Is the Multiverse Immoral?

[Warning, somewhat of a rant follows, and it’s not very original. You might want to skip this one…]

In the last week or so, I’ve run into two critiques of the currently fashionable multiverse mania that take an unusual angle on the subject, raising the question of the “morality” of the subject. The first of these was from Lee Smolin, who was here in New York last week talking at the Rubin Museum. I probably won’t get this quite right, but from what I remember he said that discussions of a multiverse containing infinite numbers of copies of ourselves behaving slightly differently made him uneasy for moral reasons. The worry is that one might be led to stop caring that much about the implications of one’s actions. After all, whatever mistake you make, in some other infinite number of universes, you didn’t do it.

Over at Scientific American, yesterday they had John Horgan’s Is speculation in multiverses as immoral as speculation in subprime mortgages?. There’s more about this in a Bloggingheads conversation today with George Johnson, where Horgan describes his current reaction to multiverse mania as “I can’t stand this shit.”

I’m in agreement with Horgan there, but my own moral concerns about the issue are different than the ones he and Smolin describe. The morality of how people choose to live their everyday lives doesn’t seem to me to have much to do with whatever the global structure of the universe might be. The world we are rapidly approaching in which a multiverse is held up as an integral part of the modern scientific world view isn’t one in which many people are likely to behave differently than before, so I don’t share Smolin’s concerns. Horgan’s exasperation with seeing the multiverse heavily promoted by famous physicists appears to have more to do with the idea that this is a retreat by physicists from engagement with the real world, something morally obtuse in an era of growing problems that scientists could help address. For what he would like to see instead, I guess a good model would be John Baez’s recent decision to turn his talents towards real-world problems facing humanity, see his blog Azimuth for more about this. Personally, I’m not uncomfortable with the fact that many mathematicians and physicists find that they don’t feel they are likely to be of much help if they go to work on the technology and science surrounding social problems. Instead, one can reasonably decide that one has some hope of making progress on fundamental issues in mathematics or physics and choose to work on that instead. One can try and justify this by hoping that new breakthroughs will somehow, someday help humanity, although this may be wishful thinking. Or one can argue that working towards a better understanding of the universe is inherently worthwhile, so pursuing this while taking some care to avoid worsening one’s local corner of the world is a morally reasonable stance.

My own moral concerns about the multiverse have more to do with worry that pseudo-science is being heavily promoted to the public, leading to the danger that it will ultimately take over from science, first in the field of fundamental physics, then perhaps spreading to others. This concern is somewhat like the one that induced Alan Sokal to engage in his famous hoax. He felt that abandonment by prominent academics of the Enlightenment ideals exemplified by the scientific method threatens a move into a new Dark Ages, where power dominates over truth. Unfortunately, I don’t think that revelation of a hoax paper would have much effect in multiverse studies, where some of the literature has already moved beyond the point where parody is possible.

For a while I was trying to keep track of multiverse-promoting books, and writing denunciatory reviews here. They’ve been appearing regularly for quite a few years now, with increasing frequency. Some typical examples that come to mind are Kaku’s Parallel Worlds (2004), Susskind’s The Cosmic Landscape (2005), and Vilenkin’s Many Worlds in One (2006). Just the past year has seen Sean Carroll’s From Eternity to Here, John Gribbin’s In Search of the Multiverse, Hawking and Mlodinow’s The Grand Design, and Brian Greene’s new The Hidden Reality. In a couple weeks there will be Steven Manly’s Visions of the Multiverse. Accompanying the flood of books is a much larger number of magazine articles and TV programs.

Several months ago a masochistic publisher sent me a copy of Gribbin’s book hoping that I might give it some attention on the blog, but I didn’t have the heart to write anything. There’s nothing original in such books and thus nothing new to be said about why they are pseudo-science. The increasing number of them is just depressing and discouraging. More depressing still are the often laudatory reviews that these things are getting, often from prominent scientists who should know better. For a recent example, see Weinberg’s new review of Hawking/Mlodinow in the New York Review of Books.

While most of the physicists and mathematicians I talk to tend towards the Horgan “I can’t stand this shit” point of view on the multiverse, David Gross is about the only prominent theorist I can think of known to publicly take a similar stand. One of the lessons of superstring theory unification is that if a wrong idea is promoted for enough years, it gets into the textbooks and becomes part of the conventional wisdom about how the world works. This process is now well underway with multiverse pseudo-science, as some theorists who should know better choose to heavily promote it, and others abdicate their responsibility to fight pseudo-science as it gains traction in their field.

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

107 Responses to Is the Multiverse Immoral?

  1. Adam F says:

    Biology has certainly not fallen into pseudoscience, but we are running into situations where a disturbingly large proportion of studies (particularly in medicine) turn out to be false. The everyday upshot of this is the constantly changing ideas about what foods and drugs are healthy. It gets very difficult to tease one small cause-effect relationship (say, genes or environmental conditions) from a huge mass of other contributing factors, all of which interact with each other. There are still plenty of areas where relationships are strong and research can be usefully done, though.

  2. Steve Baker says:

    So the multiverse theory must be wrong because it causes immorality? Wow – cogent argument!

    We can’t prove or disprove the multiverse theory – and there are other alternative theories that we also can’t prove or disprove. There is no special reason to believe or discard any of them.

    However, the moral issue is a specious one. Would you also condemn the possibility that the universe is infinite on moral grounds? You need to because in an infinite universe, there are also infinite numbers of “parallel earths” in which decisions you make come out differently.

    Really, we all need to simply talk about what is possible and what is not – and look for more clues to tell us which is true.

    Personally, I’m becoming somewhat impressed by “The Simulation Hypothesis” (look it up on Wikipedia) – it explains an awful lot about some of the seemingly arbitary things about our universe.

    However, we can’t either prove or disprove that one either.

    In the end, all we can do is to fall back on weak guidance from Occam’s razor.

  3. Peter Woit says:

    Steve Baker,

    I’ve not claimed either that the multiverse is wrong or that it is immoral or that one implies the other.

    What I do claim is that the string theory multiverse is not legitimate science since it is inherently untestable. Note that I don’t claim all multiverse theories are untestable, I’m specifically referring to the ones promoted as an implication of string theory. Also note that you need to examine exactly what these theories say to see what their problems are and just how unlikely it is that they can be overcome. This is a complicated subject, but what’s not complicated is the current situation: these theories currently make no predictions at all that can be used to test them. It’s up to proponents of the idea to come up with a plausible avenue to getting such tests. I don’t think they can, and I haven’t seen them do it when confronted with the question.

    I don’t share Smolin or Horgan’s particular concerns about morality, I was just reporting them. I do have my own concerns about the morality of how the string theory multiverse is being promoted, but this is not about the “morality of the multiverse”, it’s a moral question about how science is done.

  4. Clark says:

    The moral argument is weird. It’s akin to the argument I’ve heard some philosophers make that a block universe (i.e. taking GR seriously) must be false as there then isn’t the kind of free will many philosophers think is necessary for responsibility. It’s an odd kind of reasoning undermined further by the fact, as you note, that in practice no one is probably going to change their behavior anyways.

    It’s odd Smolin goes so much against the multiverse theory since his own quasi-evolutionary theory doesn’t seem that different. Although it isn’t quite saying that anything goes.

    What I don’t get is why folks like Greene don’t just accept that they are doing philosophy rather than science. I suspect it’s a remnant of the anti-philosophical tendency left over from Feynman’s influence. But honestly philosophy can be interesting. It can even be rigorous. It just shouldn’t be taken as a scientific statements we have relative confidence in.

  5. Brian Knoblauch says:

    While the multiverse theories may not currently (or ever) be able to be experimentally verified/debunked, I don’t buy the morality argument. As long as the science being done is still sound, and the multiverse theory is not presented as “knowledge” (merely as speculation), I have not problem with it. Speculating on all kinds of weird and exotic things has been a source of inspiration for others to make REAL discoveries/developments. Dragging morality into it this is pure silliness. If you’re doing science the right way, “morality” should not be a concern. It either is right or wrong scientifically, there is no wishy-washy emotional component. By dragging morality into it, you defeat the science.

  6. Peter Woit says:


    The problem here is that pseudo-science is being passed off as legitimate science to the public. One reason this is being done is to evade the implications of the scientific failure of string theory unification. This situation seems to me to raise moral issues.

    Just to give one example, this is from a recent article on the subject:

    “since almost all contemporary physics accepts and even demands the reality of parallel worlds. Love it or hate it, the multiverse is here to stay.”

  7. Perhaps I’ve got the wrong end of the stick here, but is the “multiverse” any different from the Many Worlds interpretation of standard quantum mechanics, which people have been kicking around for decades? If not, how exactly have things gotten any worse in the context of these new books?

  8. Peter Woit says:


    The controversy over the multiverse is about something different, the idea that string theory implies a multitude of completely separate universes with different physical laws. This is quite different than many-worlds, which is an interpretation of standard quantum mechanics, with one fixed set of physical laws. The two very different ideas are getting mixed together often, I’m not sure why.

  9. Nick says:

    Why do they get confused? Even though I have a B.S. in Math and follow science regularly, I didn’t know the distinction between multiverse theory and many-world theory until now.

    Science journalism for the general public needs an upgrade. Reputable newspapers, etc. have a difficult discerning the legit. from the pseudo science, sometimes degrees in the specific field are required. Any journalist wanting to do serious science reporting should start with a college level statistics course.

  10. Peter Woit says:


    I don’t think this can be blamed on journalists. The physicists promoting the “multiverse” tend to talk about these things together, without making the distinction, confusing the issues. On the whole I’ve found that science journalists do a pretty good job of reporting what scientists tell them.

  11. Michael T. says:

    It should come as no surprise to learn that scientists are not immune from self-promotion. The multivese is clearly a metaphysical concept at best but it does sell and the public eats it up big time. You can get a book deal, a few interviews on prime time, maybe even your own TV show. It is the era of the “scilebrity” so why shouldn’t they hustle a buck 😉

  12. Tay Owens says:

    Many concepts which are now accepted as scientific fact were once derided as unfalsifiable and hence unscientific. Just because they are unfalsifiable with today’s technology and limited knowledge doesn’t mean all multiverse theories will always be unfalsifiable, and are therefore pseudoscience. Indeed, Greene takes pains to point out which of the 8 kinds of multiverses he details may be testable in the future (and why) and which most likely will remain unfalsifiable. Perhaps then we can view some multiverse theories as scientific theories, and others as philosophical theories. I rather like that idea, particularly as it brings physics closer to its roots as Natural Philosophy–a heritage to be proud of, not ashamed of.

    Indeed, modern physics has often been preoccupied with matters that are perhaps more philosophical than strictly scientific in a Popperian sense–or, at least they seem so until much later when the math or technology or observations finally catch up to the theorizing (philosophizing). The chase after Grand Unified Theories or Theories of Everything or any unification of currently separate physical idioms is as much about producing a philosophically pleasing merge of disjointed models as it is scientific–after all, nothing precludes a universe that obeys multiple sets of laws at differing scales or states; we just don’t like the idea on philosophical grounds and seek a more elegant and parsimonious solution. We’ll probably find it eventually, at which point it will really be pure science; but until then it’s just philosophy in science’s clothing. Sounds a lot like multiverse theories to me. Does that make every physicist seeking more unified theories than current models and maths and observations accommodate a pseudoscientist?

  13. srp says:

    I’m a little puzzled by the oft-made claim that the SM is so good at matching the data that theorists have nothing to do. Lederman says that the gap between the observed vacuum energy and what theory says it should be is off by many orders of magnitude. And I seem to recall hearing about some gross anomalies in transverse-spin polarized collisions (Krisch?). Then there’s dark matter. So I guess I don’t see why everyone chases unification when more tangible prey seems to be available.

  14. Peter Woit says:


    In the SM the vacuum energy is a free parameter, you can make it anything you want. It’s only when you try and work with unified theories of the SM and quantum gravity that the CC can become in principle calculable and you have to find an explanation for why it is so small.

    There’s no observation at all of dark matter or anything related to it in a laboratory experiment. It’s an astrophysical phenomenon. The fact that standard models of cosmology and galactic structure seem to point to some sort of gravitating matter that we don’t understand is an interesting hint, but it’s not a lot to go on. In any case though, there’s a huge concentration of effort by theorists to pursue this hint.

    The anomalies in polarized proton collisions are a topic I wish I knew more about. They probably do deserve a lot more attention. Betting though seems to be that they aren’t evidence of a problem with the SM, but with the fact that our understanding of how to do non-perturbative calculations in QCD is still quite crude.

  15. Phrustrated Fizix Stoodint says:

    Peter, first thanks for your book and this weblog, it has saved me many hours of not pursuing dead ends.

    From your book, I have a quick question. On the last page of your book, you mention two directions that quantum gravity can go in, involving diffeomorphisms. A brief update on progress towards those goals since you wrote the book would be appreciated.

    From this weblog and your last reply, we’d be interested in knowing if dark matter phenomenology is still the hottest field in theoretical physics (it was last year, yes?) and whether or not in your humble opinion it would be worthwhile putting effort into it at the present time. I’m not a herd animal and wish to pursue something unique, is why I’m asking. Entering a crowded field is unappealing to me and my friends.

  16. Peter Woit says:


    Dark matter is still a very hot subject, but there’s a good reason for that: not much else in the way of data that disagrees with the standard model. Everyone is hoping that the LHC will soon change this.

    There hasn’t much new learned about the problem I mention in the book, how to handle diffeomorphism and gauge symmetries in QFT. Personally I’ve spent a lot of time learning about how to think of Langlands and geometric Langlands in terms of representation theory, and hope that this might lead somewhere, but that’s a long way off.

  17. cormac says:

    Hi Peter, how are things:
    my understanding is that,even with ST, the basic concept of the multiverse arises out of cosmic inflaton i.e. the possibity of bubble universes is raised by many of the current models of cosmic inlation.Have I got this wrong?

  18. cormac says:

    ops, that should of read “without ST”

  19. Peter Woit says:

    Hi Cormac, things are good!

    Sure, you get bubble universes in some models of inflation. But what’s new with string theory is using this as an excuse for not being able to predict anything. People making models of inflation with bubble universes can argue that their model doesn’t predict certain things because those things are different in each universe. But they have to predict something to be taken seriously.

    String theory is a unified theory, so in principle supposed to explain everything about fundamental physics. It doesn’t actually work, but instead of admitting this, string theorists try and use the multiverse to explain why they can’t predict anything at all. To me, that’s where there’s a moral problem…

  20. Stefano says:

    Peter, I enjoyed your book and visit your blog often. Now, not sure this comment belongs to this posting, but anyway… what I don’t understand is why you and Lubos are such “enemies”. Sure he goes for string theory, but on the other hand he is pretty clear about that other great bs science of day, i.e. “climate change”, just like you are about string theory. Now, I have no personal reasons to argue that climate may not be changing or even warming – but what I do see clearly is the bs articles popping up everywhere tying AGW to even the most unrelated topics. It’s clear that researchers have understood where the money is and what they need to say to get access to it. New Scientist, Scientific American and even something relatively neutral like have become nauseating in that respect. It’s just like string theory. Bottom line you and Lubos clearly disagree on a lot of things, but you both can only take so much hypocrisy and bs… I wish the two of you noticed that commonality. (Sorry for the ranting style – I think this is why I commented for this particular post.)

  21. Peter Woit says:


    Actually I do agree with Lubos about a lot. He’s not much more in sympathy with the multiverse than I am.

    About climate science though, I don’t think it’s much like string theory at all. Public discussion of the issue is even more politicized and dominated by ideologs than string theory. I’m doing my part to fight this by stopping discussion of the subject here.

  22. chris says:

    “Maybe a branch of science is ripe for infection by pseudoscience whenever it stops making enough progress to satisfy the people in that field But is this really true? Does anyone know other examples, beside the current situation in fundamental physics?”

    I don’t really know this too well, but astronomy after Ptolemy seems to fit the bill. Or actually Ptolemy can be viewed as part of the decline already.

    Another example that pops up in my mind is the scholastic period in the 13th-14th century and the associated mutation of science into comments on Aristotele.

    In modern context i’m really tempted to say economics (as a science) and psychology of the Freud and Adler kind. But i don’t know how much one can count these as sciences to begin with.

  23. Allan Rosenberg says:

    If multiverse-based theory (or its promotion) is immoral in our universe, does that mean it’s immoral in all other universes?

  24. Peter Woit says:


    If the laws of physics are different in different parts of the multiverse, I don’t see why laws of morality shouldn’t vary too…

  25. Allan Rosenberg says:

    Great, we finally have a physical definition of the term “moral landscape.” But I think we should reserve judgment on the morality of the multiverse in our universe until the string theorists can at least let us know which of the many universes we inhabit.

  26. Matt says:


    I think in your next posting on the multiverse controversy, it might be worth taking a moment to distinguish clearly between the different notions in play. There’s the very old notion of the “many worlds” interpretation of quantum mechanics, going back to the work of Everett and DeWitt, in which the different universes are branches of a wave function.

    Then there’s the more modern notion of a multiverse—the one you’re referring to in this post—in which different universes are realized at different locations in physical space. But there are two theoretical underpinnings to the idea, one coming from cosmology and the other from string theory.

    Even putting string theory aside, the most successful models of inflation, which are quantitatively consistent with observational evidence from fluctuations in the cosmic microwave background and resolve myriad inconsistencies of the old Big Bang model (horizon problem, flatness, relics, structure, etc.), require at least 60 e-foldings. That essentially inevitably produces a vast, vast space in which our entire observable universe is smaller than an atom in comparison.

    There’s no string theory here, and this simple consequence of inflationary cosmology that our observable universe really and truly is but a mere speck in a far more immense space immediately raises the question of what the heck else is out there. Is it just empty space, is it stuff that looks pretty much like our own universe, or is it totally different? Given that the early universe exhibited energies far beyond those at which the Standard Model breaks down, there’s as much reason to believe the third possibility as there is to believe that ice crystals forming at isolated, separated locations in a giant vat of water will all line up in parallel.

    Is it really unscientific to be asking these questions, or thinking about them, or looking for observational signatures (CMB patterns, evidence of bubble collisions, etc.)? What string theory brought to the conversation was merely a candidate theory of quantum gravity that leans to the third possibility, but that possibility is still hanging there whether string theory is correct or not. What do you say to that?

  27. Matt says:

    Typo— should read ” … will all line up in different directions.”

  28. chris says:


    the cosmological “multiverse” you try to construct is really not a multiverse. if you just wait for long enough (and assuming that the pressure to energy ratio of dark energy is at least not <-1) you will see. our horizon is expanding constantly and nothing too shocking is expected to come into view.


    because all relevant cosmological transitions are not first order as it seems now. there are no bubbles or cosmic strings as far as we can tell – nothing that would resemble a multiverse at all. so unless the underlying physical laws themselves are spacetime dependent, i guess it is rather boring.

    and by the way, what is the scale that the SM breaks down? last time i checked it can be somewhere beyond the Planck scale if the Higgs mass is around 150GeV.

  29. Peter Woit says:


    Some of the books I mentioned treat both many-worlds and cosmological multiverses, and the authors seem to thing these two things have something to do with each other. I’ve never been able to understand this.

    I’ve no problem with people looking for observational signatures of bubble collisions or whatnot in various inflationary models. But it seems unlikely that the size of such effects is exactly such that they’re just barely visible in the CMB data, and there’s no underlying model that has experimental support that gives any reason to believe this is something reasonable to expect. So, it seems likely that all claims about this have about as much to do with the structure of the universe as does the appearance of Hawking’s initials in the data. If people want to do this fine, but they should also be a bit careful about participating in articles in the popular press about about “scientists find evidence for the multiverse.”

    String theory doesn’t “lean to” a multiverse or anything else, it’s compatible with essentially anything. Instead of admitting that this means it’s a failure, the people backing it have adopted the morally dubious tactic of invoking a multiverse scenario carefully constructed to not predict anything. This is not science, but an attempt to use pseudo-science to prop up a failed enterprise.

  30. Matt says:

    In the interests of clarifying matters, let me ask you a series of questions:

    1. Do you agree with most cosmologists that inflation is the leading candidate for resolving the troubles with the old Big Bang model, and that its quantitative predictions about fluctuations in the CMB have been quite successful?

    2. Do you agree with most cosmologists who say that you generally need 60 e-foldings of inflation for the models to work?

    If you say yes to both 1 and 2, then you are basically led to the conclusion that the true universe is really vastly larger than the region we can currently see. You might call this vast expanse the “megaverse.”

    3. Now there are two possibilities for this megaverse—it’s either empty, or it’s not. Do you think it’s likely to have lots of other stuff in it?

    If you say yes, then you’ve already accepted that the megaverse is really a kind of a multiverse. Maybe those other regions are filled with matter similar to the stuff in our own observable region—galaxies, planets, etc.—and maybe no two regions are alike. So it’s not an “everything”-verse. But if you regard as scientifically acceptable the previous propositions, then you basically accept that we live in some kind of a multiverse.

    4. We can go further, of course. Generically, models with inflation never end—inflation is always going on somewhere, because inflationary space expands far faster than patches can freeze out and stop. This is eternal inflation. Where do you stand on this question?

    You may or not be on board with that idea, but it would obviously mean that we’ll never see most of the multiverse, no longer how long we wait. In that sense, most of it becomes unobservable.

    5. Do you regard speaking about those other parts as unscientific and out of bounds?

    6. Finally, there’s a larger notion of a multiverse where all those hypothetical low-energy vacua of string theory are realized in various places in space. Again, however, this doesn’t mean that there’s a copy of you where everything possible happens. If the set of possibilities is infinite and unbounded, then even an infinite set of bubbles won’t necessarily explore them all.

    7. Lastly, there’s the idea that every single possibility is literally realized somewhere in some bubble. That’s where the multiverse becomes an “all-verse.” I definitely know that you regard this last possibility as pseudoscience, and it’s what Smolin and others are morally objecting to.

    I think it’s important not to conflate all these ideas. There are serious scientists who find themselves at various places in this list. Where precisely do you think is too far?

  31. Matt says:


    And note that string theory only came up in question 6 on the list. The landscape is not the same thing as the multiverse, as most cosmologists will be very careful to explain.

  32. Cosmonut says:


    2. Do you agree with most cosmologists who say that you generally need 60 e-foldings of inflation for the models to work?

    If you say yes to both 1 and 2, then you are basically led to the conclusion that the true universe is really vastly larger than the region we can currently see. You might call this vast expanse the “megaverse.

    This is not really true from what I have read.
    60 e-foldings during inflation would lead to a homogenous universe of the same size, roughly, as what we see.
    That is how the “60 e-folds requirement” is obtained.

    Now, number of e-foldings to go beyond 60 — no one knows how many more, because there are dozens of models with no and no idea which is right.

    IF that happens, then our “inflationary bubble” could be much larger than the observable universe.

    But there’s no reason to believe that there would be different laws of physics operating in other parts of the bubble. The default case would be the same laws all through the bubble.

  33. Peter Woit says:


    You don’t appear to pay any attention to the answers I already gave you, so I don’t see why I should spend time answering the same questions again.

  34. Matt says:

    Peter– My point was just to make clear where you stand, but I understand if you no longer wish to discuss the matter.

    Cosmonut— No, 60 e-foldings is necessary for our observable region to have the correct properties, such as flatness, but the consequence is that it’s then a tiny patch of a far larger space. Think of what you have to do to a balloon to make a tiny surface patch look look very flat–you need to blow up the balloon to a size far more tremendous than that patch.

    One piece of advice to you folks that will help you make your case more successfully. (And I’m somewhat sympathetic to that case, by the way.) Critiques of the multiverse would be more convincing if people didn’t conflate all the relevant issues, and made more clear exactly what they accept as reasonable and what they don’t. Is that asking too much?

    Take care everybody.

  35. Emile says:

    Dear Matt,

    Science is nothing without the part where we test our hypotheses. There is nothing wrong with starting with observational evidence of the early Universe and trying to figure out the implications. You can then formulate new hypotheses that hopefully will eventually get tested. Until you have figured out how to test those ideas, they are just ideas: call it scientific speculation if you must (but it is speculation). What makes science so powerful is the testing part. I think people are doing damage to its good name when they confuse speculation with science. The lay public is on to us. Look at some comments in newspapers and elsewhere and you’ll read sentences like this one:”Modern quantum theory has gone into the never, never land of all theory and no proof, to the point where it’s become a religion unto itself, whereby its all based on faith and no fact”. That comment was related to a story about religion (I’ve started to collect these comments). It’s unfortunate that quantum theory (think of the anomalous magnetic moment of the electron) is getting dragged into this. I personally think that these ideas about multiverses (even the landscape) are quite likely. But what I think doesn’t mean squat. I’ll call it science when we have figured out how to test it. And if it is forever untestable, than it will never be science.

  36. Matt says:


    Are you a practicing physicist who works in high energy theory and cosmology? I ask not because the opinions of lay persons are inferior in any way–in particular, you make very good points–but that lay persons simply don’t get a chance to see what the people who actively work on this stuff actually think about everyday. Peter has a lot of personal bias, as do we all, and he won’t tell you.

    I can assure you from much professional experience that whatever Hawking or Greene or Susskind or Weinberg (none of whom are very active on the research side any more) say in public, the people who actively study this stuff are absolutely obsessed with observability. Tests are very much on their minds. People are constantly devising schemes to look for signatures from the CMB, or evidence to favor one measure over others for use in making probability statements about features of our observable universe. (See the measure problem.) And that’s only scratching the surface.

    We are led almost inexorably to many of the central ideas of the multiverse even without invoking string theory, as my list earlier attests, and I can assure you that many researchers were dragged along unwillingly by logic and strong indirect evidence from cosmology, which has become an extremely precise subject in that past decade. (Peter diminishes this fact because it detracts from his belief that string theory is to blame for everything he doesn’t like about contemporary high-energy theoretical physics.) Weinberg made his first use of multiverse arguments way back in the eighties when he correctly ballparked the as-yet-unobserved cosmological constant through anthropic reasoning, ten years before it was first measured and without using any string theory. Linde and Vilenkin are not and never were string theorists.

    There’s an enormous diversity of opinions on even the notion of the multiverse itself, even before connecting it to the landscape of string theory, and top people fall in many different places on my list earlier. I was hoping Peter himself would take a precise stand.

    We live in the Internet age now, so you don’t have to take my or Peter’s word for any of it. Go online and find some videos of research seminars by active people studying this stuff, and it’ll give you a real picture of what’s on their minds these days.

  37. Peter Woit says:


    From the e-mail that “Emile” left, he’s a professional high energy physicist. I gather that his opinion is that there may very well be a multiverse, but at the moment it’s not a scientific issue, and my impression is that this is probably a majority opinion among working physicists.

    Those who work on trying to figure out how to get observational evidence for a multiverse should keep in mind that most trained people in the subject are skeptical about what they are doing. This is not because they are ignorant, but because they’ve seen nothing much substantive come out of these efforts. There’s only so much mileage you can get out of the anthropic argument for the CC. People who want to work on how to turn multiverse speculations into solid science are welcome to, even if the odds of success are long, but until they’ve made some progress they might want to stop engaging in high-powered hype aimed at the public.

  38. Matt says:


    My point is simple, and I think we’re talking past each other.

    The basic reasons for the multiverse do not come from string theory or the landscape, but from cosmology. Unlike many ideas in contemporary high-energy physics that people are looking for—supersymmetry, small extra dimensions, KK modes, brane-worlds, microscopic black holes, etc.—the basic underpinnings of the idea of a multiverse were not generated in any way by string theory, but from many directions and much indirect evidence (some of which is too technical to make sense to the public, unless they know how differential geometry or Einstein’s equation work) in cosmology and inflation.

    The cosmic inflation models that we use and that agree excellently with the data generally have as a side-effect the generation of a multiverse of some kind. So writing, even in a public forum, that we’re likely living in a giant multiverse is perfectly within the realms of science. That’s what the data are telling us.

    Where I agree with you is that we don’t know very much about what else is in that giant multiverse. Here’s where string theory finally comes in: It suggests something about the multiverse, namely, that the multiverse consists of regions that look very, very different from our own region, as they have frozen out into different low-energy solutions with different low-energy particle menus and effective laws of physics.

    Some people go even a step further, and claim that the multiverse is an infinitely big space and that every conceivable (and inconceivable) possibility is realized there, with copies of you and me doing and living every possible way imaginable.

    Now that’s entirely speculative, and we have no evidence to support. If you believe that those speculations are not scientific, then I’m in agreement with you, at least until people find a way to test them.

    I worry that you are conflating the entire notion of the multiverse, which has strong backing from many directions in cosmology, with string theory and the landscape, when they’re two conceptually different (but connected) ideas. Writing publicly about the cosmic multiverse and how it follows very convincingly from what we know about cosmology (and the search for signatures like bubble collisions, etc.) is perfectly scientific. Inserting string theory and making other metaphysical claims about all possibilities being realized in the multiverse is not.

    Is that fair?

  39. Marty says:

    Hi Matt,

    (Am not sure if you’re the Matt I think you are, but if you recognize my name you probably are. In that case, Hi!)

    I’m sympathetic to your interest in hearing Peter’s stand on some of your questions to him. I too would be interested… However, in partial defense of his response, I think he probably did state his overall position; it’s just that his position kind of preempts your other questions, if I understand his point of view correctly. I think where he’s coming from is this: It’s perfectly legitimate to start with a set of assumptions and use them to make theoretical arguments for a multiverse, bubble collisions, and so on; but, without feedback from Nature (i.e., unambiguous observations of predicted consequences of those theories), there is no compelling reason to accept that the original assumptions and the theory they imply are correct. Theoretical reasons alone aren’t enough… Peter has made it clear over time that his great respect for the Standard Model comes from the huge number of predictions it makes that have been verified, and the lack of predictions it makes that have been observably shown to be false. I think he also finds certain elements of the SM to be elegant as well, but that is secondary. (Peter, if I have misrepresented you here, my apologies.)

    If you are “that Matt” 🙂 you know I have done some work with bubble collisions, the unfortunate conclusion being that they probably can’t be unambiguously observed except in very “atypical” scenarios. That was enough to discourage me from working further on the problem in any significant way. To me, it just doesn’t feel enough like “good science” to rely on theoretical “necessity” and consistency arguments; in the end, there needs to be solid experimental confirmation (as I’m pretty sure you agree), something I’m not optimistic will happen here.

    You mention,

    We are led almost inexorably to many of the central ideas of the multiverse even without invoking string theory, as my list earlier attests, and I can assure you that many researchers were dragged along unwillingly by logic and strong indirect evidence from cosmology […]

    This seems to be the standard wisdom, but I’m going to give a different take on it. First to state my own position: I personally can’t imagine how any process that generated our visible universe wouldn’t also generate other “universes,” and the number of such occurrences would almost certainly be infinite. Moreover, I find the evidence for an inflation-like era in the very early universe to be very compelling and that the idea is almost certainly correct. But… I personally don’t find compelling the argument that the second viewpoint implies the first. It all comes down to a question of theoretical uniqueness: Does an inflation-like era in the early universe imply an early de Sitter phase caused by one or more scalar fields tarrying for awhile (e.g. in a local minimum of the potential), or are there alternative pictures that can also yield something that looks like inflation and is fully consistent with all observations? It isn’t enough to dismiss this question by noting that no other promising candidates have come forth yet, since both uniqueness of the inflation picture or missing insights (or imagination) can equally well explain the lack of alternatives.

    After all, it’s not as though the traditional picture of inflation is elegant in its crucial details. You know well that it is extremely hard to come up with an inflaton potential that is at all “natural” (so hard that nobody has figured one out yet, as far as I know) — the ones that are natural don’t work, and the ones that work are contrived. (Potentials that rely on the correctness of the idea of a string theory landscape for their existence fall into the “contrived” category, in my opinion.) There’s also the issue of the inflaton itself; this particle must be postulated as the driver of inflation, but there is no other evidence for it except that it seems necessary to explain inflation. In fact, given that the search for the Higgs has turned up empty-handed thus far, we don’t (yet, anyway) have experimental evidence that any fundamental scalar particles exist in Nature. Hence, the reheating phase is also speculative; if we can’t observe the inflaton, then we must rely on our speculation that the phase even exists, and further speculate on what a QFT would look like for the inflaton. It seems apparent that none of these problems can be addressed by observation, so it’s hard for me to see how there is any inevitability to our current models of inflation or their consequences in terms of an inflation-inspired multiverse.

    My own viewpoint is that reliance on an unnatural scalar potential (or set of scalar potentials) to drive inflation in just the right way to give us the CMB we observe, as well as the postulation of scalar particle(s) whose sole purpose and consequence is to drive inflation, are significant conceptual problems. In fact, for the past couple of years or so I’ve been working hard on an alternative scenario that seems much more “natural” to me; in it, an inflation-like era is an inevitable part of a process of spacetime emergence, ending because spacetime becomes sufficiently homogeneous and isotropic and not because of certain characteristics of weird potentials. This blog isn’t the appropriate place to discuss that or other alternatives; I only bring it up to make two points. First, it suggests that the current inflation picture isn’t unique in the sense I discussed above. Second, questions like those related to the measure problem or “Boltzmann brains” aren’t even well-posed questions in the alternative scenario, suggesting that at least some of our current thorny conceptual problems are artifacts of our current (still partly speculative and non-unique) early cosmological picture rather than inevitable difficulties that will arise in any plausible inflation-like scenario.

    (Just as I was about to post this, I noticed that Matt responded to Peter. Sorry if what I just wrote doesn’t take into account anything Matt just said.)

  40. Peter Woit says:


    I mostly agree with your last comment. Hopefully we’ll learn more about inflation in coming years, but I’m skeptical we’ll learn much relevant to the question of whether inflatiion actually does produce multiple universes. Maybe someday we will, which would be a good time to write popular books on the subject, not now.

    Many thanks to Marty. He has thought a lot more about this than me, and taken the time to write out a detailed analysis of reasons to be skeptical about some parts of inflationary theory. His seems to me a better-informed version of my own skepticism about some aspects of the subject.

  41. Matt says:


    Who says arguing on the internet can’t ever produce agreement? (Or quasi-agreement; I’ll take it.)

    Just a last point. Evidence for inflation and the more general notion of a multiverse is in a different category from string theory/the landscape. There is no actual physical evidence for the latter, nor does it make any hard predictions, but several pieces of evidence and verified predictions for the former. So the former is not in the same category as the latter, even if it’s not on quite the same rigorous footing as the Standard Model. The multiverse is not the landscape, and that’s an important thing to make clear. I worry when I hear people like Horgan, who don’t seem to make that distinction.

    We have evidence that our observable universe is just a small part of a far vaster space, which we call the multiverse, even if we don’t know what the rest of it looks like. (And even if we don’t yet know exactly how inflation works in detail.)

    Would it be fair for a cosmologist to write a popular account of the logic and evidence that has led serious people to that conclusion, as long as they don’t veer off into wild speculation about the contents of the rest of that multiverse, and don’t make strong claims about the reality of string theory? Is that scientific?

  42. Emile says:

    Hi Matt,

    Peter essentially answered for me. My point is also echoed in Marty’s post above: “To me, it just doesn’t feel enough like “good science” to rely on theoretical “necessity” and consistency arguments; in the end, there needs to be solid experimental confirmation.”

  43. Peter Woit says:


    If a cosmologist would write a popular book explaining exactly what the evidence relevant to inflation is, as well as what the models people are looking at are (including ones with some sort of multiverse) and how they match the evidence, that would be great, I’d definitely buy a copy. All the ones I’ve seen though tend to go on at great length about speculative ideas that are supposed to be “exciting”, ignoring or minimising their testability problems, and making really dubious arguments in favor of these ideas.

    In more technical expository articles aimed at physicists, there’s a lot less of this, but still they’re often written in a way that makes one feel that one is only being told one half of the story.

  44. Matt says:


    Wait, so you’re upset that cosmologists who write for the public are taking their more rigorous work as springboards and then also going off in speculative directions they find exciting and inspiring? Why do you think they went into science in the first place? What do you think inspires a lot of the kids to want to do physics and cosmology?

    As long as they make very clear what’s speculation and what’s not, I think it’s perfectly fair for them to start with what we’re pretty sure about and then talk about exciting future possibilities. There’s criticism, and then there’s being a grinch. It’s a fine line sometimes, but worth keeping in mind.

    This is actually a larger problem I have with some of your own public outreach. Science needs criticism, maybe more so than it’s been getting. I would certainly concede that, and raise you a nickel. Your points are quite valid.

    But too much negativity can be extremely counterproductive, and turns off a lot of young people, not to mention potential allies. It really does. Rather than pointing students in better directions, or brining in people who are sympathetic to your point of view (as I am), it can have the effect of just driving them away from wanting to get involved altogether. (I presume that’s not your intent.) I speak from personal experience, by the way. I’ve seen fellow students react to it while I was coming up through my physics schooling.

    That’s one reason why I got so upset when I saw this posting on your blog. You seemed to be tossing in a lot of cosmology—some of it speculative—together with your diatribe against string theory. Is there any way you can be a little less negative and vitriolic, a little less broad-brush, and maybe more positive and constructive? In words that are going around a lot these days in other areas of discourse, could you disagree without being so disagreeable?

    I mean, you’re entitled to your approach and your attitudes. These are just suggestions. Take them under consideration or not.

  45. Peter Woit says:


    People can write about speculative ideas they’re excited about, I just think that they need to do so in an intellectually honest way. The popular press multiverse stuff I see, string-related or not, contains massive amounts of mis-leading, over-hyped material, aimed at getting people “excited” about some ideas which have very little evidence supporting them. As for the argument that this is the way to get kid interested in physics, I don’t accept it. It’s a good way to confuse people (of any age) about the difference between science and science fiction, and a good way to drive away people with some ability to think critically, who quickly suspect that someone is selling them a bill of goods.

    It’s true that I’ve ended up being critical of cosmologists pushing multiverse ideas independently of string theory. One reason for this is a general distaste for hype in science, although there’s plenty of that elsewhere and maybe it should just be ignored. But in this case, cosmologists hyping the multiverse have just about always joined in with string theory pseudo-science, co-promoting it and using it to justify what they are doing. I’m seriously concerned that this is doing a lot of damage to this subject, damage that may be permanent. We’re well on our way to having the dominant paradigm of the subject be “string theory is a fantastically wonderful unified theory justified by the deepest mathematics, and it shows we live in a multiverse such that we can never calculate anything beyond what we already understand, and that’s just the way the world works.”
    There are a lot of people pushing this, and a lot of cosmologists signing on for the ride. I think it’s worth my time to keep pointing out that there’s no evidence for this, that it’s an ideology driven by refusal to admit that a certain speculative idea has failed.

  46. Matt says:


    I made clear that I consider a lot of your points valid. My point was your delivery, which does you no favors and drives away a lot of people who would probably be sympathetic to your point of view. A lot of people dismiss you out of turn because of your sometimes-overwhelming negativity and the tone of your comments–some of which I’ve seen you make in person–and that ends up hurting your own cause of making physics better.

    Negativity drives people to shore up their defenses and to be less willing to take up alternative viewpoints.

    Sometimes when it seems like trends are leading to doomsday, then there’s an inclination to believe that all’s fair and there’s no reason to hold back. What was it Goldwater said, “extremism in the defense of liberty is no vice”? Well, I would have to disagree. I think it can be a vice.

  47. Matt says:

    Woops, looking at those last few sentences again—please don’t think I meant to call you an extremist! I was just taking an extreme example. My sincere apologies for seeming to construe otherwise.

  48. Peter Woit says:


    I do appreciate the advice and will keep it in mind. Sometimes the thought occurs to me “Oh, no, I’m starting to sound like Lubos!”.

    The string theory multiverse issue though I find increasingly both depressing and highly frustrating. Before about 5 years or so ago I thought it was possible to have a worthwhile discussion with string theory optimists, since the issues I disagreed with them about could be clarified by discussing the technical problems of the subject. In recent years that’s increasingly no longer the case, as the string theory multiverse ideology has been constructed to make the subject immune to the usual sort of scientific challenges. My going on about this has likely gone past the point of being useful.

  49. Matt says:


    I adore how you put that.

    Sure, I may disagree with how depressed and angry you sometimes sound (even though you’re really no cynic), but you aren’t even in the same bubble universe (sorry, couldn’t help myself) as Lubos. He’s beyond Serge Lang territory these days. As long as you retain your evident capacity for self-criticism and introspection, you don’t need to worry. I just wish you’d lighten up once in a while!

    But, honestly, what can I expect from a leftist feminazi communist university AGW idiot?


  50. Matt says:

    But, no, seriously, what I find depressing is walking around my department and seeing everybody on his blog. I think the last time I could bring myself to look at it, he was deriding the decision to hold a theoretical physics conference in Mexico, because, as he put it, Mexicans statistically have too low an average IQ to do physics.


Comments are closed.