Curt Jaimungal has a piece out, an interview with Lenny Susskind, with the title The Crisis in String Theory is Worse Than You Think…. Some of what Susskind has to say is the same as in his recent podcast with Lawrence Krauss (discussed here). These days, Susskind sometimes sounds like Peter Woit:
We live in the wrong kind of world to be described by string theory. No physicist has ever won a big prize for string theory. I can tell you with absolute certainty that it is not the real world that we live in. So we need to start over.
(interesting that Susskind seems to think the “Breakthrough Prize” is not a “big prize”, maybe because he’s one of the few well-known string theorists who hasn’t gotten one).
Susskind says he himself is working on trying to extend string theory to something different which will work in dS space, not just AdS, but he agrees with my claim that this is something the field has essentially given up on:
I actually don’t know anybody who is working, striving to try to expand the theory into either de Sitter space, which is not supersymmetric, or just more generally into an expanded version of the theory. Older people worked on it in the past. They worked on something called spontaneous breaking of supersymmetry. Don’t worry about what it means. It just means the theory wouldn’t be supersymmetric, and they failed. Now, that’s not a criticism of them. I worked on it, and I failed. That’s not a criticism of anybody, but it’s a fact that there is no precise theory which is not supersymmetric.
That is intolerable, in a sense. It can’t stay that way. We have to describe our world. That’s our purpose, and as I said, I don’t know anybody who’s actually working on that. If you were to send out a message to all the world’s theoretical physicists, anybody working on a generalization of string theory, you’d probably find some yeses, probably mostly among older people, and somehow we have to change this.
I’d argue the field has given up on it because, after decades of work, it’s clear this goes nowhere, and sooner or later Susskind will realize this.
At one point Susskind starts making an odd argument, hard to reconcile with the current state of the subject:
Look, there are still people who believe in the flat Earth, for God’s sakes. There’s people who believe all kinds of weird stuff. Don’t think about individuals. Think about the consensus of the largest fraction of physicists working on these things, and you’ll probably be right. The overall consensus of the field tends to be right. Peculiar individuals, no matter how famous they are, no matter how brilliant they are, if they’re off that consensus, and they’ve been off that consensus for a long time, they’re probably wrong. That doesn’t mean for sure that they’re wrong. Don’t look for the weirdos. Look for what the consensus of the majority of well-respected, highly accomplished physicists believe. And you’ll probably be right. There’s no guarantee of it. There are very few cases where the consensus has gone wrong for a long period of time, just where some offbeat idea of some particular individual suddenly changes everything. I’m not saying it doesn’t happen, but rarely. Penrose, what can I say? He believes all kinds of things that I wouldn’t subscribe to. But more than that, things that the consensus wouldn’t subscribe to.
Besides the weirdo Roger Penrose, he’s no fan of the ideas of another weirdo:
What is Peter Woit? If you look on the Internet, if you look on the archive, he has a small number of papers which are bad. They’re bad mathematics and bad physics. They’re just bad. I probably shouldn’t say that. I probably shouldn’t, but I’m going to say it anyway. He has nothing to offer at all. I assure you that if he had something that was compelling and interesting and that solved some problem, the physics community would notice him. I looked at his papers. I was unimpressed,
I guess his reaction is fair, partly since my own criticism of his work on the landscape is much the same (I have on the other hand had nice things to say about his textbooks, some of which appears as a blurb on the French edition of one of them).
In any case, if you start with the assumption that anything too far off the consensus is going to be unpromising, you don’t need to spend much time looking at my work to confidently evaluate it as having nothing to offer.
Jaimungal does get Susskind to realize that the “if it’s not close to the consensus, it’s probably bad” argument is a dubious one, especially at a time when the consensus research program has clearly failed:
But you’re perfectly right. We should certainly be on the lookout for ideas which are not the consensus. We should be watching for them and not immediately dismiss them because they’re not exactly the same as the ideas that we’ve been pursuing. For sure, we should be doing that. So I would agree with you about that. And maybe we haven’t been diligent enough with some of these ideas…
Most of the people I know, and that might even include myself to some extent, are derisive about a lot of these ideas. And they’re correct that there is a very strong skepticism about them, and maybe to some extent, unfounded. We all know that. There’s nothing hidden about that. The answer is I’ve looked at them, and I don’t find anything compelling about them. If you call that derision, yeah, I am a little bit derisive. However, I would say maybe there are elements in those theories which will come back, come back in some different form, which will connect better with the things which I think are right. And that’s a possibility, which I suspect most of my friends don’t entertain.
Who knows, some day Susskind may come around to the idea that one of the SU(2)s in the 4d Euclidean rotation group being an internal symmetry is not complete nonsense. Once I finish writing up a more detailed version of what I’ve been working on I’ll send him a copy. Maybe I’ll even finally figure out a way to use this to do something new with Kogut-Susskind/Kähler-Dirac versions of fermions, and he’ll be pleased that in 1977 he was on the right track…
Update: Somewhat related to the posting is this new rant from Sabine Hossenfelder. It’s motivated by this from “Professor Dave”, who has 3.4 million Youtube followers and is upset that she is hurting the credibility of scientists by criticizing what has happened in fundamental physics over the past 50 years (a topic he seems to know nothing about). While I disagree with her about some things, I strongly identify with:
Why the fuck is it my fault that cranks think I’m their best friend because I’m pointing out that there’s no progress in the foundations of physics? It’s a fact. We haven’t made progress in theory development for 50 years.
To connect explicitly to the topic of this posting, a big reason for the lack of progress is the way Susskind and other leaders of the field see things. In their minds it’s not possible that the consensus (i.e. groupthink) of GUTs/SUSY/strings of the past 50 years could be wrong. Anyone who argues otherwise is a “weirdo” who doesn’t understand the arguments behind the consensus and can’t possibly have any useful ideas about an alternative.
The link in your final sentence requires authorization to properly view, this seems to be better: https://journals.aps.org/prd/abstract/10.1103/PhysRevD.16.3031.
Anonymous,
Thanks, changed to that one.
After this declaration, why does nobody mention Edwar Witten and its role in keeping consensus in this dead-end route? Great work in mathematics, sure. Genius. Not so much on keeping string theory in life support instead of taking the much more difficult decision of unplugging it.
Prof Alexander Hamilton from Colorado has wandered into this area. In the abstract of an upcoming (1-Nov) talk (https://gaupdate.wordpress.com/2024/10/29/osmu-2024-lect-a-hamilton-string-theo-unif-geom-alg-01-nov-2024-online/) he writes “chief arguments against bosonic string theory without any supersymmetry are, first, that it does not admit fermions, and second, that its ground state is tachyonic, therefore unstable. But D-branes, discovered by Polchinski in 1995, are surfaces upon which open bosonic strings can end, and those D-branes can carry fermions.”.
Just want to acknowledge what seems to me as a mature, reasonably humble, and constructive reply to a public criticism. That’s not easy.
Being accused of doing bad physics by the person who wrote ‘Dear Qubitzers, GR=QM’ (1708.03040) has to be some sort of badge of honour
It’s so tragically sad that halfway (!) to a moment of realization that could genuinely help the field, Susskind devolves into ad hominem attacks. So many follow this route…. If he found something in Peter’s or Penrose’s work that he disagreed with, why not engage in ideas, through a letter, or in a journal? You’re a scientist; do science.
Trouble is, if you search Inspire for papers on the topic string theory the paper production isn’t tapering off. It’s just that we don’t hear about it much anymore. Swampland is tapering off. Wormholes are still going strong. And quantum simulations are at all all time high.
Hm, read Not Even Wrong when it came out – only just discovered your blog! (Yup, I’m a bit slow, but we get there in the end)
Just to repeat the previous poster, it is admirable to see ad hominem attacks politely brushed aside after finally being accepted as correct about String Theory’s limitations and cultural stagnation. It’s only taken nearly 20 years – hats off!
“Susskind says [he] is trying to extend string theory to something different which will work in dS space”
So, Suskind is trying to formulate a superstring DeSitter holographic duality? I would have thought many have tried to do that: do you know of such a research program? Are there serious blockers?
Please give us an umlaut: Kähler <— Kahler.
The part where he said that he still believes in GUT despite no observed proton decay was also interesting.
Consensus? Well, once there was one about phlogiston. The luminiferous aether had a period of reasonably broad acceptance. Observation contributed significantly to changing the consensus.
In the present, with little hope of observation as a correction for the frontiers of high energy physics, consensus seems a rather dubious lodestone. Seemingly one now just has to muddle about, with the hopes that many people doing the same on some different tack will eventually lead to someone having a breakthrough, hopefully with observable consequences. Very difficult. Extremely risky. Thousands of theorists working, most or all of them doomed to failure. I fail to see how modern academic consensus could possibly support such a diversified endeavor. Major structural change would be needed, and it seems entirely unlikely to materialize. What a prison consensus has turned out to be.
Dr. Dave,
Actually I much prefer the straightforwardness from Susskind to the usual tactic of string theorists who disagree with me, which is to do everything possible not say my name and pretend I don’t exists, acting like it would be beneath them to engage with anything I do.
This was different partly because Jaimungal went out of his way to ask repeatedly and specifically about me, partly because Susskind in general makes a point of not mincing words and saying exactly what he thinks.
The way he went on about the “consensus” was fascinating. It is exactly the way most people in this field think, although they don’t say so out loud. Because this is the way almost everyone behaves, nothing is going to change here until the day the “consensus” changes. This interview is fascinating, showing Susskind trying to reconcile two contradictory things: “the consensus is always right” and “the consensus is stuck”.
Sabine,
Yes, that’s really strange. I’m serious when I say this is a subject that has intellectually collapsed and died, but that doesn’t mean people have stopped writing papers. It really has become kind of a zombie situation.
Jim Eadon,
Susskind’s comments about other people not trying do dS holography are pretty odd. Obvious places where the string program got stuck and doesn’t do what you want (M-theory, holography with non-negative curvature, etc) have been well-known for decades. People don’t work on these because there’s no promising idea, all evidence is that these are speculative ideas which don’t work.
Sting Theory will never end until this blog ends, as this blog has talked more about string theory than any other blog/youtube channel/media outlet over the past three decades.
The sooner Peter retires, the sooner physics can move on.
“Actually I much prefer the straightforwardness from Susskind to the usual tactic of string theorists who disagree with me”
Having seen conflicts in other fields outside of physics, I tend to agree with this. The one thing more rude than having heated discussion is ignoring the person you disagree with and hoping/persuading others do the same. I’ve seen people ignore disagreements with others under the name of diplomacy, but IMO this behaviour is doing much more damage to the field.
(though on a more realistic note maybe it’s sometimes not practical to always argue, protracted arguments are expensive time and energy wise and you never know if the person you argue with will behave reasonably)
Pauline,
Maybe that’s the explanation for why people are still writing over a thousand meaningless papers about string theory each year: they’re doing it just to piss off Sabine and me, if we stopped, they’d stop.
Maybe I’m being overly generous to Susskind, but I read his comments as saying that theories outside of the consensus are usually (implying not always) wrong. And that’s correct – most theories that go against consensus turn out to be incorrect.
However, contrary to Susskind’s admitted derision of non-consensus theories, a willingness to buck consensus should be applauded even if the effort ultimately fails, especially when consensus is stuck. It’s that 1 theory out of 100 that actually works and upends consensus that makes the 99 failures all worthwhile.
This is the first time I leave a reply here but I have been a long-time reader of your blog and appreciate very much your critical comments. I am also curious about your recent papers as I work on the Osterwalder-Schrader axioms and await enthusiastically your further developments.
Apart from the interesting admission of Susskind about the failure of the string theory as a theory for the real physics, I would like to know what he means when he says “(the String theory) is a very very precise mathematical structure, it exists, it’s well-defined”. It is well-defined in which sense? Do I understand correctly that even the perturbation series for the S-matrix is not proven to be finite even on the flat 10-dimensional spacetime?
Yoh Tanimoto,
When physicists like Susskind say things like what you quote, you need to take into account that they are spending most of their time doing things like “ER=EPR” which are completely ill-defined. The minute they have any actual equations, then they have a “precise mathematical structure”. If they can compute a one-loop term, then it’s “very, very precise” and the theory is well-defined.
Possible problems with the superstring perturbation expansion in flat 10-d spacetime are not really the point. The problem is that to leading order this looks nothing like the real world. “Compactification” schemes to make it look like the real world quickly get you deep into ill-defined territory. Or, more accurately, anything well-defined doesn’t look like the real world, you need to head off into regions where there is no well-defined theory to avoid falsification of your idea. This is what happens when you start with a wrong idea, except that normally people give up (or don’t wait 40 years to do so).
Susskind’s unprofessional behavior and continuous ad-hominem attacks are probably one of (but certainly not only) the reasons nobody (but string theorists) take string theory seriously anymore. Obviously none of us can shout it at a conference because we fear funding repercussions, but we all say it privately during coffee breaks. It is fine to criticize and even trash a paper – it is not fine to use the argument “who is X” to make a point.
Ryan,
”It’s that 1 theory out of 100 that actually works and upends consensus that makes the 99 failures all worthwhile.”
No. The good theories replacing more limited theories were always due to better insight, not due to trying all possibilities. Only those few failures that increase the insight in a significant way are worthwhile. It does not help to search under 99 or 9999 lamp posts when the key was lost where there was less light.
I see… so, for mathematicians, it is not well-defined even in the flat 10d spacetime.
Well, I understand your point that the passage to 4d is much more problematic. I think I would still appreciate if something (quantum) were really well-defined even in 10d, but I’m afraid I won’t see it as no one is working on it.
Yoh Tanimoto,
I think the “consensus” view on 10 d superstring theory in flat space time is:
1. How to define the perturbation expansion for all terms in higher genus is a very technical problem, conjecturally all such terms will be finite.
2. The perturbation expansion is just an asymptotic expansion, and will not converge for non-zero coupling.
My interpretation of this has always been that, as far as mathematicians are concerned, this is not, even conjecturally, a truly well-definied theory. Susskind is using a different definition.
Is Susskind claiming that string theory is “the consensus”? Among whom?
People he talks to? Most physicists I talk to have decided long ago that string theory is just failed physics. What on earth does he mean?
Erich Weiss,
Who he talks to is theoretical physics faculty at Stanford, Harvard, IAS/Princeton and a few other places. While their colleagues in other fields of physics are realizing this is a failed project, they’re sticking to the “consensus” that string theory is “the only game in town”. For them, those in other fields who don’t agree are just basically ill-informed about the complexities of string theory and its great achievements, and not as bright as they are. Those in the same field who disagree are “weirdos”.
Many years ago, I pointed Peter to a talk by the late, great Sir Michael Atiya, given at David Gross’ group in California. I remember that he expressed his bewilderment at the baroque nature of string theory, and the amazing fact that QM depends upon the ad hoc invention of man, the square root of -1. He said, if string theory turned out to be true, he would find it entirely bizarre that nature was constructed that way. Maybe, Susskind’s recommendation to “start over” should begin there. Just saying.
Douglas Bundy,
I’d forgotten about that. For those interested, see
https://online.kitp.ucsb.edu/online/plecture/atiyah/
Since his conversation with Lawrence Krauss, I have been puzzling over Susskind’s remarks regarding supersymmetry. I still don’t really get why he’s so negative about the prospects for supersymmetry breaking within string theory. But maybe the real reason why he wants a non-supersymmetric generalization of string theory, is that his top priority is to understand quantum gravity in De Sitter space. That would make sense to me.
Just pointing out that it should be obvious that there’s a difference between the usual consensus that happens in heatlhy science (which is an essential part to drive progress), which is strongly backed by empirical findings, and its degenerate, cancerous growth relative, the group think. Of course, what Susskind sees as the former in his field, most of the rest of the community has finally reached, or is reaching, a consensus (oh, the irony! haha) that it’s actually the second thing, a tumor that has to be removed before it kills the patient.
What Susskind calls “consensus” is nothing more than a groupthink. The change will not come from those that comfortably live from the current situation.
After listening to this carefully, I am of the opinion that something extraordinary happened during the interview, and it is due to the host that things got really interesting.
As I understand it, Lenny is trying to say something radical within the established frameworks, but without losing those paradigms and questioning them. It’s like saying “We need to challenge all our assumptions and go back to square one…within String Theory.” The host then does something that I have never seen. He respectfully lets Susskind go on, but points out the contradiction that no one has been able to get at for 40 years: how can all of these theorists say with certainty that String Theory is the “Only Game in Town” when they clearly have no idea if there is anything going on outside of String Theory?
What is unique about Jaimungal as host is that he has far more information than a non-physicist like Dennis Overbye or KC Cole. But as a non-physicist, he can push back on the interviewee for the first time. When the guest says something wrong, Jaimungal is right there pointing out that the claims being made are actually wrong. This has never happened before. Either you get interviewers like S. Carroll and NDGT who are not String Theorists (but can seem like marketers for Strings), or you get people who have to take what is claimed on faith. I think what just happened is that we saw our first earthquake interview of this type. Respectful. In depth. But not Public Relations for String Theory.
The result was that Susskind actually changed his position DURING the interview. He went from saying “Follow the consensus and avoid the weirdos almost always.” at the beginning, to “You should ignore what I have to say about other researchers and other games in town. I don’t know them. I don’t know those people so I shouldn’t say. Maybe they are doing things. We don’t know. We tend to be very derisive in String Theory. This is not exactly a secret. It’s entirely possible the answer is with someone who is outside.” I have never seen this before.
Alex: Over the past year, I am hearing more and more string theorists pompously claiming that information paradox has been solved using wormholes-entanglement connection of Susskind. And I am talking here about well known string theory researchers!
Having watched the entire interview, I found it a generally laudable discussion all around. My two main quibbles are the paradoxical advice to young theorists and the terse dismissals of “bad” physics. For the latter, I should point out I greatly appreciated Susskind’s concise summary of the difficulties of the Many World’s interpretation of QM. I’ve seen all of them before, but I think Susskind has a talent for rendering a pithy verdict. So it’s disappointing, if he’s actually put in the time to read a paper (which seems to be the case) to hear it simply dismissed as “bad”. I would have genuinely been interested in his reasons for thinking so. Apparently he couldn’t be bothered.
As for advising new generations of physicists to strike out on their own, he seems to be saying they should be willing to commit career suicide en masse for the good of the discipline. That doesn’t seem to me to be very helpful. Rather, I would have liked for him implore folks with tenure to give talented young mavericks a living, and provide them the opportunity to at least fail well. It likely does really come down to the incentive structure, and high-minded exhortations to follow Truth will never be enough.
Eric+Weinstein
From the interview: “We tend to be very derisive in String Theory. This is not exactly a secret. It’s entirely possible the answer is with someone who is outside.”
Again, he can’t close the box definitively,by not saying that a completely new theory will come from outside, but instead pretending that an answer to String problems will come from outside. It is Zombieland and the Outsider.
That is almost as devastating as things could possibly be for string theory as a physical theory of nature. And his ad hominem attack is disappointing in somebody of his stature: it makes him look small-minded and vindictive.
While Susskind does acknowledge the possibility that outsiders/weirdos might do something important, it’s in the context that they might come up with an idea that will build upon the consensus research program and vindicate it. I don’t hear him anywhere acknowledging that the whole program might be a mistake. For him and the others at Stanford/IAS/Princeton/Harvard/etc. who have been doing this for forty years, it seems to be completely inconceivable that the whole thing could be wrong.
String theory has always been a deeply weird program. Obviously many technical developments “along the way” were amazing and useful (2D CFT e.g.), but the original 30 years of worldsheet theory building is one of the strangest developments I can think of in theoretical physics: trying to build a many-particle theory of everything from the vantage point of a single string, splitting and merging into a few others, operating in an ambient fixed background spacetime.
And I still don’t get why the branching polymer result due to Froehlich et al wasn’t a big alarm bell for the field back in 1984. Ivan Kostov claimed to me that the problem doesn’t resolve in the critical dimension either. If the bosonic string doesn’t exist because summing the configurations of a liquid membrane always gives you singular branching surfaces, how can one be sure the same affliction doesn’t obliterate superstrings?
The production of approximately 1000 papers per year for the last 30 years is definitely worrisome: Search on `String Theory` at INSPIRE shows the 1984 `first revolution` with persistent production around 1000 papers per year since about 1996 (the `second revolution`)
https://inspirehep.net/literature?sort=mostrecent&size=25&page=1&q=String%20Theory
In an attempt to understand more about this persistence I came across Barry Bradlyn’s analysis from 2009 (not sure if it was ever discussed on this blog as it seems relevant to the conversation)
https://web.mit.edu/demoscience/StringTheory/
Unfortunately, Bradlyn’s analysis is now 15 years old. It might be interesting to learn what an update of that analysis demonstrates.
But assuming that a career is about 40 – 45 years, the persistence of this intensity of production over 30 years suggests the possibility that the field of String Theory research escapes Thomas Kuhn’s mechanism of paradigms shifting as workers literally die off . The generation `born` in 1984 should be retiring now. If ST is a Kuhnian paradigm, then production should be dying off as the first generation retires. But that is not what seems to be happening.
`Echo chamber` effects tend to produce persistence but more practically funding produces root cause persistence. The feedback loop between funding and echo chamber effects is then a possible mechanism for the escape of ST from the natural history of a scientific paradigm.
Matthew Foster,
I tried to find the result that you’re talking about. These two 1985 papers by Ambjorn et al are still being cited:
https://inspirehep.net/literature/212827
https://inspirehep.net/literature/221042
In the second paper they explore what they call a regularization of the Polyakov action, which may avoid being dominated by the branched polymers for dimensions between 2 and 30.
I can’t say if this work ultimately converged with nonperturbative studies of string path integrals within string theory; most of its current citations are in topics like “2D Liouville quantum gravity”.
Hi Mitchell
The status is also reviewed here,
https://arxiv.org/abs/1310.8552
I actually became aware of this stuff as a cm theorist, due to applications of Liouville CFT to various condensed matter problems (although no results were directly obtained from LCFT as far as I can tell, and instead extracted from mappings to simplified glass-transition models).
I haven’t dug too deeply here but spent some time discussing with Ivan Kostov at a conference. He said that many attempts had been made to tame the branching polymer solution (by adding higher-order terms to the N-G action), but these “irrelevant” operators failed to tame the IR behavior. It’s a quite interesting statistical mechanics result, that the attempt to sum over “liquid” membranes becomes dominated by singular polymer configurations.
I can imagine various ways around this for the usual string program–these aren’t superstrings, Minkowski vs Euclidean, etc. Maybe AdS/CFT allows one to somehow claim that these problems go away in the non-perturbative limit, since the corresponding boundary SUSY CFT is weakly coupled and one can do calculations.
Then maybe it’s just dS vs AdS all over again?
Susskind is eighty-four, has a solid reputation and position, so my perception is that he can afford to say the “truth” on record, eventually. However, he chose his words very carefully, so as not to admit that an actual change of framework is needed. «We live in the wrong kind of world to be described by string theory» is, in my view, a masterpiece of linguistic manipulation, where string theory does not get to be associated with the adjective “wrong”: the world itself does. «We need to start over» sounds more like “di Lampedusa strategy” to me: keep working on the same things, while pretending that an actual self-correction has taken place. Kudos to Curt Jaimungal for pointing out that the “consensus” is biased, for showing that the leaders of the “consensus” do not actually know what the colleagues outside of said framework are working on, and for respectfully keeping the whole interview on track: that was no small feat.