- #1
- 24,752
- 795
Hartle Hawking suggest probabilities for observations (eternal inflation picture)
http://arxiv.org/abs/1009.2525
Eternal inflation without metaphysics
James Hartle, S.W. Hawking, Thomas Hertog
4 pages
(Submitted on 13 Sep 2010)
"In the usual account of eternal inflation the universe is supposed to be a de Sitter background in which pocket universes nucleate at a steady rate. However this is metaphysics because there is no way this mosaic structure can be observed. We don't see the whole universe but only a nearly homogeneous region within our past light cone. We show that we can use the no-boundary wave function to calculate small departures from homogeneity within our past light cone despite the possibility of much larger fluctuations on super horizon scales. We find that the dominant contribution comes from the history that exits eternal inflation at the lowest value of the potential and predict, in a certain class of landscape models, a tensor to scalar ratio of about 10%. In this way the no-boundary wave function defines a measure for the prediction of local cosmological observations."
They make a ton of assumptions, but they come out with [at least a symbolic calculation of] probabilities of certain features of the CMB being observed.
It changes the picture, tends to put multiverse thinking (whether or not you find it appealing) on a different footing.
Smolin also has a multiverse scheme which makes predictions about measurable physical constants that have so far not been falsified.
What H&H say here may be challenged and may turn out controversial, but having at least the shadow of a prediction will, I think, put this stuff at least temporarily in a new light.
[EDIT: after taking a closer look I cannot say for sure that the calculations they show symbolically could actually be carried out, except using questionable assumptions/radical simplification. Maybe someone else will take a look and have an opinion on it.]
[EDIT: I changed the headline from "offer probabilities" to "suggest probabilities". I am less sure now that they have a solid way to test anything here.]
http://arxiv.org/abs/1009.2525
Eternal inflation without metaphysics
James Hartle, S.W. Hawking, Thomas Hertog
4 pages
(Submitted on 13 Sep 2010)
"In the usual account of eternal inflation the universe is supposed to be a de Sitter background in which pocket universes nucleate at a steady rate. However this is metaphysics because there is no way this mosaic structure can be observed. We don't see the whole universe but only a nearly homogeneous region within our past light cone. We show that we can use the no-boundary wave function to calculate small departures from homogeneity within our past light cone despite the possibility of much larger fluctuations on super horizon scales. We find that the dominant contribution comes from the history that exits eternal inflation at the lowest value of the potential and predict, in a certain class of landscape models, a tensor to scalar ratio of about 10%. In this way the no-boundary wave function defines a measure for the prediction of local cosmological observations."
They make a ton of assumptions, but they come out with [at least a symbolic calculation of] probabilities of certain features of the CMB being observed.
It changes the picture, tends to put multiverse thinking (whether or not you find it appealing) on a different footing.
Smolin also has a multiverse scheme which makes predictions about measurable physical constants that have so far not been falsified.
What H&H say here may be challenged and may turn out controversial, but having at least the shadow of a prediction will, I think, put this stuff at least temporarily in a new light.
[EDIT: after taking a closer look I cannot say for sure that the calculations they show symbolically could actually be carried out, except using questionable assumptions/radical simplification. Maybe someone else will take a look and have an opinion on it.]
[EDIT: I changed the headline from "offer probabilities" to "suggest probabilities". I am less sure now that they have a solid way to test anything here.]
Last edited: