Dreams of a Final (or Better) Theory

John Horgan has an interesting interview with Steven Weinberg here. Weinberg isn’t very optimistic about possible progress these days:

Horgan: In 1995 you told me that it’s a “terrible time for particle physics.” Are you feeling any better about your field now? Are there any particular advances that give you hope?

Weinberg: I’m not much more cheerful.

Horgan: Do you still believe in the attainability of a “final theory” of physics, one that ends what you called “the ancient search for those principles that cannot be explained in terms of deeper principles”?

Weinberg: I still expect there to be a final theory, but I’m less confident that humans will discover it in this century.

When asked about when string theory should be abandoned as a dead end he ignores that part of the question:

Horgan: In your new book, To Explain the World, you write that “scientific theories cannot be deduced by purely mathematical reasoning.” Doesn’t that principle apply to string theory? At what point, if ever, should string theory be abandoned as a dead end?

Weinberg: String theory may be inspired by mathematical reasoning, but not deduced, and certainly not confirmed.

He defends the multiverse with

Further, if we find some future theory that does make successful predictions about a lot of things, which turn out to be true rather than false, and if that theory also predicts the existence of a multiverse, then we should take that prediction seriously even though it can’t be tested directly.

which is true enough, but doesn’t address the fact that there is no such theory. The string theory landscape “prediction” of a multiverse is exactly the opposite sort of thing, not a corollary of successful predictions, but something being invoked as an excuse for failure to make predictions about anything at all.

For something more substantive, I recommend Alessandro Strumia’s theory summary for Moriond 2015. It has a lot of interesting commentary about a range of phenomenological topics. On the multiverse and anthropics he takes a quite different point of view than Weinberg:

Nobody talked about anthropics at Moriond 2015. This has an anthropic interpretation: Moriond is not in California. Clearly, social factors are playing a role, as always when experiments cannot set the issue. On one side, `having discovered the multiverse’ is physically indistinguishable from `having pursued a failed unification program’, but sounds much better. On the other side, future physicists could consider us as crazy for not having immediately accepted anthropic arguments.

He discusses “naturalness” extensively, including explaining why anthropics is no solution, since it doesn’t explain an unnaturally small Higgs mass.

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

21 Responses to Dreams of a Final (or Better) Theory

  1. Bernhard says:

    “Further, if we find some future theory that does make successful predictions about a lot of things, which turn out to be true rather than false, and if that theory also predicts the existence of whatever” , than yes, I would also take “whatever” seriously.

    About Moriond, I think there is also another reason. This is one of the most important conferences for HEP experimentalists. I guess talking about this embarrassing multiverse businesses is something better left for more stringy-focused conferences or at least for a conference where not too many people doing real science are around.

  2. JG says:

    “Weinberg isn’t very optimistic about possible progress these days:”

    What about you?

  3. gadfly says:

    No good data -> No progress. I think we can safely say that the limits of what we can know about so called “fundamental” theories have been reached for the foreseeable future.

  4. Jacques says:

    Humans will discover the final theory this century. There is a good probability that the final theory is a theory that explains why there is only gravity and the standard model.

    The research options in this segment are not exhausted yet. There is an anthropic reason for it (-: this research segment is not attractive for researchers, as it promises few – or even no – novel particles or effects.

    The scenario might be the following: once the LHC will have shown to everybody that there is no BSM physics, the search for a final theory will be rapid, because on the one hand, the research options are not exhausted, and on the other hand, the options are not legion.

  5. Yatima says:

    In other news Quanta Magazine has the second installment of the “looking at entanglement while thinking about spacetime” series: How Quantum Pairs Stitch Space-Time, which is hard to understand but which leads to Arxiv 1306.2164: “A Practical Introduction to Tensor Networks: Matrix Product States and Projected Entangled Pair States”.

  6. TOE says:

    “The string theory landscape “prediction” of a multiverse is exactly the opposite sort of thing, not a corollary of successful predictions, but something being invoked as an excuse for failure to make predictions about anything at all.”
    Right, this is highly suggestive of the “sickness” of a theory. In fact many different theories of quantum gravity presently produce a multiverse. See
    page 21. (Very importantly, this list puts the string landscape into perspective).

  7. Bee says:

    It isn’t true that finding a multiverse is the same as having failed at unification. It means you haven’t succeeded in finding sufficiently many assumptions (axioms/constraints) that give you a unique unification. That isn’t the same thing. String theory can well have the one unification they are looking for, just that they haven’t found it.

    I’ve explaint that in great detail elsewhere, but it really isn’t so difficult to see that every theory has a multiverse if it’s underspecified for exactly the reason that Weinberg points out: You cannot deduce a theory on purely mathematical reasoning. You need a set of postulates, and if you don’t have sufficiently many, you get an ambiguous result, that’s it. In the end, all these postulates must be emprically motivated (not counting mathematical consistency itself).

  8. Eduardo Lira says:

    How much did human beings know about SR in the last years of the 19th century? Zero, right? And about GR? Zero again. What about QM, not to mention QFT? Or the SM? All ridiculous questions, aren’t they?

    Well, considering such antecedents, and with another 85 years to go before this century’s end it’s quite absurd that someone (be it on this side of ST and the MV or the other) dares to assert that humankind has hit a dead end in the foreseeable future concerning the search for a “final theory” of physics. Nature’s complexity is fully independent of human impatience. Or a physicist’s wish to find definite answers during his or her lifetime.

  9. Art Brown says:

    So we’re running out of water, Brooklyn is moving here en masse , and our physics is getting dis’ed by old Europe. Just another perfect day…

  10. JG says:

    My question wasn’t rhetorical

    What do think about possible progress?

  11. Peter Woit says:

    I’m not particularly optimistic, in some ways now even more pessimistic than ten years ago. Back then I would never have believed that multiverse nonsense would be so influential, and I hoped the LHC would discover something unexpected about the Higgs mechanism.

    On the other hand, there’s a much wider understanding that the string theory/SUSY paradigm hasn’t worked out, and this may lead to more openness to looking for other ideas. There’s a wide variety of activity in mathematics related to physics, and this may lead somewhere. I have no idea though on what time scale…

  12. Mesmar Djehha, alias EFT says:

    Dear Peter,
    When you say “There’s a wide variety of activity in mathematics related to physics, and this may lead somewhere. I have no idea though on what time scale…”
    what if the time scale becomes longer than the one already taken by string theory, would you still believe that advanced mathematics would lead to a breakthrough in physics?
    Thank you.

  13. Peter Woit says:

    Mesmar Djehha,
    I don’t think the problem with string theory is the time scale for progress, but the sign of the derivative. As more and more has been learned about string theory, it has become more and more clear that ideas for how to get unification out of it and connect to the standard model don’t work (i.e. are empty, they answer no questions about the standard model).

    On the other hand, my own personal experience is that as I have learned more and more mathematics, I see more evidence for deep connections between physics and mathematics. Maybe this is a mirage that will dissipate as more is learned about these subjects, maybe not.

    Progress doesn’t happen linearly, but in jumps when people have good ideas. The question is evaluating the prospects for good ideas solving certain problems. I’d claim the probability of finding a compelling string theory explanation for the SM has been decreasing for a long time, with the “landscape” just the claim that it has hit zero, but that’s alright. I don’t see any evidence of a comparable phenomenon if you try and evaluate prospects for the use of advanced mathematics in physics.

  14. Theo Nieuwenhuizen says:

    String theory is a consistent framework to consider questions about quantum field theory. Various insights are gained from it. It has drawbacks due to its ad-hoc assumptions like the string nature of particles, extra dimensions etc. It has failed to recover, let alone to improve, the standard model. So the program is a failure for physics, but it can be a subject for mathematical physics.
    The multiverse is a poor way of escaping to admit the failure, it is a clear sort of cheating since its claims lie out of the domain for which the theory was formulated. Incredible that people believe this fallacy.
    Moreover, if we accept the multiverse, we must give up on falsifiability, but why then still do any science at all? We do have to keep separating science from religion.
    It is amazing to see that so many people who don’t accept God do accept the multiverse. Humans do need religion, but why such an obviously failed one?

  15. Steve says:

    When I read Weinberg’s comment about taking multiverses seriously I thought of his surprisingly sympathetic treatment of many worlds in his text “Lectures on Quantum Mechanics”. He certainly doesn’t endorse many worlds, but he gives it several pages of careful thought. Maybe he’s thinking that kind of multiverse?

  16. Chris W. says:

    I think Weinberg certainly appreciates the distinction, but he is trying to give careful consideration to both. In Dreams of a Final Theory he warns against dismissing a possible way forward in physics on the grounds that it seems unacceptable on a priori philosophical grounds.

    Of course, that begs the question of why the multiverse should still be considered a possible way forward…

  17. MikeS says:

    A final theory this century?

    A final theory, one assumes, has to be so logically compelling and self-consistent that things could be no other way. I would suggest we have a difficult conceiving how such a theory might even be framed, which rather gives serious difficulty in finding what it actually is.

    Something looking similar to the Standard Model is far too arbitrary to fit the bill. General Relativity, with its geometric elegance, may give a mere glimmer of what such a theory might look like – but then again maybe not.

  18. JFD says:

    I find it hard to believe that you recommend Strummia’s comments as “more substantive”. I’ll bet that it is only because the comments reinforce your preconceived biases. He clearly says “In my opinion..” and then, amazingly, “The most likely outcome should have been…” as if he has special insight into the laws of Nature. There is not much of substance there.
    There is a demonstrable need for the weak scale (quark and gauge boson masses) and the QCD scale (the nucleon mass) to essentially overlap – this produces the complex nuclei and atoms that we see. It is a valid scientific question whether this is a consequence of Nature having a unique ground state which luckily has this property, or a multitude of ground states sufficient to invoke anthropic selection. This may be difficult to settle any time soon, but it will not be decided by mere opinions.
    I enjoy your BS detector, which is great for puncturing a lot of the hype that is out there. But “Naturalness” has been the dominant form of hype for decades, and has made clear predictions that have so far failed. Lets hope that the LHC offers us a reasonably definitive answer – finding “natural” new physics would be awesome. But if naturalness fails, it provides extra motivation to continue to explore theories with multiple ground states.

  19. Peter Woit says:


    Not sure if it qualifies as more than an opinion, but to me it seems there are no solid arguments about what Higgs mass is likely or unlikely, and that seems to put me in agreement with Strumia. This is often stated as a “measure problem”, but it seems to me worse than that: you don’t even know what space the measure you can’t calculate is a measure on (what’s the space of theories you are putting a measure on?)

    Sure, maybe a better theory has many vacua with different physics, but this postulate isn’t worth anything unless it manages to explain something, not just provide an excuse. I think what Strumia was reacting against was the point of view pushed by Arkani-Hamed that the only two possible outcomes from the LHC are vindicating one of two ideas (“naturalness” or “multiverse”), both of which are popular, but neither of which actually explains much at all.
    Glad you enjoy the BS detector!

  20. Just Another Reader says:

    “… if we find some future theory that does make successful predictions about a lot of things, which turn out to be true rather than false, and if that theory also predicts the existence of a multiverse…”
    You go on to say that he “doesn’t address the fact that there is no such theory”. But would you not agree that your criticism arises as a result of a misread of what Weinberg said. He said “if we find…”—a conditional that surely concedes there is no theory at present. So perhaps your criticism of the Weinberg is, in this case, unfair. Or rather unneeded.

  21. Peter Woit says:

    Just Another Reader,

    I don’t think it’s a misreading, note that I describe what Weinberg says as “true enough”. The conditional “if” though evades addressing the difficult question, which Horgan was explicitly asking, of evaluating claims being made about the “string landscape”, in favor of saying something obvious about a hypothetical circumstance that is not the case.

    I’d be very curious to hear Weinberg address explicitly the “dead end” question: does our current best research indicate that the string landscape is an indication of failure and that the subject is at a dead end, or are there serious indications it’s a plausible picture of reality that is our “best hope for a unified theory” as string theorists often claim these days?

Comments are closed.