String Theory Debate

Recently Curt Jaimungal offered to host a debate over string theory between me and a willing string theorist. Joe Conlon took him up on the offer and our discussion is now available. I think it turned out quite well, and gives a good idea of where Conlon and I agree or disagree, and some explanation of why we disagree when we do.

These days there’s a wide variety of different points of view about the topics we discuss among string theorists and theoretical physicists in general. A discussion with someone else would have covered some different topics. As here, I think most string theorists and I agree on quite a bit more than people expect. I’m happy this video provides a place to hear a discussion that goes beyond both the common sloganeering on the internet, and the extensive but one-sided content I’ve been providing over the years.

Update: For comments by John Baez on the “Great Stagnation” in fundamental physics, see here.

This entry was posted in Uncategorized. Bookmark the permalink.

37 Responses to String Theory Debate

  1. Deviant says:

    Is there a transcript available? I have felt for quite some time that in this sort of long discussion, a transcript helps a lot.

  2. Peter Woit says:

    Deviant,
    On youtube videos like this there’s an automatic transcript: use the “more” link, scroll down for the transcript link. It’s on a right-hand side area, keyed to the video, somewhat annoying to read independently of the video, but usable.

  3. John Hessler says:

    Peter—really nice session. I thought you really got your points across and it was nice to see such a calm and informed discussion.

  4. David says:

    Peter,
    The one clear message that I have learned from your blog (I think) is that string theory does not make any predictions that have a chance of experimental confirmation. However, when Conlon suggested that gravitational wave measurements might provide information related to string theory predictions about the early universe, I thought your response was rather tepid. Is this, in any sense, a viable experimental test of string theory?

  5. Chris Oakley says:

    I watched it to the end and was disappointed that they were only able to find you a soft-spoken, reasonable opponent. Hell – he even agreed that String Theory had a problem with hype! Next time, hopefully, it will be someone who will get the viewer numbers up like Lubos, Jacques Distler or David Gross.

  6. Eric+Weinstein says:

    Here is what I don’t understand. I watched “Quantum Gravity” get substituted for “Fundamental Physics” or “Unification” around 1984-7. Why are we not going back and saying what’s true: “THAT NEVER MADE SENSE…even at the time.” The problem with this program is that you can never say “Hey, it seems like insisting on quantizing Einsteinian gravity with zero encouragment from either theory or experiment is an abject failure of a kind never before seen. Just a thought.” The problem is, was, and remains insisting on “Quantum Gravity” as the only game in town. It’s not, and never was. Obviously.

  7. Bryan says:

    Peter,

    I’ve only watched the first half so far, but Conlon seems optimistic about the ability to experimentally test for axions. Would you consider evidence of an axion in one of these experiments (as unlikely as it may be) evidence for ST and higher dims?

    Also regarding Baez’s post, while you often point the finger as where we go off-the-track is requiring higher dims to explain a theory. Here Baez points to the problem of requiring the theory to be renormalizable. Would you agree this is a fundamental problem?

  8. Peter Woit says:

    David,
    It didn’t seem to me Conlon was claiming this as a “prediction” or “experimental test” of string theory”. From what I remember we were both responding to a question about possible future progress. Conlon was pointing to cosmology related measurements and axion searches as places where one might see “BSM” physics. If you’re a big optimist about string theory, you might hope that whatever new is seen along these lines matches somehow with some attempts to get the SM from string theory (this is kind of like the SUSY story).

  9. Peter Woit says:

    Eric Weinstein,
    I can understand why many people since the 80s have felt “we’re done with the SM, QG is the open problem, so it’s all about that”. But I’ve always thought that exactly because QG effects can’t be directly measured, the way to know if you have the right QG theory is going to be to understand how to unify it with the SM, then test the unified structure,

    Another way to say this is that the problem with QG theories is that if you believe the optimistic claims of people working in the area, we have a huge number of consistent QG theories, but to the extent they’re decoupled from the SM, no way to test any of them.

  10. Peter Woit says:

    Bryan,
    An axion field would be very interesting to see, but I don’t think such a thing in itself provides any evidence for string theory or higher dimensions. Maybe there are specific patterns of couplings to other fields for which you could make such a claim. To be clear, axion models are something I’ve never thought much about.

    I’m not sure exactly what point Baez was making about the renormalizability problem. As in my earlier comment, to me more interesting is unification. Perhaps if you understood unification with the SM degrees of freedom you would see that the renormalizability problem gets solved like in the SM (e.g. the way it is solved in non-Abelian gauge theory by asymptotic freedom).

  11. B. Malpani says:

    For those of us who prefer reading to watching television, you can use this link to download Youtube’s auto-generated transcript: https://downsub.com .

    And to make your life a wee bit easier, here’s the Youtube video’s url:
    https://www.youtube.com/watch?v=fAaXk_WoQqQ

    On Windows, you’ll need to open the file in Windows Write, and then save it (and you can simply overwrite the original file) in order to “activate” the line breaks. On Mac… don’t know, don’t care.

    I haven’t yet read the transcript, so I don’t know how accurate it is, and how many typos need to be corrected.

    As for the Baez link to “Mastodon”, well, white text on a black background is unreadable, and I didn’t bother reading or coping with it.

  12. Shantanu says:

    Peter, this is one of the first videos where I have seen you being optimistic/excited about neutrino experiments . You have written very little about neutrino experiments(or non-0 neutrino mass) in last 20 years on this blog. Anyhow we have already measured mixing angles and delta m^2. so which experimental measurement/result you are looking forward to? neutrino-less double beta decay? experiments to directly measure neutrinos.
    suprisingly (or un-suprrisingly) Joe didn’t seem too excited about neutrino results

  13. Four notes:

    1. Every video on the channel has scripts that I’ve meticulously reviewed as part of my retention process to learn from each interview. Therefore, they’re not “auto-generated.” They should be high quality, with only a few minor errors, especially compared to the significant number of errors in YouTube’s auto-generated captions.

    2. Interestingly, whenever I post anything on X regarding string theory—whether interviewing a guest who is pro string theory, against it, or neutral—there’s a slew of engagement from those who study (or have studied) string theory. However, when I posted stating I was looking for people willing to speak to Peter Woit, I had zero responses. Zero. That changed when Sabine Hossenfelder thankfully came to the rescue, retweeted it (re-X’ed?), and Joe responded. Joe deserves credit for being the only one willing to come forward.

    3. Thank you to Peter Woit for willing to come on. It was lovely seeing you in-person in Columbia two weeks ago.

    4. I’m open to hosting more conversations of this depth, length, and congeniality in the future between two sharp individuals who disagree on exigent topics in fundamental physics and philosophy.

  14. Peter Woit says:

    Shantanu,
    One reason I don’t write more about neutrino experiments is just that I don’t know that much about them and don’t have a specific example to point to.

    The general point I was trying to make in the video and have made elsewhere is just that the neutrino sector is the part of the Standard Model that is most complicated and poorly understood. At the same time, unlike the energy frontier, where we’re hitting fundamental barriers to getting to higher energy at a finite cost, studying neutrino couplings/masses doesn’t require such high energies.

    For something really exciting though, I think what is needed is both better measurements (are masses Majorana?) together with some new theoretical insight which would constrain the possibilities for neutrino couplings/masses, testable by predicting something about current or future measurements.

  15. B. Malpani says:

    @Curt Jaimungal

    Sorry; my apologies.

  16. Peter Woit says:

    Over on Twitter at
    https://x.com/quantum_geoff/status/1873582573810819161
    Geoff Pennington asks
    “Still not really sure what Peter’s point was?”
    about my comments about his new paper with Witten.

    To clarify: just before doing the video I was struck that morning by noticing
    1. The paper with Witten (https://arxiv.org/abs/2412.15549), which is (as typical for Witten) technically very impressive, leads the rest of us to think “no way I could ever do that, it’s amazing that he’s written thousands and thousands and thousands of pages of high quality papers like this”.
    2. That Witten was giving an opening talk here
    https://istringy.org/ism24/
    that was the same as he’s done multiple times in recently years. I wrote about a version of this that he published 9 years ago here:
    https://www.math.columbia.edu/~woit/wordpress/?p=8068

    I was really struck that morning by how this showed the huge difference between his research work that’s the best of the current state of where “string theory” has ended up and the dubious 40 year old story about infinities in the string worldsheet vs. particle theory worldline that he’s trying to sell as explaining why one should do string theory research. This seems to me illustrative of where the field is and what its problems are: our best theorists have abandoned work on a failed idea about fundamental physics, but are still promoting it and let it guide the field in unpromising directions.

  17. Amitabh Lath says:

    Thank you Peter for sharing this video. I found it interesting that the consensus about the deficiencies of the SM was the strong CP problem. I would have expected something about how the SM does not address Dark Matter/Dark Energy, or maybe the hierarchy problem.

  18. Peter Woit says:

    Amit,
    Conlon was the one pointing to the strong CP problem, largely because a popular way to resolve it is an axion field, and these show up in string theory. As I tried to explain, I’m in the camp of not even sure it is a problem. The problem comes about when one thinks about theta-vacua in QCD, but I’m not convinced we completely understand these, so can’t be sure they cause a strong CP problem. I don’t have a specific argument about this, but in the video mentioned someone who did, Hidenaga Yamagishi. For more about him and a link to a paper of his about this, see
    https://www.math.columbia.edu/~woit/wordpress/?p=93

  19. Amitabh Lath says:

    In that case I would like to know what Peter Voit thinks is the biggest hole in the SM. Other readers with strong (!) opinions please chime in.

  20. Peter Woit says:

    Amit,
    The problem is that, in terms of relation to the observed world, there really aren’t any holes. You experimentalists have not been doing what is needed: finding such holes. The astronomers are telling us about “dark matter”, but maybe that’s just them not understanding the composition of galaxies etc. as well as they think.
    The part of the SM experimentalists haven’t investigated that well is the neutrino sector, and maybe dark matter fits in there somehow (once you’re including right-handed neutrino fields).

    Theorists on the other hand have a bunch of holes in their SM. Besides not being able to calculate lots of things about QCD, non-perturbatively we don’t even have a completely viable version of the weak interactions (problems with chiral gauge theories), and even understanding gauge symmetry is often unclear (BRST problematic outside of perturbation theory). Then, what happens with the Higgs field at extremely short distances still unclear. And, of course, Wick rotating chiral spinor fields is not yet understood…

  21. Marty Tysanner says:

    Hi Amitabh,

    You asked for input from people with strong opinions about problems with the SM, and since my opinions on the SM don’t seem off-topic for this posting, I’ll jump in.

    First, I’ll note the answer to your question depends on what you expect from the theory. If you are looking for deviations from SM predictions, then as Peter has made clear it appears very difficult to come up with a compelling answer. On the other hand, if you think a fundamental theory should not merely be quantitatively descriptive, but also predict the structure of the SM (and even better, also give a compelling picture of the fundamental origin of properties like internal symmetries, spin, a discrete mass spectrum, quantization of electric charge, the relative strengths of the four basic interactions, and so on), then answering your question is easy!

    It’s easy because we only need to look at the parameters of the SM and ask, “Why those numbers instead of different ones, and where did those properties come from? Why is there a discrete mass spectrum, or even notions of mass and momentum at all? Why do stationary states exist?

    Sure, we can invoke anthropic arguments, but such arguments don’t get us closer to answering very basic questions like those just mentioned. Instead we keep looking for clues about a deeper theory by looking for deviations from the theories we have — and this may be the best we can do for now — but then we are limited by the predictions of purely descriptive theories like the SM.

    By contrast to what seems to be a strong consensus among theorists, I believe the most important reason fundamental physics has stalled is that we lack understanding of the underlying origin of properties of matter that we observe experimentally. And this lack, even more than the lack of interesting new data, blocks progress because the kinds of phenomena we look for experimentally are guided by, and largely limited to, the predictions of a purely descriptive theory and its wannabe (and also purely descriptive) BSM successors.

    Since my own research interests are especially directed toward creating models that can answer questions like I posed above, and many other questions, I can confirm that it is hard to come up with models of elementary particles that are consistent with all that we know. (My own focus is charged leptons, especially the electron.) It is hard enough that it is completely unsurprising that few physicists are interested in that research direction, especially anyone who doesn’t already have tenure or another secure position. But from my experience so far, I don’t think the quest is doomed from the start (I don’t expect you or others to believe that, but that’s okay).

    I’ll go out on a limb and suggest one area where I think it is possible (from the perspective of a tentative microscopic model of the photon emission and absorption processes) to discover something interesting. The model indicates photons that electrons emit are not truly indistinguishable from photons of the same frequency that a muon emits, or a tau. The distinguishability is subtle, and I haven’t thought about it enough to argue for a specific experimental setup. And of course the model might be wrong. 😉

  22. Shantanu says:

    Amitabh: I agree with Peter about dark matter. It could also be standard model particles
    See for example https://arxiv.org/abs/1410.2236. or it could be a massive graviton in which case it has nothing to do with particle Physics or BSM.
    also standard model cannot explain equation of state of neutron star or even whether a strange quark star exists. I am surprised that no one mentions this as a limitation of standard model.

  23. Amitabh Lath says:

    Hang on a minute Peter and Shantanu! What do you mean DM is not a problem for the SM? Assuming you are not bringing back MOND type theories the solution must be some sort of particle, even if the thing has a Compton wavelength the size of a Solar System. And right now there is no place for it in the posters we hand out at APS, even though DM is observably the most abundant matter there is. It would be like if Mendeleev did not have a spot on his table for Hydrogen. All we really know about this particle is that it has zero EM, Weak, and color charge. It’s a ZEPLIN sized hole in the SM!

    Marty I share your aversion to anthropic ideas. I would be interested to learn more about flavor-dependent EM, if you can point me to some reading. You might be surprised what LHC experiments can accomplish, given motivation and the right trigger schemes. But I don’t count the lack of deeper understanding of things like mass spectra (or as Peter pointed out in the podcast, why SU(2) X SU(3)?) as a hole in the SM as such. Maybe that is a problem with the “theorist” SM, but DM is a glaring problem in the “experimentalist” SM.

    Shantanu, I work with jets and I am well aware of the limitations of QCD calculations. Even at LHC energies we don’t have a leading-order understanding of what quarks and gluons are doing (we have our own “string model” of fragmentation and hadronization, and it’s laughably ad-hoc). Asking for “cold” QCD calculations such as quark stars is probably beyond hope. But I wouldn’t count that as a “hole”, just that SM gives us a non-perturbative theory that our computers are not big enough to solve. Yet.

  24. Peter Woit says:

    Amit,
    The thing is, a right-handed neutrino field is going to exactly fill that hole: no weak, EM or strong charge. Extending the SM by including these is the simplest way of giving neutrinos mass (you need the right-handed neutrino field to write down a Dirac mass term). Looking at the pattern of elementary matter fields, there’s an obvious hole where this should go. Another way to see it is to look at the way a generation appears as a spinor for SO(10) in GUT theories. The SO(10) spinor rep is 16 dimensional, the SM without a right-handed neutrino field has 15 fields in a generation, the right-handed neutrino has exactly the right properties to fill the hole and give 16.

    To me, extending the SM by Dirac mass terms for neutrinos is not much of a change so I think of this as still the SM. People like to argue whether this is “SM” or “BSM”, but that’s just an argument over terminology.

    Once you have right-handed neutrino fields, you can have massive sterile neutrinos, and these make perfectly good DM candidates.

  25. Shantanu says:

    Amitabh
    The only thing we know about Dark matter through cosmological observations is that it is not baryonic and not made up of atoms . There is no evidence for WIMP miracle or any other hint that Dark matter has anything to do with BSM particle physics.
    To give a few examples of non-BSM physics solutions to dark matter
    1. https://arxiv.org/abs/1410.2236 considers a model DM made out of only standard model particles
    2. https://arxiv.org/abs/0805.1519 (This is also modified gravity in a different avatar, different from MOND)

    3. It also could be a reformulation of GR (no modified gravity)
    See https://arxiv.org/abs/1308.5410

    There are obviously more. One thing I am also surprised that people have taken WIMP miracle so seriously even though it is only a numerical coincidence, whereas other numerical coincidences such as Koide mass formula are ignored by the particle physics community.

  26. John Baez says:

    If oracular comments are allowed here, I predict neutrinos will be fit quite well by the least surprising extension of the Standard Model, namely the one Peter is talking about, which I usually call “the Standard Model” or “the new Standard Model”.

    If so, the surprise from neutrino physics will be why it’s so unsurprising – why the Standard Model works so well.

  27. Paolo Bertozzini says:

    Dear Peter,

    I watched the video. Interesting to see finally some common agreement on some of the several points related to the crisis in fundamental physics.

    Maybe I am one of the very few that has some problems to agree on the usage of the purported “lack of new experimental results” as a “deus ex-machina” explanation/cause of the current situation: contrary to what is often claimed, theories are rarely dismissed or falsified by experimental outcomes … dominant theories are “tuned” and “adjusted” in order to fit the additional evidence; minority-supported theories are, often unfairly, “falsified” (see the case of Connes’ NCG standard model’ s original prediction of the Higgs’ mass). Furthermore, experimental physics alone is actually totally blind, if not guided by theoretical indications precisely describing what to search and where. Some examples: the mathematical apparatus and theoretical pillars of quantum theory have been put in place around 1925-1930, but the experimental evidence indicating the structural incompatibility with classical calculations where there since 30 years before; Michelson-Morley experiment did not play such a great role in Einstein’s initial elaboration of special relativity (and was peacefully coexisting with patched aether theory) and general relativity was born without any significant experimental need (that mostly came a-posteriori).

    Today there are even too many indications of experimental problems especially in cosmology that, instead of being addressed with solid theoretical proposals, have been “patched” with models of inflation, dark matter …

    My very personal opinion is that the current issues in fundamental particle physics are mainly attributable to the fact that QFT is not yet well-understood and is not yet a rigorous theory mathematically … this has been long forgotten because renormalization in the ’50-’60s has given a universal excuse to believe that such foundational problems are irrelevant (QFTs are all “effective”) and one can just define QFTs by writing clever (renormalizable) classical Lagrangians as input of a quantisation via Feynman path integration.

    As You correctly say in the video, young generations are not aware any more of the intricacies and foundational issues of QFT … but the main culprit here is not so much the “string theory cult” per-se … but rather the fact that QFT has been reduced to “Feynmanology”.

    Best Regards and Happy 2025 !

  28. Peter+Shor says:

    At one point, I briefly looked at Connes’ derivation of the Standard Model Lagrangian using non-commutative geometry. I didn’t understand it, and it seemed like black magic to me, but one interesting thing is that you can’t get all Lagrangians in this manner, and apparently one thing that is forced upon you is a right-handed neutrino field.

    I don’t know whether it constrains neutrino physics more than this, but if it does, now is the time to make predictions about neutrino physics, before they become postdictions.

  29. Peter Woit says:

    All,
    Sorry, but please try and stick to comments about the topic of the posting, this is becoming the kind of general discussion forum I can’t moderate.

  30. Marshall Eubanks says:

    Excellent discussion.

  31. John Baez says:

    My point may have been clearer by the end of that series of posts. The quest for a renormalizable theory of perturbative quantum gravity helped push physicists first into supergravity and then superstrings. These theories describe a world unlike our own – and in the end they were never shown to be renormalizable!

    In my posts I tried to explain to nonphysicists what the assumption of renormalizability means – physically, not mathematically – and why we should question it for quantum gravity. I did a middling job, and maybe sometime I’ll try again on my blog. I did not at all get into the fact that treating quantum gravity as a theory set on Minkowski spacetime, or any other background, is a conceptually incoherent blend of ideas that doesn’t deserve to work, though it’s probably fine as an approximation. I got a bit distracted thinking about new work on perturbative quantum gravity, and arguments about whether what Newton’s constant G “runs”, and what this means. I left off trying to point out that there’s a lot of unexplored room for theories that drop traditional assumptions we have no good reason to make, except for convenience.

  32. Shantanu says:

    Peter, what are your thoughts on non-commutative geometry?

  33. Max says:

    Hey Peter,

    during your discussion you mentioned open problems with the standard model that you fear are being overlooked by many young students getting into the field. As a master’s student with background in QFT and some of its non-perturbative aspects could you point me to some references that discuss these issues that you think I should be familiar with?

    Cheers,
    Max

  34. Peter Woit says:

    Shantanu,
    It depends what you mean by “non-commutative geometry”. Something I think of as absolutely fundamental to the relation between physics and mathematics is what is sometimes called “geometric representation theory”, but could be called “non-commutative geometry”. When you are doing representation theory you are studying a very non-commutative algebra (e.g. the universal enveloping algebra of a Lie algebra) and relating this to geometry (e.g. the flag variety of the Lie group) is an important tool. This is a particular kind of non-commutative algebra.

    People sometimes mean by “non-commutative geometry” the naive idea that your space-time coordinate functions become non-commutative. I haven’t seen anything interesting coming out of that naive idea.

    By “non-commutative geometry” people often mean the Connes program, which is using specific techniques. That’s a big subject I only really understand well parts of (e.g. the way the Dirac operator gets used). My guess is it needs some new ideas to be fully successful. Just today I saw there’s a new paper and announcement about recent work in this direction, see
    https://jespergrimstrup.substack.com/p/a-new-framework-of-unification
    and
    https://arxiv.org/abs/2501.00005

  35. Peter Woit says:

    Max,

    I’m not sure if I’ll have time to look up references, but here is a list of specific technical issues with the SM that everyone should be more aware of:

    1. In pure SU(3) Yang-Mills theory on the lattice we have a beautiful definition and we can do lattice calculations which conjecturally have a well-defined continuum limit (biggest problem in the field is well-known, how to find a controlled analytical calculational method for this limit). Once you try and add quarks, this becomes much less well understood. This is also a well-known problem. A lot is known and there are many proposed ways to do this, but I don’t think any is really satisfactory (an example, “rooting” problems when you try and define using not what you want, but something that gives a power of what you want).

    2. The BRST method used to get a sensible perturbative expansion in non-abelian gauge theories seems to not really work non-perturbatively. In the continuum, there’s the Gribov problem, on the lattice the 0/0 problem first pointed out by Neuberger.

    3. The standard textbook explanation of the Higgs mechanism as spontaneously breaking a gauge symmetry really doesn’t make sense (see Elitzur’s theorem). It seems you need a way to think of this that doesn’t break the gauge symmetry.

    4. We don’t seem to have a fully satisfactory non-perturbative definition of a chiral gauge theory, e.g. as a lattice theory. You’ll notice that people don’t do lattice calculations of the electroweak theory, and this is because there isn’t even a viable definition of the theory.

    Disclaimer: For all of the above, there’s a huge literature, full of all sorts of claims to have partially or fully solved the problem. In all cases I’ve tried to understand such claims, I’ve sooner or later given up in the face of a confusing and complicated situation. So, I’m in no position to say definitively that there is no known solution. In situations like this, what you need are true experts who have spent a lot of time working on the problem, know the literature, and can tell you exactly what the main problem is and the extent to which it has or hasn’t been overcome. I think a big weakness of the field right now is that for a very long time it has been extremely difficult for younger people to successfully have a career working on these problems. So the expertise resides mostly in some older people who started working on this long ago (e.g. pre-1984…) and fewer and fewer such people are active as every year goes by.

  36. Peter Woit says:

    Max,
    I should also mention what I think is the more important example here, which I do have some expertise about: the problem of Wick-rotating spinor fields. I’ve written about this elsewhere, in particular recently at

    https://www.math.columbia.edu/~woit/wordpress/?p=14279

  37. Paolo Bertozzini says:

    @Peter+Shor and those with some interest in Connes’ standard model:

    a quite good updated description of the standard model in Connes’ NCG approach can be found in the coming second edition of W. van Suijlekom’s book “Noncommutative Geometry and Particle Physics” (https://link.springer.com/book/10.1007/978-3-031-59120-4). The details on the standard model are in chapters 13, 14 (and beyond SM, in 15); some discussion of neutrinos is in section 14.1.3.

    Sorry if this is going a bit out of topic …

Comments are closed.