Many of you may know that Discover is not only a web site that hosts a diverse collection of entertaining blogs, but also publishes a monthly “magazine” printed on paper. Wild, right? Just ask this baby, who can tell you that a magazine is kind of broken when compared to an iPad.
Nevertheless, people read these things like crazy. I have recently started contributing an occasional column to the print magazine, known as “Out There.” (Our blog neighbor Carl Zimmer has been columnizing about the brain for a while now.) My first column appeared in the October issue (which comes out in September), and is now online — check it out.
The issue I’m tackling, under the draconian word count limit of an actual print magazine, is whether it’s scientific to talk about the multiverse. (Spoiler: it is!) Let me know what you think.
I agree. Let’s ignore “Dr. McGucken” from now on.
How does thinking about the multiverse effect the types of experiments we do at places like the LHC? How does it effect the way we explore the boundaries of physics? It seems like we hear a lot about things like wormholes because they capture the public’s interest, but my understanding is that they don’t actually exist. Is the multiverse like that or does the possibility give us hints about what types of experiments to try or about how to explain experiments that have already been done?
Full confession: I found this website during a late night web surfing spree that started with Plasma Cosmology. One can enjoy crackpottery to a certain limit, as long as it isn’t hurtful like anti-vaxxer conspiracy. However, it’s easy to recognize (even way outside of one’s own research field) by many symptoms. The first is an utter lack of ability, or perhaps interest, in actual communication.
Sean Carroll, have you ever considered the possibility that string/M-theory is physically wrong? SUSY is not a symmetry realized in nature, and space-time is only 4-dimensional., and GUT theories are all unrealized in nature .
“Sean Carroll, have you ever considered the possibility that string/M-theory is physically wrong? SUSY is not a symmetry realized in nature, and space-time is only 4-dimensional., and GUT theories are all unrealized in nature.”
No. He has never considered that posibility. Like EVER. Until he ran into this insightful blog comment, that is. Thank you.
Shrug.
PS: Sorry about the sarcasm, but it was irresistible to me.
To Curious Wavefunction: string theory might be over-sold, but it IS also a pretty exciting set of ideas. Nothing wrong in letting the public in on it. Afterall, public exposure of ideas is what lets people have the illusion that they are entitled to have an opinion about things they only have a foggy awareness of: like deciding what is a promising direction of research. 🙂 There is nothing like the sweet smell of napalm in the morning.
In any event, the public relations aspect might be important sociologically, but that was not my concern in my post. I was talking about where the science stood. And on that you seem to agree with me. Finally: I have always felt that the string theorists are usually more aware of the promises and limitations of their field than anyone else – whatever the public image of the subject might be. There is a touch of irony most (but not all) of us feel when we make jokes about string theory.
All of these comments are making me look arrogant and unpopular, but they are relevant and true, so they need to be stated.
somebody: to the lay public, is there anything in Sean’s article in Discover magazine, or in popular books by Kaku and Green and Randall or Hawking, that would suggest that string theory is not on the same footing as QFT and GR?
@curious Says,
As a layperson (in this area, anyway), I can say that probably every presentation I’ve seen about ST is packed full of caveats about testability. Sean’s article is more directly about Multiverse, and he provides appropriate caveats: “But the multiverse might be impossible to test directly. Even if such a theory were true, the worry goes, how would we ever know? Is it scientific to even talk about it?” (You have to consider the audience and length restrictions in deciding if this is sufficient, but I think it is.)
This point is also not understood by the public (and internet trolls): “These concerns stem from an overly simple demarcation between science and nonscience. Science depends on being able to observe something, but not necessarily everything, predicted by a theory.” I can say in studies of evolution (where I am not a layperson, and the theory is on much more solid footing than ST) that the public expectation is that we have to explain everything, perfectly, or else the theory is wrong. The theory tells me a lot of what to expect when we sequence a new animal’s genome. Nevertheless, when a scientist says “this phenomenon is contrary to expectation from evolution”, then the public and the media smell blood in the water (while a biologist sees new exciting science).
Pingback: The Eternally Existing, Self-Reproducing, Frequently Puzzling Inflationary Universe | Cosmic Variance | Discover Magazine
Phil (#19),
I turned off comments about this on my blog not because of the commenters (few if any) who disagree with me about the multiverse, but because of those who agree with me (numerous). This topic attracts comments empty of any intellectual comment, from both sides, and I just didn’t want to deal with them. One reason for this is that I’m pretty tired of the topic, nothing much has changed since Sean started on this promotional campaign even before Cosmic Variance was around:
http://www.math.columbia.edu/~woit/wordpress/?p=94
And no, string theory doesn’t make any generic predictions at ANY energy scale. Specific perturbative string theory vacua yes, but there’s an infinity of them, and the non-perturbative ones, which typically you can’t say much about at all. As “somebody” notes, string theorists do know this. They just try and ignore it when they’re writing for the public…
Peter Woit, #59,
If you’re so tired of the topic, why do you keep posting about it?
Doesn’t string theory at least predict extra dimensions to emerge at the Planck scale? Isn’t there any “stringy” signature that can be experimentally differentiated from point-particle behavior at the Planck scale?
And do you agree with the notion that there are many many quantum field theories, just as there are many many string theories, and that experiment is needed to find the correct quantum field theory that describes our low energy world? Only it’s much more difficult to build string theory vacua that resembles our own low energy universe, but there is no reason why, in principle, it cannot be done. So how is this situation different from QFT?
re: Charlie Says:
“But the multiverse might be impossible to test directly. Even if such a theory were true, the worry goes, how would we ever know? Is it scientific to even talk about it?” (You have to consider the audience and length restrictions in deciding if this is sufficient, but I think it is.)
Would this be sufficient if we consider a research program based on intelligent design?
Woit is again thoroughly missing the point when he says string theory doesn’t make any predictions at ANY energy scale. Or trying to be misleading? His posts in the past make me cynical.
It doesn’t matter what are all the solutions of string theory or how big the landscape is. Some of the solutions might not even have spacetime phases. But we know -by experiment, ha!- that WE live in a locally flat spacetime where scattering is possible and in such a phase stringiness MUST manifest at some scale below the Planck scale. Details depend on the compactification, but the generic prediction is enough as a smoking gun signature of stringiness (exponential as opposed to power law fall-offs in cross sections with momenta).
I do expect the target to move momentarily, but ‘nuf said from my side.
Ive long since abandoned any attempts to explain Peter Woit’s misleading and/or false statements by anything besides a dishonest attempt to sell books and promote himself.
I think the word “multiverse” can be very troublesome. To me the only sensible definition for the word is the space of possible histories that might be postulated to exist under the many worlds interpretation, which is by definition a metaphysical, untestable idea (not to suggest that we shouldn’t wonder about it). If instead it means something that is connected with our world in any direct way, then clearly that is part of the same “universe”, by definition.
Sean is right to urge caution in declaring a class of questions unscientific. Something is only unscientific if it cannot be decided *in principle*, and many questions are very hard and unlikely to be answered, but not unanswerable in principle. Woit is now creating an awful lot of fog by dishonestly trying to confuse the truly metaphysical questions with the hard, scientific questions of whether string theory is correct, and if so what is the true structure of our world…
Isn’t it true that theorists haven’t been able to use QCD to calculate the proton’s mass? From my understanding, this is due to the failure of perturbative QCD and only non-perturbative QCD (which we don’t know very well) can allow us to calculate the proton’s mass. Isn’t this similar to our lack of knowledge of string theory non-perturbatively?
Maybe if we understand string theory non-perturbatively, we will be able to find our low energy universe in the theory and rule out the rest of the landscape? But we don’t know if such a thing is possible, hence the continuing research in string theory.
Also, one can argue that string theory is NOT a TOE because it is only a perturbative theory of strings that postulates what the degrees of freedom are beyond the standard model and towards the Planck scale. Perhaps M-theory, or whatever the nonperturbative formulation of string theory is, is the real TOE.
If QCD is the theory of the strong interactions, why haven’t we been able to use the theory to calculate the proton mass? Because we don’t understand the theory very well non-perturbatively. Perturbative QCD, like perturbative string theory, cannot explain everything (i.e the proton mass), but a more fundamental formulation (i.e. non-perturbative QCD) can. The same may be true with string theory.
Ah, this brings back the good old days of spending huge amounts of my time trying to answer anonymous people devoted to characterizing me as a money-grubbing dishonest guy who had no idea what he was talking about. Fun times…
The best response to this is to just refer people to the old discussions, since they’re still on-line. Try for instance
http://blogs.discovermagazine.com/cosmicvariance/2005/07/21/two-cheers-for-string-theory/
and note that in comment 56 Sean agrees with me that the argument about string scattering amplitudes being a definitive prediction of string theory doesn’t work. There’s this thing called M-theory….
Peter, #65
There are many quantum field theories possible, corresponding to the many possible choices one can make for the underlying gauge groups. Similarly, there are many possible perturbative string theories, corresponding to the many possible ways of compactifying C-Y manifolds, etc. So what is the difference between these two?
The only difference I see is that we are able to perform experiments that show us the way towards the gauge groups of the standard model, but we do not yet have the technology to conduct experiments that show us what physics looks like far beyond the standard model or near the Planck scale. When string theorists get better at building stringy models that match the standard model they will have many models that not only contain the standard model, but also have different possibilities for what lies beyond the standard model and near the Planck scale. When or if our technology improves to the point which allows us to conduct much higher energy experiments, nature will guide us to the correct model, just like how nature guided us to the correct SU(3)XSU(2)XU(1) quantum field theory after we performed lots of particle physics experiments.
Do you agree with the above?
If it is shown that no stringy model contains the standard model (which hasn’t been done yet) or the results of beyond standard model experiment, then string theory will be shown to be wrong. If such a thing is not shown, then string theorists will keep on looking.
The idea of atoms was proposed over 2000 years ago. It had a ” testable” prediction, namely, that matter cannot be divided indefinitely. Nevertheless, in the absence of any viable experimental method to test it, the atomic hypothesis could hardly be called science, till relatively recently, historically speaking. Perhaps string theory’s prospects are a little better. Perhaps.
Peter,
You say QFT is predictive. Does QFT predict the SU(3) X SU(2) X U(1) group structure as the one that describes our world? No. Experiments were needed to elucidate this fact about our low energy world. Likewise, experiments are needed to elucidate the correct compactification for the extra dimensions, assuming extra dimensions and string theory describe our world. Again, experiments are needed to determine whether or not string theory and compactified extra dimensions describe our world, just as experiments were necessary to show us that group theory and the formalism of quantum field theory describe our world.
So, in that sense, QFT is just as unpredictive as perturbative string theory.
Do you agree with this? Why or why not?
Phil,
I completely agree, and this is a point that I do not think is made enough. Allow me to elaborate a little on your very nice point.
String theory and QFT are both frameworks for physics which are broad and apply in many scenarios.
A string vacuum or a quantum field theory are examples within their respective frameworks. Mathematical data is needed to specify both, and each makes predictions. Each string vacuum, for example, tells you unambiguously the gauge symmetry and matter content of the theory. If it’s SU(18) with a whole bunch of fundamentals, this string vacuum has made a prediction and it is not our world.
Other vacua may look a whole lot more like the real world. In my view, the problem isn’t whether a string vacuum makes predictions, it’s that it’s extremely difficult or impossible to determine which vacuum in the landscape is ours.
Pingback: The Eternally Existing, Self-Reproducing, Frequently Puzzling Inflationary Universe | Cosmic Variance | Theoretical Physics
When Newton gave us his 3 laws of motion, they were also just a framework for dynamics. The laws of force had to be specified separately. Newton had a law of force for gravitation, and perhaps Hooke’s law. I suppose he did not even attempt an explanation for the gravitational constant. Yet we call his laws predictive, because they could be applied to phenomena.
The same holds for quantum field theory, which is a productive framework, in exactly the same sense as Newton’s laws. The contrast of these two frameworks with string theory is stark.
I also think that the philosophy of science has undergone a huge change, not necessarily for the better. E.g., Mendeleev found the periodic table of elements in 1869 (in my analogy, akin to the Standard Model of particles). It provided a firm indication of an underlying structure, which however, was not figured out for decades. I think Moseley’s law and Bohr’s atomic model, both dating from 1913 were essential ingredients. The span of time of the Standard Model (say 1968-2011) is still less than that of the mysterious periodic table of elements (minimally, 1869-1913). But the chemists didn’t fall into anthropic reasoning or decide that the atomic weights/numbers arose from a landscape. Perhaps it is because they continued to have useful work to do, unlike the particle physicists of today?
In retrospect, during any particular era of the history of science, there were certain questions that were fruitful to ask, and a great many other questions that led nowhere – though they may have been answered in some future time. The first hard problem in the practice of science, in my opinion, is figuring out which the fruitful questions are. I think if we see ourselves flying off the rails, it may be a sign that we are not asking the questions that are right for our era.
Phil (#66, #68),
Hopefully I’m paraphrasing your point correctly. Given that both QFT and string theory can both be thought of as general frameworks that are compatible with a huge number of possible theories/solutions, observational input is needed to choose the one(s) that is/are physically relevant, why should ST be regarded as lacking predictivity, even though QFT is similarly non-predictive without the experimental input that singles out the physically correct QFTs?
I think it comes down to two words: track record. Consider QED, the most accurate physical theory ever found. As you imply, empirical evidence led to its development, and in fact was instrumental in developing the QFT framework in the first place. Given the QED Lagrangian, the QFT framework makes a large number of very accurate predictions which people take as evidence for the correctness of QFT more generally. This success has been followed by finding QFTs for the standard model and QCD, giving further evidence that QFT is a physically correct framework.
In fact, it is irrelevant whether experiments were crucial to finding the correct QFTs — once they were found, they made a huge number of accurate predictions that were in no way implied by the experimental input that originally led to them. That’s real predictivity.
Now consider string theory. One can argue that, like choosing physically relevant QFTs the space of all possible QFTs, experimental input can similarly help choose the physically relevant vacua from the space of all stringy vacua; hence. And that is true, but does it justify putting ST and QFT on similar footing? For now, I’d rather disregard arguments along the lines that there are so many candidate vacua (e.g., the 10^100 to 10^500 that get bandied about) that if you have found one vacuum that matches our world there may be a really huge number that also match our world within the precision of our experiments so that there is no way to operationally choose the correct vacuum and predict new results that weren’t used to select the ‘correct’ vacuum in the first place. That argument is probably beside the point of your question. The key thing is simply that not even one vacuum has been found yet that leads to something like our world, so it is not possible to use the string theory framework to take that vacuum and make predictions that can be checked against new experiments. That means that string theory has not been ‘battle tested,’ in very stark contrast to QFT which has been battle tested to an extraordinary degree.
Note that this comparison of QFT and ST is purely in terms of earned credibility. String theory can make many claims about explanatory power, but it has no track record at all in making accurate predictions. String theory may ‘really’ be the correct theory of nature in a philosophical sense, and the argument I gave won’t change at all. It’s like going to a well-respected doctor with a long track record of dealing with an ailment you have, versus going to a new doctor who uses flashy advertising in the local newspaper but who has no experience with your ailment: which would you choose?
Hi Arun and Marty,
I have to disagree with you both a bit, though perhaps I agree a little more with Marty.
Arun: you missed my point. Any given quantum field theory does in fact make predictions, as you point out, in analogy with Newton’s laws. Quantum field theory as a whole, which is a framework, not a theory, does not. Similarly – though you think the contrast to string theory is stark (it is not) – any given string vacuum makes predictions: it gives rise to a field theory at low energies which is often a gauge theory. Furthermore, in some sense it is MORE predictive than field theory: it predicts the gauge symmetry and representation content of the theory. As I stated, the problem isn’t predictivity of an individual vacuum, it’s separating our world out from the landscape of possibilities.
Marty: You’re point is a better one, I think. One necessary correction is that vacua have certainly been found that look something like our world, in many different contexts in string theory. See the recent work on F-theory GUTs, or slightly older work in type II orientifold compactifications or the heterotic string on orbifolds or smooth Calabi-Yau manifolds with holomorphic vector bundles. A lot of the posts in this thread come across as rants (not yours, though) from people who seem out of touch with the current status of string vacua for the purposes of unification. It’s not that there aren’t problems – there are – it’s just that they’re not the ones a lot of people are ranting about. Just thought I’d mention this correction.
You’re right to bring up the effectiveness of particular effective field theories in describing the world, in particular the importance of experiment: the experiments told us what terms to write down in the standard model Lagrangian. The issue isn’t quite as serious as you make it though, since string vacua contain field theories at low energies. If I may build on your point a little – the problem is that it’s very hard to distinguish field theory from string theory at low energies, so given new experimental input (help, LHC!) it would be difficult to say whether it’s string theory or just field theory. The typical things distinguishing between field theory and string theory – e.g. Kaluza-Klein modes or stringy excitations – typically come in at high scales. So I agree, touching low-energy experiments and distinguishing between FT and ST is difficult.
One interesting point, to me, at least, that should be mentioned. In FT, gauge theory is an input. The experiments told us what terms to put into our Lagrangian, and it turns out they’re gauge invariant. That’s a lot of structure – gauge theory! – that has been put into FT by hand. In string theory, this beautiful structure is output of the framework, not input.
J
“Furthermore, in some sense it is MORE predictive than field theory”
Well, there certainly is one huge difference between string theory and QFT, the level of hype…
One can try and make a serious argument that string theory is still not understood well-enough, that future fundamental progress in the subject will turn things around and lead to something that is predictive and competitive with QFT. But, claiming now that string theory is in the same situation or “MORE predictive” than the most successful theory ever developed by physicists is to be very, very far gone into the realm of sophistry.
I appreciate J’s point. Its one I often try to make, namely that according to Woit’s criteria, quantum field theory would itself be “unscientific” or “failed” or whatever if you imagined an analogous scenario where we had developed the general framework but had not figured out a way to determine the relevant gauge groups, etc.
But this obviously isn’t an A to B comparison. Besides its proven success, QFT operates at energy scales directly amenable to tests at particle accelerators. The main reason why I often find it hard to imagine that Woit is genuinely honest is that every serious scientist that looks at this knows that the experimental difficulties are guaranteed by the Planck scale. Anyone who promotes the idea that extreme difficulties in finding potent experiments isn’t generic to the entire project of understanding quantum gravity has either dubious honesty or dubious physics knowledge in my opinion.
Maybe we will find some potent evidence in coming years one way or the other, maybe not. But I think its ridiculous to simply throw stones at the entire project of understanding quantum gravity, and its ridiculous to expect that physicists just abandon the most promising framework for attaining that understanding without much more concrete, convincing reasons (ideally in the format of scientific papers rather than in the popular blogosphere). I dont think there is any rational reason to expect that all answers to the ultimate theory of quantum gravity could be determined without Planck-scale scattering data, but that is apparently the standard by which string theory has “failed”. Its quite simply BS.