Friday, April 15, 2016

Strominger's "new" conservation laws claimed to be "old", application problematic

Sebastian was the first one who saw a new paper
Asymptotic structure of electrodynamics revisited
by Andřej Herdegen in Krakow, Poland. The paper claims that there are some serious generalized problems with the latest "big program" by Andy Strominger and his pals (search e.g. for soft Strominger on this blog).

First of all, Herdegen claims that the bulk of the classical results by Strominger talking about new conservation laws linked to the null infinity (at least those referring to the electromagnetic field) have been known for decades, from the work by Bramson in the late 1970s, Ashtekar and Streubel in the early 1980s, and (in a more complete form) Herdegen himself in the mid 1990s.

These papers – whose content is reviewed in Herdegen's new paper – aren't easily available e.g. on the arXiv and I am not excited by the idea that I would have to study them and compare them to Andy's papers but I suspect that Herdegen knows what he's talking about. Unless someone will show me some evidence that I am wrong, I will assume that it's much more likely that Herdegen is right, the results exist, and Andy was just unaware of them for some reason; than that Herdegen is wrong and seeing a mirage.

It seems plausible to me that Andy is unfamiliar with some claims in the "classical GR" literature despite the fact that Andy Strominger is surely one of the greatest "general relativists" among the string theorists in the world. That part of the literature is almost certainly extensive and most string theorists were assuming almost all of it to be uninteresting for decades. So we could have overlooked many things that were not completely uninteresting.

In his three single-author papers and one paper with Rajzner, Herdegen also claims to have found some other results overlapping with Strominger's project between 1998 and 2011.

Aside from these issues of "credit", Herdegen discusses some problems with the interpretations that Andy tried to give to these results. There seems to be some overlap with the criticisms of Hawking+Perry+Strominger on Distler's blog but the criticisms aren't identical. If I am translating it to my language correctly, Herdegen mainly complains that the new symmetry generators (new proposed conserved charges) don't act within the physical Hilbert space of gauge-invariant states which would surely produce at least some problems for the interpretation that there are new conserved charges.

I am afraid that I won't study these matters in detail and I do believe Herdegen's claims. As I have repeatedly written in previous blog posts, Andy has basically entered a low-tech sub-industry of theoretical physics that has been studied by lots of non-stringy people for decades. If you focus on classical field theory; and/or if you assume that all of quantum gravity may be described as a full-fledged local quantum field theory on a classical background, or that the Bekenstein-Hawking entropy may be reduced to degrees of freedom that are "really" localized at the event horizon, then you are deciding to do a kind of "obvious" (and rather non-quantum, non-stringy) research that has consumed hundreds of thousands of man-hours in recent decades. Yes, they were mostly man-hours, almost no woman-hours, and they were man-hours by people who didn't interact with the string community too much.

It seems plausible that some new interesting results that haven't been found and are still waiting to be discovered – or have been found but haven't been appreciated by string theorists – exist. But I don't really believe that such things are sufficient for solving some dilemmas in quantum gravity (or in the black hole information puzzle, to be specific). As far as I can see, none of these "basically classical" results may be ingenious enough to compensate for the lethal defects resulting from the misguided physical assumption that "all degrees of freedom, including the black hole entropy, are strictly localized within a pre-existing basically classical spacetime".

This assumption is simply wrong – it has been shown wrong by many advances in quantum gravity and string theory, including those in which Andy has played an essential role. In any large enough and flat enough piece of spacetime, locality ultimately emerges and becomes exact for the infinitely large flat Minkowski background etc. But the spacetime and the locality determining its character is emergent, the emergence becomes important for smaller black holes or for the ability of the information to get out of the black hole and get imprinted to the Hawking radiation, and one can't avoid the discussion of some degrees of freedom that can't be straightforwardly localized to points or small regions of a pre-existing classical spacetime background. And the black hole entropy simply isn't composed of any "classically describable" degrees of freedom that would be localized near the event horizon and "certainly" not elsewhere. I am confident that despite the confusion that remains, all these basic theses have been firmly established and Andy is trying to return these portions of the quantum gravity research by some 25 years to the past.

You know, even though I do feel that light cones are important and new crucial realizations about quantum gravity that depend on them are waiting to be found, my relative lack of excitement about the new Strominger program could have been explained by my feeling that "it is a similar stuff that folks like Ashtekar were doing for quite some time". I wasn't terribly interested in those papers of Ashtekar and similar folks (and I convinced myself a long time ago that basically everything that these folks think about quantum gravity is fundamentally misguided – largely because they take some properties of classical GR much more seriously than they should) which is why it shouldn't have been surprising that my excitement about very similar things done by someone more famous could be limited, too.

Clearly, I am not the only one who sees the similarity.

In the new paper (and older papers mentioned in it), Herdegen also proposes some ideas that do not overlap with the content of Strominger's paper – for example a proposal that some non-local degrees of freedom should be adopted for the null infinity. In some sense, I feel that these proposals are ironically "more stringy" than the "non-emergent spacetime" picture by Andy and collaborators.

No comments:

Post a Comment