Causes of loss of interest in String program


by marcus
Tags: loss, program, string
marcus
marcus is offline
#37
Mar31-11, 01:06 PM
Astronomy
Sci Advisor
PF Gold
marcus's Avatar
P: 22,796
Quote Quote by atyy View Post
marcus, do you think ground-breaking work can really be "managed"? ...
No, but visionary leadership certainly can help.

- string remains the first field of study for anyone interested in quantum gravity:
1)integrability in AdS/CFT
2)ABJM
3)twistors
4)many, many standard model like constructions
That's not the topic issue of the thread, I'm asking why and what can be done to revitalize the program. But I'm not convinced of your general statement that anyone interested in QG shoud first study String.

I think it's fairly clear that the main testing arena for QG (as empirical science) is early universe cosmology and possibly related stuff like gammaray burst observations. If someone is interested in QG they might do well to go to Penn State's Institute for Gravitation and the Cosmos and talk to Abhay Ashtekar. They might do well to learn some cosmology and quantum cosmology. And also get some handle on the current and projected job terrain.

But that's not the topic. I'm really looking for constructive ideas about the direction of the program.
marcus
marcus is offline
#38
Mar31-11, 01:13 PM
Astronomy
Sci Advisor
PF Gold
marcus's Avatar
P: 22,796
YUK! Cosmic strings! If DESY catalogs cosmic strings papers as "string model" maybe their keywords are, in fact, deeply flawed. Thanks for pointing that out.

I will see if I can find some work-around for that.
atyy
atyy is offline
#39
Mar31-11, 01:14 PM
Sci Advisor
P: 8,004
Quote Quote by marcus View Post
That's not the topic issue of the thread, I'm asking why and what can be done to revitalize the program. But I'm not convinced of your general statement that anyone interested in QG shoud first study String.

I think it's fairly clear that the main testing arena for QG (as empirical science) is early universe cosmology and possibly related stuff like gammaray burst observations. If someone is interested in QG they might do well to go to Penn State's Institute for Gravitation and the Cosmos and talk to Abhay Ashtekar. They might do well to learn some cosmology and quantum cosmology. And also get some handle on the current and projected job terrain.

But that's not the topic. I'm really looking for constructive ideas about the direction of the program.
No, that is the topic. You claim it needs to be "revitalized". Of course everyone likes a breakthrough, but my claim is that your assumption that string isn't vital is wrong, hence any discussion that follows from it will be similarly flawed.
fzero
fzero is offline
#40
Mar31-11, 01:22 PM
Sci Advisor
HW Helper
PF Gold
P: 2,606
Quote Quote by marcus View Post
YUK! Cosmic strings! If DESY catalogs cosmic strings papers as "string model" maybe their keywords are, in fact, deeply flawed. Thanks for pointing that out.

I will see if I can find some work-around for that.
No, the cosmic string papers did not show up in the DESY keyword searches, they show up when I remove the keywords. The problem is the other 7 string papers that are not caught by your keyword search. One problem is that it is quite easy to write a paper on string theory without ever typing "string model" in the paper....
marcus
marcus is offline
#41
Mar31-11, 01:33 PM
Astronomy
Sci Advisor
PF Gold
marcus's Avatar
P: 22,796
No, the cosmic string papers did not show up in the DESY keyword searches,...
What a relief! My faith in DESY librarians is to some extent restored!

So Atyy you argue there is no pressing need for some new vision or direction in the program. On the other hand, I still suspect that there is. But I am really more interested in what can be done. So I will just acknowledge your view that nothing needs to be done, and move on.

Let's look at Suprised's first four "wrong turns" and try to figure out how they could be fixed---so the core String unification program could get back on track (if it is not already on track .)

===quote Suprised post #553===
I guess there were many potentially wrong turns - at least in the sense of bias towards certain ways of thinking about string theory. Here a partial list of traditional ideas/beliefs/claims that have their merits but that potentially did great damage by providing misleading intuition:
  1. - That geometric compactification of a higher dimensional theory is a good way to think about the string parameter space
  2. - That perturbative quantum and supergravity approximations are a good way to understand string theory
  3. - That strings predict susy, or have an intrinsic relation to it (in space-time)
  4. - That strings need to compactify first on a CY space and then susy is further broken. That's basically a toy model but tends to be confused with the real thing.
...
==endquote==

Atyy, to me this provide a glimpse of renewal in the String unification program---which so far has interesting math but no one definite physical theory. You may disagree and say it has one definite physical theory and does not need renewal etc etc. But let's consider anyway how these things translate into action:

1. We should stop thinking of compactified extra dimensions as real. The world is 4D.

3. String has no intrinsic relation to SUSY. We should cut loose from the SUSY dream and not waste so much research on "superstuff"

2. That goes for SUGRA ("supergravity") too. Supergravity approximations are not a good way to understand String.

2. continued...Furthermore perturbative is not a good way. Perturbative uses a fixed prior setup---including a prior geometry---which is then subjected to small variations. You regard this approach as deficient.

4. We are warned against confusing Calabi-Yau spaces with reality.

========================

Can anybody else come up with some ideas for programatic reconsideration and reform?
atyy
atyy is offline
#42
Mar31-11, 01:48 PM
Sci Advisor
P: 8,004
suprised's point #1 is in fact contained in David Tong's 2009 basic string theory notes, where it is given as a single sentence warning, which is where I learnt about it. What I further learnt from fzero and suprised was that in fact this warning can be useful in generating lots and lots of realistic models. Clearly the professionals have known about this for quite some time, and it's even trickled down to advanced undergrad and master level courses two years ago. Probably the only people who don't know about it in 2011 are biologists like me who read pop science books.
PAllen
PAllen is online now
#43
Mar31-11, 02:08 PM
Sci Advisor
PF Gold
P: 4,860
Purely as a point of interest, I note no change in percentage of papers by Michael Duff or Brian Greene that focus on 'core string theory'. It appears the same is true for the 'next generation' of theorists, e.g. Eva Silverstein (though, with her proclivity for humorous titles and my lack of knowledge of the field, it is a little hard for me to classify some of her papers).
fzero
fzero is offline
#44
Mar31-11, 02:26 PM
Sci Advisor
HW Helper
PF Gold
P: 2,606
Quote Quote by atyy View Post
suprised's point #1 is in fact contained in David Tong's 2009 basic string theory notes, where it is given as a single sentence warning, which is where I learnt about it. What I further learnt from fzero and suprised was that in fact this warning can be useful in generating lots and lots of realistic models. Clearly the professionals have known about this for quite some time, and it's even trickled down to advanced undergrad and master level courses two years ago. Probably the only people who don't know about it in 2011 are biologists like me who read pop science books.
But as I and Haelfix have been pointing out in the other thread, the nongeometric models are likely to all appear as special points in CY models. The best that can be said is that these points are often easier to describe from the worldsheet perspective than other points in the moduli space of the same CY. It's not obviously correct to conclude that these are new models, not already contained in the space of CY compactifications.

Now what is a rather philosophic point is whether one should consider CY directions as spacetime dimensions or just as useful descriptions of the internal degrees of freedom. There are many situations where the geometric interpretation is very useful, since it allows us to use a certain amount of intuition, but it's not necessary per se. It's also almost unavoidable to use that intuition in trying to explain things to nonexperts, because "internal degrees of freedom" is an abstraction that obfuscates the construction to a high degree.

An added complication, familiar to experts, but not otherwise well known, is that the geometric interpretation of CY models is also itself really only valid over a corner of moduli space known as the large volume limit. That is, the emergence of a smooth geometry occurs when the volumes of the CY are large compared to the string scale. When the CY itself or submanifolds within the CY are of size order one in string units, both stringy and QG corrections to the smooth geometry are expected to be large. In some cases, the physics at points like this are understood nonperturbatively in terms of topology change and string dualities. In general one expects that a complete description of small volume requires the correct nonperturbative degrees of freedom. These may or may not be geometric. We have seen examples of both in the cases of matrix theory and AdS/CFT.

So I think that a string theorist needs to be aware of these issues. I have not seen evidence that successful ones are not intimately aware of the approximations and limitations attached to any particular string description.

I will comment a bit in the future on the point "3. String has no intrinsic relation to SUSY. We should cut loose from the SUSY dream and not waste so much research on "superstuff"" This statement may actually be true, but I think most of the evidence still favors an intimate relationship between spacetime SUSY and consistency of the superstring. It would actually take more than a few words to do justice to important work like that of Gato-Rivera and Schellekens who would like to argue otherwise.
marcus
marcus is offline
#45
Mar31-11, 02:35 PM
Astronomy
Sci Advisor
PF Gold
marcus's Avatar
P: 22,796
Quote Quote by fzero View Post
... One problem is that it is quite easy to write a paper on string theory without ever typing "string model" in the paper....
I think I see your train of thought, there. Anyway we are now addressing Suprised's points and it's a more engrossing side to things.

Quote Quote by PAllen View Post
Purely as a point of interest, I note no change in percentage of papers by Michael Duff or Brian Greene that focus on 'core string theory'. It appears the same is true for the 'next generation' of theorists, e.g. Eva Slilverstein (though, with her proclivity for humorous titles and my lack of knowledge of the field, it is a little hard for me to classify some of her papers).
Thanks PAllen, I appreciate an more data like this that helps fill out the picture. You understand that by "core String" (which may have been a misnomer) I am only looking at the change over time of what Spires gives for dk "string model" and dk "membrane model".
A better classifier might be to call them DKSM papers. Maybe I'll add Michael Duff...

We are all more interested in the substantive physics correlatives here, so I'll say no more about that for now.
fzero
fzero is offline
#46
Mar31-11, 02:50 PM
Sci Advisor
HW Helper
PF Gold
P: 2,606
Quote Quote by marcus View Post
Thanks PAllen, I appreciate an more data like this that helps fill out the picture. You understand that by "core String" (which may have been a misnomer) I am only looking at the change over time of what Spires gives for dk "string model" and dk "membrane model".
A better classifier might be to call them DKSM papers.

We are all more interested in the substantive physics correlatives here, so I'll say no more about that for now.
You still refuse to explain what you think DKSM papers are representative of. My claim is that they are not representative of any trend in research interest of individuals and I have given substantive arguments in support.

Let's do another experiment. We'll search DESY for keyword "quark model":

2007-2010
http://www-library.desy.de/cgi-bin/s...=www&SEQUENCE=
462 papers

2003-2006
http://www-library.desy.de/cgi-bin/s...=www&SEQUENCE=
28 papers

1999-2002
http://www-library.desy.de/cgi-bin/s...=www&SEQUENCE=
15 papers

How can we explain the cause of this profound new interest in quark models? Conversely, why were researchers not interested in the quark model from 1999-2006?

My answer is obviously that the keyword search itself is flawed.
marcus
marcus is offline
#47
Mar31-11, 03:24 PM
Astronomy
Sci Advisor
PF Gold
marcus's Avatar
P: 22,796
Yes perhaps this indicator is dreadfully flawed . We can still see what we make of it nonetheless.
PAllen kindly suggested looking at Michael Duff papers. So in a free moment I added Michael Duff and Gary Gibbons.
Probably this should be called the DESY "string" and "membrane" timeseries. For lack of better term it counts the DKSM (DESY keyword "string model" and "membrane model") papers over the past sixteen years 1995-2010. We look for differences and changes.

          1995-1998      1999-2002      2003-2006      2007-2010
Witten         38             29              9              5
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
It looks like Witten shifted interest (in DKSM terms) sooner than some others. Here are the Spires links so you can make your own counts if you like, using DESY keywords "string model" and "membrane model". Just put in a different name instead of Witten and repeat the search.

http://www-library.desy.de/cgi-bin/s...=WWW&SEQUENCE= (38)

http://www-library.desy.de/cgi-bin/s...=WWW&SEQUENCE= (29)

http://www-library.desy.de/cgi-bin/s...=WWW&SEQUENCE= (9)

http://www-library.desy.de/cgi-bin/s...=WWW&SEQUENCE= (5)
===========================
As a reminder that citation counts could also be significant. In case anyone is coming in new to the thread I will include this

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/...1/annual.shtml (twelve)
http://www.slac.stanford.edu/spires/...3/annual.shtml (six)
http://www.slac.stanford.edu/spires/...5/annual.shtml (two)
http://www.slac.stanford.edu/spires/...7/annual.shtml (one)
http://www.slac.stanford.edu/spires/...9/annual.shtml (one)

A paper is counted as recent here if it appeared in the past five years.
atyy
atyy is offline
#48
Mar31-11, 03:25 PM
Sci Advisor
P: 8,004
Quote Quote by fzero View Post
But as I and Haelfix have been pointing out in the other thread, the nongeometric models are likely to all appear as special points in CY models. The best that can be said is that these points are often easier to describe from the worldsheet perspective than other points in the moduli space of the same CY. It's not obviously correct to conclude that these are new models, not already contained in the space of CY compactifications.

Now what is a rather philosophic point is whether one should consider CY directions as spacetime dimensions or just as useful descriptions of the internal degrees of freedom. There are many situations where the geometric interpretation is very useful, since it allows us to use a certain amount of intuition, but it's not necessary per se. It's also almost unavoidable to use that intuition in trying to explain things to nonexperts, because "internal degrees of freedom" is an abstraction that obfuscates the construction to a high degree.

An added complication, familiar to experts, but not otherwise well known, is that the geometric interpretation of CY models is also itself really only valid over a corner of moduli space known as the large volume limit. That is, the emergence of a smooth geometry occurs when the volumes of the CY are large compared to the string scale. When the CY itself or submanifolds within the CY are of size order one in string units, both stringy and QG corrections to the smooth geometry are expected to be large. In some cases, the physics at points like this are understood nonperturbatively in terms of topology change and string dualities. In general one expects that a complete description of small volume requires the correct nonperturbative degrees of freedom. These may or may not be geometric. We have seen examples of both in the cases of matrix theory and AdS/CFT.

So I think that a string theorist needs to be aware of these issues. I have not seen evidence that successful ones are not intimately aware of the approximations and limitations attached to any particular string description.

I will comment a bit in the future on the point "3. String has no intrinsic relation to SUSY. We should cut loose from the SUSY dream and not waste so much research on "superstuff"" This statement may actually be true, but I think most of the evidence still favors an intimate relationship between spacetime SUSY and consistency of the superstring. It would actually take more than a few words to do justice to important work like that of Gato-Rivera and Schellekens who would like to argue otherwise.
I see, thanks. I had understood from yours and others replies in the other thread that while many non-geometric models can be reparamterized as CY compactifications, how extensive the equivalence was wasn't yet understood. Thanks also for clarifying that even the CY compactifications do not necessarily have a geometric interpretation. What are good examples of these, say in the context of AdS/CFT that you mention?
fzero
fzero is offline
#49
Mar31-11, 04:15 PM
Sci Advisor
HW Helper
PF Gold
P: 2,606
Quote Quote by marcus View Post
Yes quite possibly this indicator is dreadfully flawed. We can see what we make of it nonetheless.


Quote Quote by atyy View Post
I see, thanks. I had understood from yours and others replies in the other thread that while many non-geometric models can be reparamterized as CY compactifications, how extensive the equivalence was wasn't yet understood.
I try to be overconservative when I make statements. In this case, I was never an expert on these matters and it's been many years since I've had a reason to be familiar with them. In the case of the original Gepner models, Greene, Vafa and Warner http://inspirebeta.net/record/266451?ln=en gave the dictionary between them and CY manifolds. I tend to believe that other nongeometric theories will also be dual descriptions, but I can't be more specific. It might still be the case that the nongeometric phase is much better understood than the CY version, owing to small volume issues or whatever. If I turn up any better explanations in the literature, I'm sure that I'll share them.

Thanks also for clarifying that even the CY compactifications do not necessarily have a geometric interpretation. What are good examples of these, say in the context of AdS/CFT that you mention?
When I mentioned AdS/CFT, I was specifically referring to the nonperturbative degrees of freedom, in the sense that the CFT degrees of freedom are not geometric. The AdS/CFT actually deals with CYs that are in a large volume limit, since in the AdS5 case, our CY is a cone over a so-called Einstein-Sasaki 5-manifold, [tex]X_5[/tex]. We can still have submanifolds going singular, like the conifold point. There the geometric naively becomes singular, but the singularity is understood in terms of a tower of wrapped D-brane states becoming light. In the dual gauge theory, this picture is understood as Seiberg duality, due to Klebanov and Strassler in this context.
marcus
marcus is offline
#50
Mar31-11, 04:47 PM
Astronomy
Sci Advisor
PF Gold
marcus's Avatar
P: 22,796
heh heh, great illustration!
Obviously I don't consider it flawed when applied in this context, but I can see how you might think like that.
marcus
marcus is offline
#51
Mar31-11, 05:14 PM
Astronomy
Sci Advisor
PF Gold
marcus's Avatar
P: 22,796
I updated the list by adding Dijkgraaf and Ooguri
Quote Quote by marcus View Post
Yes perhaps this indicator is dreadfully flawed . We can still see what we make of it nonetheless.
PAllen kindly suggested looking at Michael Duff papers. So in a free moment I added Michael Duff and Gary Gibbons.
Probably this should be called the DESY "string" and "membrane" timeseries. For lack of better term it counts the DKSM (DESY keyword "string model" and "membrane model") papers over the past sixteen years 1995-2010. We look for differences and changes.
          1995-1998      1999-2002      2003-2006      2007-2010
Witten         38             29              9              5
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
=======================
Notice that for Eva Silverstein the numbers are roughly flat. Thanks to PAllen for suggesting both Silverstein and Duff. I just edited Silverstein into the sample, at his suggestion.

It strikes me that maybe the easiest thing to do is deny there is a problem, or that anything has happened. To say the DESY librarians are inconsistent/arbitrary in their tagging. To say there is some harmless explanation, or to accuse the reporter of stupidity or bias or sinister motives And then there is nothing to talk about.

We don't want to forget about citation counts, since cites to recent papers reflect the researchers' assessment of their own colleagues' current output. So this has to be factored in with numbers of papers as an indicator of value (sometimes called the "impact" of the research.) It has gone down.

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/...1/annual.shtml (twelve)
http://www.slac.stanford.edu/spires/...3/annual.shtml (six)
http://www.slac.stanford.edu/spires/...5/annual.shtml (two)
http://www.slac.stanford.edu/spires/...7/annual.shtml (one)
http://www.slac.stanford.edu/spires/...9/annual.shtml (one)

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

Sure this could conceivably all be artifacts of some harmless/meaningless circumstance.
Paper and cite counting makes no pretense of being "science". It's just the kind of thing one normally does as part of finding out what's happening in a field.

I like what Suprised, Tom, and others are doing in that other thread though. Trying to come to grips with what may be wrong in the program. Or have been wrong but is in the process of fixing itself.
PAllen
PAllen is online now
#52
Mar31-11, 09:17 PM
Sci Advisor
PF Gold
P: 4,860
Two observations:

Eva Silverstein shows no real decline (15, 16, 10)

A key thing to independently verify is a shift in general usage of keywords. Not inconsistency of librarians, but evolution of meaning of terminology. Hopefully, the librarians respond to changing concepts. Thus, a given paper may be classified differently today than in the past. Also, as research gets more specialized, the terms applied to something 'part of the unification goal' may also become more varied and specialized. Finally, what looks like a side issue may be, to its author and other experts in the field, a promising way to get at a central issue in unification.

So though I sympathize with the attempt to come up with a simple, objective metric of activity, I really doubt it can be done. There is no alternative to a fair expert in the field judging which papers are part of the effort towards string/M as a unified theory, versus applications of its techniques to other fields; with the further complication that an 'application' paper may be intended to get at a 'central' issue by an indirect route.
marcus
marcus is offline
#53
Mar31-11, 09:57 PM
Astronomy
Sci Advisor
PF Gold
marcus's Avatar
P: 22,796
PAllen, nicely put!

There is a tradeoff between a quick and dirty indicator that is easy to use to get a rough idea of something versus careful work by a guaranteed unbiased expert. And I do suspect DESY librarians of changing how they classify papers and assign tags. Probably less so with very common terms that have been in use for a long time.

I'm glad to get your suggestion of Eva Silverstein and will add her to the list up in post #51. So far I have been putting in whoever occurred to me, without looking first---no cherrypicking --and I've been suprised at the rather consistent pattern. It is helpful now to have Silverstein as an example to show that it is possible to have a flatter output of DKSM papers. Here are the Spires links again so anyone can do it. Just put in a different name instead of Silverstein, E and repeat the search.

http://www-library.desy.de/cgi-bin/s...=WWW&SEQUENCE= (16)

http://www-library.desy.de/cgi-bin/s...=WWW&SEQUENCE= (15)

http://www-library.desy.de/cgi-bin/s...=WWW&SEQUENCE= (16)

http://www-library.desy.de/cgi-bin/s...=WWW&SEQUENCE= (10)
PAllen
PAllen is online now
#54
Apr1-11, 12:12 AM
Sci Advisor
PF Gold
P: 4,860
The question of the number of recent string papers among top cited papers for recent years (which really does seem to have gone down a lot) has many possible explanations to be sorted out, and I believe all of these are actually contributing:

1) The more mature a field is, the more of its breakthrough papers will be older. There is no shortage of string papers among the 2009 cites; just that many of them are not 'recent' (I get about 9 if I don't limit to recent).

2) There actually haven't been many breakthroughs recently, despite continued effort (nothing new of the order of ADS/CFT, dualities, black hole results).

3) Other areas have become 'hot', pushing aside recent string papers (e.g. gear up to LHC and astronomy / cosmology)

4) More active participants, less reliance on superstars, in a field of unchanged significance relative to physics as a whole, will lead to fewer top cited recent papers.

Of these, only (2) is a possible problem for a research program, and only if it continues 'too long', however that might be defined.


Register to reply

Related Discussions
Letter of Interest for an REU program? Academic Guidance 5
How to find a grad program by area of interest (math) Academic Guidance 3
Self writing program and long string assignment in C Programming & Computer Science 6