Belief in multiverse requires exceptional vision

Tom Siegfried at Science News has a new piece about how Belief in multiverse requires exceptional vision that starts off by accusing critics of multiverse mania of basically being ignoramuses who won’t accept the reality of anything they can’t see with their own eyes, like those in the past who didn’t believe in atoms, or superstrings:

If you can’t see it, it doesn’t exist. That’s an old philosophy, one that many scientists swallowed whole. But as Ziva David of NCIS would say, it’s total salami. After all, you can’t see bacteria and viruses, but they can still kill you.

Yet some scientists still invoke that philosophy to deny the scientific status of all sorts of interesting things. Like the theoretical supertiny loops of energy known as superstrings. Or the superhuge collection of parallel universes known as the multiverse.

It’s the same attitude that led some 19th century scientists and philosophers to deny the existence of atoms.

The problem with the multiverse of course is not that you can’t directly observe it, but that there’s no significant evidence of any kind for it: it’s functioning not as a testable scientific explanation, but as an excuse for the failure of ideas about unification via superstring theory. Siegfried makes this very clear, with his argument specifically aimed at those who deny the existence of “supertiny loops of energy known as superstrings”, putting such a denial in the same category as denying the existence of atoms. Those who deny the existence of superstrings don’t do so because they can’t see them, but because there’s no scientific evidence for them and no testable predictions that would provide any.

Siegfried has been part of the string theory hype industry for a long time now, and was very unhappy with my book, which he attacked in the New York Times (see here) as misguided and flat-out wrong for saying string theory made no predictions. According to him, back in 2006:

…string theory does make predictions — the existence of new supersymmetry particles, for instance, and extra dimensions of space beyond the familiar three of ordinary experience. These predictions are testable: evidence for both could be produced at the Large Hadron Collider, which is scheduled to begin operating next year near Geneva.

We now know how that turned out, but instead of LHC results causing Siegfried to become more skeptical, he’s doubling down, with superstring theory now accepted science and the multiverse its intellectual foundation.

The excuse for Siegfried’s piece is the Wilczek article about multiverses that I discussed here, where I emphasized only one part of what Wilczek had to say, the part with warnings. Siegfried ignores that part and based on Wilczek’s enthusiasm for some multiverse research takes him as a fellow multiverse maniac and his article as a club to beat those without the exceptional vision necessary to believe in superstrings and the multiverse. Besides David Gross, I’m not seeing a lot of prominent theorists standing up to this kind of nonsense, leaving those invested in failed superstring ideology with the road clear to turn fundamental physics into pseudo-science, helped along by writers like Siegfried.

Update: A commenter points to this from Wilczek, noting his lesser multiverse enthusiam than Siegfried’s.

Update: Ashutosh Jogalekar at The Curious Wavefunction has a similar reaction to the Siegfried piece.

Update: There’s an FQXI podcast up now (see here), with Wilczek discussing the multiverse.

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

73 Responses to Belief in multiverse requires exceptional vision

  1. H says:

    In a tweet, Frank Wilczek says he is less enthusiastic than Tom Siegfried.

  2. paddy says:

    “Exceptional vision” ??? Siegfried is certainly not at fault for my immediate reaction to his choice of words: that vision required to see N-rays?

  3. paddy says:

    Pardon my double post….I tried to ignore Siegfried’s choice of the word “belief” in my previous post…but now my gorge is rising..time to have another beer.

  4. Matt says:

    Cosmic inflation predicts that the actual universe is much larger than the observable part we can see, and by a truly staggering amount — the observable region is like an atom compared to the rest. The rest is called the multiverse. Other regions can fall out of inflation just like our region did.

    Cosmic inflation is now heavily supported by high-precision observational data. It also neatly resolves the horizon problem, the flatness problem, the relics problem, the structure problem, and what gave the big bang its initial oomph to begin with. And we have direct evidence today that the fundamental mechanism actually works, because of the existence of (much weaker) dark energy.

    So how is there “no significant evidence” for the multiverse? Is the evidence for inflation wrong? Is there some magical principle that prevents any other region of the multiverse from falling out of inflation like our region did, or that nonlocally forces inflation to stop everywhere simultaneously all at once? The name “multiverse” might be the trouble — would it be less controversial if we just called it “the rest-iverse”?

    Note that I’m not talking about string theory or the landscape here — those concepts are logically independent from cosmic inflation, and the enthusiasm of string theorists is not evidence against cosmic inflation. When you say there is no evidence for the multiverse, are you just referring to its fusion with the idea of the string landscape?

    I’m just trying to make sure I understand a little more precisely what you’re saying here.

  5. Peter Woit says:

    I think I’ve discussed this “inflation implies a multiverse” argument many times here. The issue is not whether there’s more than the observable universe out there. That may very well be, and maybe someday we’ll even understand inflation well enough to have a good model of what that might be. The point though is that there is zero evidence that whatever else there is has different physical laws than ours, and that is what is needed to make the whole anthropic business work. In simple models of inflation, whatever else is out there will have the same laws as ours, so yes, you can get a multiverse, but a pretty boring one. You can use string theory or something else to come up with much more complicated models that give you pretty much any physics you want in different parts of the multiverse, but there’s no evidence for these, they are untestable and explain nothing. These models function just as excuses for the failure of string theory to explain anything and that’s the pseudo-science problem here.

  6. Daniel TTY says:

    How about other kinds of many-universes theories, such as universes being created in quantum fluctuations, or via black holes?
    Do you think that our universe is the only one?

  7. Peter Woit says:

    I have no idea whether there are other universes. If you have a model that produces them via quantum fluctuations or black holes, fine. But if it’s untestable and does nothing to explain the things that the SM and GR leave as mysteries, I’m just not interested. If it’s untestable and you want to claim it shows why those mysteries can’t be explained, and it’s hideously complicated, I think you’re a pseudo-scientist (and, unfortunately, in illustrious company…)

  8. Matt says:

    Hi Peter,

    That’s exactly what I wanted to know—thanks.

    So I think it’s important make this distinction clear. You are okay with the notion of the “boring” version of the cosmic multiverse (which I don’t happen to think is very boring! I mean, come on, right!?) necessarily predicted by cosmic inflation, namely, that the actual cosmos is far vaster than our observable region (that’s the whole basis for how inflation solves the various problems like flatness, the horizon problem, etc.), and in those other regions inflation can end and produce other observable regions.

    Where you draw the line is at the pure speculation that we can attach a string landscape to this cosmic multiverse and use it to start making anthropic claims about what the low-energy physics of our own observe universe looks like—the particle menu, the coupling constants of the Standard Model, etc. So, to be clear, you are okay with the “boring” multiverse, but not okay with the string landscape and the anthropic principle.

    Is that a correct appraisal? That may seem like a subtle distinction, but I’m sure it’s pretty important to a lot of people! Maybe you could say “anthropic mania” or “string landscape mania” rather than “multiverse mania”? Because even apart from string theory or anthropic reasoning, the “boring” multiverse is a very conceptually amazing consequence of inflation!

  9. Peter Woit says:


    Sure, I’m ok with inflationary models that produce lots of unobservable universe with the same laws of physics, but, honestly I do think they’re kind of boring. The people out there selling the multiverse as a radical new advance in physics aren’t selling this kind, they’re selling the one with different physics everywhere, as an explanation for the laws of physics, and that’s what I have trouble with.

  10. Anonyrat says:

    I would find it very hard to believe that our cosmos coincides exactly with our observable region- our past light cone. This with or without inflation.

  11. Pingback: Multiverses: pro and anti | Scientific Gems

  12. Abbie Hoffman-Tegmark says:

    There are a lot of different multiverses going around these days.

  13. Casey Leedom says:

    So sad that Tom Siegfried is headed this way. Luckily his bias doesn’t seem to have overwhelmed Science News’ content or I’d have had to cancel my subscription years ago. Interestingly, I was actually turned on to your book via a short review in Science News years ago when it came out (along with Lee Smolin’s book in the same issue if I remember right).

  14. Bee says:

    The multiverse is an inevitable consequence of attempting to describe nature by relying exclusively on mathematical consistency. Whatever you do, you’ll always have to pick some axioms of your theory and these axioms will either require additional justification or, if you don’t have that justification (call it selection), they’ll imply a vast number of “universes” that are nothing like our own. That’s got nothing to do with string theory in particular, it’s a consequence of mathematical “space” being larger than our observed “space”.

    There are only two ways to deal with that. Either you believe in the multiverse as being real and say we only observe part of it. Or you select the “real” part of the mathematically possible by binding it to observation and disregard the rest.

    Which way you prefer depends on whether you believe that mathematics is a language that we use to describe nature (in which case the multiverse is an artifact of the limits of that language) or whether you believe that nature fundamentally is mathematical (in which case you’re now missing a criterion that tells you which part of the mathematics is real).

    The reason this problem occurs now is that people are attempting to construct theories without observational constraints. (One can plausibly argue that requirements like renormalizability, gauge invariance, and Lorentzian signature are ‘observational’ constraints, though they’re rarely referred to as such. But these are the requirements that dramatically reduce the space of possible theories. Alas, they’re evidently not sufficient.)

    I wrote a blogpost on that here.

    Personally, I’m in the camp of people who think that mathematics is a language and shouldn’t be confused with reality.

  15. markusm says:

    Maybe there is already evidence for the (“boring” ?) MV:

  16. imho says:

    There’s evidence for a boring multiverse??? I had no idea. Can some one give me a pointer to a good survey article appropriate for a Cond Matt theorist with a passing interest in these things.

  17. Jim says:

    I thought I might take this opportunity to test my own understanding. Apologies Peter if this is somewhat off-topic.

    Although inflation is part of the ‘standard’ lambda-CDM model of big bang cosmology, my understanding is that it is still a bit controversial. In it’s most commonly accepted form, it is based on the idea of supercooling at the end of the Grand Unification epoch, triggered by an ‘inflaton field’ responsible for symmetry-breaking leading to the separation of the strong nuclear force from the electro-weak force. It’s controversial because we don’t have a GUT that everyone can buy into, and so cosmologists have been completely free to choose a shape for the inflaton potential that is consistent with the universe we see today.

    So far as I know, there’s nothing in the theory to indicate whether inflation applies to the whole of the post big-bang universe or whether it was just one small bubble of spacetime within this that inflated. If I’m right, then debates about whether or not there’s a larger ‘boring’ universe outside of our bit of inflated spacetime are really rather moot – in any case we’ll never know one way or the other.

    Sure, inflation resolves the flatness, horizon, and other problems, but there may be lots of other ways of doing this that do not involve inflation and, if I’m being a tad skeptical, the lambda-CDM model is, for the most part, a best-fit exercise with a good number of variable parameters. It is certainly not wrong (as the fit is quite spectacular), but as we have no theory that can provide a priori predictions for these parameters (just as we can’t predict the parameters of the standard model of particle physics) then I’d argue that all bets are off.

    Inflation is a great device and it seems to work well, but I think it best that we don’t over-interpret it.

  18. Peter Woit says:


    I think very few physicist’s believe in the sort of evidence for a multiverse linked to above. For more about this, see here

  19. Low Math, Meekly Interacting says:

    “Multiverse” seems to me like a fairly meaningless term (i.e. it gets used for so many different things it’s hard to tell if it means anything). However, to the extent it is not completely bereft as noun specifying something, I don’t think it makes much sense to conflate it with “everything beyond our cosmic horizon”. It seems to get used most often as a term for the ensemble of (mostly) isolated vacua with different physical constants, e.g. the string theory landscape.

  20. Chris Oakley says:

    “Belief in multiverse requires exceptional vision” is a very 1960s thing to say, and sounds a lot like “Belief in multiverse requires exceptional amounts of LSD”. It is just more proof – if proof was needed – that the subject has been taken over by hippies.

  21. Harald Kirsch says:

    Can’t help to quote Gemany’s ex chancellor Helmut Schmidt: “People who have visions should go see a doctor.” (

  22. ED says:

    The most plausible model of the physical world could be a very bad framework for scientific investigation, and in the multiverse wars, it would be nice to see this fact kept in focus. To be worth investigating, a model must be plausible , yet a plausible model may not be worth much attention.

    If lost keys might be under a lamp post, might be in the dark beyond, and yet seem most likely to be at the bottom of a river. A wise investigator would neither let inaccessibility discredit the riverbottom-key hypothesis nor jump in the river and drown.

    In other words, “X is unobservable, probably true, and worth considering” is entirely compatible with “Assuming that X is true would destroy physics”, and both are compatible with doing experimentally oriented physics and (mostly) ignoring X.

    That said, I find Wilczek’s arguments (and lament) persuasive. I recommend the article as an exploration of the limited but non-zero utility of anthropic reasoning.

  23. Peter Woit says:


    One problem is that “plausibility” is in the eye of the beholder. Those with decades of their lives and their reputations invested in string theory may find the anthropic string landscape a plausible idea, others not so much. They also may feel regrets about putting the torch to physics as a science, that this makes them feel bad, although not as bad as they would feel if they had to admit they had been wrong.

  24. vmarko says:

    Matt said:

    “Cosmic inflation predicts that the actual universe is much larger than the observable part we can see, and by a truly staggering amount — the observable region is like an atom compared to the rest. The rest is called the multiverse.”

    If one is talking about the “boring” version of the multiverse, I think that the terminology is misguided. If there is nothing conceptually different between various inflated bubbles, then it’s just more of the same thing, sitting outside our observable space. This does not, and should not, deserve to be called “multiverse”. Rather, it is just an extension of what lies beyond our cosmic horizon. It should still be called “universe”.

    The term “multiverse” makes sense only if those other patches of space are somehow very very different than our own (like having different values for the coupling constants etc.). And *that* is the problem — there is no convincing way to establish that such things actually do exist. This is the failure of string theory that Peter is criticizing all this time.

    In this sense Peter is right to push all “multiverse” stuff into the same basket — the inflation-induced boring version of the “multiverse” should not have the “multi” part inside its name in the first place. It’s just misguided use of terminology.

    Best, 🙂

  25. Shantanu says:

    Despite much hype, I still don’t think we can conclusively say for sure inflation happened. The only thing we can say for sure from Planck is that n is different from 1.
    We still don’t know the energy scale of inflation and GUT based inflation is ruled out from Planck data. Also there are fundamental conceptual problems with single scalar field inflationary models (see

    Another thing is that there are so many models of inflation , see for example this 330 paper
    IF you have so many “models” of inflation, by random chance one of them is probably right. But that defeats the original purpose of inflation.
    Also as Peter and other mentioned there is no-observation al evidence for multi-verse and given so many models of inflation, its not obvoious which of them need a multiverse and which ones don’t.

  26. Matt says:

    That the actual “rest-iverse” is far, far vaster than all that we can directly see all the way out to the earlist galaxies is a fairly robust, model-independent consequence of inflation, regardless of the shape of the potential or the number of scalar fields. In essentially any model, you need enough e-foldings to solve the various problems with the old big bang model, and that leads to a humungous cosmos far bigger than the popular understanding of the term “universe.”

    It’s really a staggering and under-appreciated triumph of science that precision cosmology has advanced to the point at which we can confirm the general validity of cosmic inflation, even if we cannot yet pin down the specific model. And there are lots of people working on that right now, and checking their predicitons against observations carefully. That’s most definitely bona fide science, and it would be a real shame if it got lumped into the string wars in the public understanding. And one would have to be really jaded not to realize how awesome this all is.

    Some have called it the “megaverse.” But the idea that the cosmos is so unimaginably big that there are lots of other pockets just like ours, separated by inflating regions, and with each pocket containing its own galaxies and whatnot—even without the string landscape and any unfounded claims of different low-energy physics in those pockets—certainly fits my picture of what a multiverse is. It’s a less exotic notion of a multiverse than given by the string landscape, but as a vast expanse filled with little “island universes” separated by nontraversible inflating regions, it fits the bill for a lot of people when they use the term “multiverse.”

  27. Chris W. says:

    Still, Matt, you must admit that there is a certain extravagance in this consequence of inflationary models, insofar as the consequence can never be tested, directly or indirectly. String theory, and perhaps alternative quantum theories of gravity, vastly complicate the picture by introducing enormous scope for variation of low-energy physics in those inflationary pockets, setting the stage for an anthropic trivialization of our understanding of the observed universe.

    One might consider how analogous—or not—this situation is to that of the vacuum in quantum field theory. The features of that vacuum play a key role in calculating observable quantities—e.g., the well-known corrections of quantum electrodynamics.

  28. LogicFan says:

    I’m a philosopher, not a scientist, yet I cannot believe that Siegfried doesn’t understand the essential difference here; that is, the distinction between being practically unable to observe some phenomenon and that phenomenon being perforce inaccessible to experiment.

    I have heard some disparaging things (including, I believe, by the author of this blog) said about philosophers of science (I am not one), but every time I read something like this I am reminded how often even the educated make simple errors of logic, and thus how important our discipline really is.

  29. Low Math, Meekly Interacting says:

    The multiverse does seem to induce a curious form of cognitive dissonance. On the one hand, someone will propose that, say, bubble nucleation will leave an imprint on the CMB. “See? It’s testable!” So we see nothing, feel like we’ve been had yet again, and complain that this whole multiverse business is proving to be utterly unobservable, unfalsifiable, and hence maybe outside of the scope of physics. “Stop being such a positivist!”

  30. Peter Woit says:


    That’s not really the relevant distinction here. What’s more at issue is that you can test a theoretical framework not just by directly observing the fundamental entities of that framework, but instead by testing other implications of the framework. The first evidence for atoms was not from seeing atoms, but from seeing their effects on other larger objects (the phenomenon of Brownian motion).

    You can imagine a theoretical framework that predicts other universes (which you can’t observe), and if this same framework predicts indirect effects of such other universes that you can observe, or it makes new distinctive predictions (say the mass of the electron) that can be verified, you have significant evidence for the parts of the framework you don’t have direct experimental access to. This is what the string theory landscape people would like to do. Their problem is that there is no evidence at all that their framework can do this: they can’t calculate anything at all in it.

    The problem with Siegfried’s argument is a very conventional one: it’s a straw man argument. He’s claiming that the argument of multiverse critics is that the theory can’t be directly tested, and that these critics don’t understand about indirect tests. But that’s not the argument (at least not from this particular very vocal critic). The argument actually is that the string theory multiverse makes no predictions of any kind, either directly or indirectly testable. It’s not actually a scientific theory, but a pseudo-scientific construct designed to “explain” why you can’t ever predict anything testable, directly or indirectly.

  31. Paulibus says:

    Bee says all that need be said about multiverse mania. Sadly, unless someone trips over a decaying proton while walking in the woods, we’re watching the transformation of theoretical physics, sans its Baconian flavour, transform itself into a beached whale. Shades of Mark Twain’s Great Nonesuch!

  32. lucretius says:

    I have to admit I am rather new to these “string wars”. During the last 25 years or so I have listened to a number of talks by Witten at various mathematics conferences (including at least one at which Peter was present, although I only found it looking at the proceedings last night) and I have never doubted that his inspiration came form “mainstream physics”. Well, now it seems, people are claiming it is not physics at all (although nobody seems to know what else it could be). I only realised that something weird was going on behind the scenes by reading Lubos’s blog, which I visited for reasons which had nothing to do with “strings”.

    Anyway, I have decided to sit on the sidelines at least until I am confident I understand the main technical issues involved. This is still going to take time. I am somewhat surprised that not everyone thinks that is the right approach. I am also puzzled buy what exactly we are supposed to worry about. On the one hand I keep reading here alarming messages to the effect that string theorists are “putting the torch to physics as a science” and that “theoretical physics transforming itself into a beached whale” , etc. On the other I remember Peter writing that string theory (and LQG, to which all the same objections seem to apply but which rarely is criticised here) represent only a section of theoretical physics, with apparently decreasing influence. So how can these two things be true at the same time?

    On the issue of string theory’s lack of experimental predictions own naïve view at the moment is this. If we have two theories (GR and QFT, I think), which have together make a large number of confirmed experimental predictions but are mathematically inconsistent with each other then if we can construct a self-consistent theory (ST) which “in the limit” replicates GR and QM, than that constitutes scientific progress. The theory is “scientific” even if it does not make any confirmable predictions beyond those of GR and QFT since it inherits the predictions that they make and adds self-consistency. Hence the only thing that would worry me as far as ST is concerned are the two questions: 1) is it really fully self-consistent and 2) does it really give GR and QFT in the limits? So I would worry about the kind of criticism that Penrose discusses in “The Road to Reality” (finiteness or the question of stability of 10 dimensional space time) but not about the kind of things that are most often mentioned here.

    As for the “multiverse” I would be inclined to follow the method of Sherlock Holmes (as described in “The Sign of the Four”): “when you have eliminated the impossible, whatever remains, however improbable, must be the truth”. Of course to some extent this begs the question: what exactly do we mean by “impossible” in this context? Myself I am quite happy with the view that impossible means “mathematically impossible”, which means that if it takes hidden extra dimensions and multiverse to construct the unified theory and if all the mathematical problems can be sorted out – than, however improbable, that is maybe not exactly “the truth” but the best we have got.

  33. Peter Woit says:

    You’re leaving the topic of this posting, and trying to start a discussion on a range of topics about string theory and its associated hype which have been discussed ad nauseam here over the years, something I don’t want to encourage. If it’s not clear though, the on-topic problem being referred to (ok, with imprecise and inflammatory language like “putting the torch to physics”, I should try and avoid this as Lubosian) is a straightforward one. As an idea about unification, string theory has been remarkably unsuccessful, and has ended up now forced into the corner of a specific very complex and completely unpredictive and untestable proposal (the string theory “landscape”) that I claim is accurately characterized as pseudo-science designed to avoid admitting failure. Siegfried’s article is an attempt to justify this, and I’m pointing out the problems with it.

    The exact details of what the string landscape proposal is and why it is unpredictive are among the things that have been discussed ad nauseam here over the past 10 years since this proposal first surfaced. If you’re skeptical about my claims and want to devote time to this, I’d suggest trying to find a counter-example: a testable prediction, or a plausible proposal for how to get one. Many claims of such things have been discussed here on the blog. You also might want to note that even Lubos doesn’t believe in this (he thinks some new insight into string theory will make the landscape unnecessary).

    If you’re not interested in this, but just want to repeat hype about supposed wonderful properties of string theory and how it’s the only way to go to get unification, so it doesn’t matter if it’s pseudo-science, because we have no choice, then I really can’t help you, but do want to point out that you’re arguing for giving up on this kind of science, torching it if you will. Maybe better understanding fundamental physics is a lost cause, one could have argued that at anytime in the earlier history of the subject. I don’t think it was true then, or that it is true now.

  34. vmarko says:


    String theory is much more ambitious than just unifying GR with QFT — it aims to be a “theory of everything” — deducing from first principles both GR and all of the Standard Model. As far as incorporating GR into QFT, string theory has been reasonably successful, and that is all good and well (at least up to a point). But as far as reproducing the SM, all string theory attempts have been a disaster — instead of a single version of SM, it “predicts” some 10^500 different versions of it (this is a colloquial way to phrase the so-called “landscape problem”).

    So at this point string theorists were supposed to admit that ST simply fails to predict the SM, and that further work is required. Instead, they started arguing that actually all those 10^500 different “Standard Models” actually do exist in nature — the multiverse hypothesis — and that we are seeing this one particular SM in experiments (as opposed to any other) because our universe (our observable patch of the multiverse) hosts a version of the SM that supports life being created in it. This is called the “anthropic principle”.

    The issue with this argument is the following — not only that it does not predict anything, but moreover it states that one should give up trying to predict anything. Just like religion (naive interpretations of it) can answer any question with “God did it”, string theory can invoke the anthropic principle to answer any question with “we happen to live in such a patch of the multiverse”. It has 10^500 “vacuua” (or free parameters, simply put), so you can fit basically any experimental data to this set of parameters. No predictions can be made in such a setup, only interpolations and data-fitting.

    So, failing to predict the SM free parameters, string theorists should admit that ST is far too general and that further axioms/principles/whatever is needed to reduce the parameter space. But instead of doing that, they simply declare ST to be “correct”, and the SM problem unsolvable, by definition. And they get a lot of positive media attention, saying “string theory, via the multiverse and the anthropic principle, solved one of the biggest puzzles in science”, when in fact they didn’t solve anything — they *gave up* on the problem.

    It even goes as far as teaching graduate students not to look for any explanation for the SM free parameters — string theory predicts that they cannot be explained and that no explanation is necessary, so why bother?

    From a failed attempt to unify all forces of nature, string theory grew to become a fundamentalist religion of modern physics.

    All this has nothing whatsoever to do with combining GR with QFT — that can be done with some level of success in ST, in LQG, and in various other approaches.

    Best, 🙂

  35. Low Math, Meekly Interacting says:

    What strikes me as an incredible irony is that string theory was supposed to solve the problem of having to resort to anthropic reasoning. The multiverse concept (as I prefer to define it) was already quite mature in form eternal inflation. Some people were eyeing the possibility of a non-zero cosmological constant with abject terror. But once we found the principle that selected manifolds that gave us 3+1 extended dimensions, GR, and the Standard Model as a low-energy limit, we could dispense with pseudoscientific anthropic nonsense! Now the ST landscape is the foundation on which eternal inflation can stand. Nothing appears to remain of the notion of selecting anything, except maybe wistful nostalgia among those who don’t regard it now with disdain. What a terrible vision this has turned out to be.

  36. Amos Dettonville says:

    Lucretius, isn’t the landscape/multiverse idea inconsistent with what you described? It says there is no unique “limit” of string theory corresponding to the standard model and general relativity. Instead, there are a bazillion low energy limits, and we don’t actually know (and may never know) if any of them closely resemble the standard model. The hope is that some of them do, but in any case nearly all of them don’t. In this situation, it seems like a stretch to say string theory “inherits the predictions” of the standard model.

    If what you described came about, i.e., if there was a unified conceptual framework that uniquely reduced to the standard model and general relativity in some suitable “limits” (as general relativity reduces to Newtonian theory in the weak slow limit), I would agree that it would be very interesting – although if it really makes no new confirmable predictions it might be more accurate to call it an *interpretation* rather than a new theory. If that interpretation sheds light on how the conceptual inconsistencies of the two pre-existing theories can be reconciled, that would be great.

    I suppose string theorists believe there are enough qualitative similarities between the generic features of string theory solutions and the facts of the standard model and general relativity that we are warranted in claiming that string theory already constitutes a unified conceptual framework subsuming gravity and the standard model (not withstanding the fact that we don’t actually know – and may never know – if any solution of string theory actually matches the standard model quantitatively). But apparently many have concluded – some reluctantly and some enthusiastically – that a necessary ingredient of this conceptual framework is the landscape/multiverse, i.e., severe under-detemination of the specific attributes. With the landscape/multiverse, we might say the only prediction of string theory is that prediction is impossible – in the sense that there can never be any deeper explanation for the parameters of the standard model, because many different sets of parameters are equally compatible with the principles of string theory.

    So in this sense I suppose we could say string theory would be falsified by the discovery of some new constraints or inter-relationships between the parameters of the standard model. On the other hand, maybe any new constraints would simply be taken on board, and the claim updated to say there are no *more* constraints to be discovered. All we’re doing is refining the attributes of one particular solution (the one that corresponds to our universe), while retaining the belief that it is compatible with the principles of string theory (which still seem to be in flux).

  37. DN says:

    “It’s not actually a scientific theory, but a pseudo-scientific construct designed to ‘explain’ why you can’t ever predict anything testable, directly or indirectly.”

    Peter, enjoy reading your blog, opinions and the various developments you write about, even though I’m in a different field.

    Maybe you’re trying to be funny and serious at the same time up there. I get it. Anyway, help explain “explain” a bit more. Thanks.

  38. Peter Woit says:

    To really see what’s going on here, you need to get into the very complicated subject of how to construct “string vacua”. At this point I think even most string theorists have given up on working on this, with the subject left in a state where you have so many possible choices that you can get whatever you want, and evade having any testable prediction that might conflict with experiment. The “multiverse” provides the argument that any “string vacuum” might show up. Normally scientific theories explain in one way or another something we observe, this theory explains why we can’t ever test it (too many possible string vacua, we could be in any of them).

  39. lucretius says:

    Indeed I forgot about the ST probably because I am now trying to understand better QFT (for purely mathematical reasons) and the rest of the ST is just physics to me 😉
    I agree that the ideas of landscape/multiverse seem contrived and probably unnecessary (although my knowledge of this subject is very limited).

    I don’t see why anyone should be surprised today by the failure of physicists to arrive at a unique mathematical “theory of everything”, particularly in view of the experience of other disciplines.

    In mathematics, as everybody knows, Hilbert’s naïve hope that a consistent system of axiom for the whole of mathematics could be found (and proved to be such) was destroyed by Goedel’s two theorems, after Russel and Whitehead spent years trying to realise it (in volume 2 of Principia Mathematica they managed prove that 1+1 = 2).

    In option pricing (in mathematical finance) there was a period when it appeared that Nirvana had been attained: in the Black-Scholes model of stock market prices the assumption of lack of arbitrage implied that there was a unique fair price of every option, and for the simplest ones explicit formulas could be given. In spite of Nobel prizes in economics, this turned out to be an optimistic illusion: we now know that real world markets are incomplete and there are infinitely many “fair” prices of options compatible with the given price of the stock and the non-arbitrage principle. Out of these infinitely many “consistent” prices (or price processes) the market somehow chooses one – but there is no known principle that determines it.

    In microeconomics about 100 years of research into general equilibrium run into a brick wall in the form of the Sonnenschein–Mantel–Debreu theorem, which says that equilibrium is neither unique not stable.

    In none of these cases the seemingly devastatingly negative result lead to people abandoning the entire field (or even basic directions) of research. In fact, in each case pragmatic reasons were found for continuing.

    I don’t believe that string theory actually needs the landscape and multiverse to survive. The indeterminacy is disappointing but there are ways to deal with it just as there are in all these cases I have mentioned.

    Obviously the landscape and multiverse belong to the realm of speculation, even if dressed up in elegant mathematical form. In fact the title of Sigfried’s article suggests this by his use the word “belief”. Nothing in science can actually depend on a “belief”. “Belief” can be very important in motivating someone to pursue a line or research that seems unpromising to others but to accuse critics of a scientific theory or idea not having enough “belief” makes one sound like an apologist for a religion not a science.
    Hoever, I don’t see any reason why Sigfried’s obvious silliness should be taken as discrediting the rest of string theory and that is really all I wanted to say.

  40. paddy says:

    Such a comparison of theoretical physics to so-called “economic theory” might (or should) shame those who have monetized the “multiverse”.

  41. lucretius says:

    Peter, I am curious what kind of “discussion” you consider to be “on topic” in a thread like this one. I understand that you may not want to return to things that ” have been discussed ad nauseam here over the past 10 years” ( although I don’t think you are seriously expecting new readers of your blog read through 10 years of past discussions before asking a question) but then what is “on topic” except for cries of “whatever next” and “what is the world coming to” and this sort of thing? . I asked one question, which at least had the virtue of being different from what other people wrote. I repeat it: how can string theorists be increasingly irrelevant to physics research and even having difficulties getting hired (both claims I have read on your blog while trying to catch up on the past discussions) and at the same time be “putting the torch to physics”? O.K. you have now disassociated yourself from this “Lubosian” language but you still have not answered my question. If string theory is indeed as unsuccessful as you claim, and in decline then what is the purpose of posting threads in which everybody is clearly expected to express the same “anti-string” view?

  42. vmarko says:


    “I don’t see why anyone should be surprised today by the failure of physicists to arrive at a unique mathematical “theory of everything”, particularly in view of the experience of other disciplines.”

    It is not a surprise, of course. But do remember that (in light of Goedel’s incompletelness theorems), a “theory of everything” in physics actually means “theory of everything so far”, since every theory in physics can aim to describe only the so-far-known experimental facts. So we are well aware that any proposal for a “theory of everything” is necessarily incomplete from a conceptual point of view, but at the same time it should certainly be able to describe all previous established models in suitable limits.

    Also, note that the Goedel’s incompleteness theorems and the Sonnenschein–Mantel–Debreu theorem are actually theorems (such negative results in physics go by the name “no-go theorems”, and there are many examples of such results). In contrast, the multiverse idea in ST and the anthropic principle are axioms, that deny one the possibility to give any furher explanation of the SM free parameters.

    A lot of people simply refuse to accept such statements as valid axioms, precisely because of the limitations they impose. Even some string theorists are trying to steer clear of them. But there is this group of prominent string theorists who embrace them and advocate them — and they happen to be influential, vocal and widely present in the general media.

    “Hoever, I don’t see any reason why Sigfried’s obvious silliness should be taken as discrediting the rest of string theory and that is really all I wanted to say.”

    Sigfried is by no means the only one nor the first one to display such “silliness”. This “multiverse mania” (as Peter likes to call it) has been going on for quite some time now, and there is quite some amount of propaganda in the media about it. Just take a look at some past posts in Peter’s blog, you’ll find that this Sigfried’s “silliness” is just the latest in a long series.

    If other, more sensible string theorists would be as vocal against multiverse as its advocates are, the whole idea would be just one more controversial philosophical idea being debated, like so many others in history of science. But the problem is that those more sensible string theorists are keeping to themselves, while vocal advocates of multiverse are creating the illusion that there is general consensus about the topic.
    This situation is what is discrediting string theory in particular, and threatens to discredit the whole theoretical high-energy physics in general, in the eyes of general non-hep public.

    Best, 🙂

  43. Peter Woit says:

    The topic of the posting is the Siegfried piece, which, yes, I claim is arguing for putting a torch to physics by promoting pseudo-science. Some string theorists also argue for this, some don’t. The ones who do I personally think are motivated by refusal to admit failure. As for the rest of string theory’s problems, that’s another topic.

    As for insisting on people agreeing with me, I delete large numbers of comments posted here that criticize string theory but have nothing of interest to say. I very rarely delete comments from anyone who wants to disagree with me and argue for string theory. If your comments were arguing against string theory I’d probably have started deleting them on the grounds of off-topic/uninformed. Instead, here you are, admitting that you don’t actually understand the issues, but dominating the discussion. Because I’m allowing this, sensible informed people who see this won’t write in, just those who like to argue about the same old thing. You are doing everything you can to turn this comment thread into yet another tedious, uninformed and worthless set of string-war arguments, why?

  44. vmarko says:

    Also, while Peter of course has the last word in his blog about what is on/off-topic, I think that this discussion is certainly not off-topic. It certainly is a beating-a-dead-horse-discussion, yes, but off-topic, I’d say no.


  45. DrDave says:

    “If you can’t see it, it doesn’t exist. That’s an old philosophy, one that many scientists swallowed whole….After all, you can’t see bacteria and viruses, but they can still kill you.
    Yet some scientists still invoke that philosophy to deny the scientific status of all sorts of interesting things.”

    Actually, you can see bacteria and viruses with a scientific instrument called a microscope, and right there is the logic break: you can’t test for a multiverse, or the tests don’t show anything.
    This is followed by a straw man (straw person, sorry) argument of “some scientists” invoking the philosophy of the unseen. This is not what scientists do; scientists use the scientific method. The whole introduction to the article is to recast any scientists who disagree with SUSY and multiverses as the work of provincial disbelievers, as opposed to the rigorous methods and brilliant experiments developed from years of training.

  46. Chris W. says:

    More succinctly, bacteria and viruses were killing us before we even had any clue they existed, and we now know enough about them to say a lot of testable things about precisely how and when they can kill us, as well as how to produce images of them with appropriately designed equipment.

  47. lucretius says:

    @Peter Woit:

    Since you asked me a question I feel obliged to answer it. It was certainly never my intention to take part in the “string wars”. As I wrote earlier, I intend to sit on the fence until I understand things well enough to make up my mind for myself. Yes, I freely admit I do not “understand” many of the issues involved, but my standards of what constitutes “understanding” seem to be higher than yours. One of the first things I learned from my supervisor when (a long time ago) I started working on my doctorate was never to claim to understand anything unless I can reproduce all the technical arguments myself, preferably in my own way, and in this sense I certainly do not understand string theory. I noticed, however, that when Matt asked you whether you understood the technical details of string theory in one of the recent threads on this blog you refused to answer and (as you often do) declared his question off topic. I have not read your book but judging by your publications available on the Interent I can’t conclude that you apply the same standards to “understanding” as I do.

    Your claim that because you are generous enough not to delete my ignorant posts “sensible informed people who see this won’t write in” is, in the context of this particular thread, laughable. In fact, except for the replies to my posts by Marko, Amos Dettonville and a couple of other posts, the rest is information-free rhetorics.

    So this brings me to your question:
    “You are doing everything you can to turn this comment thread into yet another tedious, uninformed and worthless set of string-war arguments, why?”

    Purely out of curiosity. I wanted to find out what this blog was: a physics discussion forum or a combat training ground for the “string wars”. Lubos’ blog, besides his rants (which, in spite of their abrasive language are often witty and entertaining except perhaps to their targets) contains large amounts of technical physics, skilfully elucidated. In spite of its obviously one sided view, it is by a long distance the best place on the web to learn about certain areas of modern physics. Nobody could say this about this blog. To stick to the military metaphor: Lubos is training highly skilled commandos for the string wars while you are training suicide bombers.

    With best regards


  48. Jesper says:

    @ Lucretius

    what a load of eloquent rubbish. The problem with string theory is not string theory itself but the total dominance in theoretical physics it has had in the past decades. Although it might now face less friendly winds this general problem is still very real. Peter’s blog is a much needed vent for dissent. There is nothing suicide bombing about it, rather a forum for discussing what many feel are troubling trends in theoretical physics. I don’t see anyone bombing anyone here?

    And besides – in my views many of the unattractive sociological aspects of string theory are also found in other research directions in theoretical physics. Only, they are nowhere near as dominating as string theory.

    With best regards,


  49. Marty says:


    I have very mixed feelings about commenting on your last post, mostly because I feel like I’m contributing to the drift of the discussion from the primary topic. However, since you’ve said you haven’t followed this blog very long, hearing another perspective may be useful.

    I watched the “string wars” play out on several blogs (e.g., this one, Asymptotia and Cosmic Variance) some years ago, and needless to say the discussions turned ugly at times. Unsurprisingly, Peter was usually involved in them in some form or another, primarily because he was unusual among string theory critics in that he had actually spent a fair amount of time learning about some key aspects of string theory. (Lee Smolin was another knowledgeable critic who occasionally took part; almost always he managed to keep his cool in the face of personal attacks.) Although Peter’s level of knowledge was/is not at the level of a practicing string theorist, he had/has some technical familiarity with the topics he discussed. I think that earned him at least some respect among non-string-theorists — he didn’t post on subjects where he had no knowledge, and didn’t pretend to know something he didn’t know in responding to others. Moreover, on numerous occasions over the years he has had this blog, when it was brought to his attention that he was wrong he would generally admit it, and when he was justifiably criticized for his demeanor in responding to personal attacks he would often agree that he could have handled it more gently.

    Understandably, if you didn’t follow those “string wars” discussions, some of which contained hundreds of comments, you might not appreciate how tedious it became when a majority of comments were posted by people who were technically uninformed but who didn’t like (or did like, as the case might be) string theory for one reason or another, very often doing little more than parroting some point by Peter, Jacques Distler, Lubos or some other well-known critic or partisan — they said nothing new, but still felt that they should say something anyway. Many of the comments were personal and insulting, and for no good reason other than what appeared to be a very personal identification some people had with their own viewpoints. Practicing string theorists could also be insulting, which did nothing to polish the image of said theorists or their favorite theoretical framework.

    Returning to your questions and posts, if you keep that history in mind, it is probably obvious that Peter has no interest in repeating that experience with a “fresh crop” of blog readers. It isn’t anything personal — I really doubt he knows anything about you, so it can’t be personal — but in your last post your tone seems pretty defensive as though you took his admonition as a personal attack on you or your questions. In fact, if you compare the change in tone in your last post with some of what transpired in “string wars” discussions, you might recognize a certain commonality: at some point, a string theory partisan would feel like their views weren’t given the respect they felt was deserved and then dig in his/her heels, then resort to personal attacks on Peter’s (for example) publication record, lack of “string theorist credentials”, or other lame attempts to discredit Peter-the-person through bogus rationales, implicitly suggesting that by extension one should disregard the actual content of what he was saying. I’m not saying you are actually heading down that path, but with your most recent change of tone it is a possibility.

    (Interestingly, you mentioned Lubos’ blog as a training ground for “highly skilled commandos” and Peter’s blog as a place for training “suicide bombers.” That is such a surprising contrast that I can’t help myself in commenting on it. There is no question that Lubos knows a lot of physics; he is definitely gifted that way, certainly in my opinion and undoubtedly in the opinion of many others, say at PhysicsStackExchange. Having said that, on his blog and comments on blogs of others he freely intermixes his own biases and opinions with actual physics knowledge (or at least he did before I stopped reading him much), but often won’t tell you that is what he’s doing. And my impression/observation, again and again, was that his response when someone pointed out that he had made an error was to argue uncompromisingly, or call the person a moron (or worse) because they obviously didn’t know any physics or they would agree with him; but rarely would he admit he made a mistake. Another tactic he uses, or at least used in the past, was to delete posts that criticized the “story” he was trying to present. I stopped reading his stuff because I could no longer stand his frequent insults/intolerance of others who disagreed with him, the general low quality or sycophantic tone of comments readers made to many of his posts, and what often looked like an unwillingness to distinguish his opinions from fact. So in a sense I agree with half of what you said — his blog can act as a training ground for commandos — but only if you mean unquestioning jihadists for the string theory “cause” rather than “warriors” who do battle through rational and open minded discourse. Anyway, you apparently have a lot more patience for the ugly side of blogging than I do.)

  50. Nathalie says:

    Quite a few researchers are dissatisfied with their own (scientific) lives. They didn’t manage to dethrone Albert E., as was their original plan when they went to physics. “Everyone” had told them that they are terribly smart. Consequently, these people suffer from internal aggressions. Let them fight over Multiverse and such harmless stuff. Who cares? Nobody gets physically hurt and that is much better than them physically beating their wives and children.

Comments are closed.