Thursday, November 23, 2006

More on AdS/CFT and RHIC

Last week, Pavel Kovtun gave a seminar here on PI about the AdS/CFT correspondence and its applications for RHIC physics. For a brief introduction, see e.g. our earlier post Does String Theory explain Heavy Ion Physics?. It was the same seminar that I heard at the KITP in spring, and on which I reported in the Banana-post. If you are interested in the topic, you can now download audio, video and slides from the PI streaming seminars:

On Tuesday, I missed another seminar on the subject, but as the seminar schedule tells me, on Thursday there will be even another seminar by Andrei Starinets.

It's not that I am so tremendously interested in the topic, I just find it remarkable how much fuss there is around it. Despite the fact that Pavel started his talk with a very nice motivation about RHIC experiments, I could not avoid noticing the sharp contrast between his predictions, and the predictions I am used to from nuclear physics talks. The latter of which usually include some plots of calculated observables, how well they fit the actual data points (and error bars to both if necessary). Based on this, a conclusion on the quality of the model should be given, and how they compare to other approaches. It seems to me that the AdS/CFT calculations haven't yet quite reached this state.

I have wondered for some while how the importance of the results is seen from the nuclear physicist's community, which has gathered last week in Shanghai on the Quark Matter 2006. The Quark Matter is the largest annual meeting of the community, which I am very sorry to have missed this year. But interestingly, yesterday a friend (who kindly agreed that I post his email) reported the following on the closing talk:

----- Original Message -----
From: "******* *******"
To: [...] sabine[@]perimeterinstitute.ca
Sent: Tuesday, November 21, 2006 9:00 PM
Subject: Tearing string theory a new a*hole


This is pretty weird/interesting.
We just concluded quark matter. At the theory summary, in what was supposed to be the closing session, the guy speaking (Larry McLerran, a pretty famous person) went on a 20 minute rant on string theory in general, attempts to use it to describe RHIC in particular, and Brian Greene in personal.


See
Larry McLerran's talk (PPT), slides 19 onwards for what I am talking about (the slides do not convey his tone, e.g. giving Brian Greene a "Pinocchio award").

[...]

In any case, an amusing closing talk to conclude a somewhat politicized (no surprise) but fun (as always) conference.


[...]


------End of Original Message ---------


The powerpoint file is pretty large (11MB), so here are the last some slides as jpgs (click to enlarge), starting with the Pinocchio award for Brian Greene:











Please don't ask me details about the talk, as I mentioned, I didn't hear it. This is just to convey some skepticism from 'the other side'.

Update: See also Clifford's post Nuclear Guy goes Nuclear at Asymptotia, and Lubos' post Heavy Ion Physics and AdS/QCD.

Update: A written version of the summary talk is now available on the arxiv

Theory Summary: Quark Matter 2006
Authors: Larry McLerran
hep-ph/0702004

74 comments:

Anonymous said...

I would like to repeat the following questions, already appeared among the comments of your previous post about RHIC and Ads/CFT. The slides by McLerran give sharp answers.


True QCD (3 colors, no SUSY) is approximated with a gauge theory with many colors and 4 supersymmetries.
Why this approximation should be good, at high density? 

How accurate should it be? (100%? 30%? 1%?). 

How accurately has it been tested?

Arun said...

Bee,

Just the statement "AdsCFT MUST be accountable to the same scientific standards as are other computations" and its implications by itself is going to cause a brouhaha. I foresee many more storms in teacups, debates on the merits of "crackpot" vs. "Pinocchio", and questions will be raised about Larry McLerran's IQ.

I hope something creative also comes out of all this.

Bee said...

Hi Anonymous,

yes, I find it very helpful if accuracy is quantified through a number, such that it is easily possible to compare to other approaches. It seems to me, the approximation looks useful in a small region of the phase diagram, possibly somewhere around the RHIC dot.

Hi Arun,

I understand your concern. I certainly did not write the post with the intention to cause more bubbles of nothing. Instead what I hope to get across is that there is more communication necessary between string theorists persuing the AdS/CFT approach to RHIC physics and the heavy-ion community. This is a process that has only begun, one that I welcome very much, but one that shouldn't stop with putting the word 'RHIC' in a title of a talk.

I am not surprised by the split reaction of the heavy ion guys (I've heard criticism like the above not only from McLarren - Stefan is about to write some more about it). They are of course defending their models, and before they believe that AdS/CFT is a cool thing, it will have to compete with the established approaches. This was to expect, and I actually think it's a reasonable requirement.

On the other hand, I've not heard one single heavy-ion guy saying this is uninteresting, or nonsense. It's more like (in my impression) they are definitely interested, but also kind of frustrated because the usefulness of AdS/CFT isn't yet really good quantified (and, let me be honest, its not always really obvious what the string theorists are talking about in their seminars).

Without going into the question of what 'scientific standards' are to apply, it is certainly true that there is some gap between AdS/CFT and 'other computations' on RHIC data. This gap can be closed if both sides are aware that there is work to be done. What I don't want is AdS/CFT calculations going wild without applications, only using RHIC physics as an alibi.

Best,

B.

Anonymous said...

it would be great if somebody could explain the meaning of the point about the "difference between strongly interacting and strong coupling" in slides 21 and 22

Lumo said...

The pictures and insults by Larry MaLarren are very funny. He's good at it, and I am sure that Brian Greene will appreciate it, too. ;-)

The scientific analysis of AdS/QCD of McLarren is not as good. I am no infinite lover of AdS/QCD but McLarren's criticism is clearly pointless.

McLarren et al. have studied these problems for decades without any really new ideas, new concepts, and new zeroth approximations, and they started to believe that any new ideas are just bad. A typical example of a field that got stuck.

When he criticizes that it is N=4, what can we say. N=4 is the most beautiful gauge theory in 4D. It is not identical to QCD but it doesn't mean that it can't be a better starting point to understand any kind of physics of gauge theories than for example pure gauge theory.

These old people seem unwilling to understand that if something has a shorter Lagrangian, it is not necessary a more natural and useful zeroth approximation to understand physics of a system.

He says that in AdS/QCD we have no running coupling. Well, this is a piece of demagogy, too. N=4 of course has no running coupling because it is conformal. But more generally, the holographic approach of course has running couplings and its description in terms of diff. equations that depend on the holographic coordinates represent a large portion of the papers written in the subject.

Nevertheless, it may still be true that for many purposes, the conformal theory may be a better starting point to understand some physical phenomena than other starting points. McLarren simply wants to sell the converse statement as a dogma which is not exactly something I would call a rational approach.

He just doesn't seem to like that someone has found something more improtant than he has, namely holography and the gravitational description of gauge theories which is undoubtedly more valuable than all models of effective quarks and bags that have ever been written down.

Then he shows a picture how accurately string theory computes the density and pressure, more accurately than perturbative QCD, of course, not too surprising because this is quite a stringy regime.

Then he criticizes string theory that it "requires [strong] coupling limit". Well, that's the point of it that it wants to expand physics around strong 't Hooft coupling, not the weak one. But he is wrong that it means that one can never get anything that would involve weak coupling. For N=4, the duality holds at any coupling.

For the less constrained lower-SUSY theories, it is harder to determine exactly what's the physics at weaker coupling but there's no reason why it should be impossible. Just like perturbative QCD expands around weak coupling, there exist - at least in some highly symmetric cases - expansions around the strong coupling, via the power of duality.

The situation is analogous at weak coupling and strong coupling. This analogy hasn't been appreciated before the mid 1990s, but it is appreciated today, except for McLarren who will probably never appreciate either because he is too slow or because he simply doesn't like the idea that someone discovered something more important in his field than he has.

Then he says a lot of the cliches popular among crackpots today about equal scientific standards even though the equal standards for different theories are exactly what he wants to avoid.

The AdS description of strong interactions is a work in progress, and yes, it could need new ingredients as one goes further from the N=4 picture and closer to the real QCD and weaker coupling. It is a part of the fun that it is not yet fully known. But that's how physics has worked throughout most of its history: the theories were first incompletely known.

Let me say that I personally have no doubt that QCD is the fully correct description of the strong interactions - and from this fundamental viewpoint, we are finished in QCD. But written in its normal Lagrangian form, it doesn't have to be the best starting point to understand its physics in many complicated contexts. Physics is not just about writing a Lagrangian: it is about understanding real phenomena.

Whether or not Dr. McLarren finds an approximation controllable or uncontrollable is much less important than the question whether the physics in some regimes exhibits the concepts of the dual string theory, and be sure that it does. String theory gives predictions that would be hard to obtain otherwise.

This dual stringy picture allows the AdS/QCD researchers to make new predictions about other phenomena, and of course that it is falsifiable. But no one who is doing the work seriously is thinking about an "ultimate" falsification of the whole framework that would convince one to commit harakiri if a prediction doesn't work. When people had gotten bad predictions from some effective quark theories, they didn't abandon effective quark theories either, and McLarren is just dishonest if he suggests otherwise.

If a prediction doesn't work, one must look at all possible ways to design a new theory or fix the errors in the previous theory. The most convincing theory resulting from this constant sequence of corrections and competitions between possible new steps eventually gets the most attention. And yes, it doesn't have to be the same theory as the most straightforward naive frameworks that McLarren has had 25 years ago and hasn't changed them much since that time.

McLarren vs AdS/QCD is a classic textbook example of the tension between new exciting scientific findings and intellectually reactionary physicists who don't want any new progress.

And that's the memo.

Lumo said...

Otherwise, I don't think, Bee, that you're right that you're conveying criticism from the other side because you are clearly on the same side as these critics which is why you're constantly promoting their intellectually vacuous criticism and their not-quite-intelligent books, among many other things.

Why are you trying to paint that you are on a different side than you actually are?

Let me also say that your statement that "AdS/QCD has not reached the same point [as conventional nuclear physics]" just shows that you have some deeply held anti-theoretical-physics sentiments.

The goal in AdS/QCD is much more general and grandiose than to understand an excited state of one particular dirty nucleus or something like that. Yes, it is more theoretical and it is studied by more theoretical people than the conventional nuclear physics.

Normal theoretical physicists would tell you that the generic nuclear physicists haven't yet reached the point at which they could say something more general and important about reality than the properties of one particular nucleus or the behavior in one particular strongly coupled situation.

It is a matter of priorities, and I wonder why you chose theoretical physics if you value unstimulating analyses of a particular nucleus to be more important than the conceptual questions such as the nature and universal properties of gravity, gauge forces, and holography.

Bee said...

Dear Lubos,

He just doesn't seem to like that someone has found something more improtant than he has, namely holography and the gravitational description of gauge theories which is undoubtedly more valuable than all models of effective quarks and bags that have ever been written down.

Well, I don't want to psychoanalyse McLarren, therefore let me just say that this is a very subjective observation. Some might find effective models that allow to describe experimental facts more attractive than so far completely unverified ideas, however attractive they are. This is a very applied approach to a description of nature, and it's good that we are not all interested in the same stuff.

But no one who is doing the work seriously is thinking about an "ultimate" falsification of the whole framework that would convince one to commit harakiri if a prediction doesn't work.

I am very relieved to hear that.

Why are you trying to paint that you are on a different side than you actually are? [...] It is a matter of priorities, and I wonder why you chose theoretical physics if you value unstimulating analyses of a particular nucleus to be more important than the conceptual questions such as the nature and universal properties of gravity, gauge forces, and holography.


A) You have just insulted all theoretical nuclear physicists by calling their work generally 'unimportant' and 'unstimulating'. As always, you are impressively decent.

B) There is a reason why I did not stay in nuclear physics. The reason being that I am more interested in conceptual questions.

C) I am on the side of theoretical physics.

D) Who says what has priority?

Best,

B.

Anonymous said...

Lubos,

as a phenomenological model, AdS/QCD will probably turn out to be somewhat worse than conventional models and somewhat better than naive dimensional analysis.

You say that it is much deeper, but maybe N=4 SYM can be exactly solved just because it is a toy gauge theory: some time ago theorists found exact 3 body solutions to Newton equations with planets of equal mass at the vertices of an equilateral triangles. It was nothing deep, but at least those theorists did not insult who was testing the orbits of real planets (like the Earth and Mercury) with approximate techniques.

Lumo said...

Dear Bee, you write:

"Some might find effective models that allow to describe experimental facts more attractive than so far completely unverified ideas, however attractive they are."

Some might find anything more attractive but you don't want to see the more important part of such a possibility, namely that these people have no clue about theoretical physics. Theoretical physics is not a pile of subjective opinions, as you seem to think. Theoretical physics is the research of very objective properties of reality.

AdS/CFT is not an unverified idea. It is a very rigorous equivalence verified roughly in 5,000 articles, and even the application of its concepts on physics of QCD is documented by roughly hundred(s) of articles. Your suggestion that a random effective description of some particular nuclei could be more important than holography - just if we allow some subjective feelings in it, right - is just so incredibly lacking any sense of reality that it would probably be a good idea to follow Witten's recommendations and not respond. But you know I can't ever resist. ;-)

Have you noticed, for example, that Maldacena's correspondence, the original paper, has 4765 citations, after less than 10 years? Have you ever seen any of them? Can you compare this number with the 239 citations of the most cited quark-bag model paper that is almost 30 years old? Have you ever been brave enough to ask youself whether it could be that it is you (together with your other friends-critics of theoretical physics), not me or the theoretical physicists in general, who is brutally mis-evaluating reality?

Do I really have to be hiding that the AdS/CFT correspondence is, by orders of magnitude, more important than most frameworks to study physics of QCD approximately that have ever been written down, and to think that saying that it is equally or even less important is nothing else than a symptom of a complete misunderstanding what theoretical physics is all about?

I am no AdS/CFT partisan and in some sense, I still think that things like Matrix Theory can turn out to be more important, but saying that it is less important than a particular effective model for a nucleus is just incredibly silly.

What you're saying is just like saying that general relativity is less important than a particular model of aether because general relativity was only testable by 1 phenomenon around 1910s - and is testable on 6 phenomena or so today - while the aether was built from very material gears and wheels which is surely more attractive science. Your suggestion is just too far away from something that could lead to a rational discussion among scientists.

I don't care whether you think it is undecent. If a nuclear physicist who has studied any particular nucleus thinks that her work is as important or as stimulating as AdS/CFT, then she is simply mad. I have nothing whatsoever against any other fields of physics, but even with the existence of other fields, one must be able to compare the importance of results, at least approximately, across different fields. There has been no result in nuclear physics of the last 30 years that could match the AdS/CFT correspondence.

It would be extremely dangerous to hide this fact. Do you really want to impose the atmosphere of political correctness in which one can't even say that a paper with 1000 citations a year is most likely more important than a paper with 10 citations per year?

Do you really believe the same conspiracy theories as the bizarre critics of physics whom you keep on promoting all the time?

You are not on the side of theoretical physics. Just compare the amount of space you give to theoretical physics with the huge space you are giving to its critics. Even in this article, you only offer a few links to a rather precious review of Pavel Kovtun about AdS/QCD. On the other hand, you fully repost several silly content-free slides of a critic, probably because you find it more important, and despite the manifest fact that it is completely obvious that Kovtun's talk is a sequence of very serious and nontrivial technical results while the Larry someone slides are just non-technical emotional insults, you still have the stomach to say that AdS/QCD is not yet at the technical level of nuclear physics.

Your lack of sense of reality is just unbelievable to me.

Best
Lubos

Lumo said...

Dear Anonymous (3 comments above this one),

the precise form of the AdS/QCD model is not yet settled - and it's not guaranteed whether it ever will be. It is a work in progress and the model is only partially known, and in many contexts, approximations that are known to be inaccurate are used (e.g. the N=4 theory instead of QCD).

This is why your comment how successful it will be in the future is a completely irrelevant (and vicious) speculation that every rational person discards immediately.

This direction is studied because it is based on arguably the most important insight of theoretical physics of the last 10 years, and because the initial steps in applying the methods of AdS/CFT to the real strongly coupled physics of QCD have been unexpectedly encouraging.

So people will surely continue to study it with all possible new ideas and approaches they will see and it is rather likely that the percentage of the AdS/QCD activity within the strongly coupled nuclear phenomenology will probably keep on increasing for some time, whether you like it or not.

Best
Lubos

Lumo said...

I forgot to answer the second paragraph of the same anonymous commenter: whether or not the N=4 gauge theory is solvable is an entirely different question. And of course, if it *is* solvable, then this theory itself and its application to any problem in reality is even more important than before we knew that it was solvable.

A particular solution of Newton's equation is not important, but be sure that the N=4 gauge theory in d=4 is completely essential. It is as essential for theoretical physics as other, much simpler models such as the harmonic oscillator.

In the previous sentence, much like in your phrase "toy model", we heavily underestimate how complex the N=4 theory actually is. Look at some real papers about it: they're very complex, no real toys.

It is a complex theory but it is also a very constrained theory whose many properties we know very well, whose many properties are general for all gauge theories, and whose other properties are directly applicable to some nuclear physics phenomena.

It is this combination of theoretical depth and phenomenological relevance that will guarantee that people will continue to study the N=4 theory in hundreds of future papers, whether their motivation is purely theoretical or phenomenological in character.

If someone thinks otherwise, she just misunderstands the structure of ideas in this portion of theoretical physics.

Arun said...

Anonymous, a dense collection of weakly coupled particles may be strongly interacting, as in e.g., a particle's momentum is very rapidly effectively randomized. That is what I understood.

Bee, looking at the rest of the comments I can only :) and hope you think it is worthwhile.

stefan said...

Dear Bee,

thank you for providing the screenshots, OpenOffice running at my Mac didn't show the Pinocchio cartoon...

About the talk, I think that first of all, one should keep in mind that the AdS/QCD part was just the last quarter (unfortunately, I wasn't in Shanghai either - more eyewitnesses' accounts would be helpful.

Quark Matter Conferences are indeed, usually quite vivid, and especially the theory summary talk has a tradition of being a bit caustic. I remember Berndt Müllers summary talk at QM 99 in Torino, where he made some comments about the merits (or the lack thereof) of ever new simulation models, which left some PhD students a little bit in despair - I was to naive at that time to care much...

Anyway, just to mention that, Larry McLerran was the head of BNL's nuclear theory division, and currently is theory group leader of the RIKEN BNL Research Center.

As mentioned in the AdS/CFT-RHIC post, he is one of the inventors and main proponents of the Colour Glass Condensate picture for the hot and dense initial state of heavy-ion collisions at RHIC. The colour glass initial state creates quite a large initial pressure, such that without taking into account viscosity, the famous signal of elliptic flow as a function of transverse momentum would clearly overshoot the data. That means that the early state may have a viscosity much higher than that of the the viscosity bound conjecture, so that the question of the most ideal fluid may not be settled yet. Discussions about hydrodynamics simulations of heavy-ion collisions including viscosity seem to have been a hot topic at Quark Matter in Shanghai.

So, there are lots of things happening in heavy-ion theory, which may be not as fancy as the holographic principle, but there is progress. I can imagine that people who have done lot of hard, solid work are not happy with hegemonic declarations that string theory can explain things better than they do, even more so when such claims are still shaky. Rob Pisarski, one of the most reactionary physicists I can think of, ended the introduction of a talk in Frankfurt this summer about the perturbative calculations on the right hand side of the second slide Bee has posted by dumping two string theory books in the paper-bin - I couldn't see which books that have been, but he made his point.

To be a little more constructive, I would like to point out two recent papers (I had done it before on Clifford's blog) related to the discussion on the the viscosity bound and subtleties of the strongly coupled/strongly interacting QGP, which I found very interesting and readable:

On the Strongly-Interacting Low-Viscosity Matter Created in Relativistic Nuclear Collisions, by Laszlo P. Csernai, Joe Kapusta, and Larry McLerran (nucl-th/0604032) presents a new and fancy way to plot the phase diagram of fluids in the vicinity of the critical point: They plot the ration of shear viscosity to entropy density as a function of temperature for different pressures. These curves seem to look quite similar for all kinds of fluids, with a kink when hitting the critical point. Moreover, this kink corresponds to the absolute minimum of eta/s for the fluid under discussion, and it is always higher than the AdS/CFT bound, which seem to be quite universal indeed! This paper was published as a PRL in october, and it seems to have been directly motivated by the Kovtun-Son-Starinets PRL. However, it does not (if I did not miss that point somehow) discuss whether AdS/CFT says anything more specific about RHIC matter or not.

And, related to that, there is a strange side remark on the first of the Larry McLerran slides Bee has posted, which I do not quite understand yet: Is it true that the viscosity bound can be derived from the mean free path and the deBroglie wavelength alone? Maybe I should read the papers of Csernai, Kapusta, and McLerran more carefully ;-). But if there is a simple derivation that could have been done 80 years ago, it is quite ironic that it was re-discovered in 2004 or so using the heavy artillery of string theory ;-)

The second paper is The Letter "s" (and the sQGP), by Jamie Nagle (nucl-th/0608070). It gives a nice discussion of all the different notions of strong interactions and strong coupling involved in the debate about the hot and dense nuclear matter created at RHIC, which is a little bit confusing. The point is that with electromagnetic plasmas, one usually has weak coupling (in the particle physicists' sense that the fine structure constant doesn't run much and is low and close to 1/137, even with extreme conditions such as in the interiors of stars), but strong interaction, nevertheless. Strong interaction here means a high ratio of potential energy to kinetic energy of the particles in the plasma. Unfortunately, plasma physicist call this latter property strong coupling... It seems to me that the heavy-ion physics community is just starting learning a lot of things from plasma physics, for example exploring plasma instabilities and things like that...

Finally, I would like to comment about calling heavy-ion theory as nuclear theory. This has, in my ears, quite a strange note. Nuclear theory is, for me, juggling around with endlessly long Lagrangeans which include all kinds of phenomenological forces, with the goal to derive, using long computer code, some excitation spectra of nuclei, or to identify the island of stability. With all due respect to the Frankfurt colleagues who have embarked deeply on such calculations, to me this was allways a little dull. Heavy-ion theory, on the other hand, uses lots of crazy models and incorporates ideas from all areas of physics, including particle physics, plasma physics, thermodynamics, or hydrodynamics. So it is not getting boring soon ;-)

So, happy Thanksgiving to all American readers! (There is no Thanksgiving in Canada?)

I won't do anything else than go to bed to cure a insetting cold...


Best, stefan

Lumo said...

Dear Anonymous author of the first comment in this thread,

the real QCD has 3 colors and the N=4 theory can also have 3 colors. We normally use methods that are applicable at a large number of colors, in the 1/N expansion.

Note that 1/N is 1/3 for N=3 which makes it unsurprising that the approximation by N=infinity has been very successful: 1/3 is close to zero which is why we may expand around zero. Many quantities such as the pressure and the density in the graphs are accurate with 1% accuracy or so, even before considering 3 colors.

The N=4 theory has scalars and fermions in the adjoint representation; on the other hand, it is missing the quark flavors in the fundamental representation. These things obviously make difference in some regimes, and a more accurate theory is needed for these situations.

Nevertheless, there are situations, and they cover a large portion of the QGP/RHIC phenomena, in which these differences don't cause significant discrepancies. The physics is really dominated by the pure gauge theory sector and the addition of all other matter makes a very small difference.

The statement from the previous sentence seems to be true and people who study these things want to understand more precisely how much true it is, why it is true, and how the theory must be refined if this assumption breaks down. It's a part of their work. It's part of the physics that everyone in that field should have asked a long time ago. Some people have asked how much the quarks really matter in different situations but before AdS/QCD, there were no good tools to start answering such questions.

Some people, like that Larry, would like to transform these properties and differences to a reason not to study AdS/QCD which is of course not how science works. It doesn't matter that Dr. Larry doesn't like the N=4 large-number-of-colors theory as the starting point to describe nuclear physics. What is more important that it works for understanding of many qualitative things, and with a correct refinement, it could be used to describe an ever greater set of phenomena ever more accurately.

This whole field is about understanding real physical phenomena; the fundamental QCD Lagrangian is known. So the approaches are obviously judged by their efficiency in getting the right physics quickly and accurately and AdS/QCD is doing very well in this competition.

Best
Lubos

Lumo said...

Dear Stefan,

as you could have said, when I wrote "nuclear theory", I really meant "nuclear theory" and it was never a typo; I didn't mean "heavy ion physics" in either of these sentences.

My feelings about this field of nuclear physics are analogous to your feelings. Long Hamiltonians, a lot of technicalities, but shallow general results.

I know that the Gentleman was a boss in Berkeley. We have heard about the bad relations between high-energy theorists and non-high-energy (or) non-theorists in Berkeley many times. It's too bad, and needless to say, these bad relations are the low-energy physicists' or non-theorists' fault.

All the best
Lubos

Lumo said...

One more comment, Bee, concerning your remarks at Asymptotia.

You say "of course they are defending their models". Are you serious that this should be "of course"?

How can a scientist "of course" defend his model? If there is a better model, a scientist will accept a better model regardless who created it, and virtually all real physicists I know of - from all fields - almost perfectly satisfy this assumption.

Do you really disagree with me, or was your comment just a joke I didn't get?

You say one more thing: "and, let me be honest, its not always really obvious what you string theorists are talking about in your seminars".

Let me be honest, haven't you thought about the possibility that it may usually be your fault, not the speaker's fault?

Bee said...

Dear Lubos,

Thank you for your comments.

I will ignore your personal accusations as they are as usual besides the point.

Let me just say that I am not too fond of the heavy-ion community myself. I've had my problems with their 'scientific standards', and their eagerness to fitfitfit some data with whatever parametrization isn't a game I find very appealing. I've always been more intrigued by the elementary questions. But, maybe in contrast to you, I wouldn't say that a branch of theoretical physics that aims to characterize the properties of nuclear matter is just generally unimportant.

AdS/CFT is not an unverified idea. It is a very rigorous equivalence verified roughly in 5,000 articles,

I used the word 'unverified' with the meaning it's not verified through experimental facts that it is actually realized in nature.

You are not on the side of theoretical physics. Just compare the amount of space you give to theoretical physics with the huge space you are giving to its critics. Even in this article, you only offer a few links to a rather precious review of Pavel Kovtun about AdS/QCD. On the other hand, you fully repost several silly content-free slides of a critic, probably because you find it more important,

Are you trying to indicate I give too much space to some people's comments? *Hummm* You may be right with that.

Regarding my use of website space: I've written a summary about Pavel's talk which excited me very much at the time, and whose talk I really liked. I also learned a lot from his patient explanations. Also, we've written a longer post about the puzzling RHIC results for the viscosity, and it's possible relation to AdS/CFT. Both of these posts come with references, and the above post has links to them.

The reason why I put the jpgs in the above post was simply that it took me a rather long time to download the stupid PPT, after which it wouldn't open properly, etc.

you still have the stomach to say that AdS/QCD is not yet at the technical level of nuclear physics.

I certainly didn't say anything about a 'technical level' of nuclear physics, as I'm not entirely sure this is a desirable level to reach. What I said is that even though those pursuing AdS/CFT applications to RHIC physics aim to talk about the same experiments as the heavy-ion people, they aren't yet very efficiently communicating. This is destined to cause conflict like the above, which I think is unnecessary, not to mention a waste of time.

How can a scientist "of course" defend his model? If there is a better model, a scientist will accept a better model regardless who created it

Yes, indeed, but you'll have to proof that it is better. Unless you quantify that some model is undoubtedly better with predictions than some other one, they will try to defend their models on other grounds. The 'of course' above referred to the fact that some people just consider themselves generally to be of utmost importance.

You say one more thing: "and, let me be honest, its not always really obvious what you string theorists are talking about in your seminars".

Let me be honest, haven't you thought about the possibility that it may usually be your fault, not the speaker's fault?


I have, and I admit that there are many things in this world I'll probably never understand.

The reason for this remark was: If string theorists find it desirable to work on heavy ion physics, it is worth some effort to make the more involved sides accesible to heavy-ion physicist with little or no knowlegde about string theory, or they risk provoking silly criticism.

Best,

B.

PS: Talking to some friends and reading comments on Cliffords blog, I found that most seem to associate the Pinocchio award to 'liar'. My first association was actually a string puppet.

Bee said...

So, happy Thanksgiving to all American readers! (There is no Thanksgiving in Canada?)

There is, but it was already 6 weeks ago. If you'd read my blog, you'd have known...

Arun said...

Bee,
I too took the the Pinocchio award to mean one for lying or spinning.

Looking around on the web I see such usage

1. on freerepublic.com for the Democratic party spinmeisters.

2. in a review of a movie for a wooden performance by an actor.

3. by a motorbiker newsletter for poor riding

4. for a Catholic who dreamt that he burned his hand on the outer walls of hell - (for being inflammable like Pinocchio)

5. for various famous news anchors for spinning/lying


6. what the jury awarded Francis Coppola in his suit against Warner Bros. in a dispute about his movie "Pinocchio".

7. British political stuff where the Pinocchio award is for misrepresenting the opponents' position

8. for a newspaper for not mentioning the chief culprits in a story about pollution

9. a newspaper naming "a member of the community who tells the most whopping lie heard between January and December."

10. Conde Nast to the publisher who misses his estimates the most but in the right direction.

11. For people who poke their (long) noses into other people's business

But mostly it is used to refer to champion falsehoods.

Plato said...

Maybe, most do not know of the superfluids "relativistic nature?"

This form of matter is called quark-gluon plasma or QGP. Like its name suggests, QGP is a "soup", or plasma, of quarks and gluons.

Remember "Campbell's soup" and Horowitz. Not Broccolli :)

HDH said...

Hi Arun! Bee said... My first association was actually a string puppet. I guess that's what they mean with independent thinking *lol*. !

Bee said...

Hi hdh,

harhar, very funny.

Remark to Larry McLarren's talk:

The comment "mean free path must be bigger than DeBroglie wavelenght" probably refers to this paper:

Dissipative phenomena in quark-gluon plasmas
P. Danielewicz and M. Gyulassy


Section III A.

Best,

B.

PS: Poor Pinnochio. All that people seem to remember is that his nose grew when he was lying. His only dream was to be human.

Lumo said...

Dear Bee,

I don't think that you have given an explanation that would indicate anything less than, as you wrote, scientists should be defending their models because they are their models instead of preferring the working models. I view this opinion of yours to be very sick.

"I used the word 'unverified' with the meaning it's not verified through experimental facts that it is actually realized in nature."

This is a nonsensical assertion. AdS/CFT is a statement about the character of mathematics that describes quantum gravity, and if you agree that it is an essentially proven one, then it follows that it is realized in Nature much like 2+2=4 is realized in Nature.

Concerning nuclear physics, I didn't write it was generally unimportant. I wrote that particular questions about some concrete bound states it studies are by far less stimulating and important than the AdS/CFT correspondence and no discovery in nuclear physics in the last 30 years comes any close to the AdS/CFT as far as importance for physics and depth go.

"Are you trying to indicate I give too much space to some people's comments?"

I think that I have written very clearly - and everyone knew it before I wrote it - that you give far too much space to unqualified and silly comments of people who have nothing to say about a physical question or a set of questions. By orders of magnitude.

The reason for this remark was: If string theorists find it desirable to work on heavy ion physics, it is worth some effort to make the more involved sides accesible to heavy-ion physicist with little or no knowlegde about string theory, or they risk provoking silly criticism.

I don't think it is a primary job for scientists to do such things, and I don't think it is in fact possible. It is certainly impossible for every member of the heavy ion physics community to understand every technical aspect of the AdS/QCD papers, and it would be foolish to pretend otherwise. No clique of researchers has any right to claim "ownership of science" which is another craziness you seem to be defending.

If the AdS/QCD people get more important results and descriptions of these phenomena and if they become better than the old guys, and it is very likely, this is just a fact of life and no one in the AdS/QCD community is obliged to wait to all old members of the heavy ion community to understand. Do you really disagree with me?

Yes, indeed, but you'll have to proof that it is better. Unless you quantify that some model is undoubtedly better with predictions than some other one, they will try to defend their models on other grounds. The 'of course' above referred to the fact that some people just consider themselves generally to be of utmost importance.

Our ideas what science should be like are clearly absolutely incompatible. I think that a scientist who "will try to defend their models on other ground" has insufficient moral standards to be a scientist. It is a huge accusation against those people in heavy ion physics, and I don't believe it is a correct description in most cases.

It is up to the scientist himself to try to find the most correct answers to the questions that he studies, regardless of the author of the ideas. Otherwise he is just not doing science. Just like the string theorists who study the phenomena using the AdS/QCD correspondence had to learn what the effects are and what the normal phenomenological description of these effects are, the other people in theoretical heavy ion physics must do the same thing and look at the insights obtained via AdS/QCD or otherwise because it's their responsibility, instead of pretending that they are important and the most important thing in their field is for others to teach them.

Your idea that the goal of a scientist is to defend some old opinions and others must be bringing him the blue from the skies in order to convince him is just completely sick. What you're describing are egotistic biased bigots, not scientists.

I am personally not interested in the opinion of XY about anything if I learn that XY deliberately twists his opinions in such a way that it becomes more difficult to convince him about certain questions that differ by having different authors. In science, it is up to every individual person to decide what models are correct and/or more useful.

Your "string puppet" association is complete nonsense. Pinocchio's nose has been used as a symbol of lies for quite some time.

Best
Lubos

Moshe said...

Re: The comment "mean free path must be bigger than DeBroglie wavelength":

The point in the ads calcualtion is not so much the bound but the fact that strongly coupled the bound is nearly saturated. There is no other analytical approach with this qualitative feature, or am I wrong?

Nevertheless, the bound is equivalent to the Bekenstein bound in the bulk. Which highlights another fact: this may be a way to learn about quantum gravity using experiment, which is one reason to be excited. Even if heavy ion people find it less than useful, quantum gravity people ought to be interested.

Similar comments apply to other criticism: it is true that the coupling is of order one, no apriori reason strong coupling expansion will be successful. Of course identical comments apply to weak coupling expansion which somehow is deemed more acceptable...also, how many controlled approximations are there in heavy ion physics? I'd be enlightened to learn about one.

And finally, regarding the excitement: we all love what we do, and I hope most people will continue to be shamelessly enthusiastic. It is essential to get some feedback and compare this approach to others etc. This is of course done in many collaborations recently. I am sure this will continue despite the occasional public display of unprofessional behavior.

Bee said...

Dear Lubos,

I am really very tired of your lengthy comments that complain about my 'sick opinions', my 'lack of sense of reality' and that I am allegedly 'constantly promoting [...] intellectually vacuous criticism'. If you don't like my blog, why don't you just shut up? The only thing that your comments show is that you are searching for a reason to disagree, for which it is apparently not even necessary to read what I write.

[...] as you wrote, scientists should be defending their models because they are their models instead of preferring the working models. I view this opinion of yours to be very sick.

I never said that. In case it didn't become clear: quantitative predictions should decide scientifically which model is the better one. I said that, in absence of such, people often retreat to defending their models on other grounds (like beauty and elegance, and some maybe present deep connection to the rest of the universe, just in case you miss the point again.) I never said that they should do that.

I think that a scientist who "will try to defend their models on other ground" has insufficient moral standards to be a scientist. It is a huge accusation against those people in heavy ion physics, and I don't believe it is a correct description in most cases.

Neither do I. In most cases I find scientists are fairly reasonable people. You say: I think that a scientist who "will try to defend their models on other ground" has insufficient moral standards to be a scientist. Then how about we just stop this discussion, and instead wait until string theorists explain the structure of nuclear matter to better accuracy than all other models people have worked on.

Our ideas what science should be like are clearly absolutely incompatible.

Indeed.

Your idea that the goal of a scientist is to defend some old opinions and others must be bringing him the blue from the skies in order to convince him is just completely sick. What you're describing are egotistic biased bigots, not scientists.

Unsurprisingly, I certainly never said that this is 'my idea' of 'the goal of a scientist'. It is just a fact that established, 'old' ideas are often defended by the people who have worked on them. If that has never happened in the spheres you lived in, you are a happy guy, and I wish you a pleasant ignorance for the rest of your career. Don't get stuck on old opinions.


In science, it is up to every individual person to decide what models are correct and/or more useful.

Well, I still think it is up to nature to decide which models are correct and/or more useful.

AdS/CFT is a statement about the character of mathematics that describes quantum gravity, and if you agree that it is an essentially proven one, then it follows that it is realized in Nature much like 2+2=4 is realized in Nature.

Are you trying to say that every 'statement about the character of mathematics' has a realization in nature? And even if the AdS/CFT description should successfully predict RHIC results to high accuracy, why would that imply it necessarily had to be also part of a theory that unifies quantum field theory with general relativity?

If the AdS/QCD people get more important results and descriptions of these phenomena and if they become better than the old guys, and it is very likely, this is just a fact of life and no one in the AdS/QCD community is obliged to wait to all old members of the heavy ion community to understand. Do you really disagree with me?

No. But what I am trying to say is that the 'AdS/QCD community' should try to communicate better with the heavy ion community, not all of which are old btw. Talking to people who have already been working on issues the 'AdS/QCD community' is facing now, can certainly lead to faster progress on both sides.

I don't think it is a primary job for scientists to do such things, and I don't think it is in fact possible. It is certainly impossible for every member of the heavy ion physics community to understand every technical aspect of the AdS/QCD papers, and it would be foolish to pretend otherwise. No clique of researchers has any right to claim "ownership of science" which is another craziness you seem to be defending.

I'm not saying every member should understand every technical aspect, but I think the gap should be as small as possible. I've never said anything about some 'ownership of science', this is another craziness you have made up.

Best,

B.

Bee said...

Dear Moshe:

And finally, regarding the excitement: we all love what we do, and I hope most people will continue to be shamelessly enthusiastic. It is essential to get some feedback and compare this approach to others etc. This is of course done in many collaborations recently. I am sure this will continue despite the occasional public display of unprofessional behavior.

This is what I also hope. Feedback and constructive criticism from both sides is definitly the way to progress.

Best,

B.

Lumo said...

Dear Bee,

the reason why I care what you write is that you keep on contaminating the internet by nonsense, Pinocchios, and all this garbage often appears as hits in the search engines and it is often misinterpreted by people interested in physics as information about physics even though it is just a piece of nonsense.

Best
Lubos

Cynthia said...

Bee, my humble opinion is this. It's mildly troubling for Larry McLerran to be tacky enough to suggest that Brian Greene deserves the "Pinocchio Award". On top of that however, it's intensely troubling that you would be so tacky as to *amplify*(keyword here) McLerran's suggestion that the "Pinocchio Award" ought to be delivered to Brian Greene.

Underneath it all though, there's a subtle, yet nontrivial difference between the two brands of tackiness. Luckily, McLerran's tackiness is narrowly confined to the [more than less] private realm of an esoteric physics conference. Regrettably, your tackiness - in contrast - is widely open to the [more than less] public realm of an exoteric physics blogosphere.

Anonymous said...

Bee

I found it very valuable that you pointed out these slides, which have been made public to all by being put on the web, and hence are not (as suggested by cynthia) located in a more or less private realm. Thank you for the link.

What I do not understand is why you provide Lubos with a platform for his rantings - both because of their extraordinary unpleasantness, and because they undoubtedly seriously damage the way string theory is viewed by other physicists. Why not just delete his comments? That will greatly improve the quality of the blogoshere.

Bee said...

Dear Cynthia,

Bee, my humble opinion is this. It's mildly troubling for Larry McLerran to be tacky enough to suggest that Brian Greene deserves the "Pinocchio Award". On top of that however, it's intensely troubling that you would be so tacky as to *amplify*(keyword here) McLerran's suggestion that the "Pinocchio Award" ought to be delivered to Brian Greene.

Thanks for the open words. I admit that I've meanwhile asked myself whether mentioning McLerran's talk was a good ida. I went to Pavel's talk last week, and wrote a brief note linking to the video. It was a coincidence that I received the above email about the Quark Matter exactly the same day.

The reason why I included the slides was that there's one thing I detest more than insulting others, that's insulting them behind their back. If they want to argue, fine. But one problem with these 'sub-communities' that have formed within theoretical physics is that a strong sense of belonging to one group hinders constructive criticism, and favours nonconstructive one.

I can easily imagine how tempting it is to make a joke about someone who is not present, knowing that nobody of the people in the room would defend the person not present, I can easily imagine how tempting it is to include remarks that cause the audience to laugh and make you look witty, no matter what the content of the joke. I can easily imagine what this does to the standard of scientific argumentation.


I neither want that to happen in a room somewhere in Santa Barbara, nor in a room somewhere in San Francisco, nor in a room somewhere in Shanghai. I neither want that to happen when the person being laughed about is a string theorist, a book author, or a heavy ion physicist.

I seriously hope that making it more public causes people to rethink the way they criticise others.

Best,

B.

Bee said...

Dear Anonymous from 2 comments earlier,

I found it very valuable that you pointed out these slides,

Thank you for the feedback.

What I do not understand is why you provide Lubos with a platform for his rantings - both because of their extraordinary unpleasantness, and because they undoubtedly seriously damage the way string theory is viewed by other physicists. Why not just delete his comments? That will greatly improve the quality of the blogoshere.

Because it happens every now and then that he has something interesting to say. I try to ignore the rest, not always successfully. He's just written a post on AdS/QCD which is actually quite nice. Besides this, I'm really hesitant to delete comments. At least he's not anonymously insulting people (these comments go in the trash immediately).

Do you really think that Lubos damages the reputation of string theory? I find that hard to believe. He's just a guy with a blog who is fast with the keyboard.

Best,

B.

NO said...

Cynthia,

"Regrettably, your tackiness - in contrast - is widely open to the [more than less] public realm of an exoteric physics blogosphere."

I don't see where Bee was tacky. Read her post. Besides, that is what blogging is about. Passing on information. Widely open.

Anonymous said...

wow I have never been into physics, but reading this blog sure makes me think I should get into it. is it true that they can make matter as it was in the beginning of the universe or even a new universe in the accellerator?

tompot

Arun said...

Bee,
This is for you, in case you haven't seen it already.

Anonymous said...

If the "is it true?" statement by Greene is not true and was made in public, it deserves a public Pinocchio award. Probably mine is a minority view, but I think that honesty in science is more important than public relations.

Plato said...

Arun,

I give a "little exercise" in time travel fiction?

Reading your link to Bee, how "illogical" is my link?

Why can't we used "closed loop processes" to define the state of the universe?

stefan said...

Hi Tompot,

I have never been into physics, but reading this blog sure makes me think I should get into it.

That's encouraging feedback, indeed! And it mainly goes to Bee, since my contributions have been a little scarce lately... But I hope your excitement is not fed mainly by the distribution of Pinocchio awards ;-)

Is it true that they can make matter as it was in the beginning of the universe

Well, indeed, the quark-gluon plasma as it is created in collisions of gold nuclei at RHIC is an extreme state of matter that has existed in the first microseconds of the universe!

That's why sometimes, people refer to these collisions at RHIC as "Little Bangs", or "Mini Bangs", in analogy to the "Big Bang". Larry McLerran uses this analogy on the title slide of his Shanghai talk, and you can read more about this whole topic in the May 2006 issue of the Scientific American. In case you want to have a look at some really technical description of the early universe and its relation to particle and heavy ion physics, you may try astro-ph/0303574: The first second of the Universe, by Bee's diploma thesis advisor.

Of course, the amount of matter recreated in this "primordial" state in heavy ion collisions is tiny - it corresponds just to several thousand particles and antiparticles, and the net amount of matter is just what went into the collision: two gold nuclei, i.e., two atoms, stripped bare of all the electrons.

But besides this difference in sheer quantity, there is also a difference in quality between the Big Bang and the Little Bangs. This is often not mentioned, and that can, I guess, cause some misunderstanding.

In the Big Bang, space itself is expanding, thereby diluting and cooling the matter it is filled with. In the Little Bangs of heavy-ion collisions, the hot and dense quark-gluon plasma expands into the surrounding Minkowski space-time, which is the fixed stage of the event and not influenced at all by the collision (no black holes at RHIC ;-). So, the Little Bang corresponds to the concept of the Big Bang as an explosion localized at a point in space and time, as in the Milne universe, which today is generally thought to be wrong.

Another important difference is the time-scale involved in both kinds of Bangs: Fast as it may seem, when hearing of microseconds, the Big Bang is extremely slow compared to heavy-ion collisions, slower by roughly a factor 10^13. A heavy-ion collision is essentially over in the time light needs to travel a distance equal to the diameter of a heavy atomic nucleus - some 10^(-14) meter. So, what happens in the very early universe is for sure in local thermal equilibrium, which means that it makes perfectly sense to speak of phases such as the quark-gluon plasma, and of phase transitions, such as the "condensation" into hadrons. However, for the Little Bangs, this should not be taken for granted from the beginning, because of both the much shorter time scales and the quite tiny volumes involved. However, as it comes out, thermal equilibration seems to work...

So, I hope you get a better picture of the analogy, and of its limits, and, well, see why

or even a new universe in the accelerator?

this is, in my opinion, a bold speculation not backed by well-established facts ;-)

Best, stefan

stefan said...

Hi Bee, Moshe:

thank you for your comments about the viscosity/entropy bound!


Dear Bee,

me: (There is no Thanksgiving in Canada?)

you: There is, but it was already 6 weeks ago. If you'd read my blog, you'd have known...

too bad for me - I even wrote the first comment on that post of yours... A weighty documentation of my preoccupation with other things lately ;-)...

and even worse: I also had completely forgotten that I had already posted the references to the Nagele and Csernai/Kapusta/McLerran papers in the comments of our common post... I vow improvement :-)

Best, stefan.

Plato said...

If you move relativity to quantum examples "it's got to work?" I answered Arun's link, by answering Godel?

or even a new universe in the accelerator?

So, if such attributes are considered in context of the colliders, then what use superfluity without understanding it's viscosity? Yes, we can make wide sweeping generalizations, and limit it too.....?

I do like having this "safety net of corrections though" and the possibility of not being lead astray. Yet, "Susskind's elephant thought experiment" has not be addressed by some, while, I think I am close? :)

Been down this route Stefan by past explorations. Thanks Stefan, and I too am encouraged by newcomers statements.

Anonymous said...

Thanks Stefan, since this git a little garbled let me expand: the viscosity to entropy ratio is low experimentally, close to the bound. In the N=4 theory it exactly saturates the bound. One can re-calculate this quantity for different theories with matter content (still with susy, but less than N=4), one finds that fact, of being close to the bound, to be universal and also to be related to deep facts about semiclassical gravity. Maybe it is related to real QCD, maybe not, but if I didn't get excited I'd have to go and check my pulse...

Moshe said...

Last comment was mine.

Anonymous said...

Bee, you ask

Do you really think that Lubos damages the reputation of string theory? I find that hard to believe. He's just a guy with a blog who is fast with the keyboard

No, he's the arrogant face of string theory (with a Harvard appointment, no less) who believes that all other physicists do worthless work, and tells them so to their face. As you yourself said

You have just insulted all theoretical nuclear physicists by calling their work generally 'unimportant' and 'unstimulating'.

That has to damage string theory - or at least its standing in the physics community. The basic assumption is that what he and his colleagues do is much more important than what any other physicist does - when it's not yet even clear yet that what he and his colleagues are doing is real physics, rather than an esoteric branch of mathematics. And of course this arrogance extends to blaming everyone else for the results their own behaviour:

needless to say, these bad relations are the low-energy physicists' or non-theorists' fault.

Wow! Motl is simply unable to behave in a professionally collegial manner. When string theorists want financial support for their potentially important work, do they want other physicists to support their requests to the funding agencies? Motl is certainly not helping their cause.

other anonymous said...

Lubos also actively and merrily promotes the idea that women are inferior to men and blacks are inferior to whites.

Sure he says IQ is lower and he doesn't mention inferiority vs superiority. But given the amount of esteem this guy puts on IQ. I think the point is clear.

Anyone who has had much contact with the guy knows that one of the biggest insults for him is to say a particular group of scientists or some individual has a low IQ.

That behavior and any endorsement of it is a stain on the theoretical physics (lets face it, nobody knows the difference between a string theorist and any other kind of theoretical physicist if they know what string theory is at all).

Bee said...

Folks, The point is: Lubos is not the face of string theory per se, and I don't care very much about his opinion. I wonder why you do. In addition, I don't think the IQ is a good measure for intelligence. Thanks to Stefan, I found out yesterday what probably annoyed him: Google for AdS/CFT and RHIC. I'm actually not sure how that happened so fast. Best,

B.

Bee said...

Dear Plato,

Thanks for the link to your very interesting warp drive post. Being human, it's certainly impossible that we approach a question completely without prejudices and totally 'tabula rasa'. Our past, our education, our knowledge, and personal interests will always play a role. But as scientists, we must at least try being as objective as possible. Giving room to the opinions of others - see: diversity in science - is an easy way to do this.

Best,

B.

Anonymous said...

thanks this is helpful. does it matter what's the matter, lol. why gold? was there gold in the early universe? what special about gold, why do they use it? can they make it to a black hole is this where the universe comes from. we came out of a black hole, i think i read somewhere. do we know how the universe ends. i understand from your other writing that the adscft does not make gravity so why is this like in the beginning?

Plato said...

Bee:But as scientists, we must at least try being as objective as possible. Giving room to the opinions of others

Of course. If I was to "invasive" please always let me know. Since diversity seems to be quite a "large area" of understanding, you did it in nice way.

see: diversity in science Perimeter Institute is a prime example for an excellently balanced and managed use of job-related diversity.

This is what I liked about PI institute and the leadership taken by Smolin and others to build the program that now exists.:)The Blackberry? :)

But the universe contains everything, including all its possible observers. This means that quantum theory must be extended (or re-invented) to allow observers to be part of the system they are observing. This is presently a lively area of research, to which scientists at PI have contributed important ideas and results.

Thanks,

Cynthia said...

Stefan, thanks for the link! As far as explaining physics to the lay public is concerned, I regard the writers for "Scientific American" as perhaps some of the best in the business.

Admittedly, you and Bee, likewise, are outstanding conveyors of physics to the, more or less, common folk. Once again, thank you both!

Best,Cynthia

Arun said...

"Thanks to Stefan, I found out yesterday what probably annoyed him: Google for AdS/CFT and RHIC. "

Sorry, there is no "smoking gun" on the first four pages of search results.

stefan said...

Hi tompot,

thanks for asking... at least to some of your questions, I can try to say something ,,,

does it matter what's the matter, lol. why gold? was there gold in the early universe? what special about gold, why do they use it?

No, there was no gold in the early universe, on the contrary, all the gold that exists today, according to the current understanding of the origin of the diverse elements, comes from supernova explosions, or even more exotic astrophysical processes. In the early universe, when matter was very dense and hot, there have been electrons, protons, and neutrons (which later formed the atomic nuclei of helium and some light elements, and then, with the electrons, the first atoms) and before that, in the first microseconds, quarks and gluons, the constituents of protons and neutrons.

Now, in a heavy-ion collision, what you are really interested in is to collide as large chunks of protons and neutrons as possible, in order to create a portion of quark-gluon plasma as large as possible. But the only chunks of protons and neutrons one can handle in the lab and the accelerator are the inner parts of atoms, the atomic nuclei.

So, you have a look at the periodic table of elements and figure out which elements have the largest numbers of protons and neutrons and are comparably cheap and easy to handle in the lab. The heaviest atoms (atoms with the nuclei with most neutrons and protons) are in row 7 of the periodic table, but they are all radioactive and you don't want to use them. The right-hand heavy mass end of row 6 has also still radioactive and/or hazardous elements as the now infamous polonium. The most massive and harmless elements are then: gold and lead!

At the CERN-SPS, the nuclei of lead atoms had been used for heavy-ion collisions, and at RHIC, they use gold nuclei.

Gold has one special property (besides its colour and prize ;-) which makes it especially suited for heavy-ion collisions: it occurs in nature in only one isotope, Au-197, so you know exactly how many protons (79, because it's gold) and how many neutrons (118, the difference to 197) you have.

can they make it to a black hole?

Well, definitely not. There are considerations about possible Mini-Black holes at the LHC, but at RHIC, no black holes are produced, although there have been speculations about this before it was turned on. There is this holographically motivated dual black hole description of the quark-gluon plasma, but that does not mean that a real black hole is created in RHIC collisions.

we came out of a black hole, i think i read somewhere.

I am not sure - one reads that sometimes, but I think it is metaphoric. Or it sounds a little bit like Lee Smolin's idea of evolving universes, but that's a speculation, I would say. Anyhow, I have no idea how this black hole metric would be mapped to the FLRW metric of the universe, and how everything would still be a solution of the Einstein field euqations?

so why is this like in the beginning?

The main thing that is like in the beginning is the phase of matter: matter so extremely hot and squeezed that there are not individual gold atoms, or gold nuclei, or even individual protons and neutrons, but just one soup of quarks and gluons instead. This stuff expands, cools down, and forms hadrons. That's what is in parallel. Everything else has to be taken with large grains of salt, I would say,

Best, stefan

Bee said...

My apologies for the missing pictures on this website today. The server in Frankfurt is down - again. Unfortunately, there's nothing I can do about it except cursing. This means also I don't get your comments forwarded by email, so I'm slow with replies.

Plato,

you are always 'invasive' :-), but we like you as you are: always good for astonishing insights.

Cynthia,

Thanks. Sometimes it's quite frustrating, so I appreciate to know it's worth the effort.

Stefan,

Thanks for the terrific explanations :-)

Best,

B.

Leucipo said...

Arun: the point is that the first entry in the search is to the criticism of AdS/CFT in Backreaction. Sometimes Lubos has been know to check on google first page hits.

Leucipo said...

all the gold that exists today, according to the current understanding of the origin of the diverse elements, comes from supernova explosions, or even more exotic astrophysical processes.

Well, an skillfull alchemist can do some gold by repeating separate distillation of mercury in order to isolate the isotope 201 Hg and then using the philosophers stone to induce its decay to 197 Pt and then beta to 197 Au.

I was considering this last time the oil price went up, but you only get about 80 kwh per gram of mercury. It is better than oil, but still worse than uranium reactors. So I will to heat my home with some home made classical uranium reactor instead.

Lacking the stone, you can use the

Bee said...

Hi Leucipio,

I understood Arun's comment as that the first some pages of hits are quite reasonable, and certainly not 'anti-string'. As far as I am concerned 'anti-string' is a non-existent state of mind anyhow, specifically invented to polarize opinions - I couldn't name a single person I'd call anti-string (I don't count anonymous commenters). The above search results luckily confirm this.

What I don't understand though is why this post is the first hit after only some days. I'd have understood if our last post Does String Theory Explain heavy Ion Physics was the first hit, but it's not even the second. Instead there appears the Banana post *gna*.

I'm not among those who try to figure out how google rates pages (see also googlebombing) to improve the own hits, but I guess one factor in this game is that I can't assign meta tags (keywords) to single posts.

Sometimes Lubos has been know to check on google first page hits.

Well, yes, as he wrote above:

you keep on contaminating the internet by nonsense, Pinocchios, and all this garbage often appears as hits in the search engines and it is often misinterpreted by people interested in physics as information about physics even though it is just a piece of nonsense.

But it's likely this ranking will vanish in the next some days. At least I hope so. I've definitely written better pieces.

Best,

B.

Leucipo said...

In the arxiv there are some articles about PageRank, and even one suggesting to use it instead of citation count to select hot papers: physics/0604130

By the way, the phrase in my last comment was cut, it was supposed to be:

"Lacking the stone, you can use the SNS."

referring to the spallation neutron source (www.sns.gov), another impressive piece of experimental particle physics.

Xin-Nian Wang said...

Hi All,

As an outsider (so far) of string theory community and an involved participant of Quark Matter 2006 -Shanghai ( I served on the International Advisory Committee of the past few QM conferences and helped local organizers of this meeting with the scientific program), I would like to provide some facts about the meeting and the Theory Summary (we call it Rapporteur Talk) that has spiked so much discussions. I hope these facts will give you a full picture of the meeting and the heavy-ion physics. I think they are useful for a meaningful and informed discussion, before you reach any conclusion based on prejudice or stereotype of what heavy-ion physics is really about (for a complete program of the meeting see: http://www.sinap.ac.cn/qm2006/index.html).

(1) The QM2006 in Shanghai had about 600 participants, typical of the conference size in recent years. The conference consisted of mainly invited plenary talks. There were close to 400 abstracts submitted and only over 100 were selected for 3x4 parallel sessions.

(2) There were many young physicists in the conference. Foreign students and young postdoctoral fellows outside China who received conference support numbered over 180 this year.

(3)The Theory Summary by Larry McLerran was one of the rapporteur talks which was to summarize theoretical presentations at the meeting. The speakers are free to inject his/her own comments. Critical comments on papers presented at the meeting in those rapporteur talks are very common and are essential for the health of the field, or any scientific endeavor.

(4)As a tradition in Quark Matter conferences, no comments or questions are allowed from the audience at the end of each rapporteur talk.

(5)There are four invited plenary talks in the program that focused on or discussed about AdS/CFT and heavy-ion collisions: (a) AdS/CFT and QGP by Dam Son (b) Mach cone in quark-gluon plasma by Jorge Casalderrey-Solana (b) AdS/CFT and jet quenching by Hong Liu (c) Physics opportunities at LHC by Urs Wiedemann. Three of the above speakers have been working on heavy-ion theory for some time.

(6) In addition to the four plenary talks, Dam Son also gave an hour-long lecture on AdS/CFT and QGP in the student lectures the day before the conference. A post-conference workshop on AdS/CFT and strongly coupled quark matter was also held in Beijing (http://qm.phys.tsinghua.edu.cn/thu-henp/sqm/).

Xin-Nian Wang

stefan said...

Dear Xin-Nian,

thank you for your comments and explanations!
We try our best to convey as realistic and undistorted a picture of heavy ion phyiscs and its community as possible, but sometimes one just forgets to mention some obvious facts...

Best, stefan

Bee said...

Dear Xin-Nian,

Thank you very much for your thoughtful comment. I am very troubled by the hostility that I have noticed, from the string theory side as well as from those who follow the fashion - apparently it is currently 'in fashion' to consider string theory a waste of time. I don't think so, and I seriously hope that after the last months of discussions, we can all go back to work and lead arguments in a scientific way.

Your comment is definitely helpful in this regard, as the string theorist going to a heavy ion conference is still an exception - therefore it's probably difficult to judge on how serious McLarren's slides have to be taken. I found them slightly amusing, not very polite, and kind of inappropriate for a summary talk, but I didn't expect them to be upsetting.

One way or the other, I do think that the topic AdS/CFT in heavy ion physics is exciting, but there's also work left to do. And having seen how fast topics get overhyped, I hope to communicate that in the first line, scientists should be realists.

Best wishes,

B

Bee said...

See also

Jon's report
Clifford's post on Jon's report

and

Comments at NEW

Bill Zajc said...

Dear Bee (und Stefan): I was pleased to find this very nice blog on topics of general interest to me. Of course, that doesn't mean I was pleased with each specific entry ;-)

Eventually I will get around to commenting on the specific topic that led me to your blog, i.e., Larry McLerran's un-fortunate un-summary talk at Quark Matter 2006. But since I have a somewhat demanding day job, I can't keep up with the torrent of replies on this topic, so I'll have to do this bit-by-bit.

So Bee, you're first. I was greatly puzzled when you wrote

I could not avoid noticing the sharp contrast between his (Pavel Kovtun's) predictions, and the predictions I am used to from nuclear physics talks. The latter of which usually include some plots of calculated observables, how well they fit the actual data points (and error bars to both if necessary). Based on this, a conclusion on the quality of the model should be given, and how they compare to other approaches. It seems to me that the AdS/CFT calculations haven't yet quite reached this state.

I guess you don't go to many heavy ion talks, or perhaps you don't think of heavy ion physics as "nuclear physics" ;-)

While there are valid reasons for being skeptical of direct application of AdS/CFT results to real heavy ion data, criticizing it on the basis of error analysis seems quite strange to me. Most of the theory talks in the field could be criticized on precisely these same grounds. Did you ever see a serious discussion of the systematic error on applying a 1-d Bjorken scaling ansatz to a real (3-d!) heavy ion collision? Or an error ellipse complete with 1, 2 and 3 sigma contours for theoretically extracted quantity a versus b, a la the famous concordance plots from cosmology?

Perhaps this is why you write

Let me just say that I am not too fond of the heavy-ion community myself. I've had my problems with their 'scientific standards', and their eagerness to fitfitfit some data with whatever parametrization ...

but nonetheless it is painful for me to read. The RHIC experimental community has made a huge effort to improve the level of rigor in reporting their results. Compare the discussion of systematic errors in almost any RHIC experimental paper to almost any paper on experimental results from the pre-RHIC era- I think you'll see a very significant qualitative increase in the scientific standards. (I have no doubt the same high standards will be met or exceeded when LHC heavy ion data are published.) In an ideal world, the precision of the theory results would have responded immediately to the flood of high quality RHIC data. In the real world, causality implies a delay, but some of us are getting a little impatient.

More than two years ago, in an RBRC (RIKEN-BNL Research Center) workshop in which the experiments presented progress reports on the "white papers" then being written, we explicitly noted the need for more precision in extraction of fundamental theoretical quantities. See for example slides 40-42 in the talk I presented there.

On slide 27 of the same talk you can find a crude error estimate on one quantity (the initial gluon rapidity density), derived I believe from a guess-timate supplied by Mikos Gyulassy. Since then, the next modest step towards more quantitative error evaluation on similar parameters that I am aware of has resulted from experimentalists performing the (oh so necessary) task of evaluating sensitivity of the models to real data. See this PHENIX talk from QM06 for a nice example. Following that talk, I tried to be bemused rather than be-angried when an earnest young theorist criticized the speaker for this naïve approach which ignored all the other parameters in the theory. Well, precisely! Did he think we were not painfully aware of that? The same experimentalists who discuss (argue) for hours data point by data point over the statistical errors, the systematic error, the global scale error, how correlated are the systematic errors across the spectrum, are these 1-sigma, 90% C.L., box errors, und so wieder ?

To end this on a positive note, we are not yet in the "concordance" era of precision RHIC QCD-ology. But we are well into the "consolidation" phase of same.

Kim Jong Il said...

I wood like to caal on Moshe and other string theorist out there to *denounce* Brian Greene for shameless over-the-top hipee ("data now ... appear more accurately described using string theory methods..."). While they're at it they might as well also denounce Lubos for spout nonsense above in thread, and for being a Reactionary and Enemy-of-the-People!
If requested denunciations not forthcoming I might just have to go test more A-bomb...

Bill Zajc said...

Now in response to various comments on Larry McLerran himself, tackiness, and Pinocchio. (Thanks to Stefan for correctly spelling Larry's last name, and for noting his serious credentials.)

First, what did Larry actually say regarding the Pinocchio award? It was something like this:

"Each year at this conference, Jean-Yves Ollitraut and I get together as a committee of two to decide who told the biggest lie at the meeting. Although Jean-Yves is not here, and I haven't discussed this with him, this year I would like to publicly announce the award and present it in absentia to Brian Greene, based on what he wrote recently in the New York Times."

The precise sentence Larry found so offensive in Brian's Op-Ed article was "And in a recent, particularly intriguing development, data now emerging from the Relativistic Heavy Ion Collider at the Brookhaven National Laboratory appear to be more accurately described using string theory methods than with more traditional approaches." Is the sentence strictly true as written? It's debatable, but note that neither "string theory methods" nor "traditional approaches" is terribly specific. Would it have been accurate to say "string theory methods are providing new insights on data now emerging from RHIC"? Sure. Was Brian writing for Phys. Rev. Letters? Of course not. Did heavy ion theory gain anything by heaping opprobrium on our colleagues in a another field who are showing strong interest in our data? Well, yes, it "gained" an (in large part) undeserved recognition for insularity and insecurity.

I will comment in a later post on the substance of the AdS/CFT description of heavy ion physics, and why I agree with Brian that it's a "particularly intriguing development". But for now I'd like to focus on the style of Larry's criticism that has engendered this long discussion. In general, I would be inclined to agree with the Cynthia's first observation "It's mildly troubling for Larry McLerran to be tacky enough to suggest that Brian Greene deserves the 'Pinocchio Award'." as well as Bee's follow-up "The reason why I included the slides was that there's one thing I detest more than insulting others, that's insulting them behind their back"

Agree completely. So has anyone notified Larry- not an obvious resident of the blogosphere- about this discussion? ;-)
(Don't worry, since I'm writing critical things, I have sent him a pointer.)

On the general issue of when does someone become fair game, I would be willing to assume that most readers of this blog would not feel compelled to notify George Bush or the Pope when criticizing them. On the other hand, I would like to think that all of us would feel compelled to notify a colleague when doing so (criticizing them, not George Bush!). So resolving the issue of 'tackiness' may require determining if Brian Greene has ascended to the level of the papacy ;-) Well, I think a slightly lower standard is the determining factor- rather than a litmus test, let's call it the Letterman test: If you've appeared on David Letterman's show, you've reached the level of fame where you should 'expect' these sort of broadside (well, backside) attacks. Brian satisfies this criterion, but at the same time I'll bet he's not phased by such remarks.

Note: Just to remove any ambiguity- my Letterman test is not intended to imply that popularizers of science should be assailed; it's just that it seems to come with the territory. The "Carl Sagan = Brian Green = bad for science" dolts are the ones who are truly bad for science. If you don't like the fact that they have been successful promoting "their' science, then get out there and try promoting your science.

So it's not clear to me that Larry's criticism of Brian qualifies for 'tacky' based on the Letterman test. But don't worry, it makes the tacky grade on another criterion, i.e., being unwarranted, irrelevant and inappropriate. More later.

Bee said...

Dear Bill,

Thank you veru much for your comment and your clarifying words.

Just a general disclaimer:

So has anyone notified Larry- not an obvious resident of the blogosphere- about this discussion? ;-)

I am not aware that my post is insulting for anybody. If you think it is, please let me know. Regarding a notification: I don't send an email to Mukhanov if I mention he gave a nice colloquium, I don't send an email to Smolin when I comment on a review of his book, I don't tell Hofmann when I summarize his paper, I don't even tell my mum when I cite her. If I'd notify everybody who is mentioned on my blog, I'd never finish any post.

I DO ask people if I explicitly want to use things they told me privately.

I am also sorry if some of the comments here are perceived as anti-something. I delete the worst, but I also can't notify everybody who might feel insulted by a comment on my blog.

To add some last words: It's as interesting as frustrating that posts which I've worked on for quite some while hardly get any attention, whereas this post, which hardly has any content, causes such a response.

Best,

B.

Bill Zajc said...

To expand a little on Stefan's reply to Tompot as to why gold nuclei are the ions of choice for RHIC:

Yes, one wants the heaviest possible nucleus, so why not lead (atomic mass 208) or even uranium (mass 238) rather then gold (mass 197)? The answer really isn't due to the absence of other long-lived isotopes (which are indeed there for both lead and uranium). An accelerator complex like RHIC is an incredibly high resolution mass analyzer; one would have to work pretty hard to simultaneously accelerate two different heavy isotopes. The real answer isn't as much physics as chemistry: the nucleus to be accelerated must be from an electronegative atom, that is, an atom to which you can attach an electron.

That's because the first step of acceleration at RHIC uses a Tandem Van de Graaff accelerator. Tandems exploit a bait-and-switch tactic to double the energy of their beams: After first accelerating a negative ion (in this case gold) attracted to a high positive voltage, the atom is passed through a foil which strips (at least) two of the electrons, giving a positive ion which is then pushed away from the positive electrode. A neat trick; see this Wikipedia entry on "Tandem Accelerators".

This odd restriction on the otherwise extraordinarily flexible RHIC accelerator complex will be lifted in 2009 with the advent of the Electron Beam Ion Source. Then RHIC will be able to accelerate any species up to and including uranium. Uranium-238 is a particularly interesting nucleus because it has a very significant static deformation, roughly the same as an American football. It will be an interesting experimental challenge to see if we'll be able to sort out after the fact those collisions in which the nuclei collide with their axes aligned (leading to higher densities?) from those where their axes are perpendicular.

stefan said...

Dear Bill,

thank you very much for your clarifications about the choice of heavy ions at RHIC and the preparation of the beam!

It's great that the spokesman of one the big RHIC experiments finds some time to help us with explanations about details of RHIC physics on this blog :-)

Best, stefan

Anonymous said...

Dear Bill.

In all sincerity... Brian Greene's statement is completely wrong and blatantly misleading. There is no other way to put it. I don't think most of the people who do the AdS/CFT calculations would disagree:

AdS/CFT has so far provided a toy model theory where one can calculate things analytically in the infinitely strong coupling limit.
As a toy model, it is nice. It has provided some insights into general thermodynamics/fluid dynamics (e.g. the "eta/s bound"). Perhaps it will provide more insights closer to experiment.

But this has not happened as yet:
AdS/CFT calculations have not yielded a single experimental prediction testable at RHIC. They have not made a single
conclusion more profound than "hey, this might agree with RHIC to an order of magnitude".
They have not even approached one of the puzzles that RHIC yielded
(why does the matter produced at RHIC seem to thermalize so fast? What is it's initial energy/momentum distribution?
What are it's microscopic degrees of freedom?
What about freeze-out/HBT? Which of the models on the market best described jet quenching? Which of the models on the market best describes where does the missing jet energy go? What are the experimental signatures to look for if we want to diffeerentiate these scenarios? etc.)

When one will get an insight, from the AdS/CFT calculations into one of these questions, than you can say that
"string theory methods are providing new insights on data now emerging from RHIC"
This has not happened yet. There are good reasons to suppose it will never happen, although it's not impossible that it will happen.

To put it in a slightly informal way, Brian Greene was talking BS.
But that's somewhat OK: Theorists talk BS to "hype up" their theories all the time.
And I don't mean string theorists, I also mean heavy ion theorists and probably all other theorists. These days, it's part of their job. You, on the other hand, are an experimentalist, and part of YOUR job is to call the theorist's BS for what it is!

Experimentalists at RHIC have done an amazing job (one can measure it by discrepancy between the huge amount of very high quality data and the extremely small amount of "robust" clear explanations!) They don't need to hype up their work using statements by theorists that are, when not clearly wrong, extremely speculative. The fact that they have not always done this is something that really annoys me about heavy ion physics.

Bill Zajc said...

Dear 'Anonymous'

Normally I would not consider replying to an anonymous post. But a) on the off chance that you are a junior person posting anonymously out of concern for future job prospects (if that is the case, you have my full sympathy and understanding) and b) because what you write is so blatantly wrong when compared to current developments, I can't resist:

Q. How easy is to deconstruct your statement

AdS/CFT calculations have not yielded a single experimental prediction testable at RHIC.

?

A. Absurdly so:

To cite but two examples-

1) Hong Liu, Krishna Rajagopal, and Urs Wiedemann have calculated the screening length for charmonium suppression and found a characteristic \gamma^{1/2} factor which can be tested by examining the pT dependence of J/Psi yields.

2) Joshua Friess, Steven Gubser, Georgios Michalogiorgakis and Silviu Pufu calculate the "Mach cone" angle and its momentum dependence associated with radiation from a heavy quark traversing the medium, testable by looking at e-hadron correlations with the current detectors and (after upgrades) by heavy flavor tagged jets in the future.

More generally- the above paper cites 14 papers (Refs. 7-20) that use AdS/CFT methods to calculate energy loss and/or drag force in the thermal medium. Now, it is not clear that a calculation of say 'q-hat' is an "experimental prediction testable at RHIC". But please then level this same criticism to essentially all 'real' QCD calculations that determine q-hat, but make such unrealistic assumptions about nuclear geometry and expansion as to render them (also) untestable by direct comparison to experiment.

As for your remarks

You, on the other hand, are an experimentalist, and part of YOUR job is to call the theorist's BS for what it is!

Experimentalists at RHIC have done an amazing job (one can measure it by discrepancy between the huge amount of very high quality data and the extremely small amount of "robust" clear explanations!)

Here I am pleased to find that we are in (near) total agreement. The only point of departure is my view that the AdS/CFT crowd is actually being held to a higher standard that the 'baseline' heavy ion theory. The experimentalists I've spoken to are fully aware that the AdS/CFT methods are not a description of 'real' QCD. But at the same time, they are fully aware that the same observation applies to the majority of heavy ion calculations currently on the market.

Anonymous said...

Thanks for the reply, but I think you completely missed the point. I know that AdS/CFT can be used to calculate things related to experimental observables (and yes, the sentence you cite was misleading. But it was in a context).

My point, and I think that was Larry's point as well, is... so what?

A successful hydrodynamic description of v_2 has told us that the mean free path is small compared to the system size, and that the system thermalizes quickly, statements that we can hope to refine
(and eventually to make contact with basic theory, something that we still can't do).
Jet quenching has told us that partonic density of the system is high, statements we hope to refine using, e.g., Mach Cones.

HBT measurements have told us the system's space-time distribution at last scattering is not what we naively believe.
BS correlations tell us the degrees of freedom of the system are probably something like quarks and gluons.
etc.

Comparison between
N=4 AdS/CFT calculations tell us... what?
That the system has 4 supersymmetries?
That it's conformally invariant?
We allready know that's wrong.
At most, the fact that AdS/CFT calculations qualitatively agree with RHIC might suggest that "the coupling constant is very strong", and any differences between AdS/CFT and QCD are of secondary importance. Bust since we allready know that the degrees of freedom are very different, this in itself tells us very little.

Don't get me wrong, I think these
calculations are important and should be done, but NOT because they describe RHIC.
I think they are toy model calculations that give field theorists an idea of what a QFT looks like at strong coupling, without the need for numerical simulations. This is A LOT.
But it has, it least for now, VERY LITTLE to do with "accurately describing RHIC data" or answering to the many unsolved questions RHIC experiments are throwing at theorists This is a point which I think the authors of the papers you quote will agree on, since, if you read them, discussions of the puzzles raised by RHIC data occupy little to no space (beyond such general statements as "RHIC has proven the coupling constant is strong).

I'd be happy to be proven wrong:
In my previous post, I listed all puzzles
that arose from RHIC experiments that
I could think of.
Please point me to a paper where AdS/CFT
has been used to give a resolution to one such puzzle. If you will find one, I'll agree Brian Greene's statement is not totally wrong and misleading.

Bill Zajc said...

Dear 'Anonymous'

I see from your latest post that on an 'empirical' basis, our positions are not so far apart, for example

"I think these
calculations are important and should be done, but NOT because they describe RHIC.
I think they are toy model calculations that give field theorists an idea of what a QFT looks like at strong coupling, without the need for numerical simulations. This is A LOT."


I could agree with that. Where we differ may be on a 'philosophical' basis:

"I know that AdS/CFT can be used to calculate things related to experimental observables ... My point, and I think that was Larry's point as well, is... so what?"

Here we seem to have a profound difference, in particular regarding "so what?". I will (try to) address this in a later, longer post, perhaps this evening (real work permitting).

You nicely enumerated various puzzles from the RHIC data, to which I would add an important one, that of charm energy loss. And I would refer you to this very nice recent discussion by Steve Gubser, which does an excellent job presenting the successes, ambiguities and limitations in comparing AdS/CFT results to real data.

Bill Zajc said...

So now let's discuss the why I find AdS/CFT so fascinating (and its bashing so silly). On one level, it's just a 'toy' model that lets people efficiently calculate some things that are hard to do in 'real' QCD whenever the coupling gets strong. I agree with Anonymous that already "This is A LOT".

But it’s the precise nature of the tools being brought to bear on the problem that are important. Bekenstein's realization, and Hawking's calculation, of black hole entropy is one of the most fundamental advances in understanding made in the latter portion of the last century. (Feel free to disagree with me, but then I'll feel free to classify you with those who thought that guy Planck was on the wrong track worrying about the entropy of oscillators...) So when a connection emerges between thermal gauge theories and black hole entropy, either you sit up and take notice, or your mind is on its way to room temperature.

The application of this technology by Kovtun, Son and Starinets to obtain a (conjectured) bound on the (shear) viscosity to entropy density ratio was a breakthrough. Since the bound contains no reference to Newton's constant 'G', it is (obviously) true that the result has nothing to do with string theory- it's "just' quantum mechanics. As others have noted, it was nearly anticipated by Danielewicz and Gyulassy in 1985 (see also Hirano and Gyulassy for a modern perspective).

But think about it: We've known about entropy for nearly 200 years, viscosity nearly as long, quantum mechanics for nearly a century. Yet (as far as I know), it was only by "crossing over" from CFT to AdS that led to a sharp conjecture about the specific ratio of these two quantities. Nor am I aware of any other (non-AdS/CFT) results either validating or refuting the KSS conjecture. This seems like it should be an open challenge to those who are skeptical of string techniques: Prove or disprove KSS!

(Note: in the above I was careful to qualify the statement with the 'non-AdS/CFT' qualifier. If you are a fan (who isn't?) of the beloved 1-d hydro Bjorken scaling solution, you should look at this paper by Janik. According to Janik the Bj solution doesn't exist- at least not in N=4 SYM. If you try to impose the Bj solution with perfect fluid dynamics in the CFT, the AdS develops unphysical curvature singularities. To eliminate them requires the introduction of precisely the value of viscosity (s/4pi) calculated by KSS. While this might be directly traceable to the conformal aspect of the field theory, it is still a striking result.)

Even more striking is the notion that we might get closer to this bound in heavy ion collisions than in any other physical system. (Actually, it's nearly as amazing to me that something like helium, which depends on various 'accidents' like the residual nuclear interaction forming stable Z=2 nuclei and MassProton/MassElectron = 1836.152 and the van der Waals interaction making a fluid and ... can form a system that gets within a factor of 10 of the bound.) Presumably deconfined QCD matter near Tc has a shot at the record because a) the coupling is about as strong as allowed by unitarity and b) there are no 'extraneous' scales in the problem like 1836.152; the temperature is the only relevant dimensionful parameter.

So rather than carping that AdS/CFT isn't real QCD, I'd be much happier to hear open acknowledgment that the KSS work has transformed how we think about the sQGP, even if the bound is 'just' quantum mechanics. Measuring transport coefficients is one thing, approaching a fundamental limit is an entirely different matter.

Finally, to return to the various issues with Larry's talk that prompted me to enter the blog world (briefly!- I have no intention of living here, although I am appreciative of Bee and Stefan's hospitality):

Perhaps it was a joke that got away from the speaker, but Larry's "Bad Side/Good Side" comparison came across rather strongly, especially given that each of the two main AdS/CFT speakers at the conference(Dam Son and Hong Liu) gave appropriately nuanced talks with the caveats fully on display. And perhaps it was an unintended consequence that Larry's faint praise of KSS resulted in the perception that he was damning the entire AdS/CFT approach.

But the Pinocchio award argues against that. As does the following: Suppose the offending New York Times Op-Ed piece had been written by a 'real' plasma theorist concerned about (hypothetical) threats to the U.S. Fusion Energy Science budget, and further suppose that s/he had written "And in a recent, particularly intriguing development, data now emerging from the Relativistic Heavy Ion Collider at the Brookhaven National Laboratory appear to be more accurately described using Weibel plasma instabilities borrowed from my field rather than by more traditional approaches." (Never mind that the copy editor would remove the physics reference.) Do you think this would have been singled out for ignominy? I don't, because it doesn't have that damn 'string' attached.

And there are other issues. The comparison shown on Larry's Slide 21 of S/S_SB (deviation of entropy S from Stefan-Boltzmann value for free gas; Larry doesn't supply the reference; it's this paper by Rebhan) was presented a triumph of resummation techniques over AdS/CFT. I agree that Rebhan's application of Hard Thermal Loop technology is beautiful. It gives an essentially constant value for S/S_SB of 0.88 . Seems much better than the truly constant value of 3/4 from string theory. But the 3/4 applies in the 't Hooft limit, and we all agree neither the coupling nor the number of colors is infinite. So when Steve Gubser, in the paper I mentioned above, tries to make a realistic estimate of the coupling, then computes the AdS/CFT value for S/S_SB, he gets... 0.88! (see Eqs. 14 and 15). To be fair to Larry, Steve's paper appeared a few days after his talk. To be fair to Steve, I know he was motivated by other considerations, and, as far as I know, unaware of the nice numerical agreement, although keep in mind this is for a LQCD calculation of pure glue.

And Larry missed the opportunity to comment on a fundamental result from the AdS/CFT results on jet quenching. All calculations obtain a result proportional not to the number of degrees of freedom (unlike conventional approaches), but to sqrt{ g^2 N_C }. Either this is an essential insight, or the key to falsifying these predictions.

And by concluding with the bottom line on his penultimate slide in the last talk of the conference

"AdsCFT (sic) MUST be accountable to the same scientific standards
as are other computations,
or else it is not science. "


Larry implied there was some game going on where the AdS/CFT papers were not being held to high standards. Please. Read the papers emitted in the last year using this approach , read a random sampling of the non-AdS/CFT RHIC theory papers, and form your own conclusion.

Anonymous said...

I do not disagree, I also find AdS/CFT fascinating.
Once again, they are a toy model to get a good quantitative grip of certain physical insights;
One is the Danielewicz-Gyulassy insight about viscosity over entropy density that Larry mentioned.
Another is the paper by Janik you mention (Of course Bjorken hydrodynamics without viscosity has a singularity! One needs a finite start time for the expansion, which in heavy ion physics is put in by hand. Consistently, it should also lead to a viscosity. As in D-G, the maths is more rigorous, but the "physics" was intuited before).

The key is difference "calculating toy models" vs "describing real data".
The Weibel instability analogy is not quite correct:
Weibel instabilities are a physical insight borrowed from another field, that is being used to attempt describing data ("why is viscosity so much smaller than predicted?").
The calculation with an infinite coupling constant is just that...
a calculation with an infinite coupling constant.
It gives no feel as to why does the coupling constant appear to be infinite at RHIC.
Even as a bound on experimental viscosity it's lousy, since, until we have 3D viscous hydro AND an understanding of initial conditions (CGC vs Glauber vs ...), we actually do not know what viscosity is.
It could be one order of magnitude above s/4pi

This is why I really don't like Brian Greene's statement.
I would not have replied to your comment, except that today I learn that Polchinski wrote, in the American scientist (NOT the NY Times):
------------------
http://tinyurl.com/yh3cf8

physicists have found that some properties are better modeled via duality as a tiny black hole in a space of extra dimensions than the expected clump of particles in four dimensional spacetime.
------------------
so it's not an isolated case.
These people are either knowingly over-hyping their theory, or at best being much less careful with the truth than I would expect from a scientist.

Bee said...

Dear Anonymous, Dear Bill,

No, it's not an isolated case, otherwise it wouldn't bother me. I can't blame them for being excited about their research, I suffer from the same disease! But I do agree that it's something different as to how the research is advertised and sold in public, when the audience relies on the expertise and objectivity of scientists. I don't think this is a problem specifically of theoretical physics.

Excuse me for the crude comparison, but I could think that each time a research team claims that some tests on mice have shown that a chemical xyz affects growth of cancer cells (as a side effect, it also kills the mice), and that the result is extremely important, and we are close to curing cancer, I can imagine how colleagues of said team roll their eyes towards heaven, knowing how to judge on the results realistically. And how many times have we read news like this in the last decades? These results are definitely important, and there has been a tremendous progress esp. regarding minimizing the side effects of chemotherapy. This IS a large success, but one shouldn't sell it for what it isn't.

Regarding Joe Polchinski, I've recalled a quotation from an earlier post:

In recent years the theory has contributed significantly to heavy-ion physics, according to Joe Polchinski, a string theorist at the Kavli Institute for Theoretical Physics in Santa Barbara, California. When the Relativistic Heavy Ion Collider at Brookhaven National Laboratory in Upton, New York, first produced a hot quark gas, it was string theory that correctly predicted, retrospectively, some of the gas's properties. "In many ways, I feel the boundaries with other areas of physics are coming down," Polchinski says.

I agree with him on the latter remark, and I think that this is a very good development.

Best,

B.

Bee said...

Sorry, forgot to add, the quotation is from

"Theorists snap over string pieces- Books spark war of words in physics." by Geoff Brumfiel, Nature 443, 482, 5 Oct. 2006

Anonymous said...

Anonymous,

I would not agree that AdS/CFT only provides exactly-solvable models which elaborate on previous physical insights. In my view there are certain insights coming AdS/CFT that are complementary to the insights one gets from more traditional lines of thought.

To start, the very idea of dividing the viscosity (eta) by the volume density of entropy (s) is an insight that came from AdS/CFT. I think this ratio first appeared in a
paper
that Pavel Kovtun, Andrei Starinets and I wrote (see section 4.5). I may have overlooked something in the literature and would appreciate any pointer on an earlier appearance of the ratio. That this ratio, which seems to have no particular physical significance (see an attempt to give one here), can have a universal behavior in a class of theories remains a puzzle (at least to me), but it came out completely naturally from AdS/CFT.

To further illustrate the insights that may come from AdS/CFT, let us consider Eq.(4) in the paper by Hirano and Gyulassy that has been mentioned earlier:

eta/s > 1/15

The authors warn us that there is an uncertainty of order one on the right hand side. But suppose I foolishly ignore their warning and take the value 1/15 literally. I don't know exactly where this factor comes from, but intuition tells me it is D(D+2), where D is the number of spatial dimension. Following this logic, I would conclude that, in D spatial dimensions

eta/s > 1/ (D(D+2))

I can go further and conclude that the minimal possible value of eta/s is a decreasing function of D. Here AdS/CFT gives a valuable insight by arguing against this hypothesis: from Chris Herzog's work on the hydrodynamics of worldvolume theories on M2 and M5 branes, which live in 2+1 and 5+1 dimensions respectively, the bound on eta/s should be the same hbar/(4pi) in any number of dimensions.

It is true that many of these insights are of general character and may not be immediately important to the physics of heavy-ion collision at RHIC. However, as physicists we tend think of the more general questions, even as we study particular phenomena. Don't we think about spontaneous symmetry breaking while studying pion scatterings, and about gauge invariance while studying Josephson junctions?

I also think you misunderstood Janik's result. I know that your complain is mostly about the "more accurately described" in Brian Greene's characterization, to which I make no comment except that many people I respect (including the first director of the FTPI whose name you can learn from the last two pages of this document) are disturbed by it. But if there are any physics questions you want to discuss, I would be happy to do it privately. (You can find my email address from the arxiv or from spires.)

Dam T. Son