The Multiverse at Perimeter

This week, the Perimeter Institute is hosting a conference entitled A Debate in Cosmology – The Multiverse. Here’s the schedule, and talks have started to appear on-line here.

Some of the speakers will be discussing the Many Worlds interpretation of QM. It has always mystified me why this sometimes gets put together with the string theory landscape sort of multiverse. It will be interesting to see how many of the speakers address the fundamental problem of the subject, that of coming up with a plausible falsifiable prediction. Lee Smolin has generally put that problem front and center, but he tends to be alone in doing this. The more usual thing in this subject is to go on about what an important idea the multiverse is, then make some sort of excuse for not being able to predict anything with it.

Also dealing with the problem of multiverse predictivity is this new preprint on the arXiv about the landscape, and this laudatory commentary from Lubos. According to him, it’s not much of a problem that one is talking about measuring low energy observables to 500 digit accuracy, when one can’t now even predict their rough order of magnitude.

My colleague Brian Greene will be at Perimeter this week, giving a public talk about his new book Icarus at the Edge of Time. It is being released to bookstores today and I haven’t yet seen a copy, but it appears to be a science fiction book mainly aimed at children, illustrated with pictures from the Hubble Space Telescope. There’s a blog entry about the design of the book here, with some pages of the book here.

Update: Sabine Hossenfelder has reports from the conference and public lecture.

This entry was posted in Multiverse Mania. Bookmark the permalink.

24 Responses to The Multiverse at Perimeter

  1. Michael Bacon says:

    “It has always mystified me why this sometimes gets put together with the string theory landscape sort of multiverse. ”


    I agree with this, but I was curious why you’ve reached this conclusion. Thanks.

  2. Peter Woit says:


    They’re two completely different ideas on completely different scales addressing completely different questions. It’s not really that I’ve reached a conclusion, just that I’ve never seen anyone even try and give a reason why the two should be connected.

  3. John Baez says:

    I think one can try to combine the landscape idea and the many-worlds interpretation by imagining the universe in a superposition of quantum states corresponding to different points in the landscape. Then according to the many-worlds interpretation, all these “branches” “really exist” in some sense.

    I don’t have the energy to explain my own attitude towards this line of thought. But, I have a feeling it’s something a bunch of people have on their minds — perhaps only half-consciously. That might explain why they talk about both kinds of “multiverse” in the same breath.

    It might be good to ask them.

  4. James Robson says:

    “It might be good to ask them”, but unfortunately I don’t move in those circles!

    Surely these are two independent ideas? You may well be able to combine them: and a proper understanding of nature might require such a combination, but logically they are distinct I think?

    We could perhaps imagine a completely classical theory where our local situation depended upon the history of the way things evolved and settled down in our neighbourhood. Perhaps things could have turned out vastly differently elsewhere. So we would have some kind of multiverse where we might use an anthropic principle to explain why we see things to be the way they are around here.

    Or we could have a QM theory that was water-tight and allowed no flexibility and just told us the way things were – except that being QM things would be a superposition of all the states of that fixed world, and a QM multiverse would seek to separate these different possibilities through some kind of “decoherence” which renders the different possibilities effectively non-interacting and so different worlds.

    Don’t believe either myself, but what do I know, not my job 😉


  5. Peter Woit says:


    Since the universe is quantum-mechanical, I’d always assumed that the multiverse was too. So, yeah, you get a quantum superposition. But my understanding of Many Worlds is that it is supposed to deal with the measurement problem; you’re supposed to have Many Worlds splitting off all the time as you do measurements. That’s what seems to be something completely different.

  6. Hal says:

    Max Tegmark has identified four different types or “levels” of multiverse. The level 0 multiverse is composed of the infinity of replications of our local “Hubble bubble” within our universe if it is infinite in size. The level 1 multiverse is composed of all the local bubbles of ordinary space-time that exist within a larger universe in inflationary cosmology. The level 2 multiverse is the Everett many-worlds model. And the level 3 multiverse is the everything-exists level, where all mathematical structures have Platonic existence and structures with sufficient internal complexity correspond to physical universes. Interestingly, this taxonomy does not include the string theory landscape.

  7. Aaron Bergman says:

    But my understanding of Many Worlds is that it is supposed to deal with the measurement problem; you’re supposed to have Many Worlds splitting off all the time as you do measurements. That’s what seems to be something completely different.

    Many worlds is just ordinary quantum mechanics. New worlds don’t split off; the “worlds” are supposed to be different “branches” of the wavefunction that have decohered. It’s more of a lack of interpretation rather than an interpretation, and I don’t think it actually answers any questions, but some people seem quite fond of it.

  8. Mikael says:

    Peter, I think you are a little too obsessed with predictions regarding the discussion about the nature of physical laws and the multiverse. At least currently this discussion is within the realm of philosophy and not science. But well, we know that philosophy can help create great science and Einstein is the biggest example of all. As an example regarding the multiverse think of the hierarchy problem and naturalness. If you believe in the multiverse, you might expect other things what the LHC will see as if you believe in uniquely determined eternal laws. So as a physicist you will probably be working on different things. So in a way philosophy helps to limit the infinite search space of physical theories. Once you have the theory and it meets predictions, it gets science and independent of philosophy. You can then use this theory to refine your philosophy again and so on. Of course it also works the opposit way which you can also see from Einstein. It seems that philosophy lead Einstein into darkness, when he was searching for a unified theory at the end of his life. But this is life I guess.

  9. Peter Woit says:


    I have no problem with philosophy of science of various kinds. Some interests me, some doesn’t. Within that context, discussion of a multiverse is reasonable. My problem is with the way the multiverse is being used not by philosophers, but by physicists at the highest level of the profession. They are not doing this because they like philosophy (many of them despise it), but because they are desperate to find some way to avoid admitting the failure of string theory as an idea about unification. The string theory “landscape” is exactly the sort of thing you get when an idea fails: something you have to make more and more complicated to avoid falsification, never actually predicting anything.

    This use of philosophy by leading scientists to prop up an utterly failed scientific project is what I have a problem with. It is doing a huge amount of damage to the field.

  10. Mikael says:

    one question, do you consider loop quantum gravity (LPG) a failed project, too?

    LPG can not predict the scattering amplitude between gravitons. Why? Because there are no gravitons in the theory. Why? Because Minkowski space is not shown to be the low energy limit of the theory. Worse, one can’t see how it will happen. Not a very encouraging state of affairs I think. The only prediction of loop quantum gravity is a breaking of Lorentz invariance I think.

    On the other hand string theory does some predictions. It predicts extra dimensions. It predicts supersymmetry. Furthermore it offers a solution to the black hole information paradox based on duality.

    Let me add, that I am an outsider and I am asking from honest interest, and not in order to push any view that I have.

  11. Peter Woit says:


    As theories of quantum gravity, both LQG and string theory have their pluses and minuses, this debate has gone on ad nauseam here and elsewhere. The bottom line is that neither has produced a completely consistent and satisfactory quantum theory of 4d gravity, or come up with any way one can experimentally test them.

    My objection to string theory is that it has failed as a theory of what we can observe: the Standard Model. The unwillingness to face up to the implications of that failure are what is damaging the field.

    Saying that “string theory predicts supersymmetry” or “string theory predicts extra dimensions” is a completely empty statement until you specify the scale of supersymmetry breaking or the size of the extra dimensions. String theory has nothing to say about this, and this is why it is an empty failure as far as particle physics goes.

  12. Aaron Bergman says:

    It is not true that LQG predicts Lorentz violation.

  13. Mikael says:

    it is true, that we do not have a consistent and satisfactory quantum theory of 4d gravity. But it is also true, that this statement is in a way empty, because it does not offer any help on what to work on tomorrow. Imagine I was a young student, who wants to work on quantum gravity and I’d asked you on what approaches to study, what would you answer me? Or would you say, I was waisting my time because the problem is just too difficult?

    And I strongly disagree that predictions about supersymmetry and extra dimensions are empty, just because you do not know the scale. One thing to see this is that models based on these ideas already influence what patterns people are searching for in the LHC data. (probably even the detector optimizations). Another way to see this is that Carlo Rovelli was asking the LHC guys on strings 2008, when they will be able to tell if the world is supersymmetric, if it is on the accessable scale. He also mentioned in an interview, that he will have great respect in the intutions of string theorists, if supersymmetry is found. My very personal guess is he would even stop working on LPG in this case. How can an empty statement have such consequences? Also, how can the emptyness of a statement be dependent on our current experimental abilities?

  14. Mikael says:

    please elaborate. Predicting doubly special relativity means predicting a deviation from special relativity, right?

  15. Peter Woit says:


    If anyone wants to work on quantum gravity, fine. They should learn about what the various approaches are, what their problems are, and then see if they can make some sort of progress. The kind of hype and misleading talking points being thrown around in this subject are not helpful at all. If they are going to do this, they also need to understand and think seriously about the fact that, for very good reasons of dimensional analysis, no one knows any way to connect the subject to experiment, and learn anything about it by looking at the real world.

    As far as I know, there is no defnitive prediction of Lorentz violation, or of what the scale of it should be, from LQG. You seem to like to talk about “predictions” when there aren’t any, that’s really misleading and a bad idea. There’s a huge amount of less than honest claims being made these days about “predictions”. I would bet against the LHC discovering supersymmetry, but if it does, that will be a huge revolution in physics, and a vindication of many people’s interest in supersymmetry, most of which has nothing to do with string theory. See David Gross’s comments at Strings 2008

  16. Mikael says:

    I am sorry for the careless use of the word prediction. My point is that there is something in between a prediction and an empty statement and I tried to explain why.
    “Learn about the various approaches and try to make some progress” is an advice so true, that you can apply it to anything.

  17. Michael Bacon says:


    Back to the original topic, I listened to the Albert and Greaves talks — pretty impressive. As you may know, I feel affinity for the MWI, but in all events, I thought the level of enquiry was impressive. Hard questions, good answers all around. Just what science should be.

  18. Kea says:

    Heh, cool! The Banks talk is just fantastic – a must see for great physical intuition and a crucial alternative viewpoint with solid analysis backing it.

  19. Michael Bacon says:

    Haven’t listened to the Banks’ talk yet, but this does raise the question of the relationship between the MWI multiverse concept and the string landscape. I know John Baez mentioned that maybe you can tie them together, but I agree with Peter (if that’s what you meant) that it’s pretty odd wrapping them up in the same conference.

  20. Kea says:

    Actually, from my understanding, Banks is against both the Landscape and MWI: the former because it cannot reproduce the holographic cosmologies they consider, and the latter because he talks loosely in terms of isolated islands of ‘lonely’ universes in the dense black hole fluid, but it may be that he is headed towards a more MWI vision of the multiverse.

  21. Michael Bacon says:

    I think Banks did raise some questions (if it was him) during the Greaves talk, but as far as”interpretations” of QM go, it sounded like his view (to the extent he had one) was some sort of combination of consistent histories with wave function collapse.

  22. Kea says:

    Consistent histories, yes, but wave function collapse is a trickier question. He was careful to explain that he used ordinary vanilla QM only because he didn’t know how to do better quantitatively. On the other hand, his description of emergent geometry (as a kind of collapse from collective observers) is much more sophisticated than the naive universal wave function ideas one often hears touted by MWI proponents.

  23. Michael Bacon says:


    I’ll go back and listen, but I’m pretty certain that he acknowledge some kind of collapse as a way to get our apparently real situation to fall out of the numbers, and I think he said that once that occurs (however it happens) you just set aside the other histories and basically start the process all over again — but maybe my recollection is wrong. Anyway, I thought his questions were thoughtful and probing, as were Hartle’s (I think it was Hartle 🙂 )

  24. Rexus says:

    With all do respect, the multiverse is crap! The many worlds interpretation is the worst model for interpreting quantum theory. While the whole concept of parallel universes might make for a handy plot device to be used in a bad Michael Chrichton novel, I do not see any utility for the hypothesis. There are far better interpretations of wave-particle duality and the measurement problem. For starters, the only reason why humans find wave-particle duality to be problematic is because it is difficult for the average human to comprehend who something could be a particle and a wave. In fact, such entities are not exactly “both particles and waves” rather they are neither. These entities are wavicles, having properties of both waves and particles, but not necessarily being one or the other. Likewise, the Copenhagen interpretation is a useful starting point, despite its shortcomings, but I personally think that Penrose’s hypothesis that gravitation causes the collapse of the wave function has great potential to resolve this measurement problem. Sci-fi babble about parallel worlds violates the principle of Occam’s Razor. Whereas the Copenhagen model, Roger Penrose’s hypothesis, or any other proposed solution is infinitely better.


    1. Falsifiability
    The existence of “other universes” is not falsifiable in principle once we accept that universes besides our own might exist. After all, these parallel realities are not supposed to interact with our own, to be separate, connected only in the sense that they diverged off our universe (or we diverged off theirs). If the universes do not interact, how do we test to verify or falsify their existence. There *COULD* be other universes just as there *COULD* be a sinister conspiracy of Jews and Masons who rule the world (as rediculous as that sounds). Because these universes supposedly exist outside our universe, there is no way to test it. In this respect, arguing with ardent defenders of the many worlds interpretation is akin to arguing with creationists, Holocaust deniers, or 9-11 “truthers.” You cannot “prove” that God (or Satan) placed fossils in the ground to test our faith and that God did not create the universe like in Genesis but made it look as if evolution occurs, or that Bush did not plan the attacks.

    2. Plausibility
    I would argue that, in fact, I can “prove” that the multiverse does not exist. This might seem like a contradiction, but I only argued that the many worlds hypothesis is unfalsifiable if you open to the possibility of it being true. [Once you accept that, it becomes an argument from agnosticism (you can’t PROVE it ain’t so). But this admission that the multiverse is untestable cuts the ground from under the feet of the “many-worlders.” By admitting that the multiverse is unfalisiable, they admit that their speculation is not science, just very bad metaphysics.] However, I argue that to even grant that multiple universes CAN exist is a huge stretch. I can disprove the notion using logic alone.
    [A] The universe, by definition, is all that exists. [The prefix “uni” implies one. The entirety of reality.]
    [B] It follows that nothing else phyiscially exists beyond the universe.
    [C] The multiverse hypothesis suggests that there are other universes beyond ours.
    [D1] These “universes” do not exist (because there is by definition only one universe).
    [D2] These “universes” are not separate universes at all, just part of THE universe.
    A -> B, [(A & B) + C] -> D1 OR [(A & B) + C] -> D2.
    Assuming that the universe = all of phyisical reality, the concept of the multiverse presents an internal contradiction. Either our universe is the only one, or our “universe” is not a universe at all, just a quantum accident. This falls into the whole “reality is an illusion” problem. Curiously, the many-worlders never attempt to explain where these other universes are. If they inhabit the same space-time as our universe, then they are merely overlapping quantum states. This does not solve the problems of quantum phyisics at all!

    3. Simplicity & Parity
    According to Occam’s Razor, the simplest explanation should be expected to be the correct one. Especially because we are not supposed to multiply entities unnecessarily. The many worlds interpretation attempts to solve a mystery with an even greater mystery! Postulating an infinite (or virtually infinite) number of universes to explain an anomaly is very bad science. The Copenhagen interpretation follows logically from quantum theory without being unnecessarily complex. Likewise, Penrose’s solution might work, but we may find out as gravitation is understood in light of quantum theory.

    I think the fact that Hugh Everett’s pipe dream is even taken seriously is a sad indication of the state of modern physics and cosmology. Other than David Deutsch, I was unaware of any REAL physicist (i.e. with tenure, this excludes armchair physicists who watch Star Trek and believe it’s real) taking such speculation seriously. On the other hand, I can see the connection between multiverse quantum cosmology models and string theory, even if Dr. Woit doesn’t. I would not say that the string theory landscape is fundamentally different from the quantum “many worlds” interpretation. Even if they are on different scales, both models attempt to answer discrepancies in observations by creating more universes… Physics has strayed far from the sober analysis of Galileo, Newton, Einstein, Heisenberg, Schrodinger, Dirac, Pauli, and Fermi into the Star Trek pop-sci of Michio Kaku…

Comments are closed.