Upcoming events in and around New York, including several I’m planning to attend:
The New York Academy of Sciences is having an evening of lectures this Wednesday, hosted by Frank Wilczek, on the topic of Expanding Frontiers of Physics and Cosmology. Speakers will be Max Tegmark and Nima Arkani-Hamed.
The YITP at Stony Brook is having a symposium to celebrate its 40th anniversary, and many former students, faculty and postdocs will be in attendance. I plan to definitely spend Thursday out there, maybe also Saturday.
One reason I likely won’t be out at Stony Brook on Friday is that I’d like to attend at least some talks at another event that will be downtown at the new location of the New York Academy of Sciences. It’s the 9th Northeast String Cosmology Meeting, co-sponsored by Columbia’s ISCAP. Edward Witten will be among the four people speaking.
There will be an event entitled When the Scientist Becomes the Story at NYU next week, on May 8th, featuring a discussion about John Nash and Francis Crick with their biographers.
Much farther in the future will be next year’s program on representation theory, algebraic geometry and physics at the mathematics division of the IAS in Princeton. This will include a conference November 26-30 with a title reflecting my favorite topic “Gauge Theory and Representation Theory”. Presumably much of the focus will be on the Geometric Langlands program.
Closer in time, but farther in distance, I’ll be speaking at a science festival called FEST in Trieste on May 18th. In June my book is supposed to be coming out in an Italian edition. I have to be in London the evening of May 23rd, then will head back to New York the next day. Currently trying to come up with a plan for how to spend the time in between, with the leading possibility a train trip through the Alps to Geneva, then a stop in Paris on the way to London.
In other news:
Lee Smolin has put up on his web-site a response to the review of his book and mine by Joe Polchinski.
On the Fields Medalist blogging front, there’s a report from Terry Tao about a symposium at UCLA where he and three other Fields medalists gave talks. He gives a detailed description of the talks, including one by Richard Borcherds on QFT that sounds somewhat mystifying to me. Alain Connes at his blog gives his take on some of the talks delivered at the recent conference in his honor.
I’ve recently for no particular reason run into various interesting domain-names that some mathematicians and physicists are using for one purpose or another: monodromy.com, cohomology.com, and stringvacua.org.
A couple links mentioned by commenters here that deserve more visibility:
Neutrino Unbound is a site devoted to all things neutrino.
An interesting document concerning a bet made several years ago about whether supersymmetry will be found at currently (or soon-to-be) accessible energies is available here. Maybe someone can think of a way to get more particle theorists on the record about this…
Update: For upcoming events really far afield from here, I should mention that the new Kavli Institute for Theoretical Physics in Beijing is starting to get organized. Jonathan Shock reports that there will be an opening ceremony at the end of May, a two month program on Quantum Phases of Matter starting in June, and a program on String Theory and Cosmology in the fall.
Update: I’ve just heard that Discover Magazine has chosen the finalists in its “String Theory in Two Minutes or Less” contest. No, I didn’t enter. Here they are.
Peter,
Thanks for pointing out Polchinski’s review and Smolin’s response. This is dialog that actually is instructive – and which was unfortunately missing from some earlier reviews of your book and Smolin’s by ST er’s.
If the protagonists will concentrate on the physics, as I think you and Lee did, and as Joe seems to have, the debate not only becomes more informative but more pleasant to observe.
CIP,
I think both Lee and I have been surprised at how little serious, substantive public reaction our books have gotten from string theorists, with Polchinski one of the welcome exceptions. I certainly have found that, in private discussions, most string theorists and I find lots of common ground, and that we agree about much more than we disagree. The kind of fanatical response that one sees on various blogs I don’t think reflects the views of many serious string theorists, although it does provide evidence of some of the sociological problems afflicting the subject.
Just a short request: if your stop in Paris becomes reality, and if some public event is planned for it, please don’t forget to mention the details here so that one can attend. Thanks.
Peter said:
“Maybe someone can think of a way to get more particle theorists on the record about this…” (supersymmetry).
Suppose we speak our piece here? Supersymmetry will never be found. It was one of those crazy ideas that wasn’t crazy enough.
My prediction: Low energy supersymmetry will be discovered within the next two years, as well as the Higgs. Huge triumph for the theoretical physics of the last thirty years. Certain people will be humbled.
About that Smolin’s response, you still have to admit that those top string theorists (like Greene or Kaku) are still damn good at public relations and selling string theory to the public.
I wish I had as good speaking and presentation skills as say Brian Greene..
From Smolin’s response thing:
the technical argument I had in mind… begins with the fact that supersymmetry requires that the cosmological constant be zero or negative
…wait… what?
Is this true? I’d never heard anything about this, and it seems like if it were true it would be a big enough deal to have come up before now– since there are indeed still people who expect we will see supersymmetry at the LHC, yet we’ve got this Dark Energy thing now that at is at least analogous to the positive cosmological constant. Right?
If this isn’t too big a question to ask within a a blog comment: Why does supersymmetry require that the cosmological constant be nonpositive, if indeed it does?
The condition is for unbroken supersymmetry. It arises because the only term in the SUGRA lagrangian that can give rise to a cosmological constant is negative definite. Once you break SUSY, however, you can have positive contributions and are able to get a positive cc.
That helps a lot, thanks.
Coin,
There’s more to it than what Aaron writes. If you want more of the story and some references to follow, it should be possible to extract the physics from the vitriolic attacks by multiple string theorists on me in this thread:
http://www.math.columbia.edu/~woit/wordpress/?p=454
I once saw a quote by Witten in APS News, where he said that the discovery of the negative cc was the most disturbing fact he had ever learned. My impression is that it is not so much the magnitude as the sign which is troubling; de Sitter space apparently lack the right kind of asymptotic regions where you can put holographic data, cf hep-th/0106109. You can almost hear from its name that AdS/CFT will run into problems if spacetime is de Sitter rather than anti-ditto.
Peter:
Re: String Theory Videos
I looked at a couple of the videos and they finally inspired me to ask your opinion of the following.
Assuming ST to be correct for sake of argument are particles really strings or do the physics suggest that all interactions can be modelled AS IF the particles were strings?
This has always bugged me. What is the reality? Vibrating strings, tension, total unobservable properties. I suppose that this is really a big question on what constitutes physical reality, but don’t physical theories attempt to model reality not actually describe what reality is?
In this same vein when the string is quantized it leads to the requirement of more 4 dimensions. Is there more than one way to quantize a system or is quantization unique? Do you know of any work that attempts to avoid quantization of a classical system and just start right away with an already defined quantized system? Thanks for the feedback.
You’ve got to live with it: the world is NOT holographic — it is truly 4-dimensional. QCD is NOT holographic, it is NOT topological. Standard model is NOT a topological QFT.
Welcome to the world of real dynamics where you have to work harder than just to count instantons and everything is fixed by some analycity constraints. I wished we paid more attention to the failure of (bootstrap) S-matrix theory. Actually, the most interesting non-perturbative results in QCD were obtained long ago, just after it replaced the S-matrix approach to strong interactions. Later, holomorphic SUSY took the bandwagon on another funride, diverting attention from serious dynamical problems: the bootstrap lesson was quickly forgotten.
Cecil,
People can have different philosophical viewpoints, but to me what is physically “real” at a fundamental level is whatever are the fundamental elements of our most successful physical model. Right now, that’s the standard model, so certain quantum fields are the fundamental things. If string theory were right, it would be something different, depending on what was the successful string theory model (e.g. it might be quantized fields on the space of strings, or matrices if you believe some versions of M-theory, or something else).
It’s somewhat of a mystery that our most successful quantum theories are formulated by “quantizing” a classical system. One can come up with quantum theories that don’t arise in this way, the quantum mechanical formalism is inherently more general.
Cecil,
Very good question. I think the idea of the elementary particles being extended objects is supposed to model the fact that the spacetime geometry breaks down at the Planck length. It may very well be the case that quantiziing a classical string is accurately modeling reality, provided that the different ways that it may vibrate are restricted, e.g. through compactification of extra dimensions. That is to say that the extra dimensions and their compactifed geometry aren’t actually ‘real’, but rather a mathematical device required to model the ‘true’ reality with a classical string. In this view, their will be one and only one compactification that corresponds to the true theory, while the landscape of string vacua is a mathematical artifact left over from starting with a classical object.
Cecil,
Quantization is far from being unique. One can very easily get contradictory results by choosing the “wrong” Poisson brackets to turn into quantum commutators. What rescues the situation is the restrictions on the space of possible quantum solutions owing to group-theoretical considerations. For example, the basic Poisson bracket to commutator substitution for [x, p] follows from the notion that the Q.M. momentum operator is the generator for space displacements – Planck’s constant merely being a scaling factor determined by our unit of classical momentum; similarly for time displacements and rotations to get the Q.M. Hamiltonian and angular momentum operator.
For quantum fields the basic Poisson bracket → commutator quantization fails for all but the simplest case – hence (e.g.) the need for Gupta-Bleuler quantization for helicity 1. But once again the situation is rescued by group theory – in this case the need to produce a spectrum of physical states that are an irreducible unitary representation of the Poincare group.
If you are interested, have a look at these notes on my web site that develop QFT without mentioning quantization.
Thomas,
In mentioning Witten’s remark, did you mean to write “discovery of the positive cc”? See the exchange between Coin and Aaron Bergman.
V,
Would it be fair to say, given the viewpoint you sketched in response to Cecil, that the classical string might eventually be understood as a classical image of something embedded in spacetime, which arises from some more primitive “pregeometric” object in a dynamical substrate that generates spacetime as a kind of mean-field classical approximation?
Maybe this should become a central element in the effort to push string theory forward, instead of ongoing elaboration on differential geometry and its abstract connections with other areas of mathematics, in which questions of ontology (and the early misgivings of Green and Schwarz) have been largely pushed aside as irrelevant.
Peter:
We try to approximate aspects of reality with what our senses and instruments detect and measure. We appreciate that, at best, the measurements can only be incomplete approximations, subject to measurement noise and (in spite of our best efforts), possible undetected bias.
These observations are used to try to generate self-consistent models of that reality. And we use further observations to determine the closeness of match between the models’ predictions and what we can access of that reality.
Against this background, it’s difficult to understand “what is physically ‘real’ at a fundamental level is whatever are the fundamental elements of our most successful physical model”.
Unobservable singularities and infinities, at least temporarily, would become part of your reality. That reality is a compromise of the moment (and might not easily provide motivation for going beyond present observation).
It may be the best we presently can do. (And I appreciate the compelling and amazing matches between QED predictions and magnitudes like that of the fine structure constant.) But QED still ‘only’ a model.
The Platonist believes that in the limit, the ideal model will ultimately provide more perfect descriptions of reality than our senses and instruments can ever provide; and that theoretical physics will be indistinguishable from a branch of mathematics..except for its ‘content’ of ‘physical reality’.
Your ‘definition’ above implies that you are a Platonist. This seems to conflict with my understanding of your usual positions in this blog…and in your book!?
Chris,
Yes, I’m trying to say something similiar to this. Most string theorists would say that M-theory will be such a formulation. One of the striking things about perturbative string theory is that the strings are quantized, but the spacetime they live in is purely classical.
Leonard,
Sure, I’m a Platonist, even often describing my views as radically Platonist. Many people don’t seem to have read my book very carefully, a large part of it is about the relation of mathematics and physics, claiming that they are very deeply connected, and it ends by suggesting more attention to this.
We know the world through our senses and through experiment. These are how we know that our mathematical models of physics have to do with “reality”, and are not just random abstract structures. I’ve always made clear that I don’t object to string theory on the grounds that many people do, that it is “too mathematical”. My objections to it are two-fold:
1. When you try and use string theory to make a model of reality, you are forced into huge amounts of complexity and ugliness.
2. String theory is inherently unpredictive, there is zero experimental evidence for it.
I see 1. and 2. as closely related problems.
Dear Peter,
I have to take objection with your first point that string theories models are complex and ugly. On the contrary, models in Type II theory involving intersecting D-branes are very geometric and beautiful, and the standard model is very naturally accomadated. Regarding your second point, perhaps there is some validity to this at present. However, this may not always be the case as further progress is made.
Hi Peter,
since bets on supersymmetry are mentioned here, let me mention that I had a more substantive bet open on my site last year.
Jacques Distler and Gordon Watts have bet respectively 750$ and 250$ with me that new physics (SUSY, or something else beyond a SM Higgs) will be found at the LHC within two years from having collected 10/fb per experiment. If SUSY is found, I lose $1000. If
nothing is found, I win $1000 with which I can try to cheer up my saddened soul.
Cheers,
T.
V.,
I’ll just quote Zwiebach, page 346, about these models:
“the models seem contrived, at least in the sense that they are engineered to give the physics that we observe, rather than obtained naturally as the simplest solutions to string theory. The Standard Model is an intricate construct, and present-day attempts to describe it within string theory are not simpler.”
Smolin mentions in his reply to Polchinski’s review that N=4 SYM has not been rigorously constructed yet. I wonder (and this is certainly the most stupid question ever asked in this blog) whether N=4 SYM is asymptotically free?
Nameless,
The N=4 SYM beta-function is zero, so the coupling constant doesn’t run, there’s an ultraviolet fixed point at non-zero coupling. It’s not asymptotically free, since the coupling doesn’t go to zero.
One of the big problems with using string-gauge duality to understand QCD is exactly this: you don’t get asymptotically free theories like QCD.
Peter,
If you’d ever get your hands dirty with real work, you’d actually have a better sense of these models, rather than relying on a comment in Zweibach’s book. There are many, many constraints that one must satisfy in order to build realistic models, so that it very definitely not contrived. See hep-th/0612087 for a beautiful example.
V,
Sorry, but that looks completely contrived to me. Then again, it may look less contrived to those doing the contriving….
And even in the conformal case, only the large N limit, I gather?
Thanks for your reply!
Peter,
I doubt very seriously that you have the knowledge and experience in either string theory or particle theory to make such judgments. In any case, you will claim everything is contrived since it suits your purposes, regardless of the reality. I would have some respect for you if had something postiive and constructive to say, rather than just continually nagging.
V.,
You’ve chosen to remain publicly anonymous, while making it clear to me who you are. I really don’t think you’re in a good position to attack my or anyone else’s credentials, and in any case that’s not something you have any business doing from behind the cover of anonymity. Please cut it out.
Wow V, the number of matter generations in that beautiful noncontrived model only differs from reality by 33%. You’ve made a believer out of me.
V wrote “There are many, many constraints that one must satisfy in order to build realistic models, so that it very definitely not contrived.” I want to understand the logic implied by this statement. Are you saying (where “we” just means “people who understand string theory”) “We understand genuine principles of physics (which aren’t just “we need to end up with a realistic model” conditions in disguise) that give lots of constraints. If we build a model which satisfies those constriaints, we get a realistic model” or “We know various things must be constrained to get a realistic model. If we build a model satisfying those constraints, we get a realistic model”? I’m just trying to understand precisely what is being claimed.
Peter:
By no means have I let you know who I am anonymously. You can make whatever assumptions you like. I point to this model as one that I’m familiar with which is very simple, yet from it comes the entire structure of the MSSM. All essentially from a set of numbers which are all plus/minus one.
kuos:
How do you know there won’t turn out to be four generations, with the fourth generation being heavy? In any case, is this not an experimental prediction?
Dave:
In constructing models, there are many nontrivial consistency conditions which must be satisfied such as tadpole cancellation, conditions for supersymmetry, K-theory charge cancellation, etc.. Never mind getting the right matter spectrum.
And even in the conformal case, only the large N limit, I gather?
No. The AdS/CFT conjecture in its most powerful form holds for finite N. Gauge/geometry duality is also not only restricted to CFTs.
V, I’m not a physicist so I don’t understand what those terms mean in detail, and in the details you seem to have avoided commenting one way or another on whether these conditions are _primarily understood on their own terms_, or just understood as necessary in order to avoid producing features in concrete instances of the model which are not realistic. To put it another way, suppose tomorrow we were to observe something that implied, eg, “incomplete tadpole cancellation” happened in the universe (whatever that might mean), presumably the fact that it’s a consistency condition of string theory means no variety of string theory would be unable to model this, right? (I’m just trying to clear up what seem like vagueness about which conditions are put in to ensure a match the observed world — and could be removed if experimentally falsifed — and which are logically inviolable parts of the mathematics.)
Ugh: double negative is just a typo, should be “no variety of string theory would be able to model this”.
Peter wrote:
One of the big problems with using string-gauge duality to understand QCD is exactly this: you don’t get asymptotically free theories like QCD.
Aaron wrote:
Gauge/geometry duality is also not only restricted to CFTs.
It might be worth expanding on these two statements for the benefit of the reader who doesn’t know about gauge/gravity duality. I would say they are both correct. First, there are *non*conformal theories for which the duality is really under control. These are not asymptotically free theories, but theories which are strongly coupled even in the UV. The canonical example is Klebanov-Strassler, but there are many others.
For asymptotically free theories, the trouble is that the gravity side becomes strongly coupled when the gauge side is becoming weakly coupled. One can get interesting theories in the same universality class as gauge theories of interest; e.g. Witten found a supergravity black hole background in the universality class of nonsupersymmetric Yang-Mills. This background seems to have a tunable parameter that can be adjusted to reach the true non-SUSY Yang-Mills, but the problem is that in trying to take this limit one runs into a regime where calculations cannot be done. Similar almost-dual backgrounds exist for theories with flavor (Sakai-Sugimoto, etc.).
In that sense, gauge/gravity duality provides something much like the lattice strong-coupling expansion: a way of calculating in a theory very much like the theory of interest, weakly coupled where the dual is strongly coupled, with a limit one can take to reach the theory of interest (and no phase transition when taking that limit). In both cases the desired limit is not tractable, but qualitatively most of the properties of the theories are similar. In the case of the lattice, the desired limit actually can be done (with difficulty) on a computer, at least for calculating simple enough quantities. There’s no analogue of that for the gravitational dual.
Third post: in case it’s not clear, I’m not looking for an explanation of any of these things, just a clear statement “we have most of the model conditions as an unavoidable requirements of string theory” or “most of the model conditions are motivated by the need to obtain a realistic model”, just in order to understand what level of claim for these sorts of string theory model is being made.
Thanks, Aaron.
Dave,
The requirements that I mentioned are basically for internal consistency at the quantum level, and these requirements must be satisfied apart from any considerations of matching the model to observed physics. For example, if the tadpoles are not cancelled there will be what’s call a triangle anomaly. Once these conditions are satisfied, one would like to have the gauge group and matter representations of the MSSM, a full set of Yukawa couplings (necessary for generating mass), and gauge coupling unification.
In intersecting brane models, this is accomplished by having different stacks of D-branes which wrap around the compactified dimensions. Every time the different stacks intersect, there is chiral matter localized at the intersection. The idea is to have a configuration which gives the standard model. However, as mentioned above, it is not possible to have any possible configuration, as this is seriously constrained by the need for internal consistency. The model that I pointed out has basically the simplest possible configuration, where the different D-branes wrap around each compactified direction exactly once.
V.
One of the reasons some people get frustrated with string theory
as a fundamental description of the universe is that arbitrary
scenarios are invoked. Intersecting D-branes are of academic
interest for constructing models with particles of different colors and flavors; that doesn’t mean that this is how quantum numbers arise in
nature.
The problem is that there are no dynamical calculations in string
theory. No dynamical calculation shows that there is Calabi-Yau (or any other, including KK) compactification. No dynamical calculation shows that certain brane configurations dominate (including the one you advocate). This is what I always found unsatisfying about attempts to do string phenomenolgy, even back in the eighties – there are plenty of scenarios, but no analysis. Landscape advocates go even further in this respect. They have openly given up on justifying these scenarios.
Peter,
You’re right that at present there are no dynamics that would choose the specific vacuum presently known. However, I think this is because we don’t have the complete theory. If we can find a vacuum which correspond very closely to our universe, we may be able understand why this vacuum is selected, which may in turn lead to a complete formulation of the theory. In any case, it would be remarkable if we can find a model which provides a complete description of physics as we know it, and it would be even more remarkable if this model made predicitions that were later born out.
At the present, we don’t know why the Standard Model gauge group is what it is. Afterall, a large number of gauge groups are possible and it has many free parameters. Does that stop particle physicists from using it? Absolutely not! However, some people like Peter Woit would claim that the Standard Model is ‘contrived’ and arbitrary. It may very well be, but that does not mean it’s useless or that it doesn’t help us get closer to the truth.
It will probably turn out that we will find a string theory model that completely describes our universe which may be regarded as an effective theory just as the standard model is an effective theory. However, it will be at a deeper level. Only later we will understand why it is chosen.
In mentioning Witten’s remark, did you mean to write “discovery of the positive cc”?
Oops.
V.
I don’t think the Standard Model is contrived, quite the opposite. I think all attempts to get it out of string theory are contrived.
Anon, thanks to you too. I’ll refrain from asking further questions… that’s what review articles are for.
Peter,
String theory models are no more contrived than the Standard Model itself. The standard model can only make predictions after fixing it’s free parameters by hand. We have no reason for choosing SU(3)xSU(2)xU(1) in the SM except that it seems to work. We have no explanation for the charges in the SM except they seem to work. We have no explanation for the masses in the SM except they seem to work. Any string theory model will essentially do the same thing, except that it will provide a geometric explanation of all of the above AND unify the SM with gravity. Perhaps we will not be able to understand why this structure arises for some time, but it will still be useful, especially if it can make other predictions.
There are 19 parameters in the standard model. Assuming each of these on average needs around 4 significant digits to get similar physics to ours gives one chance in 10 ^ (-84) we would have a universe looking fairly close to this one. This chance appears to be significantly better than you get with the string theory landscape.
Is this a good measure of the “contrivedness” of a theory? I don’t think so, but I don’t buy the argument that string theory has no free parameters, so it’s less contrived.
I think you guys would be a lot better off if you’d stop worrying about the ability to uniquely determine the universe we live in. We’re clearly not there, yet. The goal should be to effectively describe our world in a predictive way, regardless of the uniqueness question. Making aesthetic judgements is rather stupid at this point. If it works, that’s what’s important. We can worry about the ‘why’ question later.
V, if you require anomaly cancellation and that the global gauge group of the SM is S(U(3)xU(2))=SU(3)xSU(2)xU(1)/Z_6, you get a unique solution for the hypercharges (and hence the electric charges), assuming anomalies cancel out within a family (the “Poor Man’s Unified Field Theory”). The SU(2) and SU(3) reps of the quarks and leptons are inputs though.