- #1
- 24,753
- 794
It would be nice if we had a transcript of the very interesting KITP video of a couple of days ago (20 October) especially the remarks by David Gross about the "background independence" of string theory (by which something is meant that seems to be different from what has been usual in QG discussions).
In any case there is a chance now to clear up some confusion and see how notions of background independence differ, if they do. I strongly recommend listening to the KITP video. Here is the link to it:
http://online.itp.ucsb.edu/online/resident/johnson2/
Lee Smolin replied a few minutes ago in a post to Peter Woit's blog
===quote===
61. Lee Smolin Says:
October 22nd, 2006 at 6:26 pm
Hi, to respond to David Gross’s claim mentioned by Kris Krogh above, I have been both clear and careful in the book. For example on p 240:
“Thus many quantum-gravity theorists believe there is a deeper level of reality, where space does not exist (this is taking background independence to its logical extreme). Since string theory requires the existence of a background-independent theory to make sense, many string theorists have indicated that they agree. In a certain limited sense, if the strong form of the Maldacena conjecture (see chapter 9) turns out to be true, a nine-dimensional geometry will emerge out of a fixed three-dimensional geometry. It is thus not surprising to hear Edward Witten say, as he did in a recent talk at the Kavli Institute for Theoretical Physics at UC Santa Barbara, that “most string theorists suspect that spacetime is an ‘emergent phenomenon,’ in the language of condensed matter physics.”2
Some string theorists have finally begun to appreciate this point, and one can only hope they will follow up by studying the concrete results that have already been obtained. But in fact, most people in quantum gravity have in mind something more radical than the
Maldacena conjecture.
The starting point is nothing like geometry. What many of us in quantum gravity mean when we say that space is emergent is that the continuum of space is an illusion….”
In more technical language, what has usually been meant by background independence is that no classical metric, field or global symmetry appears in the definition of a theory. Thus, by definition, formulations of a theory with asymptotic boundary conditions are excluded. This definition, and the motivation for it is discussed many places in the literature, for example in my recent hep-th/0507235. For a discussion of why the Maldacena conjecture does not satisfy what is usually considered background independence, see pages 23 and 24. There I am responding to various discussions with string theorists including David Gross, who insisted that string theory does not need a background independent formulation, because the AdS/CFT correspondence gives string theory a non-perturbative definition.
However, in my most recent conversation with Gross, he argued instead that the AdS/CFT correspondence should be considered to satisfy a form of background independence. Recently others made the same argument to me. Their claim is that if the strong form of the AdS/CFT conjecture is true, a dynamical ten dimensional asymptotically AdS spacetime will have arisen out of a theory defined on 4d Minkowski spacetime with global Poincare symmetry. This does not satisfy the definition just given.[/color] But it is true that 6 of the ten dimensions transmuted from the space of scalar fields in the N=4 SYM theory to dimensions of space in the dual theory. In this limited sense a weak form of background independence will have been achieved. This is what I agreed to in conversation with David and others.
But, as I thought I made it clear in these discussions, there is a big difference between this and what has usually been meant. Classical GR achieves this usual meaning, as do LQG, spin foam models, causal sets, causal dynamical triangulations, and not, at least so far, string theory.
In a recent discussion, Brian Greene proposed a new terminology to straighten out the confusion. Brian proposes to use manifest background independence for what quantum gravity people and philosophers have up till now meant by background independence and background independence for the weak form satisfied by the (strong form of the) AdS/CFT conjecture.
If it helps to distinguish between a strong or manifest form of background independence and a weak form, then I can agree to use this language. But I don’t see that changing the meaning of technical terms advances the issue. Brian and others (perhaps not David) agree that string theory should have a manifest or strong background independent formulation, and that it does not yet.
By the way, each time I note that string theory does not have the property of strong or manifest background independence, this is more than anything a criticism of myself as, to my knowledge, few others took the problem of making such a formulation of string theory seriously enough to spend years working on it, as I did.
Thanks,
Lee
Ps Just in case anyone thinks it is plausible that I and other people in LQG be considered crackpots, they can consult my cv which is at www.thetroublewithphysics.com. I didn’t listen to the tape, but if someone actually said that they should be ashamed of themselves.
Indeed, what is remarkable to me is that no one in this or, to my knowledge, any other discussion of my book says something like, “On page x of Chapter 12, Smolin says A and that is false for this reason. This is not just because some string theorists have not read the book, because if no one at KITP has, I know of others who have read it in detail, and if they had found errors of fact they would have said so. I wouldn’t be surprised if this had happened, because in the chapters leading up to the evaluation of string theory I give in Chapter 12 a large body of technical results is summarized in non-technical language. It is very hard to do this and not get something wrong, even in one’s own field, and I worked very hard, and checked and double checked with experts to be sure of having the facts right. Of course, if anybody finds such an error I would be grateful.
===endquote===
the original L.S. comment is here:
http://www.math.columbia.edu/~woit/wordpress/?p=475#comment-18072
In any case there is a chance now to clear up some confusion and see how notions of background independence differ, if they do. I strongly recommend listening to the KITP video. Here is the link to it:
http://online.itp.ucsb.edu/online/resident/johnson2/
Lee Smolin replied a few minutes ago in a post to Peter Woit's blog
===quote===
61. Lee Smolin Says:
October 22nd, 2006 at 6:26 pm
Hi, to respond to David Gross’s claim mentioned by Kris Krogh above, I have been both clear and careful in the book. For example on p 240:
“Thus many quantum-gravity theorists believe there is a deeper level of reality, where space does not exist (this is taking background independence to its logical extreme). Since string theory requires the existence of a background-independent theory to make sense, many string theorists have indicated that they agree. In a certain limited sense, if the strong form of the Maldacena conjecture (see chapter 9) turns out to be true, a nine-dimensional geometry will emerge out of a fixed three-dimensional geometry. It is thus not surprising to hear Edward Witten say, as he did in a recent talk at the Kavli Institute for Theoretical Physics at UC Santa Barbara, that “most string theorists suspect that spacetime is an ‘emergent phenomenon,’ in the language of condensed matter physics.”2
Some string theorists have finally begun to appreciate this point, and one can only hope they will follow up by studying the concrete results that have already been obtained. But in fact, most people in quantum gravity have in mind something more radical than the
Maldacena conjecture.
The starting point is nothing like geometry. What many of us in quantum gravity mean when we say that space is emergent is that the continuum of space is an illusion….”
In more technical language, what has usually been meant by background independence is that no classical metric, field or global symmetry appears in the definition of a theory. Thus, by definition, formulations of a theory with asymptotic boundary conditions are excluded. This definition, and the motivation for it is discussed many places in the literature, for example in my recent hep-th/0507235. For a discussion of why the Maldacena conjecture does not satisfy what is usually considered background independence, see pages 23 and 24. There I am responding to various discussions with string theorists including David Gross, who insisted that string theory does not need a background independent formulation, because the AdS/CFT correspondence gives string theory a non-perturbative definition.
However, in my most recent conversation with Gross, he argued instead that the AdS/CFT correspondence should be considered to satisfy a form of background independence. Recently others made the same argument to me. Their claim is that if the strong form of the AdS/CFT conjecture is true, a dynamical ten dimensional asymptotically AdS spacetime will have arisen out of a theory defined on 4d Minkowski spacetime with global Poincare symmetry. This does not satisfy the definition just given.[/color] But it is true that 6 of the ten dimensions transmuted from the space of scalar fields in the N=4 SYM theory to dimensions of space in the dual theory. In this limited sense a weak form of background independence will have been achieved. This is what I agreed to in conversation with David and others.
But, as I thought I made it clear in these discussions, there is a big difference between this and what has usually been meant. Classical GR achieves this usual meaning, as do LQG, spin foam models, causal sets, causal dynamical triangulations, and not, at least so far, string theory.
In a recent discussion, Brian Greene proposed a new terminology to straighten out the confusion. Brian proposes to use manifest background independence for what quantum gravity people and philosophers have up till now meant by background independence and background independence for the weak form satisfied by the (strong form of the) AdS/CFT conjecture.
If it helps to distinguish between a strong or manifest form of background independence and a weak form, then I can agree to use this language. But I don’t see that changing the meaning of technical terms advances the issue. Brian and others (perhaps not David) agree that string theory should have a manifest or strong background independent formulation, and that it does not yet.
By the way, each time I note that string theory does not have the property of strong or manifest background independence, this is more than anything a criticism of myself as, to my knowledge, few others took the problem of making such a formulation of string theory seriously enough to spend years working on it, as I did.
Thanks,
Lee
Ps Just in case anyone thinks it is plausible that I and other people in LQG be considered crackpots, they can consult my cv which is at www.thetroublewithphysics.com. I didn’t listen to the tape, but if someone actually said that they should be ashamed of themselves.
Indeed, what is remarkable to me is that no one in this or, to my knowledge, any other discussion of my book says something like, “On page x of Chapter 12, Smolin says A and that is false for this reason. This is not just because some string theorists have not read the book, because if no one at KITP has, I know of others who have read it in detail, and if they had found errors of fact they would have said so. I wouldn’t be surprised if this had happened, because in the chapters leading up to the evaluation of string theory I give in Chapter 12 a large body of technical results is summarized in non-technical language. It is very hard to do this and not get something wrong, even in one’s own field, and I worked very hard, and checked and double checked with experts to be sure of having the facts right. Of course, if anybody finds such an error I would be grateful.
===endquote===
the original L.S. comment is here:
http://www.math.columbia.edu/~woit/wordpress/?p=475#comment-18072
Last edited:
The web certainly provides interesting insights into different cultures.
