SciPost Submission Page
Cosmic Censorship of TransPlanckian Field Ranges in Gravitational Collapse
by Himanshu Chaudhary, Chethan Krishnan
This is not the current version.
Submission summary
As Contributors:  Chethan Krishnan 
Arxiv Link:  https://arxiv.org/abs/2003.05488v2 (pdf) 
Date submitted:  20200718 02:00 
Submitted by:  Krishnan, Chethan 
Submitted to:  SciPost Physics 
Academic field:  Physics 
Specialties: 

Approach:  Theoretical 
Abstract
A classical solution where the (scalar) field value moves by an ${\cal O}(1)$ range in Planck units is believed to signal the breakdown of Effective Field Theory (EFT). One heuristic argument for this is that such a field will have enough energy to be inside its own Schwarzschild radius, and will result in collapse. In this paper, we consider an inverse problem: what kind of field ranges arise during the gravitational collapse of a classical field? Despite the fact that collapse has been studied for almost a hundred years, most of the discussion is phrased in terms of fluid stress tensors, and not fields. An exception is the scalar collapse made famous by Choptuik. We reconsider Choptuiklike systems, but with the emphasis now on the evolution of the scalar. We give strong evidence that generic spherically symmetric collapse of a massless scalar field leads to superPlanckian field movement. But we also note that in every such supercritical collapse scenario, the large field range is hidden behind an apparent horizon. We also discuss how the familiar perfect fluid models for collapse like OppenheimerSnyder and Vaidya should be viewed in light of our results.
Current status:
Author comments upon resubmission
A key point is that at late times, there are two solutions to the apparent horizon equation and only the smaller one is inside the region of interest. Since the time of submission of the original version, we have noticed that in EVERY single case that is not too close to criticality, a bigger (and therefore correct) solution forms just outside, to cover up the large field range! This is a very strong suggestion that there is a cosmic censorship like mechanism in effect during gravitational collapse, that censors large field ranges.
While this result is quite interesting, it is a sharp deviation from our original punchline. While our original numerical results for the evolution are still correct, the interpretation changes completely.
We believe this also addresses the referee's query in the second paragraph  note that it is not very easy to construct Penrose diagrams of numerical gravity solutions, so we don't have a direct answer to his/her question. But the statements we have made above about the apparent horizon, together with the fact that the apparent horizons are expected to be inside global event horizons lead to a new observation  transPlanckian field ranges during collapse are always censored by a horizon.
So the main results of the paper are now  (a) gravitational collapse of fields GENERICALLY lead to superplanckian field ranges, and (b) these field ranges are are always hidden behind apparent horizons.
The tension with the firewall is also gone.
List of changes
We have changed the title, a crucial sentence in the abstract, and removed/changed some paragraphs in the intro and conclusion to reflect our new perspective. We have also added a small section on the determination of apparent horizon, since it has turned out to be a more subtle and important issue than we expected.
Submission & Refereeing History
You are currently on this page
Reports on this Submission
Anonymous Report 2 on 202089 Contributed Report
 Cite as: Anonymous, Report on arXiv:2003.05488v2, delivered 20200809, doi: 10.21468/SciPost.Report.1905
Strengths
This is a very clear paper on an interesting and timely topic. It provides convincing numerical evidence that scalar field collapse probes superPlanckian field values, but that these values are hidden behind a horizon. It is wellwritten, informative and a pleasure to read.
Weaknesses
There are a few minor issues.
1. The Swampland Distance Conjecture refers to the distance in scalar field space, not just the field value (which is not reparametrization invariant). It would be useful to emphasize that for the theory studied in this paper, the scalar field value is the same as the distance in field space, so the discussion really is phrased in terms of a physical, invariant quantity. For instance, the first bulleted concern in the Discussions asks about coordinate choices, but this concern applies to coordinates on field space as well as on spacetime.
2. The discussion of the relationship to other work on superPlanckian spatial variations of fields could be improved. The authors mention a "Heuristic Black Hole Argument" and cite a 2008 paper of Nicolis, but they also say this is "usually attributed to T. Banks." I don't think that this attribution is quite correct. Banks discusses this at the beginning of section 3 of arXiv:1910.12817, where he says that "it was pointed out" to him at a Rutgers group meeting; unfortunately, he doesn't say who pointed it out. If there is something to cite before the paper of Nicolis, perhaps it is the related paper by Banks, hepth/0011255, but it does not contain quite the same claims. Another, more tangentially related, early paper is arXiv:0705.2768 by ArkaniHamed, Orgera, and Polchinski, which finds superPlanckian variations in some axionic wormhole solutions. (Although it predates the Nicolis paper, it mentions the basic "Heuristic Black Hole Argument" as if it was a wellknown fact at the time.)
More recently, there have been a few papers that explore how large spatial variations of fields might be censored. One of these is cited as Ref. [8] (by Klaewer and Palti), but in the context of a list of references to the Swampland Distance Conjecture. I think it should be singled out as being closer to the topic of the current paper than the other SDC references. Similarly, arXiv:1701.05572 by Dolan, Draper, Kozaczuk, and Patel studies large spatial variations of axion fields around strings; arXiv:1901.00515 by Draper and Farkas studies large spatial variations of scalar fields in KaluzaKlein bubble geometries; and arXiv:1910.04804 continues this study. The latter reference makes a conjecture about censorship of superPlanckian spatial field excursions, namely that they are bounded by $\log(R \Lambda)$, with $R$ the size of the spatial region over which the field varies and $\Lambda$ a UV cutoff on the theory. (This is consistent with examples of arbitrarily large field variations that are not screened by horizons, but arise only in tiny regions of space.)
I think that the introduction to this paper, and the paragraph in the conclusions that "raises the speculative possibility" that horizons hide large field variations, should refer to this existing body of literature. (In particular, there are known examples where the variations are not hidden behind horizons.)
3. Reference [21] should supply an author and title, not just a web address.
Report
This paper should be published, after minor revisions to address the small issues noted under "Weaknesses." It is a highquality paper that clearly meets the publication criteria of SciPost. It makes an interesting contribution to the field, in an area that many people are actively working on, and it does so in a concise, clear, convincing way.
Requested changes
See under "Weaknesses."
Anonymous Report 1 on 2020726 Invited Report
 Cite as: Anonymous, Report on arXiv:2003.05488v2, delivered 20200726, doi: 10.21468/SciPost.Report.1862
Report
In the new version of the manuscript submitted by the authors, they have argued that the gravitational solutions they found for scalar collapse always have an apparent horizon which hides the large field displacements behind it. This sufficiently addresses my question about the location of the region of large field displacement relative to the horizon of the incipient black hole, and also resolves the purported tension with the validity of EFT at the horizon in the previous version. As I stated in my previous report, the paper is clear, wellwritten, and the results are definitely of interest to the broad high energy theory community. I recommend publication of the paper without further changes.