Causes of loss of interest in String program

[smoit]

The fact is that the Wigner 15j is key to 4d spinfoam LQG, just as the 6j was key to Ponzano-Regge 3d gravity. Understanding the asymptotics of the 15j is critical for establishing the large-scale limit of LQG. Large scale means large j---so one needs to understand the limit of the 15j symbol for large j. ...

...
Is this Loop? You know from review papers that Loop draws heavily on several of the types of mathematics mentioned by the DESY librarians as keywords. But that does not make the paper Loop. I have to use an automatic criterion in order to tabulate changes---so I do not try to second-guess the DESY, I just go by what keywords they tag on the paper. They don';t say "quantum gravity, loop space" or "quantum cosmology, loop space" or "spin, foam" so I don't count it.

So, let's try your favorite approach on some of the allegedly LQG papers from Berkeley:

http://www-spires.fnal.gov/spires/find/hep/wwwtopics?key=9038191"
List of keywords assigned to the paper: Semiclassical Analysis of the Wigner $12J$-Symbol with One Small Angular Momentum: Part I

invariance, gauge
quantum mechanics
angular momentum
semiclassical
WKB approximation
invariance, gauge

http://www-spires.fnal.gov/spires/find/hep/wwwtopics?key=9030654"
List of keywords assigned to the paper: Semiclassical Analysis of the Wigner $9J$-Symbol with Small and Large Angular Momenta

invariance, gauge
phase, geometrical
angular momentum
factorization
semiclassical
rotation
spinorhttp://www-spires.fnal.gov/spires/find/hep/wwwtopics?key=8787883"
List of keywords assigned to the paper: Semiclassical Mechanics of the Wigner $6j$-Symbol

phase space, reduced
spin, network
semiclassical
mechanics
higher-dimensional
WKB approximation
angular momentum
Poisson bracket
integrability
symplectic

So, are these papers LQG or not, Marcus?

 

[smoit]

Several of the papers contain extensive references to LQG research and discuss their relevance to the Loop program. ...

According to your logic, any mathematician working on Calabi-Yau manifolds, G2 manifolds and related things, who may have attended some string conference should be classified as a string theorist, right?

 

[marcus]

Post#91 strikes me as irrelevant---argument for argument's sake. Smoit misrepresents what I have said, it's called "straw man" I use an automatic criterion to get paper counts so as to have an easy way to keep approximate track of changes.

Keeping track of developments at UC Berkeley physics department is a completely separate activity. I don't use keywordsearch statistics. I did not claim that Liang Yu papers were LQG! Nor would I have imagined adding them to a count of 2011 Loop papers. I let DESY librarians to the classification so there is some hope of consistency and changes in the numbers over time meaning something.

The situation with UC Berkeley is exactly as I have said.
UCB faculty and PhD students have either been visitors at Rovelli's LQG group on several occasions for extended periods or will attend the main Loop conference this year, or both, and collaboration with Rovelli's group has been possible. One of Littlejohn's PhD students is doing a Loop thesis and has already posted an explicitly Loop paper which I expect will be accepted for publication.

I have not said that Liang Yu's papers are specifically Loop. However all that work towards analytical understanding of the 15j symbol is enormously important to the Loop program and although Liang Yu himself may not be part of that program, he is in Littlejohn's group. And Littlejohn has at least one student doing a Loop thesis. So one has to see this as part of a pattern, in context. So that should certainly be pointed out!

The overall picture is that Berkeley is on the Loop map. That much is clear

So, let's try your favorite approach on some of the allegedly LQG papers from Berkeley:

...
...

http://www-spires.fnal.gov/spires/find/hep/wwwtopics?key=8787883"
List of keywords assigned to the paper: Semiclassical Mechanics of the Wigner $6j$-Symbol

phase space, reduced
spin, network
semiclassical
mechanics
higher-dimensional
WKB approximation
angular momentum
Poisson bracket
integrability
symplectic
So, are these papers LQG or not, Marcus?

:zzz:

I had an earlier discussion with Atyy about this. When all you are interested in is the SLOPE, basically. When you don't care about the gross amount but you want to see if something is increasing and how steeply or has leveled off, or is declining, then the convenient thing is to have some automatic yearly tally generated the same way year after year. I don't WANT my subjective judgment involved because that might influence things. So you pick some keywords and use them year after year----let DESY do the work and don't mess with it.
Here's my post to Atyy
==quote from me, page 2 of this thread==

Originally Posted by atyy
Tell me whether these papers are stringy or not:
http://arxiv.org/abs/hep-th/0501052
http://arxiv.org/abs/1010.5009

Is this an LQG paper or not:
http://arxiv.org/abs/math/0201177
​

I will do what you say, since you ask. I have not been counting "stringy" papers because the idea is too vague to make a well-defined time series. I set up a criterion "core String" depending only on the DESY library's cataloguing, so I could measure the same thing the same way year after year without and notice changes.

You mention Loop! But we are not playing some game of "string versus loop" here. It's boring when I point out something good happening in Loop and somebody immediately gets defensive and thinks they have to tell me why String is good (to keep them "even" I guess.) And if I see that the String program has a problem (which various people have attributed to various "wrong turns" and I find interesting) that is not intended as a game of competing theories, which one is "better".

I want to see as fairly and accurately how things are, not play "one-up".

People are always trying to make it seem that the two theories are "even", to balance the points. But they are not on a level. They are actually in very different circumstances as regards speed of development towards a finished formulation and testabilty and probably other things. Also the leadership style is noticeably different. And one has only about 200 active researchers who basically all know each other. And they have very different program goals.

So it seems ridiculous to try to equate the two on merits and demerits, or even spend much time comparing.

What I want to do in this thread is study the loss of expert interest in the String unification program. And hope to hear more about what the causes might be. If it has to do primarily with program management and vision then we might see a turnaround if the causes can be identified and remedied.

=================
about the papers. Here are their DESY keywords. When trying to track an index over time the thing is not to insert one's own judgment and most importantly, measure the same thing each time. So I count "core String" papers to be those the DESY librarians tag with keywords 'string model' or 'membrane model' The following two are not "core String" in that sense.:

...

The paper by Justin Roberts you mentioned is classified mathematics, not physics, and is not in Spires, so it has no keywords. Spires is basically HEP, not math. However Spires does have one paper, from the year before, by Justin Roberts!

http://www-library.desy.de/cgi-bin/spiface/find/hep/wwwtopics?key=4817060
Rozansky-Witten theory
field theory, topological
differential forms, symplectic
algebra, Lie
category
knot theory
mathematical methods

Is this Loop? You know from review papers that Loop draws heavily on several of the types of mathematics mentioned by the DESY librarians as keywords. But that does not make the paper Loop. I have to use an automatic criterion in order to tabulate changes---so I do not try to second-guess the DESY, I just go by what keywords they tag on the paper. They don';t say "quantum gravity, loop space" or "quantum cosmology, loop space" or "spin, foam" so I don't count it.
=====endquote=======

 

[mitchell porter]

String experts have decided after several decades experience that one should NOT think in terms of strings and branes in a geometry with compactified extra dimensions.

But what you get from AdS/CFT are low-dimensional field theories in flat space being equivalent to an AdS space times a compact space, containing strings and branes. The radial AdS dimension encodes the RG flow, and the compact space (and the objects with extension in it) is "made from" the space of ground states of the field theory. From this perspective, string theory is the universal theory of emergent RG geometry in quantum field theory. At the moment, it only works properly for an emergent AdS space, but if the dS/CFT correspondence can be understood, then this will be true for spaces of positive curvature as well. (In dS/CFT the boundary is purely spacelike and lies in the infinite past and future, rather than being timelike as in AdS/CFT, so it's as if the timelike direction in the Lorentzian gravitational space is emerging from Euclidean field theory on a sphere in the infinite past.)

So not only are people still doing flat-space string phenomenology, complete with branes and extra dimensions, but branes and extra dimensions have proved to be implicit in standard quantum field theory, where they emerge from the existence of a continuous degeneracy of ground states. That multidimensional moduli space of ground states is where the extra dimensions come from, in this case! Branes are domain walls separating regions in different ground states, strings are lines of flux connecting these domain walls. Furthermore, in gauge theories with a small number of colors, it looks like the extra dimensions will be a noncommutative geometry, it's only in the "large N" limit of many colors that you get ordinary space. (Consider that the noncommutative standard model of Connes et al is a theory of gravity on an "almost commutative" space - product of a Riemannian space and a finite noncommutative geometry - with the gauge bosons coming from gravity on the noncommutative part of the product geometry. This seems to be consistent with the picture coming from string theory.)

 

[smoit]

Reposting specifically for you, Marcus.

Marcus, I've tried to explain to you before that these keyword search statistics are extremely bad science. Besides not taking into account many statistical concepts such as sample size, it is not even clear that keywords are an effective substitute for just looking at the papers and recognizing what they are about. If you want to do statistics in a way that any scientist can respect, you must actually understand the properties of the sample, have some quantity in mind that is supposed to be of statistical significance, and then actually attempt to quantify the degree of correlation of your hypothesis with the sample. Absence of any of these results in a garbage in-garbage out situation.

To see how far off the mark your keyword searches were, I actually looked at all Witten papers for the periods > 2002- < 2007 and > 2006 - < 2011. These were obtained by using your date ranges in inspire, without the keywords. There is some overlap between the two periods, possibly because papers that appeared in the arxiv in 2006 also appeared in journals in 2007. Statistically, the overlap is not of much significance.

I attempted to classify papers which were string-related and those which were not. Basically my criteria were:

String: About strings, branes or 2d topological qft. Also includes AdS/CFT, the initial twistor amplitude paper and any Langlands papers that refer directly to 2d mirror symmetry.

Other: Papers about straight qft including particle physics, Chern-Simons and exotic qfts that don't directly imply string relations in their abstract or title. Also includes papers about pure 3d quantum gravity.

Results are:

2003-2006 http://inspirebeta.net/search?ln=en...+2002+AND+DATE+<+2007&f=&action_search=Search

-52 total papers

-15 conference proceedings
* 8 QFT
* 7 string

-37 journal articles
* 15 QFT
* 3 QG
* 19 string2007-2010 http://inspirebeta.net/search?ln=en...n_search=Search&sf=&so=d&rm=&rg=25&sc=0&of=hb

-25 total papers

-5 conference proceedings
* 5 QFT

-20 journal articles
* 9 QFT
* 2 QG
* 9 stringThe first thing to notice is that Witten's total output in 2007-2010 was half that for the 3 yrs prior. Measurements for this period will be less statistically significant as a result. As for relevant ratios,

2003-2006: 50% of total were string, 51% of journal articles were string
2007-2010: 36% of total were string, 45% of journal articles were string

It is amusing to do some statistics. Suppose that Witten's papers are randomly distributed between string and nonstring physics, and that the topics of papers are independent of previous papers. Then we have a binomial distribution. With 35 string papers in 77 trials, we have p=0.45. For the two periods we have

2003-2006: expected number of string papers: 52(0.45) = 24
variance: 52(0.45)(0.55) = 13
actual number of string papers: 26
expected: 24\pm 3.6

2007-2010: expected number of string papers: 25(0.45) = 11
variance: 25(0.45)(0.55) = 6
actual number of string papers: 9
expected: 11\pm 2.4

In both cases the number of string papers produced is within one standard deviation of the expected result. There is no reason to conclude that Witten has lost interest in string theory between these two periods.

You are free to conduct a similar analysis over a larger data sample, or for other notable string theorists. The results might be interesting. Keyword search results with no analysis will not be.

When you do a careful and proper statistical analysis you'll see that there is no decline of interest, as fzero has already confirmed. This is the point that YOU are deliberately ignoring. In particular, in your estimates you are completely ignoring the decline in the total number of papers put out by the same author in a given time period. It is completely natural for older professors to be less active and put out fewer papers. For instance, Eva Silverstein is from the younger generation and is still rather active, hence her absolute number of all papers, includings those on string theory, has not at all declined.

 

[marcus]

marcus, as you know, I disagree with your definitions. But I want to find out how strict they are. Let's say LQG goes in the direction of AdS/LQG, would you count that as LQG or not?

For one thing, Atyy, that's too speculative.

Another thing is we have to make a clear distinction between what gets tabulated to indicated changes, as a convenient index, versus one's own subjective casebycase assessment.

For a handy index to watch I let DESY do the work. I'm only interested in changes not absolute numbers, I want a rough idea. Maybe they overcount, maybe they undercount, but as long as they are reasonably consistent over time, their tally will PARALLEL the real world, roughly at least. So it is a convenient objective thing to refer to.

But in my own subjective mind there is no such thing as Loop apart from the self-selecting elite community of scientists. Same way with Mathematics. It has no verbal formula defining what is and what is not. No dictionary definition. No "essence". The content of Mathematics changes historically. It is what Mathematicians do.

Mathematicians are clearly identified for us by institutions. What is interesting deep valuable etc is determined by the self-selecting elite community, an aristocracy so to speak. Therefore it is capable of evolving. The field and the community co-evolve

In a lesser way perhaps, types of physics are like that. What Loop is is determined by things like the Zakopane school, and whose PhDs get postdocs where, and whose postdocs get jobs. And the lineup of invited speakers at Loops 2011, and what they talk about.

There is no "eternal essence".

Let's not waste time on "metaphysics". We know who the leaders are in LQG, if I saw them doing "AdS/LQG" whatever that means then I'd probably say yes that is included in Loop gravity. And our idea of who the leaders are, the central figures, evolves with time.
Now it's Ashtekar Rovelli Freidel Lewandowski...But in 10 years from now who can say? Could it contain Krajewski? Or Livine or Speziale or Bianchi or Dittrich or Ryan? just to name a few that happen to come randomly to mind. The trendsetter and field-changers will emerge, we will recognize them, we will see what they do. This will define the field. The field will evolve.
============================

With String the field seems to be in a crisis. Citations have fallen off sharply. Many fewer recent string papers make the top-cited list now than say in 2003. The field is in flux but there seems to be no clear direction. I'm really in suspense about where String is going.

Right now I'm thinking that this crisis of direction goes back to 2003 with KKLT, the 10^{too many} vacuums, and Susskind's panicky response. KKLT was a FAILED attempt to address the positive cosmological constant. And Susskind spooked people and they stampeded into the Multiverse. The leadership squelched Multivism in 2008, but it did a lot of damage and continues to. Recently here a highly respected Pro has said "We should not consider the Landscape to be a disaster". Translated, this means "Don't panic." I would say Landscape may indeed be a disaster, but one should remain calm in such a case and try to carry on business as usual. Panic only makes things worse.

 

[atyy]

Let's not waste time on "metaphysics". We know who the leaders are in LQG, if I saw them doing "AdS/LQG" whatever that means then I'd probably say yes that is included in Loop gravity.

OK, you said it.

 

[marcus]

You've got me curious now. Do you have anything particular in mind for AdS/LQG? Any research paper? Freidel wrote something about AdS/CFT correspondence a while back: 2008? But I would not call that AdS/LQG, not in a serious moment anyway.

==========EDIT===========
Atyy, you seem to have issues with DESY classification. I am flexible in individual cases, see connections, try to operate and discuss in good faith, and so forth. But to generate tables like this with Spires I simply pick the best available keywords and pay attention to the changes (not the absolute numbers). I do the same thing willynilly with LQG tabulations. This I think is what you consider "inflexible" and take issue with. So please tell me what your interpretation of this is! It's pretty stark so it has to mean something. If you don't think it reflects a change in the research focus/activity of these top people, then what? The classifiers used were "string model" and "membrane model"---BTW can you find another two or three DESY tags that you prefer, that you would propose running for comparison?

          1995-1998      1999-2002      2003-2006      2007-2010

Strominger     23             14             22              4
Maldacena      27             33             24              9 
Polchinski     21             17             11              4
Harvey,J       16             15              9              2
Duff,M         24             17              8              5
Gibbons,G      17             29             11              2
Dijkgraaf      18             11              9              7
Ooguri         31             18             13              8
Silverstein,E  16             15             16             10
Seiberg,N      19             16             22              1

PS: There were loud objections to having Witten on the list. So I have removed his name. We now have a sample of 10 top people. What are these people working on NOW? What is the String program about if you deemphasize basic parts of the paradigm, like brane and string (and compactified extraD)? If you want the Spires links, go back to post #63.

 

[atyy]

You've got me curious now. Do you have anything particular in mind for AdS/LQG? Any research paper? Freidel wrote something about AdS/CFT correspondence a while back: 2008? But I would not call that AdS/LQG, not in a serious moment anyway.

My point was that if you are so flexible with "LQG", then being similarly flexible with "string" will show that interest in string has not declined.

AdS/LQG doesn't really exist yet, but I am watching to see if it develops from

1) http://arxiv.org/abs/0907.2994 which links tensor networks and LQG

2) http://arxiv.org/abs/0905.1317, a speculative attempt to link tensor networks and AdS/CFT that has been cited by Vidal http://www.emergentgravity.org/drupal/sites/default/files/EGIV_presentations/Vidal.pdf and Evenbly http://pirsa.org/index.php?p=speaker&name=Glen_Evenbly in their talks and by http://arxiv.org/abs/1102.5524

3) http://arxiv.org/abs/1103.6264 which summarizes the known links between spin foams, lattice gauge theory and tensor networks

I know spin foams try to be more background independent than AdS/CFT, but if they work, I'd expect them to contain (or at least overlap) it a special case. I doubt it's a coincidence that Smolin (the LQG heretic who loves string), Freidel (who with Livine, Oriti, and Ryan represent a branch of LQG has often tried to interpret the formalism in "emergent" ways), Vidal, and Cachazo are all somewhere in Canada.

 

[marcus]

Atyy,
You seem to be taking issue with DESY cataloguing. What impresses me is the huge decline during the period 2003-2005, among a sample of top people. It would be interesting to know what other areas (whether or not you want to say they are "really" String) those people went into. Someone was claiming that the Langland's program was "really" String and "should" be counted. This is questionable. But that was just in the case of one person out of 10 or so. They pretty much all showed the same abrupt drop. So far not adequately explained away.

Still, if the DESY classification bothers you, here is a measure of declining interest that is not based on DESY. In making the following counts I tried to be inclusive in deciding what was string. I don't recall anyone pointing out a paper I missed: i.e. one not counted which they think is "really" a string paper. Make your own counts, if you want, and see if your numbers agree with mine.

Cites to recent String papers reflect the researchers' assessment of the value of their own colleagues' current output.
Spires top cited articles during odd years 2001-2009
(with number of recent string papers making the top fifty shown in parenthesis)

http://www.slac.stanford.edu/spires/topcites/2001/annual.shtml (twelve)
http://www.slac.stanford.edu/spires/topcites/2003/annual.shtml (six)
http://www.slac.stanford.edu/spires/topcites/2005/annual.shtml (two)
http://www.slac.stanford.edu/spires/topcites/2007/annual.shtml (one)
http://www.slac.stanford.edu/spires/topcites/2009/annual.shtml (one)

A paper is counted as recent here if it appeared in the past five years.
=========================

 

[atyy]

Atyy,
You seem to be taking issue with DESY cataloguing. What impresses me is the huge decline during the period 2003-2005, among a sample of top people. It would be interesting to know what other areas (whether or not you want to say they are "really" String) those people went into. Someone was claiming that the Langland's program was "really" String and "should" be counted. This is questionable. But that was just in the case of one person out of 10 or so. They pretty much all showed the same abrupt drop. So far not adequately explained away.

Still, if the DESY classification bothers you, here is a measure of declining interest that is not based on DESY. In making these counts I tried to be inclusive in deciding what was string. I don't recall anyone pointing out a paper I missed: i.e. one not counted which they think is "really" a string paper. Make your own counts, if you want, and see if your numbers agree with mine.

Cites to recent String papers reflect the researchers' assessment of the value of their own colleagues' current output.
Spires top cited articles during odd years 2001-2009
(with number of recent string papers making the top fifty shown in parenthesis)

http://www.slac.stanford.edu/spires/topcites/2001/annual.shtml (twelve)
http://www.slac.stanford.edu/spires/topcites/2003/annual.shtml (six)
http://www.slac.stanford.edu/spires/topcites/2005/annual.shtml (two)
http://www.slac.stanford.edu/spires/topcites/2007/annual.shtml (one)
http://www.slac.stanford.edu/spires/topcites/2009/annual.shtml (one)

A paper is counted as recent here if it appeared in the past five years.
=========================

Maldacena's paper was cited 392 times in 2001, and 696 times in 2009, clearly showing a decline in interest in string?

 

[mitchell porter]

AdS/LQG

Last month, MTd2 said something about https://www.physicsforums.com/showthread.php?t=484836#6". My intuition says that's another possible starting point for "AdS/LQG". The reason is just that braids mean something in (2+1) dimensions, and ABJM is the prototypical d=3 theory with an AdS4 dual.

edit: Also, a while back I https://www.physicsforums.com/showthread.php?p=3170483#8" if Marcolli's "topspin foams" might find a use in AdS/CFT.

 

[atyy]

Last month, MTd2 said something about https://www.physicsforums.com/showthread.php?t=484836#6". My intuition says that's another possible starting point for "AdS/LQG". The reason is just that braids mean something in (2+1) dimensions, and ABJM is the prototypical d=3 theory with an AdS4 dual.

Via Bilson-Thompson's ideas? That'd be sweet. I was thinking more conventionally, like using a supersymmetric spin foam to make a lattice version of ABJM.

 

[marcus]

Maldacena's paper was cited 392 times in 2001, and 696 times in 2009, clearly showing a decline in interest in string?

What I am tracking here is the value or quality of CURRENT STRING RESEARCH output.
I arbitrarily chose 5 years as the window. A paper of M from 1998 was recent in 2001 and no longer so in 2009.

I suppose one way to interpret---one way you could put it---is to say that there are a lot more string theorists now than in 2001, and their work is less significant, less interesting. They don't value each other's current research output as much as formerly was the case.

So the citation number crashed from 12 in 2001 down to 1 in 2009.
See post #100 of this thread.

BTW PAllen had some interesting reflections earlier on why this happened. I'll look back tomorrow and find his post.

I would say that there was a big decline in the interest/value of current String program research. And that is like a "leading indicator". It was followed in the next 3-5 years by many of the top people venturing out into more remote areas, with only partial or tenuous connection to program's homebase.

 

[smoit]

The numbers which Marcus is presenting here are completely meaningless. He measures people's interest in string-related papers by choosing his own phony criteria he calls "core string" and deliberately uses an absolute number of papers on a specific topic while failing to take into account the yearly change in the total number of papers output by the same author. fzero kindly went out of his way to explain to Marcus why his analysis was "garbage in => garbage out" and did a more careful job by doing some statistics on Witten's papers. As a result fzero finds no decline of interest, contrary to Marcus' allegation. Below I repost fzero's estimate.

Marcus, I've tried to explain to you before that these keyword search statistics are extremely bad science. Besides not taking into account many statistical concepts such as sample size, it is not even clear that keywords are an effective substitute for just looking at the papers and recognizing what they are about. If you want to do statistics in a way that any scientist can respect, you must actually understand the properties of the sample, have some quantity in mind that is supposed to be of statistical significance, and then actually attempt to quantify the degree of correlation of your hypothesis with the sample. Absence of any of these results in a garbage in-garbage out situation.

To see how far off the mark your keyword searches were, I actually looked at all Witten papers for the periods > 2002- < 2007 and > 2006 - < 2011. These were obtained by using your date ranges in inspire, without the keywords. There is some overlap between the two periods, possibly because papers that appeared in the arxiv in 2006 also appeared in journals in 2007. Statistically, the overlap is not of much significance.

I attempted to classify papers which were string-related and those which were not. Basically my criteria were:

String: About strings, branes or 2d topological qft. Also includes AdS/CFT, the initial twistor amplitude paper and any Langlands papers that refer directly to 2d mirror symmetry.

Other: Papers about straight qft including particle physics, Chern-Simons and exotic qfts that don't directly imply string relations in their abstract or title. Also includes papers about pure 3d quantum gravity.

Results are:

2003-2006 http://inspirebeta.net/search?ln=en...+2002+AND+DATE+<+2007&f=&action_search=Search

-52 total papers

-15 conference proceedings
* 8 QFT
* 7 string

-37 journal articles
* 15 QFT
* 3 QG
* 19 string


2007-2010 http://inspirebeta.net/search?ln=en...n_search=Search&sf=&so=d&rm=&rg=25&sc=0&of=hb

-25 total papers

-5 conference proceedings
* 5 QFT

-20 journal articles
* 9 QFT
* 2 QG
* 9 string


The first thing to notice is that Witten's total output in 2007-2010 was half that for the 3 yrs prior. Measurements for this period will be less statistically significant as a result. As for relevant ratios,

2003-2006: 50% of total were string, 51% of journal articles were string
2007-2010: 36% of total were string, 45% of journal articles were string

It is amusing to do some statistics. Suppose that Witten's papers are randomly distributed between string and nonstring physics, and that the topics of papers are independent of previous papers. Then we have a binomial distribution. With 35 string papers in 77 trials, we have p=0.45. For the two periods we have

2003-2006: expected number of string papers: 52(0.45) = 24
variance: 52(0.45)(0.55) = 13
actual number of string papers: 26
expected: 24\pm 3.6

2007-2010: expected number of string papers: 25(0.45) = 11
variance: 25(0.45)(0.55) = 6
actual number of string papers: 9
expected: 11\pm 2.4

In both cases the number of string papers produced is within one standard deviation of the expected result. There is no reason to conclude that Witten has lost interest in string theory between these two periods.

You are free to conduct a similar analysis over a larger data sample, or for other notable string theorists. The results might be interesting. Keyword search results with no analysis will not be.

 

[marcus]

OK folks, here is the revised tabulation. No one has explained why there were so many papers for so long and then suddenly, in most cases, so few:

          1995-1998      1999-2002      2003-2006      2007-2010

Strominger     23             14             22              4
Maldacena      27             33             24              9 
Polchinski     21             17             11              4
Harvey,J       16             15              9              2
Duff,M         24             17              8              5
Gibbons,G      17             29             11              2
Dijkgraaf      18             11              9              7
Ooguri         31             18             13              8
Silverstein,E  16             15             16             10
Seiberg,N      19             16             22              1

Suggestions of other choices of DESY classifications? Anyone who wants can try running other search tallies. Or do a detailed examination of this one and look for excuses in the case of individual researchers? The overall picture is what impresses me.

BTW it does look to me, just glancing at their papers that Petr Horava and Erik Verinde are writing less string these days, also possibly other stars ike Giddings, Gubser, Marolf. Don't want to bother adding more names at this point. Overall picture is clear enough.

Attendance and activity at the annual Strings conference would be another indicator. Eg. the fraction of invited talks at Strings 20xx that were actually about strings/branes and extraD--aimed at unification of fundamental physics (not some offshoot application.)
Here's a recent list of locations: Toronto, Beijing, Madrid, Geneva, Rome, College Station, Uppsala.
The last one where I could sense real excitement (via web) was Toronto, although Rovelli's talk at Geneva (Strings 2008) was notable, as was the following Q/A discussion.

 

[smoit]

OK folks, here is the revised tabulation. No one has explained why there were so many papers for so long and then suddenly, in most cases, so few:

          1995-1998      1999-2002      2003-2006      2007-2010

Strominger     23             14             22              4
Maldacena      27             33             24              9 
Polchinski     21             17             11              4
Harvey,J       16             15              9              2
Duff,M         24             17              8              5
Gibbons,G      17             29             11              2
Dijkgraaf      18             11              9              7
Ooguri         31             18             13              8
Silverstein,E  16             15             16             10
Seiberg,N      19             16             22              1

Suggestions of other choices of DESY classifications?

As has been pointed out to you on several occasions, the most glaring problem with your estimate is that you are operating in terms of absolute numbers without taking into account the sample size. Also, your keyword search does not take into account the change of perspective of what is and what is not a string related paper. People have already given you many examples of papers, which your phony "core string" criterion failed to take into account, e.g. http://www-library.desy.de/cgi-bin/spiface/find/hep/wwwtopics?key=7529236"
List of keywords assigned to the paper: Chiral symmetry breaking and intersecting D-brane systems

symmetry breaking, chiral
D-brane
holography
Jona-Lasinio-Nambu model
Gross-Neveu model

 

[marcus]

So go ahead and run your own search tally for those ten people with what you think are better terms! All the defensiveness and denial is amusing---anyone who takes an unbiased look can see the program is in trouble (with top people devoting more of their effort to straying off in different directions).

The interesting thing, I think, is WHY. I suspect that a lot of it has to do wth the 2003 KKLT paper and Susskind's panicky reaction. KKLT failed to deal with positive cosmo constant in a satisfactory way and hung the 10^{too many} vacuums nightmare out in plain view. Since then, things have pretty much gone to College Station .

 

[fzero]

So go ahead and run your own search tally for those ten people with what you think are better terms! All the defensiveness and denial is amusing---anyone who takes an unbiased look can see the program is in trouble (with top people devoting more of their effort to straying off in different directions).

Can you explain to us what different directions are being strayed to, why they are so different from what is considered string theory proper, and how the topics are unusual given the past research interests of the physicists involved?

 

[marcus]

Can you explain to us what different directions are being strayed to, why they are so different from what is considered string theory proper, and how the topics are unusual given the past research interests of the physicists involved?

Sure. Bear in mind that I am primarily interested in the (no prior geometry) QG program with only moderate curiosity about string research. I started watching around 2003 and was impressed by the KKLT paper at that time and Susskind's "anthropic" fit that occurred a few months later. A lot of rumpus on Usenet at the time. This may tend to color my understanding---so I'd like to hear other explanations.

As a retired mathematician, I perhaps see the Langlands Program differently from you. In 2006 I attended 3 ninety-minute lectures by Witten on his main research interest at the time. He was hosted by the Math and Physics departments here. He did not mention String at all, and gave a one sentence reply at the end of the last lecture when it came up in a question from audience.

As a string researcher you may have a differerent perspective in which String is a big central prominence and everything relates to it in some way. So you may perhaps NOT see the part-time dispersion of interest that I see.

But anyway you wanted some examples of top people straying off from the main program (I haven't suggested this was full-time, just a significant part-time reallocation of research activity.)

Erik Verlinde is one case.
Petr Horava is another.

And you wanted me to describe the directions. Well in Verlinde and Horava case it is about gravity in 4D There are no string/branes in the picture, and no compactified extraD.
There is no drive for unification. The emphasis is on actually understanding gravity.

That same general direction characterized Witten's talk at Strings 2007. No strings/branes/extraD but rather a focus on understanding gravity. Then he declined to give a talk at Strings 2008, and again in 2009 only gave a public lecture which was not about string. As I recall his Strings 2010 talk was in a direction similar to 2007---low dimension gravity---but I would have to check. I think now, out of simple loyalty, he HAS to say nice things in public and give a more stringy paper at Uppsala. We'll see when the program comes out. Senior people have to give some moral support to the community, in the present situation.

A lot of the talks at Strings 2010 were spinoff to condensed matter or QCD. Finding uses for the math. But one misses the old TOE-spirit. The zeal and enthusiasm of 10 years ago.

 

[fzero]

As a retired mathematician, I perhaps see the Langland Program differently from you. In 2006 I attended 3 ninety-minute lectures by Witten on his main research interest at the time. He was hosted by the Math and Physics departments here. He did not mention String at all, and gave a one sentence reply at the end of the last lecture when it came up in a question from audience.

As a string researcher you may have a differerent perspective in which String is a big central prominence and everything relates to it in some way. So you may perhaps NOT see the part-time dispersion of interest that I see.

I explained in post #24 that the only Langlands papers that were assigned as string papers were the ones where the abstract made clear that results from mirror symmetry were being used. I don't recall how many this was, perhaps 2 or 3, but the majority were assigned to pure QFT in the absence of a detailed reading.

But anyway you wanted some examples of top people straying off from the main program (I haven't suggested this was full-time, just a significant part-time reallocation of research activity.)

Erik Verlinde is one case.
Petr Horava is another.

And you wanted me to describe the directions. Well in Verlinde and Horava case it is about gravity in 4D There are no string/branes in the picture, and no compactified extraD.
There is no drive for unification. The emphasis is on actually understanding gravity.

These authors are not in your table that you keep posting. What about those authors?

That same general direction characterized Witten's talk at Strings 2007. No strings/branes/extraD but rather a focus on understanding gravity. Then he declined to give a talk at Strings 2008, and again in 2009 only gave a public lecture which was not about string. As I recall his Strings 2010 talk was in a direction similar to 2007---low dimension gravity---but I would have to check. I think now, out of simple loyalty, he HAS to say nice things in public and give a more stringy paper at Uppsala. We'll see when the program comes out. Senior people have to give some moral support to the community, in the present situation.

A lot of the talks at Strings 2010 were spinoff to condensed matter or QCD. Finding uses for the math. But one misses the old TOE-spirit. The zeal and enthusiasm of 10 years ago.

Witten has consistently published and publicized non-string work in QFT and mathematical applications. The central result of my analysis was that the ratio of string to nonstring physics hasn't changed significantly over the last six years. Witten's most recent paper is undeniably a string paper, even if it is not a phenomenological one.

As I said before, most string theory papers dating back as far as 1995 and even beyond are not about direct ToE topics. I also said that this was not unusual in physics, one often studies less realistic models to develop techniques and uncover clues about universal behavior. I cited lattice gauge theory as another example where you would find a bulk of papers that were less about the real world and more about methods and drawing lessons.

 

[marcus]

Verlinde and Horava are useful examples to illustrate where I think the program is drifting.
If you find it for some reason inconsistent that they are not on the table, I will add them.

It is hard to see a direction in the current program. One possible perspective is that M-theory has failed to emerge as a definite formulated theory. M stands for Missing, I guess. And so far the understanding of gravity appears dependent on prior geometry or some dubious excessively elaborate structure. This is not to criticize the program but to try to get an idea of directions in current research.

I see Strominger, for example, studying black holes in ordinary 3D or 4D. Talking about this at Strings 2010.
Weinberg used to do string and thought highly of it at one time, he gave an invited talk at Strings 2010 which was on a non-string approach to QG that does not involve extraD. The asymptotic safety approach.

Since I am not an insider I can only speculate as to how it looks from within the community but one suspicion I have is that people like Strominger Horava Verlinde Weinberg see a deficiency that they are trying in various ways to remedy--just getting ordinary 4D quantum gravity right.

Maybe you would like to analyze the research output of some of these people. That would be very helpful!

          1995-1998      1999-2002      2003-2006      2007-2010

Strominger     23             14             22              4
Maldacena      27             33             24              9 
Polchinski     21             17             11              4
Harvey,J       16             15              9              2
Duff,M         24             17              8              5
Gibbons,G      17             29             11              2
Dijkgraaf      18             11              9              7
Ooguri         31             18             13              8
Silverstein,E  16             15             16             10
Seiberg,N      19             16             22              1
Horava          6              5              4              6
Verlinde,E     10              7              7              2

http://www-library.desy.de/cgi-bin/spiface/find/hep/www?rawcmd=FIND+A+Verlinde%2C+E+AND+%28DK+STRING+MODEL+OR+DK+MEMBRANE+MODEL%29+AND+DATE+%3E+1994+AND+DATE+%3C+1999&FORMAT=www&SEQUENCE=

http://www-library.desy.de/cgi-bin/spiface/find/hep/www?rawcmd=FIND+A+Verlinde%2C+E+AND+%28DK+STRING+MODEL+OR+DK+MEMBRANE+MODEL%29+AND+DATE+%3E+1998+AND+DATE+%3C+2003&FORMAT=www&SEQUENCE=

http://www-library.desy.de/cgi-bin/spiface/find/hep/www?rawcmd=FIND+A+Verlinde%2C+E+AND+%28DK+STRING+MODEL+OR+DK+MEMBRANE+MODEL%29+AND+DATE+%3E+2002+AND+DATE+%3C+2007&FORMAT=www&SEQUENCE=

http://www-library.desy.de/cgi-bin/spiface/find/hep/www?rawcmd=FIND+A+Verlinde%2C+E+AND+%28DK+STRING+MODEL+OR+DK+MEMBRANE+MODEL%29+AND+DATE+%3E+2006+AND+DATE+%3C+2011&FORMAT=www&SEQUENCE=

Suggestions of other choices of DESY classifications? Anyone who wants can try running other search tallies. Or do a detailed examination of this one and look for excuses in the case of individual researchers? The overall picture is what impresses me.

 

[suprised]

Do you really believe what you write or is this just provoction, marcus?

 

[atyy]

It is true that Stominger's work is arguably not "pure string", as he himself repeatedly says. However, the non "pure string" part goes all the way back to 1997. http://arxiv.org/abs/hep-th/9712251

Furthermore, this result was first illuminated by Strominger and Vafa by strings. So if you want to define there to be a decrease in string research, it is because strings has become accepted physics, on which current quantum gravity research is built.

 

[marcus]

Do you really believe what you write or is this just provoction, marcus?

Of course I believe. I also credit what you have written (except when it is an attack or misrepresentation of me). I find some of your posts enlightening---they seem exceptionally well thought-out and have the ring of true insight. Really. This is not flattery.

Much of what we write is honest suspicion or conjecture. Perhaps your word "believe" is too strong. We cannot know for sure, or make unqualified statements. We are exploring what may be possible causes. I try to get to the truth and when I cannot be certain I qualify and say "may" or "could".

You have sometimes said that I misinterpret or misquote from your posts. I don't think so.

One thing you have suggested is that it is wrong to think of string in terms of compactified extra spatial dimensions.

You have indicated that you think the program may potentially have been damaged by taking wrong turns.

THIS RINGS TRUE TO ME. I also respect your insight as an insider of 25 years experience in string research. In fact much of my thinking in this thread is influenced directly by what you have said.

You have said that many in the program have gotten away from thinking in terms of extra spatial dimensions.

Again this rings true. It agrees with what I, as an occasional outside observer, have noticed hints of. And it makes sense, if those people realized it was wrong to think in those naive terms they would naturally want to change to a different perspective.
Moreover this is only part of the more complete message you've conveyed.

I think many of the top-level people have probably gotten away from thinking of reality as being composed of branes and strings, or having a requirement of supersymmetry.

So I am interested in what direction the program will take. Your ideas, as they come through, may be very helpful and informative about this.

 

[element4]

I have just read all 115 posts on this thread, and let me declare myself as a completely unbiased person in this topic, with no particular interest or agenda wrt. LQG vs string theory.

What strikes me the most, is Marcus' failure in understanding the critic of other people! His "analysis" regarding "decline of interest in string theory", is clearly oversimplified and gives a unreliable picture of the situation, as fzero showed by performing a more correct and detailed analysis. Furthermore in post #46, fzero demonstrated how flawed Marcus' analysis is by using the key word "quark model". In #47 Marcus replies

Yes perhaps this indicator is dreadfully flawed . We can still see what we make of it nonetheless.

which to me indicates that Marcus is more interested in pushing his own agenda, rather than performing a serious analysis. Furthermore others have been given many very good reasons why this simple analysis may give wrong results, but Marcus keep ignoring them and keep copy/pasting the same stuff in all his posts.

I really believe that a person like Marcus can be very dangerous for young students trying to get an Idea about what the current situation of QG research is. There is an incredible amount of LQG biased posts in this forum, almost entirely due to Marcus.

I am really glad I am not in the field of QG. It seems that lack of experimental input can make science into some sort of religion for some people, its mainly in these field people like Lobus Motl and Marcus can exist.

 

[marcus]

I must say that I feel rebuked and abashed to be compared with Lubos Motl. I don't seriously think that indicator is "dreadfully flawed" or that anyone rough informal indicator matters all that much. Fzero only analysed one person out of a dozen or so, and he did not contradict my finding that the number of that person's string papers had declined (even with his broader definition.)

I think the String program is in deep trouble going back maybe to around 2003 or so. I'd like to know other people's views on this. What reasons for it they can think of. One should be able to talk openly about this.

So what would you propose that I do? I don't want to have to argue all the time with people who simply deny that there is a problem.

I do want to explore the underlying reasons the program is in trouble.

On the other hand I don't want to seem like a Motl to people like you (who seem to be of good faith and not to have an ax to grind).

 

[marcus]

BTW if anyone is interested here is a link to Strings 2011.
http://www-conference.slu.se/strings2011/programme_NEW.html

This gives a list of the confirmed invited speakers (scroll down to it).
As the conference takes shape, and titles of talks are listed, this will afford a window on the current state of the string program. So it should be interesting to watch.

Here are links to the websites of previous string conferences:
http://www-conference.slu.se/strings2011/links.html
As an example of the kind of information, in this case from Strings 2004:
http://strings04.lpthe.jussieu.fr/speakers_list.php
http://strings04.lpthe.jussieu.fr/participants.php (477)
It's a way to look back into the past.
And something more recent:
http://ph-dep-th.web.cern.ch/ph-dep-th/content2/workshops/strings2008/?site=content/talks.html
http://indico.cern.ch/confRegistrantsDisplay.py/list?confId=21917

 

[negru]

Marcus, all the info and "data" you use to evaluate the level of interest in string theory could just as well be interpreted as evidence of some degenerative brain disease affecting all string theorists.

If you want to probe the interest in string theory, go around and ask string theorists if they are still interested in string theory. That's the only way to do it.

 

[marcus]

Marcus, all the info and "data" you use to evaluate the level of interest in string theory could just as well be interpreted as evidence of some degenerative brain disease affecting all string theorists.

If you want to probe the interest in string theory, go around and ask string theorists if they are still interested in string theory. That's the only way to do it.

Heh heh, that's funny! Actually what I'm looking for are the signs of coherent activity/vitality/direction in the field. The kind of thing that gives excitement and interest like what those of us who were around remember from 10 years ago. I don't know what interest means unless it shows in people's behavior.

Maybe I should have called it something else like "energy" or "focus", but I used the word interest.

So when I measure interest I look at the numbers of papers top people are producing AND how much they are cited by the rest of the community----not only activity but also an idea of the value of that activity.

There are a dozen signs that this is down--you really don't even need numbers. E.g. just listen to David Gross who for several years seemed to be repeating every chance he had "We don't know what string theory is!" and "We need a fundamentally new idea."

But if you like numbers you can look at the annual conference participants, for example:
Strings 2004 had 477 participants, Strings 2008 had 400,...want to fill in the blanks?

Strings 2004 477
Strings 2008 400
Strings 2010 193

I remember at Strings 2008 David Gross referred to the annual Strings conference as "the canary in the mine" and was urging people to come up with good ideas for future conferences. I like him, he is honest at times and is sincerely concerned with the health and direction of the field---he doesn't walk around with his eyes shut or wearing rosy glasses. He has often given the summary talk at the end of Strings, but this year he will give the opener.