Tuesday, August 21, 2018

Roger Penrose still looks for evidence of universal rebirth

Roger Penrose really should have won a Nobel Prize for something. Though I’m not sure for what. Maybe Penrose-tilings. Or Penrose diagrams. Or the Penrose process. Or the Penrose-Hawking singularity theorems. Or maybe just because there are so many things named after Penrose.

And at 87 years he’s still at it. Penrose has a reputation for saying rude things about string theory, has his own interpretation of quantum mechanics, and he doesn’t like inflation, the idea that the early universe underwent a rapid phase of exponential expansion. Instead, he has his own theory called “conformal cyclic cosmology” (CCC).

According to Penrose’s conformal cyclic cosmology, the universe goes through an infinite series of “aeons,” each of which starts with a phase resembling a big bang, then forming galactic structures as usual, then cooling down as stars die. In the end the only thing that’s left are evaporating black holes and thinly dispersed radiation. Penrose then conjectures a slight change to particle physics that allows him to attach the end of one aeon to the beginning of another, and everything starts anew with the next bang.

This match between one aeon’s end and another’s beginning necessitates the introduction of a new field – the “erebon” – that makes up dark matter, and that decays throughout the coming aeon. We previously met the erobons because Penrose argued their decay should create noise in gravitational wave interferometers. (Not sure what happened to that.)

If Penrose’s CCC hypothesis is correct, we should also be able to see some left-over information from the previous aeon in the cosmic microwave background around us. To that end, Penrose has previously looked for low-variance rings in the CMB, that he argued should be caused by collisions between supermassive black holes in the aeon prior to ours. The search for that, however, turned out to be inconclusive. In a recent paper with Daniel An and Krzysztof Meissner he has now suggested to look instead for a different signal.

The new signal that Penrose et al are looking for are points in the CMB at the places where in the previous aeon supermassive black holes evaporated. He and collaborators called these “Hawking Points” in memory of the late Stephen Hawking. The idea is that when you glue together the end of the previous aeon with the beginning of ours, you squeeze together the radiation emitted by those black holes and that makes a blurry point at which the CMB temperature is slightly increased.

Penrose estimates the total number of such Hawking Points which should be in the total cosmic microwave background is about a million. The analysis in the paper, covering about 1/3 of the sky, finds tentative evidence for about 20. What’s with the rest remains somewhat unclear, presumably too weak to be observed.

They look for these features by generating fake “normal” CMBs, following standard procedure, and then trying to find Hawking Points in these simulations. They have now done about 5000 of such simulations, but none of them, they claim, has features similar to the actually observed CMB. This makes their detection highly statistically significant, with a chance of less than 1/5000 that the Hawking Points which they find in the CMB are due to random chance.

In the paper, the authors also address an issue that I am guessing was raised by someone else somewhere, which is that in CCC there shouldn’t be a CMB polarization signal like the one BICEP was looking for. This signal still hasn’t been confirmed, but Penrose et al claim pre-emptively that in CCC there should also be a polarization, and it should go with the Hawking Peaks because:
“primordial magnetic fields might arise in CCC as coming [...] from galactic clusters in the previous aeon […] and such primordial magnetic fields could certainly produce B-modes […] On the basis that such a galactic cluster ought to have contained a supermassive black hole which could well have swallowed several others, we might expect concentric rings centred on that location”
Quite a collection of mights and coulds and oughts.

Like Penrose, I am not a big fan of inflation, but I don’t find conformal cyclic cosmology well-motivated either. Penrose simply postulates that the known particles have a so-far unobserved property (so the physics becomes asymptotically conformally invariant) because he wants to get rid of all gravitational degrees of freedom. I don’t see what’s wrong with that, but I also can’t see any good reason for why that should be correct. Furthermore, I can’t figure out what happens with the initial conditions or the past hypothesis, which leaves me feeling somewhat uneasy.

But really I’m just a cranky ex-particle-physicist with an identity crisis, so I’ll leave the last words to Penrose himself:
“Of course, the theory is “crazy”, but I strongly believe (in view of observational facts that seem to be coming to light) that we have to take it seriously.”


  1. Dam! Every time there is something wrong in the theory! I start to be afraid that the answers we are looking for, are out of reach……
    All the best to you Sabine.

  2. Sabine: do you know anything Penrose's views on dark matter and MOND?
    Since he knows you well, you could ask him.

  3. What the correct etymology of "erebons" probably is

    (also offspring of Chaos and brother to Nyx!)

    1. Yep, I remember watching an interview where Penrose stated as much (i.e. erebons named after the Greek god Erebus).

  4. Penrose's Conformal Cyclic Cosmology (CCC) is in some ways similar to the AdS/CFT correspondence. Both ideas involve equating degrees of gravitational freedom with those of quantum fields. The observable dS spacetime is a hyperboloid of one sheet, a four dimensional “sheet” in five dimensions, while the AdS spacetime is one a single surface of a 4-dimensional hyperboloid. The two surfaces in effect meet at “scri^+.” As a result the two spacetimes share related conformal structure. In the case of AdS a conformal field theory exists on the conformal boundary with degrees of freedom holographic equivalent to those of gravitation in the bulk of the AdS_n, for n = 4 or 5. The CCC posits conformal equivalency with boson fields at or near scri^+. This means the two have some correlation as methods connecting gravitation with quantum field theory.

    The correspondence between the one and sheet hyperboloids might also be thought of as a type of mass-gap. The two hyperbolod types are connected by a null sheet or light-cone. These structures exist in a spacetime of one dimension larger than the dimension of the hyperboloids. The momentum-energy form of the spacetime and the existence of hyperboloids bears a similarity to the mass-gap in the Dirac equation. The two sheet hyperboloids are then surfaces bounded above the origin by the cosmological constant in a way similar to the gap in the Dirac equation. The one sheet hyperboloid is similarly so bounded away from the origin in a spacelike direction.

    So the two approaches to quantum gravitation are maybe not that unrelated. The de Sitter vacuum is not completely stable and has some inconsistencies with string theory. In particular type I string has negative vacuum or zero point energy that is consistent with AdS spacetime. Of course Penrose has a disdain for string theory, which he makes clear in his Road to Reality. This instability of the dS spacetime may have some connection with what Penrose proposes.

  5. In Penrose's book "Fashion, Faith, and Fantasy in the New Physics of the Universe" (2016) there is no mention of Milgrom. I say that Milgrom is the Kepler of contemporary cosmology, and those who ignore Milgrom are likely to have a somewhat wrong theory of quantum gravity.

  6. My guess would be that he isn't in favour of MOND.

  7. I think Penrose's theory is crazy, and the Big Bang itself. However I used to think multiple universes was crazy, until I started studying it myself.

  8. As a big fan of Sir Roger Penrose, I too have wondered how one gets from one aeon to another, particularly without invoking erebons, dark matter or colliding supermassive black holes.

    In the far distant future, when the universe has become a boring, featureless, matter-free rarified gas of stray low-energy photons and nothing else, what might prevent the universe from "forgetting" that it has reached maximum entropy, then starting over again in a subsequent Big Bang via a quantum fluctuation (please forgive my use of that overworked term)?

  9. CCC has always been my favorite inflation buster. Going to keep a close eye on this one.

  10. Right or wrong, Penrose is a great writer. Have read several books, all good reads!

  11. GR and QM are incompatible. Separate predictions verify to ~15 significant figures. As with Newton (lightspeed, Planck's constant, Boltzmann's constant) and Euclid (Fifth Postulate), mathematical postulates can empirically fail. Physics derives then verifies; curve fits and log scale dives, but avoids falsification - risk of failure!

    GR and QM predict cryogenic molecular beam microwave rotational spectra that are measurably wrong. Said experiments, performed and published using other molecules for other reasons, are definitive.

    White swan science aesthetically excludes Australia. Stop lamenting comfortable failures. Observe outside postulates.

  12. I've read most of Penrose's books (including The Big One), some of his papers, heard him give several talks, chatted with him once. My impression is that, as a mathematician doing cosmology, he is rather far removed from galaxies, which is where MOND (or something with similar effects) comes in.

  13. Sabine:
    Penrose does explain the initial conditions. That was actually the reason for coming up with theory in the first place! And I think it's indeed a very elegant theory, the only ad hoc assumption being the asymptotic vanishing of the mass. You don't need the remaining massive particles to decay, only to gradually loose their mass. That's possible: recall that in De Sitter space mass is not a conserved quantum number. And the model has many virtues; it gives a dynamical origin to the Weyl Curvature Hypothesis; it explains dark matter as rescaled gravitational radiation coming from the previous aeon; it uses established physics: Hawking radiation, Weyl Curvature Hypothesis, a positive cosmological constant, conformal symmetry of massless fields, techniques from the conformal treatment of infinity (eg a theorem from Friedrich), gravitational radiation, etc

    As for the alledged evidence, I do think that the first analyses were inconclusive, but subsequent ones from the Polish tean (K. Missner et al.) seem more serious. As for the Hawking points, or erebon decay, only time will tell. But at least we've got a very elegant mathematical model of the Universe, which adresses some very pressing issues in cosmology, while sticking to established physics a great deal more than alternative proposals (Smolin's Cosmological Natrual Selection; the Inflationary Fractal Multiverse with False Vacuums; 10-dimensional colliding branes (Steinhardt & Turok), Boincing models (how on earth do you explain the vanishing of the Weyl Tensor near the Big Bang?) Etc

  14. I just read through the paper published on arXiv. I found that the argument interesting and will require a more though study to really grasp how reaches that assertion from the technique he employs.

    At least he makes a prediction that can be investigated. Question: Is there enough meat in the analysis to prompt someone investigate the regions he identifies specifically? Is investigation even possible given the current state of the art.?

  15. It’s not surprising Penrose has his favorite ideas as practically every other physicist does. There is a plethora of circumstantial evidence, and historical patterns of human behavior to suggest the most probable approach is to consider that models such as GR need to be rewritten. Yes, GR is the best explanation of gravity we have to date, and yes it taught us much more than we could have ever conceived under Newtonian mechanics, yet that doesn’t mean we should stubbornly stick to it for centuries like we did with Newton as we continually see kinks in its predictions.

    I find it frustrating continually seeing unexplained observations and compatibility issues squeezed into models with ever more fantastical speculations; it’s the same behavior physicists displayed with unexplained issues of their time under Newtonian mechanics. Does an objective analysis suggest that ever more enigmatic pet peeve speculations are not the most probable place to find the answers? There is some very sound empirical evidence, axioms, and fundamental assumptions that are ignored because they don’t fit the current models, yet they could make a strong foundation in rewriting them. While that may be hardest and most dangerous thing for a physicist to attempt, I also think history and current evidence suggests it is the most probable place to look first before making more wild speculations.

  16. Dear dr.Hossenfelder,

    " A cranky ex particle physicist with an identity crisis". :-) :-( You're in the middle of the stream, the toughest spot, you know what you've lost, you can't yet see what you've gained. It helps to reflect on challenges past, how they were faced, and overcome.

    Someone believing there's only physics/tech/whatever in life is likely to be a very impoverished person. I know, it's very easy for me to say that.

    As you wrote in an earlier posting:" something will come to mind". Most likely, many things already have.Trust yourself.

  17. Am curious about the notion that traces/fingerprints/mini-structures from a previous aeon would transfer over to a subsequent aeon. Seems to be at the heart of CCC.

    I connect this to the common question posed by some about how our current universe started at a lower state of entropy and has subsequently continued to become more entropic.

    The question I'm raising is, why not entertain the notion that the universe began (in the instant before the bang) with zero entropy? Thus, entropy emerges from the start of the bang.

    In that scenario, there'd be no room for traces of any former universe... all would have gone still and non-structured for an instant... before the next cycle begins (by whatever mechanism that would be! hah!)

    Thanks for the interesting post.

  18. Carlo Rovelli has his own version of 'erebons' and he seems to agree with Penrose (2 arxiv papers from 2018). Btw Penrose's new paper does not mention them and they were not originally part of CCC. Being completely crazy is, of course, great fun.

  19. Like Parallel Universes, this mathematical elaboration of a realm that lies beyond the empirical reach of science suggests the possibility, if not likelihood, that the expanding universe model has turned out to be a scientific dead end wherein only fantasies provide an outlet.

  20. sabine,
    "Penrose … wants to get rid of all gravitational degrees of freedom"

    I thought that the idea was only to get rid of the mass component of the 10 gravitational degrees of freedom, resulting in masslessness and conformality. Penrose hopes, or expects, that mass will decay away or fall into black holes as the justification.

    In response to the Shantanu remark about Dark Matter I thought that was where the Erebons come in, but I don't know how they are progressing in his theory either.

  21. If there is to be any evidence of a theory it has to be able to be placed in the CMB?

  22. Penrose is one of the best physicists alive together with Dyson and Gell-Mann. Among his many theories, he explained relativistic jets of black holes in terms of “frame dragging” of spacetime. He was also the teacher of the late Stephen Hawking at Oxford.

    I think his cosmology is inspired by the Mandelbrot set. In his talks, he likes to show an artwork with infinite fractal boundary to represent the universe. His conformal symmetry is independent of scale, only the shape matters. There are many crazy ideas but to paraphrase Bohr, his theory is crazy enough to be true.

  23. Bee

    like shantu, do you know Penrose's views on LQG and AS as of 2018? i know in road to reality way back when penrose was skeptical of string theory supersymmetry and kaluza klein and thought LQG was impressive

  24. MJ,

    He doesn't, he selects some trajectories via the requirement that they ought to match end-to-beginning. Look, if you start with a universe resembling ours but without Hawking Points, the next one will have Hawking Points, so what's the initial condition? There's an additional criterion hidden here, which is that somehow our condition is - I dunno - likely in the infinite number of cycles? I just wrote a whole book about how arguments from probability can go wrong, so this issue makes me feel uneasy. I'm not saying it's wrong, I'm just saying it doesn't quite feel right to me. Best,


  25. neo, shantanu,

    No, I don't know what Penrose thinks about MOND and LQG and AS, whatever that is (ASG?).

  26. I love your blog.

    That's all.

  27. My guess is that we don't have the time to determine whether or not Dr. Penrose is correct.

  28. As usually in big claim about anything matter, the title is overexposed:
    "Physicists Think They've Spotted the Ghosts of Black Holes from Another Universe"
    but... we already spot the ghost of the Big Bang - every day long before LIGO.
    We did not spot ghost of earlier - or future - Big Bang's, and that might be a problem.
    On a scale of 1 to 10 - that must be the most energetic imprint in CMB pattern (5 sigma). If earlier universe timeline is a closed path and if universe is cyclic it must be that - then the BH ghosts from an (previous/earlier universe) will be shadowed by BB's starting point of that universe cycle).

  29. "I say that Milgrom is the Kepler of contemporary cosmology,"

    Maybe he is, in some sense.

    You've been saying it, here and elsewhere, for a long time now. Saying it doesn't make it true. It doesn't make it false either, but it might create the impression that all you have are sound bites.

  30. Sabine: as far as I understand, there is no fine tuning in CCC. You just have a positive cosmological constant which prevents the Universe from collapsing to one single monstrous black hole.... Then the inevitable process of Hawking Evaporation + loss of mass + conformal symmetry breaking (this two processes should also be somehow inevitable if CCC is right) takes over, and a new Aeon, satisfying the WCH, is born.

  31. My thought after reading this, "Look, if you start with a universe resembling ours but without Hawking Points, the next one will have Hawking Points, so what's the initial condition?" is that Sabine is a very careful thinker.

    However, is this true? It certainly seems to be reasonable, but without the complete dynamical laws how can we say for sure that without the remnants left over from a previous aeon and the gravitational radiation that it provides that supermassive blackholes would form and provide radiation for the next aeon?

  32. If universes go through cycles ( CCC) of trillions of years then what are the odds that we are in thr first 1% of any given cycle? Actually..the first.01% of a cycle? QM and GR incimpatability ( or lack of evidence of it) is dwarfed by the issue of of models on the universe based on time scales that are out of whack with probabilities of where existence should be along the time scale.

  33. "...cranky ex-particle-physicist..."

    Why ex, Bee?

  34. Everyone who has recently complained that their comments do not appear:

    I do not approve comments that promote your personal theory. Neither do I approve links to personal websites or youtube or vixra. I cannot edit comments, so it's all or nothing. If you have a comment on Penrose's paper, stick with it and scrape the rest.

  35. manyoso,

    We know that because you don't need it for structure formation. What do you think people who work on LCDM have been doing the last 20 years? They've secretly all the time sprinkled Hawking Points on their simulations just never told us about it?

  36. MJ,

    I hope we agree that our universe contains matter besides a cosmological constant.

  37. Sabine,

    Is generic structure formation enough or is it a particular structures (ie., galaxies) that could give rise to supermassive blackholes? All of those LCDM models have dark matter included, yes? In Penrose' formulation that dark matter is provided by the previous Aeon via supermassive blackhole evaporation.

    What I'm asking is if you take out the dark matter, do the remaining dynamics always give rise to super massive blackholes forming? Maybe they do, but I don't know this for sure. If not, then it is strictly not true that in a, "universe resembling ours but without Hawking Points, the next one will have Hawking Points."

  38. Thanks for bringing this to my attention. I like his crazy ideas. Hope I still have crazy ideas at 89. We are in the same place on Inflation, String theory and the role of black holes in the cosmos.
    This is not the first aeon. No need for initial assumptions.
    Yes everything is going to end up in a big black hole out of there in the cosmos. Now it could be a long dark end of things but it could also be part of the cycle that has repeated itself over time. Our universe created from the explosion of a black hole.
    The first thing to do is to find out if that are massive black holes out there. It may be in the CMB. Maybe from gravitational waves from those big black holes banging together. We need to be looking further. If the cosmos really is that big and our universe is just an explosion in space it simplifies a lot of physics.
    The universe has always been much bigger than we thought.
    Dark Energy is just gravity. The gravitational pull from the cosmos. No need for Inflation theory. That Dark Matter? It's what comes out of black holes.
    Black holes are the crucible of the cosmos.

  39. In his own lecture, Penrose calls his “conformal cyclic cosmology” "crazy cyclic cosmology" he has a wondeful sense of humor ;-D

  40. Penrose road map is quite interesting. Define possible the most violent events of previous eon and find echoes of those on CMB photo plus check on artificialy generated maps. He is like a radio ham.

  41. Stardust to stardust, why not energy to energy?

  42. @M.J. Glaeser : No matter what the cosmological constant is, the universe cannot collapse to a single black hole. It can collapse completely, all of it, to a final singularity, or it will come to a point when there are many black holes going away from each other.

  43. Penrose holds that GR is more fundamental than QM, and I'm told that most physicists disagree. I suspect that Penrose is right!

  44. If you don't occasionally consider crazy ideas, you'll miss some very important theories. For example, quantum mechanics was clearly a crazy idea. And Feynman diagrams were also clearly crazy, but they work. Later, Feynman studied negative probabilities to try to resolve the measurement problem, another clearly crazy idea (and in this case, one that didn't work).

    So Penrose's idea of CCC is clearly crazy, and the odds are that it probably doesn't work. But this should not be taken as a reflection on Penrose's intelligence.

  45. @ Peter Shor. Can't edit my responses here, but upon reflection, have realized that I mostly agree with you, except for where you said "Penrose's idea of CCC is clearly crazy".

    No,it's not clear. No more so than other theories not supported by evidence, of which there are quite a few.

  46. Thanks for providing some much needed context. The paper appeared in the media, but without any indication to the layman as to how solid (or fickle) the result was.

    There is a vast scale in science between the theories that are backed up by gazillions of observations and agreed upon by everyone at the one end, and wild speculation at the other end of the spectrum. That nuance sadly gets lost in any scientific news item outside of the ivory tower of science, time and again.

  47. I saw the recent TV biopic of Stephen Hawking and starring Roger Penrose. It was an impressive turn. It was much better than the movie biopic.

    Not only was QM considered crazy, but so was Special Relativity and probably General Relativity. Einstein after all only won the Nobel prize for the photoelectric effect - a much more prosaic physical phenomena!

    I do hope the identity crisis passes and that the crankiness does not.

  48. And here I thought you were “that German lady who says rude things about other peoples theories."

  49. "I saw the recent TV biopic of Stephen Hawking and starring Roger Penrose. It was an impressive turn. It was much better than the movie biopic."

    I'm not sure which one you are referring to. There is an older (15 years?) biopic which is also quite good. The chap playing Penrose really looks the part.

    "Not only was QM considered crazy, but so was Special Relativity and probably General Relativity. Einstein after all only won the Nobel prize for the photoelectric effect - a much more prosaic physical phenomena!"

    In his excellent biography of Einstein, Pais makes the point that Einstein's explanation of the photoelectric was far from prosaic---quite the opposite. Einstein himself said that this was the only time he had been really radical in his entire life, and that is saying a lot!

  50. Tom Aaron - according to Penrose the vast majority of the universes "lifespan" wouldn't be in a state that supports life so it is not at all surprising that we live in that very narrow percentage.

  51. My problem with CCC, a theory I find attractive, is the carryover of features from prior aeons, as for instance Hawking points-- over a finite number of cycles would these not accumulate? How many cycles have we had, an infinite number? Should not the universe consist entirely of carry-over artifacts? I'm not sure how he deals with this, or whether this is actually a problem in the theory.

  52. I will go on to say that Penrose finds the initial conditions in the universe to be very special, VERY unlikely. Inflation isn't adequate to explain the initial uniformity because there is no a priori reason to expect any small region of an initial state to have the requisite uniformity, such that, inflated, you get the intial conditions implicit in the CMB. Also, he observes that reversing time's arrow at any point in the universe's evolution, will not recover sufficiently uniform initial conditions. Long before the point that inflation "ends" one would have a chaotic mess. I don't really see that any of these objections have been properly answered by the critics of CCC.

  53. Only just discovered your marvelous blog, Sabine, thank you (and luv your slightly 'irreverent' sensibility)! Also perhaps one of these days someone can explain to a layman a bit better exactly what happens at the 'juncture' between these 'eons' in Roger's CCC. Especially how a relatively 'massless' universe that's eventually evaporated into photons becomes the basis for a new Big Bang, again emerging from a new presumable singularity... all of which still seems unclear. Or does the concept of 'scale' even at the quantum level become meaningless at each successive 'restart'?

  54. Roger Penrose deserves the Nobel prize next year - I can find many reasons, but his contribution on black holes could be the one after this years first “picture” of a black hole.

  55. Does decay of erebons explain acceleration of expansion of the universe? The half life of erebons hypothesized by Penrose conforms to the change in the expansion. As DM loses its grip through decay, the masses constrained by DM would begin a process of flying apart.

  56. Would the discrepancy in the observation of the Hubble constant from 73.6 km/sec/Mpc in the near universe versus 66.9 km/sec/Mpc at the CMB be an indicator of the decay of DM?



Comment moderation on this blog is turned on.
Submitted comments will only appear after manual approval, which can take up to 24 hours.
Comments posted as "Unknown" go straight to junk. You may have to click on the orange-white blogger icon next to your name to change to a different account.