Rethinking Cosmology (Steinhardt)

[marcus]

http://www.perimeterinstitute.ca/seminar/rethinking-cosmology

In five days Paul Steinhardt (Princeton) will give a colloquium talk at Perimeter. He has a 90 minute timeslot. Steinhardt has been a proponent of "brane-clash" cosmology. Sometimes called Cyclic or Ekpyrotic. Two spatial hypersurfaces floating in higher dimensional surroundings repeatedly bump each other.

Steinhardt is worth paying attention to as someone who has been both instrumental in the development of stringy ideas and inflation models, and also a mathematically perceptive critic of same. He has shown by fairly general arguments that "curled-up" extra dimensions are unstable during inflation and that prevailing assumptions about inflation make a satisfactory inflationary episode highly unlikely unless there is drastic fine-tuning.

==quote==
RETHINKING COSMOLOGY
Colloquium
Speaker: Paul Steinhardt
Wednesday, March 6, 2013 - 14:00 to 15:30
Location: Lazaridis Theatre
This talk will begin by discussing one by one the various
reasons why cosmologists today consider the big bang inflationary cosmology to
be the leading, if not proven, theory of the universe and

then explaining why each of these reasons is flawed.

This leads naturally to the question: what is the alternative? Understanding the flaws helps point the way.
==endquote==

I have a high regard for the challenges to conventional inflation ideas which Steinhardt has posed, even though I do not especially favor the answers he has come up with.

 

[Chronos]

Yes, inflation has warts. No news there. But, it works so well in so many regards it is difficult to resist. As you said, there is no good alternative - yet. My issue with inflation is it seems so easy to get it going [at least in theory], but, not so easy to stop.

 

[Mordred]

Yes, inflation has warts. No news there. But, it works so well in so many regards it is difficult to resist. As you said, there is no good alternative - yet. My issue with inflation is it seems so easy to get it going [at least in theory], but, not so easy to stop.

Now that's a statement I heartily agree with.

 

[marcus]

Past stuff from Steinhardt:

6 page SciAm article by S. (April 2011) at his Princeton website:
http://www.physics.princeton.edu/~steinh/0411036.pdf
"The Inflation Debate"
basically a critique of inflation
8 pages if you count some lavish illustration.

An earlier PF thread about Steinhardt critique (July 2011) with a lot of links
and careful outline of his arguments by PF member Ben Crowell:
https://www.physicsforums.com/showthread.php?t=514186

A Perimeter conference focusing on the problems with Inflation and more generally with Early Universe Cosmology as a whole:
http://pirsa.org/C11008 (Conference: "Challenges for Early Universe Cosmology")
Among the 30-some video talks there is one by Neil Turok, the main organizer
http://pirsa.org/11070044/ ("Overview of the Challenges")
and one by Steinhardt, who has co-authored a lot with Turok
http://pirsa.org/11070029/ ("Meeting the Challenges")

 

[marcus]

What could be interesting about this upcoming colloquium talk by Steinhardt is that it is almost 2 years later---what in his perspective will have changed?
Will he have new results to present?
Will his proposed alternatives, his ways to meet the challenges, have developed any from 2 years ago?

He clearly thinks the prevailing concepts about the start of expansion are flawed,
so what alternative does he now favor, to put in their place?
Here again is the abstract for his Wednesday talk:
==quote==
This talk will begin by discussing one by one the various
reasons why cosmologists today consider the big bang inflationary cosmology to
be the leading, if not proven, theory of the universe and

then explaining why each of these reasons is flawed.

This leads naturally to the question: what is the alternative? Understanding the flaws helps point the way.
==endquote==

 

[marcus]

A possibly relevant 2008 Steinhardt Wesley article
http://arxiv.org/abs/0811.1614
Dark Energy, Inflation and Extra Dimensions
Paul J. Steinhardt, Daniel Wesley
(Submitted on 11 Nov 2008)
We consider how accelerated expansion, whether due to inflation or dark energy, imposes strong constraints on fundamental theories obtained by compactification from higher dimensions. For theories that obey the null energy condition (NEC), we find that inflationary cosmology is impossible for a wide range of compactifications; and a dark energy phase consistent with observations is only possible if both Newton's gravitational constant and the dark energy equation-of-state vary with time. If the theory violates the NEC, inflation and dark energy are only possible if the NEC-violating elements are inhomogeneously distributed in thecompact dimensions and vary with time in precise synchrony with the matter and energy density in the non-compact dimensions. Although our proofs are derived assuming general relativity applies in both four and higher dimensions and certain forms of metrics, we argue that similar constraints must apply for more general compactifications.
26 pages, 1 figure. Phys.Rev.D79:104026,2009
It would be interesting, if this topic comes up in the Wednesday colloquium, to see if Steinhardt has revised or extended any of the conclusions of this paper.

Again the video for Wednesday's talk is expected to be posted here:
http://www.perimeterinstitute.ca/seminar/rethinking-cosmology

 

[Jorrie]

Past stuff from Steinhardt:

6 page SciAm article by S. (April 2011) at his Princeton website:
http://www.physics.princeton.edu/~steinh/0411036.pdf
"The Inflation Debate"
basically a critique of inflation
8 pages if you count some lavish illustration.

This question has no direct bearing on inflation, but is about something that puzzles me in the top part of the first figure in the mentioned SciAm article. Of the two dotted limit lines, I understand the bottom one as last scattering, but what does the top one "Limit of indirect observation" mean. I guess it could be information gained out of the CMB anisotropy, but will someone who knows please help?

 

[marcus]

This question has no direct bearing on inflation, but is about something that puzzles me in the top part of the first figure in the mentioned SciAm article. Of the two dotted limit lines, I understand the bottom one as last scattering, but what does the top one "Limit of indirect observation" mean. I guess it could be information gained out of the CMB anisotropy, but will someone who knows please help?

I can't say I'm "someone who knows" but I'll take a guess as to what he means. (I believe it is not a technical term in common use so one has to figure out what the author means by it.)

It seems to be in the right place for the PARTICLE HORIZON. That is the limit of the in principle observable part, it is the limit of causality. Nothing beyond that can have affected us or what we are able to measure. Because in a manner of speaking "causation travels at the speed of light." And that's how the particle horizon is defined.

I think he is being pedagogically adroit, given the wide SciAm audience, and not using technical terms like
"surface of last scattering" and "particle horizon"
because then he would have to go into a long spiel to explain them.
Instead he is paraphrasing the basic idea (directly visible, indirectly determinable by caused effects) which is what matters---saying what they really are in everyday language.

 

[atyy]

Can one get the fit of Fig 1 http://arxiv.org/abs/1212.5226 without inflation?

 

[Jorrie]

... Of the two dotted limit lines, I understand the bottom one as last scattering, but what does the top one "Limit of indirect observation" mean.

It seems to be in the right place for the PARTICLE HORIZON. That is the limit of the in principle observable part, it is the limit of causality. Nothing beyond that can have affected us or what we are able to measure. Because in a manner of speaking "causation travels at the speed of light." And that's how the particle horizon is defined.

This definition does not seem to fit that line of the http://www.physics.princeton.edu/~steinh/0411036.pdf graph. Should the particle horizon not be right up there, near the end of inflation, T ~ 10^-32 sec, z ~ 10²⁵+? After all, if the universe was transparent then, we should in principle have been able to get photons from there. If the scale in the SciAm article is anything to go by and http://www.physicsoftheuniverse.com is reliable, then the 'indirect limit' must be somewhere around the "era of nuclei" (3 minutes to 500,000 years) of this diagram:

I reckon this corresponds to T ~ 10⁴ sec, z ~ 10⁷ on the 'Steinhardt diagram'. I'm not sure of my interpretation, but this does not look like the particle horizon to me. If not, then what is it? End of nucleosynthesis perhaps? But I'm speculating.

 

[marcus]

I'll see if I can follow your reasoning. 378000 years is about 10¹³ seconds. So let's see if Steinhardt's "directly observable" line corresponds.

Well, sort of, it is roughly a quarter of the way from 10¹⁵ down to 10⁵ so we can call that about 10¹³.

Now that other line "limit of indirect observation" looks like it's around 1000 seconds.
About a fifth of the way from 10⁵ to 10^-5. That would be 10³. So about 15 minutes.

Yes, I agree with you. that is too recent to correspond to the particle horizon. So I don't understand what that line could represent.

Could 15 minutes be roughly when NEUTRINOS started free streaming? No, that is said to have occurred around 3 seconds, not 15 minutes. But could that be an important moment for Steinhardt? Analogous to the surface of last scattering, but for neutrinos instead of photons? Was the line badly placed by the graphic artist? Too late at night for me to puzzle over these things. Time to head for bed.

 

[Chronos]

It would be interesting to hear Steinhardt's explanation. I agree with Jorrie.

 

[Jorrie]

Now that other line "limit of indirect observation" looks like it's around 1000 seconds.
About a fifth of the way from 10⁵ to 10^-5. That would be 10³. So about 15 minutes.

Yes, I agree with you. that is too recent to correspond to the particle horizon. So I don't understand what that line could represent.

Could 15 minutes be roughly when NEUTRINOS started free streaming? No, that is said to have occurred around 3 seconds, not 15 minutes. But could that be an important moment for Steinhardt? Analogous to the surface of last scattering, but for neutrinos instead of photons?

T=1000 sec is near the end of nucleosynthesis, magnitude-wise. What sort of observable imprint could that have left on the CMB radiation? I guess if he meant neutrinos or gravitons, that should have been classified as "direct observation".

 

[marcus]

Just a general reminder, Steinhardt's talk is tomorrow. I should emphasize that personally I'm no fan of his alternative comology models, just hold him in high regard as a critic. So I want to listen carefully to what he has to say about the problems without necessarily buying his solution.

 

[skydivephil]

I have a few issues with Steinhardts criqtuqe of inflation , see here:
www.physics.princeton.edu/~steinh/0411036.pdf

and wonder what your thoughs were.
1. He states the as inflation is eternal it can't make any predictions, but then he says that the issue will be decided by the data i.e primoridal gravity waves.
How are these statements not contradictory?
2. How does an infinity of universes that are sperated in time i.e the cyclic universe differ (in terms of the measure problem) from an infinity of universes separated by inflating space?
3. Why doesn't he address the prior proability argument for inflation presented by Ashtekar and Sloan? Turoks reaction to it at the PI conference was embarassing.
4 If bubble collisions are found by Planck will he accept eternal inflation or not? Why are these not mentioned in his critique of eternal inflation and our ability to assess it. With Planck data coming out in April I think it would be better to address these concerns now so that any response doesn't look ad hoc.

 

[marcus]

I'm just about to watch today's talk by Steinhardt!
http://pirsa.org/13030079/

I may be able to respond more to the point after listening to what he has to say to fellow scientists at Perimeter Institute. (Can't always rely on the version of someone's thoughts that comes thru in the Scientific American. Notwithstanding that SciAm is a great outlet.)

 

[marcus]

Great talk!
He never mentioned branes bumping as in some earlier Steinhardt and Turok talk.
His bounce model looks objectively very much like an improved version of the LQG bounce.
The difference being that it does not need auxiliary inflation.

Around minute 80, as I recall, he answers the SECOND LAW objection that some people occasionally raise as an objection to the Big Bounce idea.

The LQG bounce (e.g. work by Ashtekar, Nelson, Agullo, Wilson-Ewing, Vidotto, Sloan etc...) involves the whole universe in a crunch and rebound, and, as some of the Loop cosmology people have calculated, it naturally produces a brief period of its own inflation without any inflaton field. But the Loop bounce does not produce ENOUGH inflation to achieve homogeneity and isotropy and other predictions...

Now Steinhardt has improved the bounce idea so that it involves contraction of a limited region, and so that the homogeneity and isotropy arise from the contraction. As he explains, in this picture we do not NEED the 60 e-fold expansion. So the Loop bounce, by adapting Steinhardt's idea, would avoid the need for an auxiliary inflation episode put in by hand.

There are amendments to the bounce that he outlines and argues for as reasonable (more so than conventional inflation notions), and it looks to me as if these could be incorporated into the Loop bounce framework.

It's a really impressive talk. Forcefully argued, and with a lot of audience participation in the last 20 or 25 minutes. Largely from the young Perimeter folk. Good Q&A.

I hope everybody interested in Cosmology watches at least the first 60 minutes, before the Q&A:
http://pirsa.org/13030079/

 

[Chronos]

Yes, I liked the talk, but, I think he glossed over some fine points. I'm not convinced writing off homogenity to a pre bounce contracting phase is as simple as suggested without any fine tuning. His explanation of CMB anisotropy was also unconvincing. I also found problems with his dismissive regard for the second law of thermodynamics objection - which I see as a big problem for any bounce model. Still, he offered a reasonable alternative to inflation, which I agree is suspect.

 

[skydivephil]

Will definitley watch this when I get the chance, not sure when that will be though.
What was his argument about the 2nd law?
Does he envoke lqc to get the bounce or some other mechanism or no mechanism?
If he does it envoke it what does he say to the period of inflation that arises from it?

 

[Jorrie]

What was his argument about the 2nd law?
Does he envoke lqc to get the bounce or some other mechanism or no mechanism?
If he does it envoke it what does he say to the period of inflation that arises from it?

As I understood it in his talk, only a small "empty" region of an accelerating (low energy) universe collapses, reducing the Hubble sphere for that region, keeping its entropy at the low value till the bounce. After the bounce, its Hubble sphere grows again and that region's energy is kept low. But, maybe I understood wrongly...

I do not understand the bounce without some form of inflation anyway.

 

[Mordred]

Watched the vid thanks for supplyin that. I also find myself agreeing with Chronos take on it.

 

[Haelfix]

This is just a pedagogical introduction to the cyclic universe proposal.

The questions at the end were good, b/c they are essentially the usual critiques of that theory (the one missing is that the cyclic model is also geodesically incomplete) and he didn't really answer them convincingly.

The modern viewpoint is that its unclear exactly what you gain from all of this. Even if you ignore the various theoretical consistency problems (violations of the null energy condition, the instabilities in the potential, the past singularity etc) At best, you temporarily save the volume measure, but the cost is that you push Penrose's finetuning problem back into the distant past, which you eventually have to address regardless.

 

[marcus]

There's something peculiar going on in the language here because looking up "ekpyrosis" in Wikipedia one sees the association with the earlier (eg 2001) picture of colliding braneworlds is very strong, whereas in the talk there is nothing of that.

Likewise if one looks at a recent paper http://arxiv.org/abs/0910.2230 one sees no brane-banging. The contracting phase prior to the bounce can apparently arise from a scalar field.

It seems that Steinhardt has managed to disassociate his ekpyotic and "cyclic" ideas from colliding worlds adrift in an unobservable higher dimension.

=================

BTW the explanation of why second law poses no obstacle to Steinhardt's "big bounce" is simple and comes around minute 79 and 80.
This is at the end of the prepared talk and at the start of the Questions. It is in response to what I think was the first, or one of the first questions at the end of the talk.
=================

I'll have to get some links to post 2009 Steinhardt papers. It could turn out that the disassociation with brane-bumping complete---one does not depend logically on the other.

 

[Chronos]

I agree his idea is not coherent.

 

[Haelfix]

The relevant paper is the following one (discussed in the talk)
http://arxiv.org/abs/arXiv:1105.3606

They derive their equations from a Braneworld model (which has been problematic in the past, as they had instabilities, and wildly departed from string theory consistency conditions eg were in the swampland), although of course it really doesn't matter. As long as you have an effective field theory of a scalar field with that specific potential, you will get cyclic solutions.

It is important to note that all of this remains semiclassical by construction. The scenario is very much tuned to avoid dealing with quantum gravity and/or any singularities during the bouncing phase.

 

[Chronos]

I have issues with the treatment of entropy. It is basically ignored in this model, which I find unrealistic. This approach suffers from the same objections as the usual inflation model, IMO.

 

[skydivephil]

I watched the video over the weekend.
Some thoughts:
1 He claimed inflation has failed becuase its been shown to be eternal and eternal inflation can't make contact with the data.
I still think this is a contradiction to his claim that one can tell inflationary and cyclic/ekypyrotic models apart from their gravitational wave signal. It may be true that inflation has not yet succeeded, but to say that it has already failed seems too much.
2 He needs to talk about the bubble collisions claims, would he accept this as evidence for eternal inflation if it is found in Planck data or not? If not , why not? If so , then how cna he make the claims mentioned above?
3 Prior proability of inflation: what about Ashtekar and Sloans paper? Again no mention.
4 Calling his model a big bounce leads to confusion. I Idont believe this model is the same as the LQC bouce, interested to see if others agree. So using the same terminology adds to confusion in my opinion.

 

[skydivephil]

Great talk!
He never mentioned branes bumping as in some earlier Steinhardt and Turok talk.
His bounce model looks objectively very much like an improved version of the LQG bounce.
The difference being that it does not need auxiliary inflation.

Around minute 80, as I recall, he answers the SECOND LAW objection that some people occasionally raise as an objection to the Big Bounce idea.

The LQG bounce (e.g. work by Ashtekar, Nelson, Agullo, Wilson-Ewing, Vidotto, Sloan etc...) involves the whole universe in a crunch and rebound, and, as some of the Loop cosmology people have calculated, it naturally produces a brief period of its own inflation without any inflaton field. But the Loop bounce does not produce ENOUGH inflation to achieve homogeneity and isotropy and other predictions...

Now Steinhardt has improved the bounce idea so that it involves contraction of a limited region, and so that the homogeneity and isotropy arise from the contraction. As he explains, in this picture we do not NEED the 60 e-fold expansion. So the Loop bounce, by adapting Steinhardt's idea, would avoid the need for an auxiliary inflation episode put in by hand.

There are amendments to the bounce that he outlines and argues for as reasonable (more so than conventional inflation notions), and it looks to me as if these could be incorporated into the Loop bounce framework.

It's a really impressive talk. Forcefully argued, and with a lot of audience participation in the last 20 or 25 minutes. Largely from the young Perimeter folk. Good Q&A.

I hope everybody interested in Cosmology watches at least the first 60 minutes, before the Q&A:
http://pirsa.org/13030079/

I don't recall the bit where he said it was a contraction froma limited region, I don't suppose you know where in the lecture that was?