram GR effects into anything like Newton's spacetime. GR really is geometrical. Whether that picture will hold up well on the tiny scale on which quantum gravity becomes important, I don't know.

The existence of gravitons by itself would *not* suffice to fit GR comfortably with quantum mechanics. The people who work on this say that the problem is that just trying to do a routine quantization like that leads to infinities for all sorts of calculated quantities. Some sort of deeper fix seems to be needed. They say that one possibility that avoids the infinities is to extend the number of spatial dimensions to 9 (for string theory) or 10 (for M-theory).

Your argument about how the structure of GR doesn't require quantization looks ok to me. The requirement for quantization comes from all the rest of physics, which loses its self-consistency if mixed with any non-quantum ingredients.

Mike W. (posted without checking until Lee gets back)

* There exists a Hermitian operator with integer eigenvalues, which plays the role of a particle count.

*(published on 05/26/2013)*

## Follow-Up #2: GR and QM

- Andy Kelsall (age 46)

Melbourne (Australia)

Thanks Andy, You've got what I was saying exactly right. Sorry I couldn't make it simpler- that's probably an indication that I don't understand it as deeply as the real experts.

Mike W.

*(published on 05/27/2013)*